1Numerous studies have demonstrated the important role played by the Catholic symbolic system in explaining political attitudes and behaviors. By “symbolic system”, we are referring to an organized body of representations, beliefs, norms, and values linking all individuals who subscribe to them into a single community. The affective component of this system is, in our view, essential. Yet one can more or less intensely belong to a community or group; belonging is graduated. For example, one is not a believer or nonbeliever, practicing or non-practicing, Catholic or not; one is more or less a believer, more or less practicing, more or less Catholic. If one observes that the order of variations of a dependent variable (attitudes or behaviors, for example) corresponds to that of an independent variable such as level of group belonging, one confirms the hypothesis according to which these attitudes or behaviors indeed belong, in terms of probabilities, to the symbolic system specific to this group.
2To pursue this analysis, we have chosen frequency of mass attendance as a marker of the level of Catholic integration. [1] More specifically, in order to construct this indicator, we have combined answers to two questions, one regarding religious belonging, the other regarding Sunday service attendance. [2] It has been observed that, among the five dimensions by which Charles Glock defined religion as an object, [3] the ritualist (measured by a scale of religious practices), ideological (scale of beliefs) and experiential (feelings relating to the personal experience of spiritual life) dimensions were strongly interrelated. And it has been shown that the frequency of religious practice constitutes a good indicator of all beliefs and practices specific to Catholicism and, by extension, of the degrees of belonging to its symbolic system. [4] In practical terms, this means that, each time we encounter a variable exhibiting a relationship of analogous order to that of the level of Catholic integration (in decreasing order: weekly churchgoers, monthly churchgoers, occasional churchgoers, non-churchgoers and the non-religious), we will consider it as belonging to the Catholic symbolic system.
The relationship between level of Catholic integration and rightwing voting
3By comparing census maps of religious practice with election result maps, André Siegfried and the field of electoral geography [5] demonstrated the strong relationship uniting high levels of religious practice and voting for the right. Analyzing the results of the 1956 legislative elections, Mattei Dogan and Jacques Derivry [6] confirmed the existence of this correlation by relating the proportion of communist votes cast at the cantonal level to the proportion of Sunday churchgoers recorded by canon Fernand Boulard. [7] A regression analysis showed the preponderant influence of the religious variable, which “explained” 40 per cent of the variation. [8] Drawing upon the same “Boulard data”, this salient relationship has also been observed for the results of the 1958 legislative elections and 1974 presidential election.
4These results were challenged on the basis of ecological statistics: did correlations obtained by observing the behavior of collectivities account for the actual behavior of the individuals comprising them? Additional confirmation – this time on the basis of individual data – was supplied by the use of polling surveys of nationally representative samples conducted on the occasion of the 1965 presidential election. [9] Seeking to test the hypothesis that the superstructure could predominate over the infrastructure, it was also shown that the positive relationship between degree of religious practice and voting for the right could not fundamentally be ascribed to the effect of an infrastructural variable. The introduction of a third, socio-economic or socio-demographic type variable such as sex, age, education level, income or profession did not eliminate the relationship between practice and voting. [10]
5Research carried out on the political universe of French people on the basis of qualitative data (non-directive interviews) and a questionnaire distributed to a representative sample (1966) revealed the existence of two, diametrically opposed models of political sub-cultures.
6The first is organized around religion. It gives priority to the person (inseparable from family, home, and the spiritual values that supply the foundation of this sub-culture) and the at once material and symbolic heritage that is to be passed on. This system of representations and valuations is most often accompanied by voting for the right. The second model is organized around social class, economic realities, and the antagonisms and solidarities with which they are associated. The use of data from the questionnaire allowed these results to be more deeply explored and also confirmed that a relationship indeed existed between level of integration to Catholicism and voting for the right. [11] The continuity of this relationship was once again confirmed in 1978, 1988, 1995, [12] 1997 [13] and 2002. [14]
7This is not a French exception. In the twelve countries that made up the European Community, only a very small proportion of churchgoers support leftwing parties, a finding that holds regardless of the source (Eurobarometers, [15] EVS 1990, [16] ISSP 1998, [17] EVS 1999, [18] ESS 2004 [19]), country, or denomination under consideration (Catholic, Protestant or Orthodox). The proportion is higher among non-churchgoers and reaches its highest level among the non-religious. [20] As early as 1944, researchers at Columbia University in the United States showed the importance of religious affiliation in predisposing voter choice. [21] And more recent studies have shown the additional role played by the frequency with which one attends service. [22] In a collection of socially and economically comparable countries, all Christian denominations thus seem to be affected by the association between a high level of religious integration and social, cultural, and political conservatism. A leftwing political orientation predominates among immigrant-origin French people with roots in North Africa, sub-Saharan Africa, and Turkey and varies little with the degree of their affiliation to Islam. [23]
The permanence of the relationship between a high degree of integration to Catholicism and voting for the right
8For long periods of French history, religious and political divisions have coincided: “There used to be no surer touchstone in dividing the right and the left.” [24] Opposed to the political changes that stemmed from the French Revolution, the Catholic Church was long a faithful ally of conservative parties, at least until the mid-twentieth century. However, the early emergence of a left-leaning Catholicism – admittedly, a minority phenomenon – must not be forgotten.
9But the historical conditions that presided over the birth and preservation of this association between the Church and conservatism no longer exist and it now appears anachronistic in the view of some. Indeed, French society as a whole experienced and continues to undergo violent upheavals, whether at the socio-economic level or that of values. The form and nature of the issues have changed, particularly those that simultaneously engage the Catholic Church and society as a whole. Evolving mores, in particular, have affected traditional conceptions of the family. The problem of secularism, which, more than any other issue, long served as a focus of conflict, has today taken new forms. It is often claimed that individual political choice is today less dependent on social and ideological identifications than in the past and that the transmission of attitudes and behaviors from one generation to the next faces greater obstacles. Individual political choice appears less and less to obey group norms and the recommendations of organizations. Finally, the Catholic Church has not given voting recommendations since 1958.
10We are also witnessing major transformations in the religious field itself, with the consequences of the Vatican II Council interpreted as a shift from a Catholicism dominated by the institution to a Catholicism in which emphasis is placed on the believer’s individual act of faith. Secularization is first and foremost reflected in a drop in the number of adherents to the Catholic faith, which has in the past 35 years lost a third of its population in France. [25] Second, Catholic regular churchgoers represent only between 6 per cent and 8 per cent of the population of registered voters, with occasional churchgoing Catholics representing between 12 per cent and 16 per cent. The majority of Catholics are thus non-practicing to the degree that they no longer attend church except for such social events as baptisms, marriages and funerals. These non-practicing Catholics today represent more than a third of the population, equivalent to the population of French people who describe themselves as “non-religious”. Regardless of their internal transformations, by contrast, the share of the population consisting of Protestants and Jews – around 2 per cent in total – has remained unchanged for 35 years. While estimates of the number of Muslims vary significantly depending on the study and how it is conducted, Islam today does indeed seem to be France’s second-largest religion. [26]
11These quantitative changes, in turn, have been accompanied and are no doubt partly explained by a profound transformation in our societies’ relationship to the sacred in recent decades. In keeping with major trends of social change, one observes what might be called an individualization of belief. This tendency reflects a form of religious deinstitutionalization, particularly in what concerns Catholicism. [27] The British sociologist Grace Davie has identified a similar development in the British case, which she summarizes in the expression “believing without belonging”. [28] To describe Europeans’ new relationship to religion, moreover, Davie advances the notion of “vicarious religion”: [29] it is as if the population delegated religious practice to a subgroup while at the same time maintaining ties to this minority, ties that are revealed on important occasions.
12Finally, recent studies have underscored the diversity of contemporary French Catholicism. [30] A few years ago, Jean-Marie Donegani drew upon this claim to contest the relationship between Catholicism and rightwing voting in France. [31] In his view, one must not ignore the freedom of choice that has been granted Catholics since Vatican II proclaimed the legitimate authority of earthly entities in the Gaudium et spes constitution. But a portion of the faithful long ago ceased to give much or any heed to the parish priest’s guidance from the pulpit on how to vote – recommendations that, in any case, have become ever less explicit over the years. [32] There has thus long been a minority of leftwing Catholics. Yet it remains the case in Europe that, as Donegani has underscored, [33] “the more practicing one is, the more one tends to vote for the right”.
Voting in the 2012 presidential election and integration to Catholicism
13Did the 2012 election represent a break with this remarkable continuity? That is the question we propose to examine here. The specificity of the presidential election – a nationwide ballot for a single office – goes hand in hand with an increased personalization of candidates and the growing importance placed on their image. [34] What’s more, one must consider the particularities of the historical and political context in which the election took place. In the seventeen years following the end of François Mitterrand’s second seven-year mandate, France had been continuously governed by successive rightwing presidents. First, then, was the possibility that the election would bring a president of a different political complexion to power. The increasing complexity of the political sphere must also be taken into account: in addition to the two traditional poles of the right-left axis, the National Front had emerged as a third force in French politics. [35] Recent years, finally, have seen a proliferation of studies claiming to identify growing individualization in voting behavior and a declining influence of so-called “heavy variables”.
