¹ Hoffmann, Stanleyet al., In Search of France (Cambridge: Harvard University Press, 1963)Google Scholar; Luethy, Herbert, France Against Herself (N. Y.: Meridian Books, 1957)Google Scholar; Williams, Philip M., Crisis and Compromise (Hamden, Conn.: Archon Books, 1964)Google Scholar; and Jacques Fauvet, La France déchirée (Paris: Fayard, 1957)Google Scholar, which might be translated as France Torn. (An English translation is less literal and less revealing: The Cockpit of France [London: Harvill, 1960].Google Scholar)

² In addition to the works cited in footnote 1, see Crozier, Michel, The Bureaucratic Phenomenon (Chicago: University of Chicago Press, 1964), Part FourGoogle Scholar; Lipset, Seymour Martin, The First New Nation (N. Y.: Basic Books, 1963), chs. 6, 8, 9Google Scholar; and de Tocqueville, Alexis, The Old Regime and the French Revolution (N. Y.: Anchor Books, 1955)Google Scholar. For two analyses that minimize the lack of consensus in France, see Leites, Nathan, On the Game of Politics in France (Stanford: Stanford University Press, 1959)Google Scholar; and Williams, Philip M., “Political Compromise in France and America,” The American Scholar, 26 (Summer 1957), pp. 273–288Google Scholar.

³ For a partial bibliography of French electoral sociology, see Goguel, François and Grosser, Alfred, La politique en France (Paris: Armand Colin, 1964), pp. 101–02Google Scholar; and Lancelot, Alain and Ranger, Jean, “Développements récents de la recherche électorate en France,” Il Politico, 29 (12 1964), pp. 763–787Google Scholar.

⁴ Bettelheim, Charles and Frère, Suzanne, Une ville française moyenne: Auxerre en 1950 (Paris: Armand Colin, 1951)Google Scholar; Marie, Christiane, Grenoble 1871–1965, l'évolution du comportement politique d'une grande ville en expansion (Paris: Armand Colin, 1966)Google Scholar; Pitt-Rivers, Julian, “Social-Class in a French Village,” Anthropological Quarterly, 33 (01 1960), pp. 1–13CrossRefGoogle Scholar; Josserand, Roger, “Rapport d'enquête sur la commune de Marigny en Charolais” (unpublished Mémoire, Institut d'Etudes Politiques de Grenoble, n.d.)Google Scholar; Bernot, Lucien and Blancard, René, Nouville: Un village français (Travaux et Mémoires de l'Institut d'Ethnologie, Vol. 57; University of Paris, 1953)Google Scholar; Mendras, Henri, Etudes de sociologie rurale: Novis et Virgin (Paris: Armand Colin, 1951)Google Scholar; Wylie, Laurence, Village in the Vaucluse (Cambridge: Harvard University Press, 1957)Google Scholar, Chanzeaux, A Village in Anjou (Cambridge: Harvard University Press, 1966)Google Scholar, and “Social Change at the Grass Roots,” in Hoffmann et al., In Search of France; and Clément, Pierre and Xydias, Nelly, Vienne sur le Rhône (Paris: Armand Colin, 1955)Google Scholar.

⁵ For example, according to the reports of Pitt-Rivers, op. cit., and Wylie, Village in the Vaucluse, local politics in Magnac is quite different from local politics in Peyrane. Further study would be needed to determine which village is more nearly typical of most French communes.

⁶ The leading journals are La revue administrative and Départements et communes; two important books are Schmitt, Charles, Le maire de la commune rurale (Paris: Berger-Levrault, 1959)Google Scholar; and Chapman, Brian, Introduction to French Local Government (London: Allen and Unwin, 1953)Google Scholar.

⁷ Chapman, op. cit., p. 219.

⁸ Siegfried, André, France: A Study of Nationality (New Haven: Yale University Press, 1930), p. 7Google Scholar. Also see Touchard, Jean, in the preface to L'association française de science politique, Le référendum de septembre et les élections de novembre 1958 (Paris: Armand Colin, 1960), p. xxiiGoogle Scholar.

⁹ Chapman, op. cit., and Wylie, Village in the the Vaucluse. See the reviews of these books in Revue française de science politique, 6 (07–September 1956), pp. 672–674Google Scholar; and ibid., 7 (October–December 1957), pp. 951–953, respectively.

