Electoral Studies 36 (2014) 15e27

Contents lists available at ScienceDirect

Electoral Studies

journal homepage: www.elsevier.com/locate/electstud

Ethnicity and electoral manipulation in Russia

* Regina Goodnow a, Robert G. Moser b, , Tony Smith b a Department of Political Science and Public Administration, University of Wisconsin-La Crosse, La Crosse, WI 54601, USA b Department of Government, University of Texas at Ausin, Austin, TX 78712, USA article info abstract

Article history: To what extent does electoral manipulation follow ethnic lines in Russia? Using an original Received 30 January 2013 dataset based on raion-level data, we find that the “ethnic component” of electoral Received in revised form 30 March 2014 manipulation is more nuanced than previous studies have suggested. Electoral manipu- Accepted 21 May 2014 lation was most prevalent in majority-minority raions across ethnic and non-ethnic as well Available online 16 June 2014 as richer and poorer regions. We argue that concentrations of ethnic minorities provide: (1) greater incentives for electoral manipulation by the central state and regional elites in Keywords: order to signal political dominance and (2) greater capacity to carry out electoral Electoral manipulation Ethnicity manipulation through networks of local co-ethnic elites. However, multilevel analyses Russia suggest that the extent of electoral manipulation was also strongly contingent on regional context. Electoral manipulation was significantly higher in the more politically volatile Muslim regions, while socioeconomic differences among regions, by contrast, had no discernible effect. © 2014 Elsevier Ltd. All rights reserved.

