European Electorates Explored (I): Who is voting for which group?

The European Parliament elections, taking place between 6 and 9 June, are fast approaching. Understanding who votes for which political group can help us better understand European politics. In this series, “European Electorates Explored”, we dive into the data with a series of graphs that reveal several characteristics of the voters supporting different political groups in the European Parliament. Today, we examine the voter bases of these groups by focusing on socio-demographics such as gender, age, self-perceived class and wealth, education, place of residence and religiosity. What differences emerge across these categories?

In a couple of weeks, more than 400 million people will have the opportunity to cast their votes for members of the European Parliament (EP). Yet although Europeans elect their MEPs directly, fully-fledged political parties don’t exist at the European level. Instead, citizens vote for their national parties, which then join broad ideology-based groups within the EP.

There are eight main groups in the EP: from left to right these are the far-left “The Left” group, the Greens – European Free Alliance, which unites greens with smaller pro-European left-wing parties, the social-democratic Progressive Alliance of Socialists and Democrats (S&D), the liberal Renew Europe group, the European People’s Party (EPP), that unites Christian-democrats and pro-European conservatives, the European Conservatives and Reformists (ECR), that brings together Eurosceptic conservatives and a number of far-right parties, and the fully-fledged far-right Identity and Democracy (ID) group. There is also a group of parties that do not belong to one of these groups: the Non-Inscrits.

In 2019, during the previous elections, only half of the eligible voters turned out in the European Parliament elections. It is not uncommon that in second order elections – like the EP elections – more voters stay at home. More than national elections, European elections are turn-out elections where parties focus on mobilizing their constituencies instead of convincing voters. The low turnout partially explains the problems with representation in the European Union: groups that do not turn out to vote will also not see their interests represented in the European Parliament.

Although much is known about these groups and their member parties, we know only very little about the overall characteristics of their voter bases. In this short series of blog posts, we will provide an overview using data from the European Election Study – a large-scale survey conducted just after the previous EP elections in 2019 (for more information, see the methodological justification below).

Let’s start with one of the key socio-demographics: gender.

Figure 1: Gender

Note: 29 respondents (across 27 countries) identified as neither male nor female. We have not included them in the analysis because this number is too low to draw any meaningful statistical conclusions.

Figure 1 shows that women are slightly overrepresented among the electorates of left-wing groups in the EP. Specifically, the voter base of the Greens consists of 54% women, while The Left has 52% women. In contrast, there is a noticeable overrepresentation of men among the constituencies of the strongly right-wing groups. The ECR has a voter base that is 55% male, and the ID group sees an even higher percentage, with 59% of its voters being men.

This pattern is in line with the reputation of far-right parties as a Männerparteien (men’s parties). Although the tendency of far-right parties to primarily attract men is diminishing in many countries, this graph indicates that, on average, this tendency still persisted during the previous European elections.

When it comes to turn-out, women also tend to stay home more often than men. This difference was present in previous elections as well. Ruth Dassonneville and Filip Kostelka explain this difference by referring to differences in political interest between men and women. 

Figure 2: Age

Figure 2 illustrates the age distribution of voters across different groups in the EP. It shows that Greens and Renew voters are relatively young, with a substantial portion under the age of 35. This pattern is also visible among those who did not vote. Additional analyses show that in particular non-voters and supporters of the Greens have the lowest mean ages (47 and 50, respectively).

In contrast, the S&D and EPP groups have the highest percentages of older voters, with mean ages of 55 and 56, respectively. This is consistent with the view that established mainstream parties are more popular among older citizens, while Greens attract younger voters. It also indicates that older citizens are more loyal to the established parties than young citizens.

Younger voters also tend to stay home more often than older voters. It is a well-known pattern in election research that younger voters are less likely to turn out.

Figure 3: Self-perceived class

Note: The original scale consists of six categories. We have recoded it into three categories. Those who indicated that they belonged to an “other” class have been left out of the analyses.

Participants in this study were asked to which class they believe they belong. Figure 3 shows that voters for the ideologically more extreme groups more frequently identify as belonging to the working or lower middle class. Specifically, 55% of voters for The Left categorizes themselves this way, as do 47% of the supporters of the ID group. Also 50% of those who did not vote perceive themselves as hailing from the working or lower middle class. Relatively few supporters of the mentioned groups identify with the upper middle or upper class.

This contrasts with in particular supporters of the Renew group: only 30% identify with the working or lower middle class, while 19% say they belong to the upper middle or upper class.

Working-class voters are also less likely to turn out. This pattern is well documented in the literature: understanding what is at stake in elections requires more effort for working-class voters because politics (and European politics in particular) is often presented as complex, academic, and elite-driven.

Figure 4: Wealth

Note: The survey question concerns someone’s family’s standard of living. The original scale consists of seven categories, ranging from 1 (poor family) to 7 (rich family). We have recoded this into three categories.

The pattern in Figure 4 mirrors that of the previous graph. Voters for The Left and, to a lesser extent, the ID group, are more likely to say they come from poorer families. Specifically, 44% of Left voters classify themselves this way, and 33% of ID supporters. Those who did not vote are, on average, similar in this respect: 39% of them claim they come from a poorer family.

Conversely, relatively few respondents in these three categories claim to hail from wealthy families. In contrast, supporters of parties in the Renew Europe group are most strongly overrepresented in the rich family category.

When it comes to turn-out we see the same pattern as we saw for class: poorer citizens are less likely to vote.

Figure 5: Education

Note: The EES does not ask questions about actual education levels. Instead, it is examined how many years of formal education respondents have had. This variable is recoded into a three-point scale: low (15 or fewer years), medium (16-19 years), and high (20 or more years).  

