Voting ‘against all’ in postcommunist Russia

Abstract

Since the early 1990s voters in Russia (and most of the other post-Soviet republics) have been offered the opportunity to vote ‘against all’ parties and candidates. Increasing numbers have done so. The evidence of two post-election surveys indicates that ‘against all’ voters are younger than other voters, more urban and more highly educated. They do not reject liberal democracy, but are critical of the contemporary practice of Russian politics and find no parties that adequately reflect their views. With the ending of the ‘against all’ facility in 2006 and other changes in the Russian electoral system under the Putin presidency, levels of turnout are likely to fall further and the protest vote will seek other outlets within or outside the parliamentary system.

To vote means to ‘send signals to government’ (Miller & Harrop 1987, ch. 4). For decades, voters in the USSR could send only a single message: of unqualified approval. At the last elections conducted on this basis in March 1984, turnout was 99.99% and the vote in favour of the single list of candidates ranged from 99.91 up to 100% (Soobshchenie 1984); the results were so predictable that newspapers could be prepared the day before with pictures of the winners. 1 In December 1988 an entirely new electoral law introduced important changes, and in March 1989 the first elections took place that allowed a choice of candidate, if not yet of party; nearly 40 leading party functionaries were defeated and the Kremlin's leading opponent, the future Russian president Boris Yel’tsin, romped home in Moscow with almost 90% of the vote (White 1991). The elections, Gorbachev told the Politburo, had been a ‘major step in carrying out political reform and in the further democratisation of society’ (Itogi 1989); his leadership colleagues saw them more accurately as ‘political shock therapy’ (Ligachev 1992, p. 75). Either way, they were arguably the decisive moment in the end of communist rule as ordinary people took advantage of their opportunities to elect a new-style legislature that reflected their own preferences rather than those of the regime itself.

From 1993, in what was now an independent and postcommunist Russia, another set of electoral arrangements took their place. A new parliament, the Federal Assembly, came into existence; half the seats in its lower house, the State Duma, were to be elected by a national contest among party lists, the others by an equal number of single-member constituencies throughout the country. Elections to the Duma were conducted on this basis in 1993, 1995, 1999 and 2003; there were parallel elections to the presidency in 1991, 1996, 2000 and 2004. Nominally, power had been placed in the hands of voters themselves: they could choose to vote or otherwise, and if they chose to do so they could select from a variety of parties and candidates. And yet turnouts fell sharply: more than 89% had voted in the first largely competitive elections in March 1989, but only 56% did so in December 2003. At the same time, a considerable body of survey evidence suggested there was little faith that competitive elections had given ordinary Russians an effective means of influencing the government that was nominally accountable to them. Instead of eliminating the gulf between regime and society, they seemed to have ‘only deepened it’ (Zinov’ev & Polyashova 2003, p. 8; Skosarenko 2007, p. 58); and when asked in 2007 to look forward to the Duma elections in December, just 35% thought they would reflect public opinion as compared with 50% who took the opposite view—a reversal of the position that had obtained four years earlier (Vovk 2007).

Russian voters, in the early years of the new century, had a variety of means at their disposal by which they could signal their preferences. They could abstain; they could vote for oppositional parties or candidates; they could spoil their ballot paper, or fail to make use of it. 2 But in addition, Russian electors could vote ‘against all’ the party lists or candidates, and increasing numbers did so—in the Duma election of December 2003, nearly three million did, which put ‘against all’ ahead of all but four of the party lists with almost enough votes to claim representation in its own right. In the parallel series of contests in the single-member constituencies, more than seven million, or nearly 13%, voted against all the candidates, which put it ahead of all but one of the parties. How had this distinctive electoral option come into existence? How many made use of it? What kind of voters made use of it, as compared with those who voted for a party or a candidate or simply abstained? And what are the consequences likely to be of the ending of the ‘against all’ facility in 2006 and related changes to the electoral system under the Putin presidency? We offer some responses to these questions in what follows, based primarily on individual-level data collected in national surveys in 2004 and 2005. 3

Voting against: the legislative framework

There had been no provision for a vote against all the candidates in Soviet electoral legislation, although a voter who crossed out the name of the single candidate had, in effect, voted against all the candidates on the ballot paper (White et al. 1997, ch. 1). At the last elections of this kind, in March 1984, no more than 100,000 out of an electorate of more than 180 million exercised their right to do so, and there were 17 invalid votes (Soobshchenie 1984, p. 200). Reform of this system of elections without choice was being discussed within the leadership as early as January 1987—the ‘main thing’, participants agreed, was that ‘the elections should be elections, and we don't have them at the moment’ (Obsuzhdenie 1987, p. 102). Following further discussion, an entirely new electoral law was approved in December 1988 that provided for the nomination of independent candidates, and for a genuinely secret ballot in which all voters had to make use of the polling booth (O vyborakh narodnykh deputatov SSSR 1988). In March 1989, in the first of these new-style elections, more than two thirds of all seats were contested, in some cases by a dozen candidates. 4 The law, however, provided only for a vote for or against each of the candidates; it did not allow a vote ‘against all’ of them, still less a vote ‘against all parties’ (there was still only one).

There were elections the following year in all of the 15 Soviet republics, including the largest of them, the Russian Republic. Again, they were largely competitive, with more than 6,700 candidates for the 1,068 seats that were available, and up to 28 nominations for a single seat. 5 But once again, there was no explicit facility for a vote ‘against all’: electors were required to cross out the names of all the candidates except the ones they were voting for, and the winner was the candidate who secured more than half the vote, provided at least half the registered electorate had taken part (O vyborakh narodnykh deputatov RSFSR 1989, arts 46 – 48). The first formal provision for a vote ‘against all’ came the following year, in the law on the election of the Russian President that was adopted in April 1991. Article 14 of the law made clear that a vote could be cast ‘against all candidates’, not just in favour or against each of them individually, and also established that the vote against all the candidates would be reported, even though it would have no direct significance for the result (O vyborakh Prezidenta RSFSR 1991). In the event, more than a million and a half exercised this option, which was 1.9% of the total. 6

The ‘against all’ facility was extended further when a newly established State Duma was elected in December 1993, in what was now a postcommunist Russia. The architects of a new and rather different representative system, introduced in controversial circumstances after the outgoing parliament had been bombed into submission, were understandably anxious to attract the highest possible level of participation. The new election regulations accordingly retained the ‘against all’ option, so that voters could express their dissatisfaction with all of the parties or candidates on offer by taking part rather than abstaining. Article 35 of the regulations, which specified the form of the ballot paper, made clear that it should not only allow voters to choose an electoral bloc or candidate, but that it should also allow them to vote ‘against all lists’ or ‘against all candidates’ by ticking the corresponding box. In the original version of the regulations, elections in the single-member constituencies would be invalid if the vote ‘against all’ exceeded the vote in favour of the most successful candidate (Polozhenie 1993); this provision was, however, removed in an amendment of 6 November (Ob utochnenii 1993). The same provisions applied to the upper house, the Federation Council, which was directly elected in 1993 for the first and (so far) only time.

