Russian and Ukrainian courts played an important role in shaping the final roster of candidates who contested a single-mandate district seat in the 2002 and 2003 parliamentary elections. At least 134 candidates in Ukraine and 118 candidates in Russia went to court to either defend their registration or challenge an opponent’s registration. Moreover, most of these candidates were not also-rans, but viable contenders for seats. In fact, 18 percent of the single-mandate-district winners in both Russia and Ukraine were involved in electoral registration court cases¹.

In comparative terms, these numbers are massive. Electoral registration disputes are exceedingly rare in consolidated democracies. Apart from Alan Keyes’s lawsuit, which challenges Barack Obama’s eligibility to be the U.S. president on the basis of purported evidence that Obama was born in Kenya, rather than the United States, one would be hard-pressed to find a U.S. politician who has had to defend his right to stand in an election in court. Any type of election-related litigation appears to be much rarer in consolidated democracies than in emerging democracies, such as Russia and Ukraine. U.S. courts at all instances received 341 election-related lawsuits in total during 2004, which was a presidential election year (Hasen, Reference Hasen2005). The total number of electoral disputes in France for 2002 was 143 (Wright, Reference Wright2003). Czech courts adjudicated 12 cases in 2002 and a record 70 cases in 2006 (Williams, Reference Williams2007). By contrast, Russian courts heard 1,739 electoral cases in 1999 (Sudebnyi Departament pri Vekhovnom Sude RF, 2000). Ukrainian courts heard 200 electoral cases in 1994, 1,753 cases in 1998, and 5,896 in 2002 (). In 2004, the year of the Orange Revolution, Ukrainian courts received 134,000 election-related petitions! (Verkhovnii Sud Ukraini, Reference Verkhovnii Sud2004)

Given their substantial involvement in the electoral process, one can hardly overstate the importance of finding out whether the courts were impartial adjudicators in electoral disputes. This question is a central component of the debate about the fairness of the election process and outcome. If the courts were independent in reaching their decisions, then those who tried to get ahead in the horse-race by violating electoral laws and procedures rightfully suffered the consequences and had their registration canceled. Even if the courts did not go after all transgressors, the punishments that they did mete out would create an incentive for other candidates in future campaigns to abide by the law. The end result would be a cleaner campaign and a long-term trend toward fair elections. However, if the courts acted as instruments in the incumbents’ attempt to hold on to power, then the elections were deeply flawed and unfair. Moreover, the long-term consequences of such judicial behavior include the gradual hollowing out of elections of their meaning and legitimacy and the undermining of the rule of law project.

The 2002–3 elections are a useful tool for comparing the level of judicial independence from politicians in the two countries because they are, at the same time, similar to each other and typical, rather than idiosyncratic. First, each country had previously held at least one parliamentary election under the same electoral system (225 SMD seats and 225 seats filled through PR lists), which means that all actors involved in the electoral process were already accustomed to the formal electoral rules. Moreover, in both countries, candidates and the “political technologists” they employed used the same campaign techniques: oligarchic media-orchestrated “black PR,” technical and clone candidates, deregistration, and administrative resources. Finally, although the Ukrainian election was clearly more competitive than the Russian election (which is the main independent variable under investigation), both elections were fairly uneventful, that is, none involved idiosyncratic factors, such as the mass-scale protests of the Ukrainian Orange Revolution (which briefly empowered the Ukrainian Supreme Court) or the violent executive–legislative confrontation that preceded the 1993 Russian parliamentary election (which also threatened the integrity of the Russian judiciary). Therefore, we can be confident that conclusions about the level of judicial independence that stem from the comparison of these particular two elections are likely to reflect the effects of political competition rather than ad hoc circumstances.

In this chapter, I argue that both countries’ courts failed to meaningfully constrain the incumbent regime. Both the 2002 Rada and the 2003 Duma electoral campaigns showcased widespread use of administrative resourcesby the incumbent Kuchma and Putin regimes. “Administrative resources” is a term that refers to the benefits that come with holding administrative or elected office. When used by the office/position holder during the campaign, those benefits give him or her an unfair advantage over the other participants in the race. For example, university rectors used large auditoriums to hold rallies or allegedly threatened to evict students from dormitories if the rector’s preferred candidate did not garner the most votes in the dormitories polling station; military personnel coerced subordinates into campaigning or voting for a particular candidate; and enterprise directors provided employees with bonuses or paid leave in exchange for participation in campaign events (for discussions of administrative resources, see, e.g., Fish, 2005 on Russia; Kuzio, 2003 on Ukraine). If the courts were independent from the incumbents, we should expect to observe a lower win-rate in court for progovernment candidates and a higher win-rate for all other candidates. Neither the Ukrainian nor the Russian data show this result, which means that in both countries progovernment candidates bolstered their reelection chances through violations of the electoral laws with impunity. This result is in line with the prediction of the strategic pressure theory that judicial independence will be low in both countries. In addition, and again in line with the strategic pressure theory’s predictions, Ukrainian judicial output in electoral registration disputes displays larger systematic bias toward progovernment candidates than the output of Russian courts. Since the size of the progovernment bias reflects the level of judicial dependence from incumbent politicians, I conclude that Ukrainian courts were more dependent on politicians than Russian courts were.

I arrive at this conclusion through statistical analysis of the candidates’ litigation experience. First, I compare the Russian and Ukrainian legislative framework on the issue and explain what electoral registration disputes are. The next section of the chapter describes the data that I collected. The third section discusses my hypothesis about the relationship between candidates’ political affiliation and judicial independence and the control variables needed to assess it. The fourth section describes the measures of the independent variables, a candidate’s political affiliation, his or her viability at the outset of the campaign, and district competitiveness. The last section presents the statistical analysis of the data and summarizes the predicted probabilities of victory in an electoral registration case for various groups of Russian and Ukrainian plaintiffs.

What are Electoral Registration Disputes?

In any democracy, citizens who wish to run for elected office need to take formal steps in order to appear on the ballot on election day. In Canada, for example, each prospective candidate files a Nomination Paper, as required by Art. 66 of the Canada Elections Act. The form contains the prospective candidate’s personal information, declarations on his or her eligibility to run for office, as well as the names, addresses, and signatures of 50–100 constituents in the district where the candidate intends to run. The Nomination Paper is accompanied by a $1,000 deposit, which is returned to the prospective candidate should Elections Canada, the straightforwardly named administrative agency that oversees elections in Canada, reject the application for nomination or should the candidate die between registration and election day (Elections Canada, 2011).

This administrative process of becoming an officially recognized candidate is usually called electoral registration. Electoral registration disputes are court cases that arise in the course of the electoral registration process. Russian and Ukrainian courts adjudicated three types of electoral registration disputes during the parliamentary election campaigns of 2002–3. The first type involved candidates who submitted documents declaring their intention to run in a single-mandate district, but were denied registration by a District Election Commission (DEC). Many of these individuals filed court appeals asking the court to direct the DEC to register them. The second type of dispute involved candidates who initially became registered by the DEC, but who were later deregistered (i.e., their registration was revoked or canceled) by a DEC. Some of these individuals went to court seeking to overturn the DEC decision and get back on the ballot. The final category of plaintiffs were candidates, ordinary citizens, and election commission officials who appealed to the court to take down a candidate’s registration (or to deregister a candidate) for alleged election law violations.

In all case types, the plaintiffs relied on electoral law legislation that gave them the right to challenge decisions of the election commissions in a court of law. In Russia, the relevant pieces of legislation were the Federal Law on the Election of Deputies of the State Duma of the Federal Assembly of the Russian Federation, the Federal Law on Basic Guarantees of Electoral Rights and the Rights of Citizens of the Russian Federation to Participate in a Referendum, as well as the Civil Procedural Code of the Russian Federation, which governs the mechanism through which citizens file court grievances against the action (or inaction) of state actors. The Ukrainian legislation that regulates the same issues was the Law on the Election of People’s Deputies of Ukraine and the Civil Procedural Code of Ukraine. Both Ukrainian and Russian electoral law stipulated that any DEC decision or action could be reviewed by a district court. Both laws also provided plaintiffs with the option of challenging any perceived failure to act on the part of DEC in court (Law on the Election of People’s Deputies of Ukraine, Art. 29; Law on the Election of Deputies of the State Duma of the Federal Assembly of the Russian Federation, Art. 94 and Art. 95). For example, if a candidate filed with the DEC a complaint that one of his opponents had been violating election law provisions, but the DEC simply ignored the complaint or decided that it was unfounded, the unsuccessful complainant could take his or her case to court.

