The Mosque Nearby: Visible Minorities and Far-Right Support in France

Abstract

How is support for right-wing populist parties affected by exposure to Muslim visibility? Using an original database on French mosques, this article analyzes the relationship between the presence of mosques and support for the Front National at the polling station level in the late 2000s. It finds that the propensity to vote for the Front National increases in polling stations up to intermediate distances from mosques and then decreases, suggesting a spatial mechanism known as the halo effect. The analysis also shows that larger mosques and those with minarets are associated with an accentuated halo effect, suggesting the importance of the salience of minority groups rather than their relative size in influencing political behavior.

Keywords

radical right mosque immigration France halo effect

Introduction

The rise of populist radical right (PRR) parties has dramatically altered Europe’s political landscape since the turn of the century, making the far-right family of parties its fastest-growing force (Golder, 2016).¹ The Front National in France epitomizes the different stages of this electoral consolidation: from political marginalization in the early 1970s, the Front National had its major breakthrough when it reached the second round of a presidential election for the first time in 2002. It has now consolidated its role as a major force in the French electoral arena, winning 41% of the vote in the second round of the 2022 presidential election and 31% in the 2024 European elections.²

A number of contextual factors have been proposed to explain this increase, such as deprivation and unemployment (Arzheimer, 2009; Dustmann et al., 2013; Dustmann et al., 2016), exposure to crime (Dinas & van Spanje, 2011; Jardin et al., 2021), or concerns about the provision of public services and welfare (Cavaillé & Ferwerda, 2023; Kavanagh et al., 2021). Among these factors, the presence of immigrants looms large, with PRR parties framing immigration as a threat to Western culture. Anti-immigrant attitudes crystallized further with the refugee crisis that hit Europe in 2014, placing cultural concerns and Muslim immigration at the center of PRR political platforms (Golder, 2016).

The complex relationship between immigration and support for far-right parties has been rationalized through three competing conceptual frameworks: competition theory, group threat theory, and intergroup contact theory (Alesina & Tabellini, 2024). Competition theory postulates that anti-immigrant sentiment is based on material conflict between native and immigrant groups over scarce resources such as jobs, housing, or welfare benefits (Olzak, 1992). Accordingly, higher immigration fosters support for anti-immigrant parties, especially among lower-class natives, because the unemployment effects of immigration may be detrimental to their well-being.³ Alternatively, group threat theory suggests that immigrants pose a threat to national identity and culture. In this perspective, motives are ideational, non-economic determinants predominate, and particular cultural characteristics of the minority outgroup play a critical role in majority reactions. Group threat theory has found support in both cross-national and sub-national contexts.⁴ Nevertheless, more localized analyses indicate that (quality) contact between minority and majority members can mitigate prejudice against minority group members, thereby reducing support for anti-minority policies.⁵ In France, while large immigrant populations are associated with greater electoral success for the Front National at the département level (Edo et al., 2019), this association is reversed at the more disaggregated municipal level (Della Posta, 2013; Vasilopoulos et al., 2022; Vertier et al., 2023). Accordingly, intergroup contact theory posits that long-term exposure to out-groups shapes more positive views and altruistic behaviors toward these groups (Allport, 1954; Pettigrew & Tropp, 2006; Quillian, 1995).

The image of a halo has been proposed to rationalize these mixed findings.⁶ According to halo theory, “individuals living adjacent to ethnically diverse areas experience sporadic contact with immigrants through daily commuting and retail activities, but lack quality contact and therefore will be more likely to perceive those groups as a threat, resulting in higher support for the PRR” (Evans & Ivaldi, 2021, p. 825). Halo theory has been used to better understand the electoral success of the Front National since the mid-1980s, at both the local and national levels (Bon & Cheylan, 1988; Della Posta, 2013; Etchebarne, 1996a; Perrineau, 1985; Rey & Roy, 1986; Schwengler, 2003). More recently, Evans and Ivaldi (2021) have examined the spatial mechanisms at work in the halo effect, finding that individuals in locations with dense immigrant communities are less predisposed to vote for a Front National candidate than those in locations within traveling distance of such dense immigrant areas.

At the heart of halo theory is the nature of interactions between minority and majority group members. Under what conditions do social interactions lead to backlash reactions on the part of majority members? Conversely, what socioeconomic contexts facilitate intergroup appeasement and prejudice reduction? This article takes a fresh look at these questions by empirically examining the relationship between exposure to Muslim visibility and voting behavior at the polling station level in France during the 2007 presidential, 2009 European, and 2010 regional elections. Building on an original dataset of French mosques and combining election results with infra-municipal socio-economic data, we make three contributions.

First, our analysis focuses on the Muslim affiliation of immigrant populations rather than their ethnicity or nationality of origin. Indeed, Islam increasingly crystallizes fears and anxieties in European societies, with individuals perceived as Muslims facing particularly strong hostility and discrimination compared to other minority groups in France (Adida et al., 2017; Bleich, 2009). Moreover, Islamophobia has gradually become the primary populist paradigm of European far-right movements (Brubaker, 2017), with 9/11 serving as a turning point in this renewed agenda (Kallis, 2018). Responding to the need for a better understanding of the religious dimension of the anti-migrant backlash (Aranguren & Madrisotti, 2019; Choi et al., 2019), we examine the relationship between electoral support for the Front National and the presence of mosques, a visible marker of Islam at the neighborhood level.

Second, we explore the electoral implications of the differential visibility of the minority group rather than its relative size. In doing so, our work complements a recent trend in immigration research that pays attention to the salience of immigrant populations rather than their size (Newman & Velez, 2014), focusing on visual cues of Muslim presence (Bornioli et al., 2023) rather than other forms of salience such as media attention (Couttenier et al., 2024; Hopkins, 2010). Close to our approach, Colussi et al. (2021) analyze the electoral effects of the increased salience of Muslim communities during Ramadan in Germany. Closer still, Gravelle et al. (2021) measure how spatial proximity to mosques (especially those with minarets) shapes individuals’ support for the radical right in the Netherlands.⁷ Our research follows this approach by focusing on visible, distinctive institutions – mosques or prayer houses – that gather worshipers for daily Islamic prayers and during Friday congregations, and that constitute key features of the built environment. Rather than the mere presence of immigrants, we show that the salience of cultural difference has significant implications for ethnocentric attitudes. Our work is thus close to Overos and Sauer (2023), which analyzes the linear relationship between Islamic and Catholic religious buildings – based on volunteered geographic information – and commune-level voting patterns in the 2017 French presidential election.

Third, we shift the analysis to the neighborhood level. Since contextual explanations of far-right support focus on social interactions, it is important to focus on the micro level where (lack of) intergroup contacts operate. In doing so, we consider polling stations, which are smaller levels of aggregation than both municipalities and IRISs, the infra-municipal statistical unit in France—in comparison, Colussi et al. (2021) focus on German municipalities, Overos and Sauer (2023), on French municipalities, while Gravelle et al. (2021) use Dutch four-digit postal codes to localize their respondents.⁸ Moreover, our granular approach enables us to address the literature on the implications of neighborhood-level ethnic composition on individual voting for far-right parties (de Blok & van der Meer, 2018; Fremerey et al., 2024; Savelkoul et al., 2017).

Using variation in distance to the (same) nearest mosque across polling stations and controlling for the presence of local immigrants, we find that the propensity to vote for the Front National increases in polling stations up to intermediate distances from mosques (16 km) and then decreases, enacting the distinctive curvilinear relationship implied by the halo hypothesis. Moreover, we find that mosque visibility matters: buildings with a minaret and a larger surface area are associated with accentuated far-right support in intermediate polling stations up to 10–14 km away from mosques. Thus, our findings help reconcile both intergroup contact theory – significant, high-quality interactions between Muslims and non-Muslims within neighborhoods where mosques are located lead to lower shares of the far-right vote – and competition and group threat theory—rare or fleeting contacts between Muslims and non-Muslims in neighborhoods at intermediate distances from the mosque lead to higher shares of the far-right vote.

The remainder of this article is organized as follows. We first document the rise of the far-right party in France and the concomitant emergence of Islamic places of worship. We then describe the data and present the empirical strategy and results. Finally, we provide an interpretation and discussion of our findings.

Background

In this section, we review the rise of the Front National in France and examine the concomitant settlement of Muslim populations on the French territory, with Muslim identity increasingly presented by far-right politicians as a distinct and incompatible cultural trait.

The Rise of the Front National

Beginning in the early 1980s, a new wave of far-right activism swept across Europe, with far-right parties participating in coalition governments in Austria, Croatia, Estonia, Finland, Italy, Latvia, the Netherlands, Poland, Serbia, Slovakia, and Switzerland. Other far-right movements played a key role in national politics in France, Belgium, and Hungary (Golder, 2016).

In France, the main PRR party – the Front National – was formed in 1972 from a disparate coalition of ultra-nationalist groups. It gradually emerged as a viable political force in the 1980s, making its first breakthrough in the 1984 European elections. Its electoral platform quickly stabilized around a few guiding principles: welfare-chauvinist policies, opposition to immigration, rejection of European integration, and virulent anti-Semitism (Camus, 1996). Throughout the 1980s and 1990s, the politicization of postcolonial immigration, crime, and insecurity became the main drivers of electoral support for the Front National. Yet no single explanation can account for its steady rise: while core supporters of the Front National generally vote on ideological grounds, a majority uses this vote as a means to protest against traditional political elites (Etchebarne, 1996b). In fact, Front National voters are far from socially homogeneous, although they share some characteristics (Mayer, 2015): a strong attachment to conservative moral values on the family, same-sex relations, and religion; low educational attainment; low-skilled occupational status, especially blue-collar workers and small shopkeepers; and a male majority. The Front National’s electoral breakthrough began in large urban areas and suburbs, especially in industrial regions. Since the early 2000s, however, a shift has occurred, with the party attracting more and more voters from rural areas (Gombin, 2015b; Huc, 2019).

Of particular interest to our analysis, the Front National has increasingly mobilized a distinctly anti-Muslim rhetoric, defining the cultural enemy in religious rather than racial terms. This exclusivist vision of national citizenship consists in presenting Islam as fundamentally incompatible with liberal-democratic values, with the notion of “Islamization” becoming a regular feature of the party’s xenophobic discourse since the 1990s (Alduy & Wahnich, 2015; Mudde, 2013). This development fits into a broader Western European populist conjuncture characterized by “civilizationalism” and the notion of a civilizational threat from Islam (Brubaker, 2017).

Mosques, Immigration, and Islam in France

The historical process of mosque construction in France reflects the progressive settlement of Muslim populations in the country. A handful of mosques were built in the first half of the 20th century to accommodate colonial subjects, most notably the Grande mosquée de Paris in 1926 (Boyer, 1992; Davidson, 2009; Sbaï, 2006). With the arrival of immigrants from North Africa after World War II, mosques and Islamic prayer rooms spread throughout the French territory. In the 1970s, the construction of mosques was supported by the government in order to dampen return migration by nurturing workers’ identification with Islam (Davidson, 2009). At the time, the vast majority of mosques were located in factories and migrant workers hostels (foyers), out of sight of the majority population. A shift occurred in the late 1970s and 1980s, reflecting the permanent settlement of Muslim communities through family reunification regulations in the late 1970s (Cesari, 1994) and the 1981 law authorizing the creation of associations by foreigners. From then on, mosques flourished in neighborhoods with a high concentration of immigrant populations (Jouanneau, 2013), especially near subsidized public housing (HLM, or habitations à loyer modéré). From 131 mosques and prayer rooms in 1976 – most of which were located out of sight in foyers– the number of mosques rose to 941 in 1986 (Legrain, 1986), 1,590 in 1997, and 2,130 in 2012. Figure 1 displays the evolution of the number of mosques from the 1920s to 2012, making apparent the upward trend in the proliferation of mosques since the 1970s.