¹⁰ See the works cited in footnotes 1–2 above.

¹¹ The act of voting is used to measure citizen involvement because turnout can be ascertained and measured relatively easily. Of course, voting alone may demonstrate only minimal political involvement. In The Civic Culture (Boston: Little, Brown, 1965), p. 131Google Scholar, Gabriel A. Almond and Sidney Verba regard voting “as a relatively passive form of participation in community life, though a form of participation it certainly is.”

¹² In Why Europe Votes (Chicago: University of Chicago Press, 1930)Google Scholar, Harold F. Gosnell found that turnout and competition vary directly in England and France. In England, “… the closer the contest the greater is the interest in the election”: ibid., p. 14. In France, “the closeness of the contest has a great deal to do with the size of the votes cast….”: ibid., p. 49. Studying turnout in France, Jean Meynaud and Alain Lancelot also find that abstentions “tend to diminish when the electoral outcome appears uncertain: the elector becomes more aware of the importance of his participation”: La participation des Français à la politique (Paris: Presses Universitaires de France, 1965), p. 18Google Scholar. In Swedish local elections, according to Donald R. Niemi, “… the presence of competing political lists was related to higher voter participation….” Niemi, , “Sweden's Municipal Consolidation Reforms,” (unpublished Ph.D. thesis, University of Chicago, 1966), p. 84Google Scholar.

¹³ Rémond, René, “Participation électorate et participation organisée,” in Vedel, Georges (ed.), La dépolitisation: mythe ou réalité? (Paris: Armand Colin, 1962), p. 78Google Scholar.

¹⁴ Fauvet, op. cit., p. 56.

¹⁵ Meynaud and Lancelot, op. cit., p. 18.

¹⁶ Wahl, Nicholas, “The French Political System,” in Beer, Samuel H. and Ulam, Adam B. (eds.), Patterns of Government (2d ed.; New York: Random House, 1962), p. 357Google Scholar.

¹⁷ Stoetzel, Jean, “Voting Behaviour in France,” The British Journal of Sociology, 6 (06 1955), p. 105CrossRefGoogle Scholar.

¹⁸ Goguel and Grosser, op. cit., p. 58.

¹⁹ Wylie, , Village in the Vancluse, p. 233Google Scholar; and Busnicourt, Jacques, “Un canton rural du Santerre,” in Fauvet, Jacques and Mendras, Henri (eds.), Les paysans et la politique (Paris: Armand Colin, 1958), p. 474Google Scholar.

²⁰ Ridley, F. and Blondel, J., Public Administration in France (London: Routledge and Kegan Paul, 1964), p. 99Google Scholar.

²¹ Lancelot, Alain, in the colloquium, “Au lendemain de l'élection présidentielle,” Revue française de science politique, 16 (02 1966), p. 178Google Scholar.

²² Gosnell, op. cit., p. 153.

²³ It is not easy to obtain local election returns in France. Prefectures only send the returns for large cities to the Ministry of the Interior in Paris. For electoral data in all other communes, one must consult the records at each prefecture.

²⁴ Turnout in the legislative elections of November 23, 1958 is noted in L'année politique, 1958, p. 145. This represents the percentage of valid votes cast (exprimés) in relation to registered voters (inscrits). For the 1959 municipal elections, the Ministry of the Interior estimated turnout by using all votes cast (votants), whether valid or not: Le Monde, March 17, 1959. Since the difference between these two measures of turnout is usually about 2 per cent, the Ministry's figure of 73.4 per cent was reduced by about 2 per cent to determine exprimés.

²⁵ Although sampling procedures were not used, an attempt was made to choose departments representative of varying conditions in France. However, as Jean Blondel has noted, “In the case of France, it is very difficult to state whether [a] particular département is typical or not: is there anything, from the point of view of social, political or economic life, which can be said to be typical of the whole of France?”: “Local Government and the Ministries in a French Département,” Public Administration (London), 37 (Spring 1959), p. 66Google Scholar.

