How occupation shapes awareness and preferences for European funding

Abstract

The European Union’s (EU) Cohesion Policy is one of the main instruments to tackle the economic struggle of left-behind regions, however, only little is known about the awareness and the opinion of citizens towards this regional policy. This article examines how awareness of EU regional policies and spending preferences varies across occupations. We argue that differences in awareness across occupations are attributable to skill and responsibility differences and divergent work logics. Our analysis of a harmonized Eurobarometer dataset ( $N=$ 82,365) provides evidence for differences in EU funding awareness and spending preferences across occupational groups, which correspond to their own material interests. In addition, citizens indicate preferences for funding decisions to be made at the most local level.

Introduction

Occupation shapes how citizens engage with broader political processes such as European integration, public investment, or globalization (Balestrini, 2012; Hooghe et al., 2007; Kitschelt and Rehm, 2014). It is also associated with the electoral success of populist radical-right parties (PRRPs) in Europe. Blue-collar workers and the unemployed are more likely to vote for such parties (Gidron and Hall, 2017; Oesch and Rennwald, 2018). This occupational pattern is also reflected in a spatial dimension. Regions with former high industrial activity are particularly susceptible to the populist backlash (Broz et al., 2021). In response to the economic hardship of these regions, place-based policies like the EU Cohesion Policy aim to foster economic development and, as a side effect, enhance public opinion towards the European Union (EU).

With an approximate budget of 392 billion euros in the 2021–2027 funding period, it constitutes a substantial component of the economic policy of the EU and one of the largest place-based policies in the world. Scholars and policy-makers have considered regional policies as a working tool to tackle challenges such as the rise of PRRPs (Albanese et al., 2022) or the general increase of political discontent in left-behind regions (Gold and Lehr, 2024). However, the effect of EU Cohesion Policy on EU attitudes is a topic of vivid scholarly debate. While various studies find a positive association between place-based policies and improved attitudes towards the EU (Borin et al., 2021; Gold and Lehr, 2024; Rodríguez-Pose and Dijkstra, 2021), there is also contrasting evidence challenging these findings (Di Cataldo, 2017; Ward et al., 2025).

Ward et al. (2025) provide evidence that, in the case of Wales, there is no positive association due to the lack of citizens’ awareness of European projects. Surprisingly, we know little about the drivers of such awareness (but see Natili et al., 2023; Osterloh, 2011) or the preferences of citizens regarding these funding schemes. Nevertheless, public opinion towards such schemes is likely to be an important component in better understanding how they translate into attitudinal changes towards the EU. Given the wide range of programmes and policy domains supported by the Cohesion Policy, we argue that there should be substantial variation in awareness between funding schemes. To address this research gap, this article investigates citizens’ awareness and preferences toward EU Cohesion Policy, highlighting occupation as a central explanatory factor.

Relying on the theoretical framework by Oesch (2006), we expect a significant difference in awareness of EU Cohesion Policy across occupational groups. We outline how occupational differences on a vertical dimension relating to hierarchy and skill, as well as a horizontal dimension relating to various work logics, drive variation in awareness. Based on this, we illustrate our expectations in Figure 1, stating our theorized expectations for levels of EU funding awareness. To the best of our knowledge, this is the first time that Oesch’s framework has been applied to the context of EU place-based policies. In addition, we argue that occupation not only influences awareness of EU Cohesion Policy but also affects citizens’ spending preferences and what regions they prefer as recipients. Existing research has demonstrated a link between occupation and attitudes toward redistribution (Busemeyer and Sahm, 2022; Thewissen and Rueda, 2019). Thewissen and Rueda (2019) find that occupational risk due to automation is associated with higher demand for redistribution. Similarly, we expect citizens to follow their material self-interest and exhibit preferences that likely benefit themselves. Finally, we expect that occupation also plays a role in shaping preferences for the decision-making level for EU projects. Here, we build on the literature that relates individuals’ skill level and their response to globalization pressure (Kuhn and Stoeckel, 2014; Walter, 2017) and literature on general EU support (Hakhverdian et al., 2013; Hobolt, 2016). Specifically, we argue that individuals in higher-skilled occupations are more likely to support the EU and, consequently, to favour decision-making on EU place-based policies at the supranational level. In contrast, individuals in lower-skilled occupations are expected to exhibit stronger preferences for local decisions on such matters.

Figure 1.

Awareness of European Union (EU) Cohesion Policy across occupational classes based on Oesch (2006). Note: This figure illustrates our ideal-typical expectations how the vertical and horizontal dimensions by Oesch (2006) shape awareness of the EU Cohesion Policy. Source: Own illustration.

Our analysis relies on three harmonized Flash Eurobarometer datasets (European Commission, 2015, 2017, 2020) covering all EU member states between 2015 and 2019 ( $N=$ 82,365). We estimate a set of logistic and multinomial regression models controlling for age, gender, and education as well as survey and region fixed effects. Additional analyses include controls for regional differences and regional corruption perceptions.