The Russian and Ukrainian election laws that were in force in 2002–3 were similar to each other and largely consistent with their counterpart legislation in consolidated democracies. According to the Office for Democratic Institutions and Human Rights (ODIHR) subunit of the Organization for Security and Cooperation in Europe (OSCE), the leading international organization that monitors democratic elections around the world, both Russian and Ukrainian electoral legislation in the early 2000s met democratic standards. Specifically, OSCE/ODIHR believes that the Ukrainian electoral law “provides an adequate framework for the conduct of democratic elections” (OSCE/ODIHR, 2002, p. 6). The OSCE’s assessment of the Russian laws is virtually identical – “[the] legal framework is generally consistent with OSCE commitments and other international standards relating to democratic elections” (OSCE/ODIHR, 2004, p. 4).

The OSCE/ODIHR’s positive and similar assessment of Russian and Ukrainian electoral legislation suggests that neither Russian, nor Ukrainian legislation overtly favored progovernment candidates or hurt opposition-affiliated candidates. This is an important point for our analysis of judicial behavior, because the content of the laws on the books can be highly consequential to the types of decisions that courts deliver in the disputes that may arise in the process of electoral registration. Imagine, for example, that one law were drafted in such a way that oppositionists would have to jump through hoops to secure registration, whereas registration for progovernment candidates was facilitated. Or that one law gave progovernment candidates carte blanche to use administrative resources, while the other law stipulated that such practice would lead to the immediate deregistration of the candidate. In such scenarios, differences that we observe between the two countries in the win-rates of different types of plaintiffs are likely to stem as much from the bias built into the law as from any bias at the court. Thus it is important to examine the electoral legislative framework in both countries.

So, first, how did citizens register to run in an SMD in Russia and Ukraine in the early 2000s? The steps and necessary documentation stipulated in the two electoral laws are very similar. Table 4.1 lists the documents that comprise the electoral registration package in each country and illustrates the significant overlap between the two laws.

Table 4.1. Required registration documents for SMD candidates

The comparison reveals two interesting differences in the registration process in Russia and Ukraine, but those differences do not suggest that the electoral laws in any country gave an implicit advantage to progovernment candidates over opposition candidates. Russian SMD candidates could choose whether to register by collecting signatures or by submitting an electoral deposit. Ukrainian legislators implemented OSCE/ODIHR’s advice to remove the signature collection option (OSCE/ODIHR, 2001, p. 7), and candidates had only the second option. Second, Russian candidates were required to disclose only their personal income and property information, whereas Ukrainian candidates had to provide the same information about family members as well. The signature collection option may appear to be democracy enhancing because it allows candidates with limited finances to achieve registration through door-to-door signature collection drives by committed volunteers. Thus, signature collection may contribute to increased levels of political participation and to the establishment and maturing of grass roots political organizations – both democracy-enhancing (or -entrenching) developments. The drawback of the signature collection route to registration is that the authenticity of the collected signatures (usually collected by volunteers and without witnesses) is harder to ascertain than the authenticity of the official documents that make up the rest of the registration application package. Thus, the signature option may tempt political opponents to try to sideline each other by challenging the authenticity of the signatures. In OSCE/ODIHR’s words, the signature system created “very many practical problems of regulating and verifying in a fair manner the collection of signatures in support of a candidate or party” (OSCE/ODIHR, 2001, p. 7). The OSCE/ODIHR concern was well-founded, as indeed, invalid signatures was a leading motivation that Russian DECs cited in denial of registration decisions, and it also figured prominently in the electoral registration disputes that made it to court (Lubarev, Reference Lubarev2006). It is important to stress again that there is no reason to believe that the signature collection option in the Russian law favored progovernment candidates over opposition candidates in the registration process – both types of candidates faced the possibility of having to defend the authenticity of the signatures that their campaign had collected. Hence, we could maybe expect Russian courts to receive more complaints than Ukrainian courts, but we have no reason to expect that they would favor progovernment candidates in those disputes, simply as a result of the content of the electoral law.

After receiving the entire list of documents along with the signatures and/or electoral deposit, the district election commission decided whether to officially register the candidate. Both the Russian and the Ukrainian election laws ensured that progovernment and opposition parties got DEC representation (OSCE/ODIHR, 2002, p. 6; OSCE/ODIHR, 2004, p. 4). This guarantee suggests that DECs in both countries behaved similarly during the registration process. For example, we have no reason to believe that Russian DECs were more likely than Ukrainian DECs (or vice versa) to deny registration to opposition candidates on shaky grounds, which would in turn result in oppositionists in one country going to court with consistently stronger cases, than oppositionists in the other country.

Achieving registration did not mean that a candidate was guaranteed a spot on the ballot on election day. Candidates in both countries could have their registration canceled for electoral law violations committed in the course of the campaign. The grounds for cancellation identified in Russia’s Federal Law on Basic Guarantees of Electoral Rights and the Rights of Citizens of the Russian Federation to Participate in a Referendum and in Ukraine’s Law on the Election of People’s Deputies of Ukraine were virtually identical.

First, candidates could lose their registration upon the discovery of vital inconsistencies in the information that they submitted as part of their electoral registration application (Art. 76 of the Russian law; Art. 49, par 8 of the Ukrainian law). The implication of these provisions was that the veracity of all components of a candidate’s biography and his/her property and income statements could be challenged by opponents who sought to take the candidate out of the election race. In both countries, these provisions became the grounds for numerous lawsuits, some containing rather comical allegations. For example, in St. Petersburg’s SMD#207, a similar lawsuit led to the deregistration of the main district’s front-runner, incumbent municipal council deputy, Alexander Morozov. His opponent, Irina Rodnina, filed a complaint alleging that Morozov lied about holding higher degrees from two universities, when, in fact, he held only one. The court sided with Rodnina and deregistered Morozov, citing the discovery of “vital inconsistencies” in his registration application. Curiously, even after successfully taking her main opponent out of the race, Irina Rodnina failed to secure the Duma seat – on election day, the largest share of votes in SMD#207 went in the “against all” category (Novaya Gazeta, December 1, 2003; www.cikrf.ru). In Ukraine, the State Tax Authorities discovered inconsistencies in the income declarations of over 600 candidates (both SMD and party list) (Kyiv Post, March 14, 2002). Dozens of SMD candidates were deregistered as a result and had to fight to get back on the ballot.

Second, both the Russian and the Ukrainian law contained explicit and detailed formal measures designed to minimize the use of administrative resources and the outright bribery of candidates. Elected incumbents and administrative leaders in both countries could not recruit campaign volunteers from among their subordinates or employees; could not use the premises of state institutions and/or the organizations they managed to hold rallies, unless they provided the exact same premises under the same conditions to their political competitors; could not use their office phones, faxes, and other office resources in their campaign; could not use free transportation or preferential rates available to them to campaign; could not use their the access to state media granted to their office to spread their message to voters or to organize the collection of signatures (in the Russian case) (Art. 40 of the Russian law; Art. 49, par. 10 of the Ukrainian law). In addition, candidates could not provide voters with “money, goods, services, securities, credits, lottery tickets or other things or material values for free or at discount prices” (Art. 56 of the Russian law; Art. 49, par. 9 of the Ukrainian law). The formal provisions thus allowed courts to curb the use of administrative resources and buying of votes by government-affiliated candidates. Finally, both the Russian and the Ukrainian law stipulated that candidates whose campaign spending exceeded the legally provided maximum would also lose their registration (Art. 76 of the Russian law; Art. 49, par. 12).

