Do European Parliament elections foster challenger parties' success on the national level?

Abstract

European Parliament elections create structural advantages for challenger parties. Building on the second-order elections theory, this article argues that European Parliament elections foster challenger parties' success on the national level by increasing their visibility and offering an opportunity structure for domestic politicisation of Europe. I test this proposition by exploiting the quasi-exogenous timing of European Parliament elections and the variation in national electoral cycles since 1979. A country-fixed effects model and two placebo-tests show that populist radical right parties gain momentum in the supranational contest, particularly when coinciding campaigns increase the domestic salience of Europe. Considering their antagonism to an integrative Europe, it seems ironic that the European Parliament elections foster the ascendency of just these opponents of the European idea.

Keywords

European integration European Parliament elections political parties radical right second-order elections

Only eight months after having narrowly missed the five percent threshold in the German Federal Election in 2013, the populist right ‘Alternative für Deutschland’ (AfD) gained more than 7% of the German votes in the European Parliament (EP) elections. Immediately after this success, nation-wide opinion polls reported a surge in public support to eight percentage points, indicating that the party would pass the national threshold if elections were to take place. Meanwhile, the ‘Sverigedemokraterna’ (SD) doubled their result in the Swedish ‘Riksdag’ election four months after their unexpected success in the 2014 EP elections. Born only a couple of months prior to the EP elections 2014, also the Spanish ‘Podemos’ movement drew crucial momentum from the broad media coverage related to their European success, helping the young party to become the third largest party in the Spanish general election a year later.

According to the second-order elections theory (Reif and Schmitt, 1980), challenger parties are likely to be successful in European elections. While the EP election is supra-national in nature, the related campaigns still take place on the national level, and national parties run for office in the European contests. Within each country the party system, media, and electorate are virtually identical in the domestic and European arena. Offering structural advantages to challenger parties, the institution of EP elections may have unanticipated consequences for national party competition (van der Brug and de Vreese, 2016). Although the literature has established that second-order elections facilitate the success of challenger parties (Hix and Marsh, 2007), it is not fully understood how their success in the second-order arena relates to their national performance (Somer-Topcu and Zar, 2014). Despite low levels of voter turnout, the very institutional existence of the EP elections offers challenger actors a forum to promote their policy-demands and to attract national attention.

This article argues that challenger parties gain momentum in EP elections. Building on the second-order elections theory, it posits that the EP elections foster challenger parties' success on the national level. I test this proposition by exploiting the variation in national electoral cycles and the quasi-exogenous timing of EP elections since 1979. The results show that particularly populist radical right parties draw crucial momentum from the supranational contest. Their national gains are greatest when the European and the national election are close in time. By changing the focus from the European to the national arena, the paper contributes to an emerging research agenda on the national implications of EP elections (Dinas and Riera, 2018; Franklin, 2016; Franklin and Hobolt, 2011; Markowski, 2016; van der Brug and de Vreese, 2016). The study disentangles the spillover effect from alternative explanations and sheds light on the underlying mechanisms of the spillover, establishing that: (a) the impact of EP elections on the national fortune of the radical right does not only stem from congruent voter preferences across governance levels; and that (b) the mere event of the EP contest benefits radical right actors when the national election is close in time. The EP elections offer an opportunity structure for the populist right to make their antagonism towards further integration domestically salient, potentially imperilling the European project.

In times of growing nationalism and the rise of populism across Europe, it is important to understand the implications that EP elections have for challenger parties' national success. Shedding light on the mechanisms that augment the domestic prospects of challengers, this paper contributes to uncover the unintended, disintegrative consequences of the European direct elections.

Electoral success for challenger parties in EP elections

Existing research shows that challenger parties have higher chances of electoral success in EP elections than in national elections because of: (a) the secondary character of the EP elections; (b) their stances on Europe in their policy proposals; and (c) the permissiveness of the electoral system in the European arena.

First, according to the seminal second-order elections theory, challenger parties have better prospects to succeed in EP elections since the elected representatives in the European arena do not decide about government formation and no immediate policy-implications accrue out of the EP result. This renders the EP elections secondary to the national elections (Reif and Schmitt, 1980; van der Eijk et al., 1996), which bears implications for citizens' voting rationale. Voters use the supranational elections instrumentally to express dissatisfaction with their national governments (Hix and Marsh, 2007). Moreover, voters are likely to defect from their national party choice due to the lack of parties' mobilisation efforts during European campaigns (Weber, 2007). Second, the EP elections are favourable to challenger parties as some voters engage in EU-issue voting. Those voters, in turn, are inclined to support a challenger in the EP elections since mainstream parties are commonly more pro-European than their average supporters are (Hobolt et al., 2009; Irwin, 1995; Reif and Schmitt, 1997). Many radical parties have a particularly strong anti-European position (Hooghe et al., 2002), contributing to the politicisation of Europe (Grande and Hutter, 2016; Halikiopoulou et al., 2012). Considering that they are not internally split on European stances as many mainstream parties are, they systematically perform better in EP elections (Ferrara, 2004). With voters being less supportive of European integration than mainstream political elites are, radical parties representing anti-European attitudes and making European issues salient enjoy an advantage (van der Eijk and Franklin, 2004; van Egmond, 2007). Third, EP elections augment challengers’ prospects based on the proportional electoral system applied. While some of the member states use a majoritarian electoral system nationally, as of 1999, all European representatives are elected proportionally. Thus, challenger parties enjoy also ‘mechanical’ advantages in European elections as opposed to some national contests (Oppenhuis et al., 1996).

