6 sub-national regions - 1995-2022. Latest country-level SCI: 75.2 (2022).
The most corrupt region of Latvia is Pieriga region (73.4/100), and the least corrupt is Kurzeme region (77.5/100), on a 0-100 scale where 0 is most corrupt and 100 is least. All 6 regions are ranked below.
The Subnational Corruption Index is scored 0–100 on the same orientation as Transparency International's Corruption Perceptions Index (higher = cleaner). The difference: the CPI and the World Bank measure the perceptions of experts and business executives and report one national number, while the SCI is built from what ordinary households perceive and experience — and it breaks Latvia into 6 regions.
| Index | Latest | Scale | Based on |
|---|---|---|---|
| Subnational Corruption Index (SCI) | 75.2 | 0–100 (100 = cleanest) | Households — perception & experience |
| Transparency International CPI | 59 | 0–100 (100 = cleanest) | Experts & business surveys |
| World Bank Control of Corruption | 0.7 | ≈ −2.5 to 2.5 | Aggregated expert sources |
Most recent year — SCI 2022, CPI 2022, CCI 2022. The SCD also publishes region-anchored SUB-CPI and SUB-CCI series so the CPI and World Bank measures can be read at the regional level.
Country-level mean of the Subnational Corruption Index, averaged across all 6 regions. Higher values mean cleaner governance (0 = most corrupt, 100 = least corrupt).
Source: SCD v1.0 (Comprehensive Dataset). Years where the panel was filled by inter/extrapolation are included in the country mean.
Each row is a sub-national region with its latest SCI score. Click into any row for the region's full time series.
| # | Region | GDL code | Years | Latest SCI |
|---|---|---|---|---|
| 1 | Pieriga region | LVAr102 | 1995-2022 | 73.4 |
| 2 | Riga region | LVAr101 | 1995-2022 | 74 |
| 3 | Zemgale region | LVAr105 | 1995-2022 | 74.5 |
| 4 | Vidzeme region | LVAr103 | 1995-2022 | 76 |
| 5 | Latgale region | LVAr106 | 1995-2022 | 76.2 |
| 6 | Kurzeme region | LVAr104 | 1995-2022 | 77.5 |
Per-country panel slice in CSV format. Columns: iso, country, gdl, region, year, sci.
Direct dataset citations should reference the Figshare deposit (see How to cite).