Autonomy or Independence: An Analysis of the 2010 Catalan Regional Election

Abstract

This paper analyzes voting for nationalist parties, Convergència i Unió (CiU), Esquerra Republicana de Catalunya (ERC), and Solidaritat Catalana per la Independència (SI), as well as the nonnationalist Ciutadans-Partit de la Ciutadania in the 2010 Catalan autonomous election with data from the Centro de Investigaciones Sociológicas (CIS) postelection survey. It uses the multistage explanatory model of voting behavior to show that nationalist voting can be explained by long-term factors, notably being from Catalonia, as well as short-term variables such as evaluations of the incumbent governments at the central and regional levels. The analysis finds that long-term variables explain more of the vote than short-term political and economic evaluations. It presents typical profiles of voters of the major nationalist parties and of Ciutadans. Finally, this paper shows simulations of voting probabilities given changing political and economic contexts. It shows that the nationalist vote of native Catalans is more resistant to changing evaluations than that of nonnative Catalans. This implies that the challenge posed to the Spanish State of the Autonomies by Catalan nationalism is relatively enduring regardless of the short-term economic and political contexts.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

The election to the Parliament of Catalonia held on November 28th 2010 signaled the return to government of the conservative nationalist coalition Convergència i Unió (CiU). It also saw the emergence of new separatist parties Solidaritat Catalana per la Independència (SI) and Reagrupament Independentista (RI). Esquerra Republicana de Catalunya (ERC), which had been advocating independence for Catalonia since 1989, also participated in the election. At the same time, the nonnationalist Ciutadans-Partit de la Ciutadania (C’s), which first ran in the 2006 regional election, retained its status as a minor party with parliamentary representation.

This election followed the June 2010 judgment by the Spanish Constitutional Tribunal that parts of the new regional autonomy statute adopted in 2006 were unconstitutional. This decision provoked a massive protest in the streets of Barcelona and seemed to be linked to a radicalization of the nationalist movement in Catalonia, with increasing demands for the independence of the region. There was also great dissatisfaction with the Spanish government’s handling of intergovernmental relations, notably with respect to financial transfers to the regions. Moreover, the incumbent tripartite leftist coalition of the Partit dels Socialistes de Catalunya (PSC), Iniciativa per Catalunya Verds (ICV), and Esquerra Republicana had been widely criticized by the opposition, especially for its failures in defending Catalonia’s self-government. Thus, this election is interesting not only because several parties challenged the current structure of the Spanish state but also because the short-term political context seemed to be favorable to these parties.

This paper uses the postelection survey conducted by the Centro de Investigaciones Sociológicas (CIS). It first shows that both long-term and short-term variables contribute to explaining Catalans’ vote choice in 2010 but enduring predispositions have a more powerful impact than short-term evaluations of incumbent governments and of the economy. It then presents profiles of typical voters of the three most significant nationalist parties, CiU, ERC, and SI and of the nonnationalist C’s. Too few people supported RI. It was, therefore, impossible to include it in the analysis. This paper ends with simulations showing that the nationalist vote of native Catalans is more resistant to changing evaluations of incumbent governments and of the economy than that of nonnative Catalans. This implies that the challenge posed to the Spanish State of the Autonomies by Catalan nationalism is relatively enduring regardless of the short-term political context.

The Multistage Model of Nationalist Voting

The analysis presented here is based on the multistage model of voting widely used in election studies (Blais et al. 2002). This type of analysis identifies variables at different degrees of proximity to the vote. Those that are most distant, especially social background and values, are quite fixed in the long term. Others, such as evaluations of government performance and of the economy, change over time as the economic and political contexts vary. More distant variables can affect the vote both directly and via more proximate variables. The following figure shows a simplified representation of the model applied to nationalist voting in Catalonia. It shows the factors that are associated with voting for the nationalist coalition CiU (Fig. 1).

