Skip to main content

Economic performance and turnout at national and local elections

Abstract

This paper analyzes the impact of economic performance on voter turnout. We estimate an economic turnout model in which local economic variables are included in quadratic form, so that non-linear effects can be taken into account. We use panel datasets covering municipalities, from 1979 to 2005, and cross-sections of parishes (freguesias) to analyze the determinants of turnout at Portuguese municipal and legislative elections. The empirical results indicate that the performance of the national economy is important only in legislative elections and that the regional and local unemployment rates tend to have non-linear relationships to turnout. The results obtained for Flemish municipalities also provide evidence in favor of a non-linear effect of unemployment on turnout.

This is a preview of subscription content, access via your institution.

Fig. 1

Notes

  1. This study also investigates the effects of standard socio-demographic and institutional variables. Although there are some studies that examine turnout at Portuguese elections from a sociological perspective (see, among others, Freire and Magalhães 2002), to the best of our knowledge, there are no articles focusing on the economic dimension nor using panel data techniques.

  2. An early discussion of expressive voting can be found in Buchanan (1954). Fiorina (1976) distinguished expressive voting from instrumental voting (suggested by Downs 1957). For a recent survey of expressive political behavior, see Hamlin and Jennings (2011).

  3. Hillman (2010) presents total utility as the sum of material utility and expressive utility. Voting implies a material loss, given the costs of voting and the low probability of affecting the outcome of the election, and a positive expressive utility, derived from a conception of civic duty or expressive confirmation of identity. Since the cost of voting is very small, the net effect of expressive voting on total utility may be positive.

  4. Due to the consumption cost of information, voters tend to be, on average, ignorant about the economy (see Aidt 2000). But, they may have a stronger incentive to become informed or experience larger changes in their personal finances, when the economy is performing better or worse than usual. Furthermore, the media may devote more attention to economic events in such situations. Thus, voters are more likely to be aware of economic conditions in good or bad times than when economic performance is as usual. In the latter case, voters are expected to be, on average, ignorant about the economy and it will be a less salient issue.

  5. Mainland Portugal is divided into 278 municipalities (municípios or concelhos). Usually, a municipality has the name of its biggest town or city or, in some cases, of its historically most important town or city. The municipality is, usually, considerably larger than the city or town after which it is named. Regarding our data set, between the restoration of democracy in 1974 and 1979 only one election was held at the municipal and legislative levels (both in April 1976). Data regarding the independent variables for the 1976 elections are effectively lost in the estimations due to the autoregressive nature of our model.

  6. Each municipality is subdivided into a variable number of parishes (freguesias). These are the lowest administrative units in Portugal.

  7. For municipal elections we use the turnout for Town Council elections, which are the ones that determine the mayor. Voters also cast their votes on the same day for the Municipal Assembly and for the Assembly of the freguesia.

  8. NUTS is a geocode standard for referencing the subdivisions of countries for statistical purposes, used within the European Union. In Portugal, a hierarchy of three NUTS levels is established. Continental Portugal corresponds to a NUTS I region, which is subdivided into five NUTS II regions. These five regions are then subdivided into 28 sub regions (NUTS III), each one comprised of several municipalities.

  9. This method is a bootstrapping-based algorithm particularly suited for tackling the problem of missing observations in panel data when t<n. It extracts relevant information from the observed portions of a data set by estimating a statistical model that includes a set of explanatory variables. This method was not used to solve the missing data problem with our demographic variables because the number of missing data points was very high and we had hardly any time consistent data for the independent variables.

  10. This is a yearly mandatory employment survey that covers almost all privately owned firms employing paid labor in Portugal (public servants and own employment are not included).

  11. More concretely, the Sales Index (SI) is constructed according to the following equation: \(\mathit{SI}_{m} = 0.2\mathit{Pop}_{m} + 0.8( \sum_{j = 1}^{5} W_{mj} )/5\), where Pop m is the share of municipality m in the national population, and W mj is the weight of municipality m in the country total regarding variable j (the five variables considered are the fiscal burden, electricity consumption, number of cars sold, number of bank agencies, and number of retail commercial establishments). The values of the index are then normalized so that a value of 100 corresponds to the country average.

  12. This refers to the classification of economic activities in three sectors: the primary, the secondary and the tertiary. The tertiary sector is characterized by the production of services instead of end products or raw materials.

  13. However, some comparative studies report the opposite relationship. For instance, Franklin (2004) found a negative effect on turnout of the percentage of the population with a college degree.

  14. Two-step results using robust standard errors corrected for finite samples are reported. The number of observations, municipalities and instruments, and the results of the Hansen and autocorrelation tests are reported at the foot of the table. It is worth noting that, in all estimations of Tables 1, 2 and 3, the Hansen test never rejects the validity of the over-identifying restrictions, and autocorrelation of second order is always rejected. Thus, all our system-GMM estimations meet the requirements set forth by Arellano and Bond (1991) for the validity of GMM estimations of dynamic panel data models.

