Harmony and dissonance: Unveiling issue linkages between voters and parties across EU democracies

Abstract

This research delves into the abundant landscape of party positioning data, highlighting a scarcity of interconnected voter and party positioning sources. Leveraging Voting Advice Applications (VAAs), this study explores congruence from a comparative standpoint, capturing political preferences on the same scale and timeframe. Analysing data from two pan-European VAAs, EU Profiler and euandi, congruence is evaluated based on voters' alignments with their preferred party across three issue dimensions. Findings underscore varying linkage strengths, particularly evident in a narrower representation on cultural and gender issues compared to economic ones. The study underscores the advantages of VAA data, urging a shift beyond the left-right continuum for a comprehensive understanding of party-voter congruence in a dynamic political landscape.

Keywords

voting advice applications issue congruence voters European parliament issue dimensions

Introduction

Political parties are channels of expression for citizen preferences and demands. Accordingly, political representation is the ultimate goal in the relationship between parties and citizens (Sartori, 2005). A higher degree of representation would suggest that citizen preferences are better reflected in the political stances of the parties, with spillover effects into increased democratic satisfaction (Dalton et al., 2000). The distance between parties and voters is used as a measure of the quality of the political representation process and ultimately, of the good functioning of a democracy.

The match between voters and their representatives lies at the heart of a good democratic process. The last decades of political representation studies present varied results and different portrayals of the linkage between citizens and political elites. On one hand, there is a line of studies that shows that congruence on the left-right (hereafter RiLe) dimension between voters and parties is high (Dalton, 1985, 2013; Powell, 2000; Thomassen, 1994), suggesting that the representation process is successful. On the other hand, various strands of representation studies show weakening ties between citizens and their representatives on the ideological RiLe, but also on narrower policy dimensions (Bremer, 2018; Clements et al., 2018). Replicating findings from one of the first studies to compare issue congruence across EU countries (Dalton, 1985), Dalton (2017) shows that while there are high levels of congruence on the RiLe scale between citizens and their preferred party, the strength of this linkage varies across issue dimensions. While the linkage is strong on policies that belong to the RiLe dimension, it is weaker on policies related to immigration and gender issues. This research builds on these findings and explores issue congruence in contemporary democracies, extending the analysis to all countries of the European Union using VAA-generated data.

The aim of this study is to investigate the linkage between parties and their supporters, in EU nations, beyond the RiLe continuum. Congruence is assessed as the distance between the position of the party and the position of the party supporters on three issue dimensions: cultural, economy, and gender. The study will also look into the determinants of issue congruence, focusing on the features of the political parties, namely ideological leaning, governmental status, and the niche/mainstream party dichotomy while accounting for the age of the party and features of the electoral system. Congruence is referred to as the match between prospective voters in the elections to the European Parliament (hereafter EP) and their most preferred party, on a set of 11 policy issues. The data were collected with the help of two pan-European VAAs, the 2009 EU Profiler and 2014 euandi, in the months leading up to the elections.

EP elections are not used here to inspect the level of support for the EU. Similar to Dalton (2017), they are used as an event when a large number of countries hold elections simultaneously and voters reconsider their policy preferences. Although treated as second-order elections, voters prioritize their true party preferences (Reif, 1984), and make choices based on the national context rather than on the parties’ stances on EU issues Clark and Rohrschneider (2009). Recent findings on European identity show that European attachments exist in combination or as an alternative to national identities (Dalton, 2021). These identities overlap with partisan attachments, creating a solid base for the study of congruence with data from EU elections.

This article proceeds in five steps. First, we discuss the political representation literature that lays the theoretical foundations of this study. Second, we present the empirical aspects of this work, introducing the data, followed by the measures of congruence on three issue dimensions. Third, we consider the effect of institutional factors. Fourth, we inspect the effect of the determinants of issue congruence at party level. The last section presents the findings and implications for congruence studies and VAA research.

Studying issue congruence

In the tradition pioneered by Huber and Powell (1994, 2000) congruence is assessed as the distance between the position of the citizens and the representatives on the RiLe scale . The use of the RiLe scale in a comparative perspective is based on two assumptions. First, the meaning of RiLe is the same across national contexts. Second, asking voters to define their policy preferences is too demanding of an exercise (Lax and Phillips, 2012). While it shaped the study of representation, the use of RiLe to assess congruence presents a few caveats. First, RiLe is an ambiguous concept, with its use based on the assumption that in most modern democracies, the left and right represent a unidimensional discourse that is familiar to voters (Powell, 2000). However, while congruence on the RiLe dimension captures the degree of ideological consistency across policy issues, it says very little about citizens’ views on issues themselves (Broockman, 2016). While the process of political representation might prove successful on the RiLe dimension it does not prove that all relevant policy dimensions are constrained by it (Thomassen, 2020). Second, the meanings of RiLe differ across electoral contexts (Benoit and Laver, 2006). Comparing values across 20 European nations, Piurko et al. (2011) show that for the most part, both left and right have coherence, but less so in the post-communist democracies of Eastern Europe, making its use in comparative research highly problematic. Third, RiLe fails to reflect the wide array of issues that shape modern-day politics. Traditionally, the issues embedded in RiLe were of economic nature. More recently, major societal changes have pushed new issues onto the political agenda, such as immigration, same-sex marriage, or the legalisation of the use of soft drugs, that shape the political discourse. With the shift to issue voting, political parties try to dominate the electoral competition with issues they own, such as economic, environmental, or immigration policies, and the nature of the electoral competition forces parties to take a stance on issues they would have previously ignored Thomassen et al. (2005).

