* I wish to express my gratitude to Henry S. Albinski for his cogent comments on an earlier draft.

¹ “Approaches to an Understanding of Modern France,” in Earle, Edward Meade, ed., Modern France (New York, 1964), 10.Google Scholar

² On the history of Catholicism and its relationship to French political life see Dansette, Adrien, Histoire religieuse de la France contemporaine (Paris, 1951)Google Scholar, 2 vols.; and Latreille, André, Histoire du catholicisme en France (Paris, 1962)Google Scholar, 3 vols. See also Bosworth, William, Catholicism and Crisis in Modern France (Princeton, 1962).CrossRefGoogle Scholar

³ This view is essentially that of Professor Maurice Duverger, Duverger, however, is somewhat optimistic that a weak consensus is possible today, making the balancing role of the centre superfluous. On this point see his La Démocratic sans le peuple (Paris, 1967). For François Goguel's thoughts, see his France under the Fourth Republic (Ithaca, 1952), chap. 5.

The use of such simplistic dichotomies as left and right can be highly ambiguous analytically, masking a variety of subtle differences and internal dimensions. This has been one of the major criticisms of traditional French political analysis. However, with respect to France, the basic poles of left and right constitute a political reality of great importance in both the popular and sophisticated languages of politics. Also the employment of such categories can be useful heuristically. In a recent study of attitudinal patterns with respect to politics in France, it was revealed that notions of “left” and “right” are highly significant in orienting political behaviour, even though their precise empirical definitions remain vague. For example, only 10 per cent of those surveyed in 1966 could not situate themselves on a general left/right political continuum. Therefore there is some justification for the use of these categories as political referents. See Emeric Deutsch, Denis Lindon, and Weill, Pierre, Les families politiques aujourd'hui en France (Paris, 1966), 12–14.Google Scholar For a criticism of the left/right dichotomy see Davis, Morris, “French Electoral Sociology,” Public Opinion Quarterly, XXII (spring 1958), 35–55.CrossRefGoogle Scholar

⁴ Fougeyrollas, Pierre, “La crise de la conscience politique en France,” Preuves, 138 (Aug 1962), 3–15.Google Scholar

⁵ Macrae, Duncan Jr., Parliament, Parties, and Society in France, 1946–1958 (New York, 1967), 176–226.Google Scholar Although limited to the period of the Fourth Republic, MacRae's analysis is highly relevant to an understanding of the political forces that have shaped the Fifth Republic.

⁶ For some empirical evidence of this see Ibid., 15–35.

⁷ Lebras, Gabriel, Etudes de sociologie religieuse (Paris, 1955), I, 367.Google Scholar

⁸ French Politics and Political Institutions (New York, 1968), 15. This does not imply that religion is the only element dividing left and right in France. However, as Philip E. Converse has observed: “Numerous independent inquiries in France… seem to have brought to the surface the same finding: that religious attitudes (here clericalism vs. anti-clericalism) appear to predict party or so-called ‘left-right’ locations more frequently than status.” See his Some Priority Variables in Comparative Electoral Research, Occasional Paper no. 3 (Glasgow: University of Strathclyde, Survey Research Centre, 1968), 8.

⁹ Mattei Dogan has proposed that the absence of a strong relationship between areas of working-class strength and the incidence of left (communist and socialist) voting in France is due primarily to the religious issue. The controversy over religion has bifurcated the working class into religious and secular elements, despite the obvious similarity in economic and social goals. See his comments in Hamon, Leo, ed., Les nouveaux comportements politiques de la classe ouvrière (Paris, 1962), 101–43Google Scholar; and his “Party Cleavage and Social Stratification in France and Italy,” in Lipset, Seymour M. and Rokkan, Stein, eds., Party Systems and Voter Alignments (New York, 1967), 134–5.Google Scholar

¹⁰ On this point see Aron, Raymond, The Opium of the Intellectuals, trans. Kilmartin, Terence (Garden City, NY, 1957), 30Google Scholar; Fougeyrollas, Pierre, La conscience politique dans la France contemporaine (Paris, 1963), 93Google Scholar; and Micaud, Charles, Communism and the French Left (New York, 1963), 207–15.Google Scholar

¹¹ Le Monde, June 20–1, 1965.

