Do reforms reduce corruption perceptions? Evidence from police reform in Ukraine

ABSTRACT

Corruption is a ubiquitous problem, but implementing serious anti-corruption measures is politically hard – the backlash from the corrupt is certain, but the broader political benefits are uncertain. Consequently, reformers face a dilemma in that even if they can pass reforms, it is not clear that the public will notice or reward them. To date, we have limited evidence on whether citizens notice anti-corruption reforms, how such reforms shape corruption perceptions, and whether there are real political benefits to reformers. In this article, we exploit a quasi-experiment embedded in the rollout of traffic police reform in Ukraine following the Euromaidan Revolution of 2013–14 to address each of these issues in turn. We find that citizens do respond quite accurately and durably in their evaluations of the most immediately affected institutions, but the spillovers to other political institutions are limited, short-lived, and offer little electoral gain.

KEYWORDS:

• Corruption
• Ukraine
• reforms
• Euromaidan

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/1060586X.2024.2321833

Notes

1. http://www.interaksyon.com/article/113602/ukraine-tackles-graft-with-new-us-style-police-force, accessed 6 January 2024.

2. In Table A10 of the online Appendix, we show that our findings are robust to excluding these “irregular” reform sites.

3. IRB approval for the study was obtained from UNC IRB #149841.

4. To account for possible biases resulting from uneven attrition for different types of respondents, all the results presented are based on weights designed to be proportional to attrition probabilities.

5. The survey question asked about “the police” in general rather than the traffic police, but this should bias against our finding an effect.

6. Given our focus on corruption perceptions in this paper, and because our data are not particularly well suited for doing so, we do not test the effects of the reforms on actual police corruption. However, additional analysis presented in the online Appendix (Figure A3) suggests that the improving corruption perceptions presented in the paper were based on actual reductions in bribe payments in places with police patrol reforms.

7. We report standard errors clustered at the locality level.

8. The differences between recent and older reforms were marginally significant (at .06) for Parliament but only at (p < .21) for the Presidency.

9. In Figure A2 in the online appendix, we test these hypotheses using an alternative measure of government support/opposition: the extent to which respondents identified as partisan supporters of governing parties, opposition parties, or as non-partisans. The results are broadly similar.

10. All DVs are rescaled from 0 to 1 to facilitate interpretability.

11. For space reasons the full regression results for Tables 1 and 2 are presented in the online appendix Tables A12 and A13, respectively.

12. Results are available upon request.

Additional information

Funding

The work was supported by the Smith Richardson Foundation [2015-0807].