In sum, the distinct subordinated character and the salience of European policies prompt a different voting rationale among voters who turn out in European elections.¹ Many of these voters express their dissatisfaction with their national governments or align their vote closely with their policy preference (being European, domestic, or Eurosceptic in nature). For both reasons, challenger parties enjoy advantages in the EP elections, which may be further amplified by the permissive electoral system in EP elections. In the following, I contend that the benefits for challenger parties in the European arena also boost their domestic prospects. European electoral successes may heighten a challenger party's visibility in the domestic arena – in particular, if the temporal proximity between both elections increases the domestic salience of European integration.

Domestic momentum and the effect of electoral timing

I argue that challenger parties gain momentum (Holbrook, 1996: 130; Mutz, 1997) through successful performance in EP elections. Virtually the same national parties and major national actors contest both elections (van der Eijk and van der Brug, 2007: 7), even if the results in the respective elections are determined by a different voting rationale. Therefore, competing parties and media may consider a challenger party's success in the EP arena an indicator for its likely next national performance. Success of a challenger party in the second-order arena leads to increased national media attention, a heightened domestic visibility of the party, and greater attention levels by party elites (Oppenhuis et al., 1996: 302). This reaction of media and competitors is particularly pronounced if the supranational performance has domestic significance in potentially polarising national party competition (Vasilopoulou, 2017).

A strong EP performance coupled with an increased visibility of the party may heighten the chances that individuals vote for the party in the next national election. Research on United States (US) primary elections shows that information on mass support for a certain candidate does not only impact strategic vote considerations (Zech, 1975), but even evokes attitudinal change among some voters (Mutz, 1997). Confronted with information on high support levels for a certain candidate, so-called ‘consensus cues’, individuals re-evaluate the candidate based on this information. They rehearse their political views in light of the arguments that they deem explanatory for the high mass support levels. Importantly, this process involves priming of the perceived others’ political views and cognitive engagement with arguments that ‘would not otherwise have come to mind’ (Mutz, 1997: 105). After successfully competing in the EP elections, a challenger party and its policy positions are primed in the minds of voters.

Yet, the proposed effects presume that the EP election is cognitively available to voters and national party actors. European politics, however, tend to take place in the shadow of national politics (Beaudonnet and Franklin, 2016). Only when the temporal distance between both elections is short, European issues enjoy some prominence in national elections. Analysing the impact of exogenous events on EU news coverage across seven EU members, Boomgaarden et al. (2010) find that media coverage increases strongest during the EP elections and the following installation of a new EU commission. Rauh (2015) points towards greater levels of domestic politicisation of EU affairs in parliamentary debates around the period of national and European elections. The character of the EP election as exogenous event prompts partisan competition and draws domestic actors’ attention to this issue, resulting in a potential contention around the issue. Media and party elites are more attentive to the supranational contest when the two elections are close in time (Oppenhuis et al., 1996: 301; Somer-Topcu and Zar, 2014), which should increase the momentum that challenger parties draw from their European success. Temporal proximity between the two elections encourages evaluations and political judgements of challenger parties based on the information pertaining to this European campaign. This should increase the chance that individuals base their vote decision on the ‘consensus cues’ taken from the supranational campaign. Importantly, voters gain such cues and may accordingly rehearse their domestic vote choice irrespective of their actual participation in the EP elections, which is important in light of the low turnout levels at those secondary contests.

Consequently, I posit that the momentum effects of success in the EP arena are greatest when the two elections are close in time. Domestic campaigns that coincide with the event of EP elections are more permissive to the issue of European integration and authority transfer to the supranational level. I expect that the potential for spillover of electoral success depends on the domestic attention levels towards the European performance of a challenger party and on the general salience of European integration during a national campaign. Both the former and the latter are greatest when the temporal distance between both elections is small.

H1: The higher the vote share of a challenger party in the EP elections, the greater the increase in national electoral gains.

H2: The closer in time national and EP election take place, the stronger the effect of the vote share in the EP election on the increase in national electoral gains.

Design and data

If EP elections foster the success of challenger parties, a strong EP electoral result should be associated with an increase in the national performance, in particular if the two elections are close in time. To test this, I create a dataset including the national election results of European member states² since the first EP election on 10 June 1979, the respective European election results, the dates of both elections, and the temporal distance between them (Döring and Manow, 2018.