In addition to their proliferation, mosques and Islamic prayer rooms gradually took on new social functions: not only ritual purposes (daily Islamic prayers and gatherings for religious holidays), but also the provision of Islamic education for children and youth on weekends, sports activities, family mediation, and vocational training (Jouanneau, 2013). These dynamics undoubtedly increased the visibility of mosques at the neighborhood level and attracted a diverse range of worshipers and beneficiaries to their premises on a regular basis. Finally, the late 1990s and 2000s witnessed the construction of purpose-built mosques – some with domes and minarets – that are more conspicuous than their predecessors, both because of their location and their architecture. As of 2014, France counted 2,502 mosques and Islamic prayer rooms on its territory, gathering some 426,000 (mostly male) Muslim worshipers each Friday—attending Friday prayers in congregation is not a religious obligation for Muslim women.

As elsewhere in Europe (Allievi, 2010), the construction and presence of mosques in a given neighborhood tend to trigger anti-Muslim protests that far-right parties both orchestrate and capitalize on in their electoral campaigns. Cases of vehement local opposition have been documented by several qualitative studies, with mosques being portrayed by far-right activists as concrete threats to security and national identity (Allen, 2013; Faury, 2024; van Es, 2020). Pushing this agenda in the media (Amengay, 2020), the European PRR frames mosques and minarets as symbols of “islamization,” making visible the supposed “Muslim enemy within” (Hafez, 2014). In France, political opposition to mosques by Front National supporters has been documented in electoral polls.⁹

Data

The analysis in this article takes advantage of two rich datasets: a relatively untapped dataset that combines election results at the polling station level along with corresponding socio-economic and geographic information, and an original dataset we constructed on mosque locations and characteristics. These data enable us to conduct a fine-grained analysis of voter radicalization according to the local salience of Muslim communities by calculating the exact distance between each polling station and its nearby mosque.

Election Data

Our analysis focuses on the Front National’s electoral performance in three relatively recent elections: the first round of the 2007 presidential election, the 2009 European elections, and the first round of the 2010 regional elections. In contrast to Evans and Ivaldi (2021), Vasilopoulos et al. (2022), and Overos and Sauer (2023), which consider a single election – the 2017 presidential election – we aim to examine a range of different elections because the Front National’s performance has historically varied across different types of elections. Indeed, the Front National first gained political legitimacy through municipal elections, as well as second-order elections such as European and regional elections (Ignazi, 1996). Presidential elections, on the other hand, are higher stakes and more difficult for fringe parties to enter. The 2002 presidential election was a milestone in this regard, marking the electoral zenith of the Front National’s historic leader, Jean-Marie Le Pen, and the first time a far-right candidate made it to the presidential runoff.¹⁰

To study the results of these elections, we rely on a unique and relatively untapped data source: the CARTELEC database, which compiles electoral data for the three aforementioned elections at the level of polling stations in 2007 geography (Beauguitte & Colange, 2013; Jadot et al., 2010).¹¹ In 2007, there were about 36 thousand municipalities in metropolitan France, of which 6 thousand had more than one polling station, resulting in about 65 thousand polling stations. The number of polling stations in a municipality together with their constituencies are determined by the département-level authorities with the aim of drawing polling stations with 800 to 1,000 voters. All voters residing within the boundaries of a polling station are required to vote at that station, so there is a direct correspondence between a polling station and the residents of its constituency.¹²

Although the set of addresses that are part of a polling station is public, the Ministry of the Interior does not have a file that centralizes this information (Jadot et al., 2010, pp. 86–7).¹³ As a result, the CARTELEC project had to aggregate this scattered information by going through each polling station in each département. Given the difficulty of the task, the compatibility problems with the data formats received, and the reluctance of the mayors of some relatively large municipalities to share this (albeit public) information, CARTELEC was only able to construct the geometries of 50,576 polling stations, including 742 municipalities divided into multiple polling stations.¹⁴ CARTELEC then matched these polling stations with the results of the 2007, 2009, and 2010 elections.¹⁵ Figure 2 displays the spatial distribution of the Front National’s vote shares in the first round of the 2007 presidential election at the level of polling stations, based on the CARTELEC database.¹⁶

Socio-Economic Data

To capture the socio-economic environment at the level of polling stations, we also rely on the CARTELEC database, which complemented its election data with contextual information from the 2007 and 2008 censuses at the level of IRISs, equivalent to census tracts. To match polling station polygons to IRIS-level data, CARTELEC intersected both geometries and ventilated the data across polling stations, assuming a constant distribution of population across polygon areas (Beauguitte & Colange, 2013, pp. 10–2). As census information is not published for municipalities with less than 100 inhabitants due to confidentiality considerations, the CARTELEC database contains socio-economic information for 46,723 of the 50,576 polling stations for which a geometry is available. This information includes the distribution of the population by age, educational attainment, and housing type.¹⁷ Essential for our purposes, it also provides the number of immigrants by polling station, that is, the foreign population without the French nationality at birth.¹⁸

Geographic Data

We further complement the CARTELEC data with geographic information based on the location of polling stations within their territorial administrative framework. Since our hypothesis concerns how ordinary interactions between majority (non-Muslim) and minority (Muslim) group members shape political behavior, we are interested in capturing the areas where most daily social interactions are likely to occur. To this end, we use four statistical zonings defined by the National Institute of Statistics and Economic Studies (INSEE): life basins (bassins de vie), urban units (unités urbaines), urban areas (aires urbaines), and sensitive urban zones (zones urbaines sensibles).

First, we match polling stations to their respective statistical zoning into life basins (bassins de vie). These zones represent the smallest areas within which residents have access to most public services and facilities, and capture the perimeter around which residents organize their daily lives (Brutel & Levy, 2012). In total, the French territory is divided into 1,641 life basins.¹⁹ Because we consider these zones to be the relevant spaces where individuals experience most daily social interactions, we conduct the baseline analysis based on voters’ proximity to mosques within these life basins—although we relax this constraint to test the robustness of our results.

To zoom in on spaces of more intense social interaction and to account for the largely urban location of mosques, we also match polling stations to their respective zoning into urban units (unités urbaines). Urban units are composed of spatially contiguous residential units with at least two thousand inhabitants. Although an urban unit includes at least one municipality, it often includes urban extensions covering several neighboring municipalities. A total of 2,233 urban units are defined over the territories of 7,224 municipalities, which host 77% of the population of mainland France.²⁰ Urban units are particularly interesting in our context because they are spaces where individuals experience intense social interactions due to residential proximity.

To further observe local social interactions, we also match polling stations to their respective zoning into urban areas (aires urbaines). Larger than urban units, urban areas capture intensive exchanges between places of residence and work. They concentrate at least fifteen hundred jobs and usually contain an urban ring. A total of 771 urban areas encompass 18,180 municipalities and 85% of the population of mainland France.²¹

Finally, we also collect information on whether a polling station contains a sensitive urban zone (zone urbaine sensible). These areas are defined by public authorities as high priority targets for urban policy. They are characterized by a high percentage of public housing, low home ownership, high unemployment, and a low percentage of high school graduates—all sorts of conditions that disproportionately affect immigrant populations. These 717 sensitive urban zones cover only a small fraction of the territory, concerning 4.4 million inhabitants. They are also generally much smaller than polling stations.²²

To assess heterogeneity across areas with more intense social interactions among residents, we run the analysis sequentially on the subset of polling stations that are outside urban areas (which we define as rural), within urban areas but outside urban units (which we define as peri-urban), and within urban units (which we define as urban). We display the distribution of life basins, rural, peri-urban, and urban areas in Appendix Figure A.3.

Mosques Data

We create an original dataset of mosque locations based on two confidential files produced by the French Ministry of the Interior, which provide a census of all mosques present in the metropolitan territory in 1997 and 2012. These censuses offer significant advantages over other sources of data on mosques, whether from web-scraping or produced by Muslim or far-right anti-Muslim websites, which have been used in the few quantitative studies on the spatial distribution of Muslim presence in Europe (Colussi et al., 2021; Drouhot, 2020; Gravelle, 2021; Gravelle et al., 2021; Ivaldi & Dutozia, 2018; Overos & Sauer, 2023) but which suffer from significant quality shortcomings (Basiri et al., 2019; Shelton et al., 2012; Sui et al., 2013). First, these are administrative files: they have been compiled by local intelligence officers – civil servants – who collect information on the ground as external observers.²³ Second, they provide substantial information about each mosque. The 1997 file includes the following information for the 1,589 mosques that existed at that time: their names, addresses, years of establishment, and attendance (number of worshipers). The 2012 file includes the following information for the 2,130 mosques that existed at that time: their names, addresses, sizes (in square meters), attendance (number of worshipers), and whether they had a minaret. Importantly, the availability of addresses in these files enables us to geocode the exact location of mosques using the API available through adresse.data.gouv.fr, which is based on the central database of addresses in metropolitan France, the Base Adresse Nationale.²⁴

While the locations of mosques that existed at the time of the 2007, 2009, or 2010 elections remain unknown, we build on these administrative files in two ways to construct a realistic approximation of the mosques that existed then. A first approach is to match the 1,589 mosques in the 1997 file with the 2,130 mosques in the 2012 file, and retain the subset of 1,053 mosques that are present in both files.²⁵ These matched mosques correspond to those that existed in 2012 and that were already established in 1997. This approach is conservative because it underestimates the number of mosques in 2007, 2009, or 2010 by excluding those established after 1997. A second approach is to keep all 2,130 mosques in the 2012 file, assuming that no mosque was established between 2007 and 2012. Given the upward trend in mosques during this period, this approach provides an upper bound on the number of mosques that existed in 2007, 2009, or 2010. Figure 3 displays the distribution of mosques resulting from both strategies. As expected given the urban settlement of Muslim immigrants and their descendants, mosques are overrepresented in urban areas—we describe below the socio-economic characteristics of the polling stations where mosques are located.

Table 1 reports summary statistics on the characteristics of mosques. Mosques in the 1997 file gathered an average of 92 worshipers and up to three thousand. The first was founded in 1926 (the Grande mosquée de Paris) and 36% hosted a Quranic school with an average of 58 students. Attendance increased over time, as the mosques in the 2012 file gathered an average of 187 worshipers. They also showed a high variability in their surface area, with a standard deviation of 673 square meters for a mean of 373. Finally, 83 mosques had a visible minaret. Panel B shows that these characteristics are broadly similar across the subset of mosques that are present in both the 1997 and 2012 files, suggesting limited selection bias in the matching process. The baseline analysis focuses on the sample of matched mosques because it provides a more conservative approximation of the distribution of mosques during the relevant elections. The information available for this set of matched mosques is also richer, allowing for a more fine-grained analysis. However, we test the robustness of our results using the set of mosques from the 2012 file.

Table 1.

Summary Statistics of Mosques Characteristics.

	A. Raw Files					B. Matched Files
	Mean	S.d.	Min.	Max.	N	Mean	S.d.	Min.	Max.	N
1997 mosques file
Attendance (worshippers)	92	156	1	3,000	1,497	103	183	10	3,000	994
Year established	1986	7	1926	1997	1,517	1986	7	1926	1997	1,014
Has a quranic school	0.36	0.48	0.00	1.00	1,589	0.39	0.49	0.00	1.00	1,053
Students \| quranic school	58	69	5	800	542	58	63	5	400	396
2012 mosques file
Attendance (worshippers)	187	297	3	4,600	2,065	202	297	3	3,000	1,032
Surface area (m²)	373	673	10	11,100	2,034	366	684	12	11,100	1,023
Has a minaret	0.04	0.19	0.00	1.00	2,130	0.05	0.21	0.00	1.00	1,053

Notes. This table summarizes the characteristics of French mosques in terms of visibility and congregation. Panel A reports summary statistics for the 1,589 mosques listed in the 1997 file and for the 2,130 mosques listed in the 2012 file of the Ministry of the Interior. Panel B reports summary statistics for the 1,053 mosques that are present in both files. The number of observations (N) does not always add up to these totals due to missing values in the original files. S.d. denotes standard deviation.

