Drivers

Anti-Immigration Sentiment## **What this driver is** Anti-Immigration Sentiment is a continuing, time-varying factor within rights, social stability, and the information environment. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Anti-Immigration Sentiment; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Anti-Immigration Sentiment is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Protection of Minority Rights: the two may interact or share a proxy, yet they represent different causal questions. Anti-Immigration Sentiment is therefore a persistent analytical node, not a label for every adjacent development. Anti-Immigration Sentiment must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Net migration or erase distinctions from Protection of Minority Rights. The temporal record attached to Anti-Immigration Sentiment uses Net migration (SM.POP.NETM) as a disclosed proxy for 2020-2025. The proxy is an observable lens, not a complete operational definition. The stored scale is inverted because a higher raw observation represents less of Anti-Immigration Sentiment; after inversion, a higher normalized value means a stronger level of the named factor. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Anti-Immigration Sentiment begins with a change in the factor itself and then moves through civic participation, access to trustworthy information, exposure to coercion, social trust, and the ability of groups to organize. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Anti-Immigration Sentiment, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Anti-Immigration Sentiment may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Anti-Immigration Sentiment. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Net migration dataset](https://api.worldbank.org/v2/country/all/indicator/SM.POP.NETM?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2025 Driver series. Across that available record, Anti-Immigration Sentiment finished level with its first normalized observation: 1 in 2020 versus 1 in 2025. The minimum was 1 in 2020, the maximum was 1 in 2020, and the observed sequence contained 0 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. The zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. For Anti-Immigration Sentiment, the scale is inverted because higher raw values indicate less of the named factor. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Anti-Immigration Sentiment, the provider describes the selected series in these terms: Net migration is the net total of migrants during the period, that is, the number of immigrants minus the number of emigrants, including both citizens and noncitizens. The cited producer is World Population Prospects, United Nations (UN), publisher: UN Population Division. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [Universal Declaration of Human Rights](https://www.un.org/en/about-us/universal-declaration-of-human-rights). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Anti-Immigration Sentiment caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Anti-Immigration Sentiment should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [WJP Rule of Law Index](/indicators/justice-wjp-rule-of-law-index-global). For Anti-Immigration Sentiment, expected first-order direction is **lower**. The channel to test is changes in civic freedoms, equal treatment, due process, and effective remedies. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to several years. Confidence: medium. A same-period correlation would not prove this impact. - [Internet penetration](/indicators/tech-internet-penetration-global). For Anti-Immigration Sentiment, expected first-order direction is **lower**. The channel to test is network shutdowns, affordability, platform access, censorship, and willingness to participate online. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to years. Confidence: low to medium and conditional on access restrictions. A same-period correlation would not prove this impact. - [People internally displaced by conflict and violence](/indicators/sec-conflict-violence-internal-displacement). For Anti-Immigration Sentiment, expected first-order direction is **higher**. The channel to test is escalation from exclusion or repression into direct threats, communal violence, and flight. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to years. Confidence: low for information effects alone and higher when coercion escalates. A same-period correlation would not prove this impact. First-order effects for Anti-Immigration Sentiment follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Anti-Immigration Sentiment can reinforce or offset other Drivers in rights, social stability, and the information environment, including Protection of Minority Rights, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Anti-Immigration Sentiment, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2025 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Anti-Immigration Sentiment.Rights, Social Stability, and the Information Environment
Climate-Finance Availability## **What this driver is** Climate-Finance Availability is a continuing, time-varying factor within climate, environment, and disasters. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Climate-Finance Availability; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Climate-Finance Availability is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Climate-Adaptation Readiness: the two may interact or share a proxy, yet they represent different causal questions. Climate-Finance Availability is therefore a persistent analytical node, not a label for every adjacent development. Climate-Finance Availability must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Personal remittances, received (current US$) or erase distinctions from Climate-Adaptation Readiness. The temporal record attached to Climate-Finance Availability uses Personal remittances, received (current US$) (BX.TRF.PWKR.CD.DT) as a disclosed proxy for 2020-2025. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Climate-Finance Availability, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Climate-Finance Availability begins with a change in the factor itself and then moves through physical exposure, ecosystem services, asset losses, adaptation costs, and the resilience of essential infrastructure. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Climate-Finance Availability, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Climate-Finance Availability may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Climate-Finance Availability. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Personal remittances, received (current US$) dataset](https://api.worldbank.org/v2/country/all/indicator/BX.TRF.PWKR.CD.DT?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2025 Driver series. Across that available record, Climate-Finance Availability finished above its first normalized observation: 0.98369327 in 2020 versus 0.99211152 in 2025. The minimum was 0.98369327 in 2020, the maximum was 0.99439858 in 2024, and the observed sequence contained 4 increases and 1 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Climate-Finance Availability, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+1000000000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Climate-Finance Availability, the provider describes the selected series in these terms: Personal remittances comprise personal transfers and compensation of employees. Personal transfers consist of all current transfers in cash or in kind made or received by resident households to or from nonresident households. Personal transfers thus include all current transfers between resident and nonresident individuals. Compensation of employees refers to the income of border, seasonal, and other short-term workers who are employed in an economy where they are not resident and of residents employed by nonresident entities. Data are the sum of two items defined in the sixth edition of the IMF's Balance of Payments Manual: personal transfers and compensation of employees. Data are in current U.S. dollars. The cited producer is IMF balance of payments data, International Monetary Fund (IMF); Staff estimates, World Bank (WB). That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [IPCC Sixth Assessment Report](https://www.ipcc.ch/assessment-report/ar6/). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Climate-Finance Availability caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Climate-Finance Availability should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global CO2 emissions per capita](/indicators/env-co2-emissions-per-capita-global). For Climate-Finance Availability, expected first-order direction is **lower**. The channel to test is changes in fossil-energy use, land cover, carbon sinks, and the emissions intensity of production and recovery. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: years to decades. Confidence: medium for emissions-producing activity and low for any single local intervention. A same-period correlation would not prove this impact. - [Population-weighted PM2.5 exposure](/indicators/env-pm25-air-pollution-exposure-global). For Climate-Finance Availability, expected first-order direction is **lower**. The channel to test is wildfire smoke, combustion, dust, atmospheric chemistry, and pollution-control capacity. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Climate-Finance Availability, expected first-order direction is **lower**. The channel to test is heat stress, injury, infectious disease, food insecurity, unsafe water, displacement, and health-system disruption. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium in directly exposed populations. A same-period correlation would not prove this impact. First-order effects for Climate-Finance Availability follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Climate-Finance Availability can reinforce or offset other Drivers in climate, environment, and disasters, including Climate-Adaptation Readiness, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Climate-Finance Availability, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2025 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Climate-Finance Availability.Climate, Environment, and Disasters
State Fragility in Conflict Zones## **What this driver is** State Fragility in Conflict Zones is a continuing, time-varying factor within international security and war. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to State Fragility in Conflict Zones; a national or sectoral observation must not be silently generalized to the whole world. The boundary of State Fragility in Conflict Zones is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Interstate Armed-Conflict Intensity: the two may interact or share a proxy, yet they represent different causal questions. State Fragility in Conflict Zones is therefore a persistent analytical node, not a label for every adjacent development. State Fragility in Conflict Zones must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Internally displaced people (IDP) by country or territory of asylum / origin or erase distinctions from Interstate Armed-Conflict Intensity. The temporal record attached to State Fragility in Conflict Zones uses Internally displaced people (IDP) by country or territory of asylum / origin (SM.POP.IDPC) as a disclosed proxy for 2020-2025. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of State Fragility in Conflict Zones, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for State Fragility in Conflict Zones begins with a change in the factor itself and then moves through security expectations, displacement, trade continuity, public expenditure, and the physical safety of affected populations. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For State Fragility in Conflict Zones, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in State Fragility in Conflict Zones may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for State Fragility in Conflict Zones. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Internally displaced people (IDP) by country or territory of asylum / origin dataset](https://api.worldbank.org/v2/country/all/indicator/SM.POP.IDPC?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2025 Driver series. Across that available record, State Fragility in Conflict Zones finished above its first normalized observation: 0.95907864 in 2020 versus 0.97953662 in 2025. The minimum was 0.95907864 in 2020, the maximum was 0.98330896 in 2024, and the observed sequence contained 4 increases and 1 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For State Fragility in Conflict Zones, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+100000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For State Fragility in Conflict Zones, the provider describes the selected series in these terms: Internally displaced people (IDPs) are persons or groups of persons who have been forced or obliged to flee or to leave their homes or places of habitual residence, in particular as a result of, or in order to avoid the effects of armed conflict, situations of generalized violence, and violations of human rights, and who have not crossed an internationally recognized State border. The cited producer is Refugee Population Statistics Database, UN High Commissioner for Refugees (UNHCR), uri: https://www.unhcr.org/refugee-statistics/download; Global Internal Displacement Database, Internal Displacement Monitoring Centre (IDMC), uri: https://www.internal-displacement.org/database/displacement-data/. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [United Nations Charter](https://www.un.org/en/about-us/un-charter/full-text). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that State Fragility in Conflict Zones caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which State Fragility in Conflict Zones should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global deaths from organized violence](/indicators/sec-global-armed-conflict-deaths). For State Fragility in Conflict Zones, expected first-order direction is **higher**. The channel to test is changes in the frequency, geographic reach, and lethality of organized violence. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium to high when exposure is direct. A same-period correlation would not prove this impact. - [People internally displaced by conflict and violence](/indicators/sec-conflict-violence-internal-displacement). For State Fragility in Conflict Zones, expected first-order direction is **higher**. The channel to test is threats to life, destruction of housing, and loss of access to livelihoods and services. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to months, with persistence measured in years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For State Fragility in Conflict Zones, expected first-order direction is **higher**. The channel to test is disruption of food, sanitation, vaccination, primary care, and household income. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to years. Confidence: medium and strongly exposure-dependent. A same-period correlation would not prove this impact. First-order effects for State Fragility in Conflict Zones follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** State Fragility in Conflict Zones can reinforce or offset other Drivers in international security and war, including Interstate Armed-Conflict Intensity, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from State Fragility in Conflict Zones, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2025 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of State Fragility in Conflict Zones.International Security and War
Nuclear Deterrence Stability## **What this driver is** Nuclear Deterrence Stability is a continuing, time-varying factor within international security and war. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Nuclear Deterrence Stability; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Nuclear Deterrence Stability is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from NATO Deterrence Capacity: the two may interact or share a proxy, yet they represent different causal questions. Nuclear Deterrence Stability is therefore a persistent analytical node, not a label for every adjacent development. Nuclear Deterrence Stability must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Military expenditure (% of GDP) or erase distinctions from NATO Deterrence Capacity. The temporal record attached to Nuclear Deterrence Stability uses Military expenditure (% of GDP) (MS.MIL.XPND.GD.ZS) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. The stored scale is inverted because a higher raw observation represents less of Nuclear Deterrence Stability; after inversion, a higher normalized value means a stronger level of the named factor. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Nuclear Deterrence Stability begins with a change in the factor itself and then moves through security expectations, displacement, trade continuity, public expenditure, and the physical safety of affected populations. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Nuclear Deterrence Stability, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Nuclear Deterrence Stability may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Nuclear Deterrence Stability. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Military expenditure (% of GDP) dataset](https://api.worldbank.org/v2/country/all/indicator/MS.MIL.XPND.GD.ZS?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Nuclear Deterrence Stability finished below its first normalized observation: 0.97685872 in 2020 versus 0.97526424 in 2024. The minimum was 0.97526424 in 2024, the maximum was 0.97832272 in 2021, and the observed sequence contained 1 increases and 3 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. The zero-to-one conversion is mechanical and preserved in lineage: Published percentage divided by 100 and clamped to [0,1]. For Nuclear Deterrence Stability, the scale is inverted because higher raw values indicate less of the named factor. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Nuclear Deterrence Stability, the provider describes the selected series in these terms: Military expenditure by country as percentage of gross domestic product The cited producer is SIPRI Military Expenditure Database, Stockholm International Peace Research Institute (SIPRI), uri: https://www.sipri.org/databases. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [United Nations Charter](https://www.un.org/en/about-us/un-charter/full-text). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Nuclear Deterrence Stability caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Nuclear Deterrence Stability should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global deaths from organized violence](/indicators/sec-global-armed-conflict-deaths). For Nuclear Deterrence Stability, expected first-order direction is **lower**. The channel to test is changes in the frequency, geographic reach, and lethality of organized violence. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium to high when exposure is direct. A same-period correlation would not prove this impact. - [People internally displaced by conflict and violence](/indicators/sec-conflict-violence-internal-displacement). For Nuclear Deterrence Stability, expected first-order direction is **lower**. The channel to test is threats to life, destruction of housing, and loss of access to livelihoods and services. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to months, with persistence measured in years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Nuclear Deterrence Stability, expected first-order direction is **lower**. The channel to test is disruption of food, sanitation, vaccination, primary care, and household income. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to years. Confidence: medium and strongly exposure-dependent. A same-period correlation would not prove this impact. First-order effects for Nuclear Deterrence Stability follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Nuclear Deterrence Stability can reinforce or offset other Drivers in international security and war, including NATO Deterrence Capacity, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Nuclear Deterrence Stability, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Nuclear Deterrence Stability.International Security and War
Household Purchasing Power## **What this driver is** Household Purchasing Power is a continuing, time-varying factor within macroeconomics and finance. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Household Purchasing Power; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Household Purchasing Power is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Labour-Market Tightness: the two may interact or share a proxy, yet they represent different causal questions. Household Purchasing Power is therefore a persistent analytical node, not a label for every adjacent development. Household Purchasing Power must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Households and NPISHs Final consumption expenditure per capita (constant 2015 US$) or erase distinctions from Labour-Market Tightness. The temporal record attached to Household Purchasing Power uses Households and NPISHs Final consumption expenditure per capita (constant 2015 US$) (NE.CON.PRVT.PC.KD) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Household Purchasing Power, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Household Purchasing Power begins with a change in the factor itself and then moves through real incomes, credit conditions, employment, fiscal capacity, investment decisions, and the distribution of purchasing power. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Household Purchasing Power, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Household Purchasing Power may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Household Purchasing Power. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Households and NPISHs Final consumption expenditure per capita (constant 2015 US$) dataset](https://api.worldbank.org/v2/country/all/indicator/NE.CON.PRVT.PC.KD?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Household Purchasing Power finished above its first normalized observation: 0.9428349 in 2020 versus 0.95814086 in 2024. The minimum was 0.9428349 in 2020, the maximum was 0.95814086 in 2024, and the observed sequence contained 4 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Household Purchasing Power, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Household Purchasing Power, the provider describes the selected series in these terms: This field includes expenditure on goods and services by the Household and NPISH sector for the direct satisfaction of human needs or wants, whether individual or collective. The core indicator has been divided by the general population to achieve a per capita estimate.This indicator is expressed in constant prices, meaning the series has been adjusted to account for price changes over time. The reference year for this adjustment is 2015. This indicator is expressed in United States dollars. The cited producer is Country official statistics, National Statistical Offices (NSOs); National Accounts data files, Central Banks; Staff estimates, World Bank (WB). That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [IMF Data Standards Initiatives](https://www.imf.org/en/About/Factsheets/Sheets/2023/Standards-for-data-dissemination). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Household Purchasing Power caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Household Purchasing Power should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global consumer price inflation](/indicators/econ-global-consumer-price-inflation). For Household Purchasing Power, expected first-order direction is **lower**. The channel to test is changes in demand, financing costs, exchange rates, wages, expectations, and supply costs. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to roughly two years. Confidence: medium, varying with pass-through and policy response. A same-period correlation would not prove this impact. - [Income inequality](/indicators/econ-us-income-inequality-gini). For Household Purchasing Power, expected first-order direction is **lower**. The channel to test is unequal exposure to prices, employment losses, asset returns, taxes, transfers, and credit constraints. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: one to several years. Confidence: low for a universal sign and medium within a defined policy setting. A same-period correlation would not prove this impact. - [Learning poverty rate](/indicators/edu-learning-poverty-rate). For Household Purchasing Power, expected first-order direction is **lower**. The channel to test is household income shocks, nutrition, attendance, public budgets, and the opportunity cost of schooling. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: one school year to a generation. Confidence: low to medium. A same-period correlation would not prove this impact. First-order effects for Household Purchasing Power follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Household Purchasing Power can reinforce or offset other Drivers in macroeconomics and finance, including Labour-Market Tightness, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Household Purchasing Power, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Household Purchasing Power.Macroeconomics and Finance
Combat-Drone Deployment Intensity## **What this driver is** Combat-Drone Deployment Intensity is a continuing, time-varying factor within international security and war. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Combat-Drone Deployment Intensity; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Combat-Drone Deployment Intensity is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Missile-Threat Level: the two may interact or share a proxy, yet they represent different causal questions. Combat-Drone Deployment Intensity is therefore a persistent analytical node, not a label for every adjacent development. Combat-Drone Deployment Intensity must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Military expenditure (current USD) or erase distinctions from Missile-Threat Level. The temporal record attached to Combat-Drone Deployment Intensity uses Military expenditure (current USD) (MS.MIL.XPND.CD) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Combat-Drone Deployment Intensity, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Combat-Drone Deployment Intensity begins with a change in the factor itself and then moves through security expectations, displacement, trade continuity, public expenditure, and the physical safety of affected populations. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Combat-Drone Deployment Intensity, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Combat-Drone Deployment Intensity may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Combat-Drone Deployment Intensity. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Military expenditure (current USD) dataset](https://api.worldbank.org/v2/country/all/indicator/MS.MIL.XPND.CD?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Combat-Drone Deployment Intensity finished above its first normalized observation: 0.9451189 in 2020 versus 0.95565703 in 2024. The minimum was 0.9451189 in 2020, the maximum was 0.95565703 in 2024, and the observed sequence contained 4 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Combat-Drone Deployment Intensity, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000000000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Combat-Drone Deployment Intensity, the provider describes the selected series in these terms: Military spending USD, in current prices (converted at the exchange rate for the given year) The cited producer is SIPRI Military Expenditure Database, Stockholm International Peace Research Institute (SIPRI), uri: https://www.sipri.org/databases. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [United Nations Charter](https://www.un.org/en/about-us/un-charter/full-text). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Combat-Drone Deployment Intensity caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Combat-Drone Deployment Intensity should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global deaths from organized violence](/indicators/sec-global-armed-conflict-deaths). For Combat-Drone Deployment Intensity, expected first-order direction is **lower**. The channel to test is changes in the frequency, geographic reach, and lethality of organized violence. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium to high when exposure is direct. A same-period correlation would not prove this impact. - [People internally displaced by conflict and violence](/indicators/sec-conflict-violence-internal-displacement). For Combat-Drone Deployment Intensity, expected first-order direction is **lower**. The channel to test is threats to life, destruction of housing, and loss of access to livelihoods and services. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to months, with persistence measured in years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Combat-Drone Deployment Intensity, expected first-order direction is **lower**. The channel to test is disruption of food, sanitation, vaccination, primary care, and household income. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to years. Confidence: medium and strongly exposure-dependent. A same-period correlation would not prove this impact. First-order effects for Combat-Drone Deployment Intensity follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Combat-Drone Deployment Intensity can reinforce or offset other Drivers in international security and war, including Missile-Threat Level, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Combat-Drone Deployment Intensity, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Combat-Drone Deployment Intensity.International Security and War
