Introduction
Though the single market has long been the center of the project of European integration, its relationship to most Europeans has also long been ambiguous. It is no secret that the strategy taken by the founders of the European Union – the so-called “Monnet method” – was to employ market integration as a mechanism to circumvent Europeans’ presumed nationalist reflexes. By focusing first on the low-salience technical-regulatory terrain of rules for cross-border market access, the European institutions would lure citizens with functional economic gains toward deeper political union (Majone, Reference Majone2006; Genschel and Jachtenfuchs, Reference Genschel and Jachtenfuchs2016).
In terms of achieving the upward delegation of policy-making power, this strategy worked. For several decades, the project enjoyed a “permissive consensus” of largely inattentive publics. By the early 2000s, EU authority over market regulation had grown so extensive that scholars increasingly analyzed it in quasi-federal terms. Much less clear, however, was the success of the Monnet method in terms of the public support that was supposed to follow. From the 1990s into the 2000s the “permissive consensus” broke down into increasing politicization and a rising wave of explicit Euroscepticism. This shift paralleled treaty revisions that extended EU powers beyond market integration into higher salience issue areas like the single currency, oversight of national budgets, and immigration into the EU (as opposed to within it). As a result, it is difficult to discern how much contemporary politicization reflects dissatisfaction with market integration itself, or is primarily a response to the EU’s deeper pooling of sovereignty outside the single market.
This paper takes up that question directly. Drawing on original survey data from eight EU member states, we isolate public attitudes toward the governance of the single market and evaluate the extent to which they are structured by utilitarian concerns, ideological commitments, and the so-called “new politics” of identity-based predispositions. Most existing measures and models do not allow us to specifically address support for the single market, because they conflate support for economic integration with broader sentiments toward the EU as a political actor, with some exceptions on the more con-tentious objective of internal migration (e.g., Karstens (Reference Karstens2020a)). Rather than treating support for the EU as a catch-all attitude, we focus specifically on citizen views of single market regulation – the core policy instruments of economic integration.
Our focus on the single market presents a particularly hard case for theories of identity-driven contestation. The rules governing market openness are complex and relatively insulated from direct political control, and thus would seem least likely to activate mass ideological or cultural predispositions. But our design allows us to ask whether the same “new politics” cleavages that shape attitudes toward high-salience integration domains also organize support for technocratic, regulatory domains (Hooghe and Marks, Reference Hooghe and Marks2009). In doing so, we move beyond debates about whether the public supports “more Europe” in general, and instead test whether the original logic of the Monnet method – quiet functionalism in service of integration – holds on single-market issues. Even the leading theorists of the “new politics” in EU opinion, Liesbet Hooghe and Gary Marks, have suggested that the “old politics” of either utilitarian or classic left-right concerns is likely to prevail on this economic terrain (Hooghe et al. (Reference Hooghe, Marks and Wilson2002); also Bremer et al. (Reference Bremer, Genschel and Jachtenfuchs2020)).
We find that it does not. The politicization of EU authority now extends even to low-salience domains like single market governance, suggesting that the postfunctionalist turn in public opinion theory has outgrown its origins in highly salient issue areas. That is, the EU’s single market is no longer buffered from identity-driven conflict, but instead shaped by the same cultural and ideological cleavages that animate opposition to other forms of supranational authority. We show that support for single market policies is systematically related not only to utilitarian and economic factors, but also – and more powerfully – to postmaterialist values and the cultural dimension of political conflict. Specifically, we find that attitudes aligned with the TAN (traditional/authoritarian/nationalist) pole of the GAL/TAN axis predict opposition to the core policy objectives and techniques of single market governance.
These findings carry theoretical implications for both European integration studies and broader research on public opinion in multilevel systems. While we do not observe change over time, our evidence indicates that even functionally oriented domains of governance are structured by symbolic and identity-laden frameworks alongside utilitarian considerations – a pattern that echoes the broader shift in the literature toward postfunctionalist explanations of European integration.
This is a consequential finding. If identity-based, postfunctionalist dynamics structure opinion even on the technical and relatively low-salience terrain of single-market governance, then their reach extends well beyond the highly politicized arenas that dominate contemporary debates. Despite its technical, often obscure politics, the fact that that it is nonetheless organized by identity-based cleavages suggests that postfunctionalist theory captures a deeper structural feature of EU politics rather than a phenomenon confined to moments of crisis or specific cultural politics. In this sense, our findings strengthen the case for postfunctionalism as a general framework for understanding mass attitudes toward multilevel governance. In practical terms, any attempt to reform or revive the single market (e.g., Letta (Reference Letta2024); Draghi (Reference Draghi2024)) must not solely attend to the economic performance of member states, but also to the symbolic and cultural meanings that citizens attach to market governance.
The paper proceeds as follows. In the next section, we situate our argument within the literature on European integration, postfunctionalist theory, and public opinion. We then describe our data and empirical strategy, including our original survey fielded across eight EU member states. The results section presents our main findings, followed by a discussion of their theoretical implications. We conclude by reflecting on how the politicization of the single market reframes longstanding assumptions about technocratic governance and the limits of public contestation in the EU.
Understanding attitudes towards the single market
Public opinion research about European integration has understandably focused on attitudes toward the most visibly politicized dimensions of EU authority, such as external immigration, fiscal transfers, and sovereignty disputes. These issues, especially after Maastricht, shifted scholarly attention toward identity and postmaterialist values as key drivers of EU support, challenging earlier models that emphasized economic utility and the classic dimensions of economic liberalism and economic conservatism (Gabel, Reference Gabel1998; Hooghe and Marks, Reference Hooghe and Marks2005; Hobolt and de Vries, Reference Hobolt and de Vries2016).