The availability of mosque characteristics also enables us to construct measures that capture the visibility of each mosque, with the hypothesis that the more visible a mosque is in its neighborhood, the stronger the effect its presence has on voting behavior (Gravelle et al., 2021). Relevant characteristics include a mosque’s attendance, size, and whether it has a minaret.

Distance to the Nearest Mosque

Key to our approach, we compute the exact distance from each polling station to its nearest mosque. Unfortunately, while the CARTELEC database contains the geometries of polling stations, it does not contain the addresses of the corresponding polling booths. To improve the precision of our distance measures at the local level (and to avoid systematically assigning the locations of polling booths to the centroids of their polling stations), we match each polling station to the corresponding location of its associated polling booth—the procedure is detailed in the Online Appendix. Because we are interested in assessing the role of social interactions in the relationship between Muslim visibility and voting patterns, our baseline analysis restricts the set of candidate mosques to those located in the same life basin as the polling station. For each polling station, we then calculate the distance in meters to the nearest mosque.

Summary Statistics

Combining these data and omitting polling stations for which we have no electoral or socio-economic information, our working dataset contains 46,297 polling stations. Given that 337 out of 1,637 life basins contain a mosque from the matched 1997–2012 file, our baseline regression sample contains 24,698 polling stations.²⁶ Panel A of Table 2 provides summary statistics for characteristics of polling stations in our baseline regression sample in terms of socio-economic information, urban localization, and electoral data. Overall, 4% of polling stations (988 polling stations) contain a mosque. The average polling booth is located 5 km from a mosque, but the distance to the nearest mosque varies widely, as displayed in Figure 4. The Front National’s average vote share ranges from 8 to 13% in different elections, with low abstention rates in the 2007 presidential election (15%) and high abstention rates in the 2009 European (58%) and 2010 regional (54%) elections. The share of the foreign population averages 6% and ranges from 0 to 55%. Hence, we can expect that a non-marginal part of the population has no post-colonial background and therefore may potentially feel threatened by the presence of a mosque. In comparison, the unrestricted sample, which also includes polling stations located in a life basin that does not contain a mosque, is more rural: these polling stations are less populated and less likely to be located in an urban area (Panel B of Table 2). We will test the robustness of our results to this sample selection procedure.

Table 2.

Summary Statistics of Polling Station Characteristics.

	A. Polling Stations in Life Basins Where Mosque Present				B. All Polling Stations
	Mean	S.d.	Min.	Max.	Mean	S.d.	Min.	Max.
Population	1,671	2,425	100	89,259	1,310	2,071	100	89,259
Area (km²)	6,844	11,078	9	206,204	11,116	14,120	9	372,898
Share foreign population (percent)	5.69	6.03	0.00	54.53	4.18	5.13	0.00	54.53
Unemployment rate (percent)	8.69	4.51	0.00	52.67	7.92	4.03	0.00	52.67
Share with no diploma (percent)	18.16	8.43	0.00	70.61	18.75	7.74	0.00	70.61
Share households in HLM (percent)	14.49	19.76	0.00	99.09	9.45	16.15	0.00	99.09

In urban area	0.87	0.33	0	1	0.66	0.47	0	1
In urban unit	0.68	0.47	0	1	0.45	0.50	0	1
Within municipality	0.54	0.50	0	1	0.31	0.46	0	1
Contains sensitive urban zone	0.08	0.27	0	1	0.04	0.20	0	1

Contains a mosque	0.04	0.19	0	1	0.02	0.14	0	1
Number of mosques	0.04	0.23	0	6	0.02	0.17	0	6
Distance to closest mosque (km)	5.34	5.54	0.00	52.48	14.22	13.29	0.00	78.65

Abstention rate (percent)
Presidential 2007	14.80	4.69	0.00	51.86	13.92	4.42	0.00	51.86
European 2009	58.06	8.45	0.24	89.53	56.40	8.21	0.24	89.53
Regional 2010	53.56	8.27	0.00	88.20	51.44	8.22	0.00	88.20

Front national vote share (percent)
Presidential 2007	11.10	5.12	0.59	44.83	11.63	5.12	0.00	44.83
European 2009	7.67	4.69	0.00	45.83	7.95	5.06	0.00	47.50
Regional 2010	12.97	7.17	0.00	59.02	12.74	7.10	0.00	59.02

Number of polling stations	24,698				46,297

Notes. This table reports summary statistics for polling stations located in a life basin with at least one mosque in Panel A and for all polling stations in Panel B. S.d. denotes standard deviation. HLM denotes public housing.

How do polling stations with a mosque compare to those without a mosque? Table 3 compares characteristics between these two types of polling stations. This simple comparison reveals dramatic differences: polling stations with a mosque are significantly more urban, and their populations are relatively more likely to be unemployed, less educated, and live in public housing. They are also more likely to be located in an urban area and to contain a sensitive urban zone. However, although they have comparatively higher abstention rates, they exhibit little difference in the electoral performance of the Front National. The empirical strategy discussed below takes these differences into account in order to obtain credible estimates of the hypothesized halo effect.

Table 3.

Summary Statistics of Polling Station Characteristics.

	A. Polling Stations With No Mosque				B. Polling Stations With a Mosque				C. Difference
	Mean	S.d.	Min.	Max.	Mean	S.d.	Min.	Max.	Mean	S.d.	p -Value
Population	1,524	1,803	100	75,765	5,504	7,670	114	89,259	3,980	78	0.00
Area (km²)	6,736	10,631	9	206,204	9,662	19,143	30	182,930	2,927	372	0.00
Share foreign population (percent)	5.43	5.79	0.00	54.53	12.28	8.19	1.13	44.21	6.85	0.20	0.00
Unemployment rate (percent)	8.50	4.34	0.00	44.11	13.48	5.88	3.04	52.67	4.97	0.15	0.00
Share with no diploma (percent)	17.78	8.12	0.00	70.61	27.92	10.40	5.00	69.75	10.14	0.28	0.00
Share households in HLM (percent)	13.78	19.28	0.00	99.09	32.86	22.92	0.00	97.04	19.07	0.65	0.00

In urban area	0.87	0.34	0	1	0.96	0.20	0	1	0	0	0.00
In urban unit	0.67	0.47	0	1	0.98	0.13	0	1	0	0	0.00
Within municipality	0.54	0.50	0	1	0.63	0.48	0	1	0	0	0.00
Contains sensitive urban zone	0.07	0.26	0	1	0.27	0.45	0	1	0	0	0.00

Number of mosques	0.00	0.00	0	0	1.12	0.41	1	6	1.12	0.00	0.00
Distance to closest mosque (km)	5.53	5.56	0.05	52.48	0.53	0.60	0.00	6.54	−5.00	0.18	0.00
Abstention rate (percent)
Presidential 2007	14.65	4.63	0.00	51.86	18.46	4.73	6.47	40.09	3.81	0.16	0.00
European 2009	57.76	8.33	0.24	89.53	65.68	7.96	43.49	89.09	7.92	0.28	0.00
Regional 2010	53.28	8.15	0.00	88.20	60.88	8.01	27.59	86.94	7.60	0.27	0.00

Front national vote share (percent)
Presidential 2007	11.11	5.14	0.59	44.83	11.02	4.48	1.79	30.27	−0.09	0.17	0.60
European 2009	7.65	4.71	0.00	45.83	8.14	4.28	0.00	32.69	0.48	0.16	0.00
Regional 2010	12.95	7.19	0.00	59.02	13.43	6.73	0.00	39.93	0.47	0.24	0.05

Number of polling stations	23,781				917				24,698

Notes. This table reports summary statistics for polling stations located in a life basin with at least one mosque. S.d. denotes standard deviation. HLM denotes public housing.

Empirical Analysis

Empirical Strategy

Our baseline empirical strategy attempts to capture a potential halo effect surrounding mosques on the Front National electoral performance through a quadratic term in distance, as is common in the literature (e.g., Evans & Ivaldi, 2021, p. 833). It follows from the expectation that the vote share of the Front National should initially increase as distance increases, but then decrease as distance increases further. Specifically, we estimate the following OLS specification:

{F N}_{ijl} = α + β_{1} {d i s t}_{i} + β_{2} {d i s t}_{i}^{2} + δ_{1} X_{i}^{'} + δ_{2} X_{j}^{'} + η_{l} + ε_{ijl},

(1)

where i indexes a polling station, j, the polling station in which the nearest mosque to station i is located, and l, the life basin in which polling station i is located. The outcome FN_ijl is the vote share of the Front National in the 2007 presidential election in percent in polling station i. We focus on this one election because it is the most salient and is characterized by a fixed political supply across the territory, which facilitates the interpretation of the results—we analyze the 2009 European and 2010 regional elections below.

The coefficients of interest β₁ and β₂ capture the halo effect, or, more precisely, the quadratic relationship between the distance of polling station i to the nearest mosque (located in polling station j in the same life basin l) in kilometers and the Front National vote share. Namely, our hypothesis is that ${\hat{β}}_{1} > 0$ and ${\hat{β}}_{2} < 0$ . Importantly, because social interactions may be affected by unobservable characteristics specific to life basins, we include life basin fixed effects η _l in order to compare polling stations located in the same life basin. We cluster standard errors at the level of polling stations—we test the robustness of this clustering procedure below.

Of course, any correlation between the distance to the nearest mosque and the vote share of the Front National could be spurious and instead capture the effect of the foreign population share in polling station i as well as the foreign population share in polling station j where the nearest mosque to i is located, a phenomenon that has been repeatedly demonstrated in the literature (e.g., Evans & Ivaldi, 2021). We therefore control for the observable characteristics of both polling stations, which are contained in vectors $X_{i}^{'}$ and $X_{j}^{'}$ , respectively. These vectors include potential confounders of the relationship between the distance to the nearest mosque and the Front National vote share: the share of foreign population, but also the unemployment rate, the share of population with no diploma, the share of population living in public housing, the log population, and the area of the polling station to account for population density.²⁷ This set of controls captures the typical profile of far-right voters (Arzheimer, 2009; Bowyer, 2008).

However, this strategy may still fail to provide credible estimates of the halo effect if unobservable characteristics of the polling station where the nearest mosque is located, or the characteristics of the mosque itself, vary systematically with distance. In particular, the demographic composition of the polling station where a mosque is located could confound the results. To mitigate this potential issue, our preferred empirical strategy includes a set of nearest mosque fixed effects. Specifically, we estimate the following specification:

{F N}_{ikl} = α + β_{1} {d i s t}_{i} + β_{2} {d i s t}_{i}^{2} + δ X_{i}^{'} + ω_{k} + η_{l} + ε_{ikl},

(2)

where k indexes the nearest mosque to polling station i and ω _k are nearest mosque fixed effects. In this specification, the coefficients of interest β₁ and β₂ are identified from variations in the Front National vote shares across polling stations in the same life basin that share the same nearest mosque but are located at different distances from that mosque.

Identifying the Halo Effect

Main Results

We report OLS coefficients from estimating Equations (1) and (2) in Table 4. Regressing the Front National vote share on a quadratic term in distance to the nearest mosque without other controls in Column (1) yields in a significant concave polynomial relationship, suggesting the reality of a halo effect: as distance to a mosque increases, the Front National vote share increases, but then declines as distance increases further, with an apex of the halo at 20 km from the nearest mosque

(= - {\hat{β}}_{1} / 2 {\hat{β}}_{2})

. Including further controls for both the polling station and the one where the nearest mosque is located produces qualitatively similar results in Column (2) and (3), albeit less pronounced. As expected, the foreign population shares in the polling station and in the polling station where the nearest mosque is located are negatively correlated with Front National vote share. Importantly, including 1,010 nearest mosque fixed effects in Column (4) produces similar results, with an apex at 16 km—about twice as close as the findings in Evans and Ivaldi (2021, p. 840).