Electoral results and the European cycle

The analysis considers the national performance of all challenger parties in European member states starting from 1979. Challenger parties are broadly defined as non-mainstream parties, i.e. green, radical left and populist radical right party actors (e.g. Hernández and Kriesi, 2016). The classification follows expert surveys (Bakker et al., 2015; Benoit and Laver, 2006; Castles and Mair, 1984; Huber and Inglehart, 1995). Given different degrees of party (system) institutionalisation across the European member states and corresponding different lengths of party survival, the main analysis considers the respective party family within each national election of a country as the unit of observation (see for a similar approach Oppenhuis et al., 1996: 291). Table 1 indicates the robustness of the respective results to using the individual parties within each election as the unit of analysis. The position of a national election within a European electoral cycle is the difference in days between the national and the European election divided by the overall length of the European legislation period. EP elections are held every five years, which means that the denominator is approximately equal to 1825 days. As discussed above, European issues are most salient in national campaigns that coincide with an EP campaign. Hence, I expect the effect of EP performance on subsequent national gains to be moderated by the temporal proximity of the two elections.

Cycle = \frac{N E_{t} - E P_{t}}{E P_{t + 1} - E P_{t}}

(1)

where NE_t is the national election date, EP_t the date of the preceding EP election, and

E P_{t + 1}

the date of the next EP election.

Table 1.

Fixed-effects regression results on national vote share by party families.

	Populist	Green	Populist
	radical left	parties	radical right
	Linear	Linear	Quadratic
EP vote	0.636***	0.377***	0.905***
	(0.078)	(0.081)	(0.133)
Cycle	2.360*	1.840**	0.767
	(1.223)	(0.809)	(4.798)
Cycle²>			2.106
			(5.070)
EP vote × cycle	−0.412***	−0.209**	−2.302***
	(0.130)	(0.100)	(0.659)
EP vote × cycle²			2.295***
			(0.739)
Government at EP election	2.432	0.535	2.469
	(2.503)	(0.831)	(1.664)
Midterm EP	0.655	−0.799**	1.278*
	(0.597)	(0.365)	(0.726)
Unemployment	0.407***	0.009	0.272***
	(0.088)	(0.051)	(0.091)
EP elections participated	1.254**	0.456	0.405
	(0.487)	(0.295)	(0.533)
EP threshold	3.003***	−0.314	−0.600
	(0.520)	(0.324)	(0.557)
National district magnitude	−0.284	0.079	0.775
	(0.933)	(0.569)	(1.066)
European district magnitude	1.036	0.468	−0.050
	(0.791)	(0.484)	(0.850)
p-value Wald test (LIE)	0.789	0.430	0.001
Decade fixed-effects	✓	✓	✓
Robustness of (non-) significance of interaction term
Cluster robust SE	✓	x	✓
Pairs cluster bootstrapped SE	✓	x	✓
Jackknife parties	✓	x	✓
Party fixed-effects	✓	x	✓
BIC	901.136	730.484	936.139
N	174	174	174

*p < 0.10, **p < 0.05, ***p < 0.01. Standard errors in parentheses. Robustness of the interaction effect: pairs cluster bootstrapped SE to account for small cluster size. Jackknife reruns analysis while omitting one party each regression. Party-fixed effects use party-specific dummies instead of country dummies.

There are some factors that facilitate challenger parties' success, which vary across the 27 European member states in the analysis, particularly the degree of party system institutionalisation or authoritarian legacies (Kriesi, 2016). This kind of heterogeneity between countries may systematically relate to challenger parties' success on both the European level and the national level and bias the point estimates. Country-fixed effects hold observed and unobserved variance between countries constant if this variance is stable over time. We assume that authoritarian legacies and the institutionalisation of a party system are time-constant after conditioning on decade dummies. All country-specific covariates that do not vary within decades and might both influence the vote share of a challenger party in the national and in the European context are controlled for by cluster ‘de-meaning’ the data in the fixed-effects model. The model estimates the national performance of challenger parties as a function of the interaction between the EP electoral result, the cycle variable, and the respective constitutive terms. The interaction coefficient can thus be interpreted as the impact of the EP performance moderated by the position of the cycle variable. The model equation can be formalised as follows:

{\overset{··}{y}}_{it} = β_{0} + β_{1} {\overset{··}{x}}_{is} + β_{2} {\overset{··}{c}}_{it} + β_{3} ({\overset{··}{x}}_{is} * {\overset{··}{c}}_{it}) + β_{k} {\overset{··}{z}}_{it} + {\overset{··}{u}}_{it}

(2)

for i = 1…27 EU member states; t = 1…n national elections in EU member state i, and s = 1…8 EP elections preceding the national elections; where

\overset{··}{y} = y_{it} - {\bar{y}}_{i}

(correspondingly for x,c,z,u); y = vote share of respective party in national election, x = vote share in EP election, c = position of the national election within a European electoral cycle (see equation (1)), z = vector of control variables. Unit of analysis is the country-election level, regression estimation per party family (variance of the residuals varies across party families).

Figure 1.

Number of national elections following an EP election within the same European cycle.