This article shows that awareness levels differ significantly across regions. As the literature often relies on single-country case studies (e.g., Bachtrögler and Oberhofer, 2018; Debus and Schweizer, 2024; Di Cataldo, 2017; Di Cataldo and Renzullo, 2025), our research has substantial implications for case selection and generalizability of results, because research focussing on countries with a low level of awareness might underestimate the general effect. Also, we provide evidence for a substantial occupational gap in awareness within regions. Citizens who are out of labour or work in manufacturing know significantly less about EU funding activities than managers, professionals, farmers, and shop owners. This supports our theoretical expectation that awareness varies across an occupational vertical and horizontal dimension. Further, we find that citizens follow egocentric considerations and prefer spending targets that are likely to benefit them the most. Finally, citizens favour funding projects to be decided upon at the local level across all occupations. Contrary to our expectation, individuals in higher-skilled occupations are also more likely to prefer decisions to be made at the local rather than the European level.

The literature on EU place-based policies has so far only marginally engaged with public opinion towards these schemes (cf. Ciornei, 2025) and even less with citizens’ awareness of them (cf. Osterloh, 2011). This article advances this research agenda by arguing that both awareness and preferences are central to understanding the political consequences of EU place-based policies. Empirically, the study shows significant differences across occupational groups in their awareness of EU-funded projects and in their preferences for Cohesion Policy spending and uses categories by Oesch (2006) to explain this variance. By identifying occupation as a key explanatory factor for awareness and preferences, the article offers a novel account for the mixed evidence on the political effects of place-based policies.

Theory

Individuals’ occupation influences their political attitudes (Balestrini, 2012; Hooghe et al., 2007; Kitschelt and Rehm, 2014) and political knowledge (Carl et al., 2019; Clark, 2014). Beyond income or education, occupation reflects a mixture of social experiences, economic considerations, and interactions. Citizens’ occupational positions determine their exposure to state institutions and information channels, as in the case of civil servants and employees in unionized workplaces, or to globalization pressures, as among manufacturing workers. These examples also speak to two mechanisms at play. First, occupation influences interpersonal interactions and exposure to information, thereby shaping how citizens form political views. For instance, political knowledge is higher in union workplaces due to political discussions (Macdonald, 2021). Second, it is related to economic risks and opportunities. For example, working in occupational sectors exposed to automation is associated with a higher demand for redistributive policies (Busemeyer and Sahm, 2022; Thewissen and Rueda, 2019). In contrast, the effect of exposure to globalization is conditional on individuals’ occupational skills. The lower the skills, the higher the support for redistribution (Walter, 2017).

Occupation and awareness of EU cohesion policy

Most empirical studies implicitly assume that citizens are aware of EU Cohesion Policy when studying its effect on outcomes such as support for further European integration or electoral success of Eurosceptic parties and do not investigate awareness as a dependent variable in its own right. In contrast, Natili et al. (2023) incorporate individuals’ awareness of EU social programmes explicitly in their theoretical framework, investigate antecedents of this awareness, and point out the low awareness of EU funding across Europe, despite significant country variation. Awareness is comparatively higher in countries that receive more funding, such as Hungary or Poland. More than half of the Hungarian and close to half of the Polish respondents heard about EU funding, whereas less than one out of five respondents in France, Spain, or Sweden is aware of funding. Dabrowski et al. (2021) add more empirical evidence on citizens’ low awareness of EU funding in the Netherlands.

Natili et al. (2023) find that individuals who are employed are more likely to be aware of EU Cohesion Policy, based on original survey data collected in 2019. Similarly, Osterloh (2011) investigates predictors of awareness for the 1994–1998 funding period. Awareness is significantly associated with citizens’ socio-demographic characteristics. Specifically, Osterloh focuses on education. However, he also provides evidence of occupational effects, i.e., retired citizens are less likely to be aware of EU transfers and individuals in executive positions are more likely to be aware of such transfers.

Beyond awareness of EU funding, occupation has been studied as a predictor for political knowledge about the EU more in general (Carl et al., 2019; Clark, 2014). These studies find no robust evidence for an association based on occupation, however, they rely on more abstract knowledge items, for example, questions about specific countries’ EU membership. We argue that the mechanism behind awareness or knowledge about EU Cohesion Policy is a different one compared to factual knowledge about the EU. Specifically, such knowledge is more a function of education, whereas we link awareness of EU funding to individuals’ work logic and (material) self-interest.

We derive our hypotheses for the association between occupation and awareness about EU Cohesion Policy drawing on the occupational class scheme by Oesch (2006, 2013). This scheme captures both the vertical dimension of employment relations (e.g., autonomy, authority, and control over work) and the horizontal dimension of individuals’ work logic (e.g., technical, interpersonal, organizational, or independent). The vertical dimension refers to where an occupation is situated in terms of class hierarchy or skill level. One can differentiate between employers, self-employed, and employees. The category of employees can be even further differentiated. The higher an individual is located on the vertical dimension, the higher her or his position or skill level. The horizontal dimension refers to the type of work performed or the logic of the job. Individuals can work either independently, in a technical profession, in a line of work that is primarily concerned with organizational tasks, or engage in interpersonal tasks.

Awareness of EU Cohesion Policy is unlikely to be uniform. By combining the vertical and horizontal dimensions, Oesch’s scheme can offer a nuanced way to understand variation in awareness across occupations. Individuals in high-skill, organizational occupations (e.g., public administrators, managers in firms benefiting from EU programmes) may be systematically more informed than routine interpersonal workers (e.g., retail or care workers), simply because it is their job to do so. Similarly, self-employed artisans or small business owners may have specific awareness as they directly interact with EU-funded projects.