Data used
14In order to analyze the role played by degree of Catholic integration in voter choice during the presidential election, we drew upon Présidoscope 2012 [36] to identify a sample of 2911 registered voters representative of the French population aged 18 and older.
First-round voting and degree of Catholic integration
15In the first round of the election, votes for the parliamentary right regularly increased in relation to the level of integration to Catholicism, from only 23 per cent among the non-religious to 71 per cent among regular churchgoers (Nicolas Sarkozy 24-51 per cent and François Bayrou 9-18 per cent) (Table 1 and Figure 1). Regular churchgoers, by contrast, were the group that voted least often for Marine Le Pen, suggesting that the standoffish position adopted by the Catholic Church towards her party had influenced the faithful. Among lapsed Catholics, on the other hand, the National Front received a larger share of the vote. All told, as groups with ideological reasons to oppose the rejection of the Other, both regularly practicing Catholics and the non-religious tended to refuse to vote for Le Pen.
16Votes for the left evolved in the opposite direction, diminishing as integration to Catholicism increased. Limited to 19 per cent of regular churchgoers, they rose to 56 per cent among the non-religious. This trend was even more marked for Jean-Luc Mélenchon (from 2 to 17 per cent) than for François Hollande (12 to 33 per cent).
1st and 2nd round voting in presidential elections by religious practice (in %)
1st and 2nd round voting in presidential elections by religious practice (in %)
The numbers in parentheses represent the total number on the basis of which the percentages have been calculated: here, the total number of individuals corresponding to each level of religious observance is indicated along the horizontal axis. At the same time, the numbers in parentheses also indicate the direction in which the tables are to be read (read horizontally by line: for example, 71 per cent of regular churchgoers state that they voted for the parliamentary right while 19 per cent state that they voted for the left).Voting by religious practice (1st and 2nd rounds)
Voting by religious practice (1st and 2nd rounds)
17In the second round (Table 1, Figure 1), the number of votes cast for Nicolas Sarkozy increased with degree of religious integration: from 27 per cent among the non-religious to 67 per cent among regular churchgoers. At the same time, those who voted for François Hollande diminished from 61 per cent to 24 per cent in the same groups. The relationship between degree of integration to Catholicism and propensity to vote for the right does not therefore seem to be called into question: 40 points (and even 50 points in the first round) separate the most integrated from the non-religious. And that despite the fact that, in the decisive round, a quarter of regularly practicing Catholics nevertheless voted for Hollande, with a quarter of the non-religious choosing Sarkozy.
18Few differences are to be observed in comparison with the previous presidential election. In 2007 and 2012, the first-round votes received by Nicolas Sarkozy exhibit very similar curves according to religious practice. The same holds for all candidates of the left. By contrast, the 2012 Bayrou vote clearly differs from that of 2007 (Figure 2). Although the Bayrou vote was closely related to Catholic integration in 2002, this was no longer the case in 2007; it thus lost its “religious texture”. [37] In 2012, the Bayrou vote was once again closely associated with religious integration but at the cost of an overall reduction in this electorate. In 2007, in fact, Bayrou had benefitted from a number of votes foreign to traditional centrism, many of them cast by left-leaning voters not won over to the candidacy of Ségolène Royal. This explains why, compared to most elections, the votes cast for the MoDem candidate in 2007 exhibited such an atypical religious profile. The phenomenon obviously did not repeat itself in 2012.
1st round voting for Bayrou and Le Pen in 2007 and 2012
1st round voting for Bayrou and Le Pen in 2007 and 2012
19Finally, despite what may have been written regarding the consequences of Jean-Marie Le Pen’s replacement as candidate by his daughter Marine, the variations observed in 2007 are once again to be found in 2012, only in more distinct fashion. While the Le Pen vote remained infrequent among the non-religious and, even more so, among regular churchgoers, it continued to flourish among lapsed Catholics, who are largely unreceptive to the views of the institution.
The effects of the electoral campaign
20In explanations of voting behavior, a distinction is often made between issue voting and cleavage voting. Studies that seek to assess the influence played by degree of integration to Catholicism obviously present the phenomenon as falling under the latter category. But the data upon which we worked – the responses of voters questioned on twelve occasions over the course of the electoral campaign (November 2011 to June 2012) – allowed us to evaluate both the influence of degree of religious integration and the impact of issues in determining voter choice in relation to the religious divide.
21Indeed, it could be argued that the purpose of the electoral campaign is to emphasize the issues that are important at the time and the candidates who champion them. If it is to be said that some issues play a role in determining the voter’s final decision, one must be able to observe a transformation in intentions to vote from the start of the campaign (when only cleavage voting considerations would have had an effect) to the moment when ballots are actually cast (after current issues have come into play). If the issues have interfered with the effect of Catholic integration, one should be able to identify disruptions in the significant ordinal relationship of the religious effect.
22However, the curves of voting intention and actual votes cast always take the same form: whether it be in the first interview or the tenth, intentions to vote for the parliamentary right always increase with degree of Catholic integration while those for the left decrease. The choice of Marine Le Pen, meanwhile, is always more frequent among lapsed Catholics. Nevertheless, a closer look allows one to qualify these results somewhat (Table 2). The intention to vote for the left is fairly constant between the first and final waves. Over the course of the campaign, however, the intention to vote for Mélenchon was confirmed as religious integration diminished (practicing Catholics: 54 per cent; the non-religious: 85 per cent) whereas confirmation of the Hollande vote diminished (31 per cent; 11 per cent). Confirmation of the Hollande vote seemed to vary little with degree of religious integration (from 77 to 73 per cent). However, one observes that the shift in favor of Mélenchon increased from 6 per cent (churchgoers) to 12 per cent (non-religious) while the transformation into a vote for Bayrou diminished (7 to 4 per cent).
Proportion of voting intentions (v1) expressed in actual votes cast (v10) by religious practice and electorates (in %)
Proportion of voting intentions (v1) expressed in actual votes cast (v10) by religious practice and electorates (in %)
23The intention to vote for Sarkozy strengthened as religious integration increased (from 73 per cent to 86 per cent). As the latter diminished, by contrast, this intention was transformed into a vote for Le Pen (from 3 to 9 per cent). At the same time, the intention to vote for Bayrou was confirmed (from 55 per cent to 71 per cent) the greater the degree of Catholic integration (or turned into a vote for Le Pen: from 3 per cent to 6 per cent); it turned into a vote for Mélenchon as religious integration diminished (from 0 per cent to 14 per cent). As for the intention to vote for Le Pen, it was somewhat confirmed as religious integration diminished (from 68 per cent to 72 per cent) while at the same time being transformed into a vote for the left (from 4 per cent to 9 per cent); by contrast, the transformation into a vote for Sarkozy diminished from 17 per cent to 7 per cent.
24It therefore appears as if the transformation of intentions to vote into actual votes in November 2011 expressed strengthening influence of degree of religious integration in determining the choice between right and left.
25Given what is known about the system of values specific to Catholics, we have sought to determine the themes that, as they developed over the course of the campaign, could have played a role in the choice or rejection of a given candidate. Of the proposals put forward by François Hollande, that regarding homosexual marriage and adoption was by far the most sensitive to degree of religious integration. The proportion of those who considered it “not desirable” ran from 29 per cent of the non-religious to 79 per cent of Sunday churchgoers. There is no surprise here, of course, for the proposal combined rejection of the Christian conception of the family and acceptance of homosexuality, a practice that is rejected in the traditional Catholic symbolic system. This result might be supposed to stem from the fact that churchgoers are older but such an interpretation is not supported by the data: while it is true that, among those 59 or older, 75 per cent of churchgoers disapproved of the proposal (as compared to 38 per cent of the non-religious), their disapproval was shared by 42 per cent of young churchgoers as well (as compared to 20 per cent of the non-religious).
26The left, by contrast, clearly approved of Hollande’s proposal: among those who considered it both desirable and realistic, 61 per cent voted for a candidate of the left compared to only 23 per cent of those who did not consider it desirable. On the contrary, the more the proposal was rejected, the greater the vote for Sarkozy (from 11 per cent to 35 per cent) and Le Pen (from 10 per cent to 22 per cent). The rejection of this proposal in itself and degree of religious integration both influenced the choice of candidate and their effects were cumulative. Thus, 51 per cent of the practicing Catholics who opposed homosexual marriage [le mariage pour tous] voted for Nicolas Sarkozy while only 7 per cent of non-religious voters who considered it desirable and realistic did so (Table 3). By contrast, 71 per cent of the non-religious voters who most enthusiastically supported this measure opted for a candidate of the left.
“Marriage and adoption for homosexual couples” by religious practice and 1st round voting (in %)
“Marriage and adoption for homosexual couples” by religious practice and 1st round voting (in %)
Understanding the impact of the religious variable on voting: the point of view of socio-demographic differences
27Significant differences relating to levels of integration to Catholicism are therefore to be observed in political behavior. How is one to account for the distance separating the vote of regularly practicing Catholics from that of the non-religious? In what respect are they different? Is it due to their sociological composition? For example, do practicing Catholics more often vote for the right because they are older than the non-religious? Or is it because they include more self-employed people amongst their number? Or is it instead because practicing Catholics and the non-religious participate in very different symbolic systems, as is shown by their systems of attitudes?