For decades of representation research, self-positioning on the RiLe continuum was the only survey component that allowed cross-national comparisons (Blais and Bodet, 2006; Golder and Stramski, 2007). A unidimensional measure does not accurately describe the modern-day political space Dalton, (1985). Even when issues are separated between RiLe and TAN/GAL, the latter fails to form a coherent dimension when viewed from a demand-side perspective (Wheatley and Mendez, 2021). To address this, we focus on issue, rather than ideological congruence.¹

This work contributes to the political representation literature in two ways. First, it uses issue congruence, expanding on the understanding of a political landscape that is continuously evolving. Second, it introduces a new way of studying issue congruence, using VAA-generated data. The measure of agreement between voters and parties is based on a large set of issues, moving away from congruence on the RiLe dimension. We address Golder and Stramski (2007)’s critique that research fails to consider the way in which congruence is most appropriately conceptualised given the research goals and that the process is driven by data availability and convention. Additionally, the reliability of congruence estimates is positively affected by the number of statements and the diversity of the topics included (Lesschaeve and Padmos, 2023).

The measure of congruence developed here builds on the concept of representative deficit, which illustrates the citizens’ political preferences that are not mirrored by the parties’ electoral promises. Introduced by Alvarez et al. (2014), the concept is used in work using VAA-generated data (Dinas et al., 2014; Lacey et al., 2014; Bright et al., 2020; Costello et al., 2021). We use two measures of congruence: centrism, and responsiveness Dalton (1985, 2017). This work adds to the representation literature by introducing a straightforward, cost-effective way of investigating congruence, using VAA data.

Measuring issue congruence

Voting advice applications data

While there is no gold standard in assessing congruence, decades of representation studies identify several desirable characteristics: the alignment of political preferences between citizens and representatives to be measured on the same issues, scale, and at the same point in time. Data used for analysis was collected through two pan-European VAAs developed for the European Parliament Elections of 2009, respectively 2014 (Reiljan et al., 2020b). EU Profiler included 255 political parties from 27 EU nations and euandi included 242 parties from 28 countries. Based on their self-positioning on 30 issue statements, the respondents receive a rank list of matches with the political parties. Euandi 2019 could not be used because it does not contain any information on the respondents’ propensity to vote. VAAs generate a convenience sample, with younger, better educated, highly interested in politics respondents (Marschall, 2014) (see Appendix).

While self-selection bias is the main drawback of VAA-generated data, the advantages outweigh the limitations. Decades of scholarly research in representation studies show that the main limitations relate to the small number of common statements, different question wording, and different measurement scales, as well as data being collected at different points in time, especially when the party positioning relies on expert surveys (Powell, 2009). VAA data offers a generous battery of questions that capture party and voter preference on the same issues, same scale, at the same point in time, across EU member states (Garzia and Marschall, 2019). These characteristics help alleviate a key problem in empirical work on congruence, known as Differential Item Functioning (DIF), used to assess whether survey scales have different measurement properties for different groups of respondents (Hare et al., 2015). As an alternative to procedures that can estimate and correct for voters’ misperceptions, De Angelis (2020) encourages the use of VAA data for the study of congruence, as it can reduce DIF and produce comparable latent policy scores.

Measures of issue congruence

To create the party-party supporter dyads, we used the user’s highest PTV score, reported on a 10-point scale. Where the user indicated the same PTV value for more than one party, we selected the party with the highest agreement score for each user². For reliability purposes, we kept only the parties with at least 50 party supporters. The final data sets include 223 political parties and 889,127 responses for EU Profiler and 163 parties and 413,726 for euandi³. To be able to compare the results with previous studies on issue congruence (Dalton, 2017), we limited the analysis to 11 questions. Comparisons between the two elections have to be treated with caution since the data is not representative of the countries’ voting populations. Table 1 outlines issue statements, policy dimensions, and principal component analysis (PCA) results. The left side details political party outcomes: cultural (e.g., immigration, criminal penalties), economic (government spending, income redistribution, private enterprise), and gender-related issues. Similar patterns, aligned with prior research (Dalton 2017), emerge in public preferences: cultural issues dominate, followed by economic and gender concerns.

Table 1.

Issue dimensions.

Issue	Parties						Party supporters
	Cultural		Economy		Gender		Cultural		Economy		Gender
	2009	2014	2009	2014	2009	2014	2009	2014	2009	2014	2009	2014
Criminals should be punished more severely	0.72	0.75	0.25	0.09	0.36	0.28	0.73	0.73	0.19	−0.28	0.10	0.07
Immigrants should accept our culture and values	0.78	0.72	0.18	0.20	0.23	0.27	0.70	0.60	0.14	0.15	0.02	0.25
Restrictions of civil liberties to fight terrorism	0.59	0.39	0.23	0.10	0.34	0.19	0.61	0.32	0.03	−0.42	0.18	0.17
Immigration made more restrictive	0.87	0.85	0.05	0.08	0.12	0.15	0.78	0.75	0.19	−0.02	0.05	0.25
Decriminalisation of soft drugs use	0.30	0.27	0.15	−0.03	0.77	0.81	0.25	0.14	−0.11	−0.07	0.71	0.75
Euthanasia should be legalised	0.24	0.17	0.13	0.18	0.88	0.87	−0.12	−0.08	0.05	0.06	0.84	0.82
Legalisation of same-sex marriages is a good thing	0.57	0.17	0.13	0.18	0.61	0.70	0.27	0.38	0.26	0.06	0.61	0.63
Social programmes maintained, higher taxes	0.30	0.37	0.82	0.74	0.03	0.22	0.17	0.56	0.75	0.41	0.12	0.13
Healthcare services should be privatised	0.08	0.20	0.83	0.79	0.10	0.19	0.09	0.24	0.65	0.64	0.12	0.21
State subsidies for childcare to be increased	−0.04	−0.18	0.60	0.74	0.28	−0.07	0.02	0.03	0.53	0.65	−0.08	−0.04
Government spending should be reduced	0.33	0.45	0.75	0.68	0.05	0.07	0.32	0.65	0.63	0.15	0.02	−0.12
SS loadings	2.92	2.65	2.50	2.30	2.15	2.30	2.29	2.51	1.84	1.87	1.67	1.30
Proportion explained	0.39	0.37	0.33	0.32	0.28	0.31	0.39	0.44	0.32	0.33	0.29	0.23
N EU profiler	230						889,127
N euandi	165						413,726