Defence Expenditure## **What this driver is** Defence Expenditure is a continuing, time-varying factor within international security and war. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Defence Expenditure; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Defence Expenditure is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Durability of Allied Military Assistance: the two may interact or share a proxy, yet they represent different causal questions. Defence Expenditure is therefore a persistent analytical node, not a label for every adjacent development. Defence Expenditure must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Military expenditure (current USD) or erase distinctions from Durability of Allied Military Assistance. The temporal record attached to Defence Expenditure uses Military expenditure (current USD) (MS.MIL.XPND.CD) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Defence Expenditure, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Defence Expenditure begins with a change in the factor itself and then moves through security expectations, displacement, trade continuity, public expenditure, and the physical safety of affected populations. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Defence Expenditure, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Defence Expenditure may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Defence Expenditure. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Military expenditure (current USD) dataset](https://api.worldbank.org/v2/country/all/indicator/MS.MIL.XPND.CD?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Defence Expenditure finished above its first normalized observation: 0.9451189 in 2020 versus 0.95565703 in 2024. The minimum was 0.9451189 in 2020, the maximum was 0.95565703 in 2024, and the observed sequence contained 4 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Defence Expenditure, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000000000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Defence Expenditure, the provider describes the selected series in these terms: Military spending USD, in current prices (converted at the exchange rate for the given year) The cited producer is SIPRI Military Expenditure Database, Stockholm International Peace Research Institute (SIPRI), uri: https://www.sipri.org/databases. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [United Nations Charter](https://www.un.org/en/about-us/un-charter/full-text). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Defence Expenditure caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Defence Expenditure should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global deaths from organized violence](/indicators/sec-global-armed-conflict-deaths). For Defence Expenditure, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is changes in the frequency, geographic reach, and lethality of organized violence. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium to high when exposure is direct. A same-period correlation would not prove this impact. - [People internally displaced by conflict and violence](/indicators/sec-conflict-violence-internal-displacement). For Defence Expenditure, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is threats to life, destruction of housing, and loss of access to livelihoods and services. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to months, with persistence measured in years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Defence Expenditure, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is disruption of food, sanitation, vaccination, primary care, and household income. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to years. Confidence: medium and strongly exposure-dependent. A same-period correlation would not prove this impact. First-order effects for Defence Expenditure follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Defence Expenditure can reinforce or offset other Drivers in international security and war, including Durability of Allied Military Assistance, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Defence Expenditure, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Defence Expenditure.International Security and War
Durability of Allied Military Assistance## **What this driver is** Durability of Allied Military Assistance is a continuing, time-varying factor within international security and war. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Durability of Allied Military Assistance; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Durability of Allied Military Assistance is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Defence-Industrial Production Capacity: the two may interact or share a proxy, yet they represent different causal questions. Durability of Allied Military Assistance is therefore a persistent analytical node, not a label for every adjacent development. Durability of Allied Military Assistance must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Military expenditure (current USD) or erase distinctions from Defence-Industrial Production Capacity. The temporal record attached to Durability of Allied Military Assistance uses Military expenditure (current USD) (MS.MIL.XPND.CD) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Durability of Allied Military Assistance, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Durability of Allied Military Assistance begins with a change in the factor itself and then moves through security expectations, displacement, trade continuity, public expenditure, and the physical safety of affected populations. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Durability of Allied Military Assistance, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Durability of Allied Military Assistance may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Durability of Allied Military Assistance. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Military expenditure (current USD) dataset](https://api.worldbank.org/v2/country/all/indicator/MS.MIL.XPND.CD?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Durability of Allied Military Assistance finished above its first normalized observation: 0.9451189 in 2020 versus 0.95565703 in 2024. The minimum was 0.9451189 in 2020, the maximum was 0.95565703 in 2024, and the observed sequence contained 4 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Durability of Allied Military Assistance, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000000000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Durability of Allied Military Assistance, the provider describes the selected series in these terms: Military spending USD, in current prices (converted at the exchange rate for the given year) The cited producer is SIPRI Military Expenditure Database, Stockholm International Peace Research Institute (SIPRI), uri: https://www.sipri.org/databases. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [United Nations Charter](https://www.un.org/en/about-us/un-charter/full-text). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Durability of Allied Military Assistance caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Durability of Allied Military Assistance should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global deaths from organized violence](/indicators/sec-global-armed-conflict-deaths). For Durability of Allied Military Assistance, expected first-order direction is **lower**. The channel to test is changes in the frequency, geographic reach, and lethality of organized violence. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium to high when exposure is direct. A same-period correlation would not prove this impact. - [People internally displaced by conflict and violence](/indicators/sec-conflict-violence-internal-displacement). For Durability of Allied Military Assistance, expected first-order direction is **lower**. The channel to test is threats to life, destruction of housing, and loss of access to livelihoods and services. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to months, with persistence measured in years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Durability of Allied Military Assistance, expected first-order direction is **lower**. The channel to test is disruption of food, sanitation, vaccination, primary care, and household income. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to years. Confidence: medium and strongly exposure-dependent. A same-period correlation would not prove this impact. First-order effects for Durability of Allied Military Assistance follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Durability of Allied Military Assistance can reinforce or offset other Drivers in international security and war, including Defence-Industrial Production Capacity, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Durability of Allied Military Assistance, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Durability of Allied Military Assistance.International Security and War
Missile-Threat Level## **What this driver is** Missile-Threat Level is a continuing, time-varying factor within international security and war. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Missile-Threat Level; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Missile-Threat Level is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Security of Maritime Trade Routes: the two may interact or share a proxy, yet they represent different causal questions. Missile-Threat Level is therefore a persistent analytical node, not a label for every adjacent development. Missile-Threat Level must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Military expenditure (current USD) or erase distinctions from Security of Maritime Trade Routes. The temporal record attached to Missile-Threat Level uses Military expenditure (current USD) (MS.MIL.XPND.CD) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Missile-Threat Level, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Missile-Threat Level begins with a change in the factor itself and then moves through security expectations, displacement, trade continuity, public expenditure, and the physical safety of affected populations. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Missile-Threat Level, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Missile-Threat Level may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Missile-Threat Level. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Military expenditure (current USD) dataset](https://api.worldbank.org/v2/country/all/indicator/MS.MIL.XPND.CD?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Missile-Threat Level finished above its first normalized observation: 0.9451189 in 2020 versus 0.95565703 in 2024. The minimum was 0.9451189 in 2020, the maximum was 0.95565703 in 2024, and the observed sequence contained 4 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Missile-Threat Level, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000000000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Missile-Threat Level, the provider describes the selected series in these terms: Military spending USD, in current prices (converted at the exchange rate for the given year) The cited producer is SIPRI Military Expenditure Database, Stockholm International Peace Research Institute (SIPRI), uri: https://www.sipri.org/databases. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [United Nations Charter](https://www.un.org/en/about-us/un-charter/full-text). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Missile-Threat Level caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Missile-Threat Level should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global deaths from organized violence](/indicators/sec-global-armed-conflict-deaths). For Missile-Threat Level, expected first-order direction is **higher**. The channel to test is changes in the frequency, geographic reach, and lethality of organized violence. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium to high when exposure is direct. A same-period correlation would not prove this impact. - [People internally displaced by conflict and violence](/indicators/sec-conflict-violence-internal-displacement). For Missile-Threat Level, expected first-order direction is **higher**. The channel to test is threats to life, destruction of housing, and loss of access to livelihoods and services. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to months, with persistence measured in years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Missile-Threat Level, expected first-order direction is **higher**. The channel to test is disruption of food, sanitation, vaccination, primary care, and household income. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to years. Confidence: medium and strongly exposure-dependent. A same-period correlation would not prove this impact. First-order effects for Missile-Threat Level follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Missile-Threat Level can reinforce or offset other Drivers in international security and war, including Security of Maritime Trade Routes, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Missile-Threat Level, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Missile-Threat Level.International Security and War
NATO Deterrence Capacity## **What this driver is** NATO Deterrence Capacity is a continuing, time-varying factor within international security and war. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to NATO Deterrence Capacity; a national or sectoral observation must not be silently generalized to the whole world. The boundary of NATO Deterrence Capacity is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Military-Alliance Cohesion: the two may interact or share a proxy, yet they represent different causal questions. NATO Deterrence Capacity is therefore a persistent analytical node, not a label for every adjacent development. NATO Deterrence Capacity must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Military expenditure (current USD) or erase distinctions from Military-Alliance Cohesion. The temporal record attached to NATO Deterrence Capacity uses Military expenditure (current USD) (MS.MIL.XPND.CD) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of NATO Deterrence Capacity, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for NATO Deterrence Capacity begins with a change in the factor itself and then moves through security expectations, displacement, trade continuity, public expenditure, and the physical safety of affected populations. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For NATO Deterrence Capacity, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in NATO Deterrence Capacity may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for NATO Deterrence Capacity. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Military expenditure (current USD) dataset](https://api.worldbank.org/v2/country/all/indicator/MS.MIL.XPND.CD?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, NATO Deterrence Capacity finished above its first normalized observation: 0.9451189 in 2020 versus 0.95565703 in 2024. The minimum was 0.9451189 in 2020, the maximum was 0.95565703 in 2024, and the observed sequence contained 4 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For NATO Deterrence Capacity, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000000000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For NATO Deterrence Capacity, the provider describes the selected series in these terms: Military spending USD, in current prices (converted at the exchange rate for the given year) The cited producer is SIPRI Military Expenditure Database, Stockholm International Peace Research Institute (SIPRI), uri: https://www.sipri.org/databases. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [United Nations Charter](https://www.un.org/en/about-us/un-charter/full-text). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that NATO Deterrence Capacity caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which NATO Deterrence Capacity should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global deaths from organized violence](/indicators/sec-global-armed-conflict-deaths). For NATO Deterrence Capacity, expected first-order direction is **lower**. The channel to test is changes in the frequency, geographic reach, and lethality of organized violence. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium to high when exposure is direct. A same-period correlation would not prove this impact. - [People internally displaced by conflict and violence](/indicators/sec-conflict-violence-internal-displacement). For NATO Deterrence Capacity, expected first-order direction is **lower**. The channel to test is threats to life, destruction of housing, and loss of access to livelihoods and services. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to months, with persistence measured in years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For NATO Deterrence Capacity, expected first-order direction is **lower**. The channel to test is disruption of food, sanitation, vaccination, primary care, and household income. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to years. Confidence: medium and strongly exposure-dependent. A same-period correlation would not prove this impact. First-order effects for NATO Deterrence Capacity follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** NATO Deterrence Capacity can reinforce or offset other Drivers in international security and war, including Military-Alliance Cohesion, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from NATO Deterrence Capacity, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of NATO Deterrence Capacity.International Security and War
Nuclear Escalation Risk## **What this driver is** Nuclear Escalation Risk is a continuing, time-varying factor within international security and war. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Nuclear Escalation Risk; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Nuclear Escalation Risk is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Nuclear Deterrence Stability: the two may interact or share a proxy, yet they represent different causal questions. Nuclear Escalation Risk is therefore a persistent analytical node, not a label for every adjacent development. Nuclear Escalation Risk must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Military expenditure (current USD) or erase distinctions from Nuclear Deterrence Stability. The temporal record attached to Nuclear Escalation Risk uses Military expenditure (current USD) (MS.MIL.XPND.CD) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Nuclear Escalation Risk, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Nuclear Escalation Risk begins with a change in the factor itself and then moves through security expectations, displacement, trade continuity, public expenditure, and the physical safety of affected populations. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Nuclear Escalation Risk, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Nuclear Escalation Risk may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Nuclear Escalation Risk. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Military expenditure (current USD) dataset](https://api.worldbank.org/v2/country/all/indicator/MS.MIL.XPND.CD?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Nuclear Escalation Risk finished above its first normalized observation: 0.9451189 in 2020 versus 0.95565703 in 2024. The minimum was 0.9451189 in 2020, the maximum was 0.95565703 in 2024, and the observed sequence contained 4 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Nuclear Escalation Risk, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000000000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Nuclear Escalation Risk, the provider describes the selected series in these terms: Military spending USD, in current prices (converted at the exchange rate for the given year) The cited producer is SIPRI Military Expenditure Database, Stockholm International Peace Research Institute (SIPRI), uri: https://www.sipri.org/databases. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [United Nations Charter](https://www.un.org/en/about-us/un-charter/full-text). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Nuclear Escalation Risk caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Nuclear Escalation Risk should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global deaths from organized violence](/indicators/sec-global-armed-conflict-deaths). For Nuclear Escalation Risk, expected first-order direction is **higher**. The channel to test is changes in the frequency, geographic reach, and lethality of organized violence. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium to high when exposure is direct. A same-period correlation would not prove this impact. - [People internally displaced by conflict and violence](/indicators/sec-conflict-violence-internal-displacement). For Nuclear Escalation Risk, expected first-order direction is **higher**. The channel to test is threats to life, destruction of housing, and loss of access to livelihoods and services. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to months, with persistence measured in years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Nuclear Escalation Risk, expected first-order direction is **higher**. The channel to test is disruption of food, sanitation, vaccination, primary care, and household income. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to years. Confidence: medium and strongly exposure-dependent. A same-period correlation would not prove this impact. First-order effects for Nuclear Escalation Risk follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Nuclear Escalation Risk can reinforce or offset other Drivers in international security and war, including Nuclear Deterrence Stability, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Nuclear Escalation Risk, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Nuclear Escalation Risk.International Security and War
Labour-Market Tightness## **What this driver is** Labour-Market Tightness is a continuing, time-varying factor within macroeconomics and finance. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Labour-Market Tightness; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Labour-Market Tightness is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Productivity-Growth Potential: the two may interact or share a proxy, yet they represent different causal questions. Labour-Market Tightness is therefore a persistent analytical node, not a label for every adjacent development. Labour-Market Tightness must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Unemployment, total (% of total labor force) (modeled ILO estimate) or erase distinctions from Productivity-Growth Potential. The temporal record attached to Labour-Market Tightness uses Unemployment, total (% of total labor force) (modeled ILO estimate) (SL.UEM.TOTL.ZS) as a disclosed proxy for 2020-2025. The proxy is an observable lens, not a complete operational definition. The stored scale is inverted because a higher raw observation represents less of Labour-Market Tightness; after inversion, a higher normalized value means a stronger level of the named factor. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Labour-Market Tightness begins with a change in the factor itself and then moves through real incomes, credit conditions, employment, fiscal capacity, investment decisions, and the distribution of purchasing power. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Labour-Market Tightness, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Labour-Market Tightness may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Labour-Market Tightness. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Unemployment, total (% of total labor force) (modeled ILO estimate) dataset](https://api.worldbank.org/v2/country/all/indicator/SL.UEM.TOTL.ZS?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2025 Driver series. Across that available record, Labour-Market Tightness finished above its first normalized observation: 0.93412311 in 2020 versus 0.95208667 in 2025. The minimum was 0.93412311 in 2020, the maximum was 0.95208667 in 2025, and the observed sequence contained 5 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. The zero-to-one conversion is mechanical and preserved in lineage: Published percentage divided by 100 and clamped to [0,1]. For Labour-Market Tightness, the scale is inverted because higher raw values indicate less of the named factor. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Labour-Market Tightness, the provider describes the selected series in these terms: Unemployment refers to the share of the labor force that is without work but available for and seeking employment. The cited producer is ILO Modelled Estimates database (ILOEST), International Labour Organization (ILO), uri: https://ilostat.ilo.org/data/bulk/, publisher: ILOSTAT, type: external database, date accessed: January 17, 2026. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [IMF Data Standards Initiatives](https://www.imf.org/en/About/Factsheets/Sheets/2023/Standards-for-data-dissemination). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Labour-Market Tightness caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Labour-Market Tightness should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global consumer price inflation](/indicators/econ-global-consumer-price-inflation). For Labour-Market Tightness, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is changes in demand, financing costs, exchange rates, wages, expectations, and supply costs. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to roughly two years. Confidence: medium, varying with pass-through and policy response. A same-period correlation would not prove this impact. - [Income inequality](/indicators/econ-us-income-inequality-gini). For Labour-Market Tightness, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is unequal exposure to prices, employment losses, asset returns, taxes, transfers, and credit constraints. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: one to several years. Confidence: low for a universal sign and medium within a defined policy setting. A same-period correlation would not prove this impact. - [Learning poverty rate](/indicators/edu-learning-poverty-rate). For Labour-Market Tightness, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is household income shocks, nutrition, attendance, public budgets, and the opportunity cost of schooling. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: one school year to a generation. Confidence: low to medium. A same-period correlation would not prove this impact. First-order effects for Labour-Market Tightness follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Labour-Market Tightness can reinforce or offset other Drivers in macroeconomics and finance, including Productivity-Growth Potential, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Labour-Market Tightness, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2025 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Labour-Market Tightness.Macroeconomics and Finance