There was also no shortage of lawsuits filed using these provisions. In Russia, for example, incumbent, oppositionist Duma deputy Viktor Cherepkov was taken to court by a rival (Evgenii Kolupaev) in his Primorskii Krai SMD#52. Kolupaev argued that Cherepkov was abusing his entitlement to free transportation for the purposes of his campaign (Kommersant, no. 220, 2/12/2003). Despite being an outspoken opposition figure, Cherepkov defended himself successfully in court and remained on the ballot. In Ukraine, Evgenii Sukhin, a businessman rumored to have spent over a million dollars on his campaign in provincial Vinnitsya’s SMD#18, had to defend himself against allegations of voter bribery. Sukhin reportedly paid for a new roof for a village church, provided new equipment to a state orthodontal office, serving a rural district, and, at a campaign rally with female voters, gave out terry towels to all attending. Despite his “generosity” on the campaign trail, he lost out to the pro-government candidate (Parkhomchuk, Reference Parkhomchuk2002).

Unfortunately, as the statistical analysis described later in this chapter shows, neither the Russian nor the Ukrainian courts took advantage of the opportunity given to them by the law on the books. Despite ample evidence collected by journalists and documented by plaintiffs that progovernment candidates regularly used administrative resources in their campaign, progovernment candidates, on average, did not lose their electoral registration cases more frequently than other candidates.

In sum, both my analysis and the OSCE/ODIHR reports suggest that Russian and Ukrainian laws were adequate, on paper, in providing electoral rights from the standpoint of democratic standards (OSCE/ODIHR, 2002, p. 6; 2004, p. 4). It is the similarity in the election laws that allows me to draw reliable conclusions about the independence of the courts in the two countries on the basis of a systematic comparison of their decisions, which follows.

The Datasets

To answer the question of whether the Ukrainian and Russian courts applied election law provisions impartially in electoral registration disputes, I started with data sets that included all candidates who vied for a seat in one of the 225 SMDs – 3,084 people in Ukraine and 3,018 in Russia. Information on all candidates, rather than only those who participated in court proceedings, is essential to the process of evaluating court bias. In order to understand whether electoral registration lawsuits were politicized, we need to know who were those not involved in court cases in the first place.

In both Russia and Ukraine, the Central Election Commission (CEC) provides information on the political affiliation, previous election experience, age, gender, and occupation of each candidate. The CEC also lists the final vote tally in each SMD. In Ukraine, the CEC compiled a set of court decisions delivered by district courts during the campaign (Tsentral’na Viborcha Komisiya, 2002). Unfortunately, the CEC electoral registration dispute collection does not provide court decisions from 10 of Ukraine’s 25 regions, which all together contain 90 of the 225 SMDs. Fortunately, the incomplete information should not skew the results of the analysis in any significant way because the excluded districts come from all parts of the country – the industrial, propresidential East, the nationalist West, and the rural, communist- and socialist-leaning South. After this exclusion, my dataset includes information on all 1,953 candidates who set out to run in 135 SMDs and participated in 134 electoral registration disputes².

In Russia, the CEC does not provide systematic information on electoral registration disputes. What it does provide is a media coverage section, which is an extensive compilation of articles from the national and regional media that deal with campaign issues. The section provides information on quite a few court cases, but its collection could be biased as it may cover only the cases involving more famous candidates. To avoid this potential pitfall, I contacted the election commissions in each of the 89 subjects of the Russian Federation and asked for information on electoral disputes adjudicated in their region during the 2003 campaign. I also contacted all regions’ highest courts and searched the decisions database of the Supreme Court of the Russian Federation. Finally, I scoured reports by election monitoring nongovernment organization NGO Assotsiatsiya “Golos”. These information-gathering techniques yielded information on 118 cases decided in 141 of the country’s 225 SMDs, where a total of 1,974 candidates ran for a seat. As in the Ukrainian dataset, the excluded regions represent the entire political spectrum – from reformist regions, through moderate and Red Belt communist regions, to authoritarian ethnic republics.³

Finally, I collected information on the viability of each of the candidates who were initially registered in the SMDs, whose electoral dynamics I analyze. During the roughly two months between registration and election day, both the regional and national press covered the SMD races and identified the main contenders for most seats. Many local newspapers ran series of articles on their region’s SMDs and pointed out who were viable (prokhodnye) candidates in each district. In addition, think tanks and pollsters conducted weekly public opinion polls in many districts and sifted the candidates who had a shot at a victory from the also-rans. Using a diverse set of sources, I identified 283 candidates in Russia (14 percent of the sample) and 255 candidates in Ukraine (13 percent of the sample) as viable contenders with a realistic chance of winning their district seats.

At the end of the data-gathering process, I ended up with detailed information about 1974 Duma candidates and 1953 Rada candidates – their political affiliation, their viability at the start of the campaign, and their experience with going to court over an electoral registration dispute. Table 4.2 lists the main cumulative figures that describe the two samples. The data are highly comparable as each country’s sample includes roughly the same number of SMDs and the same number of candidates. Moreover, the rates at which candidates go to court are roughly the same. This similarity suggests that Russian and Ukrainian candidates have roughly the same proclivity toward using the courts to resolve electoral disputes, rather than extrajudicial methods and venues. It also suggests that the courts in both countries are roughly equally used as an important arena for the implementation of “electoral technologies.” The general win-rate is also similar at around 50 percent in both countries. The one major difference is that there were more competitive districts⁴ in Ukraine than in Russia, which is additional evidence that political competition, exemplified by electoral uncertainty, was higher in Ukraine than in Russia in the early 2000s.

Table 4.2. Descriptive statistics of the electoral registration datasets

What Can a Plaintiff’s Political Affiliation Tell us About Judicial Independence?

I hypothesize that when judicial independence from politicians is low, the political affiliation of litigants is a significant predictor of success in court. In the case of electoral registration disputes, if judicial independence from politicians is low, progovernment candidates will win more often than anybody else, ceteris paribus, and opposition candidates will win less often than anybody else, again ceteris paribus. The size of the difference between the chances of victory of progovernment and opposition candidates is thus a direct measure of the level of judicial independence from politicians – the larger the difference, the lower the level of judicial independence. To examine the relationship between candidates’ political affiliation and their chances of success in court, I control for two other factors that may affect judicial decision making: a candidate’s viability and the competitiveness of the district.⁵

Candidate viability. Viable candidates are both more likely to participate in a court case and more likely to eventually win in court. By viable candidates, I mean those who were widely believed to be among the top contenders for a given SMD seat at the outset of the campaign, before any registration disputes could arise. Candidates who have a realistic chance of winning the seat from the outset of the campaign should definitely be more likely than the average candidate to fight a decision by the DEC to cancel their registration. In addition, viable candidates probably also become the target of court challenges by their opponents more often than candidates who stand no chance of gathering a significant percentage of the vote. For starters, it seems wasteful to try to take down the registration of someone who does not threaten your chances of winning the seat. Moreover, even if the challenge of a powerful opponent’s registration fails in court, the plaintiff can probably claim that the defendant escaped punishment either due to a technicality or thanks to corrupting or pressuring the judge into delivering an unfair ruling. Given the low esteem in which the public holds both judges and politicians, many voters would probably be inclined to believe the alleged misdeeds even if the charges do not stick. Therefore, lawsuits and counterlawsuits could be an effective “black PR” strategy for viable candidates. In other words, viable candidates should be more likely to become involved in electoral registration lawsuits.

Electable candidates probably also have a higher probability of winning in court than also-rans. The first reason why this advantage may exist is that even the most impartial judges would probably be reluctant to deregister a candidate who appears headed for an election victory because such an act seems countermajoritarian and somewhat undemocratic. Of course, it is crucial to find out whether only propresidential candidates enjoyed this advantage or whether it extended to opposition nominees. Another reason why viable candidates might have a better batting average in court is that they probably, on average, devote more resources to the legal fight. Since their stakes in the ultimate outcome are higher, it seems reasonable to assume that they would try harder to influence the race (Priest & Klein, Reference Priest and Klein1984).

District Competitiveness. Finally, I hypothesize that the probability of being involved in a court case in the first place is affected not only by the candidate’s characteristics but also by environmental factors. Specifically, it seems that competitive districts should yield more court cases than districts where the campaign is but a chronicle of an election victory foretold. Both meritorious and frivolous lawsuits should be more numerous in competitive districts. “True” cases should abound because candidates are more likely to push the limits of acceptable campaigning when they are facing stiff competition. For example, all leaders in the polls with access to administrative resources would mobilize them if their opponents were breathing down their necks. Frivolous lawsuits would also occur more frequently, since, as I previously mentioned, they can be an effective campaign strategy.

