At what point might the publicity that comes with holding office amount to undue influence at the next election?
Strakonice, a Czech town of around 23,000 people set in the lovely south Bohemian landscape, was known in the past for making fezes, firearms and motorcycles, and more recently its Dudák (Bagpiper) beer. The first president of Czechoslovakia, Tomáš Masaryk, was its representative in the Austrian imperial parliament in 1891–93. The current president of the Czech Republic, Miloš Zeman, won 56 percent of the vote there in January 2018, five points above his national share. But it was the local election ten months later that made headlines by raising hard questions of how the advantages of incumbency can be adjudicated, if at all.
Electoral litigation in the Czech Republic is plentiful owing to procedures that allow voters, candidates and parties to file complaints at little cost. Most are quickly dismissed, but a few result in nullifications and do-overs: at the national level, a Senate race was voided in 2016 by the Supreme Administrative Court, on the grounds that the winner was not the agreed nominee of a party torn into two bitterly opposing factions. A year later, the same court ordered a re-ranking of candidates elected from the Civic Democratic Party to the Chamber of Deputies, owing to clerical errors in tabulating preferential votes in 100 precincts. As a result of the decision, one incumbent deputy had to surrender his mandate.
As is obvious to anyone who has lived in a small town or seen a pastoral comedy like My Sweet Little Village (Vesničko má středisková, filmed about an hour up the road from Strakonice), there can be as much intrigue and scheming in the provinces as in the toughest city wards. Each round of Czech local elections produces several cases of vote buying, tampering or undue influence such as treating. As long as sufficient evidence is provided, these are judicially manageable questions of corruption and other irregularities that plainly break a law.
The Strakonice dispute is less easily judged, because at its crux is the claim that the party in power, a local movement called Strakonická Veřejnost (The Strakonice Public, hereafter SV), exploited its occupancy of town hall in ways that, while legal, gave it an unfair edge in the October 2018 election.
- Three months before the election, the town’s seven-person executive committee (of which four were from SV, including the mayor) forbade the posting of party promotional materials on 46 notice boards managed by the municipal cultural centre. The opposition parties assumed this meant they would have to pay for the use of commercial boards, of which there were only 11.
- However, the ban did not apply to other surfaces in public ownership, such as lamp posts and railings, but non-governing parties were apparently unaware of the right to use them under a 2004 ordinance. Only SV and one other local party requested permission to post on them.
- The town executive committee forbade partisan ads in the local free paper, Zpravodaj města Strakonice, during the month before the election. However, the paper did run a column by the mayor (and SV leader) claiming credit for the city’s robust fiscal health. Other parties had no chance to reply. The executive committee in 2017 had ordered Zpravodaj to carry no stories critical of SV’s record in office. Community television had likewise given exclusive, favourable attention to the mayor and his party.
SV went into the 2018 election with seven of the council’s 21 seats and came out with 16, on a vote share that rose from 28.9 percent to 51 percent. The second-place party, a regional movement called Jihočeši (South Bohemians) 2012, trailed far behind on 10 percent. Convinced that the pre-election environment caused them to underperform or fall short of the 5-percent threshold for seats, Jihočeši 2012 and other parties challenged the election at the regional court in České Budějovice.
At issue here are incumbency advantage and specifically “advertising” in the broader sense used by political scientist David Mayhew, as “any effort to disseminate one’s name among constituents in such a fashion as to create a favorable image but in messages having little or no issue content”. It relates to credit-claiming and attention-seeking outside of formal electioneering. To give one example of its potential power, an ingenious priming study estimated that the margin of victory in the reelection of an incumbent secretary of labor in North Carolina could be attributed almost wholly to the appearance of her photo and name in all elevators as the official responsible for their safety certification.[i] American courts have seen little reason to overturn elections for this; as the leading authorities on US election law argue, what is the harm in assuring officeholders that “their accomplishments in behalf of the public will become known and be rewarded politically, thus creating an incentive for high quality public service?”
The regional court in the case from Strakonice felt otherwise, and agreed with the complainants that their constitutional right to free party competition had been violated. It voided the election, and the interior ministry scheduled a new one to be held in March 2019.
Technically, there is no right to appeal a decision of a regional court in an election dispute. To get around that, SV and its leader, Mayor Břetislav Hrdlička, filed a complaint at the Constitutional Court alleging that the regional court, by reaching its decision without quantified evidence of the impact on voters of the usage of notice boards or local media coverage, had violated the winners’ rights to participate in competitive elections and take up their seats.