1. Ethnicity and electoral manipulation: the case of presumption that the main threat from ethnic diversity is Russia conflict and competition between ethnic groups. In this paper, we offer a different perspective on the Scholars have long noted the dilemmas that an ethni- interaction between ethnicity and failed democratization. cally divided society faces when trying to democratize. We argue that, under certain conditions, geographically Ethnic divisions can promote armed conflict, mass violence, concentrated ethnic minorities may contribute to authori- discriminatory state policies, and regime illegitimacy. tarianism through collusion with (rather than competition Moreover, recent scholarship has suggested that democ- against) central authorities, especially in the practice of ratization may exacerbate these problems by providing manipulating elections. Ethnic minorities may become an elites with incentives and opportunities to use nationalism important “constituency” for a competitive authoritarian for electoral gain (Mansfield and Snyder, 2005; Snyder, regime due to the political economy of elections dominated 2000; Wilkinson, 2004). Those hoping to manage these by patronage. Under these types of elections, central au- problems often look to institutions to either co-opt mi- thorities and regional elites enter into a patronage rela- norities by offering security through greater autonomy and tionship in which the former provides economic and representation (Lijphart, 2004; Saideman et al., 2002)or political benefits to targeted regions in exchange for votes bridge ethnic differences through institutions designed to delivered by regional authorities during national elections. foster multi-ethnic electoral coalitions (Horowitz, 2003; Regions with large concentrations of ethnic minorities are Reilly, 2002). Throughout this literature, there is a particularly susceptible to these political arrangements because central authorities may view such regions as a particularly formidable potential threat to their authority. * Corresponding author. E-mail addresses: [email protected] (R. Goodnow), rmoser@ Central authorities may thus have additional incentives to austin.utexas.edu (R.G. Moser), [email protected] (T. Smith). manipulate elections in these regions in order to signal http://dx.doi.org/10.1016/j.electstud.2014.05.005 0261-3794/© 2014 Elsevier Ltd. All rights reserved. 16 R. Goodnow et al. / Electoral Studies 36 (2014) 15e27 political dominance and deter opposition (see Simpser, electoral systems in divided societies need to provide in- 2013). Moreover, regions with high concentrations of centives for voters to pool their votes behind moderate, ethnic minorities may provide more conducive environ- multi-ethnic coalitions (Horowitz, 1985, 2003; Reilly, ments for electoral manipulation because they tend to be 2002). Regardless of whether the institutional remedy isolated regions with low levels of socioeconomic devel- calls for greater inclusion or incentives for multi-ethnic opment, weak civil societies, and long histories of patron- coalition-building, both consociational and centripetal ap- eclient relations. Finally, local elites in minority regions proaches view the problem of ethnic division in terms of have resources, in the form of ethnic networks, that may ethnic groups mobilizing against one another or the state in make it easier to engage in patronage and other forms of exclusive organizations. By examining the relationship electoral manipulation (see Hale, 2003). Thus, our argu- between ethnic minorities and electoral manipulation ment is two-fold. In Russia, there tends to be more electoral benefiting the dominant party in a competitive authori- manipulation in regions with concentrations of ethnic mi- tarian system, the Russian experience with ethnic electoral norities because: (1) the state feels particularly vulnerable mobilization provides an alternative dilemma e the pros- in minority regions and thus has more political reasons to pect of patronage systems and ethnic electoral machines engage in electoral manipulation and (2) it is easier to that work in tandem with a dominant party to undermine conduct electoral manipulation in minority regions due to political competition. the existence of minority-based machine politics. Russia is a particularly interesting and valuable case for Russia offers an excellent example of this phenomenon. the study of ethnic politics because it contains many Despite centralizing policies that undermined regional different ethnic groups that vary considerably along de- autonomy and a devastating war in Chechnya that has mographic, socioeconomic, and cultural lines. Moreover, spawned violence across the Caucasus, non-Russian re- Russia's system of ethnic federalism politicized and insti- gions have provided some of the strongest support for tutionalized ethnic identity by promoting geographic con- Vladimir Putin and his “party of power” (Marsh and centration and reduced assimilation of the groups with Warhola, 2001). Scholars have noted that evidence of ethnic homelands, which served to enhance their political electoral malfeasance, such as extremely high levels of capacity (Gorenburg, 2003). Finally, the country's system of voter turnout and support for the party of power, has been ethnic federalism promoted a degree of regional autonomy more endemic in Russia's ethnic republics than in its and minority control over regional politics, particularly Russian-dominated oblasts (see, in particular, Myagkov during the 1990s, when President Boris Yeltsin ruled over a et al., 2009). However, analyses of electoral fraud in system of asymmetrical federalism that provided rather Russia have tended to emphasize the detection of fraud and widespread autonomy to regional elites in non-Russian its magnitude on a national scale. In this paper, we examine regions (Stoner-Weiss, 2001). Subsequent recentralization geographic patterns to try to uncover the determinants of of state authority under President Vladimir Putin has electoral manipulation. While various forms of electoral greatly curtailed the power of regional elites but also manipulation have reportedly become widespread in arguably increased incentives for electoral manipulation Russia, they are not uniformly experienced across Russia's along ethnic lines to signal political control in previously vast territory. By carefully examining where electoral restive regions. All of these factors provide an interesting manipulation has and has not tended to occur within context in which to examine the interplay between Russia's 2700þ raions (which are roughly equivalent to U.S. ethnicity and elections in an atmosphere of competitive counties), we hope to gain some leverage on how socio- authoritarianism. economic and demographic factors affect electoral manip- We will ascertain the differences (or lack thereof) in the ulation. In particular, we use raion-level electoral and incidence of electoral manipulation between Russian and ethnicity data to explore the potential correlation between non-Russian regions using electoral and census data dis- the demographic concentration of ethnic minorities in aggregated to the raion level, which is the smallest Russia and irregular electoral patterns. geographic unit at which data are available. Empirically, we The linkage between ethnicity and electoral manipula- examine a relatively basic set of questions: Are raions with tion in Russia is important in practical terms for the high concentrations of non-Russian minorities more likely development of the postcommunist Russian state, but it to show signs of electoral manipulation than raions where also sheds light on broader issues of democratic consoli- Russians predominate? If so, are these patterns mitigated dation in ethnically divided societies. Many scholars focus by contextual factors such as rural/urban divisions or on the interrelationship between ethnic conflict and regional characteristics such as levels of economic devel- democratization (Snyder, 2000; Wilkinson, 2004) and a opment, resource-dependent economies, or ethnic differ- thriving body of literature has arisen around how to design ences between Muslim and non-Muslim minority regions? democratic institutions to mitigate the prospect of inter- ethnic conflict. Two competing perspectives have domi- 2. Electoral manipulation in Russia nated this debate. Scholars led by Arend Lijphart have recommended proportional representation, decentraliza- It is important to demarcate what types of phenomena tion, and the regional autonomy of ethnic minorities as a we are trying to capture when studying electoral manipu- method of cooptation within a broader system of con- lation in a competitive authoritarian regime such as Russia. sociationalism (see Lijphart, 1977, 2004). A rival school of Electoral manipulation is a multifaceted enterprise that thought known as centripetalism has countered that comes in a wide variety of forms ranging from illicit acts of consociational systems lock in ethnic divisions and that electoral fraud such as ballot-box stuffing, misreporting R. Goodnow et al. / Electoral Studies 36 (2014) 15e27 17 electoral returns, vote-buying, and the illegal denial or re- Therefore, we follow the example of scholars (e.g., Powell, striction of access to polling stations to acts that are not 1989; King, 2001; Myagkov et al., 2009; Mebane and necessarily illegal but nonetheless use unacceptable means Kalinin, 2009; Lukinova et al., 2011) who have adopted an to change electoral outcomes in favor of a particular indirect approach to gauge electoral misconduct based contestant such as various forms of pressure on voters and largely on quantitative indicators, usually unrealistically various forms of patronage (see Lehoucq, 2003; Simpser, high voter turnout levels and anomalous distributions of 2013). votes across parties. There are several advantages to this In this paper, we consciously employ the broader term approach. First, the evidence of fraud is mined from official electoral manipulation to capture the wide range of activ- electoral results and is thus available across a wide range of ities that render an electoral context manipulated (as countries as well as within countries over time, providing opposed to clean) (Simpser 2013: 32); rather than the opportunities for comparative analyses. Second, this narrower term electoral fraud, which is often defined as approach provides a systematic and uniform application of actions that attempt to affect elections and break the law criteria presumed to indicate manipulation for all electoral (Lehoucq, 2003: 235). Academics, journalists, and activists administrative units (in our study raions) rather than a have provided extensive indirect and direct evidence that method dependent on either uneven coverage by monitors both types of activities e electoral manipulation such as or the opinions of respondents to a survey whose answers coercive pressure and threats toward voters as well as may depend on interpretations of the question or willing- illegal acts of electoral fraud such as ballot-stuffing, mul- ness to respond candidly. Third, it provides a way to tiple voting, and officials miscounting votes and changing examine variation across time and space in the occurrence electoral results e have occurred in Russia (Myagkov et al., of electoral manipulation by differentiating between re- 2009; Fish, 2005; White, 2011).1 Our analyses using indirect gions where fraud seemed more prevalent and regions that measures of electoral manipulation cannot differentiate had cleaner elections. This allows for the investigation of between illegal acts of electoral fraud (e.g., ballot-stuffing causal relationships linking environmental factors found or tampering with results) and acts of electoral manipula- within electoral units and the occurrence of irregular tion (e.g., inappropriate pressure, threats, or bribes to vote a election results. Of course, such indirect proxy measures certain way). Instead, our goal is to analyze geographic are not perfect and suffer from some of the same limita- variation across raion-level electoral contexts that have tions as other measures of electoral manipulation, such as results indicative of some form of electoral manipulation approximations that under- or over-report the magnitude and use various socioeconomic and demographic correlates of the behavior. However, in this study we are not trying to to ascertain where relatively manipulated electoral con- measure how many votes were “stolen” in a given election tests tend to occur in Russia. but rather are striving to capture general patterns of where electoral manipulation seemed more or less prevalent 2.1. High voter turnout as an indicator of electoral across Russia and then test whether those regions deemed manipulation in Russia to experience more manipulated elections shared crucial common characteristics. With this in mind, we argue that While the myriad methods of electoral manipulation as long as our proxy measure provides a sufficiently complicate its conceptualization, its clandestine nature also convincing approximation of manipulation, we can use it to presents scholars with severe obstacles to empirical explore patterns of electoral manipulation, even if it does observation and measurement. As Lehoucq (2003: 235) not fully capture the extent and complexity of the notes, instances of electoral manipulation are acts “that phenomenon. only its victims want publicized.” Therefore, sources that We use very high voter turnout as our proxy for elec- allow scholars a window into these behaviors are severely toral manipulation. There are two reasons to suspect that limited and often biased in one way or another. Ethno- very high turnout is a good indicator of electoral manipu- graphic observation, opinion surveys, and legal challenges lation. First, as Myagkov et al. (2009) note, in the absence of of electoral results and practices offer direct measures of compulsory voting, it is quite plausible that regions that the behavior in question but suffer from problems of experience extremely high percentages of voters going to partisan bias, misreporting by respondents and observers, a the polls are experiencing some form of electoral manip- lack of comprehensive coverage of the variety of ways to fix ulation e voter pressure, ballot-stuffing, vote-buying, an election, as well as logistical problems of observers not tampering with results, patronage-based mobilization, being everywhere at all times (Lehoucq, 2003; Carothers, and/or some combination of these practices e especially if 1997).2 many other regions within the same country experience Direct measures of electoral manipulation are not much lower voter turnout. Moreover, the distribution of available in any systematic fashion for Russian elections. votes across parties also strongly suggests that electoral manipulation played a role in high voter turnout in Russia because a single party, United Russia, benefitted dis- 1 Western media outlets have provided extensive coverage of instances proportionally from large increases in turnout (see of Russian electoral fraud and manipulation. Within Russia, Golos, an Myagkov et al., 2009 on this point as well). NGO devoted to exposing electoral irregularities, has provided extensive As the empirical data plots in Figs. 1a and b, and 2a and and systematic evidence of electoral manipulation, including video evi- fi dence of ballot-stuffing. Author interview in Moscow (June 2014). b demonstrate, higher voter turnout only bene tted the 2 e For a defense of using legal challenges to electoral results as a proxy party of power United Russia received an increasing for electoral fraud see Ziblatt (2009). share of the vote as reported turnout increased, in contrast 18 R. Goodnow et al. / Electoral Studies 36 (2014) 15e27 to other parties, whose vote shares were relatively stable at lower turnout levels but actually decreased with higher turnout levels. As Myagkov et al. (2009) and Simpser (2013: 173) have pointed out, there is no logical reason why vote share for individual parties or candidates should vary in a systematic way with voter turnout. More voters at the polls should mean more supporters for all parties in competition at roughly the same vote share they garnered in other, lower-turnout contexts. But in Russia, as Fig. 1a and b starkly display, United Russia saw a dramatic and consistent increase in vote share with increases in voter turnout. Given these voting patterns, we operationalize electoral manipulation in a dichotomous fashion, classi- fying raions as fraudulent when reported voter turnout within the raion was at least as high as 75%. The 75% turnout threshold is a rather conservative estimate of potentially manipulated compared to non-manipulated raions for a context without mandatory voting laws. It also covers a greater number of potential instances of manipulation across a greater diversity of regional con- texts than would a higher turnout threshold. Moreover, a t-test confirmed the statistical significance of the different (and larger) mean in United Russia's vote share for the group of raions with more than 75% turnout compared to the group of raions with less than 75% turnout.3 Finally, the 75% threshold offers a significant distinction in the nature of voter turnout within Russian elections. For the Fig. 1. (a and b) Turnout and United Russia's vote share for 2007 and 2011. 2007 election, 26% of the raions in our dataset had turnout levels above the 75% threshold; the average turnout for raions above that threshold was 86% while the average turnout for raions under the threshold was 61% e a 25% Putin. Indeed, one of the persistent characteristics of difference. For the 2011 election, 20% of the raions in our Russian politics under the leadership of Vladimir Putin has dataset had turnout levels above the 75% threshold; the been his consistently high approval ratings hovering at 70- average turnout for raions above that threshold was 87% 80% throughout the period, according to opinion polls while the average turnout for raions under the threshold (Treisman, 2012). This has led even scholars who have was 57% e a 30% difference. asserted that the Russian state has perpetuated large-scale Of course, there is an alternative explanation for higher electoral manipulation also to concede that Putin and his than normal voter turnout in Russia. Rather than being a party would have won national elections during the 2000s product of electoral manipulation, it could be the case that quite easily and decisively in completely clean electoral raions with a combination of very high voter turnout and contests (Fish, 2005; Myagkov et al., 2009; Mebane and very high vote share for United Russia are simply areas Kalinin, 2009; Treisman, 2012). characterized by a high concentration of voters who are Why not attribute extremely high voter turnout in energized supporters of the regime and President Vladimir certain raions in Russia to positive signs of genuine ideo- logical or personal support for Putin and his party and overall regime support rather than our pessimistic asso-