Source: EU profiler, euandi.

Note: The table presents the results of Principal Component Analysis (varimax rotation and pairwise deletion using the psych package in R) on the parties and voters data sets.

To assess the link between party stances and their supporters’ political preferences, we employ two metrics. Centrism (Achen 1977; Dalton 1985), quantifies the distance between a party’s position on issue j and the mean position of its supporters on the same issue:

c e n t r i s m_{p, j} = | p a r t y p o s i t i o n_{j} - \bar{p a r t y s u p p o r t e r s_{i, j}} |

(1)

A small value indicates proximity to the center of party supporters’ preferences. The second metric, political responsiveness (Birch 1971; Achen 1977), based on the assumption that the preferences of the party supporters should be relative to the preferences of the parties they support, is gauged through the regression equation:

p a r t y p o s i t i o n_{p, j} = a + b * {\bar{X}}_{i, j}

(2)

Here,

{\bar{X}}_{i, j}

is the mean position of party supporters on issue j. A coefficient (b) of 1 signifies a one-to-one match between voter preference shifts and party position shifts. A coefficient greater than 1 implies party polarization exceeding that of supporters, while a coefficient less than 1 indicates party unresponsiveness to shifts in supporter preferences.

Figure 1 depicts congruence between parties and their supporters on the economy dimension in 2009. Each dot represents a party dyad, with party positions on the horizontal axis and averaged supporter positions on the vertical axis. The figure reveals a moderate correlation (r = 0.49), and the regression line (b = 1.12) indicates a 1.12-point shift on the x-axis for each unit increase in party supporter opinion. This aligns with findings from Dalton (2017). Although parties and voters use the same 5-point scale, their positions are not directly comparable due to the comparative nature of this study, potentially influenced by systematic respondent-level bias (Lo et al. 2014).

Figure 1.

Congruence economy dimension 2009. Source: EU Profiler Note: Each dot signifies a party dyad, with the x-axis indicating party positions and the y-axis reflecting the positions of party supporters

Table 2 expands the analysis to include issues beyond the economy. Centrism and political responsiveness measures are presented for 11 issue statements across three dimensions. The voter-party link is weakest on cultural issues and strongest on the economy. Dimension scores are averages of respective issue scores. Notably, restrictions in VAA issue selection (excluding EU themes) will not impact congruence results as scores are dimension-specific.

Table 2.

Issue congruence.

Issue			Responsiveness
	Centrism		a		b		Pearson’s R
	2009	2014	2009	2014	2009	2014	2009	2014
Criminals should be punished more severely	3.62	3.75	−1.94	−2.02	1.46	1.42	0.75	0.76
Immigrants should accept our culture and values	3.53	3.6	−1.81	−2.18	1.49	1.48	0.70	0.81
Restrictions of civil liberties to fight terrorism	2.29	2.41	−0.87	−0.93	1.30	1.26	0.58	0.68
Immigration made more restrictive	3.05	3.13	−1.33	−1.55	1.37	1.36	0.74	0.87
Cultural dimension	3.12	3.22	−1.67	−1.90	1.10	1.33	0.47	0.50
Decriminalisation of soft drugs use	3.16	2.96	−1.32	−1.49	1.60	1.65	0.72	0.79
Euthanasia should be legalised	2.50	2.46	−1.41	−0.45	1.87	1.50	0.78	0.70
Legalisation of same sex marriages is a good thing	2.65	2.48	−1.17	−0.80	1.46	1.33	0.85	0.82
Gender dimension	2.77	2.63	−1.15	−0.68	1.34	1.18	0.39	0.36
Social programmes maintained, higher taxes	2.59	2.74	−1.27	−0.97	1.49	1.34	0.70	0.81
Healthcare services should be privatised	2.33	3	−1.1	−1.49	1.48	1.39	0.68	0.8
State subsidies for creches and childcare to be increased	2.36	1.98	−0.96	−0.9	1.19	1.30	0.41	0.65
Government spending should be reduced to lower taxes	3.32	2.35	−2.11	−0.71	1.51	1.53	0.63	0.77
Economy dimension	2.65	2.51	−0.69	−0.76	1.12	1.25	0.49	0.48

Source: EU profiler, euandi.

Note: Centrism is the absolute difference between the positions of the parties and of party supporters (averaged for each party); responsiveness is based on OLS estimates.