¹² Aron, The Opium of the Intellectuals, 305–24; and Lipset, Seymour M., “The Changing Class Structure and Contemporary European Politics,” Daedalus, 93 (Winter 1964), 271–303.Google Scholar For additional views on this controversial point see Lapalombara, Joseph, “Decline of Ideology: A Dissent and an Interpretation,” American Political Science Review, LX (March 1966), 5–16CrossRefGoogle Scholar; and Lipset, “Some Further Comments on ‘The End of Ideology,’” Ibid., 17–18. For a discussion of the “end of ideology” phenomenon with particular reference to France see Waterman, Harvey, Political Change in Contemporary France (Columbus, 1969), 55–107.Google Scholar

¹³ For an excellent description of this process of change by a participant-observer see Wylie, Laurence, “Social Change at the Grass Roots,” in Hoffmann, Stanley, ed., In Search of France (New York, 1965), 159–234.Google Scholar See also Brown, Bernard E., “The French Experience of Modernization,” World Politics, XXI (April 1969), 366–91CrossRefGoogle Scholar; and Dupeux, Georges, La Société française, 1789–1960 (Paris, 1964).Google Scholar

¹⁴ Le Monde, April 19, 1968.

¹⁵ Ibid. See also L'année politique, économique, sociale et diplomatique en France, 1966 (Paris, 1966), 400. Since 1963, the annual volumes of L'année politique have included comparative circulation data for the major elements of the French press.

¹⁶ See Rémond, René, ed., Forces religieuses et attitudes politiques (Paris, 1965)Google Scholar; and Coutrot, A. and Dreyfus, F. G., Les Forces religieuses dans la société française (Paris, 1966).Google Scholar

¹⁷ The results of this study, conducted during September 1966, were first published in La vie catholique illustrée, 1116 (Dec. 28, 1966 and Jan. 3, 1967), and later included in the Institut française de l'opinion publique (IFOP) publication “Religion et Politique,” Sondages: Revue française de l'opinion publique, 2 (1967).Google Scholar The findings of an IFOP survey on Catholicism and politics were first published in “Les attitudes politiques des catholiques,” Le Nouvel Observateur, 102 (Oct. 26 and Nov. 1, 1966), and 106 (Nov. 22 and 29, 1966). This study was also included in Sondages, 2 (1967). For the results of previous studies see “Le catholicisme en France,” Sondages, 4 (1952), and “Les attitudes religieuses de la jeunesse,” Sondages, 3 (1959).

¹⁸ During the first postwar municipal elections (April–May 1945), the MRP had actually agreed to join the Communists in common electoral lists on the first ballot. See L'année politique, 1944–45, pp. 187–202.

¹⁹ The divided nature of Catholicism is best discussed in Rémond, René, “Droite et gauche dans le catholicisme française contemporaine,” Revue française de science politique, VIII, 3 and 4 (1958), 529–54 and 803–20.CrossRefGoogle Scholar

²⁰ “The Problem of Distances in Models of Voting Change,” in Jennings, M. Kent and Ziegler, L. Harmon, eds., The Electoral Process (Englewood Cliffs, 1966), 182.Google Scholar

²¹ Perhaps the most detailed statistics on national religious behaviour in France have been compiled by Professor Gabriel LeBras of the University of Paris. Most other analysts have focused upon religious behaviour and its relationship with socio-political and socio-economic variables within individual departments.

²² See Macrae, Duncan Jr., “Religious and Socio-economic Factors in the French Votes, 1946–56,” American Journal of Sociology, 64 (Nov. 1958), 290–8.CrossRefGoogle Scholar Electoral sociology or aggregate ecological analysis at the departmental level has had a firm tradition in the methodology of French social science research. Given the limitations of survey research in France, Gabriel LeBras has urged more emphasis upon ecological analysis to analyse the linkage between religion and political behaviour. For his position on this matter see “Géographie electorale et géographie religieuse,” Etudes de sociologie religieuse, II, 526—45.