Yet much of this work has treated the single market as a background condition – either presumed to be “permissively” supported or folded into general measures of EU sentiment. Even if the single market continues to impact the food available to consumers (Vaqué, Reference Vaqué2016), what jobs they can do where (Koumenta and Pagliero, Reference Koumenta and Pagliero2019) and what banking services they can access (Mugarura, Reference Mugarura2016), there is scant evidence on how EU citizens think of each of its facets. Beyond sporadic and relatively thin Eurobarometer surveys (European Commission, 2013, 2023), few have attempted to survey popular attitudes towards EU single market policies, with the exception of several studies of support for the free movement of people (Vasilopoulou and Talving, Reference Vasilopoulou and Talving2019; Karstens, Reference Karstens2020a, Reference Karstens2020b). As a result, we still know little about whether attitudes toward the single market itself are shaped by utilitarian considerations, ideological preferences, or identity-based predispositions.
Utility and its impact on popular support for the EU
The earliest accounts of support for European integration emphasized material self-interest. Rooted in the economic voting literature (e.g., Baccini and Weymouth Reference Baccini and Weymouth2021), this “pocketbook European” perspective treated attitudes toward the EU as shaped by citizens’ perceived economic benefit (Inglehart and Rabier, Reference Inglehart and Rabier1978; Gabel and Palmer, Reference Gabel and Palmer1995; Gabel, Reference Gabel1998). The premise is straightforward: individuals who gain from market openness are more likely to support further integration. This framework continues to inform recent studies linking economic positioning to support for the EU (Foster and Frieden, Reference Foster and Frieden2021) and for the Euro specifically in the post-crisis period (Hobolt and Wratil, Reference Hobolt and Wratil2015).
Following this logic, individuals with higher levels of income and education are expected to benefit more from the single market. They are better positioned to take advantage of cross-border mobility, less vulnerable to wage competition, and more likely to work in sectors that benefit from integration (De Vries et al., Reference Vries, Catherine, Hobolt and Walter2021). We thus hypothesize:
H1: Higher-income individuals will express stronger support for the free movement of goods, workers, and services
Education is likely to exert a similar influence. Prior work shows that higher education reduces Euroscepticism (Kunst et al., Reference Kunst, Kuhn and van de Werfhorst2020), potentially because it insulates workers from low-wage competition or enhances their labor mobility (Kriesi, Reference Kriesi2016). Education may also expose individuals to pro-integration norms or increase transnational experiences that foster support for market openness (Hainmueller and Hiscox, Reference Hainmueller and Hiscox2006; Hakhverdian et al., Reference Hakhverdian, van Elsas, van der Brug and Kuhn2013; van Mol, Reference Van Mol2018). In other words, education may do its work in shaping views of the single market through two different mechanisms: First, it may make citizens more likely to adopt liberal views on postmaterial issues. It may also, in a move that is more relevant to the utilitarian mechanism of preference formation, provide people with formal credentials and social networks that can, in turn, bolster their ability to withstand structural changes or economic downturns in the labor market as a result of increasing economic integration (Kriesi et al., Reference Kriesi, Grande, Dolezal, Helbling, Höglinger, Hutter and Wüest2012). In other words, education can work to shape attitudes towards the single market along two axes: one values-based axis, in which support for European integration varies with education as a function of how education produces more politically liberal worldviews (Cavaille and Marshall, Reference Cavaille and Marshall2019) and one economic axis, in which those with higher education become more favorable towards the single market because it places them in a position of being able to take advantage of the labor mobility that is implicit to it. Despite the different causal mechanisms, education will in both cases have a positive relationship to support for the single market. Accordingly:
H2: Higher levels of education will be associated with greater support for market integration.
Finally, irrespective of income or education, individuals employed in firms that engage in or aspire to cross-border activity should be more supportive of the single market. These individuals are likely to see integration as directly benefiting their employer – and, indirectly, themselves (Gabel, Reference Gabel1998; Foster and Frieden, Reference Foster and Frieden2021). Cross-national operations increase the salience of regulatory integration and amplify perceived benefits. We therefore expect:
H3: Individuals employed in firms that operate or aim to operate across EU borders will express greater support for the single market than those in firms without such exposure.
Left-right ideological expectations
Another line of research retains a focus on economic thinking but moves beyond simple utilitarian considerations to evaluations of market integration through the lens of economic ideology – specifically citizens’ placement along the left-right spectrum. From this view, attitudes toward the EU reflect not direct economic positioning, but broader ideological commitments to the role of markets and state intervention (Hooghe et al., Reference Hooghe, Marks and Wilson2002; Elsas and van der Brug, Reference Elsas and van der Brug2015).
In the EU context, left-right ideology is often described as a diminishing force in shaping public opinion about European integration – particularly since the 1980s, as the EU moved beyond its original market-building remit. But scholars have consistently argued that this “old politics” dimension still matters for EU policy areas where redistribution, regulation, or competition remain central – especially the single market (Hooghe et al., Reference Hooghe, Marks and Wilson2002; Kanthak and Spies, Reference Kanthak and Spies2018). Citizens with right-leaning, pro-market views may be generally more supportive of European-level action to liberalize markets and reduce regulatory barriers. This expectation is reinforced by the prominence of ordoliberal traditions in continental Europe, which frame central governing authority not as a threat to markets but as a necessary guarantor of competition (Hien and Joerges, Reference Hien and Joerges2017).Footnote 1 Accordingly:
H4: Individuals with more economically right-wing views will be more supportive of single market integration.