A panel of the Constitutional Court heard arguments from the lawyers for the two sides, Zdeněk Koudelka and Aleš Gerloch, both of them unsuccessful nominees for places on the Court in 2012 and 2019, respectively. Koudelka was also the architect of an attempted reform of the national electoral system that was struck down by the Constitutional Court in 2001 because it would have so disproportionately favoured large parties. On the brief for SV, however, Koudelka prevailed, as he had for previous clients accused of vote buying and treating. The Constitutional Court panel followed the standard test established by the Supreme Administrative Court (NSS) in 2004, which I summarized in an earlier article:
What made the judges even more uncomfortable with voiding the Strakonice election was a change in the wording of section 60 of Act 491/2001 on local elections, sponsored in 2016 by Social Democrat Jan Chvojka. Previously, the law allowed a petitioner to request that an election be voided if the law was violated in a way that “could have influenced” (mohl ovlivnit) the outcome; in the new wording, it must be said that a violation “grossly influenced” (hrubě ovlivnil) the results. The Court interpreted this amendment as setting a much higher evidentiary bar, which had not been met in this case. It sent the matter back to the regional court and suspended the order for a new election.
Showing that something “grossly” influenced a local Czech election is complicated by the hybrid nature of the local electoral system, combining
- At-large voting: the voter has as many votes as there are seats to fill, up to 21 in the case of Strakonice. In 2018, 9,035 voters cast ballots, with a potential total of 189,735 votes; almost all — 179,827 — were in fact allocated.
- Flexible party lists: parties offer a slate of candidates in a descending order chosen by the party, but candidates placed lower down can move to the top if they get 10 percent above the party’s average (the party’s total votes divided by the number of candidates slated). The eventual distribution of seats is determined by applying D’Hondt divisors to the total number of votes obtained by parties that clear the 5-percent threshold. (The potentially perverse mechanics of this process, which voters might not fully grasp, are set out by Lebeda; it resembles, with some adjustments, the regime for national elections.)
- Free-list panachage: voters can also distribute their votes across party lists. They could vote a straight ticket of one party and leave it at that, or they could distribute them for preferred candidates on any of the 14 lists slated in Strakonice in 2018, so long as no more than 21 candidates were chosen.
It was the panachage element that the Constitutional Court singled out in its rebuke of the regional court: there had been insufficient consideration of how many people would have to have been swayed by the purported undue influence to make a difference, for example, to prevent the ANO party of Prime Minister Andrej Babiš from making it over the 5-percent threshold. If voters were opting for straight party tickets, it might be as few as nine people, but if they were spreading their votes widely across lists, it could be 177. Czech political scientists have detected a trend of declining use of the opportunities afforded under panachage, meaning that more voters just work within, not across, party slates, which in turn means that fewer individuals would have to be swayed, but this empirical finding does not seem to have been factored into the legal proceedings.
In the Strakonice case the Constitutional Court was probably right to remand the matter back to regional court for reconsideration. It would also not be surprising if courts generally adopted an agnostic position on whether a manageable standard for judging lawful exercise of the spoils of incumbency exists (much as American courts punt on partisan gerrymandering). If judges and attorneys turned to academic literature for guidance, it could easily become an exercise in frustration: Incumbency is held up by political scientists as the most important variable in election outcomes, except by those who insist it does not matter at all. Add to this the distinct Czech dynamic of national parties competing alongside regional and local ones and loose combinations of independents for places on around 6,400 municipal and ward councils ranging in district magnitude from 5 to 70, and it might seem that no generalizable rule of thumb could be arrived at.
There are, however, some tools that could help courts make sense of the incumbency problem.
First of all, there is an established statistical method — regression discontinuity design (RDD) — to isolate the effect of incumbency from other variables, such as candidate quality or partisanship, by focusing on races decided by a narrow margin. This could be conducted on data from all local elections to determine whether, as a general rule, incumbency has any advantages in Czech politics. The finding could be adopted as a starting assumption for courts, which could then demand confirming or confounding evidence from the jurisdiction in question.
Judging by the voluminous literature on incumbency, several possible conclusions could emerge from a RDD analysis.