28We must first identify the socio-demographic variables that are strongly associated with voting behavior. Moreover, we must determine which of them are also susceptible to varying with degree of Catholic integration. We will then be in a position to confirm that the relationship we observe between religious practice and voting for the right cannot be reduced to some of these socio-demographic variables.
Religious practice by socio-demographic variables (in %)*,**,***,****
Religious practice by socio-demographic variables (in %)*,**,***,****
*PPI = profession of person interviewed. The socio-professional categories identified by the Insee (The French National Institute of Statistics and Economic Studies) are: 1. Agriculteurs exploitants; 2. Artisans, commerçants et chefs d’entreprise; 3. Cadres et professions intellectuelles supérieures; 4. Professions intermédiaires; 5. Employés; and 6. Ouvriers. Although such categories cannot be transposed wholesale into an Anglo-Saxon context, here they are translated as (1) farmers; (2) craftsmen, tradesmen, retailers and company directors; (3) executives and intellectual professions; (4) intermediate professions; (5) white-collar workers [office workers]; and (6) manual workers.** Tax-free up to a certain amount.
*** PATRI6 At least 4 of the 6 types of assets listed
**** PATRI4 More than one type of asset from the following four: Home; second home; securities, stocks and shares; rental property
29According to our data, the two socio-demographic type variables most responsible for variation in rates of religious practice (we will take as our criterion observed differences between regular churchgoers and the non-religious) seem to be age and ownership of assets (patrimoine, meaning property and other types of wealth) (Table 4). For the others, the differences are weaker. Thus, there is little variation by sex or education level and, from the point of view of profession, 31 per cent of regular churchgoers are senior managers compared to only 19 per cent of the non-religious. In regards to income (by CU), 37 per cent of regular churchgoers fall within the upper quartile compared to only 24 per cent of the non-religious.
30It is age that most clearly differentiates the various levels of integration to Catholicism. Frequency of religious practice increases with age: the proportion of those 60 and older increases from 22 per cent among the non-religious to 58 per cent among regularly practicing Catholics. [38] It is sometimes claimed that religious practice tends to increase with the approach of death after a certain biological age. In the case at hand, it is more of a generational issue. For half a century, secularization has been reflected in the gradual weakening of Christian religious involvement and a gradual withdrawal from religious practice among the youngest generations.
31However, voting for the right also becomes more frequent with age: while 33 per cent of 18-34 year olds voted for Nicolas Sarkozy in the second round of the 2012 presidential election, 51 per cent of those over age 60 did so. It must therefore be verified that the relationship between level of religious practice and voting for the right is not due to the fact that seniors are at once the most religious and most conservative portion of the electorate. But this is not the case: the religious practice/voting relationship is in no way affected by the introduction of age. It is just stronger among seniors (Table 6).
2nd round voting for Sarkozy by religious practice and age (in %)
2nd round voting for Sarkozy by religious practice and age (in %)
32One also observes important differences of type and level of asset ownership according to degree of integration to Catholicism. The responses given to a question regarding possession of six asset types [39] reveal that, in the case of four of them, ownership is related to degree of integration to Catholicism (the only exception: savings account ownership). An index of the number of asset types that are owned (Table 4, PATRI6) varies from 17 per cent among the non-religious to 30 per cent among regular churchgoers. It is true that, as with integration to Catholicism, asset ownership tends to increase with age. But the relationship remains at any age (Table 6): it is thus confirmed here that integration to Catholicism is indeed associated with a specific relationship to assets. [40]
More than 3 asset types (/6) according to religious practice and age (in %)
More than 3 asset types (/6) according to religious practice and age (in %)
33The fact that asset ownership clearly differentiates regularly practicing Catholics from the non-religious is not surprising. Indeed, it seems to be at the very heart of the highly affectivized symbolic system characteristic of Catholics. These assets simultaneously consist of a set of religiously structured values of which the individual is the guardian and must pass on in his or her turn – and the material assets inseparable from it, which constitutes its outward sign and foundation: “individual, family, affective relations and spiritual values are integrally related and cannot be separated from the possession and transmission of family wealth”. [41]
34This same system of meanings which appeared in the 1966 interviews was also expressed in those of 1978. [42] An extract from the latter is particularly revealing.
“For people today, the family is the father, the mother and the children before they leave home. In my view, the family is not just an entity in space, it’s an entity in time also. Society is based on the family and society lasts through the centuries; the family must also last through the centuries. We are in solidarity with our ancestors and we have lost this notion in France… And if we abandon what our ancestors left us, I’m not just talking about material goods here, […] then we are traitors […] Let’s now discuss the material point of view, if you like. The family is based on one thing: assets. These assets – and here I’m horribly bourgeois – […] I don’t own them, they’re not even for my use and enjoyment, I manage them. That is, I received them from my ancestors and I am obliged to leave them to my children so that my family – the notion of the family – continues. And that’s tremendously important to me.” (60-year-old man, a retired senior manager and regularly practicing Catholic)
36This conception of assets is very much in the spirit of one of Leo XIII’s encyclicals [43] and is to be found in the Catechism of the Catholic Church:
“The ownership of any property makes its holder a steward of Providence, with the task of making it fruitful and communicating its benefits to others, first of all his family.” [44]
38Proof of the symbolic importance of assets among Catholics can be found in the fact that, while it is true that the proportion of those with significant assets (the upper third) increases with income, it remains the case that, whatever one’s income, that proportion increases with the level of integration to Catholicism. This is more marked among those in the highest income brackets: 50 per cent of regularly practicing Catholics have significant assets compared to only 30 per cent of the non-religious (Table 7).
Significant assets (4-6 types) by religious practice and revenue (in %)
Significant assets (4-6 types) by religious practice and revenue (in %)
39Catholic opposition to the idea of redistribution seems to originate in a fear of asset confiscation – and, as we have just seen, assets occupy a central place in their shared symbolic system. [45] This fear was clearly expressed in the 1978 interviews in response to mention of the round of nationalizations planned under the Common Program, which was seen as potentially resulting in property seizure (for where would the acquisitiveness of the state end?). In these interviews, the respondents’ ambivalence was clear: a Catholic necessarily supports voluntary sharing but must oppose forced sharing harmful to what are seen as the foundations of family life.
40When respondents were asked the degree to which they agreed with the proposition, “To establish social justice, one must take from the rich and give to the poor”, the higher their degree of integration to Catholicism, the more overwhelming their rejection: 59 per cent of weekly churchgoers disagreed with it compared to only 21 per cent among the non-religious. The same held true when the number of asset types they owned increased. All in all, those with the most assets among regularly practicing Catholics were also the most hostile to this proposition (60 per cent) compared to 13 per cent of the non-religious with few assets.
41Of course, it is not surprising that the most integrated Catholics, who were also those with the greatest assets, should be hostile to such a policy. Yet it remains the case that, whatever the extent of respondents’ assets, hostility to redistribution increased with degree of integration to Catholicism. The privileged place occupied by material and symbolic assets in the system of attitudes specific to Catholicism is thereby confirmed (Figure 3).
“Taking from the rich…” (strongly disagree) by religious practice and number of asset types
“Taking from the rich…” (strongly disagree) by religious practice and number of asset types
42Moreover, we know the relationship that exists between asset ownership and voting. In the French case, this relationship has been demonstrated by Jacques Capdevielle and Élisabeth Dupoirier. [46] Since the assets of wealthy individuals are typically diversified, the number of asset types one owns is considered an indicator of level of wealth. The analysis that was thus carried out found that the greater the number of asset types, the more likely one is to vote for the right. [47]
43And that is what we observe here in regards to the first-round voting in the 2012 presidential election (Table 8). The vote for the parliamentary right thus increased from 23 per cent among those who own no asset types to 48 per cent among those who possess at least two (Bayrou: 9 per cent to 14 per cent; Sarkozy: 14 per cent to 34 per cent). By contrast, the score of candidates of the left diminished from 50 per cent among the least well-off to 35 per cent among the most. This result was confirmed in the second round of the election, with the Sarkozy vote moving from 30 per cent to 53 per cent and that of Hollande from 57 per cent to 39 per cent.
The 1st and 2nd round Sarkozy vote by property ownership (in %)
The 1st and 2nd round Sarkozy vote by property ownership (in %)
44As in the case of age, we must thus verify that the religious practice-voting relationship is not due to the effect of asset ownership on the vote (Table 9).
1st round voting by religious practice and asset ownership (in %)
1st round voting by religious practice and asset ownership (in %)
45In the first round, votes for the parliamentary right (François Bayrou and, even more markedly, Nicolas Sarkozy) increased with the degree of integration to Catholicism, whatever the extent of one’s assets. However, while the Sarkozy vote increased in a consistent and significant manner with the number of asset types, this was not the case for the Bayrou vote, which remained stable whatever one’s assets. The relationship was significant but inverted in the case of votes for the left (Jean-Luc Mélenchon and François Hollande). The decision to vote for Marine Le Pen, for its part, depended little on degree of Catholic integration apart from the fact that it was more common among lapsed Catholics. By contrast, it diminished with asset ownership.