We argue that awareness of EU funding varies across occupations and that it is a function of exposure to such funding due to the vertical and horizontal dimensions associated with particular roles. Figure 1 presents our ideal-typical expectations, illustrating how the vertical and horizontal dimensions shape awareness of EU Cohesion Policy. We further map the occupations observed in our data onto this two-dimensional grid to offer a visual aid that supports and clarifies our theoretical expectations.

First, we expect that individuals in occupations placed higher along the vertical dimension are more likely to be aware of EU funding. On the one hand, these occupations are more likely to involve higher levels of decision-making or financial planning. This includes positions at a managerial level, business owners, or other high-responsibility roles. We argue that individuals in these roles are more likely to be exposed to and to benefit from EU funding. Such benefits may manifest indirectly, for example, through competitive or operational advantages for these individuals’ employers, or directly in the case of self-employed individuals. It is important to note that the two-dimensional grid (Figure 1) serves foremost as a visual aid and should not be regarded as a definitive analytical tool. For example, farmers are a prominent example of an occupational group deviating slightly from our visualization. Farmers have high exposure to EU funding, particularly through the Common Agricultural Policy, which supports their economic viability. In the case of farmers, awareness of these funding mechanisms is not only advantageous but arguably indispensable for the sustainability of their livelihood. Hence, farmers should be among the occupational groups with the highest awareness of EU funding.

H1a: The higher an individual is placed on the occupational vertical dimension, the more likely he or she is to be aware of EU funding.

Second, we expect differences in awareness based on variation on the horizontal dimension. More specifically, the different work logics on the horizontal dimension – the independent, technical, organizational, and interpersonal work logics – imply varying degrees of relevance and visibility to EU funding. Self-employed individuals operating under an independent work logic (e.g., small business owners, farmers) are key targets for EU entrepreneurship and agricultural funding programmes. In contrast, occupations following an interpersonal work logic should be least likely to be exposed to EU funding. We argue that funding is both less relevant for their professional success and less visible for these occupational groups. Finally, we propose that occupations characterized by organizational or technical work logic should be placed between the independent and interpersonal work logic. Occupations following an organizational work logic (e.g., managers, administrators) should be more directly involved in applying for and managing external funding, whereas technical workers (e.g., technicians, skilled trades) may benefit from vocational funding programmes or work on projects funded by the EU. Consequently, we argue that these benefits result in higher exposure and higher visibility of EU funding.

H1b: EU funding awareness varies by horizontal occupational work logic, ranging from lowest among the interpersonal work logic to highest among independent work logic, with technical and organizational occupations in between.

Regional and policy spending preferences

Occupation should not only shape awareness of EU funding, but also individuals’ preferences regarding how such funding should be spent. We know that occupation can influence political attitudes (Balestrini, 2012; Hooghe et al., 2007; Kitschelt and Rehm, 2014) and more specifically attitudes towards redistributive policies (Busemeyer and Sahm, 2022; Thewissen and Rueda, 2019). This article investigates individual spending preferences when citizens face choices across different regional and policy goals. We expect that citizens act in line with rational self-interest, particularly when the allocation of funding directly affects their livelihoods or regional development.

Existing research supports the notion that citizens think about European funding in a rational way. For example, they prefer targeted EU investments, that is, support for the EU is higher when a regional economic need is met (Dellmuth and Chalmers, 2018). In addition, we know that occupation shapes policy preferences based on material self-interest on a more general scope. For example, working in polluting industries makes it less likely to support climate co-operation (Bechtel et al., 2019) or vote for parties with a green or environmentalist agenda (Cavallotti et al., 2025) and makes it more likely to support parties opposing a green transition (Heddesheimer et al., 2024). Also, Unger et al. (2023) highlight the importance of citizens' rational calculation of their country's benefit, for the policy measures that were taken in response to COVID-19. Last, Reinl and Braun (2023) show that highlighting personal advantages is most effective in increasing support for further EU integration.

In this context, however, a different strand of literature argues that social identity plays a crucial role in the formation of redistribution preferences in a European setting (e.g., Bechtel et al., 2014; Kuhn et al., 2018; Van der Waal et al., 2010). For instance, Bechtel et al. (2014) show that in the case of the Eurozone bailouts in 2011, it is social dispositions such as altruism and cosmopolitanism that have stronger predictive power on support for redistribution than economic self-interest. This is in line with general literature on economic shocks, which shows a tendency towards sociotropic behaviour in citizens in times of crisis (Colantone and Stanig, 2018). Kuhn et al. (2018) build on this argument with evidence from a survey experiment, demonstrating that, beyond economic shocks, self-interest plays a less significant role in shaping redistribution preferences within the EU than a cosmopolitan identity. However, Kuhn et al. (2018) do not distinguish between different policy domains in their analysis, but instead ask respondents about general redistribution preferences within the EU. We seek to extend this approach by arguing that, when preferences are considered across specific policy fields, respondents are likely to favour redistribution within the EU in areas from which they expect personal benefits.