3 Using other turnout thresholds (e.g., 80%, 85%, and 90%) produced ciation of higher turnout with electoral malfeasance? comparable results in our quantitative analyses of the determinants of Several patterns militate against the more positive view of electoral manipulation. More specifically, the significant regional effects extremely high turnout in Russia and point toward our from using the 75% threshold (presented in Tables 2 and 3 in the position that high turnout is a reasonable and even empirical section) were also significant at higher thresholds. Similarly, compelling proxy for electoral manipulation. First, average the variables with no significant effects at the 75% threshold, including the intercept effects for regional levels of economic development and voter turnout nationwide was not overly high (64% in natural resources wealth, were also insignificant when using higher 2007 and 60% in 2011), suggesting that high approval thresholds. (The one exception was a significant negative effect for the ratings for Putin and his party did not uniformly translate level of economic development when using a 90% threshold). We also into extremely high voter turnout across the country. In noted one deviation from the otherwise consistently significant key raion-level “percent minority” variable; it was no longer significant at the fact, the raions we identify as having high turnout are a 90% threshold when we accounted for whether or not raions were in relatively small minority (about 20%) of the total raions in ethnic federal regions. This could be explained by the much smaller the country. Second, patterns of very high voter turnout in number of raions meeting that very high threshold and their concen- Russia did not correspond to socioeconomic trends that tration within non-Russian regions. On the whole, even with these dif- typically drive greater voter participation e for example, ferences, the conclusions that we draw from using the 75% threshold are consistent with results from even more conservative indicators of sus- excessively high turnout was not positively correlated pected manipulation. with more prosperous raions. Finally, raions with R. Goodnow et al. / Electoral Studies 36 (2014) 15e27 19