46The second round confirmed the role played by assets in the decision of Catholics to vote for the right’s candidate. Whatever one’s assets, the Sarkozy vote increased with the degree of integration to Catholicism (particularly where there were extensive assets). The effects of assets and religion combined to push voter choice rightward. The Sarkozy vote increased from 19 per cent among non-religious voters without assets to 74 per cent among practicing Catholics with significant assets.
2nd round voting by religious practice and number of asset types
2nd round voting by religious practice and number of asset types
47The logistic regression method reinforces these various observations and supplies complementary elements of interpretation. If one takes religious integration and the various socio-demographic affiliations into account in the analysis – and independently of whether or not one introduces number of asset types – sex is not a significant factor in explaining the French 2012 presidential vote. The effect of age, for its part, is equivocal. It should nevertheless be pointed out that 50-64 year olds significantly differed from their elders in voting less often for Nicolas Sarkozy, perhaps thereby retaliating for the controversial decision taken during Sarkozy’s five-year tenure as president to raise the minimum age of retirement with pension. By contrast, the lesser tendency among those aged 18 to 24 to vote for Sarkozy disappears when one takes number of asset types into account: it is thus in fact a consequence of this age group’s still limited wealth.
48In 2012, the sway exerted by Sarkozy’s discourse over French people with little formal education is clear. Only those with more than 2 years of post-secondary education resisted it. And, from the socio-professional point of view, the election was confirmation of the profound transformations that have taken place since the postwar years. The self-employed traditionally prefer the candidate of the right. Above all, however, the voting behavior of employees, workers and intermediate professionals can today no longer be differentiated.
49By contrast, the impact of religious integration on voting, its consistency from one election to the next, and the regularity of banding are remarkable. Thus, once socio-demographic characteristics have been taken into account, a regularly practicing Catholic is 5.8 times more likely than a non-religious voter to cast a ballot for Nicolas Sarkozy than for François Hollande, an occasionally practicing Catholic is 3.7 times more likely to do so, and a non-practicing Catholic 2.4 times more likely to do so. Relative to the variables usually employed to categorize the French population, religion thus has a considerable effect in France. It is completely congruent with the left-right distribution of voting behavior.
Logistic regression for the 2nd Sarkozy vote in 2012 (v11) [48]*,**,***
Logistic regression for the 2nd Sarkozy vote in 2012 (v11) [48]*,**,***
Note: All things being equal – that is, once the effects of sex, age, education, and religion have been excluded – the likelihood of a self-employed voter (farmer, craftsmen, tradesmen, retailers and company directors) voting for Nicolas Sarkozy in the 2nd round of the 2012 presidential election is 2.53 times greater than that of a manual worker.What’s more, if one excludes the effects of assets (since the self-employed often have more diversified – and therefore greater – asset holdings than manual workers), a self-employed worker is 2.28 times more likely to vote for the candidate of the right than is a worker (model 2).
1. ***: significant at threshold 0.001.
**: significant at threshold 0.005.
*: significant at threshold 0.01.
50Above all, it is clear that Catholic voting for the right is not the simple consequence of greater ownership of assets. The propensity to vote for the candidate of the right hardly varies when this variable is taken into account. The only identifiable change concerns practicing Catholics: if greater ownership of assets is taken into account, instead of voting for Sarkozy 5.8 times more often than the non-religious, they now choose him 5.6 times more often.
Psycho-social attitudes: economic liberalism and ethno-authoritarianism
51In considering what might explain the constant relationship that has until now been observed between level of integration to Catholicism and conservative voting propensity, we hypothesized that responsibility may lie with certain variables that differentiate Catholics from the non-religious and also encourage voting for the right. Our analysis showed that, while some socio-demographic variables are indeed simultaneously related to voting for the right and degree of integration to Catholicism, they do not erase the political influence of religious practice. At most, their effects combined with and reinforced those of the religious variable, maximizing the vote for Sarkozy.
52Since the relationship between Catholic integration and voting for the right does not seem to be explained as a masked effect of socio-demographic variables, we have taken an analogous approach in order to explore the universe of attitudes and single out those that may have an influence on political behavior. Moreover, the extent to which some of these variables are present in the Catholic symbolic system must be identified. As we have shown for socio-demographic variables, we will then be able to determine whether taking these attitudes into account tends to nullify the effect of integration to Catholicism on conservative political behaviors, which would suggest that it is these attitudes that are in fact truly responsible for the observed relationship.
53Originating in a survey carried out on the occasion of the 1988 elections, L’électeur français en questions [49] presents ten attitude scales to account for the universe of socio-political attitudes. [50] Eight of them can be grouped into two major categories: socio-economic attitudes [51] and ethico-cultural ones. [52] Four of the contributions drew upon them. [53]
54Since then, the fundamental role of these two groups of attitudes in structuring the political field has been illustrated and confirmed on multiple occasions. [54] This is why we have chosen to construct tools to measure these two dimensions: one for economic liberalism and the other for ethno-authoritarianism. [55]
Differences of attitude and voting
Voting and the economic liberalism scale
55Voter choice is directly related to level of economic liberalism (Figure 5). Higher levels of economic liberalism are accompanied by increased voting for the parliamentary right: from 5 per cent to 52 per cent for the Sarkozy vote and from 6 per cent to 13 per cent for the Bayrou vote. At the same time, the propensity to vote for the left diminishes (Mélenchon from 23 per cent to 3 per cent, Hollande from 35 per cent to 10 per cent). The Le Pen vote, however, bears little relationship to this attitude.
1st and 2nd round voting on the economic liberalism scale
1st and 2nd round voting on the economic liberalism scale
56Altogether, 57 per cent of the most economically liberal voters (the upper quarter of the population) voted for Nicolas Sarkozy and 31 per cent voted for François Bayrou. Only 9 per cent voted for François Hollande and 7 per cent for Jean-Luc Mélenchon.
Voting and the ethno-authoritarian scale
57Voting behavior also very clearly depends on the ethno-authoritarianism that on the whole divides right and left. Thus, in the first round, the Sarkozy vote increased from 8 per cent among the not very authoritarian to 35 per cent among the most authoritarian (Figure 6). A similar but more marked tendency is encountered in the Le Pen vote, which increases from 1 per cent to 7 per cent. At the same time, votes for Hollande evolved in the opposite direction – from 49 per cent to only 10 per cent – while those for Mélenchon diminished from 18 per cent to 3 per cent. The Bayrou vote occupied an intermediary position.
58Altogether, 83 per cent of Le Pen voters and 56 per cent of Sarkozy voters are located in the upper third of the ethno-authoritarianism scale compared with just 13 per cent of Hollande voters (Mélenchon, 10 per cent; Bayrou, 24 per cent). The second round confirmed these results: 66 per cent of the most ethno-authoritarian chose Sarkozy versus 21 per cent who chose Hollande.
1st and 2nd round voting on the ethno-authoritarian scale
1st and 2nd round voting on the ethno-authoritarian scale
Voting as a combination of the two attitudes: economic liberalism and ethno-authoritarianism
59Not only did attitudes of economic liberalism and ethno-authoritarianism encourage voting for Sarkozy; their effects compounded and added to one another as well. Thus, this vote increased from 0 per cent among voters who were neither economically liberal nor ethno-authoritarian to 85 per cent among those located at the highest level of both scales (Table 15 and Figure 7).
Level of integration to Catholicism and attitudes: ethno-authoritarianism, economic liberalism
60Given the importance of the relationship that combines each of these two attitudes in orienting voting behavior, one may wonder whether they do not themselves vary according to degree of integration to Catholicism. Indeed, the Catholic Church holds that its mission requires it to insist upon its positions, whether in regards to ethnocentrism, authoritarianism, or economic liberalism. It therefore intervenes in public debate, often forcefully. Yet one should not conclude from this that the positions of a church are directly and completely reflected in the opinions and behaviors of its congregation. There can be significant discrepancies between the two, as the cases of family morality and sexuality would seem to suggest.
2nd round Sarkozy vote, scales of economic liberalism and ethno-authoritarianism combined
2nd round Sarkozy vote, scales of economic liberalism and ethno-authoritarianism combined
Integration to Catholicism and economic liberalism
61Figure 8 shows that economic liberalism regularly decreases as degree of integration to Catholicism diminishes. Thus, the proportion of the most economically liberal voters (26 per cent of the entire population) among the non-religious (19 per cent) is less than half that found among regularly practicing Catholics (46 per cent).
Scales of economic liberalism (++) and ethno-authoritarianism (+)
Scales of economic liberalism (++) and ethno-authoritarianism (+)
62This figure only displays the extreme degrees (+) of the scales, corresponding to the upper third of ethno-authoritarians (ETHNAUT+) and a quarter of economic liberals (ECLIB++).
