¹ The German ballot consists of a single sheet of paper on which are printed two parallel columns of contestants. Each voter casts a first vote from the column on the left side of the ballot for the party-sponsored or independent candidate whom he favors in his local election district; and casts a second vote from the right-hand column for the Land (state) list of the political party which he favors nationally. His two votes need not coincide and need not both be exercised. Since each vote is tallied separately, an invalid vote in one column does not invalidate the vote in the other column. For a detailed description of Germany's mixed system of “personalized proportional representation,” see Kitzinger, Uwe, German Electoral Politics: A Study of the 1957 Campaign (London, Oxford University Press, 1960), pp. 17–37 Google Scholar.

² Based on official publications of: Germany (Federal Republic), Statistisches Bundesamt, Wiesbaden (hereafter referred to as “Bundesamt”). Data for the 1949, 1953 and 1957 elections appeared as Statistik der Bundesrepublik Deutschland, Vol. 10 (1949)Google Scholar; Vol. 100, Nos. 1–2 (1953); and Vol. 200, Nos. 1–4 (1957). Data for the 1961 election appeared in the new electoral series, under Fachserie A: Bevoelkerung und Kultur, Reihe 8, Wahl zum Deutschen Bundestag, as Wahl zum 4. Deutschen Bundestag vom 17. September 1961, Nos. 1–4. (Hereafter these documents will be cited as “Bundesamt: 1949,” ”Bundesamt: 1953 (2),” etc.). For a detailed bibliography of official electoral statistics published by the Bundesamt and state statistical offices, see Bundesamt: 1961(1), pp. 59–67. A useful compilation of federal and state election statistics, 1946–1960, is contained in Faul, Erwin, ed., Wahlen und Waehler in Westdeutschland (Villingen-Schwarzwald, Ring-Verlag, 1960), pp. 321–363 Google Scholar.

³ German invalid vote data for 1871–1919 are found in Germany, Reichsamt, Statistisches, Statistisches Jahrbuch fuer das Deutsche Reich, Vol. 42 (Berlin, 1921/1922), pp. 354–355 Google Scholar; and for 1919–1932, ibid., Vol. 52 (Berlin, 1933), p. 539. Some comparative data are given by Richard M. Scammon, “Postwar Elections and Electoral Processes,” in Litchfield, E. H., ed., Governing Postwar Germany (Ithaca, N.Y., Cornell University Press, 1953), p. 505 Google Scholar. For France see Goldey, David B., “The French Referendum and Election of 1962: The National Campaigns,” Political Studies, Vol. 11 (10, 1963), p. 302 CrossRefGoogle Scholar; and Williams, Philip, Politics in Post-War France (2d ed., New York, Longmans, 1958), p. 324 and pp. 446–447 Google Scholar.

Comparative data on invalid votes in parliamentary elections in fifteen Western European countries since 1945 are found in the appendix on post-war election statistics in Kitzinger, Uwe, Britain, Europe and Beyond: Essays in European Politics (Leyden, A. W. Sythoff, 1964), pp. 183–217 Google Scholar. See also Kitzinger's discussion of invalid votes in Belgium, and in the cantons of Switzer-land and states of Austria where voting is compulsory, in his chapter on “The Belgian Electoral System,” ibid., pp. 174–175. For an analysis of the cleavages in Belgian politics, see Val Lorwin's chapter on Belgium in Dahl, R. A., ed., Oppositions in Western Democracies (New Haven, Yale University Press, 1965)Google Scholar, forthcoming.

Herbert Tingsten, in an extensive examination of the effects of compulsory voting in several European countries before the Second World War, demonstrated that compulsory voting—which, if effectively enforced, inevitably involves the electoral mobilization of some citizens who are more or less unwilling or apolitical—was in most cases accompanied by increases in both the level of participation and the proportion of invalid votes. See his Political Behavior: Studies in Election Statistics (Totowa, New Jersey, The Bed minister Press, 1963 Google Scholar; first published as Stockholm Economic Studies, Vol. 7, 1937), ch. 4Google Scholar.

Besides Belgium, only Luxembourg has consistently recorded an invalid vote well in excess of 3 % of the poll; clues for the reasons underlying it may perhaps be found in highly effective compulsory voting regulations and the relatively complex electoral system, described by Kitzinger, “The Luxembourg Electoral System,” in Kitzinger, loc. cit., pp. 181–182.

Italy, in 1953, completes the list of deviant Western European examples: 4.6% of the voters spoiled their ballots in that year's parliamentary election. The chief motivation seems to have been popular resentment of a specially rigged electoral law pushed through parliament just prior to the election and designed to insure continued dominance by Italy's four-party governing coalition. The law provided that any associated group of parties which received 50% plus one of the votes cast was to obtain two-thirds of the parliamentary seats; thus it represented an attempt to overthrow Italy's traditional proportional representation electoral system. For many Italians the law was unhappily reminiscent of Mussolini's Acerbo Law of 1923. Dubbed the “swindle law” by the Communists, it elicited a great deal of bitter debate both in parliament and in the election campaign itself. Cf. Mangone, Gerard J., “Italy,” in Cole, Taylor, ed., European Political Systems (2d ed. rev.; New York, 1959), pp. 508–509 Google Scholar.