3. Determinants of electoral manipulation

Lehoucq's (2003) review of the scholarly literature on electoral fraud notes that there is very little systematic analysis of the determinants of electoral manipulation and even less general consensus regarding common causes that can explain variation in these practices. For example, some scholars have found that complaints of legal infractions of electoral laws are more common in urban areas than rural areas (Lehoucq, 2003). However, Lehoucq (2003: 251) rightly notes that this may be a function of the propensity of urban populations to mobilize for their political rights and elites to engage in more detectable forms of electoral manipulation in urban settings than in rural ones where united networks of political machines and economic elites may be able to use more subtle modes of patronage and electoral “influence” to produce the results they desire. Most commonly, scholars assume that elites manipulate elections in order to win and thus expect manipulation to occur in situations where it will have the greatest impact on election results (Lehoucq, 2003: 249e251; Ziblatt, 2009: 2). Therefore, manipulation is expected to be most preva- lent in regions that have greater electoral competition, assuming that elites' efforts to skew results in their favor in competitive electoral environments will have the greatest impact on the distribution of seats and thus political power to parties. Electoral manipulation in regions where the incumbent party already enjoys widespread support pre- sumably would be unnecessary and, conversely, efforts to manipulate results in situations where victory is a long- shot would be risky (because it would take more egre- gious forms, making it more easily detected). Therefore, fixed elections are expected to occur more often in more Fig. 2. (a and b) Turnout and vote share for select parties in 2007 and 2011. competitive districts and under electoral rules such as majoritarian or plurality systems where they can have the extremely high turnout were located in precisely those greatest impact on who wins and loses elections (Ziblatt, regions that one would expect to experience some of the 2009: 3). greatest degree of discontent and disillusionment with the Simpser (2013) challenges the notion that electoral Russian state. It is difficult to imagine that results showing manipulation is primarily an “election-winning” phenom- greater than 95% turnout and 95% support for United enon. Rather, he argues that often elites manipulate elec- Russia in places like Chechnya or Dagestan, where violent tions in contexts and under circumstances that do not attacks against the state are a regular occurrence, are substantially contribute to simply winning the election at driven by anything other than massive electoral manipu- hand. Electoral manipulation is often carried out in cir- lation. As will be explained below, the more persuasive cumstances in which the incumbent authorities would explanation for the geographic distribution of extremely have won the election without resorting to cheating. high turnout is that these are areas where elections are Moreover, many acts of manipulation are brazen and not highly controlled and manipulated. Simpser asserts that concealed from public view. Simpser argues that such the very fact that Putin did not need to fix the elections in behavior serves important functions by signaling political order to win them is evidence that the flagrant manipu- dominance of the incumbent forces, making apparent the lation of elections in Russia is used to signal dominance power disparity between those in power and those in the rather than win elections (Simpser, 2013:176e183). We nascent opposition, and, thus, deterring future political agree. Our analysis showing the correlation between mobilization against the state (Simpser, 2013: pp. concentrations of ethnic minorities and electoral manip- 276e283). As noted above, scholars observing Russian ulation at the raion level in Russia suggests that ethnicity elections have also claimed that Vladimir Putin and his plays a crucial role in the targeting of electoral manipu- party would have won elections in the 2000s without lation for political purposes. Higher concentrations of resorting to electoral manipulation (Fish, 2005; Myagkov ethnic minorities in raions increase the incentives for the et al., 2009; Mebane and Kalinin, 2009; White, 2011; state to signal dominance due to perceived vulnerability Treisman, 2012), raising the question of why and under and increase the capacity to perpetuate electoral manip- what conditions such manipulation has taken place. ulation due to tendencies toward ethnic-based political Ziblatt (2009) notes that economic inequality was a key machines (Hale, 2003). feature of the regions that experienced the highest number 20 R. Goodnow et al. / Electoral Studies 36 (2014) 15e27 of legal challenges against electoral malfeasance in 19th- state into 15 successor states (Beissinger, 2002)aswellas century Germany. He argues, in part, that powerful eco- drives for autonomy by ethnic republics in the 1990s nomic interests had great resources at their disposal to (Treisman, 1997; Gorenburg, 2003). perpetuate electoral fraud, thus introducing the role of the Second, as Hale (2003, 2007) argues, the institutional capacity of powerful interests to translate political and legacy, political economy, and potential for ethnic mobili- economic advantages into electoral levers of control and zation found within ethnic republics in Russia have pro- manipulation (Ziblatt, 2009: 3). vided these regions with greater resources for patronage- In a related body of literature, scholars on American based machine politics than other areas of the country politics have long noted the connections between ethnic with much smaller minority populations. While regional minority groups, urban political machines, and electoral governors across Russia had significant access to resources manipulation. Patronage-based political machines in that could be exchanged for electoral support, minority American cities were often, although not exclusively, based elites in ethnic republics had additional incentives and on clientelistic ties with ethnic groups in which the party mechanisms to “buy” electoral support. Governors in provided patronage in the form of jobs, social services, and ethnic regions could provide specific benefits in education, opportunities for social mobility in exchange for loyalty to employment, contracts, and status to the titular minority the party at the polls (Dahl, 1961; Cornwell, 1964; group (e.g., Tatars in Tatarstan, Bashkirs in Bashkortostan) Wolfinger, 1972; Inglot and Pelissero, 1993). The key as- in exchange for electoral support from the target group. pects of this relationship included the party's access to state Moreover, this exchange could be monitored and enforced resources that could be distributed to loyal groups, the more easily since geographic concentrations of different socioeconomic vulnerability of target groups making them ethnic groups allowed clear indicators of the “ethnic vote” likely targets of machine politics, and the influence of (Hale, 2007: 231). More importantly, the bonds of ethnicity ethnic voting that made it easier to mobilize as a voting produced norms and expectations on the part of minority bloc (Cornwell, 1964). elites and minority voters within ethnic regions that made strict enforcement less necessary. Since ethnicity was 3.1. Why might regions with minority concentrations in politicized and institutionalized in ethnic-based regions, Russia experience more manipulation? there was an understanding on both sides of the exchange that the minority elites in ethnic regions would provide As exemplified by Kanchan Chandra's (2004) study of advantages for their co-ethnic voters and, in exchange, ethnic parties in India and the scholarship on American could count on strong electoral support from their clients urban political machines (see, especially, Cornwell, 1964), (Hale, 2007: 231). Given this environment, national elites patronage and ethnicity often appear to go hand in hand. had greater incentives to co-opt the minority elites in Chandra offers a compelling model for explaining this link ethnic regions within their own clientelistic exchange e between ethnicity and patronage based on notions of a economic and political benefits in exchange for electoral patronage-democracy, marked by a large public sector in support in national elections e not only to send a signal of which the state has control over jobs and services, and an control alluded to above but also because the minority electoral environment of limited information. Under such elites could deliver the votes. conditions, which arguably also characterizes the post- In short, beginning well before the emergence of communist Russian state beginning under Yeltsin, voters widespread electoral manipulation under Vladimir Putin in and elites develop reinforcing expectations of ethnic the 2000s, the political context within ethnic republics favoritism. In a self-reinforcing cycle, voters tend to support provided both the incentives (in the form of conveying co-ethnics because they expect politicians to provide ma- political dominance in regions seen as potentially volatile terial benefits to members of their own group and politi- and threatening to state power) and resources (in the form cians feel compelled to fulfill this expectation of ethnic of elite networks, political machines, and co-ethnic codes of favoritism in order to attain and retain office (Chandra, loyalty and deference) necessary for the coordination of 2004: Chs. 3e4). higher levels of electoral manipulation. Building from the work on ethnic-based patronage and the determinants of electoral manipulation, one might 4. Ethnic federalism and ethnic electoral politics in expect that geographically concentrated ethnic minorities Russia in Russia may provide a favorable environment for patronage and electoral machinations undermining Russia has a significant minority population e close to genuine electoral competition for reasons related to both 20% e similar in size to many countries in the region such incentives on the part of national and regional elites for as Bulgaria, Romania, and Slovakia. However, unlike these control over electoral processes as well as opportunities states, Russia lacks a single major minority group. Instead, and resources that make such control more possible. First, the non-Russian population is composed of literally dozens as suggested by Simpser (2013), national elites in Russia of very small groups. The largest minority group of the may have increased incentives to engage in widespread Russian Federation, Tatars, makes up less than 4% of the fixing of election results in ethnic regions as opposed to population and only three groups (Tatars, Ukrainians, and regions dominated by the majority ethnic group due to a Chuvash) comprise over 1% of the country's population. perceived need for greater control in ethnic regions. After The composition of Russia's ethnic population cannot be all, ethnic mobilization within the non-Russian union re- properly understood without acknowledgment of its publics was a principle cause of the collapse of the Soviet institutionalization within an ethnofederal structure. R. Goodnow et al. / Electoral Studies 36 (2014) 15e27 21