A more general ‘framework law’, which was intended to govern the conduct of elections at all levels, was adopted a year later (Ob osnovnykh 1994). It confirmed that electors should continue to receive a ballot paper that allowed them to vote against all candidates or party lists (art. 30), and that electoral commissions would be obliged to report these totals as well as those for individual parties and candidates (art. 31). These provisions were repeated in a new framework law, adopted in 1997, and in its successor, adopted in 2002. However, the 1997 law, and its successor, introduced an important change. From this point an election would be considered invalid, not just if the turnout fell below the level that had been specified, but also if the vote against all candidates was greater than the vote in favour of the most successful. These new arrangements applied to elections in the single-member constituencies, and to presidential contests (Ob osnovnykh 1997, art. 58.2). 7 The 2002 law made clear in addition that elections in multi-candidate constituencies—if any were established—would be governed by the same procedures, and that successful candidates in a runoff as well as a first-round presidential election would be required to secure more votes than had been cast against both of them (Ob osnovnykh 2002, arts 70.5, 71.2).

More detailed provisions have been made in successive versions of the law on elections to the State Duma, adopted before each of the relevant contests. The 1995 law, the first to take the form of a statute, provided as in 1993 that the winning candidate in a single-member constituency would be the one who secured the largest number of votes, regardless of the vote against all (which would nonetheless be reported). In the party-list election, the other seats available would be distributed among all the contenders that secured at least 5% of the vote, provided turnout was at least 25% (O vyborakh deputatov Gosudarstvennoi 1995, arts 61, 62). The next Duma election, however, followed the adoption of the 1997 framework law, which had insisted that winning candidates in individual constituencies would be obliged to secure more votes than had been cast against all of them. The 1999 election law, accordingly, continued to require that the vote against all candidates and party lists should be collected and reported, but also made clear that the winning candidate in a single-member constituency would be required to secure more votes than had been cast ‘against all the candidates’, failing which a runoff would be held on the same basis to fill the vacancy (O vyborakh deputatov Gosudarstvennoi 1999, arts 79.2b, 83).

The 2002 and 2005 election laws have followed the same guidelines, but with some variations. In the 2002 law, which governed the elections of December 2003, the winning candidate in a single-member constituency was again required to secure more votes than had been cast against all of them (O vyborakh deputatov Gosudarstvennoi 2002, art. 83.2). In the 2005 law, which was to govern the Duma election of December 2007, there was no longer a facility to vote against all the candidates in single-member constituencies as the contest had been converted into one that was exclusively among party lists. It was still possible to vote against all the party lists, however, although it would continue to have no direct implications for the validity of the election as a whole unless—improbably—it secured such a proportion of the vote that all the parties among them were left with less than 60% of the total (O vyborakh deputatov Gosudarstvennoi 2005, arts 73.5, 82.4). There had been suggestions in discussion of the new law that ‘against all’ might disappear entirely, and it did the following year; in the meantime it was agreed that individual regions should be allowed to eliminate the provision if they wished, but it remained obligatory at the federal level.

Individual laws have also been produced for the periodic elections that take place to the Russian presidency. In the law of 1995 that governed the election of the following year voters could cast a ballot ‘against all the candidates’ (art. 50), but this had no direct bearing on the result; the winner, as before, was the candidate who secured at least 50% of the vote, provided at least 50% of the electorate had taken part (art. 55). If a runoff election was necessary, however, the successful candidate was required to secure more votes than had been cast against both candidates (O vyborakh Prezidenta Rossiiskoi 1995, art. 56). The 1999 presidential election law extended the same principle to the first round, reflecting the framework law that had been adopted two years earlier. As before, the successful candidate was the one who had secured at least half the votes cast; this time, however, the election would be considered invalid if more votes had been cast ‘against all’ than in favour of the candidate that had secured the largest number of votes in either round of the contest (O vyborakh Prezidenta Rossiiskoi 2000, arts 72.3, 73.4). The same principles were followed in the election law that was adopted at the start of 2003, which was to govern the presidential election that would take place the following year (O vyborakh Prezidenta Rossiiskoi 2003, arts 76.4, 77.4). The entire sequence of changes is set out in Table 1.

TABLE 1 Voting‘Against All’and Russian Election Legislation

Election law	‘Against all lists’ facility?	‘Against all candidates’ facility?	Affects result?
1991 presidential	–	Yes	No
1993 Duma ^a	Yes	Yes	No
1994 framework	Yes	Yes	No
1995 Duma	Yes	Yes	No
1995 presidential	–	Yes	No^b
1997 framework	Yes	Yes	Yes
1999 Duma	Yes	Yes	Yes
1999 presidential	–	Yes	Yes
2002 framework	Yes	Yes	Yes
2002 Duma	Yes	Yes	Yes
2003 presidential	–	Yes^c	Yes^c
2005 Duma	No^d	–^e	No^d
Notes: The 1991 presidential law applied to the RSFSR, subsequent laws to the Russian Federation; the 1993 election was governed by a ‘regulation’ not a formal law. Dates refer to the year of adoption of the relevant legislation.
^aThe same principles applied to the election of the Federation Council.
^bSuccessful candidates had to win more votes than those cast against all in runoffs.
^cThis provision applied to the presidential election of 2004, but was removed from the law by amendment in July 2006.
^dOriginally included, but removed by amendment in July 2006; it remains the case that all the parties must collectively secure at least 60% of the (valid and invalid) vote, otherwise the result is void.
^eFrom 2005 onwards there has been no provision for single-member constituencies.