This hypothesis treats ideology as distinct from income or education, but they could be complementary. Left-right attitudes may mediate how individuals interpret their economic positioning, or serve as stable heuristics that guide opinion even when personal stakes are unclear. In testing this, we assess whether economic ideology predicts support for market integration net of utilitarian considerations – and whether it interacts with economic position to structure attitudes more powerfully in some contexts than others.
Postmaterialism and the EU’s single market
A now large body of literature suggests that public attitudes towards the EU are structured not only by economic positionality and ideology, but also by citizens’ ideological preferences on a second set of non-economic policy issues. Postfunctionalist theory argues that identity-based contestation has become the dominant logic of European integration, especially as EU authority has expanded into politically salient domains (Hooghe and Marks, Reference Hooghe and Marks2009; Hobolt and de Vries, Reference Hobolt and de Vries2016; De Vries et al., Reference Vries, Catherine, Hobolt and Walter2021). These cultural predispositions are often captured through the GAL/TAN framework – a second dimension of political conflict that runs orthogonal to traditional left-right economic ideology (Marks and Steenbergen, Reference Marks and Steenbergen2002).
The TAN (traditional/authoritarian/nationalist) pole of this cleavage has been consistently associated with opposition to European integration (Hooghe and Marks, Reference Hooghe and Marks2018; Gallina, Reference Gallina2023). This effect is especially pronounced when reinforced by elite cueing: right-wing challenger parties frequently emphasize cultural threat and national sovereignty, deepening the politicization of EU authority (Beaudonnet and Gomez, Reference Beaudonnet and Gomez2024). We test whether this axis also helps explain opposition to single market regulation – a domain often presumed to be shielded from contestation due to its technocratic and economic character.
H5a: Orientations toward the TAN side of the GAL/TAN axis will be associated with lower support for the single market.
These effects may also differ by policy area. A long tradition in social psychology suggests that cultural threat is especially potent when it implicates group membership and social identity (Tajfel and Turner, Reference Tajfel and Turner2004). This has been confirmed in EU-specific studies showing that the free movement of labor – more than goods or services – activates nationalist resistance (Karstens, Reference Karstens2020a). Socially conservative voters are particularly sensitive to migration as a symbolic threat to national identity, especially when politicized by party elites (Hameleers et al., Reference Hameleers, Bos and de Vreese2017). Thus we expect:
H5b: TAN orientations will have a stronger negative effect on support for the free movement of workers than on support for the free movement of goods or services.
A related observation is that stronger national identity – particularly when prioritized over European identity – has consistently been associated with Euroscepticism (McLaren, Reference McLaren2002; Hooghe and Marks, Reference Hooghe and Marks2005). This point is clearly related to the “N” in TAN but specifies the source of attitudes somewhat differently, such that national identification can be interpreted as a complementary measurement strategy for capturing the salience of the TAN cleavage. While nationalist identification can of course be exacerbated by economic precarity (Ponte, Reference Ponte2021), and is most likely to correlate to opposition to Euro-pean integration if it co-exists with socially conservative views, the relationship between GAL/TAN, national identity and support for the EU is somewhat orthogonal: A strong belief in the nation-state as the legitimate locus of power will be strongly correlated with nationalist self-identification. This view is likely to be an independent source of opposition to the EU single market, even among social liberal and high income individuals (Hooghe and Marks, Reference Hooghe and Marks2009). In other words, in postfunctional thinking identity could amplify the “N” in TAN by tapping into ideas about the nation-state as the legitimate source of decision-making that does not always correspond to socially conservative views of the other issues that make up the GAL/TAN cleavage. We therefore expect:
H6a: Stronger national (relative to European) identification will be associated with lower support for the single market.
H6b: This negative effect will be strongest for the free movement of workers.
By distinguishing between postmaterialist predispositions (as captured by GAL/TAN) and national identity (as a related but distinct measure of cultural orientation), we evaluate two complementary expressions of the “new politics” cleavage. This allows us to assess not only whether cultural values structure attitudes toward economic governance, but also how different expressions of cultural conflict map onto specific components of the single market. In doing so, we contribute to a growing literature on ambivalent or multidimensional public attitudes toward the EU (Stoeckel, Reference Stoeckel2013; Moland, Reference Moland2023), and help clarify whether the single market has remained sheltered from the broader “constraining dissensus” that now limits deeper integration (Hooghe and Marks, Reference Hooghe and Marks2009).
Data and methods
Our paper uses data from eight EU member states. The survey was fielded in the summer of 2023 through Ipsos, a survey company. We used a quota sampling strategy to ensure that the sample is representative of the population with respect to age and gender, in Austria, France, Germany, Ireland, Italy, Poland, Romania and Spain (N = 26,283). To standardize variables like income and education we used Ipsos’ country-specific scales. This allows us to use comparable scales to test the effect of these differently-scaled variables, which is necessary to test our utilitarian hypotheses.
Our sampling strategy also ensures that the sample is broadly representative of EU member states along several relevant dimensions. First, our sample covers a wide range of the many different regimes known from the varieties of capitalism literature (Hall et al., Reference Hall, Soskice, Hall and Soskice2001). Such variations in state-market interactions are likely to matter for public contestation of economic issues inside the EU (Wueest, Reference Wueest2013). Second, the countries in our sample include both “core” countries of the EU, such as Germany and France, and more peripheral members like Austria, Spain, Ireland and Romania. This is important because of the well-known differences in how citizens of the center and periphery relate to the EU (Bølstad, Reference Bølstad2015). The variations in both geography and EU membership tenures found in our sample allow us to paint a generalizable picture of support for the single market after the sovereign debt crisis. To further strengthen the representativeness of the sample, we also reweighted respondents to achieve representativeness on combinations of age and gender. We weighted each age-gender group relative to its share in the population as defined by Eurostat. Composite weights within markets also account for the relative distribution of the country’s population to the sampled countries’ totals within the EU.