We test this self-interest argument for spending preferences in terms of policy programmes and regional characteristics using a selection of occupational groups in our data. Specifically, we rely on occupations with an expected link between their occupational characteristics and spending preferences. In addition, we provide an explorative analysis of occupational spending preferences in cases where no clear theoretical expectation exists.

The article examines spending preferences across policy programmes and regional characteristics. Compared to self-employed professionals such as lawyers or accountants, (unskilled) manufacturing workers are expected to show stronger support for spending on vocational or employment training and social projects, as these areas are more likely to benefit them directly. Self-employed professionals and shop owners, by contrast, are likely to be most supportive of spending targeted at small and medium-sized enterprises. Regarding regional characteristics, farmers are expected to favour spending directed toward rural regions, due to their higher likelihood of residing and working in these areas compared to other occupational groups.

H2: The more likely it is that spending benefits individuals’ own occupational group, the more likely they are to prefer such spending.

Preferred levels of decision-making for EU funded projects

Beyond awareness and spending priorities, a third dimension of public attitudes toward EU place-based policies concerns the preferred level of decision-making authority, that is, whether projects should be decided at the local, national, or European level. We expect preferences for where project decisions should be made are deeply intertwined with individuals’ exposure to a globalized economy.

Our argument builds on existing research concerning how individuals with different skill levels respond to the pressures of globalization (cf. Kuhn and Stoeckel, 2014; Walter, 2017). This also aligns with Oesch’s vertical dimension (Oesch, 2006, 2013). In particular, we argue that occupational groups that require higher skill levels, such as technical professionals, managers, and self-employed entrepreneurs, should perceive EU-level governance as advantageous because they are more likely to benefit overall from regulatory stability, broader market access, or funding opportunities. Consequently, we expect individuals in high-skilled occupations to also be more likely to prefer decision-making for EU funding at the European level. Moreover, we expect this pattern to be observed as typically high-skilled occupations and citizens with higher education are more likely to be generally supportive of the EU (Hakhverdian et al., 2013). Higher trust likely translates into higher support for decision-making at the EU level among these citizens.

Occupational groups with lower skill levels, such as routine manual workers or routine service workers, in contrast, should be more likely to support national or regional control over EU place-based policies. Local decision-making offers the potential for greater responsiveness to their material and symbolic needs, and can act as a buffer against being “left behind” in processes of economic restructuring. Hegewald (2024) shows that citizens with high place-based resentment also exhibit increased trust in local institutions. Hobolt (2016), in turn, highlights the importance of occupational skills for support of the EU, underlining why citizens with lower skills can be expected to be more favourable of decisions on EU projects at the local level.

To summarize, we suggest that due to the overall importance of the socio-economic background of citizens in the preference formation process regarding the EU (Hakhverdian et al., 2013; Hobolt, 2016), we can also expect occupational differences to occur when it comes to the preference on the decision level for EU place-based policy projects. Thus, we hypothesize:

H3: Individuals with higher skill levels are more likely to prefer European decision-making on EU projects, while those with lower skill levels are more likely to prefer more local decision-making.

Research design

Data

The analysis combines and harmonizes three waves of the Flash Eurobarometer surveys conducted between 2015 and 2019 (European Commission, 2015, 2017, 2020). These surveys ask citizens in 26 EU countries about their awareness and opinions on EU regional policies. In Table 2 in the Online appendix, we provide a full table of the number of observations for each country. We exclude subsequent Eurobarometer waves from the main sample because their response categories for the dependent variables deviate from earlier waves. To test whether our findings persist, we analyze the 2023 Flash Eurobarometer (European Commission, 2023) separately and present the results in the Online appendix.

We examine four different dependent variables: Project knowledge, spending preferences, regional preferences, and decision-level preferences. Table 1 summarizes all our dependent variables, with their respective answer categories.

Table 1.

Dependent variables and their answer categories.

Dependent variable	Description	Answer categories
Project knowledge	Knowledge aboutEuropean Union (EU) regional policy	1 $=$ Yes, 0 $=$ No
Spending preferences	Item battery of 7 possible investment fields for regional policy	1 $=$ More important, 0 $=$ Less important
Regional preferences	Forced choice item battery with regional preferences (max 3)	1 $=$ Mentioned, 0 $=$ Not Mentioned
Decision level	Forced choice item asking about the preferred decision level	1 $=$ Local, 2 $=$ Regional, 3 $=$ National, 4 $=$ EU

Note: Spending Preferences and Regional Preferences represent sets of related survey items using the same response formats as shown above.

In addition to these variables, the data also contains socio-demographic information about respondents. This includes gender, age, education, and occupation. For our analysis, we summarized the occupation category from 22 to 8 categories, resulting in the following: farmers, shop owners, managers, manufacturing workers, professionals, skilled office workers, routine service workers, and out of labour. In the Online appendix, we provide a detailed list illustrating the mapping of the occupation categories. When aggregating the occupations, we followed Oesch’s class scheme as closely as possible. Finally, the data also contains spatial information about the region of a respondent. For most countries, this information is available on the NUTS-2 level. However, for Germany and Italy, this information is only available at the NUTS-1 region level, while for Estonia and Latvia, NUTS-3 information is available. Either way, the aggregation level of the regional units for our respondents is arguably sufficient from a theoretical standpoint, as large portions of EU regional policies are allocated based on the economic performance at the NUTS-2 level.