Voting against all and the Russian electorate

Russians have certainly shown no reluctance to use the facility they have been granted, in various forms, to vote against all candidates and parties. At least three trends are clearly apparent (see Table 2). First of all, the propensity to vote against all has steadily increased at every election since 1995; the election before that took place in unusual circumstances, after the Russian parliament had been unconstitutionally dissolved and its leadership arrested, but even so the vote against all party lists and presidential candidates has never been higher than at the most recent elections, in 2003 and 2004. A second pattern has been apparent at every election since the outset, which is the tendency for the vote ‘against all’ in the single-member constituencies to be much larger than the vote ‘against all’ in the parallel party-list contest. A third trend has also been apparent, which is the tendency for the vote against all candidates in presidential elections to be lower than the vote against all single-member constituency candidates or party lists, at least in the first round of voting. At their highest point, in the single-member constituencies in December 2003, nearly 13% voted against all the other candidates, party-sponsored or independent; and this put ‘against all’ in second place, behind only the Kremlin-sponsored party, United Russia (Edinaya Rossiya).

TABLE 2 The Vote‘Against All’, 1991 – 2004 (percentage of votes cast)

	Duma party list	Duma single member seats	Presidential election
1991			1.9
1993	3.9	14.8
1995	2.8	9.6
1996			1.5 (I), 4.8 (II)
1999	3.3	11.6
2000			1.9
2003	4.7	12.9
2004			3.5
Sources: For 1991, Pravda, 20 June 1991, p. 1; for 1993, T. F. Remington, Machine readable file obtained from the Central Electoral Commission, 1993, used with permission; for other years, compiled from constituency data in Vybory deputatov (1996, 2000, 2004) and Vybory Prezidenta (1996, 2000, 2004). The 1996 presidential election figures show the results in the first (I) and second (II) rounds.

A vote of this kind averaged across the entire country meant that ‘against all’ could sometimes achieve a much larger proportion of the vote in individual constituencies; indeed it was sometimes the largest vote of all. In 1993, there was a bigger vote ‘against all’ than for the most successful candidate in 29 of the 225 single-member constituencies (Remington 1993). In the 1995 Duma election, which was conducted without the same restrictions on party and media activity, ‘against all’ came first in three constituencies, although again it had no direct consequence for the result and the most successful candidates were duly returned (Vybory deputatov 1996, pp. 178, 182). In 1999 ‘against all’ won eight constituencies, but this time it was a result that invalidated the outcome and new elections had to be held to fill the vacancies (Vybory deputatov 2000, p. 181). By the following Duma election, in December 2003, ‘against all’ had moved ahead of all but four of the 23 parties in the national party-list contest and it ‘won’ three of the single-member seats, taking as much as 26% of the vote in one Moscow constituency (Vybory deputatov 2004, pp. 192, 194). An enterprising newspaper even published an interview with the ‘leader of the not yet established “Against All Party”’, who was ‘already preparing for the 2007 elections’ with every expectation of a ‘great future’ (Zaitseva 2003). 8

New elections to fill the three vacancies took place in March 2004, and in two of them a deputy was finally elected. But in the Volga city of Ul’yanovsk, where ‘against all’ had led the poll in December 2003, it came first once again and even increased its notional ‘majority’. 9 Focus group discussions suggested that this non-compliant behaviour was related more than anything else to social conditions in the area, and to the failings of the city authorities. 10 According to local survey evidence, 41% of respondents confirmed that voting ‘against all’ was most likely to be inspired by the ‘crisis in the economy of the town or region’ and the ‘unsatisfactory performance of the local authorities’, while 32% suggested ‘poor living standards’. 11 Voting ‘against all’ could reach even higher levels, especially when a popular candidate was removed by what appeared to be dishonest means: the vote ‘against all’ reached 40.2% at an election to a Duma vacancy in Yekaterinburg at the end of the 1990s (Il’ichev 2007), and as many as 69.9% rejected all the candidates in a gubernatorial election in Krasnodar krai in December 2004. 12

Voting against all: the Russian debate

As one of the Russian papers remarked in July 2004, ‘something like a public debate [was] gradually emerging’ about electoral reform, and particularly about the retention of the provision for voting against all. 13 It was a debate that continued up to the adoption of a new version of the electoral law, a year later, and only came to an end after an amendment to the new law in the summer of 2006 that removed the ‘against all’ provision. This was introduced after the Constitutional Court had ruled in a controversial judgement that there could not only be voting but also campaigning in favour of ‘against all’, using private funds in a manner that appeared impossible to regulate. The amendment was an unpopular decision, and possibly not a final one, in that it might successfully be argued in the courts that the constitutional right of citizens to express themselves politically had been improperly restricted; and it was still theoretically possible for an election to be invalidated if more than 40% of voters cast an invalid ballot. But a much wider range of issues was raised in the discussion, with pro-Kremlin parties broadly favourable to the amendment and oppositional parties broadly negative, with considerable uncertainty on all sides about the likely implications of such a change.

Those who argued in favour of removing the ‘against all’ facility included Igor’ Bunin, director of a prominent Moscow think tank. For Bunin, the provision was an ‘anachronism, a product of the late 1980s’. Back then, he explained, elections were between two candidates, both from the Communist Party, and people had to be given the option of not voting for either one. Today that opinion has become harmful: it dissipates positive voting and turns it into something negative, the negative of a civic position. It prevents society from uniting around any one candidate. It would be better if voters who want to register a protest vote for a candidate, even a populist candidate, the way some do for [the right-wing nationalist Vladimir] Zhirinovsky. 14 If Russians wanted to express their dissatisfaction with the candidates, Bunin suggested, they could do so in the same way that Europeans and Americans did—by staying at home, tearing up their ballots, or crossing out all the names. 15 International observers took a similar view; the OSCE monitoring team, at least, recommended the removal of the ‘against all’ facility in their report after the December 2003 election, on the ground that it appeared ‘impossible to take into account the political will of the “against all” voters in the proportional race’ (OSCE 2004a, p. 27). Their report on the presidential election of 2004 made a similar recommendation (OSCE 2004b, pp. 11 – 12, 27).

