Opinion & Analysis

There is little evidence European integration has created a representation gap between politicians and voters

Recent research suggests that European integration may have led political parties to adopt increasingly similar positions. This ideological convergence among parties could pose a democratic challenge if it also increases the gap between citizens and their representatives. Drawing on a new study, Daniel Devine and Raimondas Ibenskas find little evidence that the integration process has reduced the level of congruence between parties, governments and parliaments on one hand, and citizens on the other.

A key concern about European integration is the consequences for the democratic legitimacy of EU member states. Whilst this has been most evident in Eurosceptic discourse across Europe – encapsulated by the ‘take back control’ slogan of the UK’s Vote Leave campaign – it has been a longstanding topic in academic research. The late Peter Mair’s book Ruling the Void, for instance, casts the European Union as a body that is designed to evade the constraints of representative democracy through non-majoritarian institutions removed from the domestic political sphere – and the influence of citizens.

These normative judgements aside, academic research has investigated how European integration impacts an important part of domestic politics: the party system. The party system, and the parties within it, aggregate and mobilise the preferences of the public, and form governments and parliaments.

Whilst early research suggested there was no effect of integration, there is evidence that European Parliament elections give a leg-up to challenger parties which increases the number of parties in domestic elections. There is also evidence that European integration has led to parties becoming more similar over time, but then polarising as European policymaking becomes more constraining.

In a new study, we argue that these developments are less important if political elites are congruent (aligned) with public preferences. Put differently, it would be much more concerning if, alongside these developments, European integration also led to parties, governments and legislatures being less representative of their publics. This was raised as a potential issue in the earlier debates about the European Union’s ‘democratic deficit’, in which it was suggested that there was ‘policy drift’ in which policies are adopted away from (most) voters’ ideal positions.

Our theoretical argument, and more so our empirical evidence, challenge the stylised argument that integration leads to a loss of congruence. Starting with the assumption that integration leads to the convergence of mainstream parties, as empirical evidence so far suggests, we argue that this may actually increase the congruence between the median voter on the one hand and governments and the median parliamentary party on the other. However, integration should still decrease the ‘many-to-many’ congruence between the party system and the electorate, as only centrist voters will be well-represented.

Additionally, the EU does not have the same influence across issue areas – we would expect a greater impact on economic regulatory and immigration policies than, say, redistribution. Issue areas should further differ on whether congruence affects more the left/liberal or right/conservative voters.

To test the effect of European integration on citizen-elite congruence, we match public opinion data on left-right ideological placement and five issue areas (attitudes to redistribution, integration, the environment, immigration, and social liberalism) with expert survey data on parties’ positions on the same, largely using the Chapel Hill Expert Survey.

To measure European integration, we use six different measures capturing different aspects of the integration process and themselves measured in different ways, such as economic integration through intra-EU trade flows (‘EU Index’), an expert survey, legislative output (Directives, Regulations and Decisions), and treaty readings (ENLABASE). Our analysis includes all EU member states and, whilst the time period varies depending on the measure, it ranges from 1980 to the late 2010s at its longest.

Our results indicate that European integration has no noticeable impact on congruence between citizens and elites across any of our measures. No measure of integration has an impact that is statistically distinguishable from zero, and most coefficients are centred over zero. Moreover, the direction of the effects – whether integration increases or decreases congruence – has no pattern, which suggests we are not just limited by our relatively small sample size.

An example of our results is presented in Figure 1, which shows the coefficients of the different measures of integration (on the left) on parliament-public congruence across different issue areas, including left-right position. As noted, it is clear that most coefficients are over zero (i.e., have no effect) and there are no clear patterns regarding direction. In addition to this analysis, we present an extensive range of alternative tests that lead to the same conclusion.

Figure 1: Effect of European integration on congruence between public and parliament

Note: For more information, see the authors’ accompanying paper in European Union Politics

Overall, our results lead us to conclude that integration has not had an effect on the congruence between the public and elites – our uniquely large range of measures of congruence and integration provide us with additional confidence.

There are a few reasons why this might be the case. First, challenger parties – new parties on the left and right – have increased in electoral strength in many EU countries; whilst they do not often gain control of government, it may be that these parties are sufficiently influential to force governing parties to adopt policy positions they may not otherwise, or simply fill the gap left by converging mainstream parties.

Second, it might be that whilst parties have indeed become more similar, as shown by previous work, so have voters. In other words, there may have been policy changes caused by integration, but this has not led to a loss of congruence between public and elites. Another explanation may be rooted in that our analysis has focused on averages, such as the median legislator or voter, but there may be a divergence between the median voter and an increasingly disaffected minority of citizens which we have not identified.

These explanations aside, our results provide some rebuttal to the ‘democratic deficit’ fears of European integration. Whilst there may be many other problematic features, we suggest that a loss of congruence – of a match between governing elites and the governed public – is not one of them: domestic elites are still broadly representative of their citizens.

For more information, see the authors’ accompanying paper in European Union Politics

©European Union, 2021, Source: EP – Multimedia Centre