Defence-Industrial Production Capacity## **What this driver is** Defence-Industrial Production Capacity is a continuing, time-varying factor within international security and war. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Defence-Industrial Production Capacity; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Defence-Industrial Production Capacity is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Sanctions Pressure: the two may interact or share a proxy, yet they represent different causal questions. Defence-Industrial Production Capacity is therefore a persistent analytical node, not a label for every adjacent development. Defence-Industrial Production Capacity must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Arms exports (SIPRI trend indicator values) or erase distinctions from Sanctions Pressure. The temporal record attached to Defence-Industrial Production Capacity uses Arms exports (SIPRI trend indicator values) (MS.MIL.XPRT.KD) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Defence-Industrial Production Capacity, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Defence-Industrial Production Capacity begins with a change in the factor itself and then moves through security expectations, displacement, trade continuity, public expenditure, and the physical safety of affected populations. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Defence-Industrial Production Capacity, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Defence-Industrial Production Capacity may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Defence-Industrial Production Capacity. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Arms exports (SIPRI trend indicator values) dataset](https://api.worldbank.org/v2/country/all/indicator/MS.MIL.XPRT.KD?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Defence-Industrial Production Capacity finished above its first normalized observation: 0.94645116 in 2020 versus 0.94829848 in 2024. The minimum was 0.94645116 in 2020, the maximum was 0.95220943 in 2021, and the observed sequence contained 1 increases and 3 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Defence-Industrial Production Capacity, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+100000000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Defence-Industrial Production Capacity, the provider describes the selected series in these terms: Arms transfers (exports) cover the volume of transfers of major arms through sales and gifts, and those made through manufacturing licenses. Data cover major conventional weapons such as aircraft, armored vehicles, artillery, radar systems, missiles, and ships. Figures are SIPRI Trend Indicator Values (TIVs). A '0' indicates that the volume of deliveries is between 0 and 0.5 million SIPRI TIV. The cited producer is Arms Transfers Programme, Stockholm International Peace Research Institute (SIPRI), uri: https://armstransfers.sipri.org/ArmsTransfer/. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [United Nations Charter](https://www.un.org/en/about-us/un-charter/full-text). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Defence-Industrial Production Capacity caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Defence-Industrial Production Capacity should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global deaths from organized violence](/indicators/sec-global-armed-conflict-deaths). For Defence-Industrial Production Capacity, expected first-order direction is **lower**. The channel to test is changes in the frequency, geographic reach, and lethality of organized violence. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium to high when exposure is direct. A same-period correlation would not prove this impact. - [People internally displaced by conflict and violence](/indicators/sec-conflict-violence-internal-displacement). For Defence-Industrial Production Capacity, expected first-order direction is **lower**. The channel to test is threats to life, destruction of housing, and loss of access to livelihoods and services. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to months, with persistence measured in years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Defence-Industrial Production Capacity, expected first-order direction is **lower**. The channel to test is disruption of food, sanitation, vaccination, primary care, and household income. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to years. Confidence: medium and strongly exposure-dependent. A same-period correlation would not prove this impact. First-order effects for Defence-Industrial Production Capacity follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Defence-Industrial Production Capacity can reinforce or offset other Drivers in international security and war, including Sanctions Pressure, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Defence-Industrial Production Capacity, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Defence-Industrial Production Capacity.International Security and War
Greenhouse-Gas Emissions Intensity## **What this driver is** Greenhouse-Gas Emissions Intensity is a continuing, time-varying factor within climate, environment, and disasters. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Greenhouse-Gas Emissions Intensity; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Greenhouse-Gas Emissions Intensity is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Pace of Climate Transition: the two may interact or share a proxy, yet they represent different causal questions. Greenhouse-Gas Emissions Intensity is therefore a persistent analytical node, not a label for every adjacent development. Greenhouse-Gas Emissions Intensity must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Total greenhouse gas emissions excluding LULUCF (Mt CO2e) or erase distinctions from Pace of Climate Transition. The temporal record attached to Greenhouse-Gas Emissions Intensity uses Total greenhouse gas emissions excluding LULUCF (Mt CO2e) (EN.GHG.ALL.MT.CE.AR5) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Greenhouse-Gas Emissions Intensity, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Greenhouse-Gas Emissions Intensity begins with a change in the factor itself and then moves through physical exposure, ecosystem services, asset losses, adaptation costs, and the resilience of essential infrastructure. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Greenhouse-Gas Emissions Intensity, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Greenhouse-Gas Emissions Intensity may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Greenhouse-Gas Emissions Intensity. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Total greenhouse gas emissions excluding LULUCF (Mt CO2e) dataset](https://api.worldbank.org/v2/country/all/indicator/EN.GHG.ALL.MT.CE.AR5?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Greenhouse-Gas Emissions Intensity finished above its first normalized observation: 0.93797301 in 2020 versus 0.94519359 in 2024. The minimum was 0.93797301 in 2020, the maximum was 0.94519359 in 2024, and the observed sequence contained 4 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Greenhouse-Gas Emissions Intensity, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+100000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Greenhouse-Gas Emissions Intensity, the provider describes the selected series in these terms: A measure of annual emissions of the six greenhouse gases (GHG) covered by the Kyoto Protocol (carbon dioxide (CO2), methane (CH4), nitrous oxide (N2O), hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), and sulphurhexafluoride (SF6)) from the energy, industry, waste, and agriculture sectors, standardized to carbon dioxide equivalent values. This measure excludes GHG fluxes caused by Land Use Land Use Change and Forestry (LULUCF), as these fluxes have larger uncertainties. The measure is standardized to carbon dioxide equivalent values using the Global Warming Potential (GWP) factors of IPCC's 5th Assessment Report (AR5). The cited producer is EDGAR (Emissions Database for Global Atmospheric Research) Community GHG Database, Joint Research Centre (JRC) - European Commission, uri: https://edgar.jrc.ec.europa.eu/dataset_ghg2024, publisher: JRC European Commission, date published: 2024; International Energy Agency (IEA), uri: https://edgar.jrc.ec.europa.eu/dataset_ghg2024, publisher: JRC European Commission, date published: 2024. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [IPCC Sixth Assessment Report](https://www.ipcc.ch/assessment-report/ar6/). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Greenhouse-Gas Emissions Intensity caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Greenhouse-Gas Emissions Intensity should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global CO2 emissions per capita](/indicators/env-co2-emissions-per-capita-global). For Greenhouse-Gas Emissions Intensity, expected first-order direction is **higher**. The channel to test is changes in fossil-energy use, land cover, carbon sinks, and the emissions intensity of production and recovery. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: years to decades. Confidence: medium for emissions-producing activity and low for any single local intervention. A same-period correlation would not prove this impact. - [Population-weighted PM2.5 exposure](/indicators/env-pm25-air-pollution-exposure-global). For Greenhouse-Gas Emissions Intensity, expected first-order direction is **higher**. The channel to test is wildfire smoke, combustion, dust, atmospheric chemistry, and pollution-control capacity. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Greenhouse-Gas Emissions Intensity, expected first-order direction is **higher**. The channel to test is heat stress, injury, infectious disease, food insecurity, unsafe water, displacement, and health-system disruption. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium in directly exposed populations. A same-period correlation would not prove this impact. First-order effects for Greenhouse-Gas Emissions Intensity follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Greenhouse-Gas Emissions Intensity can reinforce or offset other Drivers in climate, environment, and disasters, including Pace of Climate Transition, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Greenhouse-Gas Emissions Intensity, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Greenhouse-Gas Emissions Intensity.Climate, Environment, and Disasters
Productivity-Growth Potential## **What this driver is** Productivity-Growth Potential is a continuing, time-varying factor within macroeconomics and finance. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Productivity-Growth Potential; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Productivity-Growth Potential is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Income Inequality: the two may interact or share a proxy, yet they represent different causal questions. Productivity-Growth Potential is therefore a persistent analytical node, not a label for every adjacent development. Productivity-Growth Potential must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into GDP per person employed (constant 2021 PPP $) or erase distinctions from Income Inequality. The temporal record attached to Productivity-Growth Potential uses GDP per person employed (constant 2021 PPP $) (SL.GDP.PCAP.EM.KD) as a disclosed proxy for 2020-2025. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Productivity-Growth Potential, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Productivity-Growth Potential begins with a change in the factor itself and then moves through real incomes, credit conditions, employment, fiscal capacity, investment decisions, and the distribution of purchasing power. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Productivity-Growth Potential, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Productivity-Growth Potential may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Productivity-Growth Potential. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [GDP per person employed (constant 2021 PPP $) dataset](https://api.worldbank.org/v2/country/all/indicator/SL.GDP.PCAP.EM.KD?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2025 Driver series. Across that available record, Productivity-Growth Potential finished above its first normalized observation: 0.93225594 in 2020 versus 0.94095403 in 2025. The minimum was 0.93225594 in 2020, the maximum was 0.94095403 in 2025, and the observed sequence contained 5 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Productivity-Growth Potential, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+100000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Productivity-Growth Potential, the provider describes the selected series in these terms: GDP per person employed is gross domestic product (GDP) divided by total employment in the economy. Purchasing power parity (PPP) GDP is GDP converted to 2021 constant international dollars using PPP rates. An international dollar has the same purchasing power over GDP that a U.S. dollar has in the United States. The cited producer is Staff estimates, World Bank (WB), note: Estimates are based on employment, population, GDP, and PPP data obtained from International Labour Organization, United Nations Population Division, Eurostat, OECD, and World Bank., type: estimates based on external database; International Labour Organization (ILO); United Nations (UN), publisher: UN Population Division; Eurostat (ESTAT); Organisation for Economic Co-operation and Development (OECD); World Development Indicators database, World Bank (WB). That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [IMF Data Standards Initiatives](https://www.imf.org/en/About/Factsheets/Sheets/2023/Standards-for-data-dissemination). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Productivity-Growth Potential caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Productivity-Growth Potential should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global consumer price inflation](/indicators/econ-global-consumer-price-inflation). For Productivity-Growth Potential, expected first-order direction is **lower**. The channel to test is changes in demand, financing costs, exchange rates, wages, expectations, and supply costs. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to roughly two years. Confidence: medium, varying with pass-through and policy response. A same-period correlation would not prove this impact. - [Income inequality](/indicators/econ-us-income-inequality-gini). For Productivity-Growth Potential, expected first-order direction is **lower**. The channel to test is unequal exposure to prices, employment losses, asset returns, taxes, transfers, and credit constraints. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: one to several years. Confidence: low for a universal sign and medium within a defined policy setting. A same-period correlation would not prove this impact. - [Learning poverty rate](/indicators/edu-learning-poverty-rate). For Productivity-Growth Potential, expected first-order direction is **lower**. The channel to test is household income shocks, nutrition, attendance, public budgets, and the opportunity cost of schooling. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: one school year to a generation. Confidence: low to medium. A same-period correlation would not prove this impact. First-order effects for Productivity-Growth Potential follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Productivity-Growth Potential can reinforce or offset other Drivers in macroeconomics and finance, including Income Inequality, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Productivity-Growth Potential, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2025 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Productivity-Growth Potential.Macroeconomics and Finance
Forced-Displacement Scale## **What this driver is** Forced-Displacement Scale is a continuing, time-varying factor within health, demography, and the humanitarian environment. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Forced-Displacement Scale; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Forced-Displacement Scale is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Humanitarian Access: the two may interact or share a proxy, yet they represent different causal questions. Forced-Displacement Scale is therefore a persistent analytical node, not a label for every adjacent development. Forced-Displacement Scale must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Refugees by country of origin or erase distinctions from Humanitarian Access. The temporal record attached to Forced-Displacement Scale uses Refugees by country of origin (OWID:refugee-population-by-country-or-territory-of-origin:Refugees by country of origin) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Forced-Displacement Scale, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Forced-Displacement Scale begins with a change in the factor itself and then moves through mortality and morbidity risks, care capacity, population structure, displacement, and access to essential services. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Forced-Displacement Scale, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Forced-Displacement Scale may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Forced-Displacement Scale. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Refugees by country of origin dataset](https://ourworldindata.org/grapher/refugee-population-by-country-or-territory-of-origin.csv). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Forced-Displacement Scale finished above its first normalized observation: 0.91365455 in 2020 versus 0.93586743 in 2024. The minimum was 0.91365455 in 2020, the maximum was 0.93705318 in 2023, and the observed sequence contained 3 increases and 1 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Forced-Displacement Scale, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+100000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Forced-Displacement Scale, the provider describes the selected series in these terms: Our World in Data republishes the dated series with source-specific metadata and processing notes on the chart data page. The selected column is "Refugees by country of origin". The cited producer is Our World in Data and the named original provider. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [WHO Global Health Observatory](https://www.who.int/data/gho). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Forced-Displacement Scale caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Forced-Displacement Scale should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Forced-Displacement Scale, expected first-order direction is **higher**. The channel to test is prevention, diagnosis, treatment, nutrition, maternal health, sanitation, and continuity of care. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: weeks to several years. Confidence: medium to high for direct health-system and disease channels. A same-period correlation would not prove this impact. - [First-dose measles vaccination coverage](/indicators/ph-measles-vaccine-coverage-global). For Forced-Displacement Scale, expected first-order direction is **lower**. The channel to test is procurement, cold chains, staffing, access, trust, risk communication, and missed-dose recovery. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to several years. Confidence: medium where delivery capacity or acceptance is directly affected. A same-period correlation would not prove this impact. - [Old-age dependency ratio](/indicators/mig-old-age-dependency-ratio-world). For Forced-Displacement Scale, expected first-order direction is **higher**. The channel to test is cohort size, survival, fertility, migration, labour-force participation, and retirement institutions. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: years to decades. Confidence: high for demographic arithmetic and low for short-run policy effects. A same-period correlation would not prove this impact. First-order effects for Forced-Displacement Scale follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Forced-Displacement Scale can reinforce or offset other Drivers in health, demography, and the humanitarian environment, including Humanitarian Access, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Forced-Displacement Scale, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Forced-Displacement Scale.Health, Demography, and the Humanitarian Environment
Genomic Epidemiological Surveillance## **What this driver is** Genomic Epidemiological Surveillance is a continuing, time-varying factor within health, demography, and the humanitarian environment. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Genomic Epidemiological Surveillance; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Genomic Epidemiological Surveillance is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Mental-Disorder Burden: the two may interact or share a proxy, yet they represent different causal questions. Genomic Epidemiological Surveillance is therefore a persistent analytical node, not a label for every adjacent development. Genomic Epidemiological Surveillance must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Scientific and technical journal articles or erase distinctions from Mental-Disorder Burden. The temporal record attached to Genomic Epidemiological Surveillance uses Scientific and technical journal articles (IP.JRN.ARTC.SC) as a disclosed proxy for 2020-2023. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Genomic Epidemiological Surveillance, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Genomic Epidemiological Surveillance begins with a change in the factor itself and then moves through mortality and morbidity risks, care capacity, population structure, displacement, and access to essential services. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Genomic Epidemiological Surveillance, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Genomic Epidemiological Surveillance may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Genomic Epidemiological Surveillance. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Scientific and technical journal articles dataset](https://api.worldbank.org/v2/country/all/indicator/IP.JRN.ARTC.SC?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2023 Driver series. Across that available record, Genomic Epidemiological Surveillance finished above its first normalized observation: 0.9225116 in 2020 versus 0.93067428 in 2023. The minimum was 0.9225116 in 2020, the maximum was 0.93067428 in 2023, and the observed sequence contained 3 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Genomic Epidemiological Surveillance, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Genomic Epidemiological Surveillance, the provider describes the selected series in these terms: Article counts refer to publications from a selection of conference proceedings and peer-reviewed journals from Scopus in science and engineering fields, according to the National Center for Science and Engineering Statistics Taxonomy of Disciplines. The cited producer is Science and Engineering Indicators, National Science Foundation (NSF), uri: https://ncses.nsf.gov/indicators. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [WHO Global Health Observatory](https://www.who.int/data/gho). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Genomic Epidemiological Surveillance caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Genomic Epidemiological Surveillance should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Genomic Epidemiological Surveillance, expected first-order direction is **higher**. The channel to test is prevention, diagnosis, treatment, nutrition, maternal health, sanitation, and continuity of care. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: weeks to several years. Confidence: medium to high for direct health-system and disease channels. A same-period correlation would not prove this impact. - [First-dose measles vaccination coverage](/indicators/ph-measles-vaccine-coverage-global). For Genomic Epidemiological Surveillance, expected first-order direction is **lower**. The channel to test is procurement, cold chains, staffing, access, trust, risk communication, and missed-dose recovery. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to several years. Confidence: medium where delivery capacity or acceptance is directly affected. A same-period correlation would not prove this impact. - [Old-age dependency ratio](/indicators/mig-old-age-dependency-ratio-world). For Genomic Epidemiological Surveillance, expected first-order direction is **higher**. The channel to test is cohort size, survival, fertility, migration, labour-force participation, and retirement institutions. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: years to decades. Confidence: high for demographic arithmetic and low for short-run policy effects. A same-period correlation would not prove this impact. First-order effects for Genomic Epidemiological Surveillance follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Genomic Epidemiological Surveillance can reinforce or offset other Drivers in health, demography, and the humanitarian environment, including Mental-Disorder Burden, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Genomic Epidemiological Surveillance, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2023 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Genomic Epidemiological Surveillance.Health, Demography, and the Humanitarian Environment
Electricity-Grid Resilience## **What this driver is** Electricity-Grid Resilience is a continuing, time-varying factor within energy and natural resources. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Electricity-Grid Resilience; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Electricity-Grid Resilience is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Energy-Storage Capacity: the two may interact or share a proxy, yet they represent different causal questions. Electricity-Grid Resilience is therefore a persistent analytical node, not a label for every adjacent development. Electricity-Grid Resilience must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Access to electricity (% of population) or erase distinctions from Energy-Storage Capacity. The temporal record attached to Electricity-Grid Resilience uses Access to electricity (% of population) (EG.ELC.ACCS.ZS) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Electricity-Grid Resilience, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Electricity-Grid Resilience begins with a change in the factor itself and then moves through energy affordability, supply continuity, infrastructure investment, production costs, and exposure to resource shocks. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Electricity-Grid Resilience, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Electricity-Grid Resilience may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Electricity-Grid Resilience. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Access to electricity (% of population) dataset](https://api.worldbank.org/v2/country/all/indicator/EG.ELC.ACCS.ZS?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Electricity-Grid Resilience finished above its first normalized observation: 0.90399547 in 2020 versus 0.91926484 in 2024. The minimum was 0.90399547 in 2020, the maximum was 0.91926484 in 2024, and the observed sequence contained 3 increases and 1 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Electricity-Grid Resilience, the zero-to-one conversion is mechanical and preserved in lineage: Published percentage divided by 100 and clamped to [0,1]. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Electricity-Grid Resilience, the provider describes the selected series in these terms: Access to electricity is the percentage of population with access to electricity. Electrification data are collected from industry, national surveys and international sources. The cited producer is SDG 7.1.1 Electrification Dataset, World Bank (WB), uri: https://trackingsdg7.esmap.org/downloads, note: Data is downloaded from ESMAP website. Data is released when a new Tracking SDG7 report is released., publisher: World Bank (WB), date accessed: 2024-05-16, date published: 2023. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [International Energy Agency data and statistics](https://www.iea.org/data-and-statistics). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Electricity-Grid Resilience caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Electricity-Grid Resilience should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Access to electricity](/indicators/energy-electricity-access-global). For Electricity-Grid Resilience, expected first-order direction is **higher**. The channel to test is generation adequacy, grid reliability, fuel availability, connection costs, and investment. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to several years. Confidence: medium for direct supply and infrastructure constraints. A same-period correlation would not prove this impact. - [Global CO2 emissions per capita](/indicators/env-co2-emissions-per-capita-global). For Electricity-Grid Resilience, expected first-order direction is **lower**. The channel to test is changes in fossil-energy use, efficiency, electrification, dispatch, and capital-stock turnover. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: one year to several decades. Confidence: medium to high for combustion and generation-mix channels. A same-period correlation would not prove this impact. - [Population-weighted PM2.5 exposure](/indicators/env-pm25-air-pollution-exposure-global). For Electricity-Grid Resilience, expected first-order direction is **lower**. The channel to test is fuel choice, combustion efficiency, power-plant and transport emissions, and pollution controls. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to years. Confidence: medium when the driver changes combustion or local emissions. A same-period correlation would not prove this impact. First-order effects for Electricity-Grid Resilience follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Electricity-Grid Resilience can reinforce or offset other Drivers in energy and natural resources, including Energy-Storage Capacity, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Electricity-Grid Resilience, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Electricity-Grid Resilience.Energy and Natural Resources