⁴ Some representative examples are: Barnes, Samuel H., Frank Grace, James K. Pollock and Sperlich, Peter, “The German Party System and the 1961 Federal Election,” this Review, Vol. 56 (12, 1962), pp. 905–906 Google Scholar; Bundesamt, , “Ungueltige Stimmen bei der Bundestagswahl 1961,” Wirtschaft und Stalistik, 1962, No. 3, pp. 145–147 Google Scholar (hereafter cited as “Bundesamt: WS 1962(3)”); and Wuerttemberg, Baden-, Landesamt, Statistisches, Statistik von Baden-Wuerttemberg, Vol. 80 (1961), pp. 10–11 Google Scholar.

⁵ Barnes, et al., op. cit., p. 906.

⁶ Bundesamt: 1961(3), p. 11. According to Paragraph 39 of the West German Federal Electoral Law, a vote is considered invalid if the voter's preference for a district candidate or party list is not unmistakably clear. In Table II, “blank” votes include a small number of crossedout vote columns as well as those which were unmarked; “multicrossed” votes are those marked for more than one candidate or party list; and “miscellaneous causes” include vote-columns defaced by marks, tears or comments, as well as those on which the voter's mark is unclear. The special count of invalid votes is described in Bundesamt: WS 1962(3), p. 145. On the sample survey, see Bundesamt: 1961(3), pp. 4–8.

⁷ Faul, ed., loc. cit. Void district votes have also been consistently higher in the southwestern part of the Federal Republic, except for 1953 when they were also high in these three north German states. With few exceptions the void district vote has decreased steadily in all states and in every election since 1953.

⁸ See, for example, the contentions of Barnes, et al., op. cit. But in 1957 void list votes increased in Baden-Wuerttemberg by 68,000 (1.3%), in Nordrhein-Westfalen by 62,000 (0.4%); the Saar-land added 38,000. Nordrhein-Westfalen increased by only 0.2% in 1953, and decreased by 0.4% in 1961. Baden-Wuerttemberg decreased by 0.9% in 1953, and increased by 0.6% in 1961. There-fore, even without the Saarland, the federal average invalid vote for 1957 represented an increase more than twice the federal increase of 1953 or 1961. Bundesamt: 1957(3), pp. 64–65; Bundesamt: 1961(1), p. 8.

⁹ Since women form a majority among voters, their share of the invalid vote is numerically even more striking. Bundesamt, “Wahlbeteiligung und Stimmabgabe der Maenner und Frauen nach dem Alter bei der Bundestagswahl 1961,” WS 1962(2), pp. 75–79.

¹⁰ Bundesamt: 1961(3), p. 11.

¹¹ Barnes, et al., loc. cit.

¹² Baden-Wuerttemberg, op. cit., pp. 9–11. For the federal trends see Bundesamt: 1961(3), p. 9; and Bundesamt: WS, loc. cit.

¹³ Bundesamt: 1957(3), pp. 21–22. Most of the districts (24 of 27) which ranked highest in 1953, 1957 or 1961 in per cent of invalid votes—and three-fourths of those with consistently high percentages—were comprised predominantly of farms, villages and small towns. Cf. Bundesamt, Statistische Berichte, Arb.-Nr. VIII/5/12 and Arb.-Nr. VIII/5/15; and Bundesamt: 1953(1), pp. 52–63. See also Bundesamt: 1961(4), p. 26, for the results of the 1961 sample survey.

¹⁴ Bundesamt: 1961(3), p. 10; 1957(3), p. 53; 1953(2), p. 47. For 1953 see also Hirsch-Weber, W. and Schuetz, Klaus, Waehler und Gewaehlte: Untersuchung der Bundestagswahlen 1953 (Berlin and Frankfurt, Verlag Franz Vahlen, 1957), pp. 322–333 Google Scholar.

¹⁵ Bundesamt: WS 1962(3), p. 146.

¹⁶ Schachtner, Richard, “Wahlberechtigte, Wahlbeteiligung, Nichtwaehler und Falschwaehler,” Bayern in Zahlen: Monatsheft des Bayerischen Statistischen Amtes, Vol. 6 (01, 1952), p. 21 Google Scholar. Cf. Lipset, S. M., Political Man (Garden City, N. Y., 1963), pp. 212–213 Google Scholar. Of course, cross-pressures could equally well lead to withdrawal and abstention. Cf. ibid., pp. 211ff.