Dependent variables
Our main dependent variables capture support for deeper single market integration, assessing first the level of agreement with two foundational principles, as well as whether respondents believe the EU institutions should do more to facilitate cross-border economic activity.
Our first dependent variable is a statement phrased, “In order to have a thriving economy, people must be allowed to move and work freely across the European countries.” The second dependent variable asks people whether they agree or disagree with the statement, “In order to have a thriving economy, companies must be allowed to do business freely across the European countries”.
The third question asks directly about support for more EU action to pursue these principles, and is phrased: “The EU should do more to make trade across Europe easier.” This measure also makes explicit that we are asking about openness within the EU single market, distinguishing that focus from international trade agreements that can also be politically salient (Dür and Schlipphak, Reference Dür and Schlipphak2021; Oleart, Reference Oleart2021).
In our view, the phrasing of these questions maximizes the chances that respondents will have answered with the EU and its core single-market policies concretely in mind. Rather than suggesting abstract categories like “migration” or the “four freedoms” of goods, services, labor and capital, we aimed to encourage respondents to picture people or businesses in movement across member-states, or EU steps to promote that movement. This concrete focus also aims to avoid priming respondents to think about non-economic issues that relate more to the general postmaterialist cleavage. These choices do not allow us to measure support for the free movement of capital, but we reasoned that we could elicit the most concrete reactions by setting aside this most abstract and technical area of internal-market regulation.
All questions use Likert-scales ranging from 1 (“Strongly disagree”) to 5 (“Strongly agree”). Both when presenting the descriptive results, as well as when running regression analyses, we recode the variables as dichotomous indicators. We do this consistently to facilitate easier interpretation: rather than treating the difference between “strongly agree” and “neither” as substantively important categories (as respondents may use them very differently), we distinguish those disagreeing (by saying they strongly or somewhat disagree) from those who do not (who either express ambivalence or more or less strong support). Those with invalid values are removed via listwise deletion.
Figure 1 breaks down the support of each variable in line with this recoding, and shows the generally high support for all three questions found across the EU. Even this preliminary analysis, however, reveals wide country-level variations: French citizens are consistently ranked as having the lowest levels of support for all policy areas, with Romanian and Polish citizens exhibiting the highest levels. This points towards a clear center-periphery dimension to support for EU single market policies, and that popular support for it may vary with how attractive it might be to take advantage of the opportunities for movement that the single market implies.
Support for EU single market.
Independent variables
To investigate the effect of income on support for the single market (H1a) we use a measure of yearly household income. We prefer this operationalization over one that asks about broad categories of occupations, often used in the existing literature (Gabel, Reference Gabel1998; Hooghe and Marks, Reference Hooghe and Marks2005; Foster and Frieden, Reference Foster and Frieden2021) because the correlation between income and occupation may be imprecise. We transform the country-specific scales provided by Ipsos, which employ scales in local currencies that are adapted to local wage levels while retaining similar intervals, to a uniform scale of income groups that we employ across all countries. Thus, each country-specific category is mapped onto one of the 16 categories making up our unified measure of household income. This facilitates cross-country comparisons.
We measure educational level (H2) by asking “What is the highest level of education you have completed?”. This scale is country-specific, but generally ranges from “Less than primary school” to “Doctoral degree”. As previously mentioned, the country-specific scales have been harmonized to facilitate cross-country comparisons. We have done this by mapping each of the country-specific categories (such as “Abitur” in Germany) to the relevant educational category from OECD’s International Standard Classification of Education (ISCED) (UNESCO, 2012). By harmonizing according to this international scale, we can compare the effect of education across countries with very different educational systems.
To operationalize exposure to the single market, we code those who state that their company i) already operates or sells its goods in another member state of the single market or ii) wants to do the same with 1, and everyone else with a 0. This measure is used to test H3 (that support will be greater among those working in transnationally oriented businesses) and allows us to perform a more nuanced test of the utilitarian framework, as we can assess how both the general and specific utility of the single market impacts citizen support.
To create the measure of economically right-wing views used to test H4, we create an equally weighted instrument measuring assent to two statements (“Government should try to reduce economic inequality” and “Government regulation of business usually does more harm than good”). These have been standardized so that agreement, in both cases, imply a more right-wing political position. We believe this offers a more robust measure of the left-right dimension of politics, which crucially structures citizens’ opinions of economic issues, than simply relying on a self-reported measure of where one positions oneself on the left-right dimension. The reason is that self-reported measures of left-right orientation may be shaped by a combination of views on both economic and non-economic issues (Freire, Reference Freire2015; Otjes, Reference Otjes2018). We have also streamlined the coding schemes of both questions so that agreement implies increasingly right-wing attitudes. The resulting scale goes from 0–4, with values closer to 4 indicating more strongly right-wing beliefs.