  15. Since regional GDP data are available only from 1991 onwards, the number of observations is considerably (about one half) smaller than in the estimation of Column 1 of Table 1.

  16. Data are available from 1985 onwards. Relative to the estimation of Column 1, we gain one election for 275 municipalities.

  17. Since the municipal sales index is available only from 1992 onwards, this is the estimation with the smallest number of observations: we lose one election relative to Column 1, and two elections relative to Column 2.

  18. The empirical evidence supporting the existence of non-linear effects of economic performance on turnout is practically the same when economic variables are mean-differenced in order to reduce the correlation between the linear and the quadratic terms. These results, not shown here, are available from the authors upon request. The turning point for the sales index is at 340, which is well above the sample maximum (160). Thus, for realistic values, turnout is increasing in the sales index and the effect essentially is linear.

  19. As our results seem to be in accordance with theoretical expectations at municipal elections, this surprising result for legislative elections may be due to the fact that we are using municipal data, but our measure of competitiveness is district-based.

  20. The turning points are at unemployment rates of 7.54 % (Column 1) and 7.18 % (Column 3). These values are very similar to those obtained in the estimations for legislative elections (see Table 2).

  21. It is necessary to account for heteroscedasticity because both the Breusch-Pagan and the White tests rejected the homoscedasticity assumption in our model.

  22. Turn=log(Turn i /(1−Turn i )). See Dubin and Kaslow (1996) for an application of this method.

  23. Starting with the estimated logistic model, we correct for other potential sources of heteroscedasticity using the feasible generalized least square method proposed by Wooldridge (2003).

  24. The turning points are at unemployment rates of 15.3 % (WLS) and 14.3 % (WLS/FGLS).

  25. In the estimations of Table 6, we used the econometric methodology adopted by Geys and Heyndels (2006), which essentially corresponds to our WLS.

References

  • Aidt, T. S. (2000). Economic voting and information. Electoral Studies, 19, 349–362.

    Article  Google Scholar 

  • Arellano, M., & Bond, S. (1991). Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. The Review of Economic Studies, 58, 277–297.

    Article  Google Scholar 

  • Arellano, M., & Bover, O. (1995). Another look at the instrumental variable estimation of error-component models. Journal of Econometrics, 68, 29–51.

    Article  Google Scholar 

  • Blais, A. (2000). To vote or not to vote: the merits and limits of rational choice theory. Pittsburgh: University of Pittsburgh Press.

    Google Scholar 

  • Blais, A. (2006). What affects voter turnout? Annual Review of Political Science, 9, 111–125.

    Article  Google Scholar 

  • Blais, A., & Dobrzynska, A. (1998). Turnout in electoral democracies. European Journal of Political Research, 33, 239–261.

    Google Scholar 

  • Blundell, R., & Bond, S. (1998). Initial conditions and moments restrictions in dynamic panel data models. Journal of Econometrics, 87, 115–143.

    Article  Google Scholar 

  • Buchanan, J. M. (1954). Individual choice in voting and the market. Journal of Political Economy, 62, 334–343.

    Article  Google Scholar 

  • Caldeira, G., Patterson, S., & Markko, G. (1985). Getting out the vote. Journal of Politics, 47(3), 490–509.

    Google Scholar 

  • Downs, A. (1957). An economic theory of democracy. New York: Harper Collins.

    Google Scholar 

  • Dubin, J., & Kaslow, G. (1996). Comparing absentee and precinct voters: a view over time. Political Behavior, 18(4), 369–392.

    Article  Google Scholar 

  • Fiorina, M. P. (1976). The voting decision: instrumental and expressive aspects. Journal of Politics, 38, 390–413.

    Article  Google Scholar 

  • Fiorina, M. P. (1978). Economic retrospective voting in American national elections: a micro-analysis. American Journal of Political Science, 22(2), 426–443.

    Article  Google Scholar 

  • Fornos, C., Power, T., & Garand, J. (2004). Explaining voter turnout in Latin America, 1980 to 2000. Comparative Political Studies, 37(8), 909–940.

    Article  Google Scholar 

  • Franklin, M. (2004). Voter turnout and the dynamics of electoral competition in established democracies since 1945. New York: Cambridge University Press.

    Book  Google Scholar 

  • Freire, A., & Magalhães, P. (2002). A abstenção eleitoral em Portugal. Lisboa: Imprensa de Ciências Sociais.

    Google Scholar 

  • Geys, B. (2006a). Explaining voter turnout: a review of aggregate-level research. Electoral Studies, 25, 637–663.

    Article  Google Scholar 

  • Geys, B. (2006b). “Rational” theories of voter turnout: a review. Political Studies, 4, 16–35.