Centrism scores indicate modest agreement between parties and supporters. Discrepancies are most pronounced on cultural issues, especially immigration and crime penalties. High disagreement is observed on the gender dimension, particularly soft drug decriminalization. Economy-related issues, like social programs and tax increases, exhibit somewhat better alignment. Patterns remain consistent across elections, except for the economy dimension. While 2014 sees increased agreement on government spending and state subsidies, there’s less consensus on privatizing healthcare compared to 2009.

Table 2 (right side) displays responsiveness scores. Unstandardized b coefficients >1 indicate parties hold more extreme positions than supporters. This is most evident on euthanasia (b = 1.87 in 2009, b = 1.50 in 2014) in the gender dimension, where parties show the highest responsiveness. Strongest correlation occurs on same-sex marriage legalization (r = 0.85 in 2009, r = 0.82 in 2014). Parties are responsive on cultural and economic dimensions, with consistent patterns across issues. Economy is unique, exhibiting a loose structure in party-supporter stances. Responsiveness patterns are similar across elections and opinion correlations. However, issue-level correlations between supporters and elites surpass those on dimensions.

Figures 1 and 2 Appendix reveal parties represent supporters well on the economy, but party opinions are more diverse. Centrism scores show greater supporter consensus on this dimension, aligning with Dalton (2017), likely due to mapping onto traditional issues familiar to supporters. On the cultural dimension (see Figures A1 and A3 in the Appendix), the linkage is weaker. Despite party responsiveness and strong issue correlations, supporters exhibit slightly larger opinion variation in both elections. Gender dimension congruence is weakest (see Figures A2 and A4, Appendix). High centrism scores indicate party and supporter opinions diverge. Despite strong issue correlations, overall correlation is weakest on this dimension, with notable policy differences not just between parties but also among supporters.

Figure 2.

Congruence economy dimension 2014. Source: euandi Note: Each dot signifies a party dyad, with the x-axis indicating party positions and the y-axis reflecting the positions of party supporters

Predicting issue congruence

In what follows, we aim to explain variation in congruence scores. We consider predictors at both the institutional level, in the tradition of representation studies, and at the party level.

Institutional factors

Due to the homogeneity of electoral rules for EP elections, the effect of PR versus majoritarian systems cannot be examined in this context. Following Dalton (2017), we rely on other features of the electoral systems. One of the most examined characteristics of a party system is counting the number of political parties. Having more party choices translates to higher levels of citizen satisfaction, as the number of channels through which citizens can express their political preferences increases (Dalton, 2008). For this purpose, we use the effective number of political parties (ENEP).

Closely related to the number of political parties are district magnitude and disproportionality (LSq) of the electoral system. Higher district magnitudes offer more opportunities for political parties, therefore more choices for voters (Powell, 2000). Small parties are particularly advantaged under systems with large district magnitude. Low disproportionality stimulates more responsiveness among political parties (Ezrow, 2010). The empirical results are mixed and often contradictory (Blais and Bodet, 2006; Golder and Stramski, 2007), which Eggers and Fouirnaies (2014) attribute to the level of analysis used, whether system- or district-level.

Although more relevant for the study of ideological congruence (Dalton, 2008), we also consider polarization, as the degree of differentiation among political parties in a given political system. Often overlooked in the favour of measures that quantify the parties in an electoral system, polarisation reflects the quality of party competition (Dalton, 2008). Higher dispersion along the ideological continuum should provide voters with distinct electoral choices, leading to more congruence.

We explore differences between post-communist and established Western democracies (Dalton, 2017; Rohrschneider and Whitefield, 2012). Expecting greater congruence in stable, long-standing democracies, where parties effectively communicate stances, we contrast this with lower party identification levels in Eastern Europe. Despite differing competition structures, Eastern Europe mirrors Western axes, emphasizing social issues over economic ones with niche parties (Rovny and Edwards, 2012). Aligning with Dalton (2017), we include the World Bank’s Voice and Accountability index for democratic development and the United Nations’ Human Development Index as a control variable.

Table 3 illustrates correlations between institutional predictors and centrism. A low score of centrism indicates more congruence. While some literature-supported hypotheses find partial backing, the effect of institutional factors is rather limited. More party options align with increased gender congruence but less on economic issues. Surprisingly, larger district magnitude correlates with reduced congruence, notably in the 2009 election. The primary consistent predictor across issue dimensions and both EP elections is old/new democracy. Parties in old democracies exhibit greater congruence with supporters, except in the 2014 economy dimension. The gender dimension consistently reflects institutional predictors’ significance and the highest agreement levels. In contrast, the cultural dimension lacks statistically significant relationships with hypothesised factors.

Table 3.

Institutional level predictors of congruence.

	Dimension
	Cultural		Economy		Gender
	2009	2014	2009	2014	2009	2014
EFNP	0.08	0.08	0.06	0.18**	−0.12**	−0.15**
LSq	−0.04	−0.03	0.01	−0.02	0.06	0.02
Polarisation	0.02	0.01	0.01	0.02	0.05	0.03
District magnitude	0.03	0.07	0.13**	−0.01	0.12*	−0.01
Voice and accountability	0.04	−0.08	−0.19**	0.15**	−0.28***	−0.22***
HDI	−0.02	−0.03	−0.18**	0.13*	−0.3***	−0.23***
Old democracy	−0.03	−0.02	−0.24***	−0.16**	−0.27***	−0.20***

Source: EU profiler (N = 230), euandi (N = 165).