Russia is composed of 83 subjects or regions that are and Ordeshook (2008) show quite clearly that evidence of divided into 21 republics, 46 oblasts, nine krais, two cities electoral manipulation was initially located in large part in (Moscow and St. Petersburg), four autonomous okrugs, and ethnic republics. For example, in the 1996 presidential one autonomous oblast (the Jewish Autonomous Oblast). election they show that suspicious voting patterns Republics, autonomous okrugs, and the Jewish Autono- (implausible shifts of votes in the second round of the mous Oblast make up the 26 regions named after a non- election that suggest vast majorities who supported the Russian ethnic group. Overlaying these federal subjects Communist challenger, Zyuganov, in the first round are seven federal okrugs, which were part of Putin's pro- switched to support President Yeltsin in the second round) gram of power verticalization. were concentrated in ethnic republics. These patterns were As Harris (1993: 571) has demonstrated, Russia's ethnic relatively rare overall compared to later elections in which federalism has had a strong influence on the ethnic identity electoral manipulation was believed to be more rampant of non-Russians. Federal ethnic groups provided with but were almost exclusively located in raions in ethnic re- “ethnic homelands” managed to retain their native lan- publics, namely three e Bashkortostan, Dagestan, and guages and resist assimilation efforts to a much greater Tatarstan. Only 12% of raions suspected of fixed elections in extent than (often larger) non-federal minority groups. 1996 were located in Russian oblasts (Myagkov and Moreover, regional governments in ethnic homelands have Ordeshook, 2008:4e5). This discrepancy in the incidence cultivated ethnic mobilization in centereperiphery re- of manipulation between republics and oblasts extends lations. The very designation of ethnic federal units has well into the Putin years. Myagkov and Ordeshook's (2008) produced incentives and resources for the promotion of an examination of the distribution of voter turnout from 1996 ethnic cadre of elites and mass mobilization tied to through 2003 clearly shows that this particular proxy for ethnicity (see Roeder et al., 1991; Treisman, 1997; electoral manipulation (excessively high turnout) was Gorenburg, 2003; Hale, 2003). found almost exclusively in ethnic republics while Russian In many cases, the titular nationality does not constitute oblasts had normal patterns of voter turnout (see also a majority within its designated ethnic region and com- Mebane and Kalinin, 2009). It was only beginning in the prises a majority in relatively few (former) single-member presidential election of 2004 that widespread evidence of electoral districts. Despite these demographic obstacles, manipulation appeared in Russian regions and even then these groups managed to attain significant representation such behavior was much more prevalent in ethnic republics in the national legislature, which suggests that the sym- (Myagkov and Ordeshook, 2008; Mebane and Kalinin, bolic power of an ethnic homeland is an important factor in 2009). the representation of minorities in the State Duma (Moser, In this paper, we utilize the ethnic and electoral varia- 2008). Moreover, raion-level analysis of voting patterns in tion within regions to more systematically examine the Russian SMD elections in 1995 shows a significant level of relationship between ethnicity and electoral fraud in Rus- “ethnic voting” driven by the demographic concentration sia's two most recent parliamentary elections in 2007 and of non-Russian minorities in single-member districts. 2011, when electoral manipulation was deemed to have Goodnow and Moser (2012) have shown that majority- spread beyond the ethnic republics. We show that even in minority districts had a significant impact on the vote these elections the incidence of electoral manipulation in share of minority candidates who won election as well as raions with large concentrations of ethnic minorities minority voter turnout. Minority winners tended to gain greatly outstripped that of majority-Russian raions, even larger vote shares in electoral districts with a majority non- when controlling for other factors such as degree of ur- Russian population and majority-minority districts tended banization, federal status (ethnic vs. Russian region), and to increase the size of the effect of raion-level minority socioeconomic development. concentration. In other words, minority voters tended to support minority candidates in larger numbers when sit- uated in majority-minority districts. Majority-minority 5. Empirical analysis districts tended to have a similar amplifying effect on mi- nority voter turnout (Goodnow and Moser, 2012). 5.1. Data: Russia as a test case of the impact of ethnicity on In short, ethnic federalism has worked in ways that electoral manipulation scholars recommending decentralization in ethnically diverse countries would expect. Ethnic-based decentral- The analyses that follow are based on electoral data ization aided the representation and empowerment of from the Central Election Commission of Russia, ethnicity non-Russian elites, especially in the 1990s when regions data from the 2002 All-Russian National Census, and so- enjoyed greater autonomy, by facilitating demographic cioeconomic data from Rosstat, the Russian State Statistics 4 concentration and providing material and symbolic re- Service. The election and census data are reported at the sources that encourage ethnic voting and other forms of level of each raion in the country, which gives us approxi- ethnic political mobilization. mately 2700 observations for each election year on which fi However, as the works of Myagkov, Ordeshook, and to base our ndings. In addition to the ethnicity data, the Shakhin (2008, 2009) and others have powerfully demon- strated, the ethnic-based regions that lie at the heart of 4 Census data on the ethnic composition of Russian raions at the raion- Russia's ethnic federalism have also produced pockets of level were collected by the author during fieldwork in Russia in 2003. electoral manipulation as well as other manifestations of Electoral results for parliamentary elections from 1995 through 2011 are authoritarian politics (McFaul and Petrov, 1997). Myagkov now available at the Russian CEC website. 22 R. Goodnow et al. / Electoral Studies 36 (2014) 15e27

Rosstat website reports the percentage of the rural popu- “special efforts” to maintain control and also may possess lation in each raion, which, in combination with informa- resources embedded in ethnic mobilization that promote tion about each region's Gross Regional Product (GRP) and the use of patronage and electoral machine politics (Hale, status as a resource-rich region, provides a measure of so- 2003). cioeconomic welfare. Similarly, we expect five other broader contextual Ethnic regions (republics and autonomous okrugs) have characteristics of raions e urban versus rural raions and a preponderance of raions with a high concentration of location in: Muslim regions, the Caucasus region, and ethnic minorities, while majority-Russian regions (oblasts resource-rich regions e to affect the likelihood that a given and krais) are predominantly made up of majority-Russian raion will experience electoral manipulation. Following raions. However, because ethnic regions often contain a Myagkov and Ordeshook (2008, 2009), we expect urban large proportion of Russians, there is a large enough raions to be less susceptible to electoral manipulation than number of raions with substantial Russian populations to rural raions (Rural) because the former provides fewer allow for meaningful comparisons between pockets of opportunities for elite intervention into the electoral pro- Russian and minority voters within different electoral en- cess than the latter. We also expect raions located in vironments. Indeed, nearly 34% of raions in non-Russian Muslim ethnic regions (Muslim) to be more manipulated regions in our sample have a majority-Russian popula- than raions in non-Muslim or Russian regions since in- tion. Pockets of minority concentration within Russian re- centives for national authorities to control Muslim regions gions are less common, with approximately 3% of raions in may be higher given the experience with the Chechen War Russian regions comprised of a majority-minority popula- and broader violence in the Caucasus as well as the prev- tion. This variation in the concentration of minority pop- alence of drives for autonomy during the 1990s, which ulations within and across different regional contexts gives were strongest among Muslim ethnic republics (Treisman, us considerable analytical leverage for isolating the po- 1997).5 For similar reasons (greater incentives and re- tential causes of electoral manipulation. We can use the sources to manipulate elections), we expect raions located variation within federal regions to examine not only in the Caucasus (Caucasus) and resource-rich regions whether there are differences in electoral patterns between (Resource) to be more likely to experience manipulated Russian and non-Russian regions but also whether there elections.6 We also control for economic development at are significant differences between raions with minority- the regional level, expecting that regions with higher levels dominant populations and Russian-dominant populations of prosperity will have higher turnout than regions with within regions. This should allow a finer grained analysis of lower levels of prosperity. whether electoral manipulation in regions with large mi- Fig. 3a and b and Table 1 provide a descriptive snapshot nority populations was a general phenomenon found in of the occurrence of electoral manipulation in Russian Russian and non-Russian raions alike or an outcome that raions possessing the characteristics we outlined above. was most common in minority-dominated raions. If it is the These descriptive statistics showcase that both regional- latter, this would indicate more clearly than region-level and raion-level factors seem to be associated with the analyses that there may be an ethnic component to elec- presence of suspiciously high voter turnout (our proxy for toral manipulation in Russia. electoral manipulation). Raions with high voter turnout were found much more often in non-Russian raions than 5.2. Independent variables: factors explaining the variation in Russian raions. More importantly, while electoral manip- electoral manipulation ulation in general also appeared to be more prevalent in those regions where we would expect (non-Russian re- As noted above, previous scholarship, especially by publics, the Caucasus, Muslim regions, resource-rich areas), Myagkov et al. (2009), has already firmly established that even in these contexts the gap between non-Russian and electoral manipulation was much more prevalent in ethnic Russian raions remained intact. minority regions than in Russian ones. Our primary intent in this paper is to study variation within ethnic regions and 5.3. Analytical approach we expect several factors to influence the degree of elec- toral manipulation found within ethnic republics. If, as we To explore these potential causes of electoral manipu- argue above, electoral manipulation is driven by incentives lation, we specify a series of logistic multilevel models that to control politics and the capacity to coordinate such ac- consider the effects of the varying proportions of ethnic tivities, we expect that raions with characteristics that in- and rural populations in raions on the probability that crease one or both of these dynamics will experience a electoral manipulation will occur (i.e., that turnout will be higher incidence of electoral manipulation. We argue that greater than 75%), in addition to the contextual influence of ethnicity may play a vital role in this process and thus particular regional characteristics on these variables' electoral manipulation will be more prevalent in local en- vironments (raions) with high concentrations of ethnic minorities. 5 Since religion and ethnicity are closely associated in Russia, we use Thus, the variable of greatest interest is ethnicity, more ethnicity as a proxy for religious tradition. Certain ethnic groups are specifically the ethnic composition of raions (% non- coded as Muslim (e.g., Tatars, Bashkirs, Ingush) while the Russian ma- jority and other non-Russian minorities are coded as non-Muslim (e.g., Russian) located within different types of federal units. Buryat, Kalmyk, Mordva). Raions with concentrations of ethnic minorities may be 6 Data on the percentage of regional economies based on the export of perceived as a potential threat to the state and thus require natural resources come from Bradshaw (2006). R. Goodnow et al. / Electoral Studies 36 (2014) 15e27 23