To investigate H5a-b, we include an index capturing views of non-economic issues like multiculturalism, environmentalism and international cooperation. These items are predictive of how one positions oneself on the GAL/TAN cleavage that separates social liberals from social conservatives (Marks and Steenbergen, Reference Marks and Steenbergen2002; Dassonneville et al., Reference Dassonneville, Hooghe and Marks2023; Gallina, Reference Gallina2023). To operationalize the three constituent components of the cleavage, we create an equally weighted index based on responses to three questions: “[COUNTRY]’s culture benefits if more people move here”, “Wealthier [COUNTRIES] have an obligation to help poorer [European countries]” and “Politicians should do more to solve the crisis of climate change, even if it may initially hurt jobs in [COUNTRY]” (factor loadings: 0.64, 0.58, 0.62, Cronbach’s alpha = 0.64). Our item measuring transnational leanings is arguably a particularly strong test of whether people hold a preference for transnationalism. In contrast to measures of trust in various international institutions, that are inevitably tied to both normative and utilitarian evaluations of the institutions themselves (De Vries et al., Reference Vries, Catherine, Hobolt and Walter2021), our measure asks about agreement with a general principle of solidarity between European nation-states. We believe agreement with such a general principle is a highly valid indicator of a cosmopolitan world view. As before, the response categories range, on a five-point Likert scale, from “Strongly disagree” to “Strongly agree”. To facilitate the construction of a scale for GAL/TAN values, we reverse the coding of all items of the scale, so that they go from “Strongly disagree” to “Strongly agree” after recoding. The final GAL/TAN scale goes from 0–4, with values closer to 4 suggesting that someone identifies more closely with conservative rather than liberal views (see Table A5 for a description of coding used for all independent variables). We also include a complementary indicator of relative territorial identity – how strongly individuals identify with the nation versus Europe (Griffiths, Reference Griffiths2023). Both measures are grounded in the postfunctionalist literature’s account of the “new politics” cleavage, and particularly the nationalist orientation embedded in the TAN pole. While the GAL/TAN scale emphasizes value-based orientations toward cross-border solidarity and social change, relative identity captures the degree to which individuals locate their political community at the national or supranational level. Including both as alternative operationalizations allows us to assess the robustness of our findings and to examine whether value-based measures or identity-based attachments more consistently predict attitudes toward market integration. Our indicator of European identity asks, “Being European is important to who I am” while our measure of national identity asks, “Being from [COUNTRY] is important to who I am.” Both measures use five-point Likert-scaled items, with response categories ranging from “Strongly agree” to “Strongly disagree.” By subtracting the latter from the former, we get a measure that indicates how much more strongly someone feels attached to Europe compared to their nation-state. We suspect that this measure better operationalizes the relative strength of one’s attachment to Europe than the dichotomous indicators frequently used in the Euroscepticism literature (Hooghe and Marks, Reference Hooghe and Marks2005; Hobolt and Wratil, Reference Hobolt and Wratil2015) as it takes into account that most people see national and European identities as complementary (Risse, Reference Risse2014). Here the values go from −4 to 4, with values closer to the −4 indicating that someone feels much more strongly attached to their nation-state than to Europe. The opposite applies to values closer to 4.
Control variables
We employ a variety of control variables. We first use the independent variables as important controls in situations where they are themselves not the focal variables of interest (see the Model estimation section below for a description of our analytical strategy). As a first set of controls, we add indicators for age, gender and whether the respondent lives close to their countries’ national borders. Our indicator for distance to the national border is operationalized by a question that asks respondents about the driving distance from their home to the border. These are known to impact attitudes towards the EU (Foster and Frieden, Reference Foster and Frieden2021).
Our next set of controls helps us measure that patterns of views on single-market governance do not simply reduce to broad attitudes about the EU or national government. A dichotomous variable codes those who state that they would vote to leave the EU in a hypothetical referendum as “1” and everyone else as “0”. We also include a theoretically important measure of whether respondents are satisfied with the democratic quality of their respective countries and the EU, operationalized by asking whether they believe people like them have a large say over how their country or the EU is governed. Previous literature finds strong correlations between people’s evaluations of their own national democratic systems and how they evaluate the EU, with citizens often using their perception of their countries’ democratic system as a benchmark for evaluating whether the EU and its policies would serve them better or worse than the status quo (Harteveld et al., Reference Harteveld, van der Meer and De Vries2013; de Vries, Reference de Vries2018).
Model estimation
Our primary theoretical aim is to assess whether attitudes toward market integration are best explained by utilitarian calculations, economic ideology, or cultural predispositions – and, critically, whether these dimensions interact in meaningful ways. Throughout, we remove individuals who responded “Don’t know” using listwise deletion. While this approach may lead to bias if some people are systematically more likely to enter invalid responses than others, the large number of respondents in different countries mitigate this concern, and ensures strong statistical power even where any of the included variables have considerable missing data.
As the dependent variables are all dummies we use logistic regression. We use country fixed effects to control for time-invariant variables like welfare systems and history of EU membership, thus reducing omitted variable bias. To account for the two-level structure of the data, with individuals nested within countries, we use country-clustered standard errors (applied with the vcovCL function from the sandwich R package first developed by Zeileis et al. (Reference Zeileis, Köll and Graham2020)). All models include the set of control variables and incorporate country fixed effects to account for time-invariant national characteristics such as EU accession timing, welfare state structure, and labor market regimes.
We begin by estimating the effects of GAL/TAN and left-right orientations independently, in order to test their baseline explanatory power and compare them to utilitarian models on their own terms. This approach allows us to evaluate the additive logic common in much of the public opinion literature, where cultural and economic attitudes are treated as competing or complementary predictors. The results of this approach (Table 1) provide the most direct test of the hypotheses we present above. However, evaluating these dimensions independently assumes that their effects operate uniformly across respondents. Yet a central debate in the literature concerns whether economic ideology retains independent explanatory power once cultural orientations are taken into account, or whether its influence varies depending on individuals’ positions along the identity-based cleavage. To assess the robustness of the additive specification, we therefore also model an interactive relationship between left–right economic ideology and GAL/TAN cultural orientations.
Results of regression analysis, all DVs. Cluster-robust SEs at country level. Country FEs
+ p < 0.1, *p < 0.05, **p < 0.01, ***p < 0.001.