²⁶ Electoral data were obtained from the prefectures of the three departments. Of the total of 1,978 communes in the three departments, the 95 communes in the arrondissement of Vire (Calvados) are excluded because it was impossible to get the necessary data. It was also impossible to obtain data for 73 other communes. Thus 1,810 communes were selected for statistical examination, nearly five per cent of all French communes.

²⁷ Figures are from the premier tour of each election. The national election of November 1958 was chosen since it was nearest in time to the most recent municipal elections—those of March 1959—at the time this study was conducted. Of course, turnout varies in successive elections for the same office. According to unpublished figures made available by the Ministry of the Interior, turnout was lower in the 1959 municipal elections than in any other municipal elections since World War II. Turnout in the 1958 legislative elections was lower than in previous legislative elections since World War II but higher than in the succeeding legislative election of 1962. Moreover, until December 1965, national elections in France meant legislative elections, which are used here for purposes of comparison. Different political patterns prevail for the French presidential elections. For example, turnout was 84.1 per cent in the premier tour of the December 1965 presidential elections: L'année politique, 1965, p. 105. Therefore, the hypotheses presented in this study should be tested further.

²⁸ It is significant that when the percentage of turnout in national elections is averaged by the communes in the sample, or totaled for all France and the percentage taken, the two methods of calculating turnout produce similar results. The same is not true for turnout in local elections. This suggests that the size of commune affects turnout in local elections more than in national elections, an hypothesis that will be supported by evidence presented below.

²⁹ Unpublished election returns were made available by the Bureau of Elections, Ministry of the Interior. Aggregate returns (i.e., those compiled by adding together all returns from all communes) were used for comparison rather than average-based figures (i.e., figures obtained by averaging the turnout percentages for all communes taken one by one).

³⁰ The proportion was reversed for turnout based on aggregate totals in the three departments chosen for close study. In two, turnout was higher in national elections than in local. This suggests that the finding regarding higher turnout in local elections holds elsewhere in France, where even aggregate-based turnout percentages are higher in local than in national elections.

³¹ Turnout also appears to be affected by a regional factor independent of the size of communes. Communes in the department of the Nord have a far higher average turnout rate in both local and national elections than communes in either of the other two departments. See Table 1, above.

³² Taylor, Edmond, “French Politics at the Municipal Level,” The Reporter, 04 8, 1965, p. 32Google Scholar; and Goguel, François, “Les élections municipales des 14 et 21 mars 1965: la signification de la consultation,” Revue française de science politique, 15 (10 1965), p. 912CrossRefGoogle Scholar.

³³ L'année politique, 1965, p. 20.

³⁴ In the 1959 municipal elections, average turnout for the 8 communes of the sample over 30,000 was 72.5 per cent, as compared with 81.8 per cent average turnout for all communes in the sample under 30,000. The comparison with the Ministry's figures is limited, however, because turnout was calculated differently in the two cases. In the research reported here, turnout for small and large communes was determined by averaging the turnout rate for all communes in each of the two groups. The Ministry of the Interior determined the turnout rate by adding together those voting in communes over 30,000, and dividing this figure by the total number of registered voters in the communes of that size. The Ministry calculated turnout for communes under 30,000 in a similar fashion.

³⁵ The formula for the index of multipartyism was suggested by Duncan MacRae, Jr. It is:

where: k = number of candidates or lists; p_i, = proportion of vote for ith list; Σ_ip_i = 1. Lists or candidates receiving less than 5 per cent of the vote were eliminated from calculation of the index. The index has also been used in the social sciences by Soares, Glaucio Ary Dillon and de Noronha, Amelia Maria Carvalho, “Urbanização e Dispersão Eleitorae,” Revista de Direito Público e Ciencia Politico, 3 (07–Dec. 1960), pp. 258–70Google Scholar.

³⁶ The index does not measure conflict or competition but rather the convergence of votes among competing parties. Competition and conflict are dependent on electoral laws and other factors in addition to the distribution of votes among candidates. However, since the closeness of the election is an important part of competition, the terms “competition” and “conflict” will be used below to refer to the index of multipartyism.

³⁷ The total of the averages for the lists does not equal the total number voting. For the municipal elections of March 1959, voters in communes under 120,000 were not required to vote for an entire list and could delete or substitute names if they wished. Therefore, different candidates on a list might not receive the same number of votes. In communes under 2,500, arrangements were even more flexible. For the calculations here, the number of votes on a given list was determined by averaging the votes received by the candidate with most votes on the list, and the candidate with fewest votes on the list.