Table 4.

Vote Share for the Front National and Distance to Nearest Mosque.

Dependent Variable	Front National Vote Share (%)
Dependent Variable	(1)	(2)	(3)	(4)
Distance to nearest mosque (km)	0.272***	0.190***	0.176***	0.166***
Distance to nearest mosque (km)	[0.012]	[0.013]	[0.013]	[0.013]
Distance to nearest mosque, squared (km)	−0.007***	−0.007***	−0.006***	−0.005***
Distance to nearest mosque, squared (km)	[0.001]	[0.001]	[0.001]	[0.001]
Share foreign population (%)		−0.258***	−0.234***	−0.230***
Share foreign population (%)		[0.007]	[0.008]	[0.008]

Controls	No	Yes	Yes	Yes

Nearest mosque polling station
Controls	No	No	Yes	No
Share foreign population (%)			−0.075***
Share foreign population (%)			[0.006]

Fixed effects
Life basins	333	333	333	333
Nearest mosque	0	0	0	1,010

Polling stations	24,671	24,671	24,671	24,671

Within R²	0.036	0.215	0.227	0.130
Front national vote share mean (%)	11.10	11.10	11.10	11.10
Halo apex (km)	19.45	14.53	14.75	15.95

Notes. This table reports OLS coefficients from estimating Equation 1 in Columns (1)–(3) and Equation 2 in Column (4). The dependent variable is the vote share of the Front national in percent. The unit of observation is the polling station. The distance measure is with respect to the nearest mosque located in the same life basin as the polling station. Controls include the polling station’s share of foreign population, log population, area, average age, unemployment rate, the share of population with no diploma, and the share of population living in HLM. Standard errors are in brackets and are clustered at the level of polling stations. Estimates are calculated using Correia’s (2023 [2014]) reghdfe Stata package.

^*** Significant at the 1% level.

To get a better sense of the magnitude of this halo effect, we use estimates from Table 4 to predict Front National vote shares at distances up to 30 km from mosque locations – the distance up to which there is a sufficient density of polling stations – and report mean predictions in Figure 5 along with 95% confidence intervals. The blue curve uses estimates from Column (1) and the red curve uses estimates from Column (4). Both curves clearly show the reality of a halo effect, with an even more pronounced halo in our preferred specification represented by the red curve.

Robustness

Clustering and Spatial Correlation

Our results are robust to the choice of statistical procedure used to compute the precision of regression estimates (see Appendix Table A.1). In particular, clustering standard errors at higher levels of aggregation – closest mosque or life basin as opposed to polling station – to allow for broader spatial correlation of errors generates standard errors that are larger than for the baseline but that leave the coefficients of interest significant at the 1% level. Similarly, allowing for spatial correlation up to 30 km and temporal correlation up to 4 years leaves the results unchanged.

Sample Restrictions

To assess the role of sample restrictions in generating our results, we repeat the analysis from Equation (2) when also including polling stations located in a life basin without mosques, increasing the sample of polling stations from 25 to 46 thousand. We report the results in Column (2) of Appendix Table A.2. The halo is still present, but less pronounced, with an apex at 28 km—close to the findings in Evans and Ivaldi (2021, p. 840). We interpret this result as suggestive of the role of local social interactions, where comparing polling stations beyond life basins dilutes the chances of regular contacts between communities. As noted above, life basins define coherent units in terms of daily life, structuring intense exchanges between places of work and residence. Thus, more distant exposure to mosques but lack of quality contact with Muslims (as in the case of polling stations located in a life basin without mosques) is likely to dilute the halo effect.

We further assess the robustness of our findings to the use of distances computed on the set of all mosques in the 2012 file in Columns (3) and (4) of Appendix Table A.2, increasing the number of mosques in our sample from 1,010 to 1,905. Our results hold and are nearly identical to those based on the subset of mosques in the matched 1997–2012 file, suggesting little selection through this sample restriction procedure. To make these differences more apparent, we report mean predictions of Front National vote shares across distances to mosques in Appendix Figure A.4 under these four alternative sample restrictions—we predict vote shares across distances up to 60 km when not restricting to polling stations located in life basins where at least one mosque is present, given the spatial distribution of mosques in these samples (see Appendix Figure A.5).²⁸

Controlling for the Nationality of Foreigners

Because religious and ethnic information is not collected by administrations in France, our baseline analysis can only control for the presence of foreigners as a proxy for the presence of Muslims.²⁹ To address some concerns regarding this imperfect strategy, we conduct a robustness check in which we control for the nationality of foreigners on a subset of the data.³⁰ More specifically, we match TRIRIS-level information on nationalities from the 1999 census – the closest census for which this information is publicly available – to CARTELEC’s shapefile and calculate the share of foreigners by polling station among nationalities that may capture the potential Muslim population: Algerians, Moroccans, Tunisians, and Turks.³¹ We also collect information on the proportion of the population that is naturalized, EU citizens, and other nationalities. We provide summary statistics in Appendix Table A.3.³² We find that among the foreign population, 16% are Algerian; 14%, Moroccan; 5%, Tunisian; 5%, Turkish; 37%, from the EU; and 23%, from other nationalities. In addition, 6% of the total population are naturalized citizens.

Next, we reproduce the analysis on the subsample of the 14 thousand polling stations (out of 25 thousand) for which we have nationality information. We report the results in Column (2) of Appendix Table A.4. Results are very close to the baseline, which are reported in Column (1) for reference. Then, controlling for the share of the foreign population from the TRIRIS data instead of the CARTELEC data in Column (3) again yields similar results, suggesting that our matching strategy is sound. Finally, controlling for the nationalities of foreigners together with the share of the naturalized population in Column (4) highlights a similar halo effect, which is visible in Appendix Figure A.7. Overall, this robustness check supports the validity of our analysis, despite the unavailability of religious and ethnic information at the polling station level and thus the lack of precise identification of the Muslim population.

Relaxing Parametric Assumptions

The identification of a halo may be driven by the parametric assumptions we impose—a quadratic term in distance. To address this concern, we adopt a non-parametric approach and deploy a model with a set of indicator variables by bins of distance:

{F N}_{ikl} = α + \sum_{b = 3}^{30} β_{b} 1_{\{b \geq {d i s t}_{i} > b - 3\}} + δ X_{i}^{'} + ω_{k} + η_{l} + ε_{ikl},

(3)

where b is a 3-km bin, with the indicator variable

1_{\{b \geq {d i s t}_{i} > b - 3\}}

equal to one if the distance to the nearest mosque is in that bin—we use 3-km bins to ensure that we have enough density of polling stations to estimate each of the ten bin-specific coefficients. Other variables are similar to those in Equation (2). To simplify the interpretation, the omitted category is polling stations that contain a mosque—we normalize their distance to zero. We report the results in Appendix Figure A.8. It clearly shows that the halo holds with this non-parametric strategy, with an apex located about 16 km from the nearest mosque, similar to that obtained with the parametric strategy.

Heterogeneity

Heterogeneity Across Elections

Given the variation in the Front National’s performance across types of elections as seen in Figures 2, A.1, and A.2, we assess the heterogeneity of the halo effect by estimating Equation (2) for each election separately. We report the results in Table 5, where we pool all three elections along with election fixed effects in Column (1) for reference. We identify the presence of a halo when we pool elections as well as for the presidential and European elections, but not clearly for the regional elections. To make the halo apparent, we again report mean predictions of the Front National’s vote shares across distances from mosques in Appendix Figure A.9 for all three elections.³³ The overall halo effect appears to be driven by presidential elections, where the halo is much more pronounced than in other elections.

Table 5.

Vote Share for the Front National and Distance to Nearest Mosque Across Elections.

Dependent Variable	Front National Vote Share (%)
Election	All	Presidential	European	Regional
Election	(1)	(2)	(3)	(4)
Distance to nearest mosque (km)	0.087***	0.166***	0.074***	0.021
Distance to nearest mosque (km)	[0.013]	[0.013]	[0.014]	[0.018]
Distance to nearest mosque, squared (km)	−0.003***	−0.005***	−0.003***	−0.002**
Distance to nearest mosque, squared (km)	[0.001]	[0.001]	[0.001]	[0.001]
Share foreign population (%)	−0.180***	−0.230***	−0.138***	−0.171***
Share foreign population (%)	[0.008]	[0.008]	[0.008]	[0.011]

Controls	Yes	Yes	Yes	Yes

Fixed effects
Elections	3	1	1	1
Life basins	333	333	333	333
Nearest mosque	1,010	1,010	1,010	1,010

Polling stations	24,671	24,671	24,671	24,671
Observations	74,013	24,671	24,671	24,671

Within R²	0.054	0.130	0.088	0.042
Front national vote share mean (%)	10.58	11.10	7.67	12.97
Halo apex (km)	13.77	15.95	14.29	6.33

Notes. This table reports OLS coefficients from estimating Equation 2, separately across all three elections. The dependent variable is the vote share of the Front national in percent. The unit of observation is the polling station. The distance measure is with respect to the nearest mosque located in the same life basin as the polling station. Controls include the polling station’s share of foreign population, log population, area, average age, unemployment rate, the share of population with no diploma, and the share of population living in HLM. Standard errors are in brackets and are clustered at the level of polling stations. Estimates are calculated using Correia’s (2023 [2014]) reghdfe Stata package.

^*** Significant at the 1% level. ^** Significant at the 5% level.

These results suggests that the salience of the election may enhance the halo effect. Presidential elections are often considered high-stakes elections, in contrast to regional and European elections, which are commonly described as second order and result in lower public interest and widespread abstention (Ehin & Talving, 2021). Moreover, PRR parties benefit from increased electoral mobilization in contexts of widespread political distrust (Schulte-Cloos & Leininger, 2022). For these reasons, the halo effect is likely to be accentuated in first-rate ballots. To support this interpretation, we repeat the analysis across quartiles of abstention rates. We report the results in Appendix Table A.5 as well as the predicted Front National vote shares in Appendix Figure A.11.³⁴ Consistent with our interpretation, we identify a more pronounced halo in polling stations with lower abstention rates. For instance, in polling stations in the lower quartile of abstention rates (13% on average), although we find a halo apex located about 16 km from the nearest mosque – close to that for polling stations in the second and third quartiles – the curvature of the halo curve is much more pronounced.

There are good reasons to believe that the effects identified here hold for more recent elections in France.³⁵ In particular, Overos and Sauer (2023) have shown that the presence of mosques in rural areas was associated with greater support for the Front National in the 2017 presidential election, but decreased in densely populated urban areas. However, while their study provides insights into a new episode of the Front National’s expansion as its candidate reached the second round for the second time, it focuses on the municipal level and does not allow for the granular approach we provide here, nor for a precise estimation of the spatial effect of mosque presence on far-right support.

Heterogeneity by Urbanity

Finally, we assess the heterogeneity of the halo effect across polling stations with different population densities and urban structures. To do so, we repeat the analysis from Equation (2) on three different samples: the sample of rural polling stations (those outside urban areas), the sample of peri-urban polling stations (those inside urban areas but outside urban units), and the sample of urban polling stations (those inside urban units). We report the results in Table 6 as well as predicted Front National vote shares in Appendix Figure A.13.³⁶ The halo effect appears to be driven by urban polling stations, with the distinctive curvilinear relationship being particularly pronounced in strictly urban environments. In contrast, we observe a steady increase in the Front National vote share in peri-urban polling stations as the distance from the nearest mosque increases, as well as a reversed relationship in rural polling stations, with a steady decrease in the Front National vote share as the distance from the nearest mosque increases.