Note: Pearson R correlation with centrism scores; higher centrism scores illustrate less congruence.

*p ≤ .01,**p ≤ .05,***p ≤ .001.

The impact of electoral system features and country contextual factors is limited and varies across issue dimensions and election years. This aligns with Dalton (2017), suggesting a positive interpretation that representation is not solely tied to institutional characteristics. This concurs with Belchior (2013), who found that system-level features like electoral system proportionality and fractionalization fail to explain variation in congruence between voters and MPs, as seen in the 2009 EP elections.

Party characteristics

The electoral systems’ impact on congruence is limited, possibly due to insufficient variation in party congruence at the country level. Research indicates voters reward ideologically distinctive parties (Belchior, 2013), especially those at the extremes, which are expected to be more congruent. Party size’s effect on congruence is mixed (Dalton, 1985; Mattila and Raunio, 2006); larger parties send clearer signals but may be less responsive, while niche parties are more responsive to the issues they own (Klüver and Spoon, 2016; Hooghe and Marks, 2018). Niche-mainstream party dichotomy, as defined by Meguid (2005), suggests niche parties, like Green or nationalist parties, prioritize unique issues, potentially leading to stronger links on cultural and gender issues. Ezrow, (2010) argues citizens with access to both mainstream and niche pathways fare better. Mainstream parties, larger and centrist, exhibit higher congruence (Belchior, 2013; Holmberg, 2000), especially on economic issues.

Government parties tend to adopt a centrist position at the cost of losing their ideological distinctiveness, which leads to confusion among their voters with regard to policy stances (Belchior, 2013). Controlling for governmental status is not enough, as the timing of the elections matters as well. Coining the concept of an election cycle, Reif, (1984) showed that congruence decreases the further a government party is in an election cycle. As the elections to the European Parliament are closer to the national ones, parties with governmental status will present higher levels of congruence with their supporters. Additionally, we consider the age of the party. As a measure of party size, we look at party vote share in the EP elections.

Table 4 presents regression results examining the impact of party characteristics on centrism. A notable ideological divide emerges, with left-leaning parties demonstrating higher congruence with their supporters while right-leaning parties exhibit consistent disagreement (center parties serve as the reference). This pattern holds across issue dimensions and both elections, except for gender issues in 2014. The contrast between left and right is more pronounced when considering party distinctiveness. Ideologically edged parties, particularly far-left ones, show higher congruence with their voters, especially on cultural issues. Situated at the political spectrum’s end, these parties convey their stances to voters effectively, yielding the strongest effects on cultural issues.

Table 4.

Party-level predictors of congruence.

	Cultural		Economy		Gender
	2009	2014	2009	2014	2009	2014
Intercept	3.06*	3.22*	2.63*	2.53*	2.74*	2.14*
Centre left	−0.37*	−0.24	−0.33*	−0.16	−0.26	0.07
Left	−0.80*	−0.49*	−0.66*	−0.27*	−0.54*	0.01
Centre right	0.18	0.19	0.43*	0.19	0.37*	0.48*
Right	0.57*	0.65*	0.40*	0.10	0.65*	1.00*
Age of party	0.00	−0.00	−0.00	0.00	−0.00	−0.00
Niche	0.00	−0.09	−0.11	0.09	−0.07	−0.16
Government party	−0.18	0.06	0.02	−0.21	−0.19	0.20
Election cycle	0.05	0.05	0.05	−0.06	−0.16	0.31
Vote share EP	0.01	−0.00	0.00	0.00	0.00	0.00
Gov. x el. Cycle	0.38	0.04	−0.03	0.28	0.52	−0.32
Adj. R²	0.44	0.41	0.50	0.27	0.34	0.27
Num. Obs	223	163	223	163	223	163

Source: EU profiler, euandi.

Note: The table presents unstandardised OLS coefficients for the predictors of centrism.

*p ≤ .01,**p ≤ .05,***p ≤ .001.

Ideological distinctiveness is not rewarded for parties on the right. Figures 3 and 4 depict parties’ positions on the RiLe continuum against centrism scores. Each dot represents a party dyad. Cultural issues exhibit more consistency among right-leaning parties in both elections, with greater variation on the left. Conversely, economic issues show increased variation on the right but maintain the most consistency in representing supporter views. The gender dimension mirrors the economic pattern, with left-leaning parties demonstrating higher congruence, though some far-left parties present low congruence levels, introducing noise to the data.

Figure 3.

Centrism and Left-Right 2009. Source: EU Profiler Note: Each dot represents a party dyad. The x-axis represents the RiLe continuum from 1 (left) to 7 (right)

Figure 4.

Centrism and Left-Right 2014. Source: euandi Note: Each dot represents a party dyad. The x-axis represents the RiLe continuum from 1 (left) to 7 (right)

Expecting larger parties to be less responsive, the impact of EP vote share is minimal. Differentiating parties by mainstream or niche status, often aligned with size, reveals mainstream parties aligning more closely with supporters on economic issues, particularly in 2014. This concurs with past research indicating mainstream party congruence on the RiLe dimension, encompassing economic issues (Ibenskas and Polk 2022). While niche parties exhibit a trend of higher congruence on cultural and gender issues, the coefficients lack statistical significance.