³⁸ Since the average of 3 votes was less than 5 per cent of the total cast, the 3 votes were eliminated from calculation of the index.

³⁹ The relation between turnout and competition in the 1958 national election was less clear. Turnout and competition were inversely related for the 50 communes in the sample with highest competition and the 50 communes with lowest competition. However, multiple regression analysis showed the turnout rate to be relatively unimportant in explaining competition in the 1958 national elections.

⁴⁰ Response to a French public opinion poll furnishes additional evidence that national and local cleavages are quite distinct and that they more nearly coincide as the size of commune increases. French voters were asked, “In choosing among candidates in municipal elections, are you concerned to know their attitudes toward General de Gaulle's policy?” Far more of those questioned replied “no” than “yes.” In rural communes, three-fourths of the respondents were not concerned about local candidates' attitudes toward de Gaulle's policies. The proportion of respondents who were concerned was higher in big cities: Piret, Jeanne, “L'opinion publique au début de l'année 1965,” Revue française de science politique, 15 (06 1965), p. 536CrossRefGoogle Scholar.

⁴¹ The communes of Bordeaux and Lille, which were larger than 120,000, and consequently elected their municipal councils by proportional representation, are excluded from this calculation.

⁴² While size of commune helps explain competition in local elections, it does not nullify the previous finding that turnout is negatively related to competition. In a multiple regression analysis, both turnout and size of commune independently affect local electoral competition and, in fact, turnout plays a far more important role than size of commune.

⁴³ Cf. Campbell, Angus, Converse, Philip E., Miller, Warren E., and Stokes, Donald E., The American Voter (New York: Wiley, 1960), Chs. 13 and 15Google Scholar. The question is examined intensively in a forthcoming study of French politics and society by Duncan MacRae, Jr.

⁴⁴ The French mayor is elected for a six-year term.

⁴⁵ There was also great stability in the municipal elections of March 1965. See Goguel, loc. cit.; and Williams, Philip M., “Party, Presidency and Parish Pump in France,” Parliamentary Affairs, 18 (Summer 1965), pp. 257–265CrossRefGoogle Scholar.

⁴⁶ Dogan, Mattei, “Les candidats et les élus,” in Duverger, Maurice, Goguel, François, and Touchard, Jean (eds.), Les élections du 2 Janvier 1956 (Paris: Armand Colin, 1957), p. 446Google Scholar; and Dogan, , “Political Ascent in a Class Society: French Deputies, 1870–1958,” in Marvick, Dwaine (ed.), Political Decision-makers (New York: The Free Press of Glencoe, 1961), pp. 57–90Google Scholar.

⁴⁷ L'année politique, 1958, p. 146. One reason was that the size of the National Assembly was reduced. For a study examining the defeat of Fourth Republic politicians in Fifth Republic elections, see Dogan, Mattei, “Le personnel politique et la personnalité charismatique,” Revue française de sociologie, 6 (07–September 1965), pp. 305–324CrossRefGoogle Scholar.

⁴⁸ Moreover, the higher attrition rate took place in a shorter time. Whereas the previous elections for mayor took place six years earlier, those for deputy occurred only two years before.

⁴⁹ “Many Gaullists who swept to the Assembly in a wave of protest in 1958 were surprised to see the ‘man of the System’ they had expelled from Parliament comfortably returned to his mayoral chair six months later”: Williams, , Crisis and Compromise, pp. 331–332Google Scholar.

⁵⁰ The issues raised here are explored at greater length in Kesselman, Mark, The Politics of Local Consensus in France (New York: Knopf, forthcoming)Google Scholar.

⁵¹ Georges Lavau, “Les aspects socio-culturels de la dépolitisation,” in Vedel, op. cit., p. 197. François Goguel and André Philip have also noted that French local elections differ from national elections, Goguel, loc. cit., p. 917; Philip, , in Hamon, Léo (ed.), Les nouveaux comportements de la classe ouvrière (Paris: Presses Universitaires de France, 1962), p. 128Google Scholar.