Educational and class cleavages in voting behaviour in Belgium: The effect of income, EGP class and education on party choice in Flanders and Wallonia (original) (raw)

Abstract

This study compares the effect of EGP class, income and level of education on voting behaviour in Flanders and Wallonia and tests to what extent these effects are mediated by economic and cultural attitudes. Rather than using a left/right dichotomy or a one-dimensional left/right continuum, as is common in research on cleavage voting, we distinguish the main different parties in both regions. Using data from the 2008 European Social Survey, our Multinomial Logistic Regression analyses indicate some regional variation. We find that education generally plays a more important role in party choice in Flanders than in Wallonia, whereas income and EGP class are only relevant for party choice in Wallonia. These effects remain even when economic and cultural attitudes are controlled for. Furthermore, our analyses indicate that preferences for traditional left- or right-wing parties are influenced not only by economic concerns but also by education and cultural attitudes. Similarly, preferences for new left- or right-wing parties are affected by both education and income, as well as by attitudes towards economic and cultural issues, especially in Flanders. These results highlight the need to include all parties simultaneously when studying cleavage voting in multiparty systems.

Access this article

Log in via an institution

Subscribe and save

Buy Now

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Similar content being viewed by others

Notes

  1. It should be noted that the new left-wing has been more successful in Wallonia than in Flanders. For example, Ecolo gained 12.3 per cent of the votes in the 2010 parliamentary elections, compared to 6.9 per cent for Groen!. Yet, Ecolo does not seem to significantly influence the political agenda with so-called new politics issues (Coffé, 2009).
  2. The cartels of CD&V/N-VA and SP.A/Spirit were presented on the ballot and thus voters could not choose between the separate parties forming a cartel. They were also presented as cartel in the ESS survey.
  3. We also excluded Lijst Dedecker (_N_=31) for Flanders owing to the small numbers of respondents. We did run our analyses including both Front National and Lijst Dedecker. This did not significantly change our results and main conclusions.
  4. All models were examined for multicollinearity by means of tests of the Variance Inflation Factors (VIF). With VIF values <2 and Tolerance values >0.5, we can conclude that there was no problem of multicollinearity among the variables education, EGP class and income.
  5. We performed a Principal Axis Factoring analysis that also included items on authority and the environment. This analysis indicated that these new politics attitudes did not form a single factor with the issue of migration and could thus not be included in one scale. As the migration issue is considered a salient topic in the new politics parties, we decided to use these to measure cultural attitudes.
  6. Trade union membership is known to capture political leanings and to influence voting behaviour. Inclusion of such membership thus allows us to investigate the net effect of class, income and education on voting behaviour. We replicated our analyses without trade union membership as a control variable. The results of these analyses showed similar results to those presented in our study.
  7. One thousand one hundred and sixty-seven respondents had complete information on all independent variables used in this study.
  8. The effects of our main independent variables on party choice are similar if we run the analyses without any control variables.
  9. In some preliminary analyses, we ran each model with only the cultural and only including economic attitudes as mediating variables. The main effects of EGP class, income and level of education were similar to those presented below.
  10. The odds were computed with the values of the unstandardized _b_-coefficients (latent level).
  11. We checked the significant differences of the effects between all Models 1 and 2 with Likelihood ratio tests.
  12. For CD&V, the odds of voting decrease by 50 per cent with each unit of economically progressive attitudes. This decrease in odds is 37 per cent for culturally progressive attitudes. For Open Vld, the odds of voting decrease by 60 per cent with each unit of economically progressive attitudes. This decrease in odds is 30 per cent for culturally progressive attitudes.
  13. However, if we run this model without the EGP-class measure, income and education, the effect of cultural progressive attitudes becomes negatively significant (_P_=0.02), indicating that people with culturally progressive attitudes are more likely to vote for Groen! than for SP.A.

References

Download references

Author information

Authors and Affiliations

  1. Department of Sociology/ICS, Utrecht University, Heidelberglaan 2, Utrecht, 3584 CS, The Netherlands
    Job C van den Berg & Hilde Coffé

Authors

  1. Job C van den Berg
  2. Hilde Coffé

Corresponding author

Correspondence toJob C van den Berg.

Appendix

Appendix

Tables A1 and A2

Table A1 Descriptive statistics of all dependent and explanatory variables (Flanders)

Full size table

Table A2 Descriptive statistics of all dependent and explanatory variables (Wallonia)

Full size table

Rights and permissions

About this article

Cite this article

van den Berg, J., Coffé, H. Educational and class cleavages in voting behaviour in Belgium: The effect of income, EGP class and education on party choice in Flanders and Wallonia.Acta Polit 47, 151–180 (2012). https://doi.org/10.1057/ap.2011.34

Download citation

Keywords