Measuring the Independent Variables

Political Affiliation. To measure the main variable of interest, I created two dummy variables – a pro-government variable, where I assigned a 1 (one) to all candidates who had the incumbent regime’s backing and an opposition variable, where I assigned a 1 (one) to all candidates who openly challenged the incumbent regime. How to measure a candidate’s position vis-à-vis the incumbent regime was somewhat easier in the Russian than in the Ukrainian context.

In 2003, the Kremlin openly supported one party, United Russia, so all its candidates in the SMDs were unambiguously propresidential. United Russia ran 100 candidates in the 141 SMDs in the sample. In most of the districts where United Russia did not put forward a candidate, the federal center nevertheless threw its weight behind one of the candidates. I classified 55 candidates as unofficially supported by the Kremlin.

I was conservative in identifying such candidates because of a tendency by the regional press to exaggerate certain candidates’ ties to the center. Since Putin’s stamp of approval carried enormous political value due to his popularity, independent candidates and nominees of parties other than United Russia had a strong incentive to portray themselves as being close to the president. I expected that the regional press that often publishes prepaid articles and passes them off as reporting or editorializing would reflect this bias. Therefore, to classify independent candidates as Kremlin protégés, I used only objective criteria such as an official endorsement by the United Russia leadership or by the governor, if the latter was a member of United Russia. For example, Vyacheslav Shport in Khabarovsk and, the Nizhnii Novgorod governor’s wife, Gulii Khodyreva, fit the bill (Tselobanova, Reference Tselobanova2003; Sokolov, Reference Sokolov2003). Finally, some candidates, such as former Minister of Justice, incumbent Duma deputy and head of the Duma’s Committee on Law-making, Pavel Krasheninnikov ran both in an SMD and on United Russia’s federal list (Grankin.ru, Reference Grankin.ru2004). Thus, even though in the Magnitogorsk SMD he was officially an SPS nominee, I classified him as a candidate supported by the federal center.

I also classified as propresidential some “technical candidates” for United Russia (UR) and any existing “clones” of UR nominees’ main rivals in each district. The main goal of “technical candidates” in entering the race was either to campaign in favor of the United Russia nominee and pull out of the race in favor of the principal a few days before election day or to take away votes from a viable competitor. Such candidates were, for example, Igor’ Artemenkov in Moscow’s SMD #200, who withdrew from the race in favor of eventual winner, Vladimir Vasil’ev, and Anatolii Shiryaev in Volgograd’s SMD #72, whose task was take away votes from the communist, Aleksandr Kulikov, in order to bolster the chances of victory for United Russia nominee, Aleksandr Ageev (Artemenkov, Reference Artemenkov2004; Kuts, Reference Kuts2006). Corroboration that a candidate named “Sergei Vladimirovich Kprf” (Ryazan SMD #149) had a single-minded mission to undermine the KPRF nominee in the district seems unnecessary. “Clones,” or candidates who share the same name as a viable candidate, also register with the sole purpose of confusing voters and thus lowering the viable candidates’ final tally.

Deciding who the opposition was in Russia in 2003 also required making some analytical choices. Ostensibly and rhetorically United Russia faced opposition from all sides. From the left – in KPRF and Rodina; from the democratic right – in SPS and Yabloko; and from the far right – in LDPR. In reality, however, Rodina and the LDPR were hardly a veritable opposition to the incumbents.⁶ The voting record of the LDPR in the Third Duma shows that their leader’s inflammatory pronouncements were empty bravado and the party was more useful than harmful to the Putin administration because it provided an outlet for disgruntled voters, yet behaved loyally and predictably in parliament (Levada, Reference Levada2004, p. 49). Rodina also postured as opposition, but the extensive media coverage that it received led many observers to believe that the Kremlin saw Rodina as a useful and acceptable version of the KPRF. There were indeed reports that the Kremlin (and more specifically, deputy head of the Presidential Administration, Vladislav Surkov) created Rodina, which was to play the role of a “technical party” for United Russia.⁷

The Kremlin’s attitude toward Yabloko and SPS was ambivalent. On the one hand, many believe that the main trigger of Khodorkovski’s arrest and prosecution was his decision to fund Yabloko and the SPS, which suggests that the Kremlin perceived the democratic parties as dangerous competitors, which could undermine its goal to garner a constitutional majority for United Russia (Mereu, Reference Mereu2003). In addition, Rodina’s anti-SPS tirades also might have been indirect digs by the Kremlin. But, on the other hand, Putin appeared with SPS leader Chubais a few days before the election, which could be interpreted as a token gesture of support by the Kremlin. Khodorkovski himself actually ventured a radically different take on the Presidential Administration’s relationship with the right opposition. He argued that for the first time since 1991, the Kremlin simply refrained from actively supporting the SPS and Yabloko and both floundered in the polls as a result of their inherent unpopularity with the electorate (Khodorkovskii, Reference Khodorkovskii2004).

Following the different interpretations, I constructed four political affiliation dummy variables – two to capture opposition status vis-à-vis the regime and two to reflect propresidential bloc affiliation. The first opposition variable, KremOpp1, codes only KPRF nominees and independents openly aligned with the communists. The second one, KremOpp2, also includes SPS and Yabloko nominees, as well as the independents these parties campaigned for. The first propresidential variable, KremBacked1, codes only United Russia nominees and independents overtly supported by the Kremlin. The second variable, KremBacked2, includes also LDPR and Rodina nominees, as well as protégés of these two parties.

Classifying Ukrainian candidates according to where they stood along the main propresidential/antipresidential cleavage also required in-depth investigation of local politics and the background of smaller parties. The first complication arose from the fact that a lot of candidates registered as independents, but in effect belonged firmly to one of the two camps. An initial cut at the problem was to look at the party affiliation that each candidate was required to report in his or her registration application. It seems evident that a self-nominated candidate who is a member of SDPU(o) or one of the constituent members of the Za Edu bloc (Kinakh’s Trudova Ukraina or Yanukovych’s Partiya Regioniv, for example) would not be a “true” independent.

Another obvious step in the proper identification of the political leanings of each candidate was to code all “clones” as political opponents of the viable candidate in their district, whatever his or her political affiliation. Since clones register only to take away votes from viable contenders, they cannot possibly be independent players. The final easy step was to code the major party and bloc representatives along the progovernment/opposition cleavage. The propresidential forces included Za Edu and all its constituent parties, SDPU(o), Zhinki za Maibutne, Democratic Party-Democratic Union, and the Green Party. The opposition camp consisted of Yushchenko’s Nasha Ukraina, Simonenko’s Communist Party of Ukraine, Moroz’s Socialist Party of Ukraine, and the Yuliya Tymoshenko Bloc.

The incumbents’ desire to split every portion of the opposition vote was also fairly transparent. To this end, several party “clones” sprung up shortly before the campaign started. Narodnyi Rukh Ukrainy Bloc (which included the ironically named National Rukh of Ukraine For Unity) aimed to steal nationalist votes from Nasha Ukraina, which contained the “real” Rukh, but it was not particularly successful at the national level. Hence, the irony in the “for unity” part in the name of a party that seeks to break up rather than unite the vote. It garnered only 0.16 percent of the vote or around 40,000 votes. Komanda Ozimoho Pokolinnya (KOP) had the task of attracting young liberal voters who would otherwise probably support Nasha Ukraina. This ploy worked better as KOP got over half a million votes (2.02 percent). The Communist Party of Ukraine (renewed) was also somewhat successful in splitting the communist vote and received 1.39 percent of the vote. The 362,712 votes that went to KPU(o) constituted percent of the communist vote. Finally, the Natalia Vitrenko Bloc was designed to mop up the votes of people frustrated by the loss of the Soviet social safety net but, for some reason, reluctant to support the communists. The bloc’s leader was also a rather effective attack dog who constantly spewed conspiracy theories about Yushchenko’s ties with the United States and even the Nazis. Vitrenko’s cozy relationship with the incumbents was evident from her wide media exposure. Given her extremist antimarket rhetoric, one would think that the oligarchs close to the president would dislike her as much as they feared Communist leader Simonenko. Yet Vitrenko was all over national TV, while Simonenko was almost completely shut out (Wilson, Reference Wilson2002).

