The discursive dimension of second-order elections: the case of Czech regional elections 2012.

• Author: Vasat, Petr
• Position: POLSCI PAPERS - Report

Introduction

The regional elections of 2008 in the Czech Republic were, for the first time, framed by the Czech Social Democratic Party (CSSD) as a referendum about central government. The party focused in its campaign on the unpopular health care reform that had been pushed through by the ruling coalition. One of the coalition partners, the Civic Democratic Party (ODS), was unable to respond adequately and lost the election by a large margin. A similar scenario was anticipated in 2012. However, CSSD failed to come up with an issue of similar strength, while ODS did respond, albeit weakly. New regional movements and especially the Communist Party (KSCM) achieved surprising electoral gains. (4) What did these parties promise to their voters?

Since 2008, it seems that regional elections in the Czech Republic have exhibited a number of characteristics of second-order elections (SOE) (Eibl et al. 2009; Hoskovec and Havlik 2009; Rysavy 2013). In a simplified way, it means that the SOE elections are less important for parties, voters and the media (Reif and Schmitt 1980; van der Eijk and Franklin 1996; Marsh 1998). A specific dimension of SOE theory represents manifestos, and election campaigns in general. Choosing some central parties' unpopular decisions, the top party leaders of the opposition often use regional/local elections as a test of the central government. However, campaign and manifestos analysis are in general marginal in scholars' studies. If they focus on them, they are preoccupied with the thematic dichotomy of central ruling parties and the opposition. Moreover, when concentrating on its content, they completely omit a linguistic level of it and its potential impact.

The goal of this paper is to describe issue and linguistic differences in election manifestos of the parties involved in the 2012 regional election, and to contribute to better understanding of second-order election theory on discursive level. In particular, the paper wants to look at differences between strong well-established parties on the one side and new unsatisfied movements and the Communist party, that both experience a great voting success in the Czech Republic as well as in other countries of Central and Eastern Europe, at the other. In doing so, the article seeks to answer the following two questions: (1) Are Czech regional elections the case of second order election? (2) If they are, does it appear in the discourses of political parties during election campaigns? (3) If it does, how is it manifested? Thus, we expect that SOE theory will be supported also by the linguistic characteristics. Strictly speaking, if the top party leaders of the opposition use regional elections as a test of the central government, they will emphasise central government issues, and their language will be more strict and uncompromising toward the central government. On the other side, the central ruling party will emphasize more regional issues, and their language will be more strict and uncompromising towards the regional ruling party. We can also assume that the language of new regional movements and the Communist Party (KSCM) will be strict and uncompromising in the same manner, but with the only difference that it will address all parties regardless of their central/regional ruling distinction.

The article is divided into several sections. In the first section, we briefly account for background of regional elections in the Czech Republic and a conceptualization of the SOE theory. In the second one, a field of election manifestos analysis and our conceptualization of the discursive dimension of SOE theory are outlined. In the third one, we describe our methodological approach and specific steps of our analysis. It is mainly based on the quantitative content (issue salience) analysis (Klingemann, Hofferbert and Budge 1994; Green-Pedersen and Walgrave 2014) but combined with qualitative discourse analysis (Chouliaraki and Fairclough 1999). With regards to the quite limited sample, our salience analysis was less formal as usual. We do not formulate hypotheses, nor do we test them; it is rather an explorative process. In the last section, we present the results of our salience analysis, and describe identified discourses in the context of the SOE theory.

Czech Regions and SOE Theory

After the demise of communist regimes in post-communist countries, numerous public administration reforms took place in the 1990s. These were aimed at decentralising the public administration and renewing the autonomy of local administrations (Back, Heinelt, and Magnier 2006; Loughlin, Hendriks, and Lidstrom 2011). These post-communist countries, which aimed at joining the EU, had to deal first of all with, and adhere to, a host of recommendations and requirements. One of these requirements was adopting the EU multi-level system of governance. Part of this process of adaptation was the institutionalisation of the regions and creation of regional administrations (Brusis 2005).