However, we also choose to model an interactive relationship between left-right economic ideology and GAL/TAN cultural orientations in order to offer a more robust analysis of these competing frameworks. Building on prior work suggesting these dimensions may be orthogonal but overlapping in practice (Hooghe and Marks, Reference Hooghe and Marks2009), we test whether the effect of GAL/TAN orientations on single market attitudes is conditioned by economic views. This specification does not introduce new hypotheses, but rather evaluates whether the main effects identified in Table 1 operate consistently across ideological profiles. For instance, we expect that cultural conservatives (TAN) who also hold economically interventionist views may be particularly likely to oppose market integration, while economically liberal individuals may support market openness regardless of cultural leanings.
Formally, the interactive model we estimate for respondent i in country j takes the following form:
${Y_{ij}} = \;{\beta _1}{\rm LeftRight}{_{ij}} + {\beta _2}{\rm GAL/TAN}{_{ij}} + {\beta _3}\left( {{\rm LeftRight}{_{ij}} \cdot {\rm GAL/TAN}{_{ij}}} \right) + {\beta _4}{X_{ij}} + {\lambda _j} + {\varepsilon _{ij}}$
where
${Y_{ij}}$
is an indicator of support for single market integration,
$LeftRigh{t_{ij}}$
and
$GAL/TA{N_{ij}}$
are standardized measures of economic ideological attitudes and postmaterialist attitudes, and their interaction captures the conditional relationship between the two dimensions. The vector
${X_{ij}}$
includes additional utilitarian hypotheses (i.e., education and income) as well as controls and alternative measures of ideological and post-functional constructs (e.g., national identity, partisan alignment, age), and
${\lambda _j}$
denotes country fixed effects. The interaction term allows us to assess whether the effect of cultural orientations differs depending on economic ideology – thereby testing the additive versus interactive implications of the postfunctionalist framework.
Results and discussion
Descriptive evidence and utilitarian predictors
Simple descriptive patterns seem to show why utilitarian theories remain a durable lens for understanding support for European integration. Individuals with more income and higher levels of education – those better positioned to benefit from market openness – tend to express stronger support for the single market. This is clearly visible in Figures A1 and A2, which like all tables and figures beginning with an A can be found in the online Appendix. The figures show positive, non-linear bivariate relationships between both forms of capital and support for movement of people, firms, and EU action.
At the same time, descriptive evidence also supports the postfunctionalist perspective. As Figures A3 and A4 show, socially conservative attitudes and strong national identities are consistently associated with greater opposition to the single market. These relationships are nearly linear and robust across all domains – openness for firms, EU action, and especially personal mobility – suggesting that cultural predispositions shape opinion even in ostensibly technocratic domains.
H1-H3: Functional determinants of Single Market Support
However, implementing an extensive set of controls weakens the case for the utilitarian hypotheses. Rejecting H1, we find no consistently positive relationship between household income and support across the three facets of the single market (see Table 1). We also find no consistently positive and statistically significant relationship between educational level and support for the single market; we thus reject H2. Finally, we also show that exposure to cross-border market activity, whether current or prospective, has a heterogeneous effect on support for the single market. First, the effects are mainly focused on the comparatively less salient issue of openness for businesses, rather than movement of people or EU action. The effects of education and income are largely similar between those who work in transnational or solely national businesses (see Figures A6 and A7), suggesting that neither general nor specific utilitarian models have much explanatory power for understanding variations in views of the single market. We find that the same results hold when we treat the neutral middle category as missing data (Table A15) and as an indicator of opposition to the single market (Table A14).
Cultural predispositions and the conditional role of ideology
Given that our earlier models showed weak and inconsistent support for utilitarian hypotheses (H1–H3), we focus here on the core ideological and postmaterialist hypotheses (H4–H6), using the interactive model as the primary method of analysis. This model distinguishes between the “old politics” of economic ideology (Left-Right) and the “new politics” of postmaterial, cultural predispositions (GAL/TAN and relative territorial identity) and tests whether their effects are additive or conditional.
To visualize the interaction effects, we illustrate support between the economic left and the right (traditional ideology) for respondents that belong to the 20% most socially liberal (GAL) and 20% most socially conservative respondents (TAN). We focus on these two groups as it is where we would expect the role of the postmaterialist cleavage to be most prominent in shaping attitudes towards the single market. It thus presents a very conservative test of whether left-right attitudes have the potential to modify the impact that the postmaterialist cleavage has on support for single market goals.
H4: Left-right ideology and Support for Market Integration
Hypothesis H4 posits that individuals with more economically right-wing views – those more favorable to free markets and skeptical of redistribution – should be more supportive of EU single market integration. This expectation finds limited and conditional support.
H4 posits that individuals with more economically right-wing views – those more favorable to free markets and skeptical of redistribution – should be more supportive of EU single market integration. Table 1 shows that there is evidence that right-wing views may, on its own, condition views of both the free movement of people and goods. However, as we argue above, the effect of left-right ideology may be heterogeneous depending on one’s views also of GAL/TAN issues. In other words, the additive effect of left-right ideology may be weakened if one takes into account also individual respondents’ positioning along the GAL/TAN cleavage. Table 2 finds that this is the case. As Table 2 shows, the Left-Right coefficient is negative and statistically significant only for freedom of movement for persons (−0.75, p < 0.001). It is substantively small and statistically insignificant for support for business and trade openness. This suggests that economic ideology alone does not consistently predict support for market liberalization, even when it is framed at the EU level. These findings challenge the assumption that the single market is still seen as a broadly pro-market project by the ideological right.