The vector of control variables includes time-variant covariates that may systematically relate to the performance of challenger parties at both levels of governance. The analysis accounts for the permissiveness of the electoral system by including the logarithm of the average district magnitude in each country's national elections (Johnson and Wallack, 2012) and the logarithm of the average district magnitude in the EP elections. While the national electoral thresholds have not changed within EU member states,³ the model takes the country-specific EP electoral threshold into account, which in some countries is not time-constant. The literature shows that the extent of ‘second-orderness’ of a European election varies depending on whether or not the EP election was a ‘midterm’ election and on the experience that countries have with EP contests, i.e. the number of EP elections a country has participated in (Hix and Marsh, 2011: 6; Marsh, 1998: 597). The character of the EP elections and the legislative power of the EP itself have changed since the introduction of the EP elections in 1979. To account for these changes and for other time-specific unobserved heterogeneity within the observation period, I introduce four decade dummy variables. Yet, they do not reach conventional levels of statistical significance in any of the models. To confront the fact that some of the challengers might themselves get punished in the EP elections if they were in government before, an indicator variable measures whether the parties were part of the national executive at the time of the respective EP election. The variable, however, remains insignificant throughout all models (see Table 1). The results are robust to excluding all challenger populist left, green and populist right actors that have been in government (see the Online appendix). Finally, the model controls for the state of the economy (unemployment rate) that might contribute to a high number of protest or anti-government votes in EP and national elections (International Monetary Fund, 2016).

The analysis consists of 174 national elections of EU member states. EP elections take place every five years, while most European member states hold elections every four years. Every fourth observation in the data (24.71%) refers to the same EP election result as the previous country-specific observation (see Figure 1). Yet, while the EP vote share is equivalent for these cases, the cycle values are necessarily different from each other. This introduces greater variation among these observations and renders the central interaction term of interest independent from the previous observations. The cycle variable is very equally distributed (L-Kurtosis: 0.0116), facilitating the interpretation of the conditional marginal effects.⁴ The analysis of the central interaction term proceeds as follows. As suggested by Hainmueller et al. (2016), I first test whether the moderating effect of the cycle variable follows the linear interaction effect (LIE) assumption, which is relevant to assess hypothesis H2, positing that the marginal effect of the EP vote share is conditional on the temporal proximity between both elections. The functional form of the conditioning effect of the cycle variable does not necessarily need to be linear. To test the LIE assumption, I visualise the conditional marginal effects within four equally spaced intervals of the cycle variable using the mean conditional marginal effect of the EP vote share in each interval. To obtain these estimates, the proposed binning estimator by Hainmueller et al. (2016) jointly fits the central interaction to all four individual intervals, while allowing the marginal effects to freely vary within each interval. A simple Wald test statistic reports whether the linear interaction model and the binning model are statistically equivalent. Based on the results of the Wald test, I present the respective country fixed-effects regression results with the corresponding polynomial specification of the cycle variable. The Online appendix includes semi-parametric kernel smoothed estimates to allow for a fully flexible functional form of the marginal effect of the EP vote share with respect to the position in the electoral cycle. Those semi-parametric estimates further support the respective lower and higher-order polynomial specifications reported in the main analysis.⁵ The marginal effect plots show a histogram at the bottom of the figure to help readers assess whether the estimates are supported by data of the moderating cycle variable.

Results

For Green parties, we find a linear interaction effect. The p-value of 0.43 indicates that the flexible binning estimates are statistically equivalent to a simple linear interaction model (see column 2 in Table 1).

The conditional marginal effect size is substantively speaking rather small (see Figure 2). A one-percentage point increase in the European arena improves a Green party's national result only by a maximum of 0.37 percentage points when the national election follows very shortly after an EP election. Yet, the interaction term of the European result and the temporal distance to the EP election is not robust to using bootstrapped or cluster robust standard errors. It also turns insignificant when using party-fixed effects and when jackknifing parties. This indicates that the European result does not serve as a domestic ‘marker’ for these party actors. Scholars have argued that voters are more likely to defect from their national vote in the supranational elections by switching to Green parties if they prefer the environmental issue to be instituted at the EP level (Carrubba and Timpone, 2005: 273; Gabel, 2000). For the same underlying reasons, they might not be inclined to cast a congruent vote at the next national election even if they just had supported a Green party at the previous EP election. The result suggests that the (transnational) policy agenda of Green party actors mitigates a spillover. While green parties’ policy platform may lend itself well for a ballot on the European level, supranational success of these actors contributes only little to their domestic significance. The results show that the performance of Green parties in the European arena does not encourage bandwagon effects in the next domestic electoral contest. On the one hand, this might be because of their environmental policy-agenda, which voters perceive to be located in the supranational arena as argued in the previous literature. On the other hand, Green parties’ European success may also not attract enormous national attention because of their mostly non-radical policy stances.

Figure 2.

Marginal effect of Green vote share in EP election on subsequent national election moderated by the position of the national election within the European cycle.

For radical right party actors, in contrast, we find considerable empirical support for a non-linear conditional marginal effect of the EP election result on national gains. The binning estimates indicate that the cycle does not monotonically moderate this marginal effect, but rather follows the u-shape of a second-order polynomial (see Figure 3). Relying on the Wald test, we reject the null that a naïve linear interaction model and the binning estimates are statistically equivalent (p-value: 0.001). As opposed to Green parties, the effect size of the electoral spillover is also substantively large. During national campaigns that are close in time to an EP election, a strong second-order result provides the populist radical right with domestic advantages. Whenever the distance to an EP election is less than a year, those party actors substantially benefit from a one-percentage point increase in their European fortune by nationally gaining close to the equivalent (around 0.8 percentage points). Yet, if the temporal distance to an EP election is large and a national election falls in the middle of a European electoral cycle, a populist radical right party retrieves only small marginal gains out of its European success (around 0.3 percentage points, comparable to the size of the spillover for Green parties).⁶

Figure 3.