The case of Czech Republic was not different. We can identify two principal inducements. Firstly, the new constitution of 1992 expected the creation of regions. Secondly, the creation of a new level of territorial administration was also expected from the Czech Republic by the European Commission under the terms of preliminary EU accession negotiations (Illner 2015; see here Dostal & Hlousek 2015 as well, on attitudes towards the EU in the Czech Republic). The Czech Republic established the second tier of local government in 2000, dividing its territory into 14 higher territorial units: 13 regions, and the capital city of Prague. The process dates back to 1997, when the Parliament of the Czech Republic legislated regions. In 2000, the law came into effect, and the first regional elections took place that year. Since then, there have been regional elections every four years, the latest ones taking place in 2012 (Figure 1).

According to SOE model developed by Reif and Schmitt (1980), we can distinguish various types of elections, depending on what is at stake. There are more important national first-order elections (FOE) and less important second-order elections. These were, in its original sense, the elections to the European Parliament (Reif 1984; Marsh 1998; Hobolt and Wittrock 2011; however, regional elections are also very often seen as a case of SOE (Pallares and Keating 2003; Jeffery and Hough 2001). The qualitative difference of SOE, in comparison to FOE, can be characterised by five general dimensions: the less-at-stake, the specific arena, the institutional-procedural, the campaign and the main arena political change (Reif and Schmitt 1980). An important aspect of SOE is also "the temporal location of the SOE in the FOE cycle and its impact on the election outcome" (Koepke and Ringe 2006, 326). Opposition parties could gain more support on behalf of government parties and SOE could become the elections where voters "experiment" with smaller or new parties (Jeffery and Hough 2011, 78). In this sense, SOE becomes the arena where voters could "punish" or "warn" government for its outcomes in the time between two subsequent FOEs.

In 2000, candidates nominated by political parties ran for a total of 675 regional assembly mandates in 13 regions. Turnout was around one-third of all eligible voters as the elections were carried by, and 93% of all mandates were split between the four largest parliamentary parties of that time (ODS, CSSD, KSCM and the Christian Democrats or KDU-CSL). ODS won the elections of 2000 and 2004. The following elections in 2008 were carried by CSSD, which replaced ODS in a number of regional governments and became the strongest party in regional assemblies. It still held that the four above named parties gained overwhelming majority of the mandates--more than 93%. ODS and CSSD alone gained 68% of the mandates. Despite the fact that clearly not all of the regional elections held in the European countries could be considered to be SOE (Schakel and Jeffery 2012), these elections are considered to be the example of SOE by the Czech scholars. In prticulat the lower turnout, the higher percentage of void ballots, and the government parties' loss were characteristic for the regional election held in 2008 (Eibl et al. 2009; Hoskovec and Havlik 2009).

The fourth regional elections in the country's post-communist history took place in the autumn of 2012. These elections also exhibit characteristics of the SOE: the lower turnout, higher percentage of void ballots, the government parties' loss, and a significant part of voters favoured the small regional parties (see Cermak 2013). By the total number of candidates elected, the Social Democratic and the Communist Party were the most successful: CSSD won 205 mandates and KSCM won 182 mandates (they had their candidates in all regions). The third-ranking ODS also won 102 mandates in all regions. (5) Thus, the CSSD carried the elections overwhelmingly, forming coalitions in 12 of the 13 regions and nominating 11 out of 13 governors. Voter turnout was once again rather low. Rysavy (2013) summed up the situation of 2012 as "a mix of old and new."

In the Czech Republic, in addition to the low voter turnout coupled with a higher proportion of void ballots (Eibl et al. 2009), we can primarily see an outflow of voters on the part of the ruling parties in the central government. (6) This happened during the regional elections of 2008, as well as in 2012. The ruling coalition parties in this respect attracted only 18.91% of the voters (STAN votes included). However according to Rysavy (2013, 6), the course of the regional elections was significantly affected by the central government politicians. This was mainly in the form of disagreements within the ruling coalition and, at the same time, because of arrests and criminal investigations of some individuals. All of this was reflected by the voters' dissatisfaction and their mistrust...