63Although religious practice differs by generation (Table 11), we have verified that these results do not depend on age.
Economic liberalism (ECLIB++) and ethno-authoritarianism (ETHNAUT+) according to religious practice and age (in %)
Economic liberalism (ECLIB++) and ethno-authoritarianism (ETHNAUT+) according to religious practice and age (in %)
64Is there anything surprising about these results relative to the pronouncements of the Catholic Church? Ever since Leo XIII’s Rerum novarum encyclical of 1891 [56] and, more recently, Jean-Paul II’s definition of the “social doctrine of the Church” in the Centesimus annus, it has been clear that “the Church acknowledges the legitimate role of profit as an indication that a business is functioning well” and that its teaching “recognizes the positive value of the market and of enterprise but at the same time emphasizes that these need to be oriented towards the common good”. While capitalism, which is explicitly referred to in the Centesimus annus, is seen to possess a commendable aspect, one must not forget the harmful effects of some of its forms.
In 2013, by contrast, Pope Francis no longer spoke of the positive aspects of capitalism or, for that matter, referred to it by name. On the contrary, he condemned an economy of exclusion and the social disparities that are produced by “the idolatry of money and the dictatorship of an impersonal economy […] such an economy kills”. [58] The future will tell whether this change in the Catholic Church’s discourse will have a bearing upon its followers’ attitudes and behaviors.“Can it perhaps be said that, after the failure of communism, capitalism is the victorious social system, and that capitalism should be the goal of the countries now making efforts to rebuild their economy and society? Is this the model which ought to be proposed to the countries of the Third World which are searching for the path to true economic and civil progress? […] If by ‘capitalism’ is meant an economic system which recognizes the fundamental and positive role of business, the market, private property, and the resulting responsibility for the means of production, as well as free human creativity in the economic sector, then the answer is certainly in the affirmative […] But if by ‘capitalism’ is meant a system in which freedom in the economic sector is not circumscribed within a strong juridical framework which places it at the service of human freedom in its totality, and which sees it as a particular aspect of that freedom, the core of which is ethical and religious, then the reply is certainly negative.” [57]
Integration to Catholicism and ethno-authoritarianism
65While the relationship between economic liberalism and integration to Catholicism is straightforward (the proportion of very economically liberal voters regularly increases from the non-religious to regularly practicing Catholics, see Figure 8), the same does not seem to entirely hold for ethno-authoritarianism. The latter is much more widespread among Catholics (43 per cent) than among the non-religious (29 per cent). Among Catholics, regular churchgoers are the least ethno-authoritarian (36 per cent), in contrast to occasional churchgoers (46 per cent) and non-churchgoers (42 per cent). As in the case of economic liberalism, we have verified that age has no effect on this relationship (Table 15).
66By contrast, if one simultaneously takes both of these attitudes into account, one perceives that, as Figure 9 clearly shows, ethno-authoritarianism does indeed tend to increase with degree of integration to Catholicism but only among the most economically liberal voters.
67Among the most integrated Catholics and the non-religious alike, the existence of systems of values organized around the acceptance of the Other and welcoming foreigners may explain the relatively low scores of ethno-authoritarianism observed at the two extremes of religious integration. All in all, it is lapsed Catholics – that is, those least receptive to the Church’s pronouncements – who prove the most ethno-authoritarian.
Ethno-authoritarianism (ETHNAUT+) of the most (scale scores 4 to 6) or least economically liberal voters (scores 0 to 3) according to religious practice (in %)
Ethno-authoritarianism (ETHNAUT+) of the most (scale scores 4 to 6) or least economically liberal voters (scores 0 to 3) according to religious practice (in %)
68There is nothing ambiguous about the Church’s support for respecting and welcoming immigrants. It is directly inspired by the Old Testament:
“The stranger that dwelleth with you shall be unto you as one born among you, and thou shalt love him as thyself; for ye were strangers in the land of Egypt.” [59]
70Jean-Paul II is clear on this point:
“For the Christian, acceptance of and solidarity with the stranger are not only a human duty of hospitality, but a precise demand of fidelity itself to Christ’s teaching.” [60]
72In 2013, Pope Francis specified:
“We Christians should embrace with affection and respect Muslim immigrants to our countries in the same way that we hope and ask to be received and respected in countries of Islamic tradition.”
74As for the death penalty, in his 1998 Christmas message, Jean-Paul II stated that it should be prohibited. As far as the non-religious are concerned, their weak rejection of immigrants and opposition to the death penalty seems related to the fact that most of them are to be found on the left (64 per cent), where these positions are more appreciated and widespread, both in party programs and among individuals.
75Taken together, these results lead one to wonder whether the two attitudes in question do not combine with one another in a specific fashion according to degree of integration to Catholicism. Indeed, this is what one observes in Table 12.
ECLIB/ETHNAUT combinations according to religious practice
ECLIB/ETHNAUT combinations according to religious practice
The number of those in certain categories makes it necessary to reduce them to dichotomies. Thus, in the scale of economic liberalism, LIBECR- (scores 0-3) contrasts with LIBECR+ (scores 4-6, 55% of the population). In the ethno-authoritarian scale, ETHNOR- (scores 0-4) contrasts with ETHNOR+ (scores 5-9, 52% of the population).76The differences between the attitude combinations (economic liberalism and ethno-authoritarianism) characteristic of degrees of integration to Catholicism appear very significant: at the two extremes, regularly practicing Catholics stand in very marked contrast to the non-religious. Of them all, the former are at once the most economically liberal and the most ethno-authoritarian: 43 per cent versus 21 per cent among the non-religious. The latter radically stand out since 41 per cent of them are, on the contrary, both the least economically liberal and the least ethno-authoritarian versus 17 per cent of regular churchgoers.
77Since attitudes of ethno-authoritarianism and economic liberalism are related (however unequally) to integration to Catholicism and since voter choice seems to depend on these two scales, it remains to be determined whether voting for the right is primarily explained by level of integration to Catholicism or whether it is these attitudes that are directly responsible for it.
Voting according to degree of integration to Catholicism and psycho-social attitudes
Voting according to economic liberalism and religious practice
78Whatever the degree of economic liberalism, the vote for the parliamentary right significantly increases with the level of integration to Catholicism (Figure 9). However, it also increases with economic liberalism. This is the opposite of what one observes in the case of votes for the left. The effects of Catholic integration and economic liberalism combine to maximize the votes obtained by the parliamentary right. These increase from just 7 per cent among the non-religious/non-economically liberal to 93 per cent among very economically liberal regular churchgoers. At the same time, the electoral left falls from 78 per cent to zero. Finally, we have showed that degree of economic liberalism has little influence on the propensity to vote for Marine Le Pen. It seems that, while an increased level of economic liberalism nevertheless encourages the non-religious to vote for this candidate, it tends to put off Catholics, as if the pronouncements of the Church had some effect. Finally, the second round confirmed the importance of the impact of each of these variables on voting but also its spectacular reinforcement effect, increasing the Sarkozy vote from 7 per cent among the non-economically liberal non-religious to 81 per cent among very economically liberal practicing Catholics (Figure 10). Among the same voters, the Hollande vote diminished from 82 per cent to 13 per cent.
2nd round voting according to religious practice and economic liberalism scale
2nd round voting according to religious practice and economic liberalism scale
Voting according to ethno-authoritarianism and religious practice
79As for economic liberalism, introducing the attitude does not cancel the effect of level of Catholic integration on the Sarkozy vote. The latter increases both with frequency of religious practice and with degree of ethno-authoritarianism. At the very most, ethno-authoritarianism contributes to maximizing it.
1st round voting according to religious practice and ethno-authoritarianism scale (ETHNAUT) (in %)
1st round voting according to religious practice and ethno-authoritarianism scale (ETHNAUT) (in %)
80By contrast, the Bayrou vote clearly obeys different laws: whatever the level of religious integration, it diminishes when ethno-authoritarianism increases, as if in allergic reaction to the refusal of the Other. The effect of integration to Catholicism only remains among those who are not ethno-authoritarian. The result is that the Bayrou vote reaches its highest level among the least ethnocentric regular churchgoers – that is, among Catholics who take the Christian message seriously. It reaches its lowest level among non-ethnocentric, non-religious voters for whose vote the left competes.
81The Le Pen vote here appears as a combination of leftwing and rightwing motivations. Like the Sarkozy vote, it increases with ethno-authoritarianism but, like that of the left, it diminishes with level of Catholic integration. Its reaches its highest level among ethno-authoritarian, non-religious voters.
82The second round confirms our observations of the first. Despite its attributable effect, the introduction of ethno-authoritarianism does not abolish the effect of level of integration to Catholicism on the Sarkozy vote. The latter thus rises from 8 per cent among non-ethno-authoritarian, non-religious voters to 89 per cent among ethno-authoritarian regular churchgoers. For the Hollande vote, the opposite is true: among the same groups, this vote varies from 86 per cent to 7 per cent (Figure 11).
