Danyliw Seminar 2014 | Timothy Colton

ARCHIVES Danyliw Seminar 2014

Timothy Colton

Harvard U, US, tcolton@fas.harvard.edu

Presentation

Interview

PANEL

The 2014 Elections

PAPER

The 2014 Presidential Election

Ukraine held its sixth presidential election as an independent country on May 25, 2014. The business magnate and politico Petro Poroshenko took 54.7 percent of the popular vote and was thus declared the country’s fifth president after only one round, the first time this had happened since Leonid Kravchuk was elected in a single round in 1991. The runner-up, former prime minister Yulia Tymoshenko, lagged far in the rear, at 12.8 percent of the vote, barely a quarter of her vote share in the runoff round in 2010 against Viktor Yanukovych. Oleh Lyashko, the leader of the populist Radical Party, came in third with 8.3 percent.

Together with three Western-based scholars—Henry Hale, Nadiya Kravets, and Olga Onuch—and a number of Ukrainian colleagues, I have been involved in a survey-based study of the popular component of Ukraine’s year of turmoil. Our local partner in the project has been the Kyiv International Institute of Sociology, where we have benefited enormously from the hard work, expertise, and insights of Volodymyr Paniotto and Natalya Kharchenko, among others. The poll we have conducted is a panel survey, that is, a survey in which a sample of the population is interviewed and re-interviewed (subject to “panel mortality”), in the interests of efficiency in data collection and tracking individual-level shifts in perceptions and behavior. In wave 1, conducted in the weeks before May 25, 2,015 voting-age citizens were interviewed; of them, 1,406 were re-interviewed after the election. We will shortly go into the field In Ukraine with a wave 3, subsequent to last weekend’s election of a new Verkhovna Rada.

Orientation Points

As any reader of this brief will be aware, the 2014 presidential election was held in anything but normal circumstances. In a head-spinning sequence, the Yevromaidan protest movement challenged the country’s president and then overthrew him; Ukraine lost control of the Crimean peninsula to pro-Russian forces and then to Russia; an insurrection against Kyiv’s authority, supported by Moscow, broke out in the Donbas; and early attempts to suppress the rebellion by force failed. May 25, 2014, was a wartime election, under conditions of both extreme hostility with a powerful external neighbor and armed conflict with internal enemies of the state.

One reassuring and indeed surprising lesson of our survey data is that, despite all the problems, a majority of Ukrainians as of the summer of 2014 were optimistic about the future. Forty-seven percent of our respondents in wave 1, before the election, felt that the country was headed in the right direction, as against 32 percent who thought it was not; in wave 2, after the election, the number who agreed was up to 61 percent and the number who did not was down to 24 percent.

General satisfaction with the state of affairs is correlated with a positive evaluation of the May election. As can be seen in Table 1, about two-thirds of Ukrainians judge the presidential election to have been fairly conducted.

Table 1: Retrospective opinion about fairness of presidential election (%) Evaluation of election on 5-point scale

An overall sense that the process was fair should not be confused with the notion that it was in any way routine. In fact, Ukrainians went to the polls on May 25 without the benefit of clear-cut orienting frames that would be present in a stable democratic polity. In bullet form, this was the situation:

The election was organized on the fly, with the shortest imaginable interval for preparation and positioning.
The campaign occurred in wartime. Headlines telling of violence and mayhem in the east of Ukraine prevailed over peaceful electoral news, for understandable reasons.
The action programs for the future enunciated by the presidential candidates were for the most part extremely vague. They shared a shallow populism and, except for the East-based candidates, an anti-Russian pitch.
Few of the presidential candidates presented much by way of a prior record in high-level politics and government, which might have formed a basis for voter choice about their performance and ability to deliver the goods in office. Yanukovych, the president of Ukraine from 2010 to 2014, was discredited on all sides – by the “Oranges” from the heady days of 2004 for snubbing their concerns, by the “Blues” for fleeing the presidential compound in February 2014 and leaving them in the lurch. His four years in office thus formed almost no basis for a citizen making his or her ballot choice in May 2014.
Recent political history in Ukraine was divisive and generated not only widespread antagonism toward previous reform projects (as with the Orange Revolution of 2004) but incipient disillusionment in at least some quarters with the most recent spasm of change (as with the Yevromaidan Revolution of 2013–14).
Pre-existing political parties—which are in consolidated democracies the main aggregators of interests for electoral competition and beyond—had been shattered by the turbulence of 2013–14, and no organizations of substance had replaced them.

