¹ The emphasis on institutionalization is due primarily to Samuel P. Huntington. See “Political Development and Political Decay,” World Politics, 27 (April, 1965), 386–430Google Scholar; Political Order in Changing Societies (New Haven: Yale University Press, 1968)Google Scholar; and “The Change to Change: Modernization, Development and Politics,” Comparative Politics, 3 (April, 1971), 283–323CrossRefGoogle Scholar. Dahl, Nordlinger and Rustow are among those who argue that “institutionalization” of governmental authority, rules of electoral competition, or political parties—crucially influenced the patterns of political development. Dahl, Robert A., Polyarchy (New Haven: Yale University Press, 1971)Google Scholar; Nordlinger, Eric, “Political Development: Time Sequences and Rates of Change,” World Politics, 20 (April, 1968), 494–521CrossRefGoogle Scholar; Rustow, D., A World of Nations (Washington, D.C.: The Brookings Institution, 1967)Google Scholar.

² For the summary of arguments to this effect see Lester Milbrath, W., Political Participation (Chicago: Rand McNally, 1965), pp. 142–154Google Scholar.

³ Thus Huntington discovers the “paradox that modernity produces stability and modernization instability,” Political Order, p. 47.

⁴ Lipset, Seymour Martin, Political Man (Garden City: Doubleday, 1960)Google Scholar; Kornhauser, William A., The Politics of Mass Society (London: Routledge, 1960)Google Scholar; Dahl, Polyarchy; and many others.

⁵ Converse, Philip E., “Of Time and Partisan Stability,” Comparative Political Studies, 2 (July, 1969), 139–171CrossRefGoogle Scholar; McPhee, William M. and Glaser, William A., eds., Public Opinion and Congressional Elections (New York: Free Press, 1962)Google Scholar.

⁶ Converse, Philip E. and Dupeaux, George, “Politicization of the Electorate in France and the United States,” Public Opinion Quarterly, 27 (Spring, 1962), 1–24CrossRefGoogle Scholar.

⁷ Weber, Max, The Theory of Social and Economic Organization (New York: Free Press, 1964), particularly pp. 87–130Google Scholar; Parsons, Talcott, “Psychoanalysis and the Social Structure,” in Essays in Sociological Theory (New York: Free Press, 1954)Google Scholar.

⁸ While the original formulation of this thesis comes from Samuel P. Huntington, he is certainly not responsible for the particular interpretation and reconstruction developed here. The intent of this work is not to interpret textually any particular view but to reconstruct some themes shared in various versions within the current approaches to “political development.” Moreover, the analysis that follows is limited to Western Europe and to party systems and, hence, it should not be construed as a general test of Huntington's thesis, which was developed primarily with reference to contemporary developing societies.

⁹ Different types of mobilization are discussed in Deutsch, Karl W., “Social Mobilization and Political Development,” American Political Science Review, 55 (September, 1961), 493–514CrossRefGoogle Scholar and in Ostrowski, Krzysztof and Przeworski, Adam, “A Preliminary Inquiry into the Nature of Social Change: The Case of the Polish Countryside,” International Journal of Comparative Sociology, 8 (March, 1967), 26–43CrossRefGoogle Scholar.

The concept of “political” mobilization is defined and discussed in Rokkan, Stein, “Mass Suffrage, Secret Voting and Political Participation,” in Political Sociology, Coser, Lewis A., ed. (New York: Harper and Row, 1966), pp. 101–131Google Scholar. Given various types of mobilization and the multiplicity of political institutions, it is highly likely that the relations between mobilization and institutionalization are context specific.

¹⁰ This definition, as well as the notation, follows Przeworski, Adam and Sprague, John, “Concepts in Search of Explicit Formulation: A Study in Measurement,” Midwest Journal of Political Science, 15 (May, 1971), 187CrossRefGoogle Scholar. No detailed list of sources for the data on mobilization is given, since they consist generally of official governmental publications or reference handbooks based on such publications. In general, the M(t) series was constructed by looking for official accounts of numbers of voters, finding the population over 20, and dividing the two. The errors, other than the unreliability of the original data, may stem from (1) frequent need to interpolate population data that are given either only by the 14–19 years of age split and/or by 10 year intervals, (2) an occasional need to rely on the tautology mobilization = eligibility turnout, where the dates of eligibility do not correspond exactly to the dates of elections. Perhaps the data for the United Kingdom are least reliable for the latter reason.

¹¹ See Przeworski and Sprague for a somewhat more extensive justification of this conception.