¹⁷ For example, see Schachtner, op. cit., pp. 20–21; Baden-Wuerttemberg, op. cit., p. 11; Hessen, Landesamt, Statistisches, Staat und Wirtschaft in Hessen, Vol. 4 (01, 1949), p. 15 Google Scholar; also Friedrich Erbe, “Vierzehn Jahre Wahlen in Westdeutschland (1946–1960),“ in Faul, ed., op. cit., p. 40. However, the Bundesamt noted that introduction of the two-vote, two-column ballot in 1953 had had only a fractional impact on invalid voting; void list votes increased by only 0.2% from a sum which in 1949 (see Table I) the Bundesamt had regarded as chiefly an overt expression of political protest. Bundesamt: 1953(2), p. 47; 1949, p. 7. Already by 1957 and especially in 1961 the Bundesamt's position inexplicably changed; its new emphasis has been on technical complexity as the sounder explanation for both the size and successive increases of invalid votes. See Bundesamt: 1957(3), p. 21 and p. 52; and Bundesamt: WS, loc. cit.

¹⁸ Bundesamt, “Ergebnis der Briefwahl bei der Bundestagswahl 1961,” WS 1962(1), pp. 23–25. There are, however, special features of the postal ballot which impair its comparability with other aspects of invalid voting; for a brief discussion, see Kitzinger, , German Electoral Politics, pp. 284–285 Google Scholar.

¹⁹ Bundesamt: 1957(3), p. 54; Bundesamt: WS 1962(2), p. 79. Cf. Kitzinger, , German Electoral Politics, p. 288 Google Scholar.

²⁰ For 1961 see Emnid-Institut, , Der Prozess der Meinungsbildung (Bielefeld, 1962)Google Scholar, excerpts from which appear in Eberlein, Klaus D., “Die Wahlentscheidung vom 17. September 1961, ihre Ursachen und Wirkung,” Zeitschrift fuer Polilik, Vol. 9 (09, 1962), pp. 236–257 Google Scholar; and Schmidtchen, Gerhard and Noelle-Neumann, E. (Allensbach), “Die Bedeutung repraesentativer Bevoelkerungsumfragen fuer die offene Gesellschaft,” Politische Vierteljahresschrift, Vol. 4 (06, 1963), pp. 174–179 Google Scholar. For 1957 see Divo-Institut, Untersuchung der Waehlerschaft und Wahlentscheidung 1957 (Frankfurt, 1959), pp. 72–121 and pp. 170–235 Google Scholar. For 1953 see HirschWeber and Schuetz, op. cit., pp. 94–355.

²¹ Kitzinger, loc. cit.; and Bundesamt: 1957(3), p. 21 and p. 52. The Bundesamt's suggestion that void party list votes may have resulted from rejection by major party voters of unalterable state party lists over whose composition and order they have no influence, seems improbable in the light of these data, since it is the party and not the man for whom the voter apparently marks his ballot. Bundesamt: 1957(3), p. 21. In Baden-Wuerttemberg and Bayern, where voters have long been able to alter lists and cumulate votes, and where, therefore, one might expect such considerations to be especially relevant, the void list vote has been high, but lower than in several other states where voters are not permitted to alter lists or cumulate votes. Moreover, the data from both states suggest that the more choices a voter has, the more likely he is to cast invalid votes, rather than vice versa. See Schachtner, loc. cit.; and Baden-Wuerttemberg, loc. cit.

²² I am indebted to Dr. Karl Schwarz of the Bundesamt for a photostatic copy of the data on which this table is based. The Bundesamt's own brief analysis appears in Bundesamt: WS 1962(3), p. 146. For a description of the voting machines, see Das Parlament, September 13, 1961, p. 16.

²³ A complete list of Wahlabkommen appears in Erbe, op. cit., pp. 84–85. By means of this device the Bavarian and Center parties in 1953 and the German Party in both 1953 and 1957 managed to circumvent the electoral law's splinter party clauses and obtain representation in the Federal Bundestag.

²⁴ In 1953, 14 of the 20 districts with an invalid first vote exceeding 5% were districts in which alliances had been concluded between two or more of the bourgeois parties. The largest invalid vote was recorded in districts in which the CDU withdrew in favor of one of the smaller parties; in 20 of 22 such districts the invalid first vote exceeded 4%, while that was the case in only 8 of 23 districts in which a smaller party renounced in favor of the CDU. Notice that 27 of the 39 Wahlabkommen were in Niedersachsen, Hamburg and Nordrhein-Westfalen, which thus explains their unusually high polls of void district votes that year; see above, note 8.