While the available data offers valuable opportunities to analyze awareness and spending preferences, it is important to consider certain technical and empirical limitations when interpreting the results. Because the survey waves of our sample do not provide a harmonized measure of general EU support, we cannot directly model what is plausibly an important control variable in occupation-level comparisons. We account for this issue by reiterating our analysis for awareness and decision level preferences with a proxy variable in a 2023 Eurobarometer (European Commission, 2023), asking citizens how much EU regional policy let them feel European. We argue that this variable is likely to be strongly correlated with general EU identification. We can show that our results remain robust to the inclusion of this covariate, indicating that differences across occupations are not a result of prior EU support. Results of this can be found in the Online appendix.

Estimation strategy

In this article, our key interest is threefold: First, we aim to understand the variation of EU regional policy awareness. Second, we want to examine how spending preferences differ among citizens. Specifically, we examine spending preferences for policy programmes as well as preferences based on regional characteristics. Third, we are interested in citizens’ preferences about the level of decision-making for EU regional policies. Most of our dependent variables are coded binary (Awareness $=$ 1 (yes), 0 (no); Spending preferences $=$ 1 (important), 0 (less important)). Therefore, for models examining these dependent variables, we rely on a logistic regression including region and survey-wave fixed effects as well as age, gender, and education controls. By including fixed effects for survey and region, we aim to control for a spatial and temporal dimension. Thus, we only compare occupations within the same year and the same region. As shown below, we present the results as predicted probabilities for each outcome across occupational groups.

$$\text{logit}[Pr(y_i=1)]={\beta }_0+{\mathit{\beta }}_{\text{occ}}{\mathbf{O}}_i+{\mathit{\beta }}_{\text{ctrl}}{\mathbf{X}}_i+{\mathit{\delta }}_{\mathrm{region}}+{\mathit{\theta }}_{\mathrm{survey}}$$

(1)

${\beta }_0$ is the intercept, representing the baseline log-odds of awareness. ${\mathit{\beta }}_{\text{occ}}$ captures the effects of occupational groups, while ${\mathit{\beta }}_{\text{ctrl}}$ contains the coefficients for individual-level controls, including age, gender and education. ${\mathit{\delta }}_{\mathrm{region}}$ and ${\mathit{\theta }}_{\mathrm{survey}}$ denote the coefficients for region and survey-wave fixed effects, respectively.

Finally, the variable on the preferred decision-making level for EU projects has multiple outcomes. Accordingly, we estimate a multinomial regression, where $k$ indexes the possible outcome categories (local, regional, national, and EU-level). Apart from the functional form, all model specifications remain identical to the logistic regression described above.

$$Pr(y_i=k)=\frac{\exp ({\beta }_{0k}+{\mathit{\beta }}_{\text{occ},k}{\mathbf{O}}_i+{\mathit{\beta }}_{\text{ctrl},k}{\mathbf{X}}_i+{\mathit{\delta }}_{\mathrm{region},k}+{\mathit{\theta }}_{\mathrm{survey},k})}{{\displaystyle \sum _{j=1}^K\exp ({\beta }_{0j}+{\mathit{\beta }}_{\text{occ},j}{\mathbf{O}}_i+{\mathit{\beta }}_{\text{ctrl},j}{\mathbf{X}}_i+{\mathit{\delta }}_{\mathrm{region},j}+{\mathit{\theta }}_{\mathrm{survey},j})}},k=1,\dots ,K.$$

(2)

${\beta }_{0k}$ denotes the intercept for outcome $k$ . ${\mathit{\beta }}_{\text{occ},k}$ captures the effects of occupational groups for outcome $k$ , and ${\mathit{\beta }}_{\text{ctrl},k}$ contains the coefficients for the individual-level control variables. ${\mathit{\delta }}_{\mathrm{region},k}$ and ${\mathit{\theta }}_{\mathrm{survey},k}$ represent the coefficients for the region and survey-wave fixed effects, respectively. All coefficients are estimated relative to the reference outcome category.

A potential threat to our estimation strategy for our hypotheses 1a and 1b and hypothesis 3 lies in the missing control for general EU support. This variable is likely to have explanatory power for awareness of EU funding and preference for decision-making levels. Unfortunately, the data at hand does not contain any control that would allow us to test the substantial effect of this variable. Nevertheless, we argue that this potential issue does not apply to our other hypotheses, as regional and policy programme spending preferences are arguably independent of EU support. To tackle this issue further, we reiterate the analysis of awareness and decision-making preferences with the Flash Eurobarometer 2023, which includes a proxy question on EU identity. ¹ In a separate robustness check, we control for it in our regression, so that EU identity cannot explain differences between occupational groups. The results are included in the Online appendix.