Table 6.

Vote Share for the Front National and Distance to Nearest Mosque by Urbanity.

Dependent Variable	Front National Vote Share (%)
Zoning	All	Rural	Peri-urban	Urban
Zoning	(1)	(2)	(3)	(4)
Distance to nearest mosque (km)	0.166***	−0.093**	0.132***	0.158***
Distance to nearest mosque (km)	[0.013]	[0.045]	[0.043]	[0.019]
Distance to nearest mosque, squared (km)	−0.005***	0.001	−0.002	−0.006***
Distance to nearest mosque, squared (km)	[0.001]	[0.001]	[0.002]	[0.001]
Share foreign population (%)	−0.230***	−0.156***	−0.124***	−0.216***
Share foreign population (%)	[0.008]	[0.030]	[0.039]	[0.009]

Controls	Yes	Yes	Yes	Yes

Fixed effects
Life basins	333	232	267	302
Nearest mosque	1,010	287	461	940

Polling stations	24,671	3113	4,867	16,729

Within R²	0.130	0.070	0.117	0.152
Front national vote share mean (%)	10.58	14.36	13.30	9.91
Halo apex (km)	15.95	44.30	32.41	13.53

Notes. This table reports OLS coefficients from estimating Equation 2, separately across types of zoning. Rural refers to polling stations not in urban areas; Peri-urban refers to polling stations in urban areas but not in urban units; Urban refers to polling stations in urban units. The dependent variable is the vote share of the Front national in percent. The unit of observation is the polling station. The distance measure is with respect to the nearest mosque located in the same life basin as the polling station. Controls include the polling station’s share of foreign population, log population, area, average age, unemployment rate, the share of population with no diploma, and the share of population living in HLM. Standard errors are in brackets and are clustered at the level of polling stations. Estimates are calculated using Correia’s (2023 [2014]) reghdfe Stata package.

Significant at the 1% level. ^** Significant at the 5% level.

We propose three tentative explanations to account for these mixed results: the electoral geography of PRR parties, the greater visibility of religious buildings in rural environments, and the localism of some rural communities. First, several studies have shown that the rise of PRR support is particularly strong in areas outside but close to urban environments in France (Faury, 2024; Fourquet, 2012; Girard, 2012; Gombin, 2015b) and other European countries (van Gent et al., 2014). One reason is the class division between the “diversity-seeking” middle classes living in urban centers – who are less likely to support the Front National when in direct contact with Muslim communities, in line with contact theory – and the “traditional” middle and working classes living in peri-urban peripheries—who display a defensive attitude toward social and cultural diversity and tend to vote for PRR parties even when living near Muslim communities (Brookes & Cappellina, 2023). As for rural areas, they are characterized by a strong heterogeneity towards the Front National (Barone & Négrier, 2015; Gombin, 2015b; Huc, 2019). A second explanation for the decreasing support for the Front National in rural polling stations as the distance from the nearest mosque increases could be the more salient visibility of minority religious buildings in rural areas compared to urban and peri-urban areas. This visibility – but also the fact that communes with very few mosques tend to have stronger electoral support for the Front National (Overos & Sauer, 2023) – may explain why the polarizing effect of mosque presence on political behavior is stronger but fades more quickly in these areas. A third explanation may be localism: research has shown that small rural communities, characterized by strong feelings of local attachment, are more likely to support PRR parties (Fitzgerald, 2018). Thus, attachment to one’s community may be associated with a higher perceived threat from religious and ethnic diversification, with this place-based resentment fueling PPR parties. Overall, our findings on the heterogeneous effect of mosque presence on Front National support by urbanity are consistent with previous research showing that the relationship between immigrant presence and far-right support is reversed in rural and urban areas (Barone et al., 2016; Fremerey et al., 2024).

Visibility and Novelty of Mosques

Mosque Visibility

We now evaluate the variations of the halo effect we identify in light of the visibility of mosques. The richness of our data on mosques enables us to measure visibility along three dimensions: whether the mosque has a minaret, its attendance (number of worshipers in 2012), and its surface area (in square meters). To compare polling stations that are closer to mosques with low versus high visibility, we divide the sample into quartiles of visibility for attendance and surface area, and compare polling stations in the highest and lowest quartiles. We then repeat the analysis on these subsamples and report the results in Table 7 as well as the predicted vote shares in Figure 6. We identify a halo in all subsamples, but it is much more pronounced when the mosque is more visible in the local landscape. This is especially the case when the mosque has a minaret and when the building that houses the mosque is larger. Minarets increase the conspicuousness of mosques: in the majority gaze, they are perceived as standing out from their built environment and symbolize a disruptive threat to majority culture (Göle, 2011). As such, mosques with minarets are likely to be associated with greater polarization because both nearby and distant voters can more easily identify the building, with only the former in a position to experience regular, high-quality interactions with Muslim worshipers. Similarly, the surface area of a mosque may signal to passersby the size of the local Muslim community, while also increasing the conspicuousness of the building. These reasons are likely to lead to greater political polarization, with less exclusionary attitudes among nearby residents and stronger anti-Muslim sentiments among more distant residents. This is in line with Lubbers, Coenders, and Scheepers (2006), who show that opposition to asylum centers is stronger for larger centers compared to smaller ones. Nevertheless, we find no heterogeneity across mosques with more worshipers, suggesting that it is the visibility of the building that matters in relation to the halo effect.

Table 7.

Vote Share for the Front National and Distance to Nearest Mosque Across Visibility.

Dependent Variable	Front National Vote Share (%)
Visibility Measure	Minaret		Attendance		Surface Area
	No	Yes	Q1	Q4	Q1	Q4
	(1)	(2)	(3)	(4)	(5)	(6)
Distance to nearest mosque (km)	0.165***	0.288***	0.139***	0.144***	0.121***	0.228***
Distance to nearest mosque (km)	[0.013]	[0.048]	[0.021]	[0.027]	[0.023]	[0.029]
Distance to nearest mosque, squared (km)	−0.006***	−0.007***	−0.005***	−0.003***	−0.006***	−0.006***
Distance to nearest mosque, squared (km)	[0.001]	[0.002]	[0.001]	[0.001]	[0.001]	[0.001]

Own controls	Yes	Yes	Yes	Yes	Yes	Yes
Closest mosque’s station controls	Yes	Yes	Yes	Yes	Yes	Yes

Fixed effects
Life basins	327	27	154	73	107	112

Polling stations	22,883	1,811	6,150	6,010	5,986	5,928

Within R²	0.217	0.332	0.188	0.318	0.231	0.277
Front national vote share mean (%)	11.13	10.71	11.46	9.98	10.45	11.02
Visibility measure mean	0	1	35	576	55	1102
Halo apex (km)	14.10	21.99	13.55	21.59	10.71	20.00

Notes. This table reports OLS coefficients from estimating Equation 1 across various measures of visibility. Q indicates quartiles. The dependent variable is the vote share of the Front national in percent. The unit of observation is the polling station. The distance measure is with respect to the nearest mosque located in the same life basin as the polling station. Controls include the polling station’s share of foreign population, log population, area, average age, unemployment rate, share of population with no diploma, share of population living in HLM, and an indicator for whether it contains a sensitive urban zone. Standard errors are in brackets and are clustered at the level of polling stations. Estimates are calculated using Correia’s (2023 [2014]) reghdfe Stata package.

Significant at the 1% level.

Old Versus New Mosques

We now examine whether the timing of a mosque’s establishment has a differential effect on political polarization. The baseline analysis thus far has focused on mosques established by 1997 and still in existence in 2012, that is, older mosques. To determine whether the integration of new mosques into the urban landscape affected support for the Front National in neighboring polling stations, we compare the halo effect generated by old mosques to that of mosques established after 1997. This sample includes the 1,077 mosques that are present in the 2012 file but not in the 1997 file. Specifically, we estimate Equations (1) and (2) augmented with the two distance measures on the set of polling stations located in a life basin where both an old and a new mosque are present. The results are reported in Appendix Table A.6. When considered separately, distances to both types of mosques generate a comparable halo effect, as seen in Columns (1) and (2). However, when considered together in Column (3), we find that support for the Front National responds to the presence of an old mosque but not to that of a new mosque. This result remains consistent when we include fixed effects for combinations of old and new mosques in Column (4).³⁷

This finding somewhat contradicts predictions derived from the contact hypothesis and tested in other empirical contexts, according to which the long-term presence of immigrant groups induces more positive behavior and attitudes toward these groups (Bursztyn et al., 2024; Steinmayr, 2021). However, this could be explained by the fact that local residents may not be immediately aware of the establishment of a new mosque in their neighborhood, with this awareness materializing gradually over time. Moreover, qualitative research on far-right voting in France has shown that it is precisely when Muslim religion becomes institutionalized in the long term – rather than when religious practices remain private, discreet, and perceived as temporary – that “feelings of invasion” grow stronger among the native population (Faury, 2024).³⁸

Conclusion

Analyses of European politics have regularly shown a correlation between the presence of immigrants and support for PRR parties. They have also highlighted that the scale of analysis matters: at the municipal level, large immigrant populations are associated with lower support for the far right (Della Posta, 2013), while this association is reversed at higher administrative levels (Edo et al., 2019). Our study takes a fresh look at this puzzle by considering an infra-municipal unit of analysis: the polling station. Building on a unique dataset provided by the CARTELEC project – that matches election results in 2007–10 at the polling station level with fine-grained socio-economic indicators in France – we examine whether the distance from a mosque affects support for the Front National. Our research design contributes to the emerging literature on the spatial measurement of exposure to immigrant populations, which attempts to provide a precise assessment of the geographical distance between areas with high immigrant presence and areas with high support for PRR parties (Evans & Ivaldi, 2021; Fremerey et al., 2024; Gravelle et al., 2021). Specifically, we identify an apex at about 16 km from a mosque where electoral support for the Front National is highest. This result is consistent with research at the individual level, which emphasizes that French natives who do not frequently interact with immigrants are significantly less favorable toward immigrants from non-Western countries (Clayton et al., 2021).

Moreover, our study sheds light on the heterogeneous effects that the presence of immigrants can have on far-right support by level of urbanity. The halo effect we identify is driven by urban polling stations, with the distinctive curvilinear relationship being particularly pronounced in strictly urban environments. In contrast, we observe an opposite relationship in rural polling stations, with a steady decline in the Front National vote share as the distance from the nearest mosque increases. These results are consistent with previous findings in Germany, where refugee influx rates have a positive effect on far-right support in rural areas, but a negative effect at the neighborhood level in urban areas (Fremerey et al., 2024). These contrasting effects may be explained by a stronger sense of threat from religious and ethnic diversification among rural residents, with locally tied individuals more likely to be attracted to PPR parties (Fitzgerald, 2018). Another explanation could be that the salience of minority religious buildings (in this case, mosques) may be more pronounced in rural areas, consistent with our findings on the heterogeneous effect of mosque presence on far-right voting according to visibility.

Overall, our findings on the halo effect of mosque presence on far-right support contribute to a broader discussion on the nature of interactions between minority and majority group members (Dinas et al., 2019; Hangartner et al., 2019) and, in particular, between Muslim minorities and non-Muslim majorities in Western European contexts (Adida et al., 2016). While the available data enables us to test effects but not mechanisms, we hypothesize that proximity to a mosque (and its worshipers) facilitates the prejudice-reducing effects of outgroup contact, while moderate spatial distance produces the deleterious effects of outgroup exposure. This hypothesis builds on the classic distinction made in the literature between exposure, based on distant observation of outgroup members, and contact, based on intentional interactions (Janssen et al., 2019; Valdez, 2014). More broadly, it is consistent with a recurring finding on the non-linearity of the relationship between minority presence and far-right support, which deserves further exploration in terms of mechanisms (Janssen et al., 2019; Savelkoul et al., 2017).