Water Scarcity## **What this driver is** Water Scarcity is a continuing, time-varying factor within climate, environment, and disasters. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Water Scarcity; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Water Scarcity is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Biodiversity Loss: the two may interact or share a proxy, yet they represent different causal questions. Water Scarcity is therefore a persistent analytical node, not a label for every adjacent development. Water Scarcity must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Level of water stress: freshwater withdrawal as a proportion of available freshwater resources or erase distinctions from Biodiversity Loss. The temporal record attached to Water Scarcity uses Level of water stress: freshwater withdrawal as a proportion of available freshwater resources (ER.H2O.FWST.ZS) as a disclosed proxy for 2020-2022. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Water Scarcity, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Water Scarcity begins with a change in the factor itself and then moves through physical exposure, ecosystem services, asset losses, adaptation costs, and the resilience of essential infrastructure. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Water Scarcity, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Water Scarcity may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Water Scarcity. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Level of water stress: freshwater withdrawal as a proportion of available freshwater resources dataset](https://api.worldbank.org/v2/country/all/indicator/ER.H2O.FWST.ZS?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2022 Driver series. Across that available record, Water Scarcity finished below its first normalized observation: 0.92123357 in 2020 versus 0.91829703 in 2022. The minimum was 0.91829703 in 2022, the maximum was 0.92123357 in 2020, and the observed sequence contained 0 increases and 2 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Water Scarcity, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+100); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Water Scarcity, the provider describes the selected series in these terms: The level of water stress: freshwater withdrawal as a proportion of available freshwater resources is the ratio between total freshwater withdrawn by all major sectors and total renewable freshwater resources, after taking into account environmental water requirements. Main sectors, as defined by ISIC standards, include agriculture; forestry and fishing; manufacturing; electricity industry; and services. This indicator is also known as water withdrawal intensity. The cited producer is AQUASTAT - FAO's Global Information System on Water and Agriculture, Food and Agriculture Organization of the United Nations (FAO), uri: https://data.apps.fao.org/aquastat/, publisher: Food and Agriculture Organization of the United Nations (FAO), date accessed: 20240529. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [IPCC Sixth Assessment Report](https://www.ipcc.ch/assessment-report/ar6/). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Water Scarcity caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Water Scarcity should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global CO2 emissions per capita](/indicators/env-co2-emissions-per-capita-global). For Water Scarcity, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is changes in fossil-energy use, land cover, carbon sinks, and the emissions intensity of production and recovery. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: years to decades. Confidence: medium for emissions-producing activity and low for any single local intervention. A same-period correlation would not prove this impact. - [Population-weighted PM2.5 exposure](/indicators/env-pm25-air-pollution-exposure-global). For Water Scarcity, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is wildfire smoke, combustion, dust, atmospheric chemistry, and pollution-control capacity. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Water Scarcity, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is heat stress, injury, infectious disease, food insecurity, unsafe water, displacement, and health-system disruption. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium in directly exposed populations. A same-period correlation would not prove this impact. First-order effects for Water Scarcity follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Water Scarcity can reinforce or offset other Drivers in climate, environment, and disasters, including Biodiversity Loss, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Water Scarcity, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2022 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Water Scarcity.Climate, Environment, and Disasters
Commercial Human-Spaceflight Capacity## **What this driver is** Commercial Human-Spaceflight Capacity is a continuing, time-varying factor within science, space, and biotechnology. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Commercial Human-Spaceflight Capacity; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Commercial Human-Spaceflight Capacity is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Rocket Reusability: the two may interact or share a proxy, yet they represent different causal questions. Commercial Human-Spaceflight Capacity is therefore a persistent analytical node, not a label for every adjacent development. Commercial Human-Spaceflight Capacity must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Annual number of objects launched into outer space or erase distinctions from Rocket Reusability. The temporal record attached to Commercial Human-Spaceflight Capacity uses Annual number of objects launched into outer space (OWID:yearly-number-of-objects-launched-into-outer-space:Annual number of objects launched into outer space) as a disclosed proxy for 2020-2025. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Commercial Human-Spaceflight Capacity, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Commercial Human-Spaceflight Capacity begins with a change in the factor itself and then moves through research capacity, deployment cost, infrastructure access, knowledge diffusion, and the governance of emerging capabilities. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Commercial Human-Spaceflight Capacity, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Commercial Human-Spaceflight Capacity may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Commercial Human-Spaceflight Capacity. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Annual number of objects launched into outer space dataset](https://ourworldindata.org/grapher/yearly-number-of-objects-launched-into-outer-space.csv). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2025 Driver series. Across that available record, Commercial Human-Spaceflight Capacity finished above its first normalized observation: 0.77636912 in 2020 versus 0.91355829 in 2025. The minimum was 0.77636912 in 2020, the maximum was 0.91355829 in 2025, and the observed sequence contained 4 increases and 1 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Commercial Human-Spaceflight Capacity, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Commercial Human-Spaceflight Capacity, the provider describes the selected series in these terms: Our World in Data republishes the dated series with source-specific metadata and processing notes on the chart data page. The selected column is "Annual number of objects launched into outer space". The cited producer is Our World in Data and the named original provider. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [UNESCO Institute for Statistics science data](https://uis.unesco.org/en/topic/science-technology-and-innovation). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Commercial Human-Spaceflight Capacity caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Commercial Human-Spaceflight Capacity should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Commercial Human-Spaceflight Capacity, expected first-order direction is **lower**. The channel to test is translation of research into diagnostics, vaccines, therapeutics, delivery systems, and affordable access. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: several years to decades. Confidence: low generally and medium for validated biomedical applications. A same-period correlation would not prove this impact. - [Internet penetration](/indicators/tech-internet-penetration-global). For Commercial Human-Spaceflight Capacity, expected first-order direction is **higher**. The channel to test is coverage, backhaul, launch and equipment costs, reliability, and last-mile access. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: several years. Confidence: low to medium for communications and satellite applications. A same-period correlation would not prove this impact. - [Global CO2 emissions per capita](/indicators/env-co2-emissions-per-capita-global). For Commercial Human-Spaceflight Capacity, expected first-order direction is **lower**. The channel to test is energy demand, process emissions, substitution, learning curves, and the composition of deployed capital. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: years to decades. Confidence: low until a technology reaches material deployment. A same-period correlation would not prove this impact. First-order effects for Commercial Human-Spaceflight Capacity follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Commercial Human-Spaceflight Capacity can reinforce or offset other Drivers in science, space, and biotechnology, including Rocket Reusability, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Commercial Human-Spaceflight Capacity, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2025 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Commercial Human-Spaceflight Capacity.Science, Space, and Biotechnology
Lunar-Exploration Intensity## **What this driver is** Lunar-Exploration Intensity is a continuing, time-varying factor within science, space, and biotechnology. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Lunar-Exploration Intensity; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Lunar-Exploration Intensity is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Mars-Exploration Intensity: the two may interact or share a proxy, yet they represent different causal questions. Lunar-Exploration Intensity is therefore a persistent analytical node, not a label for every adjacent development. Lunar-Exploration Intensity concerns the cadence, scale, and complexity of scientific, commercial, and governmental activity directed at the Moon, including launch, orbital, landing, surface, communications, and sample-return missions. General launch counts are only a remote capacity proxy because most objects do not support lunar activity. The concept differs from Mars-Exploration Intensity in communications delay, mission duration, radiation exposure, logistics, scientific objectives, and the feasibility of repeated near-term operations. The temporal record attached to Lunar-Exploration Intensity uses Annual number of objects launched into outer space (OWID:yearly-number-of-objects-launched-into-outer-space:Annual number of objects launched into outer space) as a disclosed proxy for 2020-2025. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Lunar-Exploration Intensity, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Lunar-Exploration Intensity begins with a change in the factor itself and then moves through research capacity, deployment cost, infrastructure access, knowledge diffusion, and the governance of emerging capabilities. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Lunar-Exploration Intensity, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Lunar-Exploration Intensity may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Lunar-Exploration Intensity. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Annual number of objects launched into outer space dataset](https://ourworldindata.org/grapher/yearly-number-of-objects-launched-into-outer-space.csv). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2025 Driver series. Across that available record, Lunar-Exploration Intensity finished above its first normalized observation: 0.77636912 in 2020 versus 0.91355829 in 2025. The minimum was 0.77636912 in 2020, the maximum was 0.91355829 in 2025, and the observed sequence contained 4 increases and 1 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Lunar-Exploration Intensity, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Lunar-Exploration Intensity, the provider describes the selected series in these terms: Our World in Data republishes the dated series with source-specific metadata and processing notes on the chart data page. The selected column is "Annual number of objects launched into outer space". The cited producer is Our World in Data and the named original provider. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [UNESCO Institute for Statistics science data](https://uis.unesco.org/en/topic/science-technology-and-innovation). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Lunar-Exploration Intensity caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Lunar-Exploration Intensity should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Lunar-Exploration Intensity, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is translation of research into diagnostics, vaccines, therapeutics, delivery systems, and affordable access. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: several years to decades. Confidence: low generally and medium for validated biomedical applications. A same-period correlation would not prove this impact. - [Internet penetration](/indicators/tech-internet-penetration-global). For Lunar-Exploration Intensity, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is coverage, backhaul, launch and equipment costs, reliability, and last-mile access. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: several years. Confidence: low to medium for communications and satellite applications. A same-period correlation would not prove this impact. - [Global CO2 emissions per capita](/indicators/env-co2-emissions-per-capita-global). For Lunar-Exploration Intensity, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is energy demand, process emissions, substitution, learning curves, and the composition of deployed capital. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: years to decades. Confidence: low until a technology reaches material deployment. A same-period correlation would not prove this impact. First-order effects for Lunar-Exploration Intensity follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Lunar-Exploration Intensity can reinforce or offset other Drivers in science, space, and biotechnology, including Mars-Exploration Intensity, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Lunar-Exploration Intensity, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2025 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Lunar-Exploration Intensity.Science, Space, and Biotechnology
Mars-Exploration Intensity## **What this driver is** Mars-Exploration Intensity is a continuing, time-varying factor within science, space, and biotechnology. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Mars-Exploration Intensity; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Mars-Exploration Intensity is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Biotechnology Innovation: the two may interact or share a proxy, yet they represent different causal questions. Mars-Exploration Intensity is therefore a persistent analytical node, not a label for every adjacent development. Mars-Exploration Intensity concerns the cadence, scale, and complexity of robotic or eventual crewed activity directed at Mars and its environment. Launch totals are an upstream proxy rather than a direct count of Mars missions, whose narrow windows and long transit times create a distinctive rhythm. The concept differs from Lunar-Exploration Intensity because distance, communications delay, planetary protection, entry and landing, surface autonomy, return logistics, and multi-year mission horizons change both costs and plausible spillovers. The temporal record attached to Mars-Exploration Intensity uses Annual number of objects launched into outer space (OWID:yearly-number-of-objects-launched-into-outer-space:Annual number of objects launched into outer space) as a disclosed proxy for 2020-2025. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Mars-Exploration Intensity, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Mars-Exploration Intensity begins with a change in the factor itself and then moves through research capacity, deployment cost, infrastructure access, knowledge diffusion, and the governance of emerging capabilities. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Mars-Exploration Intensity, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Mars-Exploration Intensity may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Mars-Exploration Intensity. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Annual number of objects launched into outer space dataset](https://ourworldindata.org/grapher/yearly-number-of-objects-launched-into-outer-space.csv). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2025 Driver series. Across that available record, Mars-Exploration Intensity finished above its first normalized observation: 0.77636912 in 2020 versus 0.91355829 in 2025. The minimum was 0.77636912 in 2020, the maximum was 0.91355829 in 2025, and the observed sequence contained 4 increases and 1 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Mars-Exploration Intensity, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Mars-Exploration Intensity, the provider describes the selected series in these terms: Our World in Data republishes the dated series with source-specific metadata and processing notes on the chart data page. The selected column is "Annual number of objects launched into outer space". The cited producer is Our World in Data and the named original provider. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [UNESCO Institute for Statistics science data](https://uis.unesco.org/en/topic/science-technology-and-innovation). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Mars-Exploration Intensity caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Mars-Exploration Intensity should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Mars-Exploration Intensity, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is translation of research into diagnostics, vaccines, therapeutics, delivery systems, and affordable access. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: several years to decades. Confidence: low generally and medium for validated biomedical applications. A same-period correlation would not prove this impact. - [Internet penetration](/indicators/tech-internet-penetration-global). For Mars-Exploration Intensity, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is coverage, backhaul, launch and equipment costs, reliability, and last-mile access. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: several years. Confidence: low to medium for communications and satellite applications. A same-period correlation would not prove this impact. - [Global CO2 emissions per capita](/indicators/env-co2-emissions-per-capita-global). For Mars-Exploration Intensity, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is energy demand, process emissions, substitution, learning curves, and the composition of deployed capital. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: years to decades. Confidence: low until a technology reaches material deployment. A same-period correlation would not prove this impact. First-order effects for Mars-Exploration Intensity follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Mars-Exploration Intensity can reinforce or offset other Drivers in science, space, and biotechnology, including Biotechnology Innovation, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Mars-Exploration Intensity, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2025 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Mars-Exploration Intensity.Science, Space, and Biotechnology
Rocket Reusability## **What this driver is** Rocket Reusability is a continuing, time-varying factor within science, space, and biotechnology. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Rocket Reusability; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Rocket Reusability is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Satellite-Internet Coverage: the two may interact or share a proxy, yet they represent different causal questions. Rocket Reusability is therefore a persistent analytical node, not a label for every adjacent development. Rocket Reusability must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Annual number of objects launched into outer space or erase distinctions from Satellite-Internet Coverage. The temporal record attached to Rocket Reusability uses Annual number of objects launched into outer space (OWID:yearly-number-of-objects-launched-into-outer-space:Annual number of objects launched into outer space) as a disclosed proxy for 2020-2025. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Rocket Reusability, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Rocket Reusability begins with a change in the factor itself and then moves through research capacity, deployment cost, infrastructure access, knowledge diffusion, and the governance of emerging capabilities. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Rocket Reusability, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Rocket Reusability may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Rocket Reusability. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Annual number of objects launched into outer space dataset](https://ourworldindata.org/grapher/yearly-number-of-objects-launched-into-outer-space.csv). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2025 Driver series. Across that available record, Rocket Reusability finished above its first normalized observation: 0.77636912 in 2020 versus 0.91355829 in 2025. The minimum was 0.77636912 in 2020, the maximum was 0.91355829 in 2025, and the observed sequence contained 4 increases and 1 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Rocket Reusability, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Rocket Reusability, the provider describes the selected series in these terms: Our World in Data republishes the dated series with source-specific metadata and processing notes on the chart data page. The selected column is "Annual number of objects launched into outer space". The cited producer is Our World in Data and the named original provider. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [UNESCO Institute for Statistics science data](https://uis.unesco.org/en/topic/science-technology-and-innovation). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Rocket Reusability caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Rocket Reusability should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Rocket Reusability, expected first-order direction is **lower**. The channel to test is translation of research into diagnostics, vaccines, therapeutics, delivery systems, and affordable access. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: several years to decades. Confidence: low generally and medium for validated biomedical applications. A same-period correlation would not prove this impact. - [Internet penetration](/indicators/tech-internet-penetration-global). For Rocket Reusability, expected first-order direction is **higher**. The channel to test is coverage, backhaul, launch and equipment costs, reliability, and last-mile access. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: several years. Confidence: low to medium for communications and satellite applications. A same-period correlation would not prove this impact. - [Global CO2 emissions per capita](/indicators/env-co2-emissions-per-capita-global). For Rocket Reusability, expected first-order direction is **lower**. The channel to test is energy demand, process emissions, substitution, learning curves, and the composition of deployed capital. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: years to decades. Confidence: low until a technology reaches material deployment. A same-period correlation would not prove this impact. First-order effects for Rocket Reusability follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Rocket Reusability can reinforce or offset other Drivers in science, space, and biotechnology, including Satellite-Internet Coverage, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Rocket Reusability, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2025 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Rocket Reusability.Science, Space, and Biotechnology
Satellite-Internet Coverage## **What this driver is** Satellite-Internet Coverage is a continuing, time-varying factor within science, space, and biotechnology. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Satellite-Internet Coverage; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Satellite-Internet Coverage is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Lunar-Exploration Intensity: the two may interact or share a proxy, yet they represent different causal questions. Satellite-Internet Coverage is therefore a persistent analytical node, not a label for every adjacent development. Satellite-Internet Coverage must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Annual number of objects launched into outer space or erase distinctions from Lunar-Exploration Intensity. The temporal record attached to Satellite-Internet Coverage uses Annual number of objects launched into outer space (OWID:yearly-number-of-objects-launched-into-outer-space:Annual number of objects launched into outer space) as a disclosed proxy for 2020-2025. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Satellite-Internet Coverage, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Satellite-Internet Coverage begins with a change in the factor itself and then moves through research capacity, deployment cost, infrastructure access, knowledge diffusion, and the governance of emerging capabilities. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Satellite-Internet Coverage, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Satellite-Internet Coverage may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Satellite-Internet Coverage. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Annual number of objects launched into outer space dataset](https://ourworldindata.org/grapher/yearly-number-of-objects-launched-into-outer-space.csv). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2025 Driver series. Across that available record, Satellite-Internet Coverage finished above its first normalized observation: 0.77636912 in 2020 versus 0.91355829 in 2025. The minimum was 0.77636912 in 2020, the maximum was 0.91355829 in 2025, and the observed sequence contained 4 increases and 1 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Satellite-Internet Coverage, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Satellite-Internet Coverage, the provider describes the selected series in these terms: Our World in Data republishes the dated series with source-specific metadata and processing notes on the chart data page. The selected column is "Annual number of objects launched into outer space". The cited producer is Our World in Data and the named original provider. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [UNESCO Institute for Statistics science data](https://uis.unesco.org/en/topic/science-technology-and-innovation). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Satellite-Internet Coverage caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Satellite-Internet Coverage should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Satellite-Internet Coverage, expected first-order direction is **lower**. The channel to test is translation of research into diagnostics, vaccines, therapeutics, delivery systems, and affordable access. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: several years to decades. Confidence: low generally and medium for validated biomedical applications. A same-period correlation would not prove this impact. - [Internet penetration](/indicators/tech-internet-penetration-global). For Satellite-Internet Coverage, expected first-order direction is **higher**. The channel to test is coverage, backhaul, launch and equipment costs, reliability, and last-mile access. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: several years. Confidence: low to medium for communications and satellite applications. A same-period correlation would not prove this impact. - [Global CO2 emissions per capita](/indicators/env-co2-emissions-per-capita-global). For Satellite-Internet Coverage, expected first-order direction is **lower**. The channel to test is energy demand, process emissions, substitution, learning curves, and the composition of deployed capital. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: years to decades. Confidence: low until a technology reaches material deployment. A same-period correlation would not prove this impact. First-order effects for Satellite-Internet Coverage follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Satellite-Internet Coverage can reinforce or offset other Drivers in science, space, and biotechnology, including Lunar-Exploration Intensity, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Satellite-Internet Coverage, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2025 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Satellite-Internet Coverage.Science, Space, and Biotechnology

Loading drivers…

Methodology
About
Pricing
Contact
Terms
Privacy
Refunds

Preview tools

Preview features

Off — 11 preview options available.