Results

EU project awareness

First, we illustrate the spatial differences in awareness of EU funding across EU regions. Figure 2 illustrates the share of citizens who are aware of EU regional policies. The pattern we observe matches the results from Natili et al. (2023) and Dabrowski et al. (2021). It illustrates that there are considerable differences across regions. This could explain why studies focussing on different regions contradict each other. Overall, the map shows that awareness is particularly high in regions that receive relatively high levels of EU financial support. For instance, it seems that most respondents in all Polish regions have heard about EU regional policies, whereas in the South of Germany, only as little as a quarter of all respondents are aware of such funding schemes. Also, while awareness of EU projects is generally high in Italy, it can be seen that, especially in the strongly supported southern regions of Italy, awareness is higher than in the northern parts. This pattern continues when looking at the UK, where Wales benefited particularly from EU support, the region with the highest project awareness in the UK. Nevertheless, the level of awareness is overall rather low, with around 31 percent. This is roughly on par with the 25 percent awareness observed by Ward et al. (2025). At first glance, it seems that higher EU contribution in the form of regional policies is correlated with higher awareness of such supporting schemes. Most importantly, based on the results of this map, we argue that scholars working on EU place-based policies should take into account the differences in EU funding awareness when conducting case studies.

Figure 2.

Share of European Union (EU) projects heard about across regions. Note: The map illustrates the average awareness from 2015-2019 across regions in the EU.

Figure 3 shows the results of a logistic regression with region and year fixed effects and demographic controls. All results only compare individuals within the same regional unit and the same year. We present predicted probabilities for several occupational groups, in order to examine our first hypothesis regarding varying awareness about EU projects along the vertical and horizontal dimensions of Oesch’s occupational scheme. On average, and depending on occupational group, 35 to 55 percent of individuals are aware of EU Cohesion Policy. Considering how complicated the issue of regional policies can be, we consider this share to be rather high.

Figure 3.

Awareness about European Union (EU) projects by occupation. Note: This figure shows the predicted probabilities for our first hypothesis. It relies on a logistic regression model. A higher probability indicates higher level of EU funding awareness. The model includes controls for age, education and gender as well as region and survey-wave fixed effects. It also includes post-stratification weights on the country level provided by the Eurobarometer dataset.

Turning to the hypothesized awareness differences along the vertical dimension that can be understood as varying hierarchy or skill level, the results show that routine service workers and manufacturing workers are significantly less aware of EU regional policies than the other occupational groups. In addition, professionals have the second highest predicted probability of being aware of EU funds. This lends support to H1a. Regarding the horizontal dimension, the fact that routine service and manufacturer workers are least aware compared to occupational groups following an independent logic, such as shop owners or farmers, supports H1b. Overall, the ordering of the occupational groups’ predicted probabilities is roughly in line with our ideal-typical expectations visualized in Figure 1.

The results are robust to the inclusion of an EU-identity control, estimated with an additional 2023 sample to account for general EU support. This exercise confirms that our findings are stable in periods beyond the observed sample years. Thus, differences in the predicted probabilities cannot be explained by differences in EU support across occupations. Furthermore, we also test whether this awareness bias significantly differs between regions with generally high or low average awareness. While the differences between occupations are slightly stronger in regions with high average awareness, the occupation bias pattern found in Figure 3 persists. This indicates that across all EU regions, an occupational awareness gap on EU funding occurs, and that it is independent of the overall average EU funding awareness of regions. We also examine which information source citizens used when they first heard of EU regional policies. For all occupations, television was the medium where they first learned about EU regional policies the most. The second most informative medium in this regard seems to be print media. Nevertheless, we also find significant differences across occupational categories. Shop owners are more likely than manufacturing workers to have learned about EU funding through the internet. Further, managers, professionals, and skilled office workers are significantly more likely to have encountered information about EU funding in the context of their work. These occupational differences in information channels add to the idea that work logic plays an important role in individuals’ awareness of EU funding. The presentation of these additional analyses can be found in the Online appendix.

Spending preferences

The next set of results turns to the analysis of spending preferences, testing our second hypothesis regarding self-interest. Here, we expect citizens to follow egocentric preferences, thus supporting spending in policy areas and regions that are likely to be most beneficial to them. This entails the analysis of two dependent variables, as indicated in Table 1. First, we examine differences in spending preferences, asking for policy fields that respondents consider especially important for regional policy schemes. Second, we examine which regions citizens prefer money to be allocated to.

Figure 4 presents the results of the first dependent variable, policy spending preferences. The predicted probabilities of respondents viewing a policy field as important for EU funding are generally high. In addition, research expenditures appear to be at the lower end of the range of results, with even lower predicted probabilities associated with manufacturers and farmers. Support for the allocation of funds to social projects is widespread among all occupational groups. While the predicted probability of manufacturers and the unemployed is higher, almost all occupations obtain a predicted probability to favor this kind of policy, with around 90%. Overall, this provides limited support for our second hypothesis that individuals tend to favor policies from which they are likely to benefit.

Figure 4.

General spending preferences. Note: This figure shows the predicted probabilities for our second hypothesis. It relies on a multiple logistic regression model. A higher probability indicates higher personal importance consideration for European Union (EU) funding for the policy area. The model includes controls for age, education and gender as well as region and survey-wave fixed effects. It also includes post-stratification weights on the country level provided by the Eurobarometer dataset.