Figures 5 and 6 display congruence based on party family or EP group membership. The vertical axis shows centrism scores, with each point representing the average centrism scores for parties in each EP group. Colors and shapes represent different issue dimensions. In 2009, EP groups housing green and communist parties (Greens-EFA and GUE-NGL) exhibit the highest congruence across all three dimensions. Greens-EFA leads in gender issue congruence, while GUE-NGL’s communist parties excel on economic issues. The 2014 elections reflect a similar pattern. Notably, right-wing nationalist parties in the Identity and Democracy (ID) EP group defy theoretical expectations. However, this aligns with regression results (Table 4), where right-leaning parties are least congruent with supporters. For both elections, the representation link is weaker on cultural issues. This is most pronounced for euroskeptic parties (EFD in 2009 and EFDD in 2014), which sends positive signals about public opinion on issues related to immigration and crime. At the same time, agreement on economic issues is strongest for europskeptic and socialist parties.

Figure 5.

Congruence by EP Group 2009. Source: EU Profiler Note: Each point represents the average for the parties in EP groups

Figure 6.

Congruence by EP Group 2014. Source: euandi Note: Each point represents the average for the parties in EP groups

The effect of governmental status varies across dimensions and elections. While some evidence suggests that parties in government and further away in the election cycle may weaken the representational link, these relationships lack statistical significance. Besides ideological leaning, none of the other party characteristics consistently and significantly affect congruence. In essence, what Dalton (2017) calls ‘objective characteristics’ are less relevant compared to the party’s ideological edge.

Conclusion

The vast field of political representation studies benefited from the technological innovations around data collection. Increased data availability led to methodological innovations for gauging political representation in various electoral settings. At the same time, the plethora of means of assessing the link between representatives and those represented led to contradictory findings in the literature, with evidence of strong links using ideological congruence (Powell, 2000; Thomassen, 1994) and cautionary findings with high levels of representative deficit (Bright et al., 2020) and weak ties not just on the RiLe, but also on narrow policy dimensions (Clements et al., 2018; Dalton, 2017).

Much of the disagreement in the literature is attributed to the panoply of ways to measure congruence, due in part to compromises around data. These range from having to rely solely on ideological congruence and limiting the number of issues, to the number of parties, countries, and elections. Estimating the positions of the representatives from voters’ perceptions is highly problematic (Dalton et al., 2011) and looking solely at ideological congruence does not suffice, as salient issues in modern-day politics go well beyond the RiLe (Dalton 2017).

While there is no gold standard in assessing congruence, with measures ranging from simple distance measures to measures of relative congruence (Golder and Stramski, 2010), representation scholars agree on one aspect, shaped by data limitations. Congruence should be measured on the same questions, on the same scale, and at the same point in time. VAA-generated data offer the perfect opportunity to measure congruence in a comparative framework, First, VAAs are an integral part of the electoral process in an increasing number of countries. To date, there are VAA tools for three EP elections, which collected millions of responses. Because the wording of the issue policies is similar across countries, it allows for a comparative study. Second, VAAs include a large number of policy issues. Third, they include small parties, usually left out, because their supporters are more difficult to reach through traditional surveys. Fourth, it provides information on party positioning, on the same issues on which the voters are surveyed. The positioning of the parties is done by political experts, in some cases in consultation with the political parties. Last, the cost of the data is low, especially compared with traditional surveys (Cedroni and Garzia, 2010).

VAAs create a wealth of data that can be used to study various aspects of electoral behaviour, from turnout (Garzia et al., 2017b; Gemenis, 2018), to vote choice (Pianzola et al., 2019) and acquisition of information (Munzert and Ramirez-Ruiz, 2021). A less researched avenue is the study of political representation using VAA data, with the exception of Alvarez et al. (2014); Triga et al. (2018); Bright et al. (2020); Costello (2021); Costello et al. (2021), who focus on representative deficit. The main drawback in using VAA data for academic research stems from its double self-selection bias. First, as an online tool, VAA tends to attract users who are young, well-educated, and from urban areas. Second, it attracts users who are interested in politics. However, with an increase in the number of users, the first self-selection bias becomes less of a problem. The second type of bias could be used as an advantage. Bright et al. (2020) use Almond’s seminal distinction between the elites, the attentive public, and the general public, arguing that VAA users make up the attentive public, who are more informed about politics compared to the general public.

The increasing use of digital technologies in the electoral process and VAA proliferation means that more data will be readily available to study voter preferences. Some VAA developers collect demographic data and vote intention, which can be used to study congruence at the individual level. Additionally, these can be used to address the concerns around data quality, by applying more sophisticated weighting techniques, such as post-stratification.

Using data generated by EU Profiler and euandi, two VAA tools developed for the EP elections of 2009 and 2014, the results are in line with previous findings from work on political representation and issue congruence in EU context (Dalton, 2017). Institutional characteristics such as the number of electoral parties, polarisation or district magnitude do not determine the level of agreement between parties and their supporters (Belchior, 2013; Dalton, 2017). However, there are some differences in levels of congruence across parties when it comes to levels of democratisation. Parties in established democracies are more congruent with their voters, compared to parties in post-communist democracies. Party characteristics have a limited effect on congruence, apart from ideological leaning. Parties on the left show consistently higher levels of agreement with their supporters compared to those on the right. This difference becomes stronger as we move toward the ends of the ideological spectrum. Niche parties are more congruent with their supporters. Green, regionalist and ethnic parties show more congruence on issues such as euthanasia, legalisation of same-sex marriage, and legalisation of drugs, while communist parties are closer on economic issues.