Drivers

Anti-Immigration Sentiment## **What this driver is** Anti-Immigration Sentiment is a continuing, time-varying factor within rights, social stability, and the information environment. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Anti-Immigration Sentiment; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Anti-Immigration Sentiment is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Protection of Minority Rights: the two may interact or share a proxy, yet they represent different causal questions. Anti-Immigration Sentiment is therefore a persistent analytical node, not a label for every adjacent development. Anti-Immigration Sentiment must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Net migration or erase distinctions from Protection of Minority Rights. The temporal record attached to Anti-Immigration Sentiment uses Net migration (SM.POP.NETM) as a disclosed proxy for 2020-2025. The proxy is an observable lens, not a complete operational definition. The stored scale is inverted because a higher raw observation represents less of Anti-Immigration Sentiment; after inversion, a higher normalized value means a stronger level of the named factor. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Anti-Immigration Sentiment begins with a change in the factor itself and then moves through civic participation, access to trustworthy information, exposure to coercion, social trust, and the ability of groups to organize. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Anti-Immigration Sentiment, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Anti-Immigration Sentiment may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Anti-Immigration Sentiment. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Net migration dataset](https://api.worldbank.org/v2/country/all/indicator/SM.POP.NETM?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2025 Driver series. Across that available record, Anti-Immigration Sentiment finished level with its first normalized observation: 1 in 2020 versus 1 in 2025. The minimum was 1 in 2020, the maximum was 1 in 2020, and the observed sequence contained 0 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. The zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. For Anti-Immigration Sentiment, the scale is inverted because higher raw values indicate less of the named factor. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Anti-Immigration Sentiment, the provider describes the selected series in these terms: Net migration is the net total of migrants during the period, that is, the number of immigrants minus the number of emigrants, including both citizens and noncitizens. The cited producer is World Population Prospects, United Nations (UN), publisher: UN Population Division. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [Universal Declaration of Human Rights](https://www.un.org/en/about-us/universal-declaration-of-human-rights). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Anti-Immigration Sentiment caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Anti-Immigration Sentiment should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [WJP Rule of Law Index](/indicators/justice-wjp-rule-of-law-index-global). For Anti-Immigration Sentiment, expected first-order direction is **lower**. The channel to test is changes in civic freedoms, equal treatment, due process, and effective remedies. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to several years. Confidence: medium. A same-period correlation would not prove this impact. - [Internet penetration](/indicators/tech-internet-penetration-global). For Anti-Immigration Sentiment, expected first-order direction is **lower**. The channel to test is network shutdowns, affordability, platform access, censorship, and willingness to participate online. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to years. Confidence: low to medium and conditional on access restrictions. A same-period correlation would not prove this impact. - [People internally displaced by conflict and violence](/indicators/sec-conflict-violence-internal-displacement). For Anti-Immigration Sentiment, expected first-order direction is **higher**. The channel to test is escalation from exclusion or repression into direct threats, communal violence, and flight. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to years. Confidence: low for information effects alone and higher when coercion escalates. A same-period correlation would not prove this impact. First-order effects for Anti-Immigration Sentiment follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Anti-Immigration Sentiment can reinforce or offset other Drivers in rights, social stability, and the information environment, including Protection of Minority Rights, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Anti-Immigration Sentiment, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2025 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Anti-Immigration Sentiment.Rights, Social Stability, and the Information Environment
Climate-Finance Availability## **What this driver is** Climate-Finance Availability is a continuing, time-varying factor within climate, environment, and disasters. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Climate-Finance Availability; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Climate-Finance Availability is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Climate-Adaptation Readiness: the two may interact or share a proxy, yet they represent different causal questions. Climate-Finance Availability is therefore a persistent analytical node, not a label for every adjacent development. Climate-Finance Availability must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Personal remittances, received (current US$) or erase distinctions from Climate-Adaptation Readiness. The temporal record attached to Climate-Finance Availability uses Personal remittances, received (current US$) (BX.TRF.PWKR.CD.DT) as a disclosed proxy for 2020-2025. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Climate-Finance Availability, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Climate-Finance Availability begins with a change in the factor itself and then moves through physical exposure, ecosystem services, asset losses, adaptation costs, and the resilience of essential infrastructure. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Climate-Finance Availability, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Climate-Finance Availability may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Climate-Finance Availability. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Personal remittances, received (current US$) dataset](https://api.worldbank.org/v2/country/all/indicator/BX.TRF.PWKR.CD.DT?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2025 Driver series. Across that available record, Climate-Finance Availability finished above its first normalized observation: 0.98369327 in 2020 versus 0.99211152 in 2025. The minimum was 0.98369327 in 2020, the maximum was 0.99439858 in 2024, and the observed sequence contained 4 increases and 1 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Climate-Finance Availability, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+1000000000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Climate-Finance Availability, the provider describes the selected series in these terms: Personal remittances comprise personal transfers and compensation of employees. Personal transfers consist of all current transfers in cash or in kind made or received by resident households to or from nonresident households. Personal transfers thus include all current transfers between resident and nonresident individuals. Compensation of employees refers to the income of border, seasonal, and other short-term workers who are employed in an economy where they are not resident and of residents employed by nonresident entities. Data are the sum of two items defined in the sixth edition of the IMF's Balance of Payments Manual: personal transfers and compensation of employees. Data are in current U.S. dollars. The cited producer is IMF balance of payments data, International Monetary Fund (IMF); Staff estimates, World Bank (WB). That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [IPCC Sixth Assessment Report](https://www.ipcc.ch/assessment-report/ar6/). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Climate-Finance Availability caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Climate-Finance Availability should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global CO2 emissions per capita](/indicators/env-co2-emissions-per-capita-global). For Climate-Finance Availability, expected first-order direction is **lower**. The channel to test is changes in fossil-energy use, land cover, carbon sinks, and the emissions intensity of production and recovery. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: years to decades. Confidence: medium for emissions-producing activity and low for any single local intervention. A same-period correlation would not prove this impact. - [Population-weighted PM2.5 exposure](/indicators/env-pm25-air-pollution-exposure-global). For Climate-Finance Availability, expected first-order direction is **lower**. The channel to test is wildfire smoke, combustion, dust, atmospheric chemistry, and pollution-control capacity. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Climate-Finance Availability, expected first-order direction is **lower**. The channel to test is heat stress, injury, infectious disease, food insecurity, unsafe water, displacement, and health-system disruption. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium in directly exposed populations. A same-period correlation would not prove this impact. First-order effects for Climate-Finance Availability follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Climate-Finance Availability can reinforce or offset other Drivers in climate, environment, and disasters, including Climate-Adaptation Readiness, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Climate-Finance Availability, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2025 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Climate-Finance Availability.Climate, Environment, and Disasters
State Fragility in Conflict Zones## **What this driver is** State Fragility in Conflict Zones is a continuing, time-varying factor within international security and war. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to State Fragility in Conflict Zones; a national or sectoral observation must not be silently generalized to the whole world. The boundary of State Fragility in Conflict Zones is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Interstate Armed-Conflict Intensity: the two may interact or share a proxy, yet they represent different causal questions. State Fragility in Conflict Zones is therefore a persistent analytical node, not a label for every adjacent development. State Fragility in Conflict Zones must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Internally displaced people (IDP) by country or territory of asylum / origin or erase distinctions from Interstate Armed-Conflict Intensity. The temporal record attached to State Fragility in Conflict Zones uses Internally displaced people (IDP) by country or territory of asylum / origin (SM.POP.IDPC) as a disclosed proxy for 2020-2025. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of State Fragility in Conflict Zones, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for State Fragility in Conflict Zones begins with a change in the factor itself and then moves through security expectations, displacement, trade continuity, public expenditure, and the physical safety of affected populations. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For State Fragility in Conflict Zones, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in State Fragility in Conflict Zones may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for State Fragility in Conflict Zones. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Internally displaced people (IDP) by country or territory of asylum / origin dataset](https://api.worldbank.org/v2/country/all/indicator/SM.POP.IDPC?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2025 Driver series. Across that available record, State Fragility in Conflict Zones finished above its first normalized observation: 0.95907864 in 2020 versus 0.97953662 in 2025. The minimum was 0.95907864 in 2020, the maximum was 0.98330896 in 2024, and the observed sequence contained 4 increases and 1 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For State Fragility in Conflict Zones, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+100000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For State Fragility in Conflict Zones, the provider describes the selected series in these terms: Internally displaced people (IDPs) are persons or groups of persons who have been forced or obliged to flee or to leave their homes or places of habitual residence, in particular as a result of, or in order to avoid the effects of armed conflict, situations of generalized violence, and violations of human rights, and who have not crossed an internationally recognized State border. The cited producer is Refugee Population Statistics Database, UN High Commissioner for Refugees (UNHCR), uri: https://www.unhcr.org/refugee-statistics/download; Global Internal Displacement Database, Internal Displacement Monitoring Centre (IDMC), uri: https://www.internal-displacement.org/database/displacement-data/. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [United Nations Charter](https://www.un.org/en/about-us/un-charter/full-text). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that State Fragility in Conflict Zones caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which State Fragility in Conflict Zones should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global deaths from organized violence](/indicators/sec-global-armed-conflict-deaths). For State Fragility in Conflict Zones, expected first-order direction is **higher**. The channel to test is changes in the frequency, geographic reach, and lethality of organized violence. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium to high when exposure is direct. A same-period correlation would not prove this impact. - [People internally displaced by conflict and violence](/indicators/sec-conflict-violence-internal-displacement). For State Fragility in Conflict Zones, expected first-order direction is **higher**. The channel to test is threats to life, destruction of housing, and loss of access to livelihoods and services. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to months, with persistence measured in years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For State Fragility in Conflict Zones, expected first-order direction is **higher**. The channel to test is disruption of food, sanitation, vaccination, primary care, and household income. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to years. Confidence: medium and strongly exposure-dependent. A same-period correlation would not prove this impact. First-order effects for State Fragility in Conflict Zones follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** State Fragility in Conflict Zones can reinforce or offset other Drivers in international security and war, including Interstate Armed-Conflict Intensity, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from State Fragility in Conflict Zones, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2025 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of State Fragility in Conflict Zones.International Security and War
Nuclear Deterrence Stability## **What this driver is** Nuclear Deterrence Stability is a continuing, time-varying factor within international security and war. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Nuclear Deterrence Stability; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Nuclear Deterrence Stability is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from NATO Deterrence Capacity: the two may interact or share a proxy, yet they represent different causal questions. Nuclear Deterrence Stability is therefore a persistent analytical node, not a label for every adjacent development. Nuclear Deterrence Stability must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Military expenditure (% of GDP) or erase distinctions from NATO Deterrence Capacity. The temporal record attached to Nuclear Deterrence Stability uses Military expenditure (% of GDP) (MS.MIL.XPND.GD.ZS) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. The stored scale is inverted because a higher raw observation represents less of Nuclear Deterrence Stability; after inversion, a higher normalized value means a stronger level of the named factor. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Nuclear Deterrence Stability begins with a change in the factor itself and then moves through security expectations, displacement, trade continuity, public expenditure, and the physical safety of affected populations. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Nuclear Deterrence Stability, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Nuclear Deterrence Stability may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Nuclear Deterrence Stability. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Military expenditure (% of GDP) dataset](https://api.worldbank.org/v2/country/all/indicator/MS.MIL.XPND.GD.ZS?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Nuclear Deterrence Stability finished below its first normalized observation: 0.97685872 in 2020 versus 0.97526424 in 2024. The minimum was 0.97526424 in 2024, the maximum was 0.97832272 in 2021, and the observed sequence contained 1 increases and 3 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. The zero-to-one conversion is mechanical and preserved in lineage: Published percentage divided by 100 and clamped to [0,1]. For Nuclear Deterrence Stability, the scale is inverted because higher raw values indicate less of the named factor. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Nuclear Deterrence Stability, the provider describes the selected series in these terms: Military expenditure by country as percentage of gross domestic product The cited producer is SIPRI Military Expenditure Database, Stockholm International Peace Research Institute (SIPRI), uri: https://www.sipri.org/databases. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [United Nations Charter](https://www.un.org/en/about-us/un-charter/full-text). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Nuclear Deterrence Stability caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Nuclear Deterrence Stability should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global deaths from organized violence](/indicators/sec-global-armed-conflict-deaths). For Nuclear Deterrence Stability, expected first-order direction is **lower**. The channel to test is changes in the frequency, geographic reach, and lethality of organized violence. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium to high when exposure is direct. A same-period correlation would not prove this impact. - [People internally displaced by conflict and violence](/indicators/sec-conflict-violence-internal-displacement). For Nuclear Deterrence Stability, expected first-order direction is **lower**. The channel to test is threats to life, destruction of housing, and loss of access to livelihoods and services. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to months, with persistence measured in years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Nuclear Deterrence Stability, expected first-order direction is **lower**. The channel to test is disruption of food, sanitation, vaccination, primary care, and household income. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to years. Confidence: medium and strongly exposure-dependent. A same-period correlation would not prove this impact. First-order effects for Nuclear Deterrence Stability follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Nuclear Deterrence Stability can reinforce or offset other Drivers in international security and war, including NATO Deterrence Capacity, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Nuclear Deterrence Stability, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Nuclear Deterrence Stability.International Security and War
Household Purchasing Power## **What this driver is** Household Purchasing Power is a continuing, time-varying factor within macroeconomics and finance. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Household Purchasing Power; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Household Purchasing Power is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Labour-Market Tightness: the two may interact or share a proxy, yet they represent different causal questions. Household Purchasing Power is therefore a persistent analytical node, not a label for every adjacent development. Household Purchasing Power must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Households and NPISHs Final consumption expenditure per capita (constant 2015 US$) or erase distinctions from Labour-Market Tightness. The temporal record attached to Household Purchasing Power uses Households and NPISHs Final consumption expenditure per capita (constant 2015 US$) (NE.CON.PRVT.PC.KD) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Household Purchasing Power, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Household Purchasing Power begins with a change in the factor itself and then moves through real incomes, credit conditions, employment, fiscal capacity, investment decisions, and the distribution of purchasing power. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Household Purchasing Power, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Household Purchasing Power may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Household Purchasing Power. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Households and NPISHs Final consumption expenditure per capita (constant 2015 US$) dataset](https://api.worldbank.org/v2/country/all/indicator/NE.CON.PRVT.PC.KD?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Household Purchasing Power finished above its first normalized observation: 0.9428349 in 2020 versus 0.95814086 in 2024. The minimum was 0.9428349 in 2020, the maximum was 0.95814086 in 2024, and the observed sequence contained 4 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Household Purchasing Power, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Household Purchasing Power, the provider describes the selected series in these terms: This field includes expenditure on goods and services by the Household and NPISH sector for the direct satisfaction of human needs or wants, whether individual or collective. The core indicator has been divided by the general population to achieve a per capita estimate.This indicator is expressed in constant prices, meaning the series has been adjusted to account for price changes over time. The reference year for this adjustment is 2015. This indicator is expressed in United States dollars. The cited producer is Country official statistics, National Statistical Offices (NSOs); National Accounts data files, Central Banks; Staff estimates, World Bank (WB). That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [IMF Data Standards Initiatives](https://www.imf.org/en/About/Factsheets/Sheets/2023/Standards-for-data-dissemination). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Household Purchasing Power caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Household Purchasing Power should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global consumer price inflation](/indicators/econ-global-consumer-price-inflation). For Household Purchasing Power, expected first-order direction is **lower**. The channel to test is changes in demand, financing costs, exchange rates, wages, expectations, and supply costs. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to roughly two years. Confidence: medium, varying with pass-through and policy response. A same-period correlation would not prove this impact. - [Income inequality](/indicators/econ-us-income-inequality-gini). For Household Purchasing Power, expected first-order direction is **lower**. The channel to test is unequal exposure to prices, employment losses, asset returns, taxes, transfers, and credit constraints. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: one to several years. Confidence: low for a universal sign and medium within a defined policy setting. A same-period correlation would not prove this impact. - [Learning poverty rate](/indicators/edu-learning-poverty-rate). For Household Purchasing Power, expected first-order direction is **lower**. The channel to test is household income shocks, nutrition, attendance, public budgets, and the opportunity cost of schooling. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: one school year to a generation. Confidence: low to medium. A same-period correlation would not prove this impact. First-order effects for Household Purchasing Power follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Household Purchasing Power can reinforce or offset other Drivers in macroeconomics and finance, including Labour-Market Tightness, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Household Purchasing Power, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Household Purchasing Power.Macroeconomics and Finance
Combat-Drone Deployment Intensity## **What this driver is** Combat-Drone Deployment Intensity is a continuing, time-varying factor within international security and war. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Combat-Drone Deployment Intensity; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Combat-Drone Deployment Intensity is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Missile-Threat Level: the two may interact or share a proxy, yet they represent different causal questions. Combat-Drone Deployment Intensity is therefore a persistent analytical node, not a label for every adjacent development. Combat-Drone Deployment Intensity must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Military expenditure (current USD) or erase distinctions from Missile-Threat Level. The temporal record attached to Combat-Drone Deployment Intensity uses Military expenditure (current USD) (MS.MIL.XPND.CD) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Combat-Drone Deployment Intensity, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Combat-Drone Deployment Intensity begins with a change in the factor itself and then moves through security expectations, displacement, trade continuity, public expenditure, and the physical safety of affected populations. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Combat-Drone Deployment Intensity, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Combat-Drone Deployment Intensity may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Combat-Drone Deployment Intensity. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Military expenditure (current USD) dataset](https://api.worldbank.org/v2/country/all/indicator/MS.MIL.XPND.CD?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Combat-Drone Deployment Intensity finished above its first normalized observation: 0.9451189 in 2020 versus 0.95565703 in 2024. The minimum was 0.9451189 in 2020, the maximum was 0.95565703 in 2024, and the observed sequence contained 4 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Combat-Drone Deployment Intensity, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000000000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Combat-Drone Deployment Intensity, the provider describes the selected series in these terms: Military spending USD, in current prices (converted at the exchange rate for the given year) The cited producer is SIPRI Military Expenditure Database, Stockholm International Peace Research Institute (SIPRI), uri: https://www.sipri.org/databases. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [United Nations Charter](https://www.un.org/en/about-us/un-charter/full-text). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Combat-Drone Deployment Intensity caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Combat-Drone Deployment Intensity should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global deaths from organized violence](/indicators/sec-global-armed-conflict-deaths). For Combat-Drone Deployment Intensity, expected first-order direction is **lower**. The channel to test is changes in the frequency, geographic reach, and lethality of organized violence. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium to high when exposure is direct. A same-period correlation would not prove this impact. - [People internally displaced by conflict and violence](/indicators/sec-conflict-violence-internal-displacement). For Combat-Drone Deployment Intensity, expected first-order direction is **lower**. The channel to test is threats to life, destruction of housing, and loss of access to livelihoods and services. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to months, with persistence measured in years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Combat-Drone Deployment Intensity, expected first-order direction is **lower**. The channel to test is disruption of food, sanitation, vaccination, primary care, and household income. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to years. Confidence: medium and strongly exposure-dependent. A same-period correlation would not prove this impact. First-order effects for Combat-Drone Deployment Intensity follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Combat-Drone Deployment Intensity can reinforce or offset other Drivers in international security and war, including Missile-Threat Level, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Combat-Drone Deployment Intensity, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Combat-Drone Deployment Intensity.International Security and War