Bunin returned to the subject in the summer of 2006, while the Duma was considering the Kremlin-sponsored amendment that would remove the ‘against all’ facility. Repeating his earlier arguments, he suggested that the ‘against all’ provision was just a ‘sign of political immaturity’, a hangover from the late Soviet period, when many local party leaders had put themselves forward without competitors, and from elections in which there had only been an opportunity to vote for or against a single approved candidate. It was ‘not an accident’ that provisions of this kind were mostly to be found in the former Soviet republics, and almost unknown in Western democracies. Nevada (since 1976) and Massachusetts (since 2004) admittedly allowed voters to choose ‘none of the above’, but these were hardly precedents for an entire country like Russia. In any case an ‘against all’ facility made most sense in the single-member constituencies, which had been abolished by the new election law; and it would take votes away from parties that might otherwise reach the threshold, which would advantage the big battalions. If electors were dissatisfied, Bunin suggested, it would be ‘more honest’ of them to stay at home rather than to express their misgivings in this underhand way. 16

Supporters of the provision were more numerous, and they included members of the Central Electoral Commission itself. ‘You need to give [electors] a way of registering their protest’, explained Elena Dubrovina, a Commission member: ‘Voting for “none of the above” is a signal to those in power that something has to change’. 17 The chairman of the Commission, Alexander Veshnyakov, argued similarly that voting ‘against all’ was a way of showing the authorities that the situation in a local area was ‘anomalous’. It took place most often when popular candidates were arbitrarily excluded from the contest, or when the campaign had been particularly dirty, or where there had been large-scale use of ‘administrative resource’ (in other words, the power of office). In this sense it was a ‘useful’ phenomenon, as it ‘identified problems that had to be addressed’. 18 Veshnyakov described the vote against all, in other interviews, as a ‘litmus paper’: it was generally negligible if the authorities were not abusing their position, but a high level of voting against all—as had taken place in many of the Russian regions in the spring of 2006—was a ‘signal to government’ that something was wrong. 19

There were other political figures, not just among the opposition, who were also in favour of retention. Removing ‘against all’, argued Sergei Mironov, the normally loyal speaker of the upper house and leader of the Party of Life (Partiya zhizni), would be a ‘major strategic and systemic mistake’. The facility to vote against all was a ‘kind of thermometer of the political health of the society’; if it showed the patient's temperature was rising, what sensible doctor would throw it out the window? 20 Oppositional parties were even more forthright. For the liberal Yabloko party, removing ‘against all’ would be a ‘further step towards the liquidation of free elections’; for the Communists, it was a violation of the constitutional rights of citizens and an attack on democracy itself. 21 One of the most impassioned attacks came from opposition deputy Vladimir Ryzhkov, who saw these developments as part of a larger process by which first of all the other parties were denied access to the ballot, then the candidates were limited to those that had the approval of the authorities, and finally electors were deprived of the ability to vote against them. 22

Apart from party politics, the removal of ‘against all’ was likely to have more broadly systemic consequences. If voters were denied this form of expression, it was argued, their political energies might find other, not necessarily peaceful or legal outlets. The extent to which Moscow voters in the city election of December 2005 had spoiled their ballot papers—more than 5% as compared with the normal figure of between 1% and 2%—was an early indication of the forms that civic disaffection might take. 23 An electoral system that denied this form of political self-expression was also a system that would be unable to convey the real state of affairs to the governing authorities, depriving them of the ‘feedback loop’ that connected them with all aspects of the life of their society. 24 And if levels of turnout were driven down even further, the legitimacy of elected institutions might increasingly be called into question. 25 For some contributors to the discussion, the ‘coloured revolutions’ in other former Soviet republics were an indication of the kind of development that could take place if governments lost touch with their electorates, and if mass publics were left with no other way in which they could communicate their views. 26

Russians themselves were fairly evenly divided about whether they approved or disapproved of voting against all, although the survey evidence suggested that there was increasing support for the facility. In a national poll in November 2005, 38% approved of those who exercised this option, 39% disapproved and 23% had no opinion; but this was a higher level of approval than had been recorded in earlier years. Why, interviewers asked, did ‘some people’ vote against all candidates or parties? For a majority (51%), it was because there was ‘no good candidate to vote for’, and for nearly a third (31%), it was to ‘voice their protest’. Another 19% were concerned that their ‘uncompleted ballot [would] be used to rig the election result’, 18% were ‘not interested in politics’, and 17% complained of a ‘lack of information about the candidates’ (voting ‘Against All' 2005). Had they ever voted against all themselves? In October 2004, 14% said they had done so; a year later, 23% reported they had done so; and by the summer of 2006, 31% said they had done so. If nothing else, this was certainly an indication that the option was ‘rapidly becom[ing] fashionable’ (Kertman 2006).

In the end, as noted above, it was a ruling by the Constitutional Court that was decisive. A Mr Bochkov in the provincial city of Kursk had been fined for distributing an appeal to other citizens to vote ‘against all’, rather than in favour of the parties and candidates that were on the ballot paper. He had funded the appeal from his own resources, rather than from one of the special accounts that were set up at election time for the use of parties and candidates, and which allowed their spending to be monitored. Bochkov complained that his constitutional rights had been infringed, he was supported by the Commissioner for Human Rights, and the Court found in his favour. 27 But this allowed private and unregulated funds to enter the electoral contest, and raised the possibility that a well-financed campaign might encourage so many to vote against all—for instance in March 2008, when Putin's successor was being chosen—that the result would be invalid. It was almost impossible to conceive of a solution within the existing law; the Constitutional Court might change its position at some point in the future, but meanwhile ‘no column [in which to vote against all]—no problem’. 28

The presidential draft of the new electoral law, presented to the Duma in December 2004, had not originally made provision for a vote ‘against all’; its main focus was a series of changes that eliminated the single-member constituencies and raised the threshold for party-list representation from 5 to 7%. 29 However, the second reading on 15 April 2005 restored the ‘against all’ provision, but at the federal level only; at lower levels, the regions would be free to decide for themselves. The third and final reading, on 22 April 2005, retained this formulation, and it was signed by the president in this form on 18 May. 30 Reflecting after the law had been adopted, Veshnyakov accepted that ‘against all’ allowed voters to avoid the burden of choice, but insisted that it was still a useful provision as it drew the attention of government and of the wider society to regions in which there were ‘anomalies’: where there had been abuses of power, where there were no worthwhile candidates, and where there was a lack of meaningful choice. It was, again, a ‘sort of litmus paper, an indicator of the lack of wellbeing in one region or another of the Russian Federation’. As for the regions themselves, ‘let them experiment’. 31