intercepts and slopes. The regions vary according to whether they are ethnic republics or autonomous okrugs (compared to oblasts or krais), majority-Muslim, located in the Caucacus, and on the basis of their GRP's and natural resource wealth. An advantage of the multilevel approach, compared to standard logistic regression, is that the dif- ferences in intercepts and slopes from the regional control and interaction terms can vary randomly across regions. The equation for the first region type e ethnic versus non-ethnic regions e is   p ij ¼b þ b þ b þ b log 0 1minorityij 2ruralij 3ethnicj 1 pij þ b * þ b * þ 4ethnic minorityij 5ethnic ruralij uj

where, pij is the probability that the dependent variable is equal to one, i.e., an instance of potential electoral manipulation, for raion i in region j. b0 is the constant term and equal to the log-odds that manipulation will occur when the variables are equal to zero. b1 and b2 are the region-specific effects on the log-odds of manipula- tion occurring from a 1% increase in raions' minority and rural populations, respectively. b3 is the direct effect from being in an ethnic region on the log-odds that manipulation will occur, and b4 and b5 are the moderating effects on the slopes of the two respective raion-level variables from being in an ethnic region. When expo- Fig. 3. (a and b) Ethnicity and electoral fraud in 2007 and 2011. nentiated, the coefficient estimates are odds ratios, comparing the odds of a raion being manipulated to raions in the same region for each unit change in the respective independent variables. The random

Table 1 Electoral manipulation and different electoral environments in 2007 and 2011.

2007 2011

Turnout Turnout Turnout Turnout >75% <75% >75% <75%

All raions 700 (26%) 2010 (74%) 553 (20%) 2160 (80%) Majority minority raions 326 (73%) 118 (27%) 286 (65%) 157 (35%) Majority Russian raions 374 (17%) 1892 (83%) 267 (12%) 2003 (88%) Rural raions 518 (36%) 904 (64%) 401 (28%) 1019 (72%) Urban raions 182 (14%) 1106 (86%) 152 (12%) 1141 (88%)

Raions in: Ethnic federal regions 378 (65%) 203 (35%) 333 (58%) 238 (42%) Maj. minority raions 309 (79%) 80 (21%) 275 (72%) 108 (28%) Maj. Russian raions 69 (36%) 123 (64%) 58 (31%) 130 (69%) Russian federal regions 322 (15%) 1807 (85%) 220 (10%) 1922 (90%) Maj. minority raions 17 (31%) 38 (69%) 11 (18%) 49 (82%) Maj. Russian raions 305 (15%) 1769 (85%) 209 (10%) 1873 (90%) Muslim regions 234 (94%) 14 (6%) 220 (88%) 29 (12%) Maj. minority raions 206 (96%) 9 (4%) 199 (92%) 17 (8%) Maj. Russian raions 28 (85%) 5 (15%) 21 (64%) 12 (36%) Caucasus regions 165 (58%) 119 (42%) 138 (48%) 150 (52%) Maj. minority raions 99 (89%) 12 (11%) 108 (96%) 4 (4%) Maj. Russian raions 20 (12%) 153 (88%) 30 (17%) 146 (83%) Natural resource regions 193 (45%) 237 (55%) 167 (39%) 264 (61%) Maj. minority raions 131 (75%) 44 (25%) 115 (66%) 60 (34%) Maj. Russian raions 62 (24%) 193 (76%) 52 (20%) 204 (80%) Poorer regions 335 (27%) 905 (73%) 260 (21%) 981 (79%) Maj. minority raions 174 (72%) 68 (28%) 167 (69%) 76 (31%) Maj. Russian raions 161 (16%) 837 (84%) 93 (9%) 905 (91%) 24 R. Goodnow et al. / Electoral Studies 36 (2014) 15e27 component of the model, or level-two residual uj,reflects every percent increase in each of these variables, the odds the effect on the log-odds that manipulation will occur of a raion being manipulated increased by 1e3%, depending from raions located in region j. on the model. We specify the above model for both the 2007 and 2011 At the regional level, the variables accounting for parliamentary elections, in addition to four additional particular contexts are especially revealing. Surprisingly, models for each election year that account for the direct the relative financial and resource wealth of regions had no and interactive effects of the remaining region types. Tables connection to the occurrence of manipulation either inde- 2 and 3 contain the results. pendently or by moderating the effects of ethnic raions (models 4 and 5). The only effect concerned rural raions in 5.4. Discussion of the results 2011, which were at slightly higher risk of being manipu- lated in resource-rich regions than similar raions in non- fi With just one exception, the variables accounting for resource regions. In contrast, the regions classi ed as the percentage of ethnic minority and rural populations at ethnic, Muslim, or Caucasus had increased odds of the raion-level had positive and statistically significant ef- manipulation compared to other regions, with interesting fects on the chance of electoral manipulation occurring. For variations in the respective intensities of their effects.

Table 2 Multilevel logistic analysis of electoral fraud in the 2007 Russian parliamentary elections.