In summary, there is evidence of a strong link between parties and their supporters, but the levels of agreement vary across issue dimensions. These nuances cannot be observed when we study representation from a traditional perspective, relying on ideological congruence. This work is an exercise in illustrating the advantage of using VAA data to investigate issue congruence in a pan-European setting, extending the analysis to two elections and a larger number of political parties. Further research can include more policy issues and an EU issue dimension.

Supplemental Material

Supplemental Material - Harmony and dissonance: Unveiling issue linkages between voters and parties across EU democracies

Supplemental Material for Harmony and dissonance: Unveiling issue linkages between voters and parties across EU democracies by Raluca Popp in Party Politics

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

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

ORCID iD

Raluca Popp

Supplemental Material

Supplemental material for this article is available online

Notes

Author biography

Raluca Popp is a Senior Lecturer in Quantitative Politics in the School of Politics and International Relations at the University of Kent, UK. Her scholarly interests focus on comparative political behaviour, political parties and conflict analysis.

References

Achen

(1977) Measuring representation: Perils of the correlation coefficient. American Journal of Political Science 21: 805–815.

Alvarez

Levin

Trechsel

, et al. (2014) Voting advice applications: how useful and for whom? Journal of Information Technology and Politics 11(1): 82–101.

Belchior

(2013) Explaining left–right party congruence across european party systems a test of micro-, meso-, and macro-level models. Comparative Political Studies 46(3): 352–386.

Benoit

Laver

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

Birch

(1971) Representing interests. In: Representation. Berlin, Germany: Springer, pp. 72–88.

Blais

Bodet

(2006) Does proportional representation foster closer congruence between citizens and policy makers? Comparative Political Studies 39(10): 1243–1262.

Bremer

(2018) The missing left? economic crisis and the programmatic response of social democratic parties in europe. Party Politics 24(1): 23–38.

Bright

Garzia

Lacey

, et al. (2020) The representative deficit in different european party systems: an analysis of the elections to the european parliament 2009-2014. Quaderni dell’Osservatorio elettorale QOE-IJES 83(1): 45–57.

Broockman

(2016) Approaches to studying policy representation. Legislative Studies Quarterly 41(1): 181–215.

10.

Cedroni

Garzia

(2010) Voting Advice Applications in Europe. The State of the Art. Napoli: Scriptaweb.

11.

Clark

Rohrschneider

(2009) Second-order elections versus first-order thinking: how voters perceive the representation process in a multi-layered system of governance. European Integration 31(5): 645–664.

12.

Clements

Nanou

Real-Dato

(2018) Economic crisis and party responsiveness on the left–right dimension in the European Union. Party Politics 24(1): 52–64.

13.

Costello

(2021) Issue congruence between voters and parties: examining the democratic party mandate in Ireland. Irish Political Studies 36(4): 581–605.

14.

Costello

Toshkov

Bos

, et al. (2021) Congruence between voters and parties: the role of party-level issue salience. European Journal of Political Research 60(1): 92–113.

15.

Dalton

(1985) Political parties and political representation party supporters and party elites in nine nations. Comparative Political Studies 18(3): 267–299.

16.

Dalton

(2008) The quantity and the quality of party systems party system polarization, its measurement, and its consequences. Comparative Political Studies 41(7): 899–920.

17.

Dalton

(2013) Citizen Politics: Public Opinion and Political Parties in Advanced Industrial Democracies. Washington, DC, USA: Cq Press.

18.

Dalton

(2017) Party representation across multiple issue dimensions. Party Politics 23(6): 609–622.

19.

Dalton

(2021) National/european identities and political alignments. European Union Politics 22(2): 340–350.

20.

Dalton

Wattenberg

(2000) Parties without Partisans: Political Change in Advanced Industrial Democracies. Oxford: Oxford University Press.

21.

Dalton

Farrell

McAllister

(2011) Political Parties and Democratic Linkage: How Parties Organize Democracy. Oxford: Oxford University Press.

22.

De Angelis

(2020) 946How voters distort their perceptions and why this matters. In: The Oxford Handbook of Electoral Persuasion. Oxford: Oxford University Press, pp. 946–973. ISBN 9780190860806. URL. DOI: 10.1093/oxfordhb/9780190860806.013.55.

23.

Dinas

Trechsel

Vassil

(2014) A look into the mirror: preferences, representation and electoral participation. Electoral Studies 36: 290–297.

24.

Eggers

Fouirnaies

(2014) Representation and district magnitude in plurality systems. Electoral Studies 33: 267–277.

25.

Ezrow

(2010) Linking Citizens and Parties: How Electoral Systems Matter for Political Representation: How Electoral Systems Matter for Political Representation. Oxford: Oxford University Press.

26.

Garzia

Marschall

(2019) Voting Advice Applications. Oxford: Oxford University Press.

27.

Garzia

Trechsel

De Sio

(2017a) Party placement in supranational elections: an introduction to the euandi 2014 dataset. Party Politics 23(4): 333–341.

28.

Garzia

Trechsel

De Angelis

(2017b) Voting advice applications and electoral participation: a multi-method study. Political Communication 34(3): 424–443.

29.

Gemenis

(2018) The impact of voting advice applications on electoral turnout: evidence from Greece. Statistics, Politics, and Policy 9(2): 161–179.

30.

Golder

Stramski

(2007) Ideological congruence and two visions of democracy. In: Annual Meeting of the American Political. Chicago: Science Association, pp. 90–106.

31.

Golder

Stramski

(2010) Ideological congruence and electoral institutions. American Journal of Political Science 54(1): 90–106.

32.