2nd round voting according to religious practice and ethno-authoritarianism scale (ETHNAUT)
2nd round voting according to religious practice and ethno-authoritarianism scale (ETHNAUT)
83We have already observed that, when one combines the two attitudes, regular churchgoers and the non-religious are total opposites: where the former are the most economically liberal and most ethno-authoritarian, the latter exhibit the inverse characteristics. One sees the consequences of this in voting.
84Let us begin by noting that, among our respondents, the most economically liberal and most ethno-authoritarian mainly voted for the parliamentary right (67 per cent). Those with the opposite characteristics – neither very economically liberal nor very ethno-authoritarian – overwhelmingly voted (91 per cent) for the left (Table 14). Not very economically liberal but very ethno-authoritarian voters tended to cast their ballots for Marine Le Pen.
85These results are valid irrespective of degree of integration to Catholicism (Table 15). Reciprocally, the ordinal relationship associated with level of religious practice is confirmed irrespective of the combination of attitudes. Thus, among economically liberal/ethno-authoritarian voters, voting for the parliamentary right rises from 50 per cent to 72 per cent, depending on level of religious practice. Similarly, among not very ethno-authoritarian/not very economically liberal voters, voting for the left rises from 56 per cent among regular churchgoers to 86 per cent among the non-religious.
86The Le Pen vote, for its part, is only frequently encountered among the ethno-authoritarian. In their case, however, integration to Catholicism seems to somewhat attenuate the effect of their attitudes.
87Second-round voting for the candidate of the right increases with degree of integration to Catholicism irrespective of degree of economic liberalism and ethno-authoritarianism, although both encourage this vote. Figure 12 clearly shows that, depending on the level of integration to Catholicism, the curves of the votes obtained by Sarkozy exhibit exactly the same form for all combinations of attitude: they differ only in terms of the proportion of votes received. Altogether, a difference of 75 percentage points separates the Sarkozy vote among economically liberal, ethno-authoritarian practicing Catholics (79 per cent) from non-economically liberal, non-authoritarian, non-religious voters (4 per cent).
88To respond to the question we posed above, one might say that Catholics do in fact seem to vote for the right because they are Catholic. Indeed, we have seen that, the greater one’s integration to this group, the greater the probability that one shares with other members the symbolic system, attitudes and characteristics specific to it.
2nd round Sarkozy vote by combinations of economic liberalism / ethno-authoritarianism, and religious practice
2nd round Sarkozy vote by combinations of economic liberalism / ethno-authoritarianism, and religious practice
1st round voting according to economic liberalism (ECLIB) and ethno-authoritarianism (ETHNAUT) (in %)
1st round voting according to economic liberalism (ECLIB) and ethno-authoritarianism (ETHNAUT) (in %)
1st round voting according to economic liberalism (LIB), ethno-authoritarianism (ETH), and religious practice (in %)
1st round voting according to economic liberalism (LIB), ethno-authoritarianism (ETH), and religious practice (in %)
89As we have seen, however, these attitudes – economic liberalism and ethno-authoritarianism – do indeed seem to be part of the symbolic system specific to Catholics and increase the likelihood of locating oneself on and voting for the right.
90* * *
91On the basis of this analysis, it is clear that the degree of religious integration to Catholicism continues to play a very important role in voting behavior. Of course, the proportion of French people who refer to themselves as Catholic continues to diminish (from 85 per cent in 1966 to around 60 per cent today), with the most integrated Catholics’ share of the population – those who most regularly attend church – halved over the same period (from 20 per cent to 10 per cent). Yet, despite the enormous weakening of the “Catholic grip” on France, it remains the case that, “the more one is Catholic” – that is, the more one is integrated into the Catholic symbolic system – the higher the probability that one will be situated on the right of the political spectrum and share its electoral choices. Indeed, beyond political behavior properly so-called, it is to be noted that 43 per cent of regularly practicing Catholics are located on the right of the left-right dimension (versus 16 per cent of the non-religious) and only 13 per cent on the left (versus 48 per cent of the non-religious). [61]
92How is one to explain the continuity of this relationship? The affinity of the Catholic Church for conservatism was borne of bygone historical conditions. Yet their effects continue to manifest themselves in successive elections. With the advent of a new set of issues, the possibility of political changeover after more than fifteen years of center-right presidency, greater personalization of the candidates, and the spread of new means of political communication, the 2012 presidential election could have marked a break with the past. But this was not to be: the election once again confirmed that voting for the parliamentary right increases with frequency of church attendance. What variables account for this difference in voting behavior between the most integrated Catholics and the non-religious? Moreover, in what ways are they different?
93We have attempted to determine what variables might be related to both voting for the right and degree of integration to Catholicism. Initially, we thought they might consist of certain differences of a socio-demographic type. The most observant Catholics are in fact older, more often senior managers or members of the liberal professions, and have a higher income and more assets. These variables are related to a high likelihood of voting for the right. Yet our analysis shows that, after having taken into account the specific effect of these variables on voting, there are still significant differences in political behavior relating to the degree of integration to Catholicism.
94We then hypothesized that practicing Catholics and the non-religious could be characterized by particular systems of attitudes, or the very different symbolic systems that predispose them to specific votes. We have already noted that the most observant Catholics are more often asset holders. We saw that ownership of assets was directly related to a representation of the family that occupies a central place in the Catholic symbolic system. Moreover, we found that economic liberalism and ethno-authoritarianism, which accompany conservative attitudes and political behaviors, also go together with an increased degree of integration to Catholicism. Since they tend to be more economically liberal and ethno-authoritarian, it seems obvious that Catholics should more often vote for the right than do the non-religious. This is explained by major differences of symbolic system, but Catholics’ rich symbolic system cannot be reduced to these two attitudes. Indeed, it can be confirmed that, while voting for the right increases with frequency of churchgoing, this is fundamentally due to level of integration within the Catholic symbolic system itself, since one finds its influence irrespective of degree of economic liberalism and ethno-authoritarianism.
95Since the elections, the evolution of the political situation seems to further confirm our conclusions. Here we are thinking in particular of the anti-homosexual marriage demonstrations and all that they might mean. The “manifs pour tous” [62] sought to rise up against what was seen as an attack on the traditional and, more particularly, Catholic conception of the family. We have seen that the family is located at the heart of what we have called the Catholic symbolic system and great importance is attributed to its continuity as well as the wealth of its values (both spiritual and material) that it seeks to pass on.
97However, the Taubira law [63] struck at the very foundation of the traditional family, consisting of father, mother and children. This is why the objective of the “Manif pour tous” was to bring about a sort of Grenelle de la famille, [64] the official drafting of a charter for the defense of family values and a framework law to aid in combatting the bill expected by the government. According to a February 2014 poster, “angry families” called for “No More Familyphobia! Parentage, Gender, Taxes!’
98Attacks on the traditional Catholic conception of the family take various forms, all centered on the idea of a violation of natural law. The movement’s point of departure was a refusal of “homosexual marriage”: the condemnation of homosexuality is a constant of Catholicism (as in many religions).
« Homosexual acts are intrinsically disordered. » [65]
100Thus, although acceptance of homosexuality has made significant inroads throughout society, in December 2013 only 15 per cent of regularly practicing Catholics [66] found it “very acceptable” versus 62 per cent of the non-religious. To this was added a rejection of abortion, euthanasia, medically assisted pregnancy, surrogate pregnancy and, finally, “gender theory”. Le mariage pour tous “is a rupture in society,” declared Cardinal Barbarin. “Next, they will want to have couples with three or four people. Then, someday perhaps, the ban on incest will be dropped.” [67] Today’s debates seem to clearly indicate that political opposition to the government has led to the awakening of significant attitudes present among Catholics.
Content of the scales
Economic liberalism scale (ECLIBEC) [68]
101The composition of this scale seems to correspond to what is generally meant by “economic liberalism” by associating the words “profit”, “privatization”, “capitalism” and fear of a tax on profits with “liberalism” and “market economy”:
- A 75 per cent tax rate on income above 1 million Euros (not desirable (20 per cent)/desirable but not realistic, desirable and realistic).
102For you, do each of the following words evoke something very positive, somewhat positive, somewhat negative or very negative?
- Capitalism (very, somewhat positive (30 per cent)/somewhat, very negative, NR [69]);
- Profit (very, somewhat positive (39 per cent)/somewhat, very negative, NR);
- Liberalism (very, somewhat positive, somewhat negative (84 per cent)/very negative, NR);
- Privatization (very, somewhat positive, somewhat negative (85 per cent)/very negative, NR);
- Market Economy (very, somewhat positive, somewhat negative (90 per cent)/very negative, NR).
Ethno-authoritarian scale (ETHNAUT) [70]
103This scale unites two aspects, the first characterized by fear and rejection of the Other (mainly represented by immigrants and the feeling of no longer being at home as before) and a second, authoritarian aspect, which can, for example, lead one to call the abolition of the death penalty into question.
104Imagine that things could be better. Among the following items, a better world would be a world with more of what in your view? Authority (6 per cent).
105Among the following questions, which three seem to you the most worrisome for France today? Immigration (21 per cent).