Turning to regional spending preferences, Figure 5 shows that generally predicted probabilities indicate higher variance. This is likely to be a result of the forced-choice item battery that restricts respondents to choose only three categories, while for the preferences above, respondents were allowed to indicate all policies. We can observe considerable variation both across region preferences and across occupations, although the variance across region preferences is higher. Respondents most frequently prefer allocating funding to regions with high unemployment, and this preference is similarly strong across occupational groups. In turn, only very few respondents support allocating funding to developed regions or border regions. Thus, it appears that respondents focus on economically struggling regions, as indicated by their substantial preference for high-unemployment areas. However, they also want regional policies to extend beyond just poor regions. We expected farmers to favour EU regional policies that prioritize rural areas. The results confirm this expectation, showing that farmers are more supportive of directing EU funding toward rural regions. Overall, the differences between occupational categories for regional spending preferences are not as substantial as we expected them to be. While we do find significant differences, they remain small in size. To summarize, it seems that citizens have similar regional preferences as shown in Figure 5, however, they differ on policy spending preferences as illustrated in Figure 4.

Figure 5.

Regional spending preferences. Note: This figure shows the predicted probabilities for our third hypothesis. It relies on a multiple logistic regression model. A higher probability indicates higher personal perceived importance of European Union (EU) funding for regions with certain characteristics. The model includes controls for age, education and gender as well as region and survey-wave fixed effects. It also includes post-stratification weights on the country level provided by the Eurobarometer dataset.

Preferred levels of decision-making for EU funded projects

Finally, Figure 6 presents the last set of results, showing the predicted probabilities of the preferred decision-making level for EU projects, and tests our third hypothesis. To reiterate, we expect citizens with high skill occupational roles to be more likely to support decision-making on place-based policies at the EU level, whereas those in occupational roles requiring fewer skills should support more local levels. The dependent variable is a forced-choice item, so respondents had to choose one of four levels (Local, Regional, National, and EU). Substantially, if respondents selected the category ‘local’, this would indicate that they prefer the local level to decide upon the EU funding allocation.

Figure 6.

Funding decision level. Note: This figure shows the predicted probabilities for our third hypothesis. It relies on a multiple multinomial regression model with four outcomes (local, regional, national, European Union (EU)). A higher probability indicates higher preference for a certain level to decide on EU funding activity. The model includes controls for age, education and gender as well as region and survey-wave fixed effects. It also includes post-stratification weights on the country level provided by the Eurobarometer dataset.

Decision-making on place-based policy projects at the EU level received the lowest preference across all occupational groups, indicating a general tendency to favour other levels of governance for decisions on EU projects. This is in contrast to our initial expectation that individuals in high-skilled occupations would favour more supranational governance. For nearly all occupations, the gap between the EU and other decision-making levels is notable, with differences reaching up to 10–15 percentage points compared to the EU level. When looking at the most favoured level of decision-making, managers and professionals lean toward the regional level, whereas local decision-making for the allocation of EU regional policy funds is preferred by farmers, shop owners, and manufacturing workers.

Our results suggest that citizens tend to support local decision-making over EU funding allocation across all occupations. This partly contradicts our initial hypothesis, as the EU emerged as the least preferred decision level across all occupations. However, it is also in line with existing empirical evidence showing general support for regional involvement in EU policies (Chacha, 2013). We complement this analysis with an additional robustness check that aims to control for regional corruption perception. As Muñoz et al. (2011) show, citizens’ trust in the EU is substantially influenced by their perception of corruption at the national or local level. It is thus likely to influence preferences in the decision-making on EU-funded projects. To tackle this issue, we conduct an analysis using the Quality of Government (QoG) EU Regional dataset (Charron et al., 2020) to obtain a regional measurement of corruption perception. Due to data availability, we examine a reduced sample including 2015 and 2017. Note that we need to use QoG data from 2013 for 2015. The detailed results are included in the Online appendix. To summarize, if regional corruption perceptions are controlled for, the results and interpretations remain the same as in the main models.

Discussion

This article examines the effect of occupation on public awareness of EU regional policies, preferences regarding how these funds should be spent, differentiated by policy programmes and regional characteristics, and preferences for the decision-making level for EU projects. This article makes three main contributions to the literature, summarized as follows: First, there is substantial variation in awareness of EU regional policies both between EU regions and within regions across occupational groups. Second, spending preferences vary significantly across occupations and tend to align with citizens’ self-interests. Third, citizens across all occupations prefer decisions on projects to be made at the most local level. We provide a theoretical explanation for these differences by relying on occupational differences in skill and responsibility (Oesch, 2006). Our results, however, contain important nuances that need to be discussed in more detail.