¹² In many countries political parties, and particularly their names, are highly ephemeral. Fortunately, errors that may have emerged from misinterpreting their fluctuations should be relatively small since they probably involve mostly small parties. In general, the rule followed was to treat the parties as they presented themselves to the electorate rather than to look for the underlying ideological tendencies. Since the results obtained here may to some extent depend upon the particular decisions involved in measuring the voting change, the list of sources which was followed in identifying the parties is provided. These are for the United Kingdom: Kinnear, Michael, The British Voter, An Atlas and Survey Since 1885 (London: Batsford, 1968)Google Scholar; Denmark: Statistical Yearbook (Kobenhavn: Danmarks Statistik, various years); Sweden: Lewin, L., Janssen, B., and Sorbom, D., The Swedish Electorate, 1887–1968 (Stockholm: Almquist and Wiksell, 1972)Google Scholar for the pre-1911 period; Statistisk Årsbok för Sveṅge (Stockholm: Statistiska Centralbyrån, various years)Google Scholar; Norway: Statistisk Årbok for Norge (Oslo: Statistisk Sentralbyra, various years)Google Scholar; Switzerland: Statistisches Jahrbuch (Basel: Statistisches Amt, 1969)Google Scholar; Finland: Statistical Yearbook (Helsinki: Statistical Office, various years)Google Scholar; Belgium: Annuaire Statistique de la Belgique (Bruxelles, various years)Google Scholar; France: Peter Campbell, French Electoral Systems and Elections, 1789–1957 (London: Fabep and Faber, 1958)Google Scholar, and Annuaire Statistique for the 1958 and 1962 elections; Canada: Urquhart, M. C., ed., Historical Statistics of Canada (Toronto: Macmillan, 1965)Google Scholar; Germany: Statistisches Jahrbuch für das Deutsche Reich (Berlin: Statistisches Reichsamt, 1933)Google Scholar.

¹³ The entire logic of analysis underlying this investigation is based on Cortés, Fernando, Przeworski, Adam, and Sprague, John, Systems Analysis for Social Scientists (New York: Wiley, 1974)Google Scholar.

¹⁴ Given the rules of graph algebra. See Cortés, et al., Systems Analysis, Chapters 4 and 5.

¹⁵ Readers not familiar with difference equations may want to consult Goldberg, Samuel P., Introduction to Difference Equations (New York: Wiley, 1958)Google Scholar.

¹⁶ The choice of countries in which the theory was tested was dictated by availability of data or by the length of time series. Austria, Italy, and Spain were not included because their time series were too short. The United States was not included deliberately.

¹⁷ Least-squares estimation of this model presents a number of problems stemming from the fact that the dependent variable occurs with a lag as a regressor. To analyze the behavior of error terms, I have computed the Durbin-Watson statistic in those cases where the graphs of errors gave the impression that they may be autocorrelated. This test brought the conclusion that errors are not autocorrelated. However, I discovered subsequently that the Durbin-Watson statistic should not be used when some of the regressors were lagged dependent variables and the tests developed for this kind of a situation are valid only asymptotically for large samples. See Nerlove, M. and Wallis, K. F., “Use of the Durbin-Watson Statistic in Inappropriate Situations,” Econometrica, 34 (January, 1966), 235–238CrossRefGoogle Scholar; Durbin, J., “Testing for Serial Correlation in Least-Squares Regression When Some of the Regressors are Lagged Dependent Variables,” Econometrica, 38 (May, 1970), 410–421CrossRefGoogle Scholar.

Without knowing whether and how the error terms may be autocorrelated all we know is that: (1) the sample estimates of parameters are biased, and (2) they are inconsistent estimators of true parameter values. Hibbs, Douglas A. Jr. (“Problems of Statistical Estimation and Causal Inference in Time-Series Regression Models,” Sociological Methodology 1973, Costner, Herbert L., ed. [San Francisco: Jossey-Bass, 1973])Google Scholar demonstrates that values of our parameter P will generally be overestimated while the values of Q will be underestimated. Hence even the low values of P which we obtained are most likely higher than the true values. I am grateful to Douglas Hibbs for his comments and for letting me see a prepublication copy of his article.

¹⁸ I take into account the fact that the Canadian military personnel did not vote in the first election following World War I. It is significant that in Belgium, where female universal franchise was introduced in 1948, the mobilization of women had much less impact on the voting distribution than the war.

¹⁹ See Rokkan, Stein and Valen, Henry, “The Mobilization of the Periphery: Data On Turnout, Party Membership and Candidate Recruitment in Norway,” Acta Sociologica (January, 1962), 111–158Google Scholar.

²⁰ For a discussion of this concept of memory see Cortés et al., Systems Analysis, Chapter 10.