²⁵ In 1957 Wahlabkommen were concluded between the CDU and German Party (DP) in 12 districts, between the SPD and Bavarian Party (FU) in 8 others. In 8 of the 10 districts where the SPD or CDU renounced, invalid votes numbered more than 7%, ranking first through eighth among German election districts. In each of the other two, void district votes exceeded 5%. For an analysis of the cross-pressures impinging on the SPD and FU electorates, and the resultant high incidence of first vote abstentions especially among SPD voters, see the local study of Munich-Land by Keith Panter-Brick, in Kitzinger, , German Electoral Politics, pp. 338–339 Google Scholar.

²⁶ Bundesamt: WS 1962(3), p. 147.

²⁷ If the CDU's Bavarian sister-party, the CSU led by Franz Josef Strauss, should follow through on its apparent 1965 Wahlabkommen with the small right-wing Bavarian GDP (whereby the GDP is to “deliver” the right-wing vote to the CSU, and the CSU is to guarantee the GDP two direct mandates and three list mandates, thus permitting it to reenter the Bundestag; see Die Zeil, 01 22, 1965, p. 3 Google Scholar), void district votes by disenchanted CSU voters could contribute to a relatively higher invalid first vote in particular districts.

In the absence of Wahlabkommen, however, the downward trend of void district votes may even accelerate as minor and splinter party voters make the transition to voting for the major parties. In the past, these voters have often cast invalid first votes in districts where their party ran no candidate. For example, in 1953 special counts revealed that 25% of the DRP voters in Rhineland-Pfalz and over 11% of the Center Party's voters in Nordrhein-Westfalen invalidated their first vote because they had no opportunity to vote for a DRP or Center Party candidate in their district; see Hirsch-Weber and Schuetz, op. cit., p. 330 and p. 333. For 1957 see the data and analysis in Bundesamt: 1957(3), p. 53. Representative of such districts in 1961 were 153 in Rhineland-Pfalz and 198, 211, 213, 218, 221, 230, 232 and 239 in Bayern; see Bundesamt: 1961(1).

²⁸ See especially districts 196, 198, 199, 205, 208, 211, 212, 213, 217, 224, 226 and 228. Bundesamt: 1949, p. 24; 1953(1). In noting that voters did not vote invalidly in 1953 as often as had been expected owing to introduction of the new two-vote, two-column ballot, James K. Pollock implies that voter ignorance of the technical aspects of vote-casting was particularly minimized in Bayern, where the voters already had experience with a comparable electoral system. Pollock, James K., ed., German Democracy at Work: A Selective Study (Ann Arbor, Michigan, University of Michigan Press, 1955), p. 88 and p. 193 Google Scholar, note 11. Were voter ignorance actually the primary cause of invalid votes in the Federal Republic, however, Bayern should have recorded no significant increase from 1949 and should have polled among the fewest invalid votes of any of the German states. In fact, it registered a larger increase in 1953 than any other state and ranked third in per cent of invalid second votes.

²⁹ The relevant figures by district are found in Bremen, Landesamt, Statistisches, Statistische Mitteilungen, 1951, No. 3 and 1953, No. 4Google Scholar; and in Niedersachsen, , Landesamt, Statistisches, Wahlen und Abstimmungen, Reihe F, Vol. 14, Nos. 1 and 3Google Scholar. Detailed information on the SRP may be found in Otto Buesch and Furth, Peter, Rechtsradikalismus im Nachkriegsdeutschland: Studien ueber die ‘Sozialistische Reichspartei’ (SRP) (Berlin and Frankfurt, Verlag Franz Vahlen, 1957), esp. pp. 93 ff., 97 ff., 152 ff., and 156 ff.Google Scholar

³⁰ The areas are included in federal districts 151, 152, 153, 156, 157, 159 and 160. See Rheinland-Pfalz, , Landesamt, Statistisches, Statistische Mitteilungen, 11 1961, pp. 259–266 Google Scholar and May 1959, pp. 99–104. Cf. Cromwell, Richard S., “Rightist Extremism in Postwar West Germany,” Western Political Quarterly, Vol. 16 (06, 1964), pp. 284–293 Google Scholar. A contributing factor in the void list vote increase in Rheinland-Pfalz was probably the merger of the increasingly right-wing Refugee Party with the former German Party to form the All-German Party (GDP). The GDP suffered an 83% drop in 1961 from the combined 1957 strength of the two parent parties; at the same time the per cent of GDP electors casting void list votes jumped to 8.3, four times the party's national average (which in any case was up sharply from 1957). Bundesamt: 1961(3), p. 10 and p. 20; 1957(3), p. 53. On the resentment engendered among former DP supporters by the merger with the Refugee Party, see Eberlein, op. cit., pp. 245–246.