²³ This index was first suggested by Roger de Smet as an indirect measure of religiosity in Belgium. See his “La géographie electorale en Belgique,” Revue française de science politique, II (Jan–March 1952), 87–95. This index does contain a number of inherent deficiencies, especially the danger of committing the “ecological fallacy.” However, no attempt is made here to predict individual behaviour with any finality. The index is merely an indirect measure designed to associate certain sets of aggregate data at the departmental level. The index is used to characterize a specific segment of the French electorate within the geographical bounds of the department. The assumption is that some estimate can be made of a departmental electorate's religiosity by measuring the prevalence of private (church) school enrolment in the respective department. Other indicators of religiosity such as data on regular church attendance compiled by LeBras and Boulard were considered for possible use, but finally rejected. The LeBras-Boulard data are in too crude a form to provide meaningful relationships, and is not based upon units comparable with political variables.

²⁴ Various estimates have shown that more than a million Catholic families have their children enrolled in private (church) schools. Private school enrolment correlated positively with the departmental distribution of clergy and religious personnel (Pearson's r = +.72). Using ordinally ranked data, private school enrolment also showed a strong positive association with the circulation of Catholic Action publications among the rural populations. (Goodman's Γ = +.54.)

²⁵ For some discussion of the political aspects of the religious schools question see Brown, Bernard E., “Religious Schools and Politics in France,” Midwest Journal of Political Science, II (May 1958), 160–78CrossRefGoogle Scholar; “Laïcité et paix scolaire,” Esprit, XXVII (Oct. 1959), entire issue; and Vandermeersch, Edward, “La loi scolaire,” Etudes (May 1968), 611–30.Google Scholar

²⁶ Simon, Pierre-Henri, “La Querelle scolaire,” Preuves, 110 (1960), 47.Google Scholar

²⁷ Data processing was facilitated by using the services of an IBM 360/67 computer at the Pennsylvania State University Computation Center. Enrolment figures for private schools at the department level are for the 1963–4 school year, and were obtained from the 1965 volume of Annuaire statistique de la France, compiled by the Institut national de la statistique et des études economiques (Paris, 1965), 100–1. Election results were obtained from the various editions of Le Monde.

²⁸ Brulé, Michel, “L'Appartenance religieuse et le vote du décembre 1965,” Sondages, 2 (1966), 15–19.Google Scholar For post-election commentary see Thibaud, Paul, “Les catholiques et le gaullisme,” Esprit, 353 (1966), 508.Google Scholar

²⁹ See Moreau, Jacques, “Le choix du MRP,” Revue française de science politique, XV (Feb. 1965), 67–86.CrossRefGoogle Scholar

³⁰ Since we are dealing here with virtually a total population of items (all metropolitan departments with the exception of Corsica and the Paris region), statistical confidence levels as a measure of sampling probability are inappropriate and cannot be interpreted in the usual sense. However, they are presented here because they provide additional information about the strength of the reported relationship.

³¹ The correlation between Lecanuet's vote on the first ballot and de Gaulle's vote on the second ballot was r = +.673. The correlation between Lecanuet and Mitterand's second ballot vote was r = −.674. The strength of these relationships seriously questions the thesis of François Furet and Jacques Ozouf that a good portion of Lecanuet's vote accrued to Mitterand on the second ballot. For a presentation of their position see their “La France a-t-elle changé?” Preuves, 180 (1966), 47–55.

³² The results of the 1968 legislative elections do show some slight deviation from the past. This is due, perhaps, to the special circumstances under which the elections were held. Correlations between first ballot returns in 1968 and the index of religiosity are as follows: Gaullist Union pour la défense de la république (r = +.296, p < .01); centre Centriste progrès et démocratic moderne (not significant); non-communist left Fédération de la gauche démocrate et socialiste (r = −.259, p <.05); and Parti communiste (r = −.454, p < .001). For the combined French left (communist and non-communist), the correlation was moderately high (r = −.454, p < .001). The most interesting electoral group in 1968 was the PSU (Parti socialiste unifié). Although ostensibly a grouping of the non-communist left, support for the PSU differed noticeably from that of the other parties of the left. Its support bore no relationship either to factors of religiosity or anti-Gaullism, nor was it inter-related with the other parties of the French left. PSU behaviour during the strike and student unrest was oriented toward strong support of the students. This support may have been responsible for its loss of a significant proportion of the left electorate. Religiosity influenced the referendum in that the highest percentage of “no” votes came from the non-Catholic departments. This can be illustrated using LeBras’ assessment of departmental religiosity:

³³ This analysis follows that of Goguel, François in “L'élection présidentielle française de décembre 1965,” Revue française de science pohtique, XVI (April 1966), 221–54.CrossRefGoogle Scholar