Defence Expenditure## **What this driver is** Defence Expenditure is a continuing, time-varying factor within international security and war. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Defence Expenditure; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Defence Expenditure is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Durability of Allied Military Assistance: the two may interact or share a proxy, yet they represent different causal questions. Defence Expenditure is therefore a persistent analytical node, not a label for every adjacent development. Defence Expenditure must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Military expenditure (current USD) or erase distinctions from Durability of Allied Military Assistance. The temporal record attached to Defence Expenditure uses Military expenditure (current USD) (MS.MIL.XPND.CD) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Defence Expenditure, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Defence Expenditure begins with a change in the factor itself and then moves through security expectations, displacement, trade continuity, public expenditure, and the physical safety of affected populations. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Defence Expenditure, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Defence Expenditure may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Defence Expenditure. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Military expenditure (current USD) dataset](https://api.worldbank.org/v2/country/all/indicator/MS.MIL.XPND.CD?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Defence Expenditure finished above its first normalized observation: 0.9451189 in 2020 versus 0.95565703 in 2024. The minimum was 0.9451189 in 2020, the maximum was 0.95565703 in 2024, and the observed sequence contained 4 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Defence Expenditure, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000000000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Defence Expenditure, the provider describes the selected series in these terms: Military spending USD, in current prices (converted at the exchange rate for the given year) The cited producer is SIPRI Military Expenditure Database, Stockholm International Peace Research Institute (SIPRI), uri: https://www.sipri.org/databases. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [United Nations Charter](https://www.un.org/en/about-us/un-charter/full-text). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Defence Expenditure caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Defence Expenditure should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global deaths from organized violence](/indicators/sec-global-armed-conflict-deaths). For Defence Expenditure, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is changes in the frequency, geographic reach, and lethality of organized violence. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium to high when exposure is direct. A same-period correlation would not prove this impact. - [People internally displaced by conflict and violence](/indicators/sec-conflict-violence-internal-displacement). For Defence Expenditure, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is threats to life, destruction of housing, and loss of access to livelihoods and services. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to months, with persistence measured in years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Defence Expenditure, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is disruption of food, sanitation, vaccination, primary care, and household income. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to years. Confidence: medium and strongly exposure-dependent. A same-period correlation would not prove this impact. First-order effects for Defence Expenditure follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Defence Expenditure can reinforce or offset other Drivers in international security and war, including Durability of Allied Military Assistance, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Defence Expenditure, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Defence Expenditure.International Security and War
Durability of Allied Military Assistance## **What this driver is** Durability of Allied Military Assistance is a continuing, time-varying factor within international security and war. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Durability of Allied Military Assistance; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Durability of Allied Military Assistance is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Defence-Industrial Production Capacity: the two may interact or share a proxy, yet they represent different causal questions. Durability of Allied Military Assistance is therefore a persistent analytical node, not a label for every adjacent development. Durability of Allied Military Assistance must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Military expenditure (current USD) or erase distinctions from Defence-Industrial Production Capacity. The temporal record attached to Durability of Allied Military Assistance uses Military expenditure (current USD) (MS.MIL.XPND.CD) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Durability of Allied Military Assistance, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Durability of Allied Military Assistance begins with a change in the factor itself and then moves through security expectations, displacement, trade continuity, public expenditure, and the physical safety of affected populations. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Durability of Allied Military Assistance, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Durability of Allied Military Assistance may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Durability of Allied Military Assistance. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Military expenditure (current USD) dataset](https://api.worldbank.org/v2/country/all/indicator/MS.MIL.XPND.CD?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Durability of Allied Military Assistance finished above its first normalized observation: 0.9451189 in 2020 versus 0.95565703 in 2024. The minimum was 0.9451189 in 2020, the maximum was 0.95565703 in 2024, and the observed sequence contained 4 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Durability of Allied Military Assistance, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000000000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Durability of Allied Military Assistance, the provider describes the selected series in these terms: Military spending USD, in current prices (converted at the exchange rate for the given year) The cited producer is SIPRI Military Expenditure Database, Stockholm International Peace Research Institute (SIPRI), uri: https://www.sipri.org/databases. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [United Nations Charter](https://www.un.org/en/about-us/un-charter/full-text). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Durability of Allied Military Assistance caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Durability of Allied Military Assistance should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global deaths from organized violence](/indicators/sec-global-armed-conflict-deaths). For Durability of Allied Military Assistance, expected first-order direction is **lower**. The channel to test is changes in the frequency, geographic reach, and lethality of organized violence. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium to high when exposure is direct. A same-period correlation would not prove this impact. - [People internally displaced by conflict and violence](/indicators/sec-conflict-violence-internal-displacement). For Durability of Allied Military Assistance, expected first-order direction is **lower**. The channel to test is threats to life, destruction of housing, and loss of access to livelihoods and services. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to months, with persistence measured in years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Durability of Allied Military Assistance, expected first-order direction is **lower**. The channel to test is disruption of food, sanitation, vaccination, primary care, and household income. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to years. Confidence: medium and strongly exposure-dependent. A same-period correlation would not prove this impact. First-order effects for Durability of Allied Military Assistance follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Durability of Allied Military Assistance can reinforce or offset other Drivers in international security and war, including Defence-Industrial Production Capacity, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Durability of Allied Military Assistance, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Durability of Allied Military Assistance.International Security and War
Missile-Threat Level## **What this driver is** Missile-Threat Level is a continuing, time-varying factor within international security and war. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Missile-Threat Level; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Missile-Threat Level is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Security of Maritime Trade Routes: the two may interact or share a proxy, yet they represent different causal questions. Missile-Threat Level is therefore a persistent analytical node, not a label for every adjacent development. Missile-Threat Level must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Military expenditure (current USD) or erase distinctions from Security of Maritime Trade Routes. The temporal record attached to Missile-Threat Level uses Military expenditure (current USD) (MS.MIL.XPND.CD) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Missile-Threat Level, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Missile-Threat Level begins with a change in the factor itself and then moves through security expectations, displacement, trade continuity, public expenditure, and the physical safety of affected populations. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Missile-Threat Level, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Missile-Threat Level may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Missile-Threat Level. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Military expenditure (current USD) dataset](https://api.worldbank.org/v2/country/all/indicator/MS.MIL.XPND.CD?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Missile-Threat Level finished above its first normalized observation: 0.9451189 in 2020 versus 0.95565703 in 2024. The minimum was 0.9451189 in 2020, the maximum was 0.95565703 in 2024, and the observed sequence contained 4 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Missile-Threat Level, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000000000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Missile-Threat Level, the provider describes the selected series in these terms: Military spending USD, in current prices (converted at the exchange rate for the given year) The cited producer is SIPRI Military Expenditure Database, Stockholm International Peace Research Institute (SIPRI), uri: https://www.sipri.org/databases. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [United Nations Charter](https://www.un.org/en/about-us/un-charter/full-text). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Missile-Threat Level caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Missile-Threat Level should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global deaths from organized violence](/indicators/sec-global-armed-conflict-deaths). For Missile-Threat Level, expected first-order direction is **higher**. The channel to test is changes in the frequency, geographic reach, and lethality of organized violence. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium to high when exposure is direct. A same-period correlation would not prove this impact. - [People internally displaced by conflict and violence](/indicators/sec-conflict-violence-internal-displacement). For Missile-Threat Level, expected first-order direction is **higher**. The channel to test is threats to life, destruction of housing, and loss of access to livelihoods and services. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to months, with persistence measured in years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Missile-Threat Level, expected first-order direction is **higher**. The channel to test is disruption of food, sanitation, vaccination, primary care, and household income. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to years. Confidence: medium and strongly exposure-dependent. A same-period correlation would not prove this impact. First-order effects for Missile-Threat Level follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Missile-Threat Level can reinforce or offset other Drivers in international security and war, including Security of Maritime Trade Routes, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Missile-Threat Level, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Missile-Threat Level.International Security and War
NATO Deterrence Capacity## **What this driver is** NATO Deterrence Capacity is a continuing, time-varying factor within international security and war. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to NATO Deterrence Capacity; a national or sectoral observation must not be silently generalized to the whole world. The boundary of NATO Deterrence Capacity is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Military-Alliance Cohesion: the two may interact or share a proxy, yet they represent different causal questions. NATO Deterrence Capacity is therefore a persistent analytical node, not a label for every adjacent development. NATO Deterrence Capacity must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Military expenditure (current USD) or erase distinctions from Military-Alliance Cohesion. The temporal record attached to NATO Deterrence Capacity uses Military expenditure (current USD) (MS.MIL.XPND.CD) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of NATO Deterrence Capacity, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for NATO Deterrence Capacity begins with a change in the factor itself and then moves through security expectations, displacement, trade continuity, public expenditure, and the physical safety of affected populations. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For NATO Deterrence Capacity, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in NATO Deterrence Capacity may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for NATO Deterrence Capacity. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Military expenditure (current USD) dataset](https://api.worldbank.org/v2/country/all/indicator/MS.MIL.XPND.CD?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, NATO Deterrence Capacity finished above its first normalized observation: 0.9451189 in 2020 versus 0.95565703 in 2024. The minimum was 0.9451189 in 2020, the maximum was 0.95565703 in 2024, and the observed sequence contained 4 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For NATO Deterrence Capacity, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000000000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For NATO Deterrence Capacity, the provider describes the selected series in these terms: Military spending USD, in current prices (converted at the exchange rate for the given year) The cited producer is SIPRI Military Expenditure Database, Stockholm International Peace Research Institute (SIPRI), uri: https://www.sipri.org/databases. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [United Nations Charter](https://www.un.org/en/about-us/un-charter/full-text). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that NATO Deterrence Capacity caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which NATO Deterrence Capacity should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global deaths from organized violence](/indicators/sec-global-armed-conflict-deaths). For NATO Deterrence Capacity, expected first-order direction is **lower**. The channel to test is changes in the frequency, geographic reach, and lethality of organized violence. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium to high when exposure is direct. A same-period correlation would not prove this impact. - [People internally displaced by conflict and violence](/indicators/sec-conflict-violence-internal-displacement). For NATO Deterrence Capacity, expected first-order direction is **lower**. The channel to test is threats to life, destruction of housing, and loss of access to livelihoods and services. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to months, with persistence measured in years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For NATO Deterrence Capacity, expected first-order direction is **lower**. The channel to test is disruption of food, sanitation, vaccination, primary care, and household income. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to years. Confidence: medium and strongly exposure-dependent. A same-period correlation would not prove this impact. First-order effects for NATO Deterrence Capacity follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** NATO Deterrence Capacity can reinforce or offset other Drivers in international security and war, including Military-Alliance Cohesion, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from NATO Deterrence Capacity, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of NATO Deterrence Capacity.International Security and War
Nuclear Escalation Risk## **What this driver is** Nuclear Escalation Risk is a continuing, time-varying factor within international security and war. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Nuclear Escalation Risk; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Nuclear Escalation Risk is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Nuclear Deterrence Stability: the two may interact or share a proxy, yet they represent different causal questions. Nuclear Escalation Risk is therefore a persistent analytical node, not a label for every adjacent development. Nuclear Escalation Risk must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Military expenditure (current USD) or erase distinctions from Nuclear Deterrence Stability. The temporal record attached to Nuclear Escalation Risk uses Military expenditure (current USD) (MS.MIL.XPND.CD) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Nuclear Escalation Risk, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Nuclear Escalation Risk begins with a change in the factor itself and then moves through security expectations, displacement, trade continuity, public expenditure, and the physical safety of affected populations. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Nuclear Escalation Risk, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Nuclear Escalation Risk may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Nuclear Escalation Risk. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Military expenditure (current USD) dataset](https://api.worldbank.org/v2/country/all/indicator/MS.MIL.XPND.CD?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Nuclear Escalation Risk finished above its first normalized observation: 0.9451189 in 2020 versus 0.95565703 in 2024. The minimum was 0.9451189 in 2020, the maximum was 0.95565703 in 2024, and the observed sequence contained 4 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Nuclear Escalation Risk, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000000000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Nuclear Escalation Risk, the provider describes the selected series in these terms: Military spending USD, in current prices (converted at the exchange rate for the given year) The cited producer is SIPRI Military Expenditure Database, Stockholm International Peace Research Institute (SIPRI), uri: https://www.sipri.org/databases. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [United Nations Charter](https://www.un.org/en/about-us/un-charter/full-text). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Nuclear Escalation Risk caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Nuclear Escalation Risk should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global deaths from organized violence](/indicators/sec-global-armed-conflict-deaths). For Nuclear Escalation Risk, expected first-order direction is **higher**. The channel to test is changes in the frequency, geographic reach, and lethality of organized violence. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium to high when exposure is direct. A same-period correlation would not prove this impact. - [People internally displaced by conflict and violence](/indicators/sec-conflict-violence-internal-displacement). For Nuclear Escalation Risk, expected first-order direction is **higher**. The channel to test is threats to life, destruction of housing, and loss of access to livelihoods and services. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to months, with persistence measured in years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Nuclear Escalation Risk, expected first-order direction is **higher**. The channel to test is disruption of food, sanitation, vaccination, primary care, and household income. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to years. Confidence: medium and strongly exposure-dependent. A same-period correlation would not prove this impact. First-order effects for Nuclear Escalation Risk follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Nuclear Escalation Risk can reinforce or offset other Drivers in international security and war, including Nuclear Deterrence Stability, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Nuclear Escalation Risk, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Nuclear Escalation Risk.International Security and War
Labour-Market Tightness## **What this driver is** Labour-Market Tightness is a continuing, time-varying factor within macroeconomics and finance. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Labour-Market Tightness; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Labour-Market Tightness is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Productivity-Growth Potential: the two may interact or share a proxy, yet they represent different causal questions. Labour-Market Tightness is therefore a persistent analytical node, not a label for every adjacent development. Labour-Market Tightness must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Unemployment, total (% of total labor force) (modeled ILO estimate) or erase distinctions from Productivity-Growth Potential. The temporal record attached to Labour-Market Tightness uses Unemployment, total (% of total labor force) (modeled ILO estimate) (SL.UEM.TOTL.ZS) as a disclosed proxy for 2020-2025. The proxy is an observable lens, not a complete operational definition. The stored scale is inverted because a higher raw observation represents less of Labour-Market Tightness; after inversion, a higher normalized value means a stronger level of the named factor. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Labour-Market Tightness begins with a change in the factor itself and then moves through real incomes, credit conditions, employment, fiscal capacity, investment decisions, and the distribution of purchasing power. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Labour-Market Tightness, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Labour-Market Tightness may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Labour-Market Tightness. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Unemployment, total (% of total labor force) (modeled ILO estimate) dataset](https://api.worldbank.org/v2/country/all/indicator/SL.UEM.TOTL.ZS?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2025 Driver series. Across that available record, Labour-Market Tightness finished above its first normalized observation: 0.93412311 in 2020 versus 0.95208667 in 2025. The minimum was 0.93412311 in 2020, the maximum was 0.95208667 in 2025, and the observed sequence contained 5 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. The zero-to-one conversion is mechanical and preserved in lineage: Published percentage divided by 100 and clamped to [0,1]. For Labour-Market Tightness, the scale is inverted because higher raw values indicate less of the named factor. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Labour-Market Tightness, the provider describes the selected series in these terms: Unemployment refers to the share of the labor force that is without work but available for and seeking employment. The cited producer is ILO Modelled Estimates database (ILOEST), International Labour Organization (ILO), uri: https://ilostat.ilo.org/data/bulk/, publisher: ILOSTAT, type: external database, date accessed: January 17, 2026. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [IMF Data Standards Initiatives](https://www.imf.org/en/About/Factsheets/Sheets/2023/Standards-for-data-dissemination). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Labour-Market Tightness caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Labour-Market Tightness should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global consumer price inflation](/indicators/econ-global-consumer-price-inflation). For Labour-Market Tightness, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is changes in demand, financing costs, exchange rates, wages, expectations, and supply costs. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to roughly two years. Confidence: medium, varying with pass-through and policy response. A same-period correlation would not prove this impact. - [Income inequality](/indicators/econ-us-income-inequality-gini). For Labour-Market Tightness, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is unequal exposure to prices, employment losses, asset returns, taxes, transfers, and credit constraints. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: one to several years. Confidence: low for a universal sign and medium within a defined policy setting. A same-period correlation would not prove this impact. - [Learning poverty rate](/indicators/edu-learning-poverty-rate). For Labour-Market Tightness, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is household income shocks, nutrition, attendance, public budgets, and the opportunity cost of schooling. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: one school year to a generation. Confidence: low to medium. A same-period correlation would not prove this impact. First-order effects for Labour-Market Tightness follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Labour-Market Tightness can reinforce or offset other Drivers in macroeconomics and finance, including Productivity-Growth Potential, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Labour-Market Tightness, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2025 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Labour-Market Tightness.Macroeconomics and Finance
Defence-Industrial Production Capacity## **What this driver is** Defence-Industrial Production Capacity is a continuing, time-varying factor within international security and war. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Defence-Industrial Production Capacity; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Defence-Industrial Production Capacity is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Sanctions Pressure: the two may interact or share a proxy, yet they represent different causal questions. Defence-Industrial Production Capacity is therefore a persistent analytical node, not a label for every adjacent development. Defence-Industrial Production Capacity must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Arms exports (SIPRI trend indicator values) or erase distinctions from Sanctions Pressure. The temporal record attached to Defence-Industrial Production Capacity uses Arms exports (SIPRI trend indicator values) (MS.MIL.XPRT.KD) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Defence-Industrial Production Capacity, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Defence-Industrial Production Capacity begins with a change in the factor itself and then moves through security expectations, displacement, trade continuity, public expenditure, and the physical safety of affected populations. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Defence-Industrial Production Capacity, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Defence-Industrial Production Capacity may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Defence-Industrial Production Capacity. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Arms exports (SIPRI trend indicator values) dataset](https://api.worldbank.org/v2/country/all/indicator/MS.MIL.XPRT.KD?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Defence-Industrial Production Capacity finished above its first normalized observation: 0.94645116 in 2020 versus 0.94829848 in 2024. The minimum was 0.94645116 in 2020, the maximum was 0.95220943 in 2021, and the observed sequence contained 1 increases and 3 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Defence-Industrial Production Capacity, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+100000000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Defence-Industrial Production Capacity, the provider describes the selected series in these terms: Arms transfers (exports) cover the volume of transfers of major arms through sales and gifts, and those made through manufacturing licenses. Data cover major conventional weapons such as aircraft, armored vehicles, artillery, radar systems, missiles, and ships. Figures are SIPRI Trend Indicator Values (TIVs). A '0' indicates that the volume of deliveries is between 0 and 0.5 million SIPRI TIV. The cited producer is Arms Transfers Programme, Stockholm International Peace Research Institute (SIPRI), uri: https://armstransfers.sipri.org/ArmsTransfer/. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [United Nations Charter](https://www.un.org/en/about-us/un-charter/full-text). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Defence-Industrial Production Capacity caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Defence-Industrial Production Capacity should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global deaths from organized violence](/indicators/sec-global-armed-conflict-deaths). For Defence-Industrial Production Capacity, expected first-order direction is **lower**. The channel to test is changes in the frequency, geographic reach, and lethality of organized violence. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium to high when exposure is direct. A same-period correlation would not prove this impact. - [People internally displaced by conflict and violence](/indicators/sec-conflict-violence-internal-displacement). For Defence-Industrial Production Capacity, expected first-order direction is **lower**. The channel to test is threats to life, destruction of housing, and loss of access to livelihoods and services. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to months, with persistence measured in years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Defence-Industrial Production Capacity, expected first-order direction is **lower**. The channel to test is disruption of food, sanitation, vaccination, primary care, and household income. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to years. Confidence: medium and strongly exposure-dependent. A same-period correlation would not prove this impact. First-order effects for Defence-Industrial Production Capacity follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Defence-Industrial Production Capacity can reinforce or offset other Drivers in international security and war, including Sanctions Pressure, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Defence-Industrial Production Capacity, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Defence-Industrial Production Capacity.International Security and War
