Were All Ukrainians Orange?: The Political Machine and the East-West Division in the 2004 Ukrainian Presidential Elections

In democratic states, scholars often attribute geographic divisions in elections to voter identity cleavages, such as ethnicity. A sizable literature examining the Ukrainian electoral geography in the same scheme has also reached the consensus that Ukrainian voters tend to cast their ballots based on their identities demarcated by a variety of factors, such as language, ethnicity, regionalism, etc. Yet much less scholarly attention has been devoted to the difference between elections in democracy and in autocracy. In autocracy, local elites disturb electoral processes with administrative resources, so ordinary people hardly vote on the basis of their own preferences due to threats posed by local elites. Therefore, unlike in democratic countries, local dynamics, proving coercive and capable of voter mobilization, have a strong influence on electoral results under authoritarian regimes. In such a context, rather than reporting the associations between electoral results and indicators measuring identities, it is necessary to take into account how regional political machines work to investigate the electoral geography. With a careful case study and several quantitative analyses with the data of the 2004 presidential election in Ukraine, this paper attempts to explain cleavages appearing in authoritarian elections with the degree of mobilizational ability of regional elites, instead of voter preferences or their identities. The first section overviews the previous studies concerning three topics: identity, electoral results, and clientelism. Building on the nascent literature focusing on the underlying reciprocal relationship in ethnic voting, this study highlights the importance of the motivations of voters behind their voting behaviors in autocracies. Regional elites in authoritarian regimes can pose pressure on constituencies to extract support for incumbents from them by implying punishments such as cutting administrative services and demotions or dismissals in workplaces. My argument here is that, given such a nature of authoritarian elections, the electoral geography in autocracies is more likely to be determined by the mobilizational capacity of regional elites, rather than voter preferences. To clarify the point, the next section moves to the specific case of Ukraine. A great number of scholars have argued that, in terms of elections, Ukraine features a clear division between the eastern and western regions determined by identities or complicated historical experiences. However, using the district-level data in the 2004 election in Ukraine, this study provides empirical evidence that an identity variable accounts for the inter-regional variation, but not the intra-regional variation. Based on this result, I claim that inferences explaining the electoral geography in Ukraine with identities are not convincing, since they cannot explain what motivations Ukrainian voters have in voting processes or this intra-regional variation. An alternative explanation presented here is that the viability of regional political machines is of importance in understanding these factors overlooked in the previous literature on Ukrainian electoral geography. In the third section, the case study focusing on two south-eastern regions provides factors enabling regional elites to mobilize voters successfully. Employing the most-similar method of case selection, the political processes in Dnipropetrovs’k and Kherson oblasti from 2002 to 2004 are profoundly investigated. Although these two regions share a similar ethnic, linguistic, and historical identity, the electoral results in 2004 elections remarkably differed in both of the regions. In the former region, the candidate from the eastern side of Ukraine (pro-incumbent) beat the opponent (anti-incumbent), who was supported in the central-western regions, by a wide margin. In the latter region, however, these candidates were in close contest, in spite of the fact that this region’s identities are similar to those in other eastern regions including Dnipropetrovs’k. The careful case study reveals that this dissimilarity of electoral results can be explained by (1) the degree of the elite unity and (2) the degree of autonomy of voters from the control of local elites. The first factor is associated with the mobilizational capacity of regional political machines. Although the regional governors had retained uncontested control over the entire political and economic spheres in Dnipropetrovs’k, regional elites in Kherson always sought their own interests and were never completely controlled by the regional governors. As the political situation in Kherson illuminates, without consolidated political hierarchies in regions, regional political machines would struggle to perform well for incumbent candidates. Meanwhile, the second factor captures how easy it is for voters to deflect the expectations of regional elites. If the daily life of constituencies is independent and autonomous from the influence of regional elites, they can easily escape from intimidation by regional elites and support opposition candidates in elections. While the majority of ordinary people were dependent on the regional administrations in Dnipropetrovs’k, residents in Kherson were relatively free from the influence of regional elites. Consequently, they could blatantly chant slogans against the regional administration, participate in rallies to extract concessions from the regional elites, and, also, support the opposition candidate in the 2004 election. The final section tests the applicability of the argument generated through the case study, i.e., whether these two factors account for the variation in electoral results in the 2004 Ukrainian presidential election. First, with the qualitative evidence offered in the case study, it is feasible to expect that frequent replacements of regional and municipal executive chiefs are intertwined with the erosion of political machines, resulting in poor electoral performance. Exploiting the career patterns of governors and mayors from 2002 to 2004, the quantitative tendency that the replacements of top officials in regions or municipalities are positively correlated with the support for opposition candidates is disclosed. Next, the case study demonstrates that the degree of autonomy of ordinary people is critical for the electoral results in both of the regions. Using the productivity of small businesses as a proxy of the degree of autonomy of ordinary people, the quantitative test shows that the association between the productivity of small businesses and electoral results is statistically significant and a 1-standard deviation increase in the productivity of small businesses predicts an approximate 7-percentage-point increase in the support for Yushchenko. In conclusion, my argument and findings in this article have implications for the further understanding of the relationship between electoral geography and the local dynamics of political mobilization in Ukraine, and, more generally, electoral cleavages under authoritarian regimes