Fig. 1

Simplified multistage model of Nationalist Vote Choice (for CiU)

Other explanations of nationalist voting in Catalonia focus on spatial models either exclusively (e.g. Balcells i Ventura 2007) or in combination with some of the short-term variables analyzed here (e.g. Aguilar and Sánchez-Cuenca 2007). The analysis presented in this chapter presents a wider range of variables, from social background characteristics to leader evaluations, at varying degrees of proximity to the vote. Doing so allows us to explain a greater proportion of voting behavior in Catalonia. It also allows us to simulate changes in some of the variables, while keeping others fixed.

Major Determinants of the Vote

To identify the variables that help explain nationalist voting in Catalonia, a series of multinomial logistic regressions were run, using vote as the dependent variable and adding variables that are progressively closer to the individual’s voting decision. This allows us to determine which variables contribute to an explanation of nationalist voting at different levels of proximity to the vote.

The first level consists of social background variables. The most significant demographic influence on the nationalist vote is a Catalan voter’s degree of nativeness. A variable was constructed combining respondents’ answers to questions about whether they were born in Catalonia, whether their mother was born in the region, and whether their father was born there. Figure 2 displays the predicted vote choices of two ideal types. A native Catalan is someone who was born in Catalonia and whose mother and father were born there as well. A nonnative Catalan was born outside of Catalonia, as were his or her parents. As can be seen, nonnatives were significantly less likely to vote for CiU than natives, and their probabilities of voting for ERC and SI were practically nil. On the other hand, they were much more likely than native Catalans to vote for the nonnationalist Ciutadans. While it may be objected that native and nonnative are extreme categories that do not represent all Catalans, 36 % of Catalans are natives, as defined here, and 26 % of Catalans are nonnatives. In other words, these extreme categories include more than 60 % of the population. The other 40 % has intermediate levels of support for the nationalist and nonnationalist parties. This initial analysis has thus identified a major cleavage in Catalan society. Overall, this variable explains about 6 % of the variation in the vote.¹

Fig. 2

Vote choices of natives and nonnatives

Once the influence of social background characteristics has been taken into account, Catalans’ self-placement on a scale of nationalism running from one (low nationalism) to ten (high nationalism) adds significant explanatory power to our nationalist voting model. As can be seen in Fig. 3, voting for nationalist parties increases along with individuals’ self-placement on the nationalism scale. However, natives are practically always more likely to vote for nationalist parties than do nonnatives. Moreover, support for the separatist Solidaritat is concentrated among the most extreme nationalists. Voting for the nonnationalist Ciutadans follows the opposite pattern. It decreases as the level of nationalism increases and is always higher among nonnatives. Respondents’ subjective nationalism explains an additional 7 % of the vote in 2010.

Fig. 3

Impact of self-placement on scale of Nationalism

Evaluations of the economy also had an independent impact on support for the nationalist parties. However, their impact was small, explaining only about 1 % of the vote. Negative evaluations of the Spanish economy, which were very common, benefitted CiU while hurting Ciutadans. Unfavorable evaluations of the Catalan economy, which were also quite common, helped CiU, ERC, and Ciutadans.²

Adding preferences regarding the form of the Spanish state to the model explains another 2 % of the vote. While the impact on the probability of supporting the nationalist coalition Convergència i Unió was negligible, separatist Esquerra republicana’s support was strongest among those wanting at least more regional autonomy for autonomous communities in Spain, and Solidaritat’s support was practically nonexistent among those who did not support independence for the regions of Spain. Support for Ciutadans was greatest among respondents preferring greater centralization.

Adding evaluations of the Spanish and Catalan governments allows us to explain an additional 1 % of the vote. The most significant influences on the vote in this category were evaluations of the central government’s policies towards Catalonia’s autonomy and of the regional government’s handling of its own self-government. Respondents who rated the Spanish government’s treatment of Catalonia’s autonomy negatively were more likely to vote for CiU, ERC, and SI than those who had a neutral evaluation. Those who evaluated the Catalan government poorly on this issue were more likely to vote for CiU and SI but less likely to vote for ERC, which had been part of the previous governing coalition in Catalonia.