Some parties that ran quite a few candidates in the SMDs were much harder to place on the main political cleavage of 2002. The most problematic cases are Mikhail Brodskii’s Yabluko and Kyiv mayor Omelchenko’s Ednist. At first glance, Yabluko appears to be the government’s attempt to attract the votes of liberal and democratic-minded citizens by creating a party that echoes Russian Yabloko’s established oppositionist stance. Yabluko’s leader, Mikhail Brodskii, made his sizable fortune in a joint venture with the leader of SDPU(o), Viktor Medvedchuk, the grey cardinal of the Kuchma regime. Finally, Yabluko almost outspent Nasha Ukraina on TV advertising, so it is no surprise that most observers consider Yabluko to be a Za Edu satellite (Wilson, Reference Wilson2002).

If Yabluko was indeed doing the regime’s bidding, however, it is hard to explain why Mikhail Brodskii initiated a legal battle with the Za Edu right from the start of the election campaign. In late February, at the outset of fierce campaigning, Brodskii filed a complaint at the CEC arguing that the First National TV Channel, UT-1, was illegally campaigning for Za Edu. The plaintiff also alleged that the Za Edu might be providing free transportation to the Ukrainian Paralympics team in violation of Art. 29, par. 4 of the election law. Based on these charges, Brodskii demanded nothing short of the temporary closing of UT-1 and the deregistration of Za Edu’s top party-list candidates: Litvin, Kinakh, Seminozhenko, Tihipko, Sharov, Derkach and Pustovoitenko. To no one’s surprise, the CEC did not go for the bold move of disqualifying the leaders of the main propresidential force. In fact, the CEC did not even consider the merits of the complaint, but refused to hear it on a procedural technicality. Brodskii, however, took his appeal to the Supreme Court, which on February 20, 2002, passed the hot potato back to the CEC – it ordered the commission to rule on the merits of Brodskii’s complaint (Tsentral’na Viborcha Komisiya, 2002, pp. 78–9). On February 25, the CEC dismissed the complaint as unfounded and the issue was soon back at the Supreme Court. Finally, on March 11, the Supreme Court closed the case by denying Brodskii’s appeal and upholding the CEC decision (Tsentral’na Viborcha Komisiya, 2002, pp. 70–2).

The story did not end there. Only four days after the second Supreme Court decision, the CEC found significant discrepancies in the income declarations filed by both Brodskii and Yabluko’s #2 figure, Viktor Chaika, and deregistered both candidates. With the top two candidates on the party list disqualified, the party as a whole faced exclusion from the ballot. Brodskii and Chaika immediately appealed the CEC decisions to the Supreme Court, and by March 21 their registrations were reinstated (Tsentral’na Viborcha Komisiya, 2002, pp. 33–6). Since Brodskii, Chaika, and Yabluko ultimately did participate in the election, a conspiracy-minded observer could interpret the whole affair as the government’s attempt to create a Trojan horse in the opposition camp. The court cases, the conspiracy theory would go, were just instruments in the chiseling of Yabluko’s image as legitimate opposition, but once Yabluko entered parliament, they would cooperate fully with its sponsors from the propresidential camp.

This theory seems rather far-fetched because it seems somewhat risky to attract public attention to Za Edu’s abuse of administrative resources at such a high level as the Supreme Court and the CEC. Moreover, after successfully defending its leaders’ registration, Yabluko came out with a scathing declaration, which urged voters to support “all opposition forces”. Yabluko’s leadership went on to call the establishment “not a government, but a criminal regime […] which has to be replaced at every level – from lower ranks to the very top” (Unian, 2002). Yabluko did not make it into parliament, but remained in opposition throughout the rest of Kuchma’s presidency and supported Yushchenko’s presidential candidacy in 2004, so it seems that they did move into the opposition camp for real. It seems that the party was simply the personal vehicle of an opportunistic oligarch, Mikhail Brodskii, whose ego tempted him to confront the establishment and gradually left him no choice but to support the opposition. Based on this contextual analysis, I concluded that Yabluko’s candidates in the SMDs were probably the closest things to “true” independents because they were neither clearly aligned with the opposition, nor actively helped by the establishment.

Ednist is another difficult nut to crack. Oleksandr Omelchenko was certainly cozy with the president himself; however, like Brodskii his personal ambition made him a fickle ally. Unlike Brodskii, he had access to his own administrative resources, which gave him confidence in the success of his candidates on election day. It seems that Omelchenko expected that the Ednist faction in the new Rada could find itself in the position of being the tie-breaker between the propresidential and antipresidential camp. Consequently, during the campaign, Omelchenko wanted to keep his options open as much as possible. The “nonaggression pact” that Ednist and Nasha Ukraina purportedly signed at the start of the campaign seems to support such an interpretation of Omelchenko’s position (Storozhenko, Reference Storozhenko2001).

In addition, Ednist was overtly hostile to SDPU(o), Za Edu’s major ally in the propresidential camp. The Ednist faction in the outgoing Rada had initiated a petition demanding Viktor Medvedchuk’s resignation, so there was no love lost between Omelchenko and the SDPU(o) leader. However, the president himself seemed somewhat supportive of the Kyiv mayor, Omelchenko. When Brodskii charged that Omelchenko was illegally holding both the office of mayor and the office of head of the Kyiv state administration, Kuchma seemed to take Omelchenko’s side by calling for an “expert opinion” on the issue before any steps were taken (Chernaya, Reference Chernaya2001).

In conclusion, like Yabluko, the Ednist candidates seemed to be in a neutral position vis-à-vis the main political cleavage of the 2002 elections. Even though they benefited from the administrative resources available to their leader in Kyiv and Kyiv oblast, in regions controlled by Omelchenko’s foes (e.g., Donetsk), they suffered intimidation at the hands of state authorities. So I coded Ednist candidates as neutral in all regions, except for Kyiv and Kyiv oblast where judges were likely to perceive an Ednist candidate as a representative of the party of power.

Table 4.3 provides comparative descriptive statistics on the political affiliation of the candidates in the Ukrainian and Russian samples. The data show that in both countries the groups of opposition-affiliated and government-affiliated candidates are roughly equal. However, in Russia, the proportion of neutral candidates was much higher than in Ukraine, perhaps as a reflection of the lower level of political competition in Russia.

Table 4.3. Descriptive statistics of candidates’ political affiliation

District Competitiveness. It might seem reasonable to assume that the difference between the number of votes that the winner receives and the number of votes that go to the runner-up captures the competitiveness of the campaign. Given this operationalization of the concept, both countries display a wide variation in district competitiveness. In Russia, the closest majoritarian race took place in SMD #163 in Sverdlovsk Oblast, where incumbent Duma deputy Georgii Leont’ev beat provincial oligarch Aleksandr Ryavkin by only five votes! By contrast, in Saratov’s SMD#156, the deputy leader of United Russia’s Duma fraction, Vyacheslav Volodin, overtook his KPRF opponent, Ol’ga Alimova, by more than 250,000 votes, winning 82 percent of the vote to her 9 percent. In the closest Ukrainian race in Odessa’s SMD #135, propresidential candidate Ihor Reznik defeated independent Serhii Bovbalan by 0.25 percent or 260 votes. At the other end of the spectrum, in Ternopil’s district #168, Nasha Ukraina’s nominee Mihailo Polyanchich beat his closest competitor, independent Oleg Povadyuk, by 69 percent or 92,290 votes.

Instead of using the continuous variable, however, I created a DistComp dummy variable, where I assigned 1s to all SMDs where the difference between the winner and the runner-up was less than 10 percent. After applying this coding scheme, I ended up with seventy-two competitive districts in the Ukrainian sample (53 percent of the total) and fifty-two competitive districts in the Russian sample (37 percent of the total). This result is in line with all the other pieces of evidence that the Ukrainian election was more competitive than the Russian election. The rationale for using a dummy variable, which required me to choose a cut-off point, which is inherently somewhat arbitrary, is that beyond a certain threshold an increasing difference between the two top vote-getters stops reflecting competitiveness. Whether the winner beats his closest rival by 25 or 45 percentage points does not really seem to matter much – in both cases the race seems rather uncompetitive.