106Are you very favorable, somewhat favorable, somewhat opposed or very opposed to regularizing undocumented immigrants on a case-by-case basis depending on their situation (employment, housing, length of time they have resided on French territory)? (very opposed (33 per cent)/somewhat opposed, somewhat, very favorable, NR).
107The death penalty should be re-established (completely, somewhat agree (38 per cent)/somewhat, strongly disagree?).
108I no longer feel at home like before (completely, somewhat agree (47 per cent)/somewhat, strongly disagree?).
109The right to vote in local elections for non-European Union foreigners who have legally resided in France for 5 years (not desirable (56 per cent)/desirable and realistic, desirable but not realistic, NR).
110Does each of the following words evoke for you something very positive, somewhat positive, somewhat negative or very negative? Immigrants (very, somewhat negative (70 per cent)/somewhat, very positive, NR).
111Reducing by half the number of immigrants received each year by France (desirable and realistic, desirable but not realistic (82 per cent)/not desirable, NR).
112There are too many immigrants in France (completely, somewhat agree, somewhat disagree (86 per cent)/completely disagree, NR).
Notes
-
[1]
In what follows, we will use a scale of integration to Catholicism in which the lowest level consists of those without religion. Indeed, the originality of the Catholic symbolic system can only be understood by contrasting it with the system characteristic of those of our fellow citizens who reject all religious affiliation.
-
[2]
“What is your religion, if you have one?” “If yes, which?” We will define as Catholics and those without religion those who gave one of these two answers to the first question. Among the Catholics, we will distinguish between those who attend mass: every Sunday (Sunday churchgoers: the “messalisants” of Fernand Boulard); at least once a month (monthly churchgoers), occasionally, for major holidays (occasional churchgoers); never or only for ceremonies, marriages, burials, etc. (non-practicing). Today, in most survey-based publications, it has become habitual to refer to those who attend mass “at least once a month” as “regular churchgoers”. This is why, for the purposes of this article, we have chosen to distinguish between “Sunday churchgoers” and “monthly churchgoers”.
-
[3]
Charles Glock, “Y a-t-il un réveil religieux aux États-Unis?”, Archives de sociologie des religions, 12, 1961, 35-52.
-
[4]
Guy Michelat, “L’identité catholique des Français. I. Les dimensions de la religiosité. II. Appartenances et socialisation religieuse”, Revue française de sociologie, 31(3 & 4), 1990, 355-88 and 609-33; “Les indicateurs du niveau d’intégration religieuse”, in Élisabeth Dupoirier, Jean-Luc Parodi (eds), Les indicateurs socio-politiques aujourd’hui (Paris: L’Harmattan, 1997), 185-99.
-
[5]
André Siegfried, Tableau politique de la France de l’Ouest sous la IIIe République (Paris: Imprimerie nationale, 1995 [1st edn: 1913]).
-
[6]
Mattei Dogan, Daniel Derivry, “Unité d’analyse et espace de référence en écologie électorale”, Revue française de science politique, 21(3), 1971, 517-70.Online
-
[7]
Cantonal-level survey of the religious practices of Catholics in most dioceses. Cf. Fernand Boulard, Premiers itinéraires en sociologie religieuse (Paris: Les Éditions ouvrières, 1966).
-
[8]
François Isambert, Jean-Paul Terrenoire, Atlas de la pratique religieuse des catholiques français (Paris: Presses de Sciences Po / Éditions du CNRS, 1980).
-
[9]
Michel Brulé, “L’appartenance religieuse et le vote du 5 décembre 1965”, Sondages, 28(2), 1966, 15-19.
-
[10]
Évelyne Aver, Constant Hamés, Jacques Maître, Guy Michelat, “Pratique religieuse et comportement électoral à travers les sondages d’opinion”, Archives de sociologie des religions, 29, 1970, 26-52.
-
[11]
Guy Michelat, Michel Simon, Classe, religion et comportement politique (Paris: Presses de Sciences Po/Éditions sociales, 1977).
-
[12]
Daniel Boy, Nonna Mayer, “Que reste-t-il des variables lourdes?”, in Daniel Boy, Nonna Mayer (eds), L’électeur a ses raisons (Paris: Presses de Sciences Po, 1997), 101-38.
-
[13]
Guy Michelat, “Intégration au catholicisme, attitudes éthico-politiques et comportement électoral”, in Pierre Brechon, Annie Laurent, Pascale Perrineau (eds), Les cultures politiques des Français (Paris: Presses de Sciences Po, 2000), 209-39.
-
[14]
Claude Dargent, “La religion, encore et toujours”, in Bruno Cautrès, Nonna Mayer (eds), Le nouveau désordre électoral (Paris: Presses de Sciences Po, 2004), 161-83.
-
[15]
Bruno Cautrès, Christine Pina, Kristoff Talin, “L’influence de la religion sur les attitudes politiques en Europe”, in Pierre Bréchon, Bruno Cautrès (eds), Les enquêtes Eurobaromètres. Analyse comparée des données socio-politiques (Paris: L’Harmattan, 1998), 269-86.
-
[16]
Pierre Bréchon, “Les valeurs politiques en Europe: effet du contexte national et des attitudes religieuses”, Archives de sciences sociales des religions, 93, January-March, 1996, 99-128.
-
[17]
Pierre Bréchon, “Influence de l’intégration religieuse sur les attitudes: analyse comparative européenne”, Revue française de sociologie, 43(3), 2002, 461-83.Online
-
[18]
Pierre Bréchon, “Valeurs de gauche, valeurs de droite et identités religieuses en Europe”, Revue française de sociologie, 47(4), October-December 2006, 725-53.
-
[19]
Claude Dargent, “La fille aînée de l’Église dans l’isoloir: religion et politique en France et en Europe”, in Daniel Boy, Bruno Cautrès, Nicolas Sauger (eds), Les Français sont-ils des Européens? (Paris: Presses de Sciences Po, 2010), 181-213.
-
[20]
There are obviously exceptions to this rule. Thus, in Great Britain, Catholics – a very small and long marginalized population – are as likely to be leftwing as rightwing, whatever their level of churchgoing. By contrast, among those affiliated with Protestant churches, which are traditionally associated with the political establishment, regular churchgoers are more rightwing than non-churchgoers. (P. Bréchon, “Valeurs de gauche…”).
-
[21]
Paul Lazarfeld, Bernard Berelson, Hazel Gaudet, The People’s Choice (New York: Columbia University Press, 1944).
-
[22]
Jeff Manza, Clem Brooks, “The religious factor in US presidential elections, 1960-1992”, American Journal of Sociology, 103 (1), 1997, 39-81. In 2004, according to CNN.com Election Results, the vote for Bush increased from 36 per cent among those who never attend church or temple to 45 per cent (a few times per year), 50 per cent (once a month), 58 per cent (once a week) and 64 per cent (more than once a week).
-
[23]
Claude Dargent, “Les musulmans déclarés en France. Affirmation religieuse, subordination sociale et progressisme politique”, Cahiers du Cevipof, 34, 2003; Sylvain Brouard, Vincent Tiberj, “L’incorporation politique ‘à la française’: modèles explicatifs des alignements politiques des Français d’origine maghrébine, africaine et turque”, in “Citoyens et enfants de l’immigration”, Migrations Société, 19 (113), September-October 2007, 127-48.
-
[24]
René Rémond, James M. Laux (trans.), The Right Wing in France: From 1815 to De Gaulle (Philadelphia: University of Pennsylvania Press, 1969), 342.
-
[25]
In order to estimate the size of the Catholic population, we use the question, “What is your religion, if you have one?” followed by a list of religions ending with “no religion” (cf. note 1). In general, in international surveys, the question, “do you consider yourself as belonging to a religion?” precedes the list of religions. When the notion of belonging is specified, the estimation of the proportion of Catholics is around 15 points lower.
-
[26]
Claude Dargent, “La population musulmane de France: de l’ombre à la lumière?”, Revue française de sociologie, 51(2), April-June 2010, 219-46.
-
[27]
Danièle Hervieu-Léger, Le Pèlerin et le Converti. La religion en mouvement (Paris: Flammarion, 1999), and Catholicisme. La fin d’un monde (Paris: Bayard, 2003).
-
[28]
Grace Davie, La religion des Britanniques de 1945 à nos jours (Geneva: Labor et Fides, 1996).
-
[29]
Grace Davie, Religion in Modern Europe: A Memory Mutates (Oxford: Oxford University Press, 2000), and “Vicarious religion: a methodological challenge”, in Nancy T. Ammerman (ed.), Everyday Religion: Observing Modern Religious Lives (Oxford: Oxford University Press, 2006), 21-36.
-
[30]
Yann Raison du Cleuziou, Qui sont les cathos aujourd’hui? (Paris: Desclée de Brouwer, 2014).
-
[31]
Jean-Marie Donegani, La liberté de choisir (Paris: Presses de Sciences Po, 1993).
-
[32]
On the intervention of the Catholic clergy in regards to voting over the past two centuries, see Yves Déloye, Les voix de Dieu. Pour une autre histoire du suffrage électoral: le clergé catholique français et le vote, XIXe-XXe siècles (Paris: Fayard, 2006).