Finally, the short-term variables that seem to have the highest impact on the vote are evaluations of the leaders of the different parties. In fact, they explain almost as much of the vote as all the previous variables combined, about 31 %. Higher leader ratings are associated with higher probabilities of voting for the different parties. However, it is difficult to determine the direction of causality: Do people vote for a party because they like the leader, or do they like the leader of the party they vote for? Because these are not clearly causal variables, the rest of this paper will focus on the other variables we have identified.

The above findings allow us to identify typical voters for each party. Table 1 shows the values of each of the significant independent variables identified above that is associated with the highest probability of supporting each of the parties.

Table 1 Typical voters of each party

Simulations

Using the model of nationalist voting presented above, it is possible to observe how the probabilities of voting for the different parties change when evaluations of the incumbent governments and of the economy change. The probability of natives and nonnatives voting for each party is shown when evaluations change from the most favorable for each party to the least favorable. For CiU, favorable evaluations are negative with respect to the governments and economies of both Catalonia and Spain. Figure 4 shows that the probability of voting for CiU declines among natives by about 34 % points when evaluations of the government and of the economy move from most favorable to least favorable for CiU. Meanwhile, support for the three left-wing parties that formed the previous governing coalition in Catalonia, including ERC, increases.

Fig. 4

Vote probabilities for CIU given most and least favorable evaluations (natives). Note: Negative evaluations of the Catalan and Spanish governments and economies are favorable to CiU. The reverse are unfavorable

Figure 5 shows that support for CiU among nonnative Catalans declines by a similar 32 % points when evaluations go from being favorable to unfavorable for CiU, to the benefit of the Socialists. However, this decline represents a much more significant proportion of nonnatives’ nationalist vote. In fact, unfavorable evaluations are associated with a probability of voting for CiU of less than 10 %. This means that most of nonnative Catalans’ support for CiU is motivated by negative evaluations of the incumbent governments and of the economy.

Fig. 5

Vote probabilities for CIU given most and least favorable evaluations (nonnatives)

Because so few respondents voted for the other parties discussed in this paper, it was impossible to conduct similar simulations for them. Nevertheless, the finding with respect to support for the moderate nationalist Convergència i Unió is intriguing. If its support is still relatively strong amongst native Catalans who have evaluations of the economy and of government that are all unfavorable to it, this suggests that its support is relatively firm, not being subject to changes in the political and economic contexts. This suggests that any attempt to satisfy nationalists would be extremely difficult. If a significant amount of nationalist support comes from individuals who are highly committed to the party, it is unlikely that any change in policy, for example, would convince many of the native Catalans who voted for CiU to vote for another party that poses less of a challenge to the status quo. Therefore, any significant reduction in nationalist support could only be achieved by some major policy or institutional change that had an effect on how much people identify with Catalan nationalism.

Conclusion

This paper has presented a multistage model of voting for nationalist parties, as well as for the nonnationalist Ciutadans-Partit de la Ciutadania in the 2010 election to the Parliament of Catalonia. It showed that the most powerful determinants of nationalist voting are, on the one hand, whether an individual and his or her parents were born in Catalonia and, on the other hand, how nationalistic an individual considers him or herself. Short-term factors, notably preferences regarding the future of the Spanish state and economic and governmental evaluations, also help explain voting in 2010, but their impact is much smaller. Simulations using this model show that much of the vote for CiU can be explained by negative short-term evaluations. Changing these to their most positive values decreases CiU’s electoral support significantly. This paper has also identified two types of voters. Nonnative Catalans’ support for CiU almost disappears when we simulate a context unfavorable to CiU, while native Catalans’ propensity to support CiU remains relatively strong even when the context is not at all favorable for the nationalist coalition. This finding suggests that support for CiU among natives depends on their Catalan identity, as well as their evaluations of the political and economic context. Nonnatives’ nationalist support is mostly motivated by short-term economic and political considerations. Therefore, it seems that the impact of the political and economic context on nationalist challenges to the Spanish state is relatively limited. Satisfying many of those who currently vote for nationalist parties would thus be a significant challenge.