An important caveat of measuring district competitiveness through election results is that if a major competitor is deregistered by the courts and cannot participate in the election, an extremely competitive district may appear uncompetitive if we only look at the election outcome. To avoid this problem, I classified the five Ukrainian and seven Russian SMDs, where a viable candidate was deregistered, as competitive, regardless of the size of the difference between the winner and the runner-up. In Ukraine, these SMDs were #6, 71, 99, 147, and 148; in Russia, they were SMDs #9, 47, 96, 97, 120, 159, and 179. There were, of course, SMDs where nonviable candidates were deregistered. However, since these candidates were not expected to affect the overall result of the race, I assumed that their deregistration also had no effect on the district’s competitiveness.

Candidate Viability. Capturing candidate viability in a variable is even more challenging and definitely more labor-intensive than gauging district competitiveness. The most obvious operationalization based on looking at the election results and estimating with the benefit of hindsight that anyone within a reasonable shot of the winning percentage must have had a chance at victory when the campaign started could be quite misleading. For one, this method would leave out perfectly viable candidates who did not participate in the election at all as a result of a canceled registration. In Ukraine’s SMD #99 in Kirovohrad oblast, a court deregistered Batkivshchyna candidate Valerii Kalchenko on the very eve of the election. Kalchenko was running both for the parliamentary seat and for mayor of Kirovohrad, and it appears that he was the front-runner in both races. Not only did the local press identify him as clearly viable (prokhodnoi in Russian), but after the elections thousands of people protested his deregistration (Gorobets, Reference Gorobets2002; Gotsuenko, Reference Gorobets2002). In Russia, a similar fate befell former vice-president and ex-Kursk governor, Aleksandr Rutskoi, who failed to secure registration in Kursk, where he had sky-high name recognition and a very realistic chance of winning. In Nizhnii Novgorod, Krasnoyarsk, and Kurgan, shady businessmen Andrei Kliment’ev, Anatolii Bykov, and Pavel Fedulev were deregistered by court decision. Their high name recognition and virtually bottomless campaign coffers made them viable contenders for the SMD seat. Moreover, Kliment’ev had already proven his electability by winning the 1998 NizhniiNovgorod mayoral race.

Another problem with this approach to estimating viability is that election results reflect the success or failure of administrative resources and administrative resources are usually mobilized precisely to benefit or hurt viable candidates. For example, if we looked at the distribution of the vote in SMD #70 in Western Ukraine’s Zakarpattia oblast, we might conclude that the viable candidates in the district were only two: Serhii Ratushniak, an independent who scored a 35.47 percent victory, and the Nasha Ukraina nominee Ihor Kril’ who received 20.14 percent of the vote. With 9.2 percent of the vote, SDPU(o) nominee Nestor Shufrich would not appear to be among the viable candidates. However, only several months later, Shufrich, whose political base is in his native Zakarpattia, managed to win repeat elections in a district with an opposition-leaning electorate (SMD #201 in Cherkasy). Amazingly enough, despite being an outsider in the Cherkasy election, Shufrich easily defeated both the winner of the regular election, Nasha Ukraina nominee Mykolai Bulatetskii, and Natalia Vitrenko (the leader of the Natalia Vitrenko Bloc), which almost received national representation (Lubenskii, Reference Lubenskii2002). This development suggests that Shufrich was a viable candidate in his home Zakarpattia district as well, but his election strategies failed to yield the desired results.

A more reliable way of estimating candidate viability is to scour the regional and national press for assessments by people familiar with the politics of each SMD, and this is what I did. Based on a media search for each of the 1,974 Russian candidates and each of the 1,953 Ukrainian candidates, I created a dummy variable where 1 was assigned to viable (or to use the Russian language term prokhodnye), candidates and 0 to candidates who, at the outset of the campaign, stood no realistic chance of winning the seat. Whereas some publications surely exaggerate the chances of success of the candidates they support, if the exaggeration is of staggering proportions, then the candidate probably has significant resources. In a way, such exaggerations indicate that the candidate is in this race for real and acts as if he or she were a viable contender for the seat, which for the purposes of this analysis is as important as actual viability.

I supplemented the results of the press reports with data from public opinion polls conducted by regional research institutes and NGOs during the campaign. For example, the Institute for Regional Issues (Institut Regional’nykh Problem) in Odessa conducted weekly polls on a sizable sample of 2,064 respondents and calculated the approval rating of all candidates in the Odessa oblast SMDs. The institute’s reports identify the viable candidates in each SMD (Institut Regional’nykh Problem, 2002). For the Russian sample, I used the assessments contained in the Russian Regional Report⁸ and reports by Group 7/89.⁹

Administrative Resources and Case Merit: Why 2002–2003 Progovernment Candidates Should Have Been Losing in Court

Finally, before proceeding with the model that estimates the importance of political affiliation to a candidate’s victory in an electoral registration dispute, I consider case merit as a potential source of omitted variable bias. Case merit refers to the strength of a plaintiff’s case. Ideally, the rule-of-law doctrine envisions a legal system in which case merit, determined impartially and faithfully by the judiciary on the basis of the laws on the books, should be the one and only predictor of victory in court. Courts, which are independent and impartial, will grant victory to plaintiffs who bring strong cases. Plaintiffs who file frivolous or weak cases, on the other hand, will lose more often than other litigants. It is essential to know whether the political affiliation of the candidates is highly correlated with case merit or not. If candidates of one political stripe have systematically stronger cases, then we should, in fact, expect an impartial and independent court to side with them more often than with other candidates.

Unfortunately, it is all but impossible to capture case merit in an observable variable. If we could, judges and courts would be superfluous. It is also extremely hard to capture judges’ true perceptions of case merit, not least because judges themselves may find it impossible to distinguish between faithful interpretation of case merit according to the legal text and their personal prejudices on the subject or preconceptions about the litigants. Case merit is basically unobservable in individual cases, but it is important to know whether as a variable it might be correlated with our main variable of interest, candidate political affiliation, in order to avoid potential omitted variable bias. In other words, if we happen to find out that progovernment plaintiffs win more often than other plaintiffs, we want to know whether that is because they have stronger cases on average or it is because judicial independence from incumbents is low and judges have to deliver favorable rulings.

Fortunately, there is significant empirical evidence on the relationship between candidate political affiliation and case merit in the context of the 2002 Rada and 2003 Duma elections. We can assume that both in Russia and in Ukraine progovernmental candidates did not have stronger cases than other plaintiffs. If anything, as the main beneficiaries of the practice of using administrative resources (which is an electoral law violation in both countries), progovernment candidates should have weaker cases both as plaintiffs and as defendants. Finally, given electoral technologies that allowed the incumbents to stack the DECs with loyal representatives, we can also assume that most DECs would be reluctant to deregister a progovernment candidate. When they did, they were probably responding to particularly egregious electoral law violations. This dynamic should also lead to progovernment candidates having weaker cases than everybody else.

Domestic and foreign election monitors in both countries cited ample evidence to support these assumptions. The administrative resources and mechanisms employed in both campaigns to boost propresidential candidates and hurt oppositionists were very similar. Both the Committee of Voters of Ukraine (CVU), the country’s largest NGO devoted to monitoring campaigning, and the OSCE/ODIHR observers reported high levels of administrative resource use (Committee of Voters of Ukraine, 2002; OSCE/ODIHR, 2002, p. 12). In Kharkiv oblast, for example, CVU representatives noted that the oblast administration bused people to attend a public “Youth Forum for a United Ukraine,” distributed leaflets and encouraged everyone to support the For United Ukraine (Za Edu) bloc. In Chernihiv, professors and students from local universities were required to take part in a state-organized rally for Za Edu. And in Kyiv, local businesses that refused to purchase Ednist (Mayor Omelchenko’s party) election materials received warnings from city officials to expect rent increases and frequent tax and fire inspections after the election (Committee of Voters of Ukraine, 2002). OSCE observers reported that in Lviv, the head of the local branch of Za Edu distributed free coal and used state vehicles during working hours. In Kharkiv, citizens received free electrical appliances along with notes soliciting votes for Za Edu (OSCE/ODIHR, 2002, p. 13).