The Constitutional Court's decision however, led to a further reconsideration of the legislation, and a year later an amendment to the law removed ‘against all’ from elections at all levels. It was an opportunity to rehearse now-familiar arguments. Boris Gryzlov, speaker of the Duma and a Kremlin loyalist, claimed that removing the provision would ‘raise voters’ electoral responsibility’. 32 The committee on constitutional legislation, which was responsible for the detailed evaluation of proposals of this kind, argued for its part that voting ‘against all’ was more or less unknown in other countries; 33 and United Russia spokesmen insisted that the change was simply bringing Russian legislation into line with ‘European democratic norms’. 34 A first reading was approved by a large majority on 9 June, a second on 28 June, and a third and final reading on 30 June; it was approved by the upper house and signed by the president on 12 July 2006. 35 The law, accordingly, had changed once again. Even so, it was not yet certain that it would be a final change, as individual citizens might wish to appeal to the Constitutional Court on the grounds that they were being denied their right of political self-expression. 36 And public opinion was certainly in favour of a return to earlier practices. An April 2007 survey for the Levada Centre found that 65% supported the restoration of the ‘against all’ option and just 25% were against it, with the remainder undecided (Levada Centre 2007); VTsIOM reported almost identical figures (Il’ichev 2007).

Nor was it entirely clear what the effects of the change would be. There were firm predictions, for instance, that turnout would fall, perhaps by 5 – 7%. 37 This might have had implications for the Duma election, which required a minimum turnout of 25%, and certainly for the next presidential election, for which a 50% turnout would have been necessary; but further legislation in December 2006 removed the minimum turnout requirement at all levels. 38 This left the question of the destination of the vote that would otherwise have been cast ‘against all’. Perhaps a third of those who had voted against all were expected not to vote at all. The pro-Kremlin ‘party of power’, United Russia, thought they would be the main beneficiaries of those who would otherwise have voted ‘against all’ but who would now vote for one of the parties; others thought it was more likely that what was in effect a protest vote would gravitate towards the opposition. 39 But others still predicted that half of those who had voted against all would not take part in a future election, and that United Russia and right-wing nationalists would both make modest gains among those who did. 40 There was yet another, more cynical view, which was that as the results would inevitably be falsified, the removal of ‘against all’ would have very little significance either way: the authorities would simply continue to ‘allocate the vote the way they want’. 41

Voters, nonvoters and against all voters

The limited literature on voting ‘against all’ has so far concentrated almost entirely on regional differences, using aggregate rather than individual-level data. Perhaps the most widely shared conclusion is that the level of voting ‘against all’ varies considerably and consistently by the different levels of the federation, and that several factors are particularly important: constitutional status (there is less ‘against all’ voting in the republics, which have greater rights than ordinary regions and which accordingly give local elites a greater opportunity to minimise anti-regime behaviour); the share of the indigenous population that is Russian (there is generally less voting ‘against all’ in regions that have larger proportions of non-Russians); and the nature of the settlement (there is generally more ‘against all’ voting in urban than in rural areas). These factors, in turn, are often related to each other: republics, for instance, tend to have larger non-Russian populations, and non-Russians are also more likely to live outside the major conurbations (Akhremenko 2001; Akhremenko & Meleshkina 2002; Lyubarev 2003; Hutcheson 2004).

Individual-level evidence has helped to refine some of these associations and identify other vote determinants, but it also yielded rather contradictory conclusions. In the first-ever Duma elections, in December 1993, it was older age and lower social status that made a difference. Ten years later, on the evidence of an exit poll in single-member districts, the picture was rather different: this time it was younger voters, and the better educated, and students rather than pensioners who were more likely to be ‘against all’ (Mickiewicz 2006, pp. 19 – 21). The Public Opinion Foundation, in a 2005 survey, found a slightly greater propensity for males to vote ‘against all’ than females (24% compared with 21%); but this time, it was the middle-aged (26%) who were the most likely at some point to have voted against all, as compared with younger (22%) or older respondents (17%). Income made no difference; but once again, those with a higher education (34%) were much more likely to have voted against all than those with no more than a primary education (16%). Those who lived in the biggest cities (33%) were also more likely to have voted against all than those who lived in the countryside (18%) (voting ‘Against All' 2005).

We set out our own evidence of the socioeconomic characteristics of the various groups of voters in Table 3, based on post-election surveys in 2004 and 2005 respectively. 42 Voters, on this evidence, are older than nonvoters, better educated and disproportionately female—all of which is in line with the comparative literature (for instance Franklin 2004; Wattenberg 2006). Our primary focus is however on the comparisons between voters in general and ‘against all’ voters in particular, and then secondarily between ‘against all’ voters in the single-member constituencies and their counterparts in the party-list election to the other half of the Duma. ‘Against all’ voters, it emerges, are consistently different from other voters on three criteria across both surveys: age, place of residence and labour force participation. ‘Against all’ voters, on average, are in their early forties, about five years younger than the average for other voters. They are also more likely to live in urban areas and to be labour force participants: 59% of ‘against all’ voters in the presidential election, for instance, were labour force participants, compared with 48% of other voters; and they were also more likely to be urban residents, and to have a higher education—taken together, the characteristics of middle class protest rather than of marginality and alienation.

TABLE 3 Socioeconomic Characteristics of Groups of Voters, 2003 – 2004

	Against all
	All	(SMD)	(List)	Voters	Nonvoters	(Total)
	2003 Duma election
Gender (% male)	48	(47)	(43)	43	51	(46)
Age (mean years)	43**	(43)	(41)	47	38	(45)
Urban (%)	82**	(82)	(89)	73	78	(75)
Ethnic Russian (%)	88	(88)	(92)	89	90	(89)
Married (%)	60	(61)	(55)	60	57	(59)
Tertiary education (%)	20	(20)	(19)	17	8	(15)
Labour force participation (%)	52*	(48)	(59)	44	53	(47)
(N)	(264)	(234)	(88)	(1,000)	(348)	(1,612)
	2004 presidential election
Gender (% male)	52			44	55	46
Age (mean years)	41**			46	37	45
Urban (%)	85**			73	79	75
Ethnic Russian (%)	87			89	90	89
Married (%)	60			62	46	60
Tertiary education (%)	22**			16	10	15
Labour force participation (%)	59**			48	42	48
(N)	(98)			(1,552)	(289)	(1,939)
Note: **Against all voters statistically different from other voters at p < 0.01, *p < 0.05, both two-tailed.
Sources: Authors' 2004 and 2005 surveys (see fn 42 for further details).