    Article  Google Scholar 

  • Geys, B., & Heyndels, B. (2006). Disentangling the effects of political fragmentation on voter turnout: the Flemish municipal elections. Economics & Politics, 18(3), 367–387.

    Article  Google Scholar 

  • Hamlin, A., & Jennings, C. (2011). Expressive political behavior: foundations, scope and implications. British Journal of Political Science, 41(3), 645–670.

    Article  Google Scholar 

  • Hillman, A. L. (2010). Expressive behavior in economics and politics. European Journal of Political Economy, 26, 403–418.

    Article  Google Scholar 

  • Hoffman-Martinot, V. (1994). Voter turnout in French municipal elections. In L. Lopez Nieto (Ed.), Local elections in Europe (pp. 13–42). Barcelona: Institut de Ciènces Politiques I Sociales.

    Google Scholar 

  • Honaker, J., & King, G. (2010). What to do about missing values in time series cross-section data. American Journal of Political Science, 54(2), 561–581.

    Article  Google Scholar 

  • Jackman, R. (1987). Political institutions and voter turnout in the industrial democracies. American Journal of Political Science, 37, 405–423.

    Google Scholar 

  • Jackman, R., & Miller, R. (1995). Voter turnout in the industrial democracies during the 1980s. Comparative Political Studies, 27(4), 467–469.

    Article  Google Scholar 

  • Kirchgässner, G., & Himmern, A. (1997). Expected closeness and turnout: an empirical analysis for the German general elections, 1983–1994. Public Choice, 91, 2–25.

    Article  Google Scholar 

  • Kostadinova, T. (2003). Voter turnout dynamics in post-communist Europe. European Journal of Political Research, 42(6), 467–492.

    Article  Google Scholar 

  • Laakso, M., & Taegepera, R. (1979). The effective number of parties: a measure with application to Western Europe. Comparative Political Studies, 12(1), 3–27.

    Google Scholar 

  • Lau, R. (1982). Negativity in political perceptions. Political Behavior, 4, 353–378.

    Article  Google Scholar 

  • Lijphart, A. (1997). Unequal participation: democracy’s unresolved dilemma. American Political Science Review, 91(1), 1–14.

    Google Scholar 

  • Maddala, G. (1983). Limited dependent and qualitative variables in econometrics. Cambridge: Cambridge University Press.

    Book  Google Scholar 

  • Nannestad, P., & Paldam, M. (1994). The VP-function: a survey of the literature on vote and popularity after 25 years. Public Choice, 79, 213–245.

    Article  Google Scholar 

  • Norris, P. (2004). Electoral engineering: voting rules and political behavior. New York: Cambridge University Press.

    Book  Google Scholar 

  • Paldam, M. (2004). Are vote and popularity functions economically correct? In C. K. Rowley & F. Schneider (Eds.), The encyclopedia of public choice (pp. 49–59). Dordrecht: Kluwer Academic.

    Google Scholar 

  • Powell, G. (1986). American voter turnout in comparative perspective. American Political Science Review, 80(1), 17–43.

    Article  Google Scholar 

  • Radcliff, B. (1992). The welfare state, turnout, and the economy: a comparative analysis. American Political Science Review, 86(2), 444–454.

    Article  Google Scholar 

  • Rosenstone, S. (1982). Economic adversity and voter turnout. American Journal of Political Science, 26(1), 25–46.

    Article  Google Scholar 

  • Schlozman, K., & Verba, S. (1979). Injury to insult. Cambridge: Harvard University Press.

    Google Scholar 

  • Wolfinger, R., & Rosenstone, S. (1980). Who votes? New Haven: Yale University Press.

    Google Scholar 

  • Wooldridge, J. (2003). Introductory econometrics: a modern approach. Mason: South Western Publishers.

    Google Scholar 

Download references

Acknowledgements

The authors thank Jared P. Barton, two anonymous referees, and participants at the 2nd World Conference of the Public Choice Society and at the Martin Paldam Workshop for very useful comments. We also thank Benny Geys for sharing the dataset on turnout at the 2000 Flemish municipal elections. The authors are also thankful for the financial support provided by ERDF funds through the Operational Program Factors of Competitiveness—COMPETE and by the Portuguese Foundation for Science and Technology (FCT) under research grants PTDC/EGE-ECO/118501/2010 and PEst-C/EGE/UI3182/2011.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Francisco José Veiga.

Appendix

Appendix

Table 7 Descriptive statistics: panel data (legislative elections)
Table 8 Descriptive statistics: panel data (municipal elections)
Table 9 Descriptive statistics: cross-section

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Martins, R., Veiga, F.J. Economic performance and turnout at national and local elections. Public Choice 157, 429–448 (2013). https://doi.org/10.1007/s11127-012-0047-5

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11127-012-0047-5

Keywords

  • Turnout
  • Local governments
  • Elections
  • Economic conditions

JEL Classification

  • D72
  • H7