In Russia, the main beneficiary of administrative resources was, of course, United Russia. OSCE observers noted that regional governments often supplied all the equipment and most services to the local United Russia campaign headquarters. Opposition candidates, on the other hand, often did not receive permits from the regional authorities to hold rallies and could not find public organizations willing to provide them space to hold meetings. The OSCE also received numerous complaints about police detaining oppositionists’ campaign workers and impounding opposition campaign materials (OSCE/ODIHR, 2004, p. 5).

Some will argue that the OSCE was biased against the incumbents in both elections and as a result painted a skewed picture of the situation. Maybe these were no-holds-barred campaigns, in which all competitors used every dirty campaign trick in the book. If the opposition lagged behind the presidential supporters in administrative resources, it compensated in illegal campaign financing by friendly oligarchs and maybe even the United States. This theory was the leitmotif of propresidential election technologists and Commonwealth of Independent States (CIS) and Russian observers in Ukraine. The official publication of the Russian government, Rossiiskaya Gazeta, hinted that the US$5 million earmarked by the USAID for programs aimed at ensuring the transparency of the elections, was in fact spent on behalf of Nasha Ukraina’s campaign, which of course would be a violation of election law (Bogdanov, Reference Bogdanov2002). Even if this were an accurate description of both parliamentary campaigns, no one has attempted to argue that oppositionists were the main perpetrators of campaign rules violations, so we can safely dismiss the possibility that the propresidential candidates who went to court had stronger cases, all else being equal.

In short, although case merit is unobservable and omitted from the model, it is unlikely that its omission would lead us to find a progovernment bias in court decision making where none exists. On the contrary, if Russian and Ukrainian courts were adjudicating electoral registration cases independently and impartially, we should observe a lower win-rate than average for progovernment plaintiffs.

Sartori Selection Model of Court Appeal-Rate and Win-Rate

To test the proposed hypotheses and determine what candidate and district characteristics can best predict which way the court would go in a dispute over electoral registration, I estimated a Sartori selection model with candidate court experience as the dependent variable. The Sartori estimator is a selection-effects model, which is appropriate to use in cases when the dependent variables in the selection equation and the outcome equation might have correlated error terms. In the case of electoral registration disputes, the decision to pursue a court appeal and the trial outcome are probably both affected by similar unobservable variables, such as case merit. The Sartori estimator assumes that the correlation between the error terms is 1. While such an assumption is probably wrong, the estimator is very robust to that assumption being wrong, whereas the Heckman model (which is a more venerable selection model) is less robust. Moreover, unlike the Heckman, the Sartori estimator allows us to use the same independent variables in both equations. Since the decision to go to court is probably affected by the exact same factors as the ultimate court decision, and the two decisions are close together in time, the Sartori estimator is more appropriate than a Heckman estimator (Sartori, Reference Sartori2003).

The two binary dependent variables of the selection and the outcome equation (decision to go to court and trial outcome) are coded in one trichotomous variable. If a candidate did not participate in any court proceedings during the campaign, he or she received a 0 for court experience. All candidates who went to court to demand the cancellation of an opponent’s registration, to appeal their own deregistration, or to defend their registration from an opponent’s challenge but did not win in court received a 1. Finally, all candidates who scored a victory in court received a 2. The independent variables include the main variables of interest, progovernment and opposition affiliation, as well as the control variables: candidate viability and district competitiveness. The two models’ coefficients are summarized in Table 4.4¹⁰.

Table 4.4. Results of Sartori models for Russia and Ukraine

Note: *p < .1 one-tailed test; ** p < .05 one-tailed test; *** p < .01 one-tailed test

The Sartori models first suggest that judicial independence was low in both Russia and Ukraine during 2002–3. This would actually be a rather uncontroversial statement if we relied only on cross-country rule of law indices. The World Governance Indicators (WGI) puts both Russia in 2003 and Ukraine in 2002 in the bottom 25 percent in the world in terms of rule of law (Kaufmann, Kraay, and Mastruzzi, 2010). My data lend a more fine-tuned and direct confirmation of this empirical claim as they show that in neither country did progovernment candidates have a lower win-rate than other candidates. This finding indicates that the courts did not effectively constrain the incumbent regime and did not check the abuses of power associated with the widespread use of administrative resources. Progovernment candidates bolstered their reelection chances by violating the electoral laws with impunity. Had the courts been independent from the incumbents, we would have observed a lower win-rate for progovernment candidates and a higher win-rate for all other candidates. In other words, the statistical analysis of electoral registration disputes that made it to court during the campaign for the 2002 Rada and the 2003 Duma elections is in line with the charges of oppositionists and Western and local observers that these were not fair elections.

Next, the models confirm the hypothesis that viable candidates participated and won court cases more often than nonviable candidates. This relationship is strong in both countries, and we can be very confident that it exists as it is highly significant. The tendency of viable candidates both to go to court more often and to win more often could reflect these candidates’ higher level of commitment to the race. But the result also lends systematic support to much-discussed anecdotal claims that registration disputes were a prominent weapon in the black PR arsenal and that therefore viable candidates were both the main perpetrators and the main victims of such cases. Theoretically, the results confirm the contention of the strategic pressure model that in emerging democracies, like Russia and Ukraine during the late 1990s and early 2000s, justice is more politicized. The comparative data shows that in both countries, the courts are very useful to politicians as tools of political competition and hence viable politicians use them more often than everybody else.

The rest of the results indicate substantial differences between how courts decided registration disputes in Russia and in Ukraine. In Russia, candidates running in competitive districts were significantly more likely to become involved in a court case. The optimistic interpretation of this result suggests that candidates were trying in earnest to play by the rules, but those who tried to get the upper hand in a very close race, predictably crossed the line into forbidden campaign tactics more often, which resulted in the significantly larger number of disputes in competitive districts. The cynical, and arguably more plausible, interpretation is that the use of “election technologies” was more prevalent in the highly competitive districts where viable candidates dug up dirt on each other and tried to use it to derail competitors’ quest for the seat.

The Ukrainian data, however, hint at a worse situation than the cynical Russian scenario. The district competitiveness coefficient is not only insignificant but also very small. This result suggests that district competitiveness likely had no effect whatsoever on the likelihood that candidates would go to court over registration disputes. In other words, during the 2002 Rada campaign, candidates sued each other regardless of whether the race was competitive or not. There are two ways to account for this outcome. One possibility is that electoral law violations took place regardless of whether the race was close. This means that candidates had complete disregard for election law provisions and violated them even when a victory seemed secure. Another possibility is that lawsuits are not highly correlated with actual electoral law violations, but are just a method for harassing the competition. In all likelihood, both possibilities describe the 2002 Rada campaign. The point is that the Ukrainian courts were deeply embroiled in politics because they were useful tools for competing politicians.

Another difference is that in Russia, opposition candidates were significantly more likely to go to court, whereas in Ukraine political affiliation had no bearing on the decision to go to court. We should not overestimate this difference, though, because the Russian result barely meets a low significance threshold and because both coefficients are positive and of similar size. In other words, in both countries, opposition candidates were probably slightly more prone to filing lawsuits, it is just that in Russia we can make this claim with somewhat more confidence. This result may be interpreted as confirmation of anecdotal observations that opposition candidates had more registration problems than progovernment candidates and thus were more likely to end up involved in a registration dispute.

The most important result, however, is that in Russia the political affiliation of the plaintiff has a smaller effect on a candidate’s probability of victory in an electoral registration trial. The coefficients indicate that any advantage that Russian progovernment candidates may enjoy in court is about two-thirds the size of the Ukrainian progovernment candidates’ advantage. Moreover, the Russian coefficient is not statistically significant, which suggests that we are not certain that such an advantage even exists. In Ukraine, on the other hand, progovernment candidates enjoyed a significant advantage when their registration disputes end up in court. This contrast suggests that judicial independence from politicians is lower in Ukraine than in Russia, as Ukrainian judicial output consistently reflects the preference for the incumbent regime.