²¹ Actually, I estimated the parameter of the continuous function

where C is a constant. Since only three points, 1890, 1920, and 1950, were used for the estimation, the results should be treated as very rough.

²² Whatever way one tries to find a relationship between the pace and/or the timing of mobilization and the mean institutionalization, the attempts fail. These results cannot be presented in detail for lack of space, but they are highly significant in the light of the quantity of literature that posits such a relation. The thesis that relates the pace of electoral mobilization to the institutionalization, stability, or continuity of political institutions requires detailed examination within the full comparative spectrum. The analysis of selected cases, upon which this thesis has been built, relies excessively on Anglo-centrism and comparisons of the United Kingdom, Germany and France.

²³ See footnote 17 for the discussion of problems in fitting this model. The introduction of the “partially dummy” variables should not complicate matters excessively since they are linearly independent. This model, however, does contain nearly the entire bag of econometric worms, since the same variable appears on both sides, dummy-like variables are utilized, and the error may be serially dependent. Hence, the stability of the estimates of P and S should be considered as a rough criterion of the magnitude of problems introduced additionally by the separate treatment of mobilization and demobilization.

²⁴ German dominant ideology considered the mystical will of the “volk,” rather than outcomes of partisan intrigues, to constitute the source of political legitimacy. The notion of representation mediated by parties never gained ideological primacy over the concept of unmediated relation between the “community” and a leader. See the essays by Ernst Fraenkel and particularly by Sontheimer, Kurt in Stern, Fritz, ed., The Path of Dictatorship, 1918–1933 (Garden City, N.Y.: Doubleday, 1966)Google Scholar.

²⁵ This entire list follows nearly verbatim various passages in Carr, William, A History of Germany, 1815–1945 (New York: St. Martin's Press, 1969)Google Scholar.

²⁶ The numbers of uncontested constituencies were 64 in 1892, 185 in 1895, 241 in 1900, 114 in 1906, 75 in the first election of 1910, 163 in the second election of 1910. The turnout dropped in 1895, 1900, and the second election of 1910. See Michael Kinnear, The British Voter, An Atlas and Survey Since 1885; and Cornford, James, “Aggregate Election Data and British Party Alignments, 1885–1910,” in Mass Politics, ed. Allardt, Erik and Rokkan, Stein (New York: The Free Press, 1970), p. 114Google Scholar.

²⁷ This is a loose translation from Gramsci's, AntonioNote sul Machiavelli, sulla politico e sullo stato moderno (Torino: Einaudi, 1966), p. 81Google Scholar. The original reads as follows:

Si misura proprio l'efficacia e la capacita di espansione e di persuasione delle opinioni di pochi, delle minoranze attive, delle elites, delle avan-guardie, ecc. … Le idee e le opinioni non “nascono” spontaneamente nel cervello di ogni singolo; hanno avuto un centro di formazione, di irradazione, di diffusione, di persuasione, un gruppo di uomini o anche una singola individualita che le ha elaborate e presentate nella forma politica di attualita. La numerazione dei “voti” e la mani-festazione terminale di un longo processo in cui l'influsso massimo appartiene proprio a quelli che “dedicano alio Stato e alla nazione le loro migliori forze.”

²⁸ The importance of the sharp change in the relations between classes and parties is emphasized by Poulantzas, Nicos, Fascisme et dictature (Paris: Francois Maspero, 1970)Google Scholar.

²⁹ For a short history of Italy during this period see Saladino, Salvatore, Italy from Unification to 1919 (New York: Crowell, 1970)Google Scholar. Controversies concerning the Giolitti era are collected in Italy from the Risorgimento to Fascism. Salomone, A. W., ed. (Garden City, N.Y.: Doubleday, 1970)Google Scholar. See in particular the essay by Gaetano Salvemini, “Was Prefascist Italy a Democracy?” reprinted in this volume under the title “The Historian as Polemicist. ‘Giolitti, No—Democracy, Yes’,” pp. 474—490, for a description of the electoral system under Giolitti's rule. For the attitude of the working class movement toward elections see Tasca, Angelo, La Naissance du Fascisme (Paris: Galimard, 1938), Chapter IIGoogle Scholar.

³⁰ Linz, Juan, “The Party System of Spain: Past and Future,” in Lipset, Seymour M. and Rokkan, Stein, Party Systems and Voter Alignments (New York: The Free Press, 1967), p. 202Google Scholar.

³¹ Cuadrado, Miguel Martinez, Elecciones y Partidos Politicos de Espana, 1868–1931, 2 vol. (Madrid: Taurus, 1969), p. 760 and Table 28Google Scholar.