Conditional satisfaction: political support, congruence, and cabinet composition

ABSTRACT

This paper examines the relationship between citizen satisfaction with the functioning of democracy and ideological congruence. We focus on how this relationship may vary by government type, paying attention to the conditioning effects of coalition governments’ ideological make-up and individual-level education. Our analyses rely on harmonized survey data covering one million respondents in 28 countries over a 40-year period. We find limited evidence that the relationship between citizen satisfaction and ideological congruence is conditional on national government type. All coalitions are not, however, created equal. Comparing single-party governments to multi-party governments with different ideological compositions, we find striking differences, but only for the higher educated. While the negative relationship between citizen satisfaction and ideological incongruence is similar for lower-educated citizens in single-party and multi-party coalition settings (irrespective of cabinet composition), for the higher educated, the relationship weakens as a function of the ideological diversity of the coalition cabinet.

KEYWORDS:

• Ideological congruence
• satisfaction with democracy
• coalition government
• education‌

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 Even though there are intrinsic problems with comparing citizens’ left-right self-placement in a survey and expert-coded party-placement, Powell (2009) finds that alternative approaches, such as citizen perception of party position, expert surveys, and manifesto data, generate similar congruence scores. Since we are interested in within-country congruence, this eliminates some of the problems with this approach. See Lo et al. (2014) for a detailed discussion.

2 See Appendix for more information on the procedure used to match surveys and cabinets using survey field dates.

3 We also calculate ideological congruence in two other ways as (1) an unweighted average, and (2) using the left-right position of the party of the prime minister. The predicted probabilities estimated on the basis of the models using these alternative measures (see Figures A4-A7 in Appendix) are substantively very similar.

4 In our sample, respondents fall into these government types as follows: single-party majority governments (16.2%); multi-party majority governments (57.1%); single- and multi-party minority governments (26.7%), and other government types (0.1%).

5 We also ran our models using a weighted measure of cabinets’ ideological diversity using government parties’ legislative seat share. See Appendix for more information.

6 See Appendix for more information on education coding.

7 Following Fairbrother (2014), all macro-contextual variables are group-mean centered. All models are estimated using MLwiN.

8 See Table A1 in Appendix for the main regression output discussed in the text. Table A6 displays the full regression model. Table A2 reports the predicted probabilities related to individual- and macro-level covariates.

9 Our predicted probability figures use 84-percent confidence intervals. An overlap in the 84% confidence intervals of two point estimates indicates that they are not statistically different from one another at the 5% level in a two-sided test (MacGregor-Fors and Payton 2013).

10 For this, we set cabinet ideological diversity at -1, +1, and +2 standard deviation(s) below/above the sample mean for multi-party governments. A low-diversity coalition has a maximum ideological distance between governing parties of approximately 1 point on the 10-point left-right scale. The ideological distance between governing parties in a coalition with very high diversity is approximately 5 points on the same scale.

11 Predicted probabilities are estimated based on citizens with a level of education two standard deviations below/above the mean.