Greenhouse-Gas Emissions Intensity## **What this driver is** Greenhouse-Gas Emissions Intensity is a continuing, time-varying factor within climate, environment, and disasters. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Greenhouse-Gas Emissions Intensity; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Greenhouse-Gas Emissions Intensity is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Pace of Climate Transition: the two may interact or share a proxy, yet they represent different causal questions. Greenhouse-Gas Emissions Intensity is therefore a persistent analytical node, not a label for every adjacent development. Greenhouse-Gas Emissions Intensity must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Total greenhouse gas emissions excluding LULUCF (Mt CO2e) or erase distinctions from Pace of Climate Transition. The temporal record attached to Greenhouse-Gas Emissions Intensity uses Total greenhouse gas emissions excluding LULUCF (Mt CO2e) (EN.GHG.ALL.MT.CE.AR5) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Greenhouse-Gas Emissions Intensity, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Greenhouse-Gas Emissions Intensity begins with a change in the factor itself and then moves through physical exposure, ecosystem services, asset losses, adaptation costs, and the resilience of essential infrastructure. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Greenhouse-Gas Emissions Intensity, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Greenhouse-Gas Emissions Intensity may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Greenhouse-Gas Emissions Intensity. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Total greenhouse gas emissions excluding LULUCF (Mt CO2e) dataset](https://api.worldbank.org/v2/country/all/indicator/EN.GHG.ALL.MT.CE.AR5?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Greenhouse-Gas Emissions Intensity finished above its first normalized observation: 0.93797301 in 2020 versus 0.94519359 in 2024. The minimum was 0.93797301 in 2020, the maximum was 0.94519359 in 2024, and the observed sequence contained 4 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Greenhouse-Gas Emissions Intensity, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+100000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Greenhouse-Gas Emissions Intensity, the provider describes the selected series in these terms: A measure of annual emissions of the six greenhouse gases (GHG) covered by the Kyoto Protocol (carbon dioxide (CO2), methane (CH4), nitrous oxide (N2O), hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), and sulphurhexafluoride (SF6)) from the energy, industry, waste, and agriculture sectors, standardized to carbon dioxide equivalent values. This measure excludes GHG fluxes caused by Land Use Land Use Change and Forestry (LULUCF), as these fluxes have larger uncertainties. The measure is standardized to carbon dioxide equivalent values using the Global Warming Potential (GWP) factors of IPCC's 5th Assessment Report (AR5). The cited producer is EDGAR (Emissions Database for Global Atmospheric Research) Community GHG Database, Joint Research Centre (JRC) - European Commission, uri: https://edgar.jrc.ec.europa.eu/dataset_ghg2024, publisher: JRC European Commission, date published: 2024; International Energy Agency (IEA), uri: https://edgar.jrc.ec.europa.eu/dataset_ghg2024, publisher: JRC European Commission, date published: 2024. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [IPCC Sixth Assessment Report](https://www.ipcc.ch/assessment-report/ar6/). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Greenhouse-Gas Emissions Intensity caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Greenhouse-Gas Emissions Intensity should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global CO2 emissions per capita](/indicators/env-co2-emissions-per-capita-global). For Greenhouse-Gas Emissions Intensity, expected first-order direction is **higher**. The channel to test is changes in fossil-energy use, land cover, carbon sinks, and the emissions intensity of production and recovery. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: years to decades. Confidence: medium for emissions-producing activity and low for any single local intervention. A same-period correlation would not prove this impact. - [Population-weighted PM2.5 exposure](/indicators/env-pm25-air-pollution-exposure-global). For Greenhouse-Gas Emissions Intensity, expected first-order direction is **higher**. The channel to test is wildfire smoke, combustion, dust, atmospheric chemistry, and pollution-control capacity. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Greenhouse-Gas Emissions Intensity, expected first-order direction is **higher**. The channel to test is heat stress, injury, infectious disease, food insecurity, unsafe water, displacement, and health-system disruption. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium in directly exposed populations. A same-period correlation would not prove this impact. First-order effects for Greenhouse-Gas Emissions Intensity follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Greenhouse-Gas Emissions Intensity can reinforce or offset other Drivers in climate, environment, and disasters, including Pace of Climate Transition, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Greenhouse-Gas Emissions Intensity, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Greenhouse-Gas Emissions Intensity.Climate, Environment, and Disasters
Productivity-Growth Potential## **What this driver is** Productivity-Growth Potential is a continuing, time-varying factor within macroeconomics and finance. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Productivity-Growth Potential; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Productivity-Growth Potential is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Income Inequality: the two may interact or share a proxy, yet they represent different causal questions. Productivity-Growth Potential is therefore a persistent analytical node, not a label for every adjacent development. Productivity-Growth Potential must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into GDP per person employed (constant 2021 PPP $) or erase distinctions from Income Inequality. The temporal record attached to Productivity-Growth Potential uses GDP per person employed (constant 2021 PPP $) (SL.GDP.PCAP.EM.KD) as a disclosed proxy for 2020-2025. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Productivity-Growth Potential, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Productivity-Growth Potential begins with a change in the factor itself and then moves through real incomes, credit conditions, employment, fiscal capacity, investment decisions, and the distribution of purchasing power. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Productivity-Growth Potential, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Productivity-Growth Potential may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Productivity-Growth Potential. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [GDP per person employed (constant 2021 PPP $) dataset](https://api.worldbank.org/v2/country/all/indicator/SL.GDP.PCAP.EM.KD?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2025 Driver series. Across that available record, Productivity-Growth Potential finished above its first normalized observation: 0.93225594 in 2020 versus 0.94095403 in 2025. The minimum was 0.93225594 in 2020, the maximum was 0.94095403 in 2025, and the observed sequence contained 5 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Productivity-Growth Potential, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+100000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Productivity-Growth Potential, the provider describes the selected series in these terms: GDP per person employed is gross domestic product (GDP) divided by total employment in the economy. Purchasing power parity (PPP) GDP is GDP converted to 2021 constant international dollars using PPP rates. An international dollar has the same purchasing power over GDP that a U.S. dollar has in the United States. The cited producer is Staff estimates, World Bank (WB), note: Estimates are based on employment, population, GDP, and PPP data obtained from International Labour Organization, United Nations Population Division, Eurostat, OECD, and World Bank., type: estimates based on external database; International Labour Organization (ILO); United Nations (UN), publisher: UN Population Division; Eurostat (ESTAT); Organisation for Economic Co-operation and Development (OECD); World Development Indicators database, World Bank (WB). That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [IMF Data Standards Initiatives](https://www.imf.org/en/About/Factsheets/Sheets/2023/Standards-for-data-dissemination). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Productivity-Growth Potential caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Productivity-Growth Potential should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global consumer price inflation](/indicators/econ-global-consumer-price-inflation). For Productivity-Growth Potential, expected first-order direction is **lower**. The channel to test is changes in demand, financing costs, exchange rates, wages, expectations, and supply costs. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to roughly two years. Confidence: medium, varying with pass-through and policy response. A same-period correlation would not prove this impact. - [Income inequality](/indicators/econ-us-income-inequality-gini). For Productivity-Growth Potential, expected first-order direction is **lower**. The channel to test is unequal exposure to prices, employment losses, asset returns, taxes, transfers, and credit constraints. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: one to several years. Confidence: low for a universal sign and medium within a defined policy setting. A same-period correlation would not prove this impact. - [Learning poverty rate](/indicators/edu-learning-poverty-rate). For Productivity-Growth Potential, expected first-order direction is **lower**. The channel to test is household income shocks, nutrition, attendance, public budgets, and the opportunity cost of schooling. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: one school year to a generation. Confidence: low to medium. A same-period correlation would not prove this impact. First-order effects for Productivity-Growth Potential follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Productivity-Growth Potential can reinforce or offset other Drivers in macroeconomics and finance, including Income Inequality, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Productivity-Growth Potential, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2025 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Productivity-Growth Potential.Macroeconomics and Finance
Forced-Displacement Scale## **What this driver is** Forced-Displacement Scale is a continuing, time-varying factor within health, demography, and the humanitarian environment. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Forced-Displacement Scale; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Forced-Displacement Scale is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Humanitarian Access: the two may interact or share a proxy, yet they represent different causal questions. Forced-Displacement Scale is therefore a persistent analytical node, not a label for every adjacent development. Forced-Displacement Scale must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Refugees by country of origin or erase distinctions from Humanitarian Access. The temporal record attached to Forced-Displacement Scale uses Refugees by country of origin (OWID:refugee-population-by-country-or-territory-of-origin:Refugees by country of origin) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Forced-Displacement Scale, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Forced-Displacement Scale begins with a change in the factor itself and then moves through mortality and morbidity risks, care capacity, population structure, displacement, and access to essential services. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Forced-Displacement Scale, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Forced-Displacement Scale may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Forced-Displacement Scale. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Refugees by country of origin dataset](https://ourworldindata.org/grapher/refugee-population-by-country-or-territory-of-origin.csv). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Forced-Displacement Scale finished above its first normalized observation: 0.91365455 in 2020 versus 0.93586743 in 2024. The minimum was 0.91365455 in 2020, the maximum was 0.93705318 in 2023, and the observed sequence contained 3 increases and 1 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Forced-Displacement Scale, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+100000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Forced-Displacement Scale, the provider describes the selected series in these terms: Our World in Data republishes the dated series with source-specific metadata and processing notes on the chart data page. The selected column is "Refugees by country of origin". The cited producer is Our World in Data and the named original provider. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [WHO Global Health Observatory](https://www.who.int/data/gho). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Forced-Displacement Scale caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Forced-Displacement Scale should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Forced-Displacement Scale, expected first-order direction is **higher**. The channel to test is prevention, diagnosis, treatment, nutrition, maternal health, sanitation, and continuity of care. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: weeks to several years. Confidence: medium to high for direct health-system and disease channels. A same-period correlation would not prove this impact. - [First-dose measles vaccination coverage](/indicators/ph-measles-vaccine-coverage-global). For Forced-Displacement Scale, expected first-order direction is **lower**. The channel to test is procurement, cold chains, staffing, access, trust, risk communication, and missed-dose recovery. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to several years. Confidence: medium where delivery capacity or acceptance is directly affected. A same-period correlation would not prove this impact. - [Old-age dependency ratio](/indicators/mig-old-age-dependency-ratio-world). For Forced-Displacement Scale, expected first-order direction is **higher**. The channel to test is cohort size, survival, fertility, migration, labour-force participation, and retirement institutions. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: years to decades. Confidence: high for demographic arithmetic and low for short-run policy effects. A same-period correlation would not prove this impact. First-order effects for Forced-Displacement Scale follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Forced-Displacement Scale can reinforce or offset other Drivers in health, demography, and the humanitarian environment, including Humanitarian Access, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Forced-Displacement Scale, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Forced-Displacement Scale.Health, Demography, and the Humanitarian Environment
Genomic Epidemiological Surveillance## **What this driver is** Genomic Epidemiological Surveillance is a continuing, time-varying factor within health, demography, and the humanitarian environment. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Genomic Epidemiological Surveillance; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Genomic Epidemiological Surveillance is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Mental-Disorder Burden: the two may interact or share a proxy, yet they represent different causal questions. Genomic Epidemiological Surveillance is therefore a persistent analytical node, not a label for every adjacent development. Genomic Epidemiological Surveillance must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Scientific and technical journal articles or erase distinctions from Mental-Disorder Burden. The temporal record attached to Genomic Epidemiological Surveillance uses Scientific and technical journal articles (IP.JRN.ARTC.SC) as a disclosed proxy for 2020-2023. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Genomic Epidemiological Surveillance, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Genomic Epidemiological Surveillance begins with a change in the factor itself and then moves through mortality and morbidity risks, care capacity, population structure, displacement, and access to essential services. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Genomic Epidemiological Surveillance, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Genomic Epidemiological Surveillance may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Genomic Epidemiological Surveillance. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Scientific and technical journal articles dataset](https://api.worldbank.org/v2/country/all/indicator/IP.JRN.ARTC.SC?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2023 Driver series. Across that available record, Genomic Epidemiological Surveillance finished above its first normalized observation: 0.9225116 in 2020 versus 0.93067428 in 2023. The minimum was 0.9225116 in 2020, the maximum was 0.93067428 in 2023, and the observed sequence contained 3 increases and 0 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Genomic Epidemiological Surveillance, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Genomic Epidemiological Surveillance, the provider describes the selected series in these terms: Article counts refer to publications from a selection of conference proceedings and peer-reviewed journals from Scopus in science and engineering fields, according to the National Center for Science and Engineering Statistics Taxonomy of Disciplines. The cited producer is Science and Engineering Indicators, National Science Foundation (NSF), uri: https://ncses.nsf.gov/indicators. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [WHO Global Health Observatory](https://www.who.int/data/gho). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Genomic Epidemiological Surveillance caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Genomic Epidemiological Surveillance should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Genomic Epidemiological Surveillance, expected first-order direction is **higher**. The channel to test is prevention, diagnosis, treatment, nutrition, maternal health, sanitation, and continuity of care. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: weeks to several years. Confidence: medium to high for direct health-system and disease channels. A same-period correlation would not prove this impact. - [First-dose measles vaccination coverage](/indicators/ph-measles-vaccine-coverage-global). For Genomic Epidemiological Surveillance, expected first-order direction is **lower**. The channel to test is procurement, cold chains, staffing, access, trust, risk communication, and missed-dose recovery. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to several years. Confidence: medium where delivery capacity or acceptance is directly affected. A same-period correlation would not prove this impact. - [Old-age dependency ratio](/indicators/mig-old-age-dependency-ratio-world). For Genomic Epidemiological Surveillance, expected first-order direction is **higher**. The channel to test is cohort size, survival, fertility, migration, labour-force participation, and retirement institutions. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: years to decades. Confidence: high for demographic arithmetic and low for short-run policy effects. A same-period correlation would not prove this impact. First-order effects for Genomic Epidemiological Surveillance follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Genomic Epidemiological Surveillance can reinforce or offset other Drivers in health, demography, and the humanitarian environment, including Mental-Disorder Burden, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Genomic Epidemiological Surveillance, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2023 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Genomic Epidemiological Surveillance.Health, Demography, and the Humanitarian Environment
Electricity-Grid Resilience## **What this driver is** Electricity-Grid Resilience is a continuing, time-varying factor within energy and natural resources. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Electricity-Grid Resilience; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Electricity-Grid Resilience is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Energy-Storage Capacity: the two may interact or share a proxy, yet they represent different causal questions. Electricity-Grid Resilience is therefore a persistent analytical node, not a label for every adjacent development. Electricity-Grid Resilience must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Access to electricity (% of population) or erase distinctions from Energy-Storage Capacity. The temporal record attached to Electricity-Grid Resilience uses Access to electricity (% of population) (EG.ELC.ACCS.ZS) as a disclosed proxy for 2020-2024. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Electricity-Grid Resilience, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Electricity-Grid Resilience begins with a change in the factor itself and then moves through energy affordability, supply continuity, infrastructure investment, production costs, and exposure to resource shocks. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Electricity-Grid Resilience, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Electricity-Grid Resilience may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Electricity-Grid Resilience. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Access to electricity (% of population) dataset](https://api.worldbank.org/v2/country/all/indicator/EG.ELC.ACCS.ZS?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2024 Driver series. Across that available record, Electricity-Grid Resilience finished above its first normalized observation: 0.90399547 in 2020 versus 0.91926484 in 2024. The minimum was 0.90399547 in 2020, the maximum was 0.91926484 in 2024, and the observed sequence contained 3 increases and 1 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Electricity-Grid Resilience, the zero-to-one conversion is mechanical and preserved in lineage: Published percentage divided by 100 and clamped to [0,1]. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Electricity-Grid Resilience, the provider describes the selected series in these terms: Access to electricity is the percentage of population with access to electricity. Electrification data are collected from industry, national surveys and international sources. The cited producer is SDG 7.1.1 Electrification Dataset, World Bank (WB), uri: https://trackingsdg7.esmap.org/downloads, note: Data is downloaded from ESMAP website. Data is released when a new Tracking SDG7 report is released., publisher: World Bank (WB), date accessed: 2024-05-16, date published: 2023. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [International Energy Agency data and statistics](https://www.iea.org/data-and-statistics). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Electricity-Grid Resilience caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Electricity-Grid Resilience should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Access to electricity](/indicators/energy-electricity-access-global). For Electricity-Grid Resilience, expected first-order direction is **higher**. The channel to test is generation adequacy, grid reliability, fuel availability, connection costs, and investment. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: months to several years. Confidence: medium for direct supply and infrastructure constraints. A same-period correlation would not prove this impact. - [Global CO2 emissions per capita](/indicators/env-co2-emissions-per-capita-global). For Electricity-Grid Resilience, expected first-order direction is **lower**. The channel to test is changes in fossil-energy use, efficiency, electrification, dispatch, and capital-stock turnover. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: one year to several decades. Confidence: medium to high for combustion and generation-mix channels. A same-period correlation would not prove this impact. - [Population-weighted PM2.5 exposure](/indicators/env-pm25-air-pollution-exposure-global). For Electricity-Grid Resilience, expected first-order direction is **lower**. The channel to test is fuel choice, combustion efficiency, power-plant and transport emissions, and pollution controls. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to years. Confidence: medium when the driver changes combustion or local emissions. A same-period correlation would not prove this impact. First-order effects for Electricity-Grid Resilience follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Electricity-Grid Resilience can reinforce or offset other Drivers in energy and natural resources, including Energy-Storage Capacity, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Electricity-Grid Resilience, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2024 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Electricity-Grid Resilience.Energy and Natural Resources
Water Scarcity## **What this driver is** Water Scarcity is a continuing, time-varying factor within climate, environment, and disasters. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Water Scarcity; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Water Scarcity is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Biodiversity Loss: the two may interact or share a proxy, yet they represent different causal questions. Water Scarcity is therefore a persistent analytical node, not a label for every adjacent development. Water Scarcity must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Level of water stress: freshwater withdrawal as a proportion of available freshwater resources or erase distinctions from Biodiversity Loss. The temporal record attached to Water Scarcity uses Level of water stress: freshwater withdrawal as a proportion of available freshwater resources (ER.H2O.FWST.ZS) as a disclosed proxy for 2020-2022. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Water Scarcity, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Water Scarcity begins with a change in the factor itself and then moves through physical exposure, ecosystem services, asset losses, adaptation costs, and the resilience of essential infrastructure. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Water Scarcity, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Water Scarcity may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Water Scarcity. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Level of water stress: freshwater withdrawal as a proportion of available freshwater resources dataset](https://api.worldbank.org/v2/country/all/indicator/ER.H2O.FWST.ZS?date=2020:2025&format=json&per_page=20000). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2022 Driver series. Across that available record, Water Scarcity finished below its first normalized observation: 0.92123357 in 2020 versus 0.91829703 in 2022. The minimum was 0.91829703 in 2022, the maximum was 0.92123357 in 2020, and the observed sequence contained 0 increases and 2 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Water Scarcity, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+100); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Water Scarcity, the provider describes the selected series in these terms: The level of water stress: freshwater withdrawal as a proportion of available freshwater resources is the ratio between total freshwater withdrawn by all major sectors and total renewable freshwater resources, after taking into account environmental water requirements. Main sectors, as defined by ISIC standards, include agriculture; forestry and fishing; manufacturing; electricity industry; and services. This indicator is also known as water withdrawal intensity. The cited producer is AQUASTAT - FAO's Global Information System on Water and Agriculture, Food and Agriculture Organization of the United Nations (FAO), uri: https://data.apps.fao.org/aquastat/, publisher: Food and Agriculture Organization of the United Nations (FAO), date accessed: 20240529. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [IPCC Sixth Assessment Report](https://www.ipcc.ch/assessment-report/ar6/). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Water Scarcity caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Water Scarcity should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Global CO2 emissions per capita](/indicators/env-co2-emissions-per-capita-global). For Water Scarcity, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is changes in fossil-energy use, land cover, carbon sinks, and the emissions intensity of production and recovery. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: years to decades. Confidence: medium for emissions-producing activity and low for any single local intervention. A same-period correlation would not prove this impact. - [Population-weighted PM2.5 exposure](/indicators/env-pm25-air-pollution-exposure-global). For Water Scarcity, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is wildfire smoke, combustion, dust, atmospheric chemistry, and pollution-control capacity. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to years. Confidence: medium. A same-period correlation would not prove this impact. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Water Scarcity, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is heat stress, injury, infectious disease, food insecurity, unsafe water, displacement, and health-system disruption. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: days to several years. Confidence: medium in directly exposed populations. A same-period correlation would not prove this impact. First-order effects for Water Scarcity follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Water Scarcity can reinforce or offset other Drivers in climate, environment, and disasters, including Biodiversity Loss, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Water Scarcity, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2022 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Water Scarcity.Climate, Environment, and Disasters
