The 2010 Regional Election in Catalonia: A Multilevel Account in an Age of Economic Crisis

Abstract

This paper examines the context, actors, campaign, and results of the Catalan election of 2010, which took place in a context of economic recession and growing political discontent. I argue that certain circumstances concurred in this particular contest to enhance the influence of factors pertaining to the first-order (i.e. national) arena as compared with the influence of second-order (i.e. region specific) factors. As a result, regional incumbent support was found to be largely driven by evaluations of national incumbent performance, which worked against the tripartite coalition government.

Keywords:

• Second-order Elections
• Catalan Election
• Multilevel Governance
• Government Evaluations
• Voting Behaviour
• Spain

Acknowledgements

The author would like to thank Agustí Bosch, the editors, and an anonymous referee for their helpful suggestions on the manuscript. This work was supported in part by a postdoctoral mobility grant from the Ministerio de Ciencia e Innovación/Fulbright (ref. 2008-0708).

Notes

[1] Respondents were asked to state what they believed to be the first and second most important problems in Catalonia at the time of the survey. These data come from pre-electoral survey no. 2582, conducted in October 2010 by the Centro de Investigaciones Sociológicas (CIS) (N = 2,966).

[2] The percentage of respondents professing to be unaware of each of the candidates is: José Montilla, 6; Joan Laporta, 14; Artur Mas, 17; Joan Puigcercós, 34; Alicia Sánchez Camacho, 43; Joan Herrera, 47; Albert Rivera, 55; Joan Carretero, 59. See survey information in note 5.

[3] The quarterly telephone surveys of the Centre d'Estudis d'Opinió point to a steady increase in the proportion of respondents naming ‘an independent state’ as their preferred territorial scheme, a shift from 14 per cent in October 2006 to 25 per cent in October 2010. The yearly face-to-face surveys undertaken by the ICPS show an increase from 18 to 23 per cent for the same period.

[4] Data on attitudes towards the economy come from national (not regional) representative surveys, which are available on a more regular basis. No relevant differences were observed when compared with the results of similar questions in regional surveys. ‘Neither good nor bad’ responses are not included in the index.

[5] The surveys were conducted in September–October 2006 (N = 2,000) and September 2010 (N = 2,000).

[6] Note that respondents may have changed their votes within the tripartite (e.g., by supporting the PSC in one election and ICV in the other one). Such cases are considered not defections, but instances of loyalty to the tripartite coalition.

[7] The substantive results of the analysis are highly robust to alternative specifications. For example, one may suspect that the influence of regional government evaluations is being captured by the party image variable. Indeed, party images do mediate some of the influence of regional government evaluations, but this nonetheless stands significantly below that of national incumbent approval if the party image is removed from the 2010 model. It may also be argued that the strong correlation between regional and national performance evaluations biases the results. Yet, the correlation never moves beyond 0.65, which should be considered an acceptable level of collinearity, and even if evaluations are entered separately, the effect of national performance tends to be higher than the effect of regional performance in 2010, while the reverse is true in 2006. The main substantive results also stand when other control variables—such as generic party sympathy, evaluations of both the Catalan and Spanish presidents, and preferences over government composition (single-party vs. coalition)—are added to the models. Interestingly, the influence of judgments of Prime Minister Zapatero is visible higher in 2010 than it is in 2006, thus following the same trend observed in the impact of central government performance. Finally, diagnostic tests show that the findings cannot be ascribed exclusively to the Socialist vote, which arguably is the most prone to encourage voters' connection across levels of government, but extend to the minor members of the tripartite coalition. When PSC voters are dropped from the models, central government performance still has a greater influence in 2010 than it does in the previous election, while the reverse is true for of regional government performance.