Marginal effect of Populist Radical Right vote share in EP election on subsequent national election moderated by the position of the national election within the European cycle.

The temporal variation in the spillover effect suggests that the salience of European issues in domestic campaigns brings to the fore a ‘highly symbolic issue that fits [radical right parties'] traditionalist-communitarian ideology’ (Bornschier, 2010: 63). When the EP contest comes close in time to a national election, the radical right can successfully mobilise their opposition against the European project in the domestic arena. For the populist radical right, the empirical results give support to hypotheses H1 and H2. The closer the temporal distance between a first-order and a European second-order election, the higher the chances that a strong EP result of these party competitors leaves an imprint on their national fortunes.

For the radical left, these hypotheses are, in contrast, only partially corroborated (see Figure 4 and column 1 in Table 1). The Wald test of the binning estimate (p-value: 0.789) indicates that the moderating effect of the cycle variable follows a linear pattern. The decreasing effect size over time shows that a strong EP result provides radical left actors with a one-time, quickly evaporating increase in national visibility rather than with a heightened salience of their policy issues even in proximity to the next second-order election. While the radical left is positionally distinctive on the traditional left-right political dimension related to redistributional issues, their positions on the cross-cutting national demarcation vs. European integration dimension are less clear-cut. Thus, success in EP elections might make some radical left actors more visible in the short run, but it may not be likely to change the salience of their core issues in the domestic arena.

Figure 4.

Marginal effect of Populist Radical Left vote share in EP election on subsequent national election moderated by the position of the national election within the European cycle.

The latter finding is only valid for populist radical right party actors and is robust to: (a) the exclusion of single parties from the analysis (jackknife procedure);⁷ and (b) to the estimation of party-fixed effects instead of country-fixed effects to account for unobserved organisational differences between parties that might determine both their EP electoral success and subsequent national gains. The results are also not sensitive to (c) bootstrapping the standard errors to confront a possible overconfidence due to the small cluster size within the sample.⁸ The findings are (d) robust to other correlates of populist radical right success frequently discussed in the literature, which might impact these actors' success on both governance levels, namely the influx of asylum seekers, the turnout rate in a given election, or potential party-strategic advantages for radical right parties determined by the left-right position of the largest conservative mainstream competitor. Finally, the results remain unchanged if those elections that follow a first national election within the same EP electoral cycle and those elections that are held concurrently with an EP election (cycle = 0) are excluded from the analysis. The various robustness tests are reported in the Online appendix. Among the vector of controls, in contrast, most of the variables do not significantly affect challenger parties' electoral fortune across different model specifications. While some of the measures do have a significant effect in the main model reported in Table 1, they fail to reach statistical significance when pair-clustering standard errors and estimating the various alternative model specifications, like party-fixed effects (see the Online appendix). The only variable that stands out among the vector of controls is the unemployment rate that contributes to an increase in populist radical right success on the national level, confirming previous research on the macro-correlates of populist right success across Europe (e.g. Arzheimer, 2009). For the populist radical left, in turn, high unemployment does not feed into electoral success robustly across models.

In the following, I investigate the underlying mechanism driving the spillover effect for the populist right by: (a) showing that the spillover does not stem from similar levels of support for the populist right across the European and the national arena; and by (b) showing that the salience of European integration in domestic campaigns drives the spillover.

Congruence of voters' party preferences across arenas?

To identify a spillover effect of EP electoral success, I propose a placebo-test assessing whether both national and EP results are affected by the same unobserved factors rather than by European success feeding into national success. To the extent that election results measure voters' party preferences and a party's current popularity, the closer to (or further apart from) each other two elections take place, the greater (lower) the association between the results to be expected. The voluminous empirical evidence from the second-order literature suggests that different voting calculi apply to both kinds of elections, which is supported by parties' different results at concurrent national and European elections. Yet, if we still assume that voters' party preferences are partially congruent across the national and supranational arena, an alternative explanation for the cyclical spillover effect is given by potential similar popularity levels of populist radical right parties in the EP and national elections.⁹ If this were the case, however, we should find the same cyclical pattern when predicting the success of radical right parties in the European election (dependent variable) conditional on the interaction between temporal distance to the last national election and the respective electoral result.

However, the placebo-test does not give any support to a similar cyclical spillover effect. The binning estimate first suggests a linear functional form of a national spillover to the European area (p-value of the LIE assumption: 0.428). Second, the interaction term between the position of the European election in the national election and the national vote share is insignificant (and marginally positive). The full results of this placebo-test along with the respective figures are reported in the Online appendix.

Supposing that the spillover from the second-order arena to the domestic one were only driven by a high congruence of voters' preferences across governance levels, we should, however find the same decreasing strength of association the greater the temporal distance between both elections. Yet, the placebo-test shows a different pattern. This indicates that the institution of the EP elections and the salience of European integration in itself fosters the spillover of populist radical right success rather than a high correlation between voters' preferences spanning the different governance levels.