1. The incumbency advantage might in fact not be as great a fact of political life as assumed. Its apparent flourishing as an independent factor in the United States in the 1960s-80s was a product of the relatively brief heyday of ticket-splitting, which ended once realignment resulted in a new party-line voting that recoupled local and national loyalties and now neutralizes much of the benefit an incumbent receives from the use of office.
2. Incumbency might matter but to different degrees depending on the level of government, type of office or electoral system. In the United States, local elections are usually nonpartisan, thus depriving voters of the potent cue they rely upon for state and federal elections, and this could help incumbents on city councils; Czech municipal elections, however, have a partisan frame. There may be a spatial aspect of incumbency in a densely-populated, compact country like the Czech Republic, in which the average municipality has only 1,650 residents and 13 square kilometres of territory. In 2018, there were 8,349,297 registered voters, of whom 3,948,620 (47 percent) turned out, to choose 61,950 councillors, a ratio of one councillor for every 134 potential voters (and just one for every 64 participating voters). This ratio should enable councillors — even part-time, non-professional ones — to communicate with constituents and curry favour. As for electoral system, studies reach different findings even when looking at northern European countries that share open-list PR (somewhat akin to the Czech but giving voters only one vote in a multimember district). While incumbency apparently helps councillors get reelected in Sweden and Denmark, it does not in Finland. In Norwegian local elections, incumbency contributes to a councillor’s chances of reelection in that the party elite will place that person near the top of their list; their party as whole, however, might enjoy no boost from already being in office. (In the case of SV in Strakonice, 5 of their 7 incumbents from 2014 stood again in 2018, were placed within the top 6 spots and were reelected, which fits a trend identified a decade ago by Šedo.)
3. Post-Communist Europe, however, is not Scandinavian Europe, and we have to bear in mind that incumbency might have different effects in emerging or developing democracies than in older, affluent ones. As Andrew Roberts observed, incumbency could actually be more of a liability than a bonus, especially if voters suspect corruption or simply want to hedge against the likelihood that an incumbent is corrupt — which could backfire if it encourages rent-seeking politicians to steal as quickly and as much as they can before being turfed out (as found in Romania by Marko Klašnja). The question for the Czech context is whether the country still acts like a permanently emerging democracy or like an older consolidated one.
4. To low-information voters in local elections, incumbency may matter less than other ready cues and heuristics, including candidate information on the ballot (age, sex, street address, profession and academic qualifications or titles) and ballot order effects. Bernard’s analysis of the 2010 local elections identified a mix of significant factors in addition to incumbency, such as age and university degrees, but also variations in the importance of these factors depending on the size of the municipality. A study of Czech subnational results found that “slates ordered within the first three positions on ballot paper enjoy higher shares of council seats won in both regional and large-municipality election contests”, i.e. in areas with more than 5000 inhabitants (which would include Strakonice). That study, however, presumed vertical ordering, but the 2018 Strakonice ballot positioned the party lists horizontally. In that instance, it may have been more important than any bias in local press coverage that the ballot was two-sided and folded down the middle, and SV (as party number 6) was smack in the middle of the front recto page the voter would see before extending the ballot to its full length.
5. Finally, any incumbency advantage might not be undeserved and could be just a selection effect of the voters’ choice at the previous election to go with the better candidates, who can then scare off quality challengers at the next election, especially if local office is less valued as a political prize. Polls find that, in general, Czechs have much greater confidence in local government (around 64 percent) than in the regional and national tiers.
It is not clear from the text of the Constitutional Court’s decision whether any experts were called either to help the lawyers prepare their briefs or to help the judges understand the issues. The implication is that time was of the essence, so that Strakonice’s local government not be stuck in a long limbo, and the court’s opinion has a distinctly rushed, undercooked quality. Even so, there is now a sizeable community of specialists — Czech political scientists and sociologists — who know how local elections work and could be invited to assist. They have been grappling with how to measure the growth of personalization in local politics, and whether an incumbency advantage can be detected in those parts of the Czech political system, such as the Senate, that have a stronger personal dimension (so far, incumbency in that body is, if anything, a liability). Several top experts are based at Masaryk University in Brno, whose political science department is practically next-door to the Constitutional Court in Joštova street; only the Red Church of the great educator Jan Amos Komenský stands between them.
Lastly, Czech courts do not have to reinvent the wheel here: Inspiration can be drawn from other countries, such as France, whose Conseil constitutionnel and Conseil d’État provide a model of how courts can try to locate the line between information and promotion.
[i]I thank my colleague Greg Wolf for bringing this study to my attention.