³¹ Invalid list votes increased in the Saar, Baden-Wuerttemberg, Bremen, Hessen, and Nordrhein-Westfalen, invalid district votes in all but Nordrhein-Westfalen. The KPD had polled 3.4% of the vote in Hessen and 3.8% in Nordrhein-Westfalen in 1954, 6.6% in the Saar and 5.0% in Bremen in 1955, and 3.2% in Baden-Wuerttemberg as late as April, 1956. Among states with no appreciable Communist vote, only Schleswig-Holstein recorded a large invalid vote increase in 1957. See the data in Faul, ed., op. cit., pp. 321–363. The unusual parallel increase of invalid first and second votes may have been a reflection of the great regularity of KPD straight-ticket voting (exceeded only by the SPD) in 1953. See Bundesamt: 1953(2), p. 47.

³² Kitzinger, , German Electoral Politics, p. 284 Google Scholar. Moreover, the per cent of SPD electors casting void district votes rose steeply, possibly indicating a less than wholehearted commitment even among those KPD voters who did switch. Bundesamt: 1957(3), p. 53; 1953, loc. cit.

³³ Cf. especially the results in districts 74, 75, 76, 89, 90, 92, 99, 102, 111, 112, 113, 114, 116, 117. Bundesamt: 1957(1); 1953(1).

³⁴ The correlation is particularly clear in the Saar districts of St. Wendel (Federal election district 246) where the KPD had won a plurality in the Saar Landtag election of 1952: i.e., 229, 235, 249, 261, 268, 275, 282. It is noteworthy that Communist strength in postwar Saarland scarcely wavered: in 1955 the party received about the same number of votes it had in 1947 and only 7% fewer than in 1952. See Saarland, Statistisches Landesamt, Statistischer Bericht, B III 1, January 1961; and Kurzbericht, 11/10, 1955 and II/1, 1956. On the character of the German Peace Union (DFU) see Der Spiegel, August 23, 1961; and Wald, Eduard, “Communism,” in Stahl, Walter, ed., The Politics of Postwar Germany (New York, 1963), pp. 293–294 Google Scholar. On Communism and invalid voting in the Saar, cf. also Erbe, op. cit., p. 41.

³⁵ Barnes, et al., loc. cit.

³⁶ Bundesamt: 1961(3), p. 22.

³⁷ Erbe, loc. cit.; Bernhard Beger, “Anatomie einer Wahl, Gewinne und Verluste am 17. September,” Politische Meinung, Vol. 6 (October, 1961), p. 19.

³⁸ Saarland, Landesamt, Statistisches, Statistischer Bericht, B III 1, 01 1961 Google Scholar. Actually, Saarlanders had waged political battles with invalid votes even prior to the 1955 referendum, but then the source of discontent was the obverse of 1957 or 1961. In 1952, for example, almost a quarter of the Saar electorate invalidated their ballots in response to appeals for abstention by the Deutscher Saarbund and the pro-German political parties, which the fearful and mildly authoritarian Saar government had refused to license. See Hagmann, Hans-Joachim, Die Saarlaendischen Landtagswahlen vom SO. November 1962 (Cologne, 1953), pp. 111 ff. and pp. 174 ff.Google Scholar; Schachtner, Richard, Die deutsche Nachkriegswahlen (Munich, Isar Verlag, 1956), pp. 171–173 Google Scholar; Erbe, op. cit., p. 40; and esp. Freymond, Jacques, The Saar Conflict 1945–1955 (New York 1960), pp. 125–128 Google Scholar. Only in 1955, when all claimants competed openly for the first and last time in a postwar Saar election, did the invalid vote amount to no more than a “normal” 1–2%.

³⁹ The most useful compilation of election statistics for the immediate postwar period is that prepared by the Office of the U. S. High Commissioner for Germany, Elections and Political Parties in Germany, 1945–1952 (Bad Godesberg/Mehlen, Germany, 1952), pp. 37–71 Google Scholar. See also the extensive publications of state statistical offices for the state and local elections of this period, to which the bibliography in Bundesamt: 1961(1), pp. 59–67, provides a convenient guide.

⁴⁰ Erbe, lot. cit. On other aspects of anti-Occupation protest voting with invalid ballots in Hessen, see Hessen, , Landesamt, Statistisches, Staat und Wirtschaft in Hessen, Vol. 2, No. 1 (1947), pp. 20–26 Google Scholar; and Vol. 4, No. 1 (1949), p. 13 and p. 15.

⁴¹ Cf. Erbe, loc. cit. Invalid votes in all three areas may also have reflected the relatively greater social unrest in the French Zone caused by severe local economic dislocations resulting from France's harsh reparations, monetary and commercial policies, described in Willis, F. Roy, The French in Germany, 1945–1949 (Stanford, Calif., Stanford University Press, 1963)Google Scholar. For a brief description of local disruptions in Rhineland-Pfalz, cf. Freymond, op. cit., pp. 52–53 and pp. 245–248.

⁴² Scammon, op. cit., p. 505.