EP elections as quasi-exogeneous event

It is worth to exploit the quasi-exogenous nature of the EP contest as a political event. To date, any given member state has consistently participated in the EP elections, rendering the existence of this political institution and its timing largely exogenous to party actors' strategic short-term influence. Hence, a conservative test to assess whether the institution of the EP elections in itself prompts the salience of populist radical right parties' issues and fosters their visibility in the national arena, is given when we reassess the mere impact of the EP elections as an event, not considering a party's actual performance therein.

When re-estimating the model and only including the continuous cycle variable as predictor of a challenger party's national success instead of the interaction term, the quadratic cycle variable maintains its statistical significance for the radical right (see Table 2).¹⁰ The coefficient indicates the same u-curved relationship between temporal proximity of the two elections and marginal gains in the national vote shares of populist radical right parties. Exploiting the quasi-exogeneity of the temporal proximity to the EP contest, we find that the event of the EP election in itself augments the electoral prospects for populist radical right parties. This is not the case for populist radical left and green parties whose national performance is not significantly affected by a temporal proximity to the European contest. Previous research shows that individuals feel particularly negative about the EU in the year of an EP election (Beaudonnet and Franklin, 2016). Hence, when the European campaigns coincide with the national electoral campaigns, the salience of Europe in domestic politics accentuates radical right actors’ electoral gains. It appears that such domestic contests are particularly favourable to increase the radical right's mobilisation on anti-EU stances and ‘pulling voters’ who might not have supported them on the basis of their left-right concerns (van der Eijk and Franklin, 2004: 47). The European contest seems to offer party actors who represent both the anti-European and authoritarian-nationalistic attitudes of many voters a permeable forum to politicise these issues domestically. The pace of further deepening and widening of the EU has not always been accompanied by an increase of citizens’ level of support for further integration. On the contrary, citizens’ Eurosceptic attitudes have increased over the years (Eichenberg and Dalton, 2007). Whenever the European and the national campaigns coincide, issues relating to further European integration make their way into the domestic arena. This affects party competition and creates opportunities for such challengers who favour demarcation as opposed to further integration, namely actors from the populist right-wing end of the spectrum (Kriesi, 2007).

Table 2.

Fixed-effects regression results on national vote share by party families.

	Populist	Green	Populist
	radical left	parties	radical right
	Linear	Linear	Quadratic
Cycle	0.831	−0.021	−11.591**
	(1.052)	(0.720)	(5.477)
Cycle²			15.095**
			(5.811)
Controls	✓	✓	✓
Decade fixed-effects	✓	✓	✓
Robustness of (non-) significance of interaction term
Cluster robust SE	✓	✓	✓
Pairs cluster bootstrapped SE	✓	✓	✓

*p < 0.10, **p < 0.05, ***p < 0.01. Standard errors in parentheses. Robustness of interaction effect: pairs cluster bootstrapped SE to account for small cluster size.

Conclusions

Recently, scholars contended that the European direct elections are not working as elections ‘are supposed to perform’. They are second-rate in failing to achieve their supposed objectives – to provide direct policy consequences (Franklin, 2016). This different character of the European contest leaves the Union merely with the intended consequence of decreasing its ‘democratic deficit’. Following the seminal work of Reif and Schmitt (1980), there is a voluminous literature on the character of the supranational contest, the voting calculus, and the policy issues involved therein. Yet, ‘in reality, we find influences running in both directions’ (van der Eijk et al., 1996), and the secondary elections themselves impact domestic party systems. This analysis highlights that a ‘vote against Europe’, particularly once made visible in European elections, may decisively shape domestic elections. The supranational contest offers populist radical right actors an opportunity structure to mobilise voters based on their antagonism towards the elite-consensus on European integration. By shifting the focus to the national arena, this paper first shows that the direct second-order elections have important national consequences. While anecdotal evidence holds that the EP elections provided parties like AfD, Front National or the Sweden Democrats with the first favourable opportunity for gaining momentum and translating their success into national power, the present analysis offers a systematic analysis of such spillover effects across all European member states and national elections since 1979. Second, the study disentangles the mechanisms behind these electoral spillover effects, corroborating the idea that populist radical right parties draw crucial momentum from EP success. If national and European elections are close in time, the salience of European integration boosts the domestic electoral prospects of radical right parties.

Future research is necessary to explore the potential variation in the European spillover effects across different party systems. Mainstream parties’ responses to European success of a populist right challenger, the policy-influence of those actors within a country, and country-specific variation in the evolution of saliency of European integration might crucially mediate the cyclical spillover effect. This might put in motion or prevent further spillover effects from the national to the European arena. Future research should also address the underlying micro-level mechanisms. Individual-level panel data across European countries could help to assess whether individuals who turn out for a challenger party in the EP elections are also more likely to cast a similar ‘habitual’ ballot in the following national contest. If this were the case, the EP elections would contribute to individual partisan re- or dealignment, working as a ‘virtual pump’ that may pull impressionable voters from mainstream parties (Franklin, 2016). Dinas and Riera (2018) show that individuals who first became eligible for a European election are more likely to support a small party than individuals who became eligible for a national election, arguing that the act of voting socialises individuals into such voting patterns (Dinas and Riera, 2018). In light of the comparatively low turnout levels in EP elections, such habitual voting may only partially account for the electoral spillover of populist radical right success. The experimental evidence from the bandwagon literature and the empirical results in this paper support the idea that also individuals who did not participate in the European contest are encouraged to cast a ballot for a populist radical right party after its success in the supranational arena, based on the consensus cues they take from mass support levels in the supranational arena. This hypothesis should be empirically addressed by future research.