-
[33]
Jean-Marie Donegani, “Religion et politique”, in Pascal Perrineau, Luc Rouban (eds), La politique en France et en Europe (Paris: Presses de Sciences Po, 2007), 61-94 (72-3).
-
[34]
Daniel Boy, Jean Chiche, “L’image des candidats dans la décision électorale”, Revue française de science politique, 57(3-4), 2007, 329-42.
-
[35]
Gérard Grunberg, Étienne Schweisguth, “La tripartition de l’espace politique”, in Pascal Perrineau, Colette Ysmal (eds), Le vote de tous les refus (Paris: Presses de Sciences Po, 2003), 341-62.
-
[36]
In the course of the electoral campaign, the Sciences Po Centre de recherches politiques (Cevipof), Le Monde newspaper, the Fondation pour l’innovation politique (Fondapol) and the Fondation Jean-Jaures joined forces with Ipsos/Logica Business Consulting to question a panel of 6000 registered voters representative of the French population over the age of 18. This panel was entitled “Présidoscope 2012” and surveyed these voters on twelve occasions between November 2011 and June 2012. 2422 voters responded to all twelve surveys. Some questions were only asked in some surveys. Initial analyses of the evolution of voting intentions leading to the choice of a candidate were published in 2013 (Pascal Perrineau (ed.), Le vote normal. Les élections présidentielle et législatives d’avril-mai-juin 2012 (Paris: Presses de Sciences Po, 2013)). From this data, we have extracted the set of individuals questioned in surveys 1, 2, 7, 9, 10 and 11 in order to gather responses to the questions that were only asked in certain surveys. The total sample size is 2911. We have confirmed that, over the course of the twelve surveys, the sample experienced only very limited attrition.
-
[37]
Pierre Bréchon, “Un nouveau centrisme électoral?”, in Pascal Perrineau (ed.), Le vote de rupture. Les élections présidentielle et législatives d’avril-juin 2007 (Paris: Presses de Sciences Po, 2008), 175-95 (183-4); on vote transfers under this relationship between 2002 and 2007, see also Sylvie Strudel, “La tentation bayrouiste”, in Bruno Cautrès, Anne Muxel (eds), Comment les électeurs font-ils leur choix? (Paris: Presses de Sciences Po, 2009), 221-37.
-
[38]
In 1978, the corresponding proportions were 12 per cent among the non-religious and 38 per cent among regular churchgoers (Cevipof-Sofres poll).
-
[39]
Your main residence; a secondary residence; securities, stocks, mutual funds; a livret A savings account; another livret or savings account (LDD, yellow livret, home savings plans, etc.); rental property (apartment, house, land, etc.).
-
[40]
This recalls the results from the 1978 Cevipof post-election survey data. Cf. Guy Michelat, Michel Simon, “Déterminations socio-économiques, organisations symboliques et comportement électoral”, Revue française de sociologie, 26(1), 1985, 32-69.
-
[41]
Guy Michelat, Michel Simon, “Catholiques déclarés et irréligieux communisants: vision du monde et perception du champ politique”, Archives de sciences sociales des religions, 35(1), 1973, reprinted in Classe, religion…, and in “Déterminations socio-économiques…”.
-
[42]
Jean-Marie Donegani, Guy Michelat, Michel Simon, Représentations du champ social, attitudes politiques et changements socio-économiques (Lille/Paris: Institut de sociologie Lille I / Centre d’étude de la vie politique française contemporaine (CNRS-FNSP) 1980).
-
[43]
“It is a most sacred law of nature that a father should provide food and all necessaries for those whom he has begotten; and, similarly, it is natural that he should wish that his children, who carry on, so to speak, and continue his personality, should be by him provided with all that is needful to enable them to keep themselves decently from want and misery amid the uncertainties of this mortal life” (Leon XIII, encyclical Rerum novarum).
-
[44]
Catechism of the Catholic Church, para. 2404.
-
[45]
And this even if many passages of the Gospel express a radical critique of the effects of wealth. Of these, the best known is from Luke: “For it is easier for a camel to go through a needle’s eye, than for a rich man to enter into the kingdom of God” (Luke, 18.25).
-
[46]
Jacques Capdevielle, Élisabeth Dupoirier, “L’éffet patrimoine”, in Jacques Capdevielle et al., France de gauche, vote à droite (Paris: Presses de Sciences Po, 1981), 171-355.
-
[47]
This relationship between asset ownership and voting was subsequently confirmed for the United Kingdom: Monica Charlot, “Un effet patrimoine?”, in Monica Charlot (ed.), L’effet Thatcher (Paris: Economica, 1989), 33-51; David E. Butler, Dennis Kavanagh, The British General Election of 1983 (London: Macmillan, 1985). Following a curiously long period of indifference, it was once again observed in France during the 2007 presidential election: Viviane Le Hay, Mariette Sineau, “‘Effet patrimoine’: 30 ans après, le retour?”, Revue française de science politique, 60(5), October 2010, 869-900; Martial Foucault, Richard Nadeau, Michael S. Lewis-Beck, “La persistance de l’effet patrimoine lors des élections présidentielles françaises”, Revue française de science politique, 61(4), 2011, 659-80.
-
[48]
BEPC, National diploma awarded at the end of ninth grade. CAP, Certificat d’aptitude professionnelle: Certificate of professional aptitude, awarded at the completion of technical and professional studies. BEP, Brevet d’études professionnelles: Certificate of professional studies, required for a professional baccalaureate.
-
[49]
Daniel Boy, Nonna Mayer (eds), L’électeur français en questions (Paris: Presses de Sciences Po, 1990).
-
[50]
Guy Michelat, “Les échelles d’attitude et de comportement”, in D. Boy, N. Mayer (eds), L’électeur français en questions, 229-36.
-
[51]
Economic liberalism, the market economy, commitment to social gains, forms of political action.
-
[52]
Ethnocentrism, sexual permissiveness, traditions-nationalism, anti-authoritarianism.
-
[53]
See the following contributions in D. Boy, N. Mayer, L’électeur français en questions: Nonna Mayer, “Ethnocentrism, racisme et intolérance”, 45-69; Guy Michelat, “À la recherche de la gauche et de la droite”, 71-103; Jean Chiche, “L’univers idéologique et politique des Français: une exploration par l’analyse factorielle”, 219-28.
-
[54]
Gérard Grunberg, Étienne Schweisguth, “Recompositions idéologiques”, in D. Boy, N. Mayer (eds), L’électeur a ses raisons, 139-78; Vincent Tiberj, “La politique des deux axes: variables sociologiques, valeurs et votes en France (1988-2007)”, Revue française de science politique, 62(1), 2012, 71-106.
-
[55]
To create tools for measuring these attitudes, we have constructed scales using the Loevinger technique. See Guy Michelat, Éric Kerrouche, “Les échelles d’attitude”, Revue internationale de politique comparée, 6(2), summer 1999, 463-512; on the content of the scales of economic liberalism and ethno-authoritarianism that we use here, see the appendix to this article.
-
[56]
Pius XI, Quadragesimo anno, 1931.
-
[57]
Jean-Paul II, Centesimus annus, 1991, art. 35±43.
-
[58]
Francis, Apostolic Exhortation Evangelii gaudium, 2013, art. 253.
-
[59]
Leviticus, 18.34.
-
[60]
Message on the occasion of World Migration Day, 1998.
-
[61]
Compared to the similar proportions observed in 1978, the present results are of the same nature but show a leftwards shift among regular churchgoers (left: 20 per cent; right: 49 per cent) and a larger rightward shift among the non-religious (left: 66 per cent; right: 12 per cent). The gap between regular churchgoers and the non-religious has shrunk.
-
[62]
The term “manifs pour tous” (“demonstrations for all”) was coined as a response to the draft legislation permitting marriage between persons of the same sex, which was thus dubbed “marriage for all”. The demonstrations took place from the autumn of 2012 to protest against the enactment of the law. Later, they extended their scope to include opposition to gay parenting and to the teaching of gender theory. Catholic associations were a particular driving force in these protests.
-
[63]
The law permitting marriage between persons of the same sex came to be known as the loi Taubira, after Christiane Taubira, the minister of justice responsible for introducing the bill to parliament.
-
[64]
The Grenelle agreements were the outcome of the negotiations held at the end of May 1968 at the ministry of labour in the Rue de Grenelle in Paris between the government, the trades unions, and employers’ organizations. They were intended to put an end to the social protest which had accompanied the student protests. Since then the term “Grenelle” has been used in France to describe any major agreement.
-
[65]
Catechism of the Catholic Church, para. 2357.
-
[66]
BVA poll. Evidence of increased tolerance in society, this attitude of total acceptance was in 2013 shared by 43 per cent of French people but only 32 per cent of those over age 64. The discrepancy between practicing Catholics and the non-religious remains constant: 45-67 per cent among those under 40, 23-48 per cent among those over 65.
-
[67]
Radio RCF, TLM channel, 13 September 2012.
-
[68]
Loevinger coefficient: 0.53.
-
[69]
NR = no response.
-
[70]
Loevinger coefficient: 0.63.