⁴³ Comparative data are found in Kitzinger, Britain, Europe and Beyond, op. cit., pp. 183–217. For Germany, see Erwin Faul, “Soziologie der westdeutschen Waehlerschaft,” in Faul, ed., op. cit., pp. 135–315.

⁴⁴ For example: Barnes, et al., loc. cit.

⁴⁵ In 1957 less than half the Divo respondents expressed interest in the campaign; only 4 in 10 could remember having discussed the outcome of the fall election when questioned shortly before it took place; discussion was least frequent among women, older persons, CDU voters, and in outlying areas. Divo-Institut, op. cit., pp. 198–209; 72–75; 268–269. Few voters could perceive any relationship at all between the possible electoral result and their personal lives. ibid., pp. 294–295. Cf. Allensbach Institut, “SPD und oeffentliche Meinung,” Politische Meinung, Vol. 8 (07-August, 1963), pp. 103–105 Google Scholar; and Edinger, Lewis J., “Electoral Politics and Voting in Western Germany,” World Politics, Vol. 13 (04, 1961), pp. 480–481 CrossRefGoogle Scholar.

⁴⁶ Hartenstein, Wolfgang and Schubert, G., Mitlaufen oder Mitbestimmen: Eine Untersuchung zum Demokratischen Bewusstsein und Politische Tradition (Frankfurt, 1961), pp. 43–48 Google Scholar; Hartenstein, W. and Liepelt, K., “Party Members and Party Voters in W. Germany,” in Rokkan, Stein, ed., Approaches to the Study of Political Participation (Special Issue of Acta Sociologica, Vol. 6, Fasc. 1–2, 1962), pp. 43–52 Google Scholar; Sidney Verba, “Political Participation and Strategies of Influence: A Comparative Study,” ibid., pp. 22–42.

⁴⁷ Lohmar, U., Innerparteiliche Demokratie (Stuttgart, F. Enke Verlag, 1963), pp. 26–34 Google Scholar. Cf. Otto Kirchheimer, Appendix to his chapter on “Conditions of a Stable Opposition: Germany,” in Dahl, ed., op. cit.

⁴⁸ Cf. the remarks in Kirchheimer, Otto, “German Democracy in the 1950's,” World Politics, Vol. 13 (10, 1960), pp. 256–257 Google Scholar.

⁴⁹ That is, in coercive plebiscitarian voting (which particularly mobilized women and residents of village and rural areas) or in the circumspect habits acquired under the watchful tutelage of foreign military authorities. On the former see Faul, op. cit., pp. 150–151 and p. 159; on the latter see Hirsch-Weber and Schuetz, op. cit., p. 160.

⁵⁰ In one recent poll 52% of the respondents felt that even persons who were politically totally uninterested should vote; 39% considered voting a formal duty. Hartenstein and Schubert, op. cit., p. 37. For 1957 data, see Divo-Institut, op. cit., p. 295 and Table 159. Less than a quarter of the Divo respondents cited ideological or pragmatic political reasons for voting. This relationship between sense of duty and voting turnout is not necessarily a uniquely German trait; Campbell, Angus, et al. , The American Voter (New York, 1960), p. 106 Google Scholar, report a similarly close correlation between the two factors for the 1956 U. S. Presidential election. Nevertheless, as Almond and Verba have clearly demonstrated, this combination of high frequencies of formal participation and a passive orientation to that participation is an attitude and behavior pattern that reaches far more deeply into the German political culture than it does the American or British, with important consequences for democratic politics. See Almond, Gabriel A. and Verba, Sidney, The Civic Culture (Princeton, N. J., Princeton University Press, 1963)CrossRefGoogle Scholar. Cf. Verba, op. cit.; and the same author's excellent chapter, “Germany: The Remaking of Political Culture,” in Pye, Lucian and Verba, Sidney, eds., Political Culture and Political Development (Princeton, Princeton University Press, 1965)CrossRefGoogle Scholar.

⁵¹ Others, more knowledgeable in at least the technique of voting, perhaps felt competent to choose between national parties and programs but unable to decide between local personalities who were often unknown to them; according to Kitzinger, many of these voters no doubt deliberately allowed the left-hand side of their ballot paper to remain blank. Kitzinger, German Electoral Politics, p. 288.

⁵² Divo-Institut, op. cit., pp. 152–161. Sex is probably the most striking variable: in late 1963 only 15 % of the women (but over 50 % of the men) interviewed by the Allensbach Institut expressed an interest in politics. See “Umfragen: Interesse fuer Politik,” Politische Meinung, Vol. 8 (12, 1963), pp. 84–88 Google Scholar.

⁵³ Lipset, op. cit., p. 229.