The 2003 Duma election offered voters two opportunities to vote ‘against all’: in the single-member constituencies, where 12.9% did so, and in the national party-list contest, where 4.7% did so. Voters and nonvoters were consistent in both parts of the ballot: Table 4 shows that of the total sample, 21% reported abstaining in both the single-member constituency and party-list ballots, and 61% reported voting in both. Another 4% voted ‘against all’ in both ballots; more interestingly, 8% of those who voted ‘against all’ in the single-member constituencies cast a valid vote in the party-list ballot, but hardly any the other way round. The reason is likely to be that individual characteristics are of more concern to voters in the single-member constituencies, where the candidates may be known personally or at least by reputation and where the outcome can evidently be affected by a substantial vote against all of them; whereas a national party list is a more generalised option in which the outcome is most unlikely to be affected by a rejection of all the parties on the ballot. We found some confirmation of these hypotheses within our surveys: those who said they had voted ‘against all’ in the 2004 presidential election, for instance, were substantially more likely to report that single-member constituency elections were a more effective means of exercising political influence than the parallel party-list contest.

TABLE 4 Against All in SMC and List Voting, 2003 (%)

	List vote
	Nonvoter	Against all	Other voter
SMC vote
Nonvoter	21	1	2
Against all	2	4	8
Other voter	1	0.4	61
Source: 2004 survey; n = 1,634 respondents.

Explaining ‘against all’ voting

Studies of electoral participation have generally suggested that the individual decision to take part is a balance among three sets of factors. 43 First, there are the resources required to engage in the election—such as the cognitive resources that come from education, and the socioeconomic resources that come from income and wealth, all of which provide the political skills with which to make an informed choice. Second, there is the perceived relevance of the vote to the outcome of the election, which is a function of the political choices on offer and the degree of political competition (Blais 2001; Wattenberg 2002). And third, there are generational effects that have been shown to be important, with a steady decline in political interest and participation among the young (Wattenberg 2006). Since we are primarily interested in ‘against all’, which is a form of voting, rather than voting as such, we hypothesise that resources will be a less important determinant than either generation or electoral choices.

These hypotheses are tested in Table 5 in the form of a multinomial logistic regression model which reports the results of three sets of contrasts between against all voters, other voters and nonvoters, respectively. In each case, the model uses a range of variables to predict voting ‘against all’, but uses a different comparator. In the first set of results, the comparator is nonvoters; in the second, it is other voters; and in the third, voters in general are compared with nonvoters. Since our hypothesis is that ‘against all’ voters are essentially protest voters, rather than voters who are alienated from the political system as a whole, the more important comparison occurs in the second set of results. The independent variables are divided into two groups: the individual resources that underpin the act of participation, and the beliefs that voters hold about political institutions. Based on the discussion above, we would expect the resources the individual possesses to be the most important single factor in distinguishing ‘against all’ voters from nonvoters, and beliefs about political institutions to be the main predictor in distinguishing them from those who vote for the candidates and parties that appear on the ballot.

TABLE 5 Resource and Belief Explanations for‘Against All’ Voting, 2003 (multinomial logistic regression estimates)

	Against all versus nonvoters		Against all versus other voters		Voters versus nonvoters
	Est	(SE)	Est	(SE)	Est	(SE)
Resources
Gender	−0.032	(0.191)	0.187	(0.160)	−0.219	(0.152)
Age	0.033**	(0.007)	−0.015**	(0.006)	0.048**	(0.006)
Urban resident	0.060	(0.242)	0.404*	(0.201)	−0.344*	(0.185)
Ethnic Russian	0.069	(0.302)	−0.082	(0.250)	0.151	(0.244)
Tertiary education	1.014**	(0.299)	−0.049	(0.205)	1.063**	(0.261)
Married	0.045	(0.197)	−0.035	(0.164)	0.080	(0.158)
Labour force participant	0.049	(0.198)	0.119	(0.168)	−0.070	(0.157)
Beliefs about institutions
Satisfied with things in Russia	0.179*	(0.072)	−0.025	(0.060)	0.204**	(0.058)
Russia not close to rule of law	−0.027	(0.120)	0.363**	(0.100)	−0.380**	(0.096)
Corruption increased since Soviet times	0.070	(0.114)	0.131	(0.095)	−0.061	(0.093)
Prefers private versus state control of economy	−0.086*	(0.038)	−0.034	(0.032)	−0.052	(0.030)
Constant	−2.147		−1.497		−.650
Nagelkerke R-sq			0.15
(N)			(1,284)
Notes: **Statistically significant at p < 0.01, *p < 0.05, both two-tailed.
Multinomial logistic regression analyses showing parameter estimates and (in parentheses) standard errors predicting three sets of contrasts between against all, other voters and nonvoters.
Question wordings were ‘To what extent are you satisfied with how things are going at the moment in Russia?’; ‘How close do you think the idea of the rule of law is to the Russian government?’; ‘Compared with Soviet times, would you say the level of bribery and corruption has [somewhat or greatly] increased?’; ‘With which of these views are you inclined to agree? An enterprise is always better managed by a private entrepreneur or state ownership is the best means of managing an enterprise’.
Source: 2004 survey.

The results that are presented in Table 5 largely support these predictions. In the first set of results, contrasting ‘against all’ voters with nonvoters, the predominant effects are indeed associated with the resources that attach to individuals. Thus, ‘against all’ voters are older than nonvoters, and much more likely to have a university education. These are results that are consistent with most of the recent literature dealing with turnout, and underpin the importance of cognitive skills and experience in taking the decision to vote but doing so in a manner that repudiates all the parties on the ballot. There are more modest effects for beliefs about institutions, although ‘against all’ voters are somewhat more likely than nonvoters to be satisfied with the current state of affairs in Russia, and to support state rather than private control of the economy. In general, however, resources are more important than beliefs in identifying ‘against all’ voters and distinguishing them from nonvoters.