In addition, previous ecological studies of the correlation between immigrant presence and support for the Front National have taken the proportion of their foreign population as the main variable of interest (Della Posta, 2013; Lubbers & Scheepers, 2002; Vasilopoulos et al., 2022). However, given the increasing crystallization of anti-Muslim sentiments in European societies (Bleich, 2009) and the growing salience of anti-Muslim discourses in the platform of the Front National (Benveniste & Pingaud, 2016) and other populist movements (Brubaker, 2017; Hafez, 2014), we resort to a different research design to capture the anti-Muslim dimension of the nativist backlash, similar to Colussi et al. (2021) and Gravelle et al. (2021). By using mosques as a proxy for the practicing Muslim population, composed of foreigners, but also of first-generation naturalized immigrants, converts without immigrant ancestry, and second- and third-generation French Muslims, our results point to the importance of taking into account a particular cultural trait – in our case, a minority religion – of the immigrant-origin population when studying its effect on voter polarization. Indeed, when controlling for the share of the foreign population from North Africa and Turkey in neighboring polling stations – an imperfect but useful proxy for identifying Muslim population (Brown, 2000) – we still find a significant effect of exposure to the presence of a mosque, pointing to the specifically anti-Muslim dimension of contemporary Front National support.

A final contribution of this study is to focus on the visibility of immigrant-origin groups rather than their size. We postulate that changes in the visibility of these groups are more likely to affect majority members than the mere number of minority members. Mosques are indeed permanent, conspicuous marks on the urban landscape that make visible the permanent settlement of immigrant-origin Muslim populations as well as their willingness to practice their religion publicly (Becker, 2021). Exposure to these buildings and the Muslim worshipers who regularly visit them is thus likely to shape majority attitudes toward immigration and Islam (Faury, 2024). Indeed, our findings confirm the interest in focusing on the visibility of minority cultural difference, in line with the salience hypothesis (Newman & Velez, 2014; Valdez, 2014). Exposure to a mosque induces support for parties with anti-immigrant and anti-Muslim agendas in polling stations located some distance from the mosque. This effect is stronger for mosques with a minaret and for mosques with a larger surface area suggesting the importance of visibility markers in shaping political behavior. Given the hardening of exclusionary secularism in France (Esmili, 2023) and its use as an identity marker against Islam by the Rassemblement National – the new name of the Front National since 2018 – it is likely that polarization around the presence of mosques will continue to have lasting effects on French politics (Almeida, 2017; Cremer, 2023).

Supplemental Material

Supplemental Material - The Mosque Nearby: Visible Minorities and Far-Right Support in France

Supplemental Material for The Mosque Nearby: Visible Minorities and Far-Right Support in France by Margot Dazey and Victor Gay in Comparative Political Studies

Footnotes

Acknowledgments

We thank Charlotte Cavaillé, Jeffrey Friedman, Gilles Ivaldi, Antoine Jacquet, Horacio Larreguy, Sebastien Montpetit, Imil Nurutdinov, Mounu Prem, Jan Stuckatz, Daniel Tavana, and Sebastian Thieme for fruitful discussions.

Declaration of Conflicting Interests

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

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: We gratefully acknowledge funding from the Agence Nationale de la Recherche under grant ANR-17-EURE-0010 (Investissements d’Avenir program) and grant ANR-17-CONV-0001 (Institut Convergences Migrations).

ORCID iD

Victor Gay

Data Availability Statement

The reproduction package for this article is available at the following address: https://doi.org/10.7910/DVN/PP9ZJ2.

Supplemental Material

Supplemental material for this article is available online.

Notes

Author Biographies

Margot Dazey is a researcher at the CNRS, attached to the CERAPS (University of Lille). Her work focuses on Islam in France, the experience of discrimination, and social stratification within minority groups.

Victor Gay is an assistant professor of economics at Toulouse School of Economics (TSE) and a Research Fellow at the Institute for Advanced Study in Toulouse (IAST).

References

Adida

Laitin

Valfort

M.-A.

(2016). “One muslim is enough!” Evidence from a field experiment in France. Annals of Economics and Statistics, 121–2, 121–160. https://doi.org/10.15609/annaeconstat2009.121-122.121

Adida

Laitin

Valfort

M.-A.

(2017). Why muslim integration fails in christian-heritage societies. Harvard University Press.

Alduy

Wahnich

(2015). Marine Le Pen prise aux mots. Décryptage du nouveau discours frontiste. Seuil.

Alesina

Tabellini

(2024). The political effects of immigration: Culture or economics? Journal of Economic Litterature, 62(1), 5–46. https://doi.org/10.1257/jel.20221643

Allen

(2013). Between critical and uncritical understandings: A case study analyzing the claims of islamophobia made in the context of the proposed ‘super-mosque’ in Dudley, England. Societies, 3(2), 186–203. https://doi.org/10.3390/soc3020186

Allievi

(2010). Mosques in Europe. Why a solution has become a problem. Alliance Publishing Trust.

Allport

(1954). The nature of prejudice. Addison-Wesley.

Almeida

(2017). Exclusionary secularism: The Front National and the Reinvention of Laïcité. Modern and Contemporary France, 25(3), 249–263. https://doi.org/10.1080/09639489.2016.1272561

Amengay

(2020). Les thèmes de prédilection du Front national dans la presse régionale: vers une meilleure compréhension du rôle des médias dans les succès électoraux des partis de la droite radicale populiste. PhD diss. Université d’Ottawa.

10.

Aranguren

Madrisotti

(2019). Des regards aux égards: effets du foulard islamique sur le comportement visuel des passagers du métro de Paris. Hommes & migrations, 1324(1), 39–47. https://doi.org/10.4000/hommesmigrations.8276

11.

Arzheimer

(2009). Contextual factors and the extreme right vote in Western Europe, 1980–2002. American Journal of Political Science, 53(2), 259–275. https://doi.org/10.1111/j.1540-5907.2009.00369.x

12.

Audemard

Gouard

(2016). Le bureau de vote, objet et terrain d’enquêtes électorales. Pole Sud, 44(1), 5–19. https://doi.org/10.3917/psud.044.0005

13.

Barone

D’Ignazio

de Blasio

Naticchioni

(2016). Mr. Rossi, Mr. Hu and politics. The role of immigration in shaping natives’ voting behavior. Journal of Public Economics, 136, 1–13. https://doi.org/10.1016/j.jpubeco.2016.03.002

14.

Barone

Négrier

(2015). Voter Front National en milieu rural. Une perspective ethnographique. In Crépon

(Ed.), Les faux-semblants du Front national (pp. 417–434). Presses de Sciences Po.

15.

Basiri

Haklay

Fody

Mooney

(2019). Crowdsourced geospatial data quality: Challenges and future directions. International Journal of Geographical Information Science, 33(8), 1588–1593. https://doi.org/10.1080/13658816.2019.1593422

16.

Beauchemin

Hamel

Simon

(Eds.). (2018). Trajectories and origins: Survey on the diversity of the French population. Springer International Publishing.

17.

Beauchemin

Ichou

Simon

(2023). Trajectories and origins 2 (2019–2020): A survey on population diversity in France. Population, 78(1), 11–28. https://doi.org/10.3917/popu.2301.0011

18.

Beauguitte

Colange

(2013). Analyser les comportements électoraux à l’échelle du bureau de vote. Agence Nationale de la Recherche. Compte-rendu de fin de projet ANR-FORM-090601-01-01. https://shs.hal.science/halshs-00839899

19.

Becker

(2021). Mosques in the metropolis: Incivility, caste, and contention in Europe. University Press of Chicago.

20.

Becker

Fetzer

Novy

(2017). Who voted for Brexit? A comprehensive district-level analysis. Economic Policy, 32(92), 601–650. https://doi.org/10.1093/epolic/eix012

21.

Benveniste

Pingaud

(2016). Far-right movements in France: The principal role of Front National and the Rise of Islamophobia. In Lazaridis

Campani

Benveniste

(Eds.), The rise of the far right in Europe: Populist shifts and ’othering (pp. 55–79). Palgrave Macmillan UK.

22.

Berning

(2016). Contextual perceived group threat and radical right-wing populist party preferences: Evidence from Switzerland. Research & Politics, 3(1), 1–7. https://doi.org/10.1177/2053168016635670

23.

Biggs

Knauss

(2012). Explaining membership in the British National Party: A multilevel analysis of contact and threat. European Sociological Review, 28(5), 633–646. https://doi.org/10.1093/esr/jcr031

24.

Bleich

(2009). Where do Muslims stand on ethno-racial hierarchies in Britain and France? Evidence from public opinion surveys, 1988–2008. Patterns of Prejudice, 43(3–4), 379–400. https://doi.org/10.1080/00313220903109326

25.

Bolet

(2020). Local labour market competition and radical right voting: Evidence from France. European Journal of Political Research, 59(4), 817–841. https://doi.org/10.1111/1475-6765.12378

26.

Bon

Cheylan

J.-P.

(1988). La France qui vote. Hachette Education.

27.

Bonelli

(2001). Renseignements généraux et violences urbaines. Actes de la Recherche en Sciences Sociales, 136–7(1–2), 95–103. https://doi.org/10.3917/arss.136.0095

28.

Bornioli

Birenboim

Druez

Livny

van der Noll

Mayer

Bloom

P.B.-N.

(2023). Environmental appraisals in outgroup cultural landscapes: The role of muslim elements in urban settings. Cities, 143(2), Article 104579. https://doi.org/10.1016/j.cities.2023.104579

29.

Bowyer

(2008). Local context and extreme right support in England: The British National Party in the 2002 and 2003 local elections. Electoral Studies, 27(4), 611–620. https://doi.org/10.1016/j.electstud.2008.05.001

30.

Boyer

(1992). L’Institut musulman de la Mosquée de Paris. CHEAM.

31.

Bratti

Deiana

Havari

Mazzarella

Meroni

E. C.

(2020). Geographical proximity to refugee reception centres and voting. Journal of Urban Economics, 120, Article 103290. https://doi.org/10.1016/j.jue.2020.103290

32.

Bredtmann

(2022). Immigration and electoral outcomes: Evidence from the 2015 refugee inflow to Germany. Regional Science and Urban Economics, 96, Article 103807. https://doi.org/10.1016/j.regsciurbeco.2022.103807

33.

Brookes

Cappellina

(2023). Political behaviour in France: The impact of the rural–urban divide. French Politics, 21(1), 104–124. https://doi.org/10.1057/s41253-022-00204-x

34.

Brown

(2000). Quantifying the muslim population in Europe: Conceptual and data issues. International Journal of Social Research Methodology, 3(2), 87–101. https://doi.org/10.1080/136455700405163

35.

Brubaker

(2017). Between nationalism and civilizationism: The European populist moment in comparative perspective. Ethnic and Racial Studies, 40(9), 1191–1226. https://doi.org/10.1080/01419870.2017.1294700

36.

Brunner

Kuhn

(2018). Immigration, cultural distance and natives’ attitudes towards immigrants: Evidence from Swiss voting results. Kyklos, 71(1), 28–58. https://doi.org/10.1111/kykl.12161

37.

Brutel

Levy

(2012). Le nouveau zonage en bassins de vie de 2012. INSEE Première, 1425, 1–4.

38.

Bursztyn

Chaney

Hassan

Rao

(2024). The immigrant next door. The American Economic Review, 114(2), 348–384. https://doi.org/10.1257/aer.20220376

39.

Camus

J.-Y.

(1996). Le Front national. Milan.

