The effect of MEPs' local ties on perceived representation: Evidence from four EU member states

The effect of local MEPs on perceived representation

In order to test the conjecture that MEPs sharing local ties with citizens will increase feelings of representation in the latter group, the article analyzes original data collected by YouGov in May 2024 in four EU member states: Belgium, Italy, Hungary, and Romania. Sample size is approximately N = 2000 for each country; responses were obtained via computer-assisted web interviewing and used quotas for age, gender, residence (NUTS-1 regions), and education.

While case selection is driven by constraints of data availability—respondents’ place of residence was only asked in countries that held local and EP elections concurrently—it nevertheless provides a neat balance between new and old member states, as well as countries both from the EU’s core and periphery (Clasen, 2024). While both Romania and Hungary have a closed-list system with a single national constituency, Italy and Belgium use multi-member districts with preferential voting for the EP vote. Nevertheless, previous literature has found preferential voting and district magnitude to be only weakly related to geographic representativeness (Carella and Eggers, 2024), and electoral system design, in general, to be unrelated to the feeling of being represented (Holmberg, 2020).

For the outcome variable, I leverage a survey item measuring the extent to which a respondent “feels personally represented in EU decision-making by their preferred member of the European Parliament.” This item wording mirrors previous operationalizations of perceived representation (Holmberg, 2020; Tosson and Schüle, 2025), which were used to capture the “subjectively experienced degree of policy representation” (Holmberg, 2020: 414) and a “subjective feeling state of political representation” (Vik and De Wilde, 2024). This measure is often contrasted with “objective” policy representation, and can be conceptualized along the lines of descriptive representation (De Mulder, 2023: 634). Responses are measured on an 11-point-scale, where 0 means “Disagree completely” and 10 means “Agree completely.”

Our primary independent variable captures shared local ties between respondents and MEPs. Recent scholarly work used a variety of different proxies for local roots, such as residence, place of birth and local-level political experience (Kovarek, 2022; Nagtzaam and Louwerse, 2023; Put et al., 2019). This article uses the latter two operationalizations.

First, I calculate whether the respondent’s place of residence matches an incumbent MEP’s hometown. To construct this variable, I utilize a dataset on socio-demographic background of MEPs (Michon and Wiest, 2021). Alternatively, I also specify local ties in a way that it only captures a subset of MEPs: those who previously held elected political office at the local level, relying on data from Salvati (2024) in municipalities with ≥ 20,000 inhabitants. The Online appendix demonstrates that both approaches are frequently used when measuring localness in these countries. I am unaware of any dataset recording place of residence for MEPs. Given that residence is often found to be a stronger proxy for local ties than birthplace (Put et al., 2019), the relationships reported here are expected to be even stronger when domicile was used to capture MEPs’ local ties. As Figure 1 shows, respondents sharing local roots with incumbent MEPs are scattered across various municipalities within each of the four countries in our sample.

OPEN IN VIEWER

Figure 1.

Hometowns of incumbent MEPs with shared local ties.

Figure 2 presents results obtained via logistic regression. Coefficients are plotted for bivariate models (only using birthplace or local-level experience), as well as for models that include a battery of control variables frequently used by the literature measuring the feeling of being represented. Models with different specifications (“represented” or “full sample”) differ primarily in the number of observations: whereas the previous estimate the effect of MEP localness only among respondents who “feel personally represented in EU decision-making” (i.e. scoring ≥ 6 on the scale), and predict the likelihood of selecting “my preferred MEP” as an answer to the follow-up question (“By whom do you feel represented?”), the latter models estimate the same likelihood in the entire sample (i.e. even including those, who do not feel represented at all). All models include country fixed effects, which control for unobserved country-specific factors, including the aforementioned differences between electoral systems. The Online appendix offers a tabular presentation of the estimates and a detailed description of how variables were operationalized.

OPEN IN VIEWER

Figure 2.

The effect of local MEPs on perceived representation.

In the Online appendix, I also present four additional robustness checks. An additional placebo analysis increases our confidence in the results. Estimating perceived representation by heads of governments, respondents’ preferred party in the EP or the head of the European Commission, using the same set of predictors as in Figure 2, yields null results, suggesting that results are not driven by endogenous confounders. In other words, the mechanism appears to be specific to MEPs sharing local ties with their constituents. Furthermore, I also verify that this relationship is not driven by densely populated urban localities, respondent subgroups with high levels of political interest, or senior-ranking MEPs enjoying heightened media visibility (see the Online appendix Material).

Conclusion

This article investigated how sharing local ties with incumbents makes European citizens more likely to perceive themselves as being personally represented by such MEPs. In the age of the Europeanization of national parliamentary representation (Kinski, 2018), this letter demonstrated the localization of European representation. I find that respondents feel more represented when they live in places to which incumbent MEPs have personal or political ties, echoing findings of previous studies on local politicians in other institutional contexts (Campbell et al., 2019; Put et al., 2019). This increased sense of being represented might be a normatively important voter-elite linkage that could contribute to reducing the democratic deficit that has long plagued the EU (Follesdal and Hix, 2006), as representation beyond formal constituencies has been conjectured to alleviate representation deficits beyond the state and to generate democratic legitimacy for the EP (Kuyper and Moffitt, 2020).

What is presented here is nothing more than a necessary first step in studying localism in the context of the EP. In particular, more research is needed on behavioral localism: future work should probe the intersections of MEPs’ parliamentary behavior and their local roots (cf. Nagtzaam and Louwerse, 2023). The Online appendix presents an exploratory analysis of speeches by Italian, Romanian, Hungarian, and Belgian MEPs and finds that legislators’ birthplaces are more likely to be mentioned in the EP.

The mounting empirical evidence suggesting local incumbents are expected to spoil their constituents with public goods and tangible benefits (Kovarek, 2022) should also invite further study about how local roots of elected deputies interact with other motivations of sub-state governments when allocating structural funds to projects in a discretionary fashion (Dellmuth and Stoffel, 2012). Roll-call analysis on EP data (Høyland, 2010) would be well-suited not only to better elucidate the specific mechanisms at work, but also to test whether elected representatives with strong local ties are more likely to break (Euro-)party unity, similarly to national parliamentarians (Tavits, 2009).

Future studies should also try to further unpack whether local candidates running for EP mandates also enjoy an electoral bonus. Friends-and-neighbors voting in national elections was empirically documented in at least twelve EU member states (see the Online appendix), providing a nuanced vantage point.

Supplemental Material

Supplemental Material