Results of analysis showing interaction between left-right orientation and GAL/TAN views. Cluster-robust SEs at country level
+ p < 0.1, *p < 0.05, **p < 0.01, ***p < 0.001.
However, the interaction term between Left-Right and GAL/TAN orientations is positive and significant for labor mobility (0.24, p < 0.001), indicating that economic ideology has stronger effects among those who also hold socially conservative views. As shown in Figure 2, individuals who are both economically right-leaning and culturally conservative are the least supportive of labor mobility. Among culturally liberal (GAL-oriented) individuals, economic ideology has little effect on support. This confirms that economic attitudes matter not on their own, but in combination with cultural predispositions – a core insight of the postfunctionalist perspective.
Support for free movement of business as an interaction between left-right and GAL/TAN orientation. 95% confidence intervals shown.
H5a–H5b: Postmaterialist Attitudes and Single Market Support
Hypothesis H5a predicted that orientations toward the TAN side of the GAL/TAN spectrum would be associated with lower support for the single market. This receives strong support. The coefficient for GAL/TAN is large, negative, and statistically significant across all three dimensions of market integration (Table 2). The effect is strongest for freedom of movement for persons (−1.13, p < 0.001), followed by businesses and trade. These results reinforce the argument that postmaterialist cultural predispositions are the most powerful predictors of support for EU market governance, even in its most technical, regulatory forms.
H5b posited that these effects would be especially strong for labor mobility, which touches directly on national identity and symbolic sovereignty. This expectation is clearly confirmed in Figure 3, which shows steeper and more precisely estimated declines in support for labor mobility among TAN-aligned respondents. This aligns with prior work on cultural threat and group-based opposition to migration (Hameleers et al., Reference Hameleers, Bos and de Vreese2017; Hooghe and Marks, Reference Hooghe and Marks2009). That these patterns hold in the context of intra-EU migration, as distinct from non-EU immigration, underscores how deeply identity-based concerns structure attitudes even toward the internal market. The same pattern, in which the contribution of post-functional and identity-based concerns to support for the single market greatly outweigh those of material concerns, is robust also to modeling the results as OLS or ordinal logistic regressions with Likert-scaled response categories (see Tables A12, A13 and A20) as well as a three-category scale that treats the neutral “Neither” responses as a middle category (see Table A19).
Predicted probability of support as a function of GAL (low) or TAN (high) political orientation. 95% prediction intervals shown.
Importantly, we find no significant interaction between GAL/TAN and income (Figure A8), suggesting that opposition among TAN-oriented individuals is not driven by economic vulnerability. Rather, it reflects deeper cultural orientations – consistent with the “new politics” thesis.
H6a & H6b: Territorial identity as a complementary measure
To complement our attitudinal measure of TAN orientations, we also test a second hypothesis: whether relative territorial identity (i.e., stronger national vs. European identification) predicts lower support for market integration. We find strong support for H6a. Table 2 shows a positive and statistically significant effect of European identity for support for labor mobility (0.18), and smaller but significant effects for businesses and EU internal-market action.
Figure 4 further confirms H6b: the positive effect of European identity is most pronounced for labor mobility. Respondents who identify more with their nation than with Europe are thus substantially less supportive of allowing individuals from other EU countries to work in their own. These effects are substantively smaller than those associated with GAL/TAN, but their consistency across domains – especially for labor mobility – supports the view that identity-based opposition is a meaningful expression of the TAN dimension. Whereas the GAL/TAN scale captures value-based orientations toward social and political change, relative territorial identity captures the symbolic dimension of belonging. Together, they reflect different pathways through which the “nationalism” in TAN is expressed.
Predicted probability of support as a function of European identity. 95% prediction intervals shown.
It is also worth noting that these effects are not driven predominantly by potential multicollinearity between left-right orientations, nationalist identities and GAL/TAN views. Instead, as Table A16 to Table A18 show, removing each of the potentially collinear variables only increases the originally established effects.
Overall, our results suggest that citizens do not think like “pocketbook Europeans” – even in policy areas that are economic and regulatory in nature. Instead, we find that cultural predispositions dominate, particularly on labor mobility, and that their effects are contingent on economic ideology. The theoretical promise of postfunctionalist theory – that identity would increasingly shape support for integration – appears to hold even in the EU’s most technocratic and utilitarian-dominated domains. But our results go further: they show that the GAL/TAN cleavage is not just a second dimension, but a structuring force that conditions how economic ideology itself is activated. In that sense, the politics of the single market are neither technical nor merely distributive – they are cultural, contested, and conditional.
Country-level variations
Our sample contains countries that span the full economic and geographical diversity of the EU. Despite this variation, however, country-level analysis of all three freedoms (see Figures 5–7 for core independent variables, and Table A9 to Table A11 for the full models) shows that attitudes overlap strongly across countries.
Coefficient of independent variables. Support for openness for people. Heteroscedasticity-robust SEs.
Coefficient of independent variables. Support for openness for businesses. Heteroscedasticity-robust SEs.
Coefficient of independent variables. Support for more active EU trade policies. Heteroscedasticity-robust SEs.
However, while our conclusions are broadly applicable across countries, there is still much heterogeneity in the effects of individual variables. One example is the effect of working in a company that sells across countries. Those working in such companies in Romania are much more positively disposed to businesses’ open access to the single market than are their French counterparts. The social conservatism measured by the GAL/TAN indicator has, on the other hand, on average a negative effect on support for the single market across all countries. However, the statistical precision and predictive power of the effects vary strongly from one country to another.