40.

Cavaillé

Ferwerda

(2023). How distributional conflict over in-kind benefits generates support for far-right parties. The Journal of Politics, 85(1), 19–33. https://doi.org/10.1086/720643

41.

Cesari

(1994). Être musulman en France. Karthala/IREMAM.

42.

Choi

Poertner

Sambanis

(2019). Parochialism, social norms, and discrimination against immigrants. Proceedings of the National Academy of Sciences, 116(33), 16274–16279. https://doi.org/10.1073/pnas.1820146116

43.

Clayton

Ferwerda

Horiuchi

(2021). Exposure to immigration and admission preferences: Evidence from France. Political Behavior, 43, 175–200. https://doi.org/10.2139/ssrn.3268904

44.

Coffé

Heyndels

Jan

(2007). Fertile grounds for extreme right-wing parties: Explaining the Vlaams blok’s electoral success. Electoral Studies, 26(1), 142–155. https://doi.org/10.1016/j.electstud.2006.01.005

45.

Colussi

Isphording

Pestel

(2021). Minority salience and political extremism. American Economic Journal: Applied Economics, 13(3), 23–71. https://doi.org/10.1257/app.20190703

46.

Cools

Finseraas

Rogeberg

(2021). Local immigration and support for anti-immigration parties: A meta-analysis. American Journal of Political Science, 65(4), 988–1006. https://doi.org/10.1111/ajps.12613

47.

Correia

(2023). [2014]. reghdfe: Stata module to perform linear or instrumental-variable regression absorbing any number of high-dimensional fixed effects. Boston College Department of Economics. https://ideas.repec.org/c/boc/bocode/s457874.html.

48.

Couttenier

Hatte

Thoenig

Vlachos

(2024). Anti-muslim voting and media coverage of immigrant crimes. The Review of Economics and Statistics, 16(2), 276–285. https://doi.org/10.1162/rest_a_01152

49.

Cremer

(2023). The rise of the post-religious right: Christianism and secularism in the French rassemblement national. Party Politics, 29(1), 40–50. https://doi.org/10.1177/13540688211046859

50.

Dancygier

Donnelly

(2013). Sectoral economies, economic contexts, and attitudes toward immigration. The Journal of Politics, 75(1), 17–35. https://doi.org/10.1017/S0022381612000849

51.

David

Pilet

J.-B.

Van Hamme

(2018). Scale matters in contextual analysis of extreme right voting and political attitudes. Kyklos, 71(4), 509–536. https://doi.org/10.1111/kykl.12183

52.

Davidson

(2009). La Mosquée de Paris. Construire l’islam français et l’islam en France, 1926–1947. Revue des Mondes Musulmans et de la Méditerranée, 125, 197–215. https://doi.org/10.4000/remmm.6246

53.

Dazey

Gay

(2024). Replication data for: The mosque nearby: Visible minorities and far-right support in France [database]. Harvard Dataverse. https://doi.org/10.7910/DVN/PP9ZJ2

54.

de Blok

van der Meer

(2018). The puzzling effect of residential neighbourhoods on the vote for the radical right: An individual-level Panel study on the mechanisms behind neighbourhood effects on voting for the Dutch Freedom Party, 2010–2013. Electoral Studies, 53, 122–132. https://doi.org/10.1016/j.electstud.2018.04.003

55.

Della Posta

(2013). Competitive threat, intergroup contact, or both? Immigration and the dynamics of Front national voting in France. Social Forces, 92(1), 249–273. https://doi.org/10.1093/sf/sot046

56.

Dinas

Matakos

Xefteris

Hangartner

(2019). Waking up the golden dawn: Does exposure to the refugee crisis increase support for extreme-right parties? Political Analysis, 27(2), 244–254. https://doi.org/10.1017/pan.2018.48

57.

Dinas

van Spanje

(2011). Crime story: The role of crime and immigration in the anti-immigration vote. Electoral Studies, 30(4), 658–671. https://doi.org/10.1016/j.electstud.2011.06.010

58.

Drouhot

(2020). Income segregation and the incomplete integration of islam in the Paris metropolitan area. Socius, 6, 1–3. https://doi.org/10.1177/2378023119899585

59.

Dustmann

Frattini

Preston

(2013). The effect of immigration along the distribution of wages. The Review of Economic Studies, 80(1), 145–173. https://doi.org/10.1093/restud/rds019

60.

Dustmann

Schönberg

Stuhler

(2016). The impact of immigration: Why do studies reach such different results? The Journal of Economic Perspectives, 30(4), 31–56. https://doi.org/10.1257/jep.30.4.31

61.

Dustmann

Vasiljeva

Damm

A. P.

(2019). Refugee migration and electoral outcomes. The Review of Economic Studies, 86(5), 2035–2091. https://doi.org/10.1093/restud/rdy047

62.

Edo

Giesing

Öztunc

Poutvaara

(2019). Immigration and electoral support for the far-left and the far-right. European Economic Review, 115, 99–143. https://doi.org/10.1016/j.euroecorev.2019.03.001

63.

Ehin

Talving

(2021). Still second-order? European elections in the era of populism, extremism, and euroscepticism. Politics, 41(4), 467–485. https://doi.org/10.1177/0263395720986026

64.

Endrich

(2023). A gate to the world for all? The reaction of neighborhoods in hamburg to refugee housing. European Journal of Political Economy, In Press. https://doi.org/10.1016/j.ejpoleco.2023.102455

65.

Esmili

(2023). La possibilité d’un complot. Contemporary French and Francophone Studies, 27(2), 226–239. https://doi.org/10.1080/17409292.2023.2185418

66.

Etchebarne

(1996a). Le FN dans le Nord ou les logiques d’une implantation électorale. In Mayer

(Ed.), Le Front National à découvert (pp. 284–306). Presses de Sciences Po.

67.

Etchebarne

(1996b). Préface à la première édition. In Mayer

(Ed.), Le Front National à découvert (pp. 11–14). Presses de Sciences Po.

68.

Evans

Ivaldi

(2021). Contextual effects of immigrant presence on populist radical right support: Testing the ‘halo effect’ on Front National voting in France. Comparative Political Studies, 54(5), 823–854. https://doi.org/10.1177/0010414020957677

69.

Faury

(2024). Des électeurs ordinaires. Enquête sur la normalisation de l’extrême droite. Seuil.

70.

Fitzgerald

(2018). Close to home: Local ties and voting radical right in Europe. Cambridge University Press.

71.

Fourquet

(2012). Le sens des cartes: analyse sur la géographie des votes à la présidentielle. Fondation Jean Jaurès.

72.

Fremerey

Hörnig

Schaffner

(2024). Becoming neighbors with refugees and voting for the far-right? The impact of refugee inflows at the small-scale level. Labour Economics, 86, Article 102467. https://doi.org/10.1016/j.labeco.2023.102467

73.

Gamalerio

Luca

Romarri

Viskanic

(2023). Refugee reception, extreme-right voting, and compositional amenities: Evidence from Italian municipalities. Regional Science and Urban Economics, 100, Article 103892. https://doi.org/10.1016/j.regsciurbeco.2023.103892

74.

Georgiadou

Rori

Roumanias

(2018). Mapping the European far right in the 21st century: A meso-level analysis. Electoral Studies, 54, 103–115. https://doi.org/10.1016/j.electstud.2018.05.004

75.

Gessler

Tóth

Wachs

(2022). No country for asylum seekers? How short-term exposure to refugees influences attitudes and voting behavior in Hungary. Political Behavior, 44, 1813–1841.

76.

Girard

(2012). Explaining the periurban right-wing vote: ‘Social frustrations’ of low-income households or the reshaping of the working classes? Métropolitiques. https://metropolitiques.eu/Explaining-the-periurban-right.html (Accessed in July 2024).

77.

Golder

(2003). Explaining variation in the success of extreme right parties in Western Europe. Comparative Political Studies, 36(4), 432–466. https://doi.org/10.1177/0010414003251176

78.

Golder

(2016). Far right parties in Europe. Annual Review of Political Science, 19(1), 477–497. https://doi.org/10.1146/annurev-polisci-042814-012441

79.

Göle

(2011). The public visibility of islam and European politics of resentment: The minarets-mosques debate. Philosophy & Social Criticism, 37(4), 383–392. https://doi.org/10.1177/0191453711398773

80.

Gombin

(2015a). Créer et diffuser de l’information électorale au niveau des bureaux de vote. Congrès de l’Association française de science politique. https://shs.hal.science/halshs-01588887

81.

Gombin

(2015b). Le changement dans la continuité: géographies électorales du Front National depuis 1992. In Crépon

Dézé

Mayer

(Eds.), Les faux-semblants du Front National (pp. 395–416). Presses de Sciences Po.

82.

Gravelle

(2021). Explaining islamophobia in Australia: Partisanship, intergroup contact, and local context. Australian Journal of Political Science, 56(2), 132–152. https://doi.org/10.1080/10361146.2021.1884645

83.

Gravelle

Medeiros

Nai

(2021). In the shadow of the tower: Spatial proximity to mosques, visible diversity, and support for the radical right. Political Geography, 91, Article 102499. https://doi.org/10.1016/j.polgeo.2021.102499

84.

Hafez

(2014). Shifting borders: Islamophobia as common ground for building pan-European right-wing unity. Patterns of Prejudice, 5(48), 479–499. https://doi.org/10.1080/0031322X.2014.965877

85.

Halla

Wagner

Zweimüller

(2017). Immigration and voting for the far right. Journal of the European Economic Association, 15(6), 1341–1385. https://doi.org/10.1093/jeea/jvx003

86.

Hangartner

Elias

Moritz

Matakos

Xefteris

(2019). Does exposure to the refugee crisis make natives more hostile? American Political Science Review, 113(2), 442–455. https://doi.org/10.1017/S0003055418000813

87.

Harmon

(2018). Immigration, ethnic diversity, and political outcomes: Evidence from Denmark. The Scandinavian Journal of Economics, 120(4), 1043–1074. https://doi.org/10.1111/sjoe.12239

88.

Hennig

(2021). Neighborhood quality and opposition to immigration: Evidence from German refugee shelters. Journal of Development Economics, 150, Article 102604. https://doi.org/10.1016/j.jdeveco.2020.102604

89.

Hopkins

(2010). Politicized places: Explaining where and when immigrants provoke local opposition. American Political Science Review, 104(1), 40–60. https://doi.org/10.1017/S0003055409990360

90.

Huc

(2019). FN du Nord contre FN du Sud? Analyse sociogéographique des électorats Le Pen en 2017. Revue Française de Science Politique, 69(2), 223–247. https://doi.org/10.3917/rfsp.692.0007

91.

Ignazi

(1996). The crisis of parties and the rise of new political parties. Party Politics, 2(4), 549–566. https://doi.org/10.1177/1354068896002004007

92.

INSEE . (1999). Recensement de la population 1999: tableaux analyses [database]. INSEE [producer], ADISP [distributor]. https://data.progedo.fr/studies/doi/10.13144/lil-0144

93.

INSEE . (2000a). Contours IRIS 2000 [database]. INSEE [producer], ADISP [distributor]. https://data.progedo.fr/studies/doi/10.13144/lil-0262

94.

INSEE . (2000b). Correspondances IRIS - TRIRIS [database]. INSEE [producer], ADISP [distributor]. https://data.progedo.fr/studies/doi/10.13144/lil-0271

95.

Ivaldi

(2016). Le Front National français dans l’espace des droites radicales européennes. Pouvoirs, 157(2), 115–126. https://revue-pouvoirs.fr/le-front-national-francais-dans-l/.

96.

Ivaldi

Dutozia

(2018). The ‘territorialization’ of the Front National’s populist politics in France. Comparative European Politics, 16(6), 1033–1050. https://doi.org/10.1057/s41295-018-0145-y

97.