Table 2 below drives home the second-last point. A plurality of Ukrainians consider the Orange Revolution to have done more harm than good. For Yevromaidan, a plurality still favor a positive interpretation, but one citizen in five thinks Yevromaidan had negative consequences on balance and one in three thinks it had mixed effects. What no one knows for now is whether the balance for Yevromaidan will tip toward the generally negative outlook that prevails vis-a-vis the Orange Revolution, ten years after the fact.

Table 2: Evaluation of Orange and Yevromaidan revolutions (%)

Table 3 takes up the last point and related points. It relies on self-assessment by participating voters. Admittedly a less than perfect instrument—which at a minimum provides food for thought about the reasoning of the Ukrainian electorate. It records the decisive factor, from a closed list read out by the interviewer, to which participating Ukrainian voters interviewed in wave 2 of our survey ascribed their choice at the polling booth on May 25.

Table 3: Voters’ self-assessment of the single principal factor in their electoral choice (%)

The most important negative point coming out of this table is about political parties. In our survey, almost no one—1 percent—recalled having voted on the basis of partisan sentiment. This is a stunning consequence of the shattering of the mainstays of Ukraine’s party system as it had existed in equilibrium for the past decade, beginning with the hitherto “ruling” Party of Regions and Tymoshenko’s Batkivshchyna, the parties that backed the two finalists in the 2010 presidential election. To probe what we think of as “transitional partisanship,” we administered to our Ukrainian respondents a battery of questions that have been asked in numerous electoral surveys in Russia and, in this or a related form, in other formerly communist polities in Eastern Europe and Eurasia, going back to the 1990s. Essentially, respondents in the first wave of our Ukrainian election survey were asked: whether they saw a particular political party or movement as “my party”; if so, how intensely their affect for that party was; and for respondents who had answered in the negative to the first query, whether they could indicate another party or quasi-party that more closely ,et their needs. By this test, a mere 15 percent of Ukrainians testified that they had any agree of attachment to a political party.

Speaking positively about what did motivate them on May 25, according to self-analysis, the most potent factor for the electorate as a whole turns out to be the personal qualities of the candidate chosen (26 percent); that is followed by the candidate’s past record in political and nonpolitical roles (17 percent), possible benefits to the respondent’s region (at 15 percent) tied with the candidate’s potential to win in the first round (also 15 percent), and only then the candidate’s electoral “program” (at 14 percent).

There is considerable variation in these figures across the three leading candidates. The third-placed Lyashko’s personal qualities were cited by more than half of his supporters (53 percent), and another 17 percent said they voted for him mainly to send a protest message. The well-traveled Tymoshenko’s profile is distinguished by the high rate of emphasis on her past record (42 percent), followed by her personality at 25 percent. The victor, Poroshenko, has a more balanced profile, in that five motivating factors exceed 10 percent in terms of draw. The list is headed by his personal qualities (24 percent, or a hair below the average for all candidates) and his chances of winning in the first round (20 percent, which dwarfs the 2 percent who saw this attraction in Tymoshenko and Lyashko). Given Poroshenko’s lead in opinion polls throughout the campaign, and the apparent wish of many citizens to spare Ukraine further political division, the latter appeal makes a good deal of sense. What Table 3 suggests is that, without the can-win-in-the-first-round factor, Poroshenko would have fallen well short of 50 percent on May 25. Approximately 10 percent of the popular vote would have been redistributed to other candidates, and Poroshenko would have been forced into a runoff against the runner-up, presumably either Tymoshenko or Lyashko.[1] He in all likelihood would have won the runoff with ease, but his opponent in a second round would have received a large boost in authority and credibility and, one may speculate, would quite possibly be either a strong coalition partner or a kingmaker today—instead of Arseniy Yatsenyuk, whose name was not on the ballot on May 25.

Correlates of the Electoral Choice

Time has not allowed me to conduct and present the systematic modeling of Ukrainians’ electoral choices in May 2014 that would allow us to go beyond the self-analysis reported in Table 3. What I can present here is an impressionistic sketch of some of the points that will go into the analysis that will eventually emerge.