The salience of European integration seems decisive in explaining the populist right spillover effects to the domestic arena. This salience is augmented when the national election occurs in close temporal proximity to the EP contest. Neither green nor populist radical left actors are able to similarly capitalise on European success. While populist radical right parties do play a part in politicising Europe, they are also among the most hostile actors towards further European integration. In view of their opposition to an integrative Europe, it seems rather ironic that the supranational contest fosters the ascendency of just these opponents of the European idea.

Supplemental Material

eup-17-1360-File002 -Supplemental material for Do European Parliament elections foster challenger parties' success on the national level?

Supplemental material, eup-17-1360-File002 for Do European Parliament elections foster challenger parties' success on the national level? by Julia Schulte-Cloos in European Union Politics

Footnotes

Acknowledgements

I would like to thank the editors and two reviewers of EUP for excellent comments on an earlier draft. I would further like to thank Elias Dinas, Cees van der Eijk, Liesbet Hooghe, Hanspeter Kriesi and Hermann Schmitt for very insightful discussions about the arguments presented here.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Supplemental material

Supplemental material for this article is available online.

Notes

References

Arzheimer

(2009) Contextual factors and the extreme right vote in Western Europe, 1980–2002. American Journal of Political Science 53(2): 259–275.

Bakker

de Vries

Edwards

et al. (2015) Measuring party positions in Europe: The Chapel Hill expert survey trend file, 1999–2010. Party Politics 21(1): 143–152.

Beaudonnet

Franklin

(2016) The negative effect of elections to the European Parliament. In: van der Brug

de Vreese

(eds) (Un)intended Consequences of EU Parliamentary Elections, New York, NY: Oxford University Press, pp. 195–214.

Benoit

Laver

(2006) Party Policy in Modern Democracies, London: Routledge.

Boomgaarden

et al. (2010) News on the move: Exogenous events and news coverage of the European Union. Journal of European Public Policy 17(4): 506–526.

Bornschier

(2010) Cleavage Politics and the Populist Right: The New Cultural Conflict in Western Europe, Philadelphia: Temple University Press.

Carrubba

Timpone

(2005) Explaining vote switching across first-and second-order elections evidence from Europe. Comparative Political Studies 38(3): 260–281.

Castles

Mair

(1984) Left–Right political scales: Some “expert” judgments. European Journal of Political Research 12(1): 73–88.

Dinas

Riera

(2018) Do European Parliament elections impact national party system fragmentation?. Comparative Political Studies 51(4): 447–476.

10.

Döring H and Manow M (2018) Parliament and government composition database (Parlgov). Information on parties, elections and cabinets in modern democracies. Development version. Available at: http://www.parlgov.org/.

11.

Eichenberg

Dalton

(2007) Post-Maastricht blues: The transformation of citizen support for European integration, 1973–2004. Acta politica 42(2–3): 128–152.

12.

Esarey J and Menger A (2016) Practical and effective approaches to dealing with clustered data. Unpublished manuscript.

13.

Ferrara

(2004) Get your act together: Party performance in European parliament elections. European Union Politics 5(3): 283–306.

14.

Franklin

Hobolt

(2011) The legacy of lethargy: How elections to the European Parliament depress turnout. Electoral Studies 30(1): 67–76.

15.

Franklin

van der Eijk

Oppenhuis

(1996) The institutional context: Turnout. In: van der Eijk

Franklin

(eds) Choosing Europe? The European Electorate and National Politics in the Face of Union, Ann Arbor: University of Michigan Press, pp. 306–331.

16.

Franklin

(2016) EP elections as stepping-stones to Eurosceptic party success. In: Nielsen

Franklin

(eds) The Eurosceptic 2014 European Parliament Elections, London: Palgrave Macmillan, pp. 223–238.

17.

Gabel

(2000) European integration, voters and national politics. West European Politics 23(4): 52–72.

18.

Grande

Hutter

(2016) Is the giant still asleep? The politicisation of Europe in the national electoral arena. In: Hutter

Grande

Kriesi

(eds) Politicising Europe, New York: Cambridge University Press, pp. 90–111.

19.

Hainmueller J, Mummolo J and Xu Y (2016) How much should we trust estimates from multiplicative interaction models? Simple tools to improve empirical practice. In: The summer meeting of the society for political methodology, Rice University, Houston, TX, 14 July 2016.

20.

Halikiopoulou

Nanou

Vasilopoulou

(2012) The paradox of nationalism: The common denominator of radical right and radical Left Euroscepticism. European Journal of Political Research 51(4): 504–539.

21.

