How economic globalization affects satisfaction with democracy and preferences for direct democracy: Evidence from France

 

The past twenty-five years have witnessed three simultaneous trends: the intensification of economic globalization (Zohlnhöfer et al., 2017), a growing dissatisfaction with representative democracy (Bedock 2017), and the multiplication of direct democratic procedures in advanced democracies, most notably referendums and citizen initiatives (Scarrow 2001; Altman 2011; Qvortrup 2014). So far, these phenomena have mostly been studied in isolation. In this contribution, we argue that they are, in fact, causally linked: globalization leads citizens to be both more sceptical toward representative institutions and more likely to advocate democracy.


Our argument proceeds in two steps. First, we posit that globalization, by reducing elected governments’ leeway on economic issues, adversely impacts satisfaction with representative democracy. Indeed, there is evidence that constraints put upon governments lead to a convergence between parties’ positions on economic issues (Steiner and Martin, 2012) and to the hollowing out of economic issues in party appeals (Ward et al. 2015), potentially hurting the linkage between elites and citizens. Besides, there is also evidence that economic constraints alter the belief that elections matter as mechanisms of policy change, which eventually decrease electoral turnouts (Steiner, 2016). Hence, satisfaction with democracy decreases as globalization deepens because economic constraints have a negative effect on citizens’ political representation.


Second, we argue that the hollowing out of representative institutions can induce citizens to favour mechanisms that increase their influence in the decision-making process, i.e., referendums. While the idea that constraints contributes to the erosion of satisfaction with democracy is not new (Armingeon and Guthmann 2013), the relationship between globalization and demands for direct democracy has been completely neglected so far. The main mechanism driving this relationship is that referendum outcomes provide national governments with stronger mandates to bargain trade deals at the international level: direct democracy enables state leaders to credibly communicate their unwillingness to deviate from their preferred policy position, thus forcing other parties in the bargain to move closer to their ideal point. Alternatively, globalization losers may view referendums as tools to demand better compensation schemes, thus partly offsetting the negative effects of economic integration.


Thus, we expect individuals who believe their governments to be constrained by globalization to be less satisfied with the functioning of democracy (H1); and to be more favourable to direct democracy, especially referendums (H2). The 2017 French presidential post-electoral survey composed of 1830 respondents (Gougou and Sauger, 2017) provides data on perceptions of globalization constraints, satisfaction with democracy, and attitudes toward referendums, thus enabling us to test these two hypotheses at the micro-level.


To test H1, we rely on the widely used “Satisfaction with democracy” scale, which measures the extent to which respondents are satisfied with the way democracy works in their country. The variable is coded from 1 (not satisfied at all) to 4 (very satisfied). The main predictor (perceptions of globalization constraints) takes the value of 4 when respondents believe governments to be heavily constrained by economic globalization, and 0 when they disagree with the statement. We estimate ordered logistic regressions because of the ordinal nature of the dependent variables, and control for the “usual suspects”: social trust, vote for the winning party, vote for an anti-establishment party, economic assessment, age, income, education and left/right positioning.


Table 1 shows the results. Coefficients are displayed in proportional odds ratios: for a one-unit increase of “Globalization constraints,” the odds of being very satisfied with democracy decrease by a factor of 0.77 (i.e., by approximately one quarter). The effect is significant at the 1% level. Thus, our first hypothesis is confirmed: the higher respondents rate globalization constraints, the less satisfied they are with the way democracy currently works. This finding is in line with past research (Armingeon and Guthmann 2013) but provides new empirical insight on the effects of globalization on satisfaction with democracy, since this is the first time that this relationship is confirmed at the micro-level.

 

Table 1: Perceived globalization constraints and satisfaction with democracy (ordered logistic regression).

Dependent variable: Satisfaction with democracy

Globalization constraints

0.766*** (0.0450)

Social trust

1.086*** (0.0278)

Winner

1.600*** (0.237)

Economy assessment

1.875*** (0.124)

Left/right position

1.052* (0.0295)

Age

1.017*** (0.00389)

Income

1.031 (0.0277)

Male

1.035 (0.122)

Anti-system

0.631*** (0.0927)

Education: low secondary

1.092 (0.199)

Education: secondary

1.538** (0.316)

Education: tertiary

1.495** (0.295)

Constant cut 1

0.638 (0.231)

Constant cut 2

6.421*** (2.335)

Constant cut 3

206.9*** (83.05)

Observations

1,158

 Annotations: Coefficients in exponential form. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1.

 

To test H2, we rely on an item which captures respondents’ positions on referendums, and is formulated as a statement (“We should be able to make a referendum if many people demand it”)[1] on which respondents have to take position. We recoded the variable so that the response “strongly agree” takes on the maximal value (i.e., 3 on a four-point scale). We use the same control variables as previously (with the exception of economy assessment) but also add trust in politicians and respondents’ assessment of the political competence of their fellow citizens.

 

Table 2 provides evidence supporting H2: perceived globalization constraints make respondents more likely to demand referendums. Each time scores on “Globalization constraints” increase by one unit, the odds of agreeing strongly with the statement (as compared to lower response categories) increase by more than 30%; likewise, the odds of choosing response categories higher than 0 (strongly disagree) increase by 30%. The effect is significant at the 1% level. This model otherwise confirms previous findings conducted in other countries which have shown that less educated and less trustful citizens are more likely to be in favour of direct democracy mechanisms (Bengtsson and Mattila 2009; Coffé and Michels 2014; Jacquet et al. 2015). This increases our confidence that the model is correctly specified, since control variables behave as expected.

 

Table 2: Perceived globalization constraints and attitudes toward referendums (ordered logistic regression).

Dependent variable: referendums

Globalization constraints

1.309*** (0.0748)

Social trust

0.971 (0.0245)

Winner

0.649*** (0.0901)

Trust in politicians

0.766*** (0.0460)

Perceived citizens competence

0.980 (0.0508)

Left/right position

1.001 (0.0278)

Age

0.840*** (0.0456)

Income

0.916*** (0.0245)

Male

0.820* (0.0961)

Anti-system

1.479*** (0.221)

Education: low secondary

0.807 (0.147)

Education: secondary

0.662** (0.132)

Education: tertiary

0.622** (0.120)

Constant cut 1

0.0105*** (0.00392)

Constant cut 2

0.0560*** (0.0192)

Constant cut 3

0.435** (0.143)

Observations

1,162

 

Annotations: Coefficients in exponential form. Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1.

 

These findings contribute to the literature arguing that economic integration weakens the foundations of democracy by affecting the type of policies implemented, the de-emphasis of economic issues in political competition, and (ultimately) the meaning and the value citizen attribute to representative institutions. However, we add a qualification to this argument, as we show that perceived globalization constraints make citizens more likely to advocate referendums: thus, while globalization may contribute to the erosion of democratic support in the short run, it may promote a strengthening of democracy in the long run, by increasing citizens’ demands for more inclusive participation.

Authors:

Camille Bedock (Université Libre De Bruxelles),

Cal Le Gall (Salzburg Center of European Union Studies and Sciences Po Grenoble), and

Sophie Panel (Sciences Po Bordeaux) in December 2017


Kommentare: 0