Can endemic corruption actually be broken? What Georgia, Ukraine and Romania reveal

Few questions in governance are as contested as whether deeply rooted corruption can be reversed at all, or whether reform is cosmetic and temporary. Three cases in the Factrail dataset offer a partial answer: endemic corruption can be broken quickly, but the gains are reversible.

What happened

The clearest case is Georgia's post-Rose-Revolution anti-corruption reforms. After 2003, the government dismissed roughly 16,000 traffic and patrol police and rebuilt tax, customs, civil-registry and university-admission systems around transparent, technology-based services that removed discretionary contact points. Reported bribery in police encounters fell to about 1%. The reform drive is attributed in the dataset to Mikheil Saakashvili, with the World Bank acting as an independent documenter rather than an instigator.

Ukraine offers a second model. The establishment of the National Anti-Corruption Bureau (NABU) in 2015 created a dedicated, independent body to investigate high-level corruption where none existed, signed into being by Petro Poroshenko under IMF and EU conditionality. It was reinforced by Ukraine's ProZorro transparent e-procurement system, made mandatory in 2016, which publishes the entire procurement process for public scrutiny.

Romania is the cautionary tale. Under chief prosecutor Laura Codruța Kövesi, the National Anticorruption Directorate prosecuted senior officials between 2013 and 2018, with reported conviction rates above 90%. That campaign ended with Kövesi's contested removal in July 2018 — a dismissal the European Court of Human Rights later found to have violated her rights.

Why it matters and the causal chain

Factrail organizes these events around two opposing drivers. Anti-corruption enforcement capacity captures the sustained intensity of credible, institutionalized enforcement — independent bureaus, transparent procurement, and political will to prosecute. Its antagonist is state-capture pressure: the continuing effort by concentrated private or partisan interests to bend institutions to their benefit.

The Factrail model links each reform to a rise in enforcement capacity, with Georgia's reforms also directly neutralizing capture pressure during the reform decade. Enforcement capacity in turn feeds the Corruption Perceptions Index published by Transparency International. The strongest measured signal is national: Georgia's CPI rose from a recorded low of 18 in 2003 to a peak of 58 in 2018 — far above the index's de-facto integrity threshold of 50.

What changed relative to the normal line

Georgia crossed decisively above the 50-point integrity line during its reform decade. But the dataset also records partial backsliding: the score eased to 53 in 2024 and 50 in 2025, which Transparency International attributes to weakening democracy in the region. In the Factrail model this aligns with resurgent state-capture pressure, whose estimated intensity rose from a 2018 low back toward 0.62 by 2024 — mirroring the slip in enforcement capacity from a 2017 peak.

What is verified versus uncertain

The Georgia and UNCAC-era facts are well sourced and verified. The Romania case is flagged needs_review: aggregate conviction figures come from politically contested reporting, and the article asserts no individual guilt — only that prosecutions occurred and that Kövesi's removal was later judged unlawful by the ECHR. ProZorro's savings claims are reform-advocate estimates and are similarly flagged.

Alternative explanations

CPI is a perceptions measure aggregating expert and business surveys, not a direct count of corrupt acts; movements can reflect shifting expert sentiment, economic shocks, or political narratives rather than enforcement alone. Georgia's gains coincided with broader state-building and EU-integration ambitions, so attributing them solely to enforcement would overstate the case. The dataset assigns medium confidence to most driver-to-indicator links for this reason.

What may happen next

The consistent lesson is that institutions which drive gains can be captured or dismantled. Romania's rollback and Georgia's recent slide both suggest that without sustained political will protecting enforcement independence, scores drift back toward the threshold. Sources supporting these conclusions include Transparency International's CPI releases and country profile, the World Bank's 2012 assessment of Georgia, and the public records on NABU and the DNA cited in the dossier.