In the second set of results, which distinguishes ‘against all’ from other voters, there are more substantial effects for beliefs about the political system than resources, which is as we had hypothesised. Here, ‘against all’ voters are much more likely than other voters to believe that Russia is still some distance from the rule of law. Compared with others, the motivation for ‘against all’ voting is clearly a protest against what is regarded as the absence of an effective legal framework, rather than an objection to the political institutions that presently exist. The final set of results contrasts voters with nonvoters. The major factor here—indeed, more important than the other variables combined—is age, a finding that is consistent with the literature on turnout we have already cited. Tertiary education is second in importance, followed by attitudes—specifically, those who are more satisfied with conditions in Russia and more inclined to believe it is close to the rule of law are also more likely to vote.

Democracy, parties and ‘against all’

The evidence we have so far considered shows that ‘against all’ voting in Russia is related to resources, and that there is also a strong link to a general sense of disenchantment with the current state of Russian politics. However, the disenchantment of ‘against all’ voters is not with the principle of competitive elections and the political system in general. This is apparent from the responses to a series of further questions about the direction in which the country is moving, and preferred types of political system. Table 6 shows that ‘against all’ and other voters were identical in their belief that Russia was moving towards democracy—27% in each case. Nonvoters were less sure, with 22% in agreement. ‘Against all’ voters, moreover, were significantly more likely than other voters to believe that the most appropriate political system for their country was Western-style democracy: one in four took this view, compared with 15% of other voters and 23% of nonvoters.

TABLE 6 Views of Democracy and the 2003 Election

Percent who:	Against all	Other voters	Nonvoters	(Total)
Believe country moving to democracy	27	27	22	(22)
Believe most acceptable system is Western democracy	25**	15	23	(23)
Are satisfied with democracy in Russia	20**	30	26	(28)
(N)	(264)	(1,000)	(348)	(1,687)
Note: **Against all statistically different from voters at p < 0.01, *p < 0.05, both two-tailed.
Source: 2004 survey.
Question wordings were: ‘In your opinion, the direction our country is moving characterises it as … Option: Moving consistently towards the consolidation of democracy’; ‘Which of the following political systems would be the most appropriate for Russia? Option: Democracy of the Western kind’; ‘How satisfied are you with the level of democracy in Russia?’.

Apart from its association with a positive view of Western democracy and a negative view of Russian democratic practice, did ‘against all’ have any direct implications for party choices and electoral outcomes, at least until the option was removed in 2006? The latter is especially important since, as we have seen, the vote ‘against all’ has consistently exceeded the vote for most of the parties in both the party list and the single-member contests. In such a context, a relatively small bias on the part of former ‘against all’ voters could potentially have had a significant effect on the outcome of future elections, and might do so again if the option were eventually restored. The results in Table 7 suggest that there are in fact few differences between the three groups of respondents in these respects. Predictably, nonvoters are the least interested in politics—almost three quarters express no interest at all, with half of the remainder opting for the political centre. On the more important distinction, between ‘against all’ and other voters, the former are less likely to be interested in politics, but in both cases about a quarter opt for the political centre and the remainder divide almost equally between left and right.

TABLE 7 Political Orientations and Voting Behaviour, 2003

	Against all	Other voters	Nonvoters	(All)
Political orientation
Left	12	19	6	(6)
Centre	21	33	13	(13)
Right	12	16	7	(7)
Not interested in politics	55	32	74	(74)
Total	100	100	100	(100)
(N)	(170)	(717)	(259)	(1,146)
	Chi sq = 146.5, df = 6, p = <0.000
Party usually support
Communist party	16	27	14	(22)
Socialist party	6	12	5	(10)
Party supporting market economy	23	38	18	(31)
Nationalist party	2	4	2	(4)
Party supporting national minorities	1	1	0.3	(1)
Green/ecological party	3	2	3	(2)
Other	6	13	4	(10)
None	43	3	54	(20)
Total	100	100	100	(100)
(N)	(188)	(809)	(288)	(1,285)
	Chi sq = 437.9, df = 14, p = <0.000
Source: 2004 survey.

The second part of Table 7 examines the party that respondents said they usually supported, conceived in terms of ‘families’ of parties rather than the changing set of parties that contest particular elections. Not surprisingly, nearly half of all ‘against all’ voters find it impossible to identify a party they normally support; this compares with just 3% among other voters, although it is slightly less than among nonvoters. But among ‘against all’ voters who do support a party, about half choose one that favours the market economy, followed by a communist party—a breakdown very similar to that among more conventional voters. These results confirm our earlier hypothesis that ‘against all’ voters are a distinctive group: more in favour of Western democracy than other voters, but more dissatisfied with the political system that presently exists; they are similar to others in their self-placement among the political philosophies, but unable to find a means of expressing their opinions through the political parties that are actually on offer.

Conclusions

‘Against all’ has been a significant and increasing presence not only in Russian elections, up to 2006, but also in the other post-Soviet republics. Bondar’ and Dzhagaryan note the presence of this provision in the electoral codes of Armenia, Azerbaijan and Belarus (2005, pp. 168 – 69). It is also a part of electoral practice in Ukraine, although it has no direct implications for the result (Zakon Ukraïny 2005); and the right to vote against all candidates or lists of candidates was included in the convention on democratic electoral standards that was adopted by the member countries of the Commonwealth of Independent States in 2002, as one of the elements that allowed voters to express their views directly (Konventsiya 2002, art. 4.1). On our evidence, it has operated in practice as a protest vote, with many of the same characteristics as the vote for a third party in a Western democracy. ‘Against all’ voters, at least in Russia, are not hostile to competitive politics; indeed, they are more committed to Western democracy than those who vote for a party or candidate. But they are much more likely to believe the Russian government falls short of those democratic standards.

Removing the ‘against all’ facility, in these circumstances, is likely to have a number of consequences. Turnout, for a start, may fall even further, although increasing numbers are likely to take part and then spoil their ballot paper as a means of expressing generalised dissatisfaction. Another form of protest that would become more common, to judge from the Moscow regional elections in late 2006 at which there had been no ‘against all’ option, was demonstratively to hand the ballot paper back to election officials. 44 Less probable, given the Kremlin's control of the electoral arena, is the emergence of a party that bids for the support of those who are prepared to vote but who have found no political home among the parties that are presently on offer. In the absence of an improvement on the supply side, there is likely to be a further decline in support for electoral politics and some increase in the support of social movements and other agencies of political influence that operate outside the parliamentary system altogether. An electoral system that allows fewer choices is one that can more easily be controlled from above; but it is one in which citizens are less likely to engage and which is less likely to sustain their support.