Hare

Armstrong

Bakker

, et al. (2015) Using bayesian aldrich-mckelvey scaling to study citizens’ ideological preferences and perceptions. American Journal of Political Science 59(3): 759–774.

33.

Holmberg

(2000) Issue agreement matters. In: Beyond Westminster and Congress: The Nordic Experience. Ohio: Ohio State University Press, p. 155.

34.

Hooghe

Marks

(2018) Cleavage theory meets europe’s crises: lipset, rokkan, and the transnational cleavage. Journal of European Public Policy 25(1): 109–135.

35.

Huber

Powell

(1994) Congruence between citizens and policymakers in two visions of liberal democracy. World Politics 46(3): 291–326.

36.

Ibenskas

Polk

(2022) Congruence and party responsiveness in western europe in the 21st century. West European Politics 45(2): 201–222.

37.

Ilmarinen

Isotalo

Lönnqvist

, et al. (2022) Do politicians’ answers to voting advice applications reflect their sincere beliefs? comparing publicly and confidentially stated ideological positions in a candidate-centred electoral context. Electoral Studies 79: 102504.

38.

Klüver

Spoon

(2016) Who responds? voters, parties and issue attention. British Journal of Political Science 46(3): 633–654.

39.

Krouwel

Vitiello

Wall

(2014) Voting advice applications as campaign actors: mapping vaas’ interactions with parties, media and voters. Matching Voters with Parties and Candidates. Voting Advice Applications in a Comparative Perspective. Colchester: ECPR Press, 67–78.

40.

Lacey

Trechsel

. (2014) Trans-nationalising Europe’s Voting Space. Florence, Italy: European University Institute (EUI), Robert Schuman Centre of Advanced. Technical report.

41.

Lax

Phillips

(2012) The democratic deficit in the states. American Journal of Political Science 56(1): 148–166.

42.

Lesschaeve

Padmos

(2023) Measuring agreement: how to arrive at reliable measures of opinion congruence between voters and parties. Representation 59(2): 311–328.

43.

Proksch

Gschwend

(2014) A common left-right scale for voters and parties in europe. Political Analysis 22(2): 205–223. DOI: 10.1093/pan/mpt028.

44.

Marschall

(2014) Profiling users. Matching Voters with Parties and Candidates. Colchester: ECPR Press, 93–104.

45.

Mattila

Raunio

(2006) Cautious voters-supportive parties: opinion congruence between voters and parties on the eu dimension. European Union Politics 7(4): 427–449.

46.

Meguid

(2005) Competition between unequals: the role of mainstream party strategy in niche party success. American Political Science Review 99(03): 347–359.

47.

Munzert

Ramirez-Ruiz

(2021) Meta-analysis of the effects of voting advice applications. Political Communication 38(6): 691–706.

48.

Pianzola

Trechsel

Vassil

, et al. (2019) The impact of personalized information on vote intention: evidence from a randomized field experiment. The Journal of Politics 81(3): 833–847.

49.

Piurko

Schwartz

Davidov

(2011) Basic personal values and the meaning of left-right political orientations in 20 countries. Political Psychology 32(4): 537–561.

50.

Powell

(2000) Elections as Instruments of Democracy: Majoritarian and Proportional Visions. New Haven, CT, USA: Yale University Press.

51.

Powell

(2009) The ideological congruence controversy: the impact of alternative measures, data, and time periods on the effects of election rules. Comparative Political Studies 42(12): 1475–1497.

52.

Reif

(1984) National electoral cycles and european elections 1979 and 1984. Electoral Studies 3(3): 244–255.

53.

Reiljan

da Silva

Cicchi

, et al. (2020a) Longitudinal dataset of political issue-positions of 411 parties across 28 european countries (2009–2019) from voting advice applications eu profiler and euandi. Data in Brief 31: 105968.

54.

Reiljan

Ferreira da Silva

Cicchi

, et al. (2020b) EU profiler/euandi trend file (2009-2019).

55.

Rohrschneider

Whitefield

(2012) The Strain of Representation: How Parties Represent Diverse Voters in Western and Eastern Europe. Oxford: Oxford University Press.

56.

Rovny

Edwards

(2012) Struggle over dimensionality: party competition in western and eastern europe. East European Politics and Societies 26(1): 56–74.

57.

Sartori

(2005) Parties and Party Systems: A Framework for Analysis. Colchester: ECPR press.

58.

Thomassen

(1994) Empirical research into political representation: failing democracy or failing models. In: Jennings

Mann

(eds), Elections at Home and Abroad: Essays in Honor of Warren E. Miller, Chapter 10. Michigan: University of Michigan Press, pp. 237–265.

59.

Thomassen

(2020) Party government and democracy. In: Research Handbook on Political Partisanship. Cheltenham, UK: Edward Elgar, pp. 31–46.

60.

Thomassen

. (2005) The European Voter: A Comparative Study of Modern Democracies. Oxford: Oxford University Press on Demand.

61.

Trechsel

Mair

(2011) When parties (also) position themselves: an introduction to the eu profiler. Journal of Information Technology and Politics 8(1): 1–20.

62.

Triga

Marschall

Wurthmann

(2018) Measuring issue congruence between German parties and voters: evidence from vaa data in three German elections. Conference Paper. Nicosia: Cyprus University of Technology, 90–106.

63.

Wheatley

Mendez

(2021) Reconceptualizing dimensions of political competition in europe: a demand-side approach. British Journal of Political Science 51(1): 40–59.

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

8.29 MB