Hernández

Kriesi

(2016) The electoral consequences of the financial and economic crisis in Europe. European Journal of Political Research 55(2): 203–224.

22.

Hix

Marsh

(2007) Punishment or protest? Understanding European parliament elections. Journal of Politics 69(2): 495–510.

23.

Hix

Marsh

(2011) Second-order effects plus pan-European political swings: An analysis of European parliament elections across time. Electoral Studies 30(1): 4–15.

24.

Hobolt

Spoon

J-J

Tilley

(2009) A vote against Europe? Explaining defection at the 1999 and 2004 European parliament elections. British Journal of Political Science 39(1): 93–115.

25.

Holbrook

(1996) Do Campaigns Matter?, Thousand Oaks, CA: SAGE Publications.

26.

Hooghe

Marks

Wilson

(2002) Does Left/Right structure party positions on European Integration?. Comparative Political Studies 35(8): 965–989.

27.

Hosking

JRM

(1990) L-moments: Analysis and estimation of distributions using linear combinations of order statistics. Journal of the Royal Statistical Society. Series B (Methodological) 52(1): 105–124. .

28.

Huber

Inglehart

(1995) Expert interpretations of party space and party locations in 42 societies. Party Politics 1(1): 73–111.

29.

International Monetary Fund (2016) World Economic Outlook Database. Available at: http://www.imf.org/external/pubs/ft/weo/2016/01/weodata/index.aspx.

30.

Irwin

(1995) Second-order or third-rate? Issues in the campaign for the elections for the European Parliament 1994. Electoral Studies 2(14): 183–199.

31.

Johnson

Wallack

(2012) Electoral systems and the personal vote, Harvard Dataverse, V1, Available at: https://hdl.handle.net/1902.1/17901.

32.

Kriesi

(2007) The role of European integration in national election campaigns. European Union Politics 8(1): 83–108.

33.

Kriesi

(2016) The politicization of European integration. Journal of Common Market Studies 54(1): 32–47.

34.

Markowski

(2016) How European elections affect national party systems. In: van der Brug

de Vreese

(eds) (Un)intended Consequences of EU Parliamentary Elections, New York, NY: Oxford University Press, pp. 125–147.

35.

Marsh M (1998) Testing the second-order election model after four European elections. British Journal of Political Science 28(4): 591–607.

36.

Mutz

(1997) Mechanisms of momentum: Does thinking make it so?. The Journal of Politics 59(1): 104–125.

37.

Oppenhuis

van der Eijk

Franklin

(1996) The party context. In: van der Eijk

Franklin

(eds) Choosing Europe? The European Electorate and National Politics in the Face of Union, Ann Arbor: University of Michigan Press, pp. 287–305.

38.

Rauh

(2015) Communicating supranational governance? The salience of EU affairs in the German Bundestag, 1991–2013. European Union Politics 16(1): 116–138.

39.

Reif

Schmitt

(1997) Second-order elections. European Journal of Political Research 31(1–2): 109–124.

40.

Reif

Schmitt

(1980) Nine second-order national elections – A conceptual framework for the analysis of European election results. European Journal of Political Research 8(1): 3–44.

41.

Schmitt

Mannheimer

(1991) About voting and non-voting in the European elections of June 1989. European Journal of Political Research 19(1): 31–54.

42.

Somer-Topcu

Zar

(2014) European parliamentary elections and national party policy change. Comparative Political Studies 47(6): 878–902.

43.

van der Brug

de Vreese

(2016) Introduction. In: van der Brug

de Vreese

(eds) (Un)intended Consequences of EU Parliamentary Elections, New York, NY: Oxford University Press, pp. 1–16.

44.

van der Eijk

Franklin

(2004) Potential for contestation on European matters at national elections in Europe. In: Gary

Steenbergen

(eds) European Integration and Political Conflict, Cambridge, UK. New York: Cambridge University Press, pp. 32–50.

45.

van der Eijk

Franklin

Marsh

(1996) What voters teach us about Europe-wide elections: What Europe-wide elections teach us about voters. Electoral Studies 15(2): 149–166.

46.

van der Eijk

van der Brug

(2007) Introduction: Electoral politics in the European Union and the 2004 Enlargement. In: van der Eijk

van der Brug

(eds) European Elections & Domestic Politics: Lessons from the Past and Scenarios for the Future, Notre Dame, Ind: University of Notre Dame Press, pp. 1–12.

47.

van Egmond

(2007) European elections as counterfactual national elections. In: van der Brug

van der Eijk

(eds) European Elections & Domestic Politics. Lessons from the Past and Scenarios for the Future, Contemporary European Politics and Society. Notre Dame: University of Notre Dame Press, pp. 32–50. .

48.

Vasilopoulou

(2017) British Eurosceptic voting in 2014: Anti-EU or anti-government?. In: Nielsen

Franklin

(eds) The Eurosceptic 2014 European Parliament Elections, London: Palgrave Macmillan, pp. 57–81.

49.

Weber

(2007) Campaign effects and second-order cycles: A top-down approach to European Parliament elections. European Union Politics 8(4): 509–536.

50.

Zech

(1975) Leibenstein's bandwagon effect as applied to voting. Public Choice 21: 117–122.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.08 MB