⁵⁴ See the analysis of rural invalid voting in Bavarian elections by Schachtner, Richard, “Wahlberechtigte, Wahlbeteiligung, Nichtwaehler und Falschwaehler,” Bayern in Zahlen: Monatshefte des Bayerischen Statistischen Amtes, Vol. 4 (08, 1950), pp. 402–403 Google Scholar; cf. Hessen, , Landesamt, Statistisches, Staat und Wirtschaft in Hessen, Vol. 4 (02, 1949), p. 15 Google Scholar; and Faul, op. cit., p. 160. The practice of reporting election results by precinct even in sparsely populated rural areas probably tends also to increase the general social pressure to vote; cf. Kitzinger, , German Electoral Politics, pp. 35—36 Google Scholar. Concerning the factors which generally promote a combination of high electoral participation and comparatively low interest in politics in rural areas, see Linz, Juan, The Social Bases of West German Politics, Ph.D. dissertation, Columbia University, 1959 Google Scholar; Microfilm Publication #59-4075 (University Microfilms, Inc., Ann Arbor, Michigan), pp. 731–790.

Minor and radical political parties were supported chiefly by older and middle-aged voters (especially men) in rural and village areas. Bundesamt: 1957(3), pp. 46 ff. and p. 53. Significantly, the least decrease in paired invalid votes has occurred among older and middle-aged voters (esp. men); invalid list votes have been most numerous and have increased most rapidly among older voters; while the only increase of invalid first votes has occurred among the elderly and among middle-aged men in rural areas. The slow increase in the percentage of major party electors who cast valid district votes but void list votes also is indicative of the relationship between invalid voting and minor party political orientation; moreover, the increase is faster for the CDU, which absorbed most of these minor party switchovers, than for the SPD. (See above, text and references on “General Patterns” of invalid voting). Another suggestive relationship is the following: the spatial distribution of radical political party strength in postwar Germany has tended to be concentrated along roughly the same belt where the Nazis formerly polled best and where the invalid vote tended later to be highest. Cf. Faul, op. cit., p. 159.

One final remark concerning the mobilization of these peripheral voters may be relevant: in many countries the habitual, uninterested or disaffected nonvoter-type simply stays home; when forced to vote (in Belgium or Luxembourg, or in the cantons of Switzerland or states of Austria where voting is compulsory) he may vote invalidly. In Germany social compulsion replaces legal compulsion but the effect is the same: non-voters vote invalidly. Cf. above, note 3.

⁵⁵ Cf. some of the data in Habermas, J., et al. , Student und Politik: Eine Soziologische Untersuchung zum Politischen Bewusstsein Frankfurter Studenten (Neuwied: Hermann Luchterhand Verlag, 1961)Google Scholar, summarized in part by Tauber, Kurt, “Nationalism and Social Restoration: Fraternities in Postwar Germany,” Political Science Quarterly, Vol. 78 (03, 1963), pp. 83–85 CrossRefGoogle Scholar; Peter Schoenbach, “German Reactions to Anti-Semitic Incidents—A Sociological Study,” in Stahl, ed., op. cit., pp. 329–331. See also HansHelmuth Knuetter, “Extreme-Rightism,” ibid., pp. 213–228; and Jenke, Manfred, Verschwoerung von Rechts? Ein Bericht ueber den Rechtsradihalismus in Deutschland nach 1945 (Berlin, Colloquium Verlag, 1961)Google Scholar.

⁵⁶ Note particularly the sharp 14% drop in Krefeld from 1957 and 15% drop in Dortmund from 1953. See above, Table IV.

⁵⁷ For the basic data, see Faul, ed., op. cit., pp. 321–363; for the extensive literature on sample surveys conducted by state statistical offices, see Bundesamt: 1961(1), pp. 59–67. Cf. Kitzinger, , German Electoral Politics, pp. 291–295 Google Scholar. There are of course additional reasons for lower participation rates, as well as for the relatively poorer showing of the CDU; see Edinger, op. cit., pp. 436–437; and Erbe, op. cit., pp. 36–39. However, the reason cited seems to be decisive for the lower invalid vote in most cases.

⁵⁸ For a concise survey of most of the minor and splinter party groups which have appeared since 1945, see Erbe, op. cit., pp. 52–63. Cf. above, the material cited in notes 29, 30, 31, 38; such examples could be multiplied many times over. However, in many cases at the lower levels of self-government, a narrowed range of political choice again obtains and the invalid vote increases: frequent failure of the major parties to nominate candidates for local office in small towns and villages deprives many voters of the opportunity to vote for the party of their choice—in this case usually CDU or SPD—and they respond to that situation by depositing a blank or crossed-out ballot in protest. As a result, the invalid vote in peripheral areas at local elections often exceeds that cast in these same areas in federal elections. For example, in the 1960 communal elections in Hessen, the SPD put up a party list in just 900 of the 2600 communities, the CDU in 357, the BHE in 388 and the FDP in 78. Invalid votes varied directly with the number of parties competing in a given commune: the more parties, the fewer invalid votes, and vice versa. In the larger, kreisfreie cities—those detached governmentally and administratively from the county in which they are situated—all the parties put up lists of candidates; there the invalid vote was just 1.3%. The neglected areas were mainly the smaller of the kreisangehoerige towns and villages in less accessible rural areas; in these the invalid vote averaged 6.0%, depending on which party or combination of parties failed to present candidates. In communes where the parties presented lists for the first time in 1960, there was a noticeable retrenchment in the percentage of invalid votes cast in 1956. Hessen, Statistisches Landesamt, Beitraege zur Statistik Hessens, No. 124; and Staat und Wirtschaft in Hessen, Vol. 15 (10, 1960), pp. 253–256 Google Scholar.