When it comes to education (see Figure 5), supporters of the ID group stand out. 41% of them report having had only 15 years or less of formal education. This percentage is much lower among supporters of other groups and non-voters (between 17-25%). In particular, the difference with ECR voters stands out here: only 17% of the voter base of this group fall into this category. Conversely, relatively few ID supporters (21%) report having 20 or more years of education, whereas this percentage is much higher in all other groups, ranging from 37% among S&D voters to 46% among Renew voters. Non-voters also fall less often into the “high” education category, with only 30% reporting 20 or more years of education.

This finding aligns with numerous academic studies that highlight a clear correlation between education level and far-right support: in particular those with university-level education are underrepresented among the far-right voter base. An interesting result from this analysis is the significant difference between supporters of the ECR group and those who voted for the ID group. This discrepancy might be due to the higher level of diversity within the ECR, which also includes conservative parties that are not far-right. The ID faction consists of far-right parties only.

In terms of education, non-voters have an ambiguous profile. They are less often “higher educated” but are not overrepresented in the “lower educated” category. This result is puzzling at first glance. However, this can be understood by noting that younger voters are less likely to turn out. Given the increase in years of schooling in recent decades, younger voters are more likely to have middle or higher education, yet they still tend to stay home.

Figure 6: Religiosity

Note: The EES asks respondents how often they attend religious services apart from weddings and funerals. We recoded “more than once a week”, “once a week” and “once a month” into “often”, “about each 2 or 3 months” and “only on special holidays” into “sometimes” and “about once a year”, “less often” and “never” into “seldom or never”.

Figure 6 shows that, as can be expected, those who support the pro-European conservative and Christian-democratic EPP and Eurosceptic conservative ECR most often fall into the “often” category for attending religious services (36% and 37% respectively). Similarly, voters for these groups report less often that they seldom or never attend religious services. However, a notable difference between the supporter bases of these two groups emerges: the share of the least religious category is much smaller among EPP supporters (34%) compared to ECR supporters (45%). In other words, on average EPP voters are more religious than ECR voters.

The least religious are supporters of the Greens and the ID group, with only 14% and 16% respectively falling into the “often” category for attending religious services. Those who did not vote also turn out to be not very religious.

As with education, a clear difference between the supporters of the ECR and ID groups becomes apparent. Voters for the latter group are much less religious. In our next blog post in this series, we will examine the ideological consequences of this difference.

Figure 7: Place of residence

Note: The EES asks people where they live. Hence, this is how respondents themselves describe their place of residence.

When it comes to place of residence (see Figure 7), the most striking finding is that there are only modest differences between the support bases of various parties as well as non-voters. The percentage of people living in large towns is slightly higher among those who support the S&D group (47% compared to 36%-41% in other groups). Conversely, the share of people living in rural areas or villages is slightly lower among S&D supporters (23% compared to 27%-35% in other groups).

Overall, in terms of voting behavior, there doesn’t appear to be a strong rural-urban divide – at least not during the previous EP elections.

Conclusion

To sum up: how do the different voter bases stand out?

The three groups on the left share one common feature: their supporters are predominantly women. However, these parties have distinct profiles. The core constituency of The Left consists of working-class and poorer, secular voters. The Greens also mobilize women and secular voters, but their class profile is opposite to that of The Left. Their voters are also much younger. The S&D mobilizes older, urban, and working class voters.

The Renew group stands in the political center, but its voters are not average voters. The liberal group mobilizes secular, upper, and upper middle class voters, and voters with higher education and more financial resources.

The three groups on the right share one common feature: their supporters are predominantly men. But on other variables there are some interesting differences and similarities: The Christian-democratic and pro-European conservative EPP mobilizes religious and older voters. The ECR, which unites Eurosceptic conservatives and far-right parties, mobilizes a similar electorate of religious and older voters. Its electorate is also quite rural. The kind of electorate that might vote for the EPP in one country, could choose the ECR in another. ID voters are less religious and share several similarities with The Left electorate: they are lower class, less wealthy, and have the lowest education levels.

Finally, the non-voting half of the electorate also has a specific profile: they are younger and secular women, but they are also working class and tend to lack resources, such as higher education levels and higher incomes. It is challenging for parties to mobilize them for the complex, academic, elite-driven politics of the European Parliament. This helps explain why the Christian-democratic EPP and the liberal Renew group (which tend to be quite sensitive to business interests) are so strong in the EP. The voices of working class, lower educated, and poorer citizens are not heard that well in Brussels, because these people tend to stay at home during the elections.

The cynical truth is that decisions are made for those who turn out.

Notes

Usually we only publish in Dutch on Stuk Rood Vlees. This short series forms a one-time (or actually three-time) exception.

Image from Wikimedia Commons.

Methodological justification

For the analyses, we made use of the 2019 European Election Study (EES) Voter Study. This large-scale survey was conducted after the elections to the EP in 2019 in all 28 member states of the EU. The data were collected by Gallup International. The sample size is about 1,000 in each country. For more information about the dataset see here.

All participants was asked the following question: “The European Parliament elections were held on Thursday 23 May 2019. For one reason or another, some people in [country] did not vote in these elections. Did you vote in the recent European Parliament elections?”. If the answer was yes, it was followed by this question: “Which party did you vote for in the European Parliament elections?”

The answering categories were formed by the national parties people could vote for. We replaced these values by the European groups these parties are members of. We dropped the UK from our analyses, and used weights to make sure the sample is representative on key characteristics like age, gender, urbanity, region, education, turnout and vote recall.

Note that because of rounding, some of the percentages in the bar charts add up to more than 100%.