It is important to emphasize that the model indicates that political affiliation matters in Ukraine even after we control for candidate viability. In other words, propresidential candidates win not simply because a larger percentage of them have a realistic chance of winning and the courts are reluctant to foil the will of the electorate by deregistering a front-runner. Rather, Ukrainian propresidential candidates win more often simply thanks to their affiliation with the establishment.

Effect Magnitude as a Measure of Judicial Independence

The Sartori estimator not only suggests that political affiliation is a significant predictor of success in court but also allows us to predict exactly how much more likely a propresidential candidate is to win in court in comparison to other candidates. The difference in the predicted probability of court victory represents the extent to which politicians can impose their preferences on the courts. In other words, the greater the difference between the predicted win-rates of progovernment and opposition candidates, the lower the level of judicial independence from incumbent politicians.

I use the Sartori model coefficients to calculate and compare the predicted probabilities of court victory for four groups of plaintiffs, depending on their political affiliation and viability. District competitiveness is held constant at its modal value (0 in Russia; 1 in Ukraine). Figure 4.1 compares the probability of court victory for nonviable opposition-affiliated and government-affiliated candidates in Russia and Ukraine.

Figure 4.1. Predicted probabilities of success in court according to the political affiliation of non-viable candidates: Comparison between Russia and Ukraine.

The contrast between Russia and Ukraine is quite stark. In Russia, nonviable candidates have roughly the same probability of winning a registration dispute in court. If progovernment candidates have any advantage at all over opposition-affiliated candidates, it appears to be about 5 percent. In Ukraine, on the other hand, progovernment candidates enjoy a big advantage over opposition-affiliated candidates – almost 25 percent. Candidates close to the Kuchma regime were twice as likely to win a court case, as they were to lose, even if they were opportunists who did not have a realistic chance of being elected through a fair process.

I also calculated the 95 percent confidence intervals for each predicted probability in order to show the amount of uncertainty associated with each estimate. In Russia, the difference between the predicted win-rates of Russian candidates is not statistically significant – progovernment candidates have a 12 to 40 percent probability of victory, and opposition candidates have an 8 to 34 percent probability of victory. By contrast, the results suggest that Ukrainian candidates of different political stripes have statistically different chances of winning a registration lawsuit because the confidence intervals associated with the two estimates barely overlap. The confidence interval for progovernment candidates is 46 to 83 percent, and for opposition candidates it is 23 to 54 percent.

The difference between Russia and Ukraine is also significant when we compare the predicted probabilities of success of viable candidates. As Figure 4.2 shows, in Russia progovernment candidates appear to have about a 12 percentage point advantage over opposition-affiliated candidates. The confidence intervals of the two predictions overlap significantly – progovernment candidates have a 46 to 100 percent probability of victory, and opposition candidates have a 34 to 96 percent probability of victory. Ukrainian government protégés’ advantage is more than twice as big, at 27 points. In fact, viable candidates who were close to the Kuchma administration had a tight lock on the courts. They were almost assured to win any electoral registration dispute that they were involved in, their predicted probability of winning being at 89 percent. The confidence intervals also do not overlap as in the Russian model, which points to a statistically significant difference between the two predicted probabilities. Pro-government candidates have a 74 to 100 percent probability of victory, while opposition candidates have a 43 to 89 percent probability of winning in court.

Figure 4.2. Predicted probabilities of success in court according to the political affiliation of viable candidates: Comparison between Russia and Ukraine.

Implications of the Electoral Dispute Analysis for Judicial Independence Theories

The direct measure of judicial independence, which Chapter 3 proposed and this chapter applies, allows a fine-tuned comparison of the level of judicial independence in two very similar countries. The results indicate that, contrary to impressionistic accounts and presumptions, the Russian courts had slightly more independent output from politicians’ preferences, than Ukrainian courts, even if we cannot really talk about independent courts in either country.

More importantly, however, this methodological innovation facilitates a test of the predictions of existing theories of judicial independence. The traditional political competition theories predict just the opposite results. They expect that as weak incumbents, Kuchma and his political circle, would refrain from pressuring the courts and, in fact, would bolster judicial independence. The political insurance theory expects the Kuchma regime to promote independent courts as a guarantee against future persecution by the next incumbent. The strategic defection theory expects that by 2002, Ukrainian judges should have been abandoning the sinking ship of the Kuchma regime. The policy stability hypothesis posits that the Kuchma administration would try to buttress independent courts in order to make sure that after the loss of power (which by 2002 should have seemed a plausible turn of events) would not mean a complete loss of policy control. Rather Kuchmists would retain some policy control by being able to rely on independent courts. On the other hand, Putin’s electorally secure regime supposedly had none of these incentives to allow independent courts. Russian judges also purportedly had no incentive to stray from supporting the strong Putin regime through favorable rulings for pro-Kremlin candidates involved in registration disputes.

None of these predictions are borne out by the electoral registration data. On the contrary, the data show that Ukrainian courts produced more decisions in favor of the incumbent regime than Russian courts did. This finding bolsters the prediction of the strategic pressure theory that the more competitive regime would display a lower level of judicial independence.

The quantitative analysis of electoral registration trial outcomes does not in itself illustrate the mechanism of the strategic pressure theory, though. In other words, the fact that Ukrainian courts have delivered more progovernment rulings than Russian courts is not in itself proof that the Kuchma regime has pressured judges into reaching the decisions that they did. Evidence that this is the reason behind the notable progovernment bias in electoral registration cases is provided in Chapter 7. In addition, the finding that, despite a nontrivial difference, judicial independence from politicians was low in both Russia and Ukraine challenges the predictions of the institutional theories of judicial independence. According to these theories, the structural insulation of the Russian and Ukrainian judiciaries should have facilitated the emergence of independent courts, as formal institutional safeguards should have made it difficult for incumbent politicians to impose their preferences on the courts. Instead, the electoral registration data show that both the Russian and the Ukrainian courts failed to hold incumbent politicians accountable for electoral law violations. Despite incumbents’ ubiquitous use of administrative resources, neither Russian nor Ukrainian government politicians had a lower than average win-rate in court, when oppositionist candidates decided to fight back and attempted to seek redress through the courts. “Good”, purportedly rule-of-law-fostering institutions have not produced a “good” enough outcome in either Russia or Ukraine.

Because structural insulation of the judiciary is higher in Russia than in Ukraine (as shown in Chapter 3), however, it is still possible to attribute the slightly less dependent Russian judicial output to institutional constraints, rather than to strategic incentives (as the strategic pressure theory does). In other words, maybe Russian courts were slightly less dependent on politicians because the institutional setup of the judiciary made interference in judicial decision making slightly more complicated for Russian politicians? Maybe the “good,” insulated judicial institutions have begun to foster rule-of-law-abiding behavior among Russian politicians? Maybe it just takes time for these gains to be solidified, so that we can observe the entrenchment of independent courts in the postcommunist context? Maybe in time Russia’s judicial insulation will produce a high judicial independence ranking not just relative to Ukraine but in absolute terms as well? Chapter 6 addresses this alternative explanation of the empirical results presented in this chapter. It argues that the formal insulation of the Russian (and the Ukrainian) judiciary can and is routinely circumvented by enduring informal practices, which make it easy for any motivated Russian (or Ukrainian) political incumbent to intervene in judicial affairs and impose his or her preferences on individual case outcomes.

Before I get to the mechanism testing, however, the following chapter (Chapter 5) expands the analysis of judicial independence to another politically salient issue area, namely defamation lawsuits against media outlets. The goal is not only to check how independent Russian and Ukrainian courts are when it comes to guaranteeing another central civic right, namely the right to free speech. The goal is also to perform a robustness check on the comparative results presented in this chapter. If the contrast between Russian and Ukrainian courts holds beyond the electoral registration cases, then we can be more confident that these results are not some random artifact of the electoral laws. If Russian courts are more independent than their Ukrainian counterparts in two distinct types of politically important cases, we can be more certain that there is indeed an underlying difference in judicial independence between the two countries.