⁵⁹ A classic example was that of the commune of Rockensuess (Kreis or County Rotenburg), in the 1952 Hessen communal elections, where nearly 55% of the ballots deposited were invalid. Close examination revealed that most of these had been cast in a single polling district, the neighborhood of Cornberg (where 82.7% of the voters spoiled their ballots); Cornberg was campaigning for detachment from Rockensuess and status as a self-governing commune. Hessen, Statistisches Landesamt, Beitraege zur Statistik Hessens, No. 52. A number of personal and local factors in particular Saar communes in 1955 also provide instructive insight; they are described in Freymond, op. cit., pp. 252–254. Cf. especially the case of Webenheim. Schachtner's two articles on invalid voting in Bayern in the early 1950s (see above, notes 16 and 54) are also valuable, particularly for their detailed differentiation of urban and rural patterns.

⁶⁰ This can even be stated as a rule: wherever there is a perceived constriction of political choice, some invalid votes will be cast by those who feel deprived. In addition to the data on German federal, state and local elections presented in this paper, evidence in support of this rule may be found in the first (comparatively less regimented) Nazi and East German elections, in which invalid votes were often unusually frequent, and in some communes even more numerous than valid votes e.g., at the East German communal elections of 1946); in the Italian elections of 1953, cited in footnote 3; and in occasional single-member constituency elections in Switzerland in which only one candidate has been nominated. See: Germany, Reichsamt, Statistisches, Statistisches Jahrbuchfuer das Deutsche Reich, Vol. 53 (Berlin, 1934), p. 550 Google Scholar; Bundesministerium fuer Fragen, Gesamtdeutsche, Die Wahlen in der Sowjetzone: Dokumente und Materialien (5th ed.; Bonn and Berlin, Deutscher Bundes-Verlag, 1963)Google Scholar; Scammon, op. cit., pp. 526–531; and Kitzinger, Uwe, “Swiss Electoral Democracy,” Parliamentary Affairs, Vol. 13 (Summer, 1960), p. 339 Google Scholar.

⁶¹ Niemoeller, a neutralist who opposed German rearmament, was formerly director of the bureau of foreign affairs of the EKD (Evangelische Kirche in Deutschland), and president of the provincial evangelical church of Hessen-Nassau. For Niemoeller's appeal and reasons for casting invalid votes, as well as the sharply critical response of the press and leading politicians, consult: Frankfurter Allgemeine Zeitung, December 30, 1964 and January 4, 1965; Christ und Welt, January 8, 1965; Die Zeit, January 15, 1965.

⁶² Although we have stressed the dominance of older age groups, it should be pointed out that more than 4.0% of the voters in each age group of both sexes failed to cast two valid votes in both 1957 and 1961. See above, Table III.

⁶³ Cf. Kirchheimer, , “German Democracy in the 1950s,” pp. 262–265 Google Scholar.

⁶⁴ Verba, “Germany: The Remaking of Political Culture,” op. cit.

⁶⁵ In this connection, see the provocative discussion in Ulf Himmelstrand, “A Theoretical and Empirical Approach to Depoliticization and Political Involvement,” in Rokkan, ed., op. cit., pp. 83–110, and esp. p. 110; and in Lohmar, op. cit., pp. 28–31.

⁶⁶ For an intriguing letter-to-the-editor by a disenchanted German citizen, proposing such survey questions as a means of both exercising political opposition and analyzing it, see Christ und Welt, January 8, 19G5, p. 11. Unfortunately, the Bundesamt has implied its endorsement of vote-counting measures which would obfuscate rather than clarify invalid vote patterns. It suggests that a more realistic appraisal of the number of invalid votes could be gained if separate ballot papers were used for first and second votes, with the voter free to deposit one or both ballots as he sees fit. The two ballots would then be tallied separately. Abstentions would thus appear statistically as lower turnout rather than as deposited blank (invalid) votes. See: Bundesamt: WS 1962(3), p. 145. No doubt, this would be a convenient device for reducing the per cent of invalid votes, but at the expense of making it even more difficult to assess the possible political significance of invalid votes and their combinations on the ballot papers.