Model 1 Model 2 Model 3 Model 4 Model 5

Intercept 3.87 (0.353)*** 3.76 (0.303)*** 3.78 (0.336)*** 3.38 (0.369)*** 3.54 (0.345)***

Raion-level: Percent minority 0.0148 (0.006)*** 0.0256 (0.005)*** 0.0271 (0.005)*** 0.0335 (0.006)*** 0.0310 (0.005)*** OR 1.0150 OR 1.0260 OR 1.0275 OR 1.0342 OR 1.0315 Percent rural 0.0183 (0.002)*** 0.0191 (0.002)*** 0.0208 (0.002)*** 0.0179 (0.003)*** 0.0170 (0.002)*** OR 1.0186 OR 1.0194 OR 1.0211 OR 1.0182 OR 1.0172

Region-level intercept effects: Ethnic region 1.3612 (0.708)* OR 3.9011 Muslim region 4.566 (1.42)*** OR 96.2445 Caucasus region 2.466 (1.063)** OR 11.7829 Economic development -0.0798 (0.152) OR 0.9233 Natural resource wealth 0.4232 (0.939) OR 1.5270

Regionelevel slope effects: Percent minority *Ethnic region 0.0208 (0.010)** OR 1.0211 *Muslim region 0.0026 (0.017) OR 1.0027 *Caucasus region 0.0098 (0.015) OR 1.0099 *Economic development -0.0033 (0.004) OR 0.9966 *Natural resource wealth -0.0072 (0.012) OR 0.9928 Percent rural *Ethnic region -0.0010 (0.005) OR 0.9989 *Muslim region -0.0107 (0.008) OR 0.9893 *Caucasus region -0.020 (0.006)*** OR 0.9801 *Economic development -0.0003 (0.002) OR 0.9997 *Natural resource wealth 0.0070 (0.006) OR 1.0071

Between region variance: LR statistic 650.13*** 586.91*** 725.67*** 673.99*** 710.87***

Model fit: Wald statistic (df) 154.19 (5)*** 164.51 (5)*** 152.03 (5)*** 144.09 (5)*** 141.99 (5)***

Notes: The reported coefficients were estimated in STATA using adaptive Gaussian quadrature (AGQ); the dependent variable is 1 if turnout is greater than 75% in the raion, and zero otherwise; level-1 n ¼ 2710, level-2 n ¼ 84; ***p < 0.01, **p < 0.05, *p < 0.1.1. R. Goodnow et al. / Electoral Studies 36 (2014) 15e27 25

Table 3 Multilevel logistic analysis of electoral fraud in the 2011 Russian parliamentary elections.

Model 1 Model 2 Model 3 Model 4 Model 5

Intercept 4.66 (0.467)*** 4.17 (0.391)*** 4.49 (0.419)*** 4.00 (0.457)*** 3.97 (0.448)*** Raion-level: Percent minority 0.0103 (0.008) 0.0172 (0.005)*** 0.0178 (0.005)*** 0.0274 (0.007)*** 0.0242 (0.006)*** OR 1.0104 OR 1.0174 OR 1.0180 OR 1.0278 OR 1.0245 Percent rural 0.0144 (0.003)*** 0.0141 (0.002)*** 0.0197 (0.003)*** 0.0139 (0.004)*** 0.0133 (0.003)*** OR 1.0145 OR 1.0142 OR 1.0199 OR 1.0141 OR 1.0135

Regionelevel intercept effects: Ethnic region 2.647 (0.853)*** OR 14.1145 Muslim region 4.322 (1.532)*** OR 75.3893 Caucasus region 3.737 (1.393)*** OR 41.9910 Economic development -0.0007 (0.154) OR 0.9992 Natural resource wealth -0.0181 (1.100) OR 0.9820

Regionelevel slope effects: Percent minority *Ethnic region 0.0087 (0.011) OR 1.0088 *Muslim region 0.0074 (0.016) OR 1.0074 *Caucasus region 0.0173 (0.016) OR 1.0175 *Economic Development -0.0038 (0.005) OR 0.9961 *Natural resource wealth -0.0090 (0.012) OR 0.9910 Percent rural *Ethnic region 0.0074 (0.005) OR 1.0075 *Muslim region 0.01852 (0.008)** OR 1.0187 *Caucasus region -0.019 (0.006)*** OR 0.9811 *Economic development 0.0015 (0.002) OR 1.0015 *Natural resource wealth 0.0120 (0.005)** OR 1.0121

Between region variance: LR statistic 635.04*** 558.72*** 663.01*** 677.25*** 693.61***

Model fit: Wald statistic (df) 109.41 (5)*** 109.47 (5)*** 109.97 (5)*** 93.93 (5)*** 95.11 (5)***

Notes: the reported coefficients were estimated in STATA using adaptive Gaussian quadrature (AGQ); the dependent variable is 1 if turnout is greater than 75% in the raion, and zero otherwise; level-1 n ¼ 2712, level-2 n ¼ 83; ***p < 0.01, **p < 0.05, *p < 0.1.

