Does everyday corruption affect how Russians view their political leadership?

Abstract

Do Russians’ personal experiences with corruption influence how they evaluate their political leaders and, if so, in what direction? In addressing this question, we focus specifically on small-scale corruption that arises when Russians encounter employees of service provision organizations. We analyze survey data gathered in the summer of 2015 from Russia to trace the links between personal corrupt behavior and political attitudes. We show that participation in everyday corruption lowers a person’s support for the political regime, both as a bivariate relationship and in a multivariate model with controls. Being involved in corrupt transactions reduces support for the regime through two indirect mechanisms: by making the political leadership’s performance seem worse and by heightening perceptions that corruption is widespread among the country’s leaders. We find no support for arguments in the literature that bribery and other forms of bureaucratic corruption help citizens pursue their needs in the face of inefficient state institutions and less developed economies. In Russia, those who frequently encounter corruption are less, not more, happy with the regime.

Keywords:

• Russia
• corruption
• legitimacy
• political support
• public opinion

Acknowledgments

We are grateful for the assistance at varying stages of the project given by Elena Bashkirova, Maria Bashkirova, Andriy Gorbachyk, and Merab Pachulia. We appreciate the research assistance provided by Samantha Gerleman, Jenny Juehring, Hu Yue, and Joseph Zhou. Donna Bahry provided valuable suggestions on an earlier version.

Notes

1. Some observers question the validity of the reported ratings in light of the heightened repression of recent years. Yet, as Frye et al. (2016) have shown, list experiments, which allow respondents to indicate their opinion without the interviewer knowing what it is, bear out the high level of support.

2. A political regime refers to “the formal and informal organization of the center of political power, and of its relations with the broader society. A regime determines who has access to political power, and how those who are in power deal with those who are not” (Fishman 1990, 428; see similar definitions by Schmitter and Karl 1991, 76 and Munck 1996, 4–6).

3. Both Manzetti and Wilson (2006) and Klašnja and Tucker (2013) find that interactions between perceptions of corruption and evaluation of economic performance influence confidence in the political regime. This is a different question than we ask since we focus on citizens’ behavior. We do, however, relate those behaviors to perceptions of elite corruption and of the regime’s performance in the multivariate analyses that follow.

4. For reviews of the literature on this point, see Weatherford (1984), Finkel, Muller, and Seligson (1989), McAllister (1999), and Lewis-Beck and Stegmaier (2007); regarding Russia, see Treisman (2011, 2014).

5. The Appendix for this paper contains more information on the survey (see Supplementary material).

6. Based on gross regional product per capita in 2012 (Russian Federal State Statistics Service 2014, 349–350).

7. For this and subsequent multivariate analyses, we report heteroscedasticity-corrected robust standard errors to control for skewness in our measure of corruption engagement (Kaufman 2013). They do not change the results substantially for any coefficient. In addition, we reran all analyses excluding the four cases with the highest values (18 or 19 on our measure). Because doing so resulted in only trivial changes in the model fits or coefficients, we report the results using all the cases.

8. We constructed this and our other index variables using Stata’s alpha command, which first standardizes each variable and then takes the average of the standardized score across the five variables. Table A2 in the online appendix provides summary information on all the measures we employ.

9. Although the use of the word “duties” might seem to elicit views of job performance, these questions more probably elicited more broadly based reactions of support. They came early in the questionnaire, prior to questions that asked about leadership performance in specific areas.

10. Because the five component variables are strongly correlated with each other, the results we report using the combined index variable remain unchanged substantively when it is replaced with any of the component variables.

11. We cannot definitively rule out that the causal direction in this relationship is the opposite of the one we defend. It might be that supporters of the leadership have lower likelihoods of encountering street-level corruption or of reporting such involvement. For several reasons, however, we find this implausible. Tables 4–6 support the view that corruption engagement is causally prior, by revealing pathways through which corruption engagement helps explain the two attitudes that most strongly influence regime support. In addition, we can examine the corruption engagement of those who voted for Putin in the 2012 presidential election vs. others. This behavioral indication of support is a useful proxy that was not part of our attitudinal measure of regime support. We find only a small difference: 14% of those who voted for Putin reported giving a bribe at any point over the past year, while 15% of those who voted for a different candidate had done so. Using our measure of corruption engagement, which ranges from 0 to 19, the difference is also small: an average score of .69 for Putin voters vs. .82 for others. Neither difference reaches statistical significance in a t-test. We suspect that, to the extent that Putin voters report slightly less engagement in corruption, this is influenced by the higher proportion of Putin voters who are elderly and female, both of which (in our study and others) are less engaged in corruption. Regarding the possibility that regime supporters underreported their corruption engagement, the survey experiment we described earlier shows little evidence of bias in reporting overall.

12. Although not all of the organizational sectors we asked respondents about are state agencies, Russians understand them all, evidently, as being linked with the governing regime.

13. As discussed above, after asking whether respondents had contact with each type of official and prior to mentioning bribes, gifts, or favors, we asked whether that experience was satisfactory, somewhat satisfactory, somewhat unsatisfactory, or very unsatisfactory. We used the row mean of the resulting 13 variables.