The authoritarian turn in Russian politics since the accession of Vladimir Putin has imposed new choices on all Russian voters, including those who might otherwise have voted ‘against all’. Across our entire 2005 survey, a plurality approved of the move to a Duma that since December 2007 has been exclusively composed of party-list deputies (39% were in favour, 27% against). Among ‘against all’ voters, however, the proportions were reversed: 28% were in favour of the change, but 53% opposed. Across the entire sample, a majority (59%) approved of the related move away from directly elected governors; but once again, ‘against all’ voters were less supportive—just 44% approved of the change, with nearly as many (40%) against it. Russia's ‘against alls’ are young, urban, and highly educated. They like Western democracy but are dissatisfied with the political system in which they find themselves, particularly its lack of a rule of law. They have been unable to express their views in the past through the political parties, and they oppose the changes in the electoral system that will restrict future parliamentary contests to a small number of Kremlin-sponsored groupings. On this basis, they will have even more to protest about in the future.

Notes

¹ Pravda, 5 March 1989, p. 1.

²The Central Electoral Commission took the proportion of electors who had ‘taken part in the election’ to be defined as those receiving a ballot paper. Turnout had to reach a minimum of 25% in Duma and 50% in presidential elections. In every election, however, substantial numbers—in December 2003, nearly 80,000—took possession of a ballot paper but did not cast a vote of any kind (Vybory deputatov 2004, p. 106).

³Fuller details are provided in footnote 42. We acknowledge the support of the UK Economic and Social Research Council under grant RES-000-23-0146, the assistance of Julia Korosteleva, the cooperation of Derek Hutcheson, and the pioneering contribution of Oversloot et al. (2002), who were the first to draw attention to this distinctive form of voting in the Western-language literature.

⁴ Izvestiya, 5 April 1989, p. 1.

⁵ Pravda, 4 March 1990, p. 2; Pravda, 23 January 1990, p. 1.

⁶ Pravda, 20 June 1991, p. 1.

⁷These provisions of the 1997 law were challenged by the Federation Council; the Constitutional Court, however, ruled that they were constitutionally proper (Sobranie zakonodatel’stva Rossiiskoi Federatsii 25, 1998, art. 3002, 10 June 1998).

⁸There had in fact been public suggestions that ‘against all’ voters should form an ‘oppositional democratic party of a new type’ (Gushchin 2003a, p. 3) or at least an ‘against all’ electoral bloc that could compete at general elections: not to win seats, but to ‘struggle against a government that is adopting unjust laws’ (Gushchin 2003b, p. 3; similarly Vremya MN, 12 August 2003, p. 3).

⁹ Moscow Times, 24 March 2004, pp. 1, 4.

¹⁰Typical complaints were that the streets were dark, there was dirt everywhere, and payments for utilities kept increasing. The discussions took place at the end of March and the beginning of April 2004 (see White 2005, p. 1,141).

¹¹ Moskovskie novosti, 26 November 2004, p. 9.

¹² Izvestiya, 7 December 2004, p. 3.

¹³ Vremya novostei, 16 July 2004, p. 4.

¹⁴ Vremya novostei, 16 July 2004, p. 4.

¹⁵ Vremya novostei, 16 July 2004, p. 4.

¹⁶ Novye izvestiya, 28 June 2006, p. 2.

¹⁷ Izvestiya, 7 October 2004, p. 4.

¹⁸ Izvestiya, 7 October 2004, p. 4.

¹⁹ Moskovskie novosti, 16 June 2006, p. 11.

²⁰ Vedomosti, 29 June 2006, p. A6.

²¹ Sankt-Peterburgskie vedomosti, 24 June 2006, p. 4.

²² Izvestiya, 7 June 2006, p. 2.

²³ Ekspert, 5 June 2006, p. 68.

²⁴ Kommersant-vlast’, 5 June 2006, p. 34.

²⁵ Nezavisimaya gazeta, 6 June 2006, p. 4.

²⁶ Kommersant-vlast’, 5 June 2006, p. 34.

²⁷ Nezavisimaya gazeta, 15 November 2005, pp. 1 – 2; the decision is reported in Postanovlenie (2005).

²⁸ Moskovskie novosti, 16 June 2006, p. 11.

²⁹ Novye izvestiya, 9 December 2004, p. 1.

³⁰ Rossiiskaya gazeta, 24 May 2005, pp. 22 – 32, art. 73.5.

³¹ Rossiiskaya gazeta, 26 July 2005, p. 9.

³² Komsomol’skaya Pravda, 10 June 2006, p. 2.

³³ Komsomol’skaya Pravda, 10 June 2006, p. 2.

³⁴ Kommersant, 29 June 2006, p. 8.

³⁵ Rossiiskaya gazeta, 15 July 2006, p. 7.

³⁶ Izvestiya, 19 June 2006, p. 2.

³⁷ Vedomosti, 13 June 2006, p. 2.

³⁸ Rossiiskaya gazeta, 7 December 2006, p. 6.

³⁹ Komsomol’skaya Pravda, 10 June 2006, p. 2.

⁴⁰ Izvestiya, 29 June 2006, p. 1; similarly Il’ichev (2007).

⁴¹ Kommersant-vlast’, 5 June 2006, p. 34.

⁴²Our surveys were conducted by Russian Research in association with the project on ‘Inclusion without Membership? Bringing Russia, Ukraine and Belarus closer to “Europe”’, which is funded by the UK Economic and Social Research Council to Stephen White, Margot Light and Roy Allison under grant RES-000-23-0146. Fieldwork took place between 21 December 2003 and 16 January 2004 and again between 23 March and 20 April 2005. The number of respondents in each case was 2,000, selected according to the agency's normal sampling procedures; it was representative of the population aged 18 and over, using a multistage proportional method with a random route method of selecting households. Interviews were conducted face to face in respondents' homes. The sample was then weighted in accordance with sex, age and education in each region. In both cases there were 97 sampling points, and 150 interviewers were employed; the agency's standard procedures were employed to check the completion of questionnaires and the logical consistency of the data. The entire dataset may be consulted at the UK Data Archive, where it has been deposited (FN 5671).

⁴³We have chosen in what follows to focus on the 2003 Duma rather than the 2004 presidential contest, in which just 3.5% voted against all; this also enhances the comparability of our findings across parliamentary elections cross-nationally.

⁴⁴ Argumenty i fakty, 1, 2007, p. 56.