We might have expected utilitarian factors to play a different role in Spain and Italy, our two Southern European countries. During and after the sovereign debt crisis both countries were affected by austerity policies with detrimental effects, especially in areas where employment opportunities were scarce (Cremaschi et al. Reference Cremaschi, Rettl, Cappelluti and De Vries2025). Table A2 shows that whereas higher education levels increase support for free movement of workers in both the East and the West, as we would expect, it does the opposite in Southern Europe. However, as Figure A11 shows, the correlation between support for labor migration and education is actually more comparable between Southern Europe and the other regions surveyed than we would expect.
Nevertheless, when using a dummy that captures the peripheral status of the Eastern European states in our sample, the relationship changes somewhat. As Table A3 shows, there is a positive correlation between being a citizen of either state and support for the single market. Also true is that the interaction effects discussed in the previous section are stronger and more consistent in Western Europe – particularly in North and South Western regions – than in Eastern Europe (Table A6 to Table A8). This is consistent with prior findings that the GAL/TAN cleavage plays a smaller structuring role in Eastern European public opinion (Marks et al., Reference Marks, Hooghe, Nelson and Edwards2006; Rohrschneider and Whitefield, Reference Rohrschneider and Whitefield2016). Overall, the EU’s main geographical cleavage when it comes to support for the single market goes between the newer Eastern member states and the older member states of Western and Southern Europe, and not between a north and a south with very different experiences of the sovereign debt crisis.
Conclusion
This paper sheds light on the under-studied question of how people form opinions about the regulatory policies associated with the EU’s original and enduringly core project of building a single market. We find that the postmaterialist and identity-based explanations that have become increasingly important to understanding diffuse support for the EU (Hobolt and de Vries, Reference Hobolt and de Vries2016) also offer the greatest explanatory power for support toward the principles underpinning the EU’s single market. Support for cross-border openness – whether for workers, businesses, or EU action to promote it – is, like broader views of European integration, predominantly shaped by postmaterialist concerns. Contrary to common expectations, we show that the “old” politics of functional integration is just as vulnerable to politicization along identity lines as “new” integration efforts. Indeed, the new politics cleavages not only shape views of the single market on their own terms, but also interact with economic ideology in important ways: strongly socially conservative, economically right-wing respondents – those who are also typically Eurosceptic – exhibit the lowest support for market openness, suggesting that cultural predispositions condition the effect of economic views.
These findings expand our understanding of the dimensionality of public support for EU integration and challenge assumptions that the single market remains safely technocratic. We also find sharp differences in public reactions to trade and labor migration, compared to more general attitudes toward EU regulatory authority. This distinction reinforces a core insight of postfunctionalist theory (Hooghe and Marks, Reference Hooghe and Marks2009): that it is not whether a policy is “old” or “new” that determines politicization, but whether it is culturally resonant and publicly salient. Scholars seeking to understand when integration will provoke a “constraining dissensus” should therefore focus not on policy age, but on the cultural visibility of integration’s effects. Our findings affirm that even domains long presumed to be insulated from mass contestation are now interpreted through symbolic and identity-laden frames. In that sense, this paper lays the groundwork for what might be called a post-postfunctionalist research agenda, in which identity-based conflict is not the exception but the rule, even in economic governance.
Our results also intervene directly in current debates about the future of the single market. While the EU is “much more than a market,” our findings complement two recent, high-profile blueprints for its future. Both the reports from Enrico Letta (Reference Letta2024) and Mario Draghi (Reference Draghi2024) implicitly recognize a core conclusion of this paper: public support for the EU’s market project is not simply a matter of functional economic benefit, but is deeply shaped by citizens’ identities and cultural attachments. In particular, these insights resonate with Letta’s call for a reimagined single market that promotes economic growth while acknowledging Europe’s diverse cultural fabric. A robust and future-proof single market must therefore attend not only to distributive outcomes, but also to the symbolic politics that define its legitimacy.
Letta (Reference Letta2024) further notes that the single market of tomorrow will likely expand in size and scope. Our data show just how divided publics remain, even on relatively low-salience market policies. If future expansion includes countries like Serbia or Ukraine, existing cleavages – especially around internal labor migration – may intensify. As such, policymakers would be wise to consider not only institutional and economic feasibility, but the cultural meaning and potential contestation that further integration may provoke.
One important caveat to our results points towards a need for further research: Our sample does not include Nordic EU member states such as Sweden and Denmark. This is an important limitation, because Nordic elites have typically adopted a strongly economic perspective on what EU membership is (Bølstad, Reference Bølstad2015). Future work should, in other words, probe how the results of our analysis travel also to this region.
Finally, an important question for future research is how individuals think about the trade-offs between national sovereignty and international cooperation. Our findings suggest that identity-based skepticism of integration is not simply about perceived economic losses, but about symbolic control and group boundaries. Do citizens support deeper market integration when they perceive that their state retains the ability to opt out? Do preferences for harmonization versus flexibility vary by policy domain? Future work should explore how preferences for “opt-out” provisions and shared authority interact with public attitudes toward the EU. In doing so, we can better understand whether integration fatigue stems from economic concerns, cultural resistance – or, most likely, the complex interplay of both.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S1755773926100526.
Data availability statement
Full replication data can be found at https://doi.org/10.7910/DVN/NHVBGN.
Acknowledgements
Various iterations of this manuscript have benefited from the insightful feedback of many readers. We would especially like to thank the discussants and participants at the American Political Science Association 2024, the ECPR Standing Group on European Union 2024, the Conference of European Studies 2023 and an ARENA Tuesday Seminar for valuable comments and suggestions.
Financial support
The authors gratefully acknowledge funding from the Research Council of Norway, through the SINGLEMARKETS grant. The authors declare that all data collection has happened in accordance with the relevant laws.
Competing interests
Do not declare any conflicts of interest.
Open access