Jadot

Bussi

Colange

Freire-Diaz

(2010). Un outil d’analyse électorale en cours de création. CARTELEC, un SIG au niveau des bureaux de vote français. Revue du comité français de cartographie, 205, 81–98. https://www.lecfc.fr/new/articles/205-article-8.pdf.

98.

Janssen

van Ham

Kleinepier

ieuwenhuis

(2019). A micro-scale approach to ethnic minority concentration in the residential environment and voting for the radical right in The Netherlands. European Sociological Review, 35(4), 552–566. https://doi.org/10.1093/esr/jcz018

99.

Jardin

Préteceille

Robert

Zauberman

(2021). Territoires et insécurité en Île-de-France. Déviance et Société, 45(2), 319–355. https://doi.org/10.3917/ds.452.0121

100.

Josselin

Yahiaoui

Sanjuan

Marchand-Lagier

(2016). Automatic delineation of polling station areas using legal texts. Proceedings of AGILE’ 2016. https://hal.science/hal-01338079

101.

Jouanneau

(2013). Les imams en France: une autorité religieuse sous contrôle. Agone.

102.

Kallis

(2018). The radical right and islamophobia. In Rydgren

(Ed.), The oxford handbook of the radical right (pp. 42–60). Oxford University Press. https://doi.org/10.1093/oxfordhb/9780190274559.013.3

103.

Kaufmann

Goodwin

(2018). The diversity wave: A meta-analysis of the native-born white response to ethnic diversity. Social Science Research, 76(2), 120–131. https://doi.org/10.1016/j.ssresearch.2018.07.008

104.

Kavanagh

Menon

Heinze

(2021). Does health vulnerability predict voting for right-wing populist parties in Europe? American Political Science Review, 115(3), 1104–1109. https://doi.org/10.1017/S0003055421000265

105.

Kellermann

Simon

(2022). Immigration and anti-immigrant voting in the 2017 German parliamentary election. German Economic Review, 23(3), 341–401. https://doi.org/10.1515/ger-2021-0026

106.

Knigge

(1998). The ecological correlates of right–wing extremism in Western Europe. European Journal of Political Research, 34(2), 249–279. https://doi.org/10.1111/1475-6765.00407

107.

Legrain

J.-F.

(1986). Islam en France, Islam de France. Esprit, 119(10), 1–30.

108.

Lewis-Beck

Mitchell

(1993). French electoral theory: The National Front test. Electoral Studies, 12(2), 112–127. https://doi.org/10.1016/0261-3794(93)90013-A

109.

Lonsky

(2021). Does immigration decrease far-right popularity? Evidence from Finnish municipalities. Journal of Population Economics, 34(1), 97–139. https://doi.org/10.1007/s00148-020-00784-4

110.

Lubbers

Coenders

Scheepers

(2006). Objections to asylum seeker centres: Individual and contextual determinants of resistance to small and large centres in The Netherlands. European Sociological Review, 22(3), 243–257. https://doi.org/10.1093/esr/jci055

111.

Lubbers

Gijsberts

Scheepers

(2002). Extreme right-wing voting in Western Europe. European Journal of Political Research, 41(3), 345–378. https://doi.org/10.1111/1475-6765.00015

112.

Lubbers

Scheepers

(2002). French Front national voting: A micro and macro perspective. Ethnic and Racial Studies, 25(1), 120–149. https://doi.org/10.1080/01419870120112085

113.

Martig

Bernauer

(2018). The halo effect: Perceptions of diffuse threat and SVP vote share. World Political Science, 14(1), 27–54. https://doi.org/10.1515/wps-2018-0002

114.

Mayda

Peri

Steingress

(2022). The political impact of immigration: Evidence from the United States. American Economic Journal: Applied Economics, 14(1), 358–389. https://doi.org/10.1257/app.20190081

115.

Mayer

(2015). Le plafond de verre électoral entamé, mais pas brisé. In Crépon

Dézé

Mayer

(Eds.), Les faux-semblants du Front National (pp. 297–322). Presses de Sciences Po.

116.

McAvay

(2018). The ethnoracial context of residential mobility in France: Neighbourhood out-migration and relocation. Population, Space and Place, 24(6), Article e2138. https://doi.org/10.1002/psp.2138

117.

Mendez

Cutillas

I. M.

(2014). Has immigration affected Spanish presidential elections results? Journal of Population Economics, 27(1), 135–171. https://doi.org/10.1007/s00148-013-0471-y

118.

Miller

Grubesic

(2021). A spatial exploration of the Halo effect in the 2016 U.S. Presidential election. Annals of the American Association of Geographers, 111(4), 1094–1109. https://doi.org/10.1080/24694452.2020.1785271

119.

Morning

(2015). Ethnic classification in global perspective: A cross-national survey of the 2000 census round. In Simon

Piché

Gagnon

A. A.

(Eds.), Social statistics and ethnic diversity: Cross-national perspectives in classifications and identity politics (pp. 17–37). Springer International Publishing.

120.

Mudde

(2013). Three decades of populist radical right parties in Western Europe: So what? European Journal of Political Research, 52(1), 1–19. https://doi.org/10.1111/j.1475-6765.2012.02065.x

121.

Newman

Velez

(2014). Group size versus change? Assessing Americans’ perception of local immigration. Political Research Quarterly, 67(2), 293–303. https://doi.org/10.1177/1065912913517303

122.

Olzak

(1992). The dynamics of ethnic competition and conflict. Stanford University Press.

123.

Otto

Steinhardt

(2014). Immigration and election outcomes? Evidence from city districts in hamburg. Regional Science and Urban Economics, 45(1), 67–79. https://doi.org/10.1016/j.regsciurbeco.2014.01.004

124.

Overos

H. D.

Sauer

(2023). Dieu dans la Commune: Religion and Voting in the 2017 French Election. Politics and Religion, 16(4), 605–633. https://doi.org/10.1017/S1755048323000184

125.

Pardos-Prado

Xena

(2019). Skill specificity and attitudes toward immigration. American Journal of Political Science, 63(2), 286–304. https://doi.org/10.1111/ajps.12406

126.

Perrineau

(1985). Le Front national un électorat autoritaire. Revue Politique et Parlementaire, 918, 24–31.

127.

Pettigrew

Tropp

(2006). A meta-analytic test of intergroup contact theory. Journal of Personality and Social Psychology, 90(5), 751–783. https://doi.org/10.1037/0022-3514.90.5.751

128.

Pettrachin

Gabrielli

Kim

Ludwig-Dehm

Pötzschke

(2023). Did exposure to Asylum seeking migration affect the electoral outcome of the ‘Alternative für Deutschland’ in Berlin? Evidence from the 2019 European elections. Journal of Ethnic and Migration Studies, 49(2), 576–600. https://doi.org/10.1080/1369183X.2022.2100543

129.

Polavieja

(2016). Labour-market competition, recession and anti-immigrant sentiments in Europe: Occupational and environmental drivers of competitive threat. Socio-Economic Review, 14(3), 395–417. https://doi.org/10.1093/ser/mww002

130.

Quillian

(1995). Prejudice as a response to perceived group threat: Population composition and anti-immigrant and racial prejudice in Europe. American Sociological Review, 60(4), 586–611. https://doi.org/10.2307/2096296

131.

Rey

Roy

(1986). Quelques réflexions sur l’évolution électorale d’un département de la banlieue parisienne, la Seine-Saint-Denis. Hérodote: Revue de géographie et de géopolitique, 43, 6–38. http://geoprodig.cnrs.fr/items/show/180413.

132.

Rydgren

Ruth

(2013). Contextual explanations of radical right-wing support in Sweden: Socioeconomic marginalization, group threat, and the halo effect. Ethnic and Racial Studies, 36(4), 711–728. https://doi.org/10.1080/01419870.2011.623786

133.

Savelkoul

Laméris

Tolsma

(2017). Neighbourhood ethnic composition and voting for the radical right in The Netherlands. The role of perceived neighbourhood threat and interethnic neighbourhood contact. European Sociological Review, 33(2), 209–224. https://doi.org/10.1093/esr/jcw055

134.

Sbaï

(2006). La république et la mosquée: genèse et institution(s) de l’islam en France. In Luizard

P.-J.

(Ed.), Le choc colonial et l’islam (pp. 223–236). La Découverte.

135.

Schaub

Gereke

Baldassarri

(2021). Strangers in hostile lands: Exposure to refugees and right-wing support in Germany’s Eastern regions. Comparative Political Studies, 54(3–4), 686–717. https://doi.org/10.1177/0010414020957675

136.

Schulte-Cloos

Leininger

(2022). Electoral participation, political disaffection, and the rise of the populist radical right. Party Politics, 28(3), 431–443. https://doi.org/10.1177/1354068820985186

137.

Schwengler

(2003). L’ouvrier caché: le paradoxe du vote rural d’extrême droite dans la France du Nord-Est. Revue Française de Science Politique, 53(4), 513–533. https://doi.org/10.3917/rfsp.534.0513

138.

Shelton

Zook

Graham

(2012). The technology of religion: Mapping religious cyberscapes. The Professional Geographer, 64(4), 602–617. https://doi.org/10.1080/00330124.2011.614571

139.

Simon

(2010). Statistics, French social sciences and ethnic and racial social relations. Revue Française de Sociologie, 51(5), 159–174. https://doi.org/10.2307/40731133

140.

Simon

(2012). Collecting ethnic statistics in Europe: A review. Ethnic and Racial Studies, 35(8), 1366–1391. https://doi.org/10.1080/01419870.2011.607507

141.

Simon

Lopatin

Oliveri

A. M.

González-Ferrer

Fortunato

(2019). Surveying ethnic minorities and migrants. A legal framework for collecting data and other methodological issues. Sciences Po.

142.

Steinmayr

(2021). Contact versus exposure: Refugee presence and voting for the far right. The Review of Economics and Statistics, 103(2), 310–327. https://doi.org/10.1162/rest_a_00922

143.

Sui

Elwood

Goodchild

(2013). Crowdsourcing geographic knowledge. Volunteered geographic information (VGI) in theory and practice. Springer.

144.

Valdez

(2014). Visibility and votes: A spatial analysis of anti-immigrant voting in Sweden. Migration Studies, 2(2), 162–188. https://doi.org/10.1093/migration/mnu029

145.

van Es

(2020). Roasting a pig in Front of a mosque: How pork matters in Pegida’s anti-islam protest in Eindhoven. Religions, 11(7), 359. https://doi.org/10.3390/rel11070359

146.

van Gent

Jansen

Smits

(2014). Right-wing radical populism in City and Suburbs: An electoral geography of the Partij Voor de Vrijheid in the Netherlands. Urban Studies, 51(9), 1775–1794. https://doi.org/10.1177/0042098013505889

147.

Vasilopoulos

McAcay

Brouard

(2022). Residential context and voting for the far right: The impact of immigration and unemployment on the 2017 French presidential election. Political Behavior, 44, 1703–1730. https://doi.org/10.1007/s11109-021-09676-z

148.

Vertier

Viskanic

Gamalerio

(2023). Dismantling the ‘jungle’: Migrant relocation and extreme voting in France. Political Science Research and Methods, 11(1), 129–143. https://doi.org/10.1017/psrm.2022.26

149.

Weber

(2015). National and regional proportion of immigrants and perceived threat of immigration: A three-level analysis in Western Europe. International Journal of Comparative Sociology, 56(2), 116–140. https://doi.org/10.1177/0020715215571950

150.

Winter

(2021). [2014]. combomarginsplot: Stata module to combine the saved results from multiple calls to margins into one marginsplot. Statistical Software Components, Boston College Department of Economics. https://ideas.repec.org/c/boc/bocode/s457804.html

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

3.03 MB