We are of course most interested in the Poroshenko vote. In our sample, 71 percent of respondents who told interviewers they had voted on May 25 reported having voted for Petro Poroshenko. Over-reporting of the vote for the winner is well-night universal in electoral surveys, for a variety of reasons, and Ukraine in 2014 is no exception. When over-counting is combined with the fact that the actual outcome is as lopsided as it was in May 2014, the political scientist who tries to explain events after the fact is presented with an exaggerated majority of self-declared supporters of the victor and very small minorities who say they sided with other candidates—in our sample, only 7 percent with Tymoshenko and 7 percent with Lyashko. Using conventional quantitative estimation techniques, it is inherently very difficult to come to crisp conclusions about electoral support for candidates or political parties that come to considerably less than 10 percent in the overall survey sample drawn.

One general point coming out of our data is that Poroshenko managed to draw on a very broad base of support. It would be difficult to get to 55 percent of the total electorate (not to mention the 71 reported in our sample of the electorate) without such a foundation. In this regard, Poroshenko is a lot like a fairly well-known politician in a neighboring Slavic country who in his first encounter with the electorate some years before captured only slightly less of the popular vote than Poroshenko in 2014. I am referring to Vladimir Putin, who gained victory in the first round of Russia’s 2000 presidential election, also after the unscheduled departure of his predecessor, with 53.4 percent of the popular vote.[2] In both cases, there was a strong alignment around the leader as the campaign period unfolded. This dynamic was linked in each country to a traumatic pre-election experience—Boris Yeltsin’s premature resignation and Yanukovych’s ouster—and to warfare—the second Chechen War in Russia, the conflicts in Crimea (bloodless) and Donbas (bloody) in Ukraine.

Information about the generational distribution of support for the three leading candidates serves to illustrate the point (see Table 4).

Table 4: Presidential vote share for three candidates, by age group (%)

The only one of the three leading candidates to display a clearly differential pattern here is the firebrand Lyashko, who in our imperfect sample reels in 17 percent support among voters aged eighteen to twenty-nine but less than 5 percent among voters over the age of fifty. Poroshenko shows a rather weaker reverse pattern, doing noticeably better on average with citizens older than fifty. The over-representation of reported Poroshenko voters in our survey sample means that the actual amount of support for the victor, age group by age group, must have been considerably lower than indicated here. It is conceivable, for example, that Poroshenko received something like 50 to 55 percent backing among voters younger than thirty, and that Lyashko received something like 20 to 25 percent. Even if this was so, however, Poroshenko would have attracted majority or near-majority support in every age cohort, thus receiving as many votes cast as all other candidates combined. Considering this particular social category, Poroshenko comes across as an electoral omnivore, feasting on robust support in every or almost every subgroup.

This pattern reproduces itself across other social categories and subcategories, as is laid out in Table 5.

Table 5: Presidential vote share for three candidates, by social group other than age group (%)

Multivariate Analysis

To come to a comprehensive understanding of the determinants of electoral choice in Ukraine in 2014, it is necessary to look at background social characteristics, as well as other potential drivers of theoretical and practical interest, in combination, and not one by one. The standard way this is done is through multivariate regression analysis.

A full discussion has to be left to another day. Let me, though, throw out several bullet points that come out of preliminary work, limiting myself to the Poroshenko vote:

The apparent effects of most independent variables on our dependent variable of interest are quite small and are dwarfed by the residual tendency to vote for Poroshenko (captured in a visual plot as the high “intercept” of the regression line).
All things considered, the only social-background factors systematically associated with a higher probability of casting a ballot for Poroshenko are age, education, financial condition (weakly correlated), and location in the East. Poroshenko draws rather better among older, better educated, and more prosperous voters and more poorly than average among citizens in the oblasts of eastern Ukraine.[3]
The last association might be thought to be revealed by aggregate election returns, but in fact the inferences that can be made from aggregate data—to which we have usually had to limit ourselves sin the past—are limited because of aggregation effects and biases. Geographic location may in principle either exercise an autonomous effect on electoral behavior or it may be a proxy for underlying characteristics of the population in the territorial unit—for Ukraine, of course, ethnicity and especially language come to mind. What we see for Poroshenko in 2014 is region trumping both ethnicity and language.

[1] Subtracting 20 percent of the 55 percent of the tallied popular vote which Poroshenko attained in round 1 would have left him at 44 percent of the popular vote on May 25.

[2] On Putin’s support base in 2000, see Timothy J. Colton and Michael McFaul, Popular Choice and Managed Democracy: The Russian Elections of 1999 and 2000 (Washington, D.C.: Brookings Institution Press, 2003).

[3] As Sergiy Kudelia of Baylor University has shown, abstention rates were also considerably higher in the East. Some of this was due to combat conditions but some due to conscious choice on the part of alienated citizens. Our survey uncovered some signs of the latter.