The most striking result was the very large increase in comprised all raions in republics and autonomous okrugs, the odds of electoral manipulation for raions located in were still at greater risk of manipulation than raions in Muslim regions (model 2). These odds were as high as 96 in oblasts and krais, but to a much lesser extent than in the 2007 and 75 in 2011. The more rural raions in these regions narrower Muslim and Caucasus categories. The odds of also had a higher chance of manipulation in 2011 than electoral manipulation in ethnic regions were four times as similar raions in other regions. For raions in Caucasus re- large as non-ethnic regions in 2007 and 14 times as large in gions (model 3), the odds of manipulation were consider- 2011. Also of interest, the effect from the percentage of ably higher than the odds of manipulation in non-Caucasus minorities in raions on the odds that manipulation will regions, at 12 in 2007 and 42 in 2011, but these were much occur was intensified in the context of ethnic regions dur- lower than the odds of manipulation in Muslim compared ing the 2007 election. to non-Muslim regions. An interesting result for the Cau- casus regions is the diminished influence of rural raions on 6. Conclusion the odds of manipulation occurring within this area compared to the rest of the country. Lastly, raions in the In this paper we examined the relationship between broader category for ethnic regions (model 1), which ethnicity and electoral manipulation by asking whether 26 R. Goodnow et al. / Electoral Studies 36 (2014) 15e27 discernible differences exist in the incidence of suspicious manipulation still persists in those areas with the greatest electoral results between raions with different concen- incentives to convey control and the resources to coordi- trations of Russian and non-Russian populations. We also nate manipulation. Moreover, different types of elections, considered the impact of various social, political, and such as presidential elections, may also affect electoral economic factors that could affect the occurrence of manipulation by raising the stakes and thus the incentives manipulation independently and in combination with to demonstrate political control. Further analyses of more ethnicity. Our initial findings suggest that ethnicity, as Russian elections (both parliamentary and presidential) are experienced at the local (raion-level), has a prevailing ef- needed to see if these relationships bear out. fect on the occurrence of electoral manipulation. In a va- riety of contexts, raions with high concentrations of non- References Russian minorities tended to experience voting patterns Beissinger, M., 2002. Nationalist Mobilization and the Collapse of the indicative of electoral manipulation more often than Soviet State. Cambridge University Press, Cambridge. Russian raions. One implication of this finding is that Bradshaw, M., 2006. Observations on the geographical dimensions of electoral manipulation in Russia seems to be a more Russia’s resource abundance. Eurasian Geogr. Econ. 47 (6), 724e746. Carothers, T., 1997. The observers observed. J. Democr. 8 (3), 17e31. decentralized process than often depicted by region-based Chandra, K., 2004. Why Ethnic Parties Succeed. Cambridge University analyses. If regional leaders were delivering the vote on Press, Cambridge. the orders of national elites in a centralized fashion, one Cornwell, E., 1964. Bosses, machines, and ethnic groups. Ann. Am. Acad. fi Polit. Soc. Sci. 353, 27e39. might expect more uniform evidence of xing election Dahl, R., 1961. Who Governs? Yale University Press, New Haven. results within regions rather than the patchwork pattern Fish, M.S., 2005. Democracy Derailed in Russia. Cambridge University of high voter turnout in raions with greater concentrations Press, Cambridge. of ethnic minorities or more rural populations. We cannot Gorenburg, D., 2003. Minority Ethnic Mobilization in the Russian Feder- ation. Cambridge University Press, Cambridge. discern from this study the precise mechanisms of how Goodnow, R., Moser, R., 2012. Layers of ethnicity: the effects of ethnic electoral manipulation works in Russia but it hints that federalism, majorityeminority districts, and minority concentration local elites and populations are a central component of the on the electoral success of ethnic minorities in Russia. Comp. Polit. Stud. 45 (2), 238e264. process. Further research is clearly required to investigate Hale, H., 2003. Explaining machine politics in Russia’s regions: economy, these processes. ethnicity, and legacy. Post-Soviet Aff. 19 (3), 228e263. The much higher levels of manipulated raions within Hale, H., 2007. Correlates of clientelism: political economy, politicized ethnicity, and postcommunist transition. In: Kitschelt, H., Muslim versus non-Muslim regions suggest that the type of Wilkinson, S. (Eds.), Patrons, Clients, and Policies: Patterns of Dem- ethnic cleavages in Russia may play a role in this dynamic ocratic Accountability and Political Competition. Cambridge Univer- as well. This is most likely due to historical experience with sity Press, Cambridge, UK, pp. 227e250. Harris, C., 1993. The new Russian minorities: a statistical overview. Post- mobilization against the state, which increases the Soviet Geogr. 34. perceived threat posed by these regions, raising incentives Horowitz, D., 1985. Ethnic Groups in Conflict. University of California to convey political control. One possible extension of this Press, Berkeley. Horowitz, D., 2003. Electoral Systems: A Primer for Decision Makers. J. research would be to further investigate variance of elec- Democr. 14 (4), 115e127. toral manipulation among raions with concentrations of Inglot, T., Pelissero, J., 1993. Ethnic political power in a machine city: different types of non-Russian ethnic groups.7 Chicago's poles at Rainbow's end. Urban Aff. Q. 28, 526e543. ' Another avenue for further research is investigation of King, R., 2001. Counting the votes: South Carolina s stolen election of 1876. J. Interdiscip. Hist. 32 (1), 169e191. change over time. Comparison of our cross-sectional data Lehoucq, F., 2003. Electoral fraud: causes, types and consequences. Annu. suggests interesting similarities and differences between Rev. Polit. Sci. 6, 233e256. the elections of 2007 and 2011. Ethnicity appears to have Lijphart, A., 1977. Democracy in Plural Societies: A Comparative Explo- fi ration. Yale University Press, New Haven. had a signi cant impact in both elections but the gap be- Lijphart, A., 2004. Constitutional design for divided societies. J. Democr. 15 tween Russian and non-Russian raions is greater in 2011 (2), 96e109. (see Fig. 3a and b), which is surprising given that the 2011 Lukinova, E., Myagkov, M., Ordeshook, P., 2011. Ukraine 2010: were Tymoshenko’s cries of fraud anything more than smoke? Post-Soviet election actually witnessed a substantial decline in the vote Aff. 27 (1), 37e63. share for United Russia as it was under greater pressure to Mansfield, E., Snyder, J., 2005. Electing to Fight: Why Emerging. De- avoid the egregious manipulation of the previous election. mocracies Go to War. MIT Press, Cambridge. Marsh, C., Warhola, J., 2001. Ethnicity, ethno-territoriality, and the polit- Arguably, as electoral manipulation spreads more widely it ical geography of Putin's electoral support. Post-Soviet Geogr. Econ. tends to encroach on regions without concentrations of 42 (No. 4), 1e14. ethnic minorities and, conversely, when it recedes McFaul, M., Petrov, N., 1997. Russian electoral politics after transition: regional and national assessments. Post-Soviet Geogr. Econ. 38 (9), 507e549. Mebane, W., Kalinin, K., 2009. Comparative election fraud detection. In: Prepared for Presentation at the Annual Meeting of the American e 7 For example, we did a preliminary examination of Bashkortostan, Political Science Association, Toronto, Canada, Sept 3 6, 2009. which is one of the more ethnically diverse regions in Russia with large Moser, R., 2008. Electoral systems and the representation of ethnic mi- e concentrations of the titular group, Bashkirs, but also significant numbers norities: evidence from Russia. Comp. Polit. 40 (3), 273 292. Myagkov, M., Ordeshook, P.C., February 1, 2008. Russian Elections: An of Russians and other non-Russian minorities, most notably, Tatars. It was Oxymoron of Democracy. Available at: SSRN. http://ssrn.com/ found that, while the largest gap in the incidence of manipulation was abstract¼1090975. found between Russian versus non-Russian raions, there was also a Myagkov, M., Ordeshook, P., Shakin, D., 2009. The Forensics of Electoral fi signi cantly larger number of manipulated raions among the majority- Fraud: Russia and Ukraine. Cambridge University Press, New York. Bashkir regions as opposed to raions with simply a majority-minority Powell, L., 1989. Correcting for fraud: a quantitative reassessment of the population comprised of other minority groups as well. This suggests Mississippi ratification election of 1868. J. South. Hist. 6, 633e658. that titular minorities may provide even greater capacity to manipulate Reilly, B., 2002. Electoral systems for divided societies. J. Democr. 13 (1), elections than other minorities. 156e170. R. Goodnow et al. / Electoral Studies 36 (2014) 15e27 27

Roeder, P., 1991. Soviet federalism and ethnic mobilization. World Polit. Treisman, D., 2012. The Return: Russia’s Journey from Gorbachev to 43, 196e232. Medvedev. The Free Press, New York. Saideman, S., Lanoue, D., Campenni, M., Stanton, S., 2002. Democratiza- Treisman, D., January 1997. Russia's ethnic ‘revival’- the separatist activism tion, political institutions, and ethnic conflict: a pooled, cross- of regional leaders in a postcommunist order. World Polit. 49, 212e249. sectional time series analysis from 1985e1998. Comp. Polit. Stud. 35 White, S., 2011. Elections Russian-style. Europe-Asia Stud. 63 (4), (1), 103e129. 531e566. Simpser, A., 2013. More than Winning: Why Governments and Parties Wilkinson, S., 2004. Votes and Violence: Electoral Competition and Ethnic Manipulate Elections. Cambridge University Press, Cambridge. Riots in India. Cambridge University Press, Cambridge. Snyder, J., 2000. From Voting to Violence: Democratization and Nation- Wolfinger, R., 1972. Why political machines have not withered away and alist Conflict. Norton Books, New York. other revisionist thoughts. J. Polit. 34, 365e398. Stoner-Weiss, K., 2001. The Russian central state in crisis: center and Ziblatt, D., 2009. Shaping democratic practice and the causes of electoral periphery in the post-soviet era. In: Barany, Z., Moser, R. (Eds.), fraud: the case of nineteenth-century Germany. Am. Polit. Sci. Rev. Russian Politics: Challenges of Democratization. Cambridge Univer- 103 (1), 1e21. sity Press, New York.