First, there is substantial variation in awareness both between EU regions and within regions across occupational groups. While awareness is particularly high in Eastern Europe, only a minority of respondents in countries such as Germany, the UK, Denmark, Austria, and Finland are aware of these policies. Overall, awareness of EU funding ranges from 7 percent to 89 percent among respondents, illustrating the substantial differences within the EU. Further, we find significant and substantial differences in EU funding awareness on two occupational dimensions supporting our theoretical argument based on Oesch’s occupational class scheme from Figure 1. Professionals, managers, farmers and shop owners are occupations with the highest awareness of EU funding activity. While we expected skilled office workers to have higher awareness than manual and routine service workers, we were surprised to see that they even obtained a similar awareness as shop owners. Nevertheless, our results are mostly in line with our expectations depicted in Figure 1. Occupations with high levels of occupational skills and an independent work logic are most likely to be aware of EU funding activity. In comparison, citizens with interpersonal work logic and low occupational skills (e.g. routine service workers) have comparatively low awareness of funding activity. We also find these differences across all EU regions when taking regional average awareness about EU funds into account (see Online appendix). These findings carry important implications for research on EU regional policy. Many studies rely on single-country case studies (e.g., Bachtrögler and Oberhofer, 2018; Debus and Schweizer, 2024; Di Cataldo, 2017; Di Cataldo and Renzullo, 2025), but our evidence of substantial spatial and occupational variation suggests that awareness levels likely affect the strength and direction of observed policy effects. Analyses that overlook this heterogeneity risk missing key mechanisms in how citizens respond to EU place-based policies.

Second, we investigate how self-interest influences spending preferences with regard to policy programmes and regional characteristics. Citizens prefer spending on policy programmes that are likely to be beneficial to them. Farmers and shop owners are most supportive of business support, while manufacturing and routine service workers, as well as unemployed citizens, are most in favour of spending on training. It is also low-skilled occupations that have a stronger preference for social projects than shop owners or professionals. However, the differences are not as large as we would have expected, as we find that spending on social projects appears to be highly popular overall, indicating altruistic motives. This is in line with the most recent results from Italy and Spain showing that EU social spending seems to be the most popular supranational redistribution scheme among citizens (Ciornei, 2025). Also, Burgoon et al. (2022) find generally high support for EU unemployment schemes. In contrast, and surprisingly, spending on infrastructure projects was comparatively less preferred across all occupational groups. Nevertheless, our results suggest that citizens at least partly have their personal material interest in mind when indicating which policy spending they prefer. Relating this to the literature on redistribution preferences (Bechtel et al., 2014; Kuhn et al., 2018; Kuhn and Stoeckel, 2014; Van der Waal et al., 2010), this result shows that examining the average support for EU redistribution might not be sufficient, as underlying policy domain preferences hold significant variation.

With respect to regional characteristics, we also find some support for a self-interest mechanism, as farmers are by far most likely to prefer rural policies. Nevertheless, we also find high support among all occupational groups for regions with high unemployment. To our surprise, all occupations indicate a similar level of support for spending in regions like the unemployed, while we would have expected a significantly lower level. In addition, it needs to be stressed that while the differences between occupations within spending categories are statistically significant and reach up to 10 percentage points, differences between spending categories tend to be higher than the differences between occupational categories.

Third, we examine the level at which citizens prefer decisions about EU regional policies to be made. Across occupational groups, we find a strong preference for decision-making at the local or regional level. All occupations indicate the EU as the least preferred option to decide upon projects. This partly stands in contrast with our theoretical expectations, as we have argued that individuals in high-skilled occupations should be more likely to support decision-making at the EU level. Empirically, we are not able to test the motivations behind these preferences. However, the finding might be explained by a general higher trust in local institutions (Muñoz, 2017). We suggest that individuals perceive local actors to be more knowledgeable about regional needs and therefore prefer local compared to EU governance. Furthermore, based on Fitzgerald and Wolak (2016), the expectation of opportunities for citizens to participate in the decision-making process can also be a reason to prefer local-level decision-making.

In summary, this article engages with EU funding awareness, spending preferences, and preferences of EU funding decision-making levels, highlighting several important dimensions that so far are understudied in the literature on EU place-based policies. We show that both EU funding awareness and spending preferences differ systematically across occupations, in line with our theoretical expectations derived from Oesch’s occupational class scheme. Our findings suggest that a vertical dimension based on individuals’ marketable skills and a horizontal dimension reflecting workplace responsibility and daily work realities shape citizens’ orientations toward EU regional policy. Contrary to our expectations, however, citizens tend to favour local institutions when it comes to decisions on EU-funded projects. While these results offer novel insights into how occupational structures shape attitudes toward EU Cohesion Policy, the absence of detailed data on citizens’ daily job activities limits our ability to explore the underlying mechanisms more directly. Future research should address this limitation by incorporating richer information on work realities, skill levels, and responsibility at work, and by complementing observational evidence with experimental approaches.

Supplemental Material

sj-pdf-1-eup-10.1177_14651165261423537 - Supplemental material for How occupation shapes awareness and preferences for European funding

Supplemental material, sj-pdf-1-eup-10.1177_14651165261423537 for How occupation shapes awareness and preferences for European funding by Johannes Lattmann and David Schweizer in European Union Politics

Supplemental Material

sj-zip-2-eup-10.1177_14651165261423537 - Supplemental material for How occupation shapes awareness and preferences for European funding

Supplemental material, sj-zip-2-eup-10.1177_14651165261423537 for How occupation shapes awareness and preferences for European funding by Johannes Lattmann and David Schweizer in European Union Politics

Supplemental Material

sj-zip-3-eup-10.1177_14651165261423537 - Supplemental material for How occupation shapes awareness and preferences for European funding

Supplemental material, sj-zip-3-eup-10.1177_14651165261423537 for How occupation shapes awareness and preferences for European funding by Johannes Lattmann and David Schweizer in European Union Politics