Commercial Human-Spaceflight Capacity## **What this driver is** Commercial Human-Spaceflight Capacity is a continuing, time-varying factor within science, space, and biotechnology. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Commercial Human-Spaceflight Capacity; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Commercial Human-Spaceflight Capacity is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Rocket Reusability: the two may interact or share a proxy, yet they represent different causal questions. Commercial Human-Spaceflight Capacity is therefore a persistent analytical node, not a label for every adjacent development. Commercial Human-Spaceflight Capacity must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Annual number of objects launched into outer space or erase distinctions from Rocket Reusability. The temporal record attached to Commercial Human-Spaceflight Capacity uses Annual number of objects launched into outer space (OWID:yearly-number-of-objects-launched-into-outer-space:Annual number of objects launched into outer space) as a disclosed proxy for 2020-2025. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Commercial Human-Spaceflight Capacity, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Commercial Human-Spaceflight Capacity begins with a change in the factor itself and then moves through research capacity, deployment cost, infrastructure access, knowledge diffusion, and the governance of emerging capabilities. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Commercial Human-Spaceflight Capacity, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Commercial Human-Spaceflight Capacity may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Commercial Human-Spaceflight Capacity. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Annual number of objects launched into outer space dataset](https://ourworldindata.org/grapher/yearly-number-of-objects-launched-into-outer-space.csv). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2025 Driver series. Across that available record, Commercial Human-Spaceflight Capacity finished above its first normalized observation: 0.77636912 in 2020 versus 0.91355829 in 2025. The minimum was 0.77636912 in 2020, the maximum was 0.91355829 in 2025, and the observed sequence contained 4 increases and 1 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Commercial Human-Spaceflight Capacity, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Commercial Human-Spaceflight Capacity, the provider describes the selected series in these terms: Our World in Data republishes the dated series with source-specific metadata and processing notes on the chart data page. The selected column is "Annual number of objects launched into outer space". The cited producer is Our World in Data and the named original provider. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [UNESCO Institute for Statistics science data](https://uis.unesco.org/en/topic/science-technology-and-innovation). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Commercial Human-Spaceflight Capacity caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Commercial Human-Spaceflight Capacity should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Commercial Human-Spaceflight Capacity, expected first-order direction is **lower**. The channel to test is translation of research into diagnostics, vaccines, therapeutics, delivery systems, and affordable access. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: several years to decades. Confidence: low generally and medium for validated biomedical applications. A same-period correlation would not prove this impact. - [Internet penetration](/indicators/tech-internet-penetration-global). For Commercial Human-Spaceflight Capacity, expected first-order direction is **higher**. The channel to test is coverage, backhaul, launch and equipment costs, reliability, and last-mile access. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: several years. Confidence: low to medium for communications and satellite applications. A same-period correlation would not prove this impact. - [Global CO2 emissions per capita](/indicators/env-co2-emissions-per-capita-global). For Commercial Human-Spaceflight Capacity, expected first-order direction is **lower**. The channel to test is energy demand, process emissions, substitution, learning curves, and the composition of deployed capital. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: years to decades. Confidence: low until a technology reaches material deployment. A same-period correlation would not prove this impact. First-order effects for Commercial Human-Spaceflight Capacity follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Commercial Human-Spaceflight Capacity can reinforce or offset other Drivers in science, space, and biotechnology, including Rocket Reusability, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Commercial Human-Spaceflight Capacity, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2025 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Commercial Human-Spaceflight Capacity.Science, Space, and Biotechnology
Lunar-Exploration Intensity## **What this driver is** Lunar-Exploration Intensity is a continuing, time-varying factor within science, space, and biotechnology. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Lunar-Exploration Intensity; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Lunar-Exploration Intensity is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Mars-Exploration Intensity: the two may interact or share a proxy, yet they represent different causal questions. Lunar-Exploration Intensity is therefore a persistent analytical node, not a label for every adjacent development. Lunar-Exploration Intensity concerns the cadence, scale, and complexity of scientific, commercial, and governmental activity directed at the Moon, including launch, orbital, landing, surface, communications, and sample-return missions. General launch counts are only a remote capacity proxy because most objects do not support lunar activity. The concept differs from Mars-Exploration Intensity in communications delay, mission duration, radiation exposure, logistics, scientific objectives, and the feasibility of repeated near-term operations. The temporal record attached to Lunar-Exploration Intensity uses Annual number of objects launched into outer space (OWID:yearly-number-of-objects-launched-into-outer-space:Annual number of objects launched into outer space) as a disclosed proxy for 2020-2025. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Lunar-Exploration Intensity, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Lunar-Exploration Intensity begins with a change in the factor itself and then moves through research capacity, deployment cost, infrastructure access, knowledge diffusion, and the governance of emerging capabilities. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Lunar-Exploration Intensity, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Lunar-Exploration Intensity may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Lunar-Exploration Intensity. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Annual number of objects launched into outer space dataset](https://ourworldindata.org/grapher/yearly-number-of-objects-launched-into-outer-space.csv). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2025 Driver series. Across that available record, Lunar-Exploration Intensity finished above its first normalized observation: 0.77636912 in 2020 versus 0.91355829 in 2025. The minimum was 0.77636912 in 2020, the maximum was 0.91355829 in 2025, and the observed sequence contained 4 increases and 1 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Lunar-Exploration Intensity, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Lunar-Exploration Intensity, the provider describes the selected series in these terms: Our World in Data republishes the dated series with source-specific metadata and processing notes on the chart data page. The selected column is "Annual number of objects launched into outer space". The cited producer is Our World in Data and the named original provider. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [UNESCO Institute for Statistics science data](https://uis.unesco.org/en/topic/science-technology-and-innovation). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Lunar-Exploration Intensity caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Lunar-Exploration Intensity should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Lunar-Exploration Intensity, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is translation of research into diagnostics, vaccines, therapeutics, delivery systems, and affordable access. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: several years to decades. Confidence: low generally and medium for validated biomedical applications. A same-period correlation would not prove this impact. - [Internet penetration](/indicators/tech-internet-penetration-global). For Lunar-Exploration Intensity, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is coverage, backhaul, launch and equipment costs, reliability, and last-mile access. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: several years. Confidence: low to medium for communications and satellite applications. A same-period correlation would not prove this impact. - [Global CO2 emissions per capita](/indicators/env-co2-emissions-per-capita-global). For Lunar-Exploration Intensity, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is energy demand, process emissions, substitution, learning curves, and the composition of deployed capital. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: years to decades. Confidence: low until a technology reaches material deployment. A same-period correlation would not prove this impact. First-order effects for Lunar-Exploration Intensity follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Lunar-Exploration Intensity can reinforce or offset other Drivers in science, space, and biotechnology, including Mars-Exploration Intensity, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Lunar-Exploration Intensity, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2025 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Lunar-Exploration Intensity.Science, Space, and Biotechnology
Mars-Exploration Intensity## **What this driver is** Mars-Exploration Intensity is a continuing, time-varying factor within science, space, and biotechnology. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Mars-Exploration Intensity; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Mars-Exploration Intensity is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Biotechnology Innovation: the two may interact or share a proxy, yet they represent different causal questions. Mars-Exploration Intensity is therefore a persistent analytical node, not a label for every adjacent development. Mars-Exploration Intensity concerns the cadence, scale, and complexity of robotic or eventual crewed activity directed at Mars and its environment. Launch totals are an upstream proxy rather than a direct count of Mars missions, whose narrow windows and long transit times create a distinctive rhythm. The concept differs from Lunar-Exploration Intensity because distance, communications delay, planetary protection, entry and landing, surface autonomy, return logistics, and multi-year mission horizons change both costs and plausible spillovers. The temporal record attached to Mars-Exploration Intensity uses Annual number of objects launched into outer space (OWID:yearly-number-of-objects-launched-into-outer-space:Annual number of objects launched into outer space) as a disclosed proxy for 2020-2025. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Mars-Exploration Intensity, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Mars-Exploration Intensity begins with a change in the factor itself and then moves through research capacity, deployment cost, infrastructure access, knowledge diffusion, and the governance of emerging capabilities. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Mars-Exploration Intensity, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Mars-Exploration Intensity may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Mars-Exploration Intensity. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Annual number of objects launched into outer space dataset](https://ourworldindata.org/grapher/yearly-number-of-objects-launched-into-outer-space.csv). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2025 Driver series. Across that available record, Mars-Exploration Intensity finished above its first normalized observation: 0.77636912 in 2020 versus 0.91355829 in 2025. The minimum was 0.77636912 in 2020, the maximum was 0.91355829 in 2025, and the observed sequence contained 4 increases and 1 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Mars-Exploration Intensity, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Mars-Exploration Intensity, the provider describes the selected series in these terms: Our World in Data republishes the dated series with source-specific metadata and processing notes on the chart data page. The selected column is "Annual number of objects launched into outer space". The cited producer is Our World in Data and the named original provider. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [UNESCO Institute for Statistics science data](https://uis.unesco.org/en/topic/science-technology-and-innovation). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Mars-Exploration Intensity caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Mars-Exploration Intensity should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Mars-Exploration Intensity, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is translation of research into diagnostics, vaccines, therapeutics, delivery systems, and affordable access. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: several years to decades. Confidence: low generally and medium for validated biomedical applications. A same-period correlation would not prove this impact. - [Internet penetration](/indicators/tech-internet-penetration-global). For Mars-Exploration Intensity, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is coverage, backhaul, launch and equipment costs, reliability, and last-mile access. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: several years. Confidence: low to medium for communications and satellite applications. A same-period correlation would not prove this impact. - [Global CO2 emissions per capita](/indicators/env-co2-emissions-per-capita-global). For Mars-Exploration Intensity, expected first-order direction is **conditional: the indicator can move higher or lower depending on the form, scale, and governance of the change**. The channel to test is energy demand, process emissions, substitution, learning curves, and the composition of deployed capital. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: years to decades. Confidence: low until a technology reaches material deployment. A same-period correlation would not prove this impact. First-order effects for Mars-Exploration Intensity follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Mars-Exploration Intensity can reinforce or offset other Drivers in science, space, and biotechnology, including Biotechnology Innovation, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Mars-Exploration Intensity, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2025 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Mars-Exploration Intensity.Science, Space, and Biotechnology
Rocket Reusability## **What this driver is** Rocket Reusability is a continuing, time-varying factor within science, space, and biotechnology. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Rocket Reusability; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Rocket Reusability is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Satellite-Internet Coverage: the two may interact or share a proxy, yet they represent different causal questions. Rocket Reusability is therefore a persistent analytical node, not a label for every adjacent development. Rocket Reusability must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Annual number of objects launched into outer space or erase distinctions from Satellite-Internet Coverage. The temporal record attached to Rocket Reusability uses Annual number of objects launched into outer space (OWID:yearly-number-of-objects-launched-into-outer-space:Annual number of objects launched into outer space) as a disclosed proxy for 2020-2025. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Rocket Reusability, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Rocket Reusability begins with a change in the factor itself and then moves through research capacity, deployment cost, infrastructure access, knowledge diffusion, and the governance of emerging capabilities. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Rocket Reusability, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Rocket Reusability may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Rocket Reusability. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Annual number of objects launched into outer space dataset](https://ourworldindata.org/grapher/yearly-number-of-objects-launched-into-outer-space.csv). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2025 Driver series. Across that available record, Rocket Reusability finished above its first normalized observation: 0.77636912 in 2020 versus 0.91355829 in 2025. The minimum was 0.77636912 in 2020, the maximum was 0.91355829 in 2025, and the observed sequence contained 4 increases and 1 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Rocket Reusability, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Rocket Reusability, the provider describes the selected series in these terms: Our World in Data republishes the dated series with source-specific metadata and processing notes on the chart data page. The selected column is "Annual number of objects launched into outer space". The cited producer is Our World in Data and the named original provider. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [UNESCO Institute for Statistics science data](https://uis.unesco.org/en/topic/science-technology-and-innovation). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Rocket Reusability caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Rocket Reusability should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Rocket Reusability, expected first-order direction is **lower**. The channel to test is translation of research into diagnostics, vaccines, therapeutics, delivery systems, and affordable access. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: several years to decades. Confidence: low generally and medium for validated biomedical applications. A same-period correlation would not prove this impact. - [Internet penetration](/indicators/tech-internet-penetration-global). For Rocket Reusability, expected first-order direction is **higher**. The channel to test is coverage, backhaul, launch and equipment costs, reliability, and last-mile access. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: several years. Confidence: low to medium for communications and satellite applications. A same-period correlation would not prove this impact. - [Global CO2 emissions per capita](/indicators/env-co2-emissions-per-capita-global). For Rocket Reusability, expected first-order direction is **lower**. The channel to test is energy demand, process emissions, substitution, learning curves, and the composition of deployed capital. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: years to decades. Confidence: low until a technology reaches material deployment. A same-period correlation would not prove this impact. First-order effects for Rocket Reusability follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Rocket Reusability can reinforce or offset other Drivers in science, space, and biotechnology, including Satellite-Internet Coverage, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Rocket Reusability, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2025 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Rocket Reusability.Science, Space, and Biotechnology
Satellite-Internet Coverage## **What this driver is** Satellite-Internet Coverage is a continuing, time-varying factor within science, space, and biotechnology. It is modeled as a Driver because its intensity, capacity, or prevalence can change across reporting periods and can transmit the effects of multiple events, decisions, and institutions to later outcomes. It is not a dated event, a person, an organization, a welfare score, or an assertion that every movement in a correlated series was caused by the same mechanism. The relevant scope is the population, market, institution, infrastructure, or ecological system actually exposed to Satellite-Internet Coverage; a national or sectoral observation must not be silently generalized to the whole world. The boundary of Satellite-Internet Coverage is narrower than its category and broader than one headline. It covers the durable condition named by the title, but excludes downstream welfare outcomes that must be measured separately. It also remains distinct from Lunar-Exploration Intensity: the two may interact or share a proxy, yet they represent different causal questions. Satellite-Internet Coverage is therefore a persistent analytical node, not a label for every adjacent development. Satellite-Internet Coverage must be interpreted through its own named mechanism and evidence rather than inferred from the category label. The selected proxy identifies one observable facet; it does not collapse the Driver into Annual number of objects launched into outer space or erase distinctions from Lunar-Exploration Intensity. The temporal record attached to Satellite-Internet Coverage uses Annual number of objects launched into outer space (OWID:yearly-number-of-objects-launched-into-outer-space:Annual number of objects launched into outer space) as a disclosed proxy for 2020-2025. The proxy is an observable lens, not a complete operational definition. A higher normalized value means more of Satellite-Internet Coverage, while a lower value means less; this measurement direction is not a welfare verdict. Current weight is derived from the latest real normalized observation rather than entered as an editorial score, probability, forecast, or confidence estimate. Missing releases remain missing: the seed does not interpolate, forward-fill, extrapolate, smooth, or invent a 2025 value where the provider supplied none. ## **Mechanism** The mechanism for Satellite-Internet Coverage begins with a change in the factor itself and then moves through research capacity, deployment cost, infrastructure access, knowledge diffusion, and the governance of emerging capabilities. The first step is exposure: the change must reach identifiable households, firms, public bodies, infrastructure, ecosystems, or security actors. The second step is transmission through prices, incentives, rules, information, physical constraints, organizational capacity, or behavior. The third step is adaptation: exposed actors may substitute, relocate, delay decisions, change compliance, invest, seek protection, or pass costs to others. The final welfare effect is the net result after those responses, not the initial movement alone. For Satellite-Internet Coverage, direct and indirect effects must be separated. A direct effect changes safety, access, income, health, legal protection, service continuity, or environmental exposure without a long chain of assumptions. An indirect effect passes through fiscal space, expectations, legitimacy, supply networks, knowledge, or capital formation. Each extra link makes timing and magnitude more conditional and raises the evidentiary burden for a graph relation. Distribution is part of the mechanism. The same movement in Satellite-Internet Coverage may help one group and harm another because exposure, geography, wealth, age, legal status, occupation, insurance, and institutional quality differ. Analysis must identify who is exposed, who can adapt, and who bears transition costs; an aggregate average can conceal opposing effects. Timing also matters for Satellite-Internet Coverage. Safety, availability, prices, or service interruption can move quickly; budgets, investment, reform, demography, diffusion, and capital replacement take longer. Feedback may reinforce the move, while substitution, policy response, learning, or resilience may offset it. The timeline records movement, but every DriverIndicatorImpact must separately state lag, direction, strength, and evidence. ## **Evidence base** The primary quantitative record is the [Annual number of objects launched into outer space dataset](https://ourworldindata.org/grapher/yearly-number-of-objects-launched-into-outer-space.csv). It supplies the raw dated observations, provider unit, reporting scope, and aggregation note used for the 2020-2025 Driver series. Across that available record, Satellite-Internet Coverage finished above its first normalized observation: 0.77636912 in 2020 versus 0.91355829 in 2025. The minimum was 0.77636912 in 2020, the maximum was 0.91355829 in 2025, and the observed sequence contained 4 increases and 1 decreases. These are descriptive facts about the selected proxy; they are not an estimated causal effect on welfare. For Satellite-Internet Coverage, the zero-to-one conversion is mechanical and preserved in lineage: Fixed log envelope: log10(1+abs(raw))/log10(1+10000); envelope is based on the complete fetched 2020-2025 reporting scope and is recorded, never fitted per year. Adjacent real observations create a segment only when the value changes; direction follows the sign and strength is the absolute change. The latest point sets current weight. Normalization preserves reproducibility but neither makes unlike concepts interchangeable nor turns a proxy into a complete index. For Satellite-Internet Coverage, the provider describes the selected series in these terms: Our World in Data republishes the dated series with source-specific metadata and processing notes on the chart data page. The selected column is "Annual number of objects launched into outer space". The cited producer is Our World in Data and the named original provider. That qualification controls interpretation: coverage gaps, aggregation choices, revisions, reporting incentives, and the distance between the series and the Driver concept can all limit inference. Where the exact requested metric lacked sufficient observations and a documented fallback was used, the fallback remains visibly identified in the research artifact rather than presented as an exact measurement. Separate context comes from [UNESCO Institute for Statistics science data](https://uis.unesco.org/en/topic/science-technology-and-innovation). It supports domain vocabulary and setting, not hidden annual values or a manufactured coefficient. Evidence is strongest for observed proxy movement, weaker for the claim that Satellite-Internet Coverage caused an outcome, and weakest where the proxy captures a neighboring facet. Coincidence or association remains a hypothesis until analysis addresses confounding, reverse causality, selection, and measurement error. ## **Effect on welfare indicators** The following indicators are the concrete welfare-sensitive endpoints against which Satellite-Internet Coverage should be evaluated. Their links resolve to existing Factrail Indicator records. They do not create a causal graph edge by themselves: an active DriverIndicatorImpact is warranted only when a source supports the specific pathway, direction, lag, and scope. Because the two seed sources establish measurement and domain context rather than a universal effect size, the magnitude statements below remain qualitative and explicitly bounded. - [Under-five mortality rate](/indicators/ph-under-five-mortality-global). For Satellite-Internet Coverage, expected first-order direction is **lower**. The channel to test is translation of research into diagnostics, vaccines, therapeutics, delivery systems, and affordable access. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: several years to decades. Confidence: low generally and medium for validated biomedical applications. A same-period correlation would not prove this impact. - [Internet penetration](/indicators/tech-internet-penetration-global). For Satellite-Internet Coverage, expected first-order direction is **higher**. The channel to test is coverage, backhaul, launch and equipment costs, reliability, and last-mile access. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: several years. Confidence: low to medium for communications and satellite applications. A same-period correlation would not prove this impact. - [Global CO2 emissions per capita](/indicators/env-co2-emissions-per-capita-global). For Satellite-Internet Coverage, expected first-order direction is **lower**. The channel to test is energy demand, process emissions, substitution, learning curves, and the composition of deployed capital. No universal coefficient is supported by the seed evidence: plausible magnitude is small for remote exposure and potentially material where the channel is direct. Horizon: years to decades. Confidence: low until a technology reaches material deployment. A same-period correlation would not prove this impact. First-order effects for Satellite-Internet Coverage follow directly through named channels. Second-order effects depend on responses by governments, markets, organizations, communities, or households and may arrive later or with the opposite sign. Missing coefficients must not be replaced with round numbers; the relation stays draft or absent until evidence narrows the range. ## **Interactions and contested points** Satellite-Internet Coverage can reinforce or offset other Drivers in science, space, and biotechnology, including Lunar-Exploration Intensity, but shared timing is not enough to establish an interaction. Reinforcement is plausible when both factors act on the same bottleneck, exposed population, institutional rule, price, or infrastructure network. Offsetting is plausible when adaptation, substitution, redundancy, legal safeguards, fiscal support, or technological learning weakens the pathway. Conditional interaction is the default where one Driver changes the exposure or response to another rather than moving the welfare indicator independently. The strongest skeptical interpretation is that the selected series is too remote from Satellite-Internet Coverage, that observed movement is driven by omitted factors, and that the broad causal narrative cannot identify a stable sign or magnitude across countries and periods. That objection is especially important when multiple Drivers share the same public dataset or when the series measures an outcome adjacent to the concept rather than the concept itself. The strongest competing interpretation is that a transparent proxy, clear boundaries, and explicit uncertainty are still more useful than an unmeasured label, provided the proxy is never mistaken for proof and the graph does not fabricate unsupported edges. The record therefore separates what is established from what remains contested. Established here are the English canonical identity, the sourced 2020-2025 proxy observations, the disclosed normalization, and the distinction between Driver movement and welfare impact. Plausible but not automatically established are the indicator pathways listed above. Unresolved are universal causal magnitudes, responsibility, forecasts, and any claim outside the source scope. Publication or verification must not erase those boundaries; later evidence should update the relations and content through the governed pipeline rather than silently rewriting the history of Satellite-Internet Coverage.Science, Space, and Biotechnology