Published online by Cambridge University Press: 29 January 2009
Over the last twenty years there have been several interesting attempts by scholars working both in Turkey and elsewhere to offer a critical examination of the workings of the competitive party system in Turkey, and to relate changes in the parliamentary arena to the tensions created in a rapidly developing society. This emphasis on the social and economic background to political change has tended to turn attention away from the constitutional and legal arrangements on which the proper functioning of parliamentary democracy also depends. What follows tries to assess the impact of one of these factors — the electoral system — on Turkish political development since 1950. It closes with some suggestions about the possible effects on Turkish politics of some hitherto untried systems of election.
1 Duverger, Maurice, L'influence des systèmes electoraux sur la vie politique (Paris: Armand Cohn, 1950), p. 11.Google Scholar
2 For a discussion of earlier Turkish electoral practice see Turkiye'de ye Yabanci Memleketlerde Seçim Mevzuati (Ankara: Balbakanlik Deviet Matbaasi, 1949),Google Scholar and Frey, Frederick W., The Turkish Political Elite (Cambridge, Mass.: M.I.T., 1965), pp. 423–437.Google Scholar Detailed information on the original provisions of 1876 appears in Devereux, Robert, The First Ottoman Constitutional Period (Baltimore: Johns Hopkins Press, 1963), pp. 123–153.Google Scholar The first full Ottoman electoral Law, which was discussed in the parliament of 1877–1878 but not implemented until after the 1908 revolution, was published in English by the British authorities in Iraq as Ottoman Law for the Election of Deputies of Parliament (Baghdad: 1920). Condensed versions of the texts of the 1942 and 1950 electoral laws are given by Davis, Helen Clarkson, Constitutions, Electoral Laws, Treaties of State in the Near and Middle East (Durham, N.C.: Duke University Press, 1947 and 1953).Google Scholar
3 Article 3 of the 1950 Electoral Law (quoted in ibid., 1953, p. 467).
4 Nevertheless, this did occur in nine cases in 1950, in three in 1954 and in two in 1957 (calculated from 14 Ekim 1973 Milletvekili Seçimi Sonuçlari [Ankara: Başbakanlik Devlet Istatistik Enstitüsü, 1974], p. 18: hereinafter, 1973 Election Results).Google Scholar
5 Cumhuriyet Senatosu Üyelerinin Seçimi Kanunu, no. 304, 24 05 1961,Google Scholar and Milletvekili Secimi Kanunu, no. 306, 25 05 1961.Google Scholar
6 All the members of the lower house are directly elected. The original membership of the senate was 188: of the nonelected senators, fifteen are chosen by the President and the remaining 23 were ex-members of the National Unity Committee (the military rulers of 1960–61,) as of 13 December 1960. Death has since thinned their ranks. Ex-Presidents of the Republic are also cx officio senators.Google Scholar
7 If none of the parties succeeded in clearing the barrage, then this provision was ignored: seats were simply distributed to the parties according to the procedure outlined below (Law no. 306, Art. 32, paras 4 and 5).Google Scholar
8 Hypothetical, because, unfortunately, none of the actual 1961 constituency returns demonstrate all the points which are worth noticing.Google Scholar
9 Istanbul province returned ten senators, Ankara seven, Izmir six, Konya five and Adana, Sivas, Bursa, Balikesir and Samsum four each. The electoral districts into which these provinces were divided were allotted a maximum of three senators each (Abadan, Nermin, 1965 Seçimlerinin Tahlili [Ankara University, Political Science Faculty, 1966], p. 87 n. 34).Google Scholar
10 Secimlerle ilgili Kanunlarin bazi Hükümlerinin Değiştirilmesi hakkinda Kanun, No. 533, 13 02 1965.Google Scholar
11 See Dodd, C. H., Politics and Government in Turkey (Manchester University Press, 1969), pp. 182–183,Google Scholar and Soysal, Mümtaz, Anayasaya Giris, 2d ed. (Ankara University, Political Science Faculty, 1969), p. 258.Google Scholar
12 Calculated from 1973 Election Results, pp. 6–7.Google Scholar
13 In fact, the proportion of their seats which the smaller parties won from the pool was as follows: Nation Party, 25 out of 31: New Turkey Party, 16 out of 19, Turkish Worker's Party, 13 out of 15: Republican Peasant's Nation Party, 11 out of 11 (Landau, Jacob M., Radical Politics in Modern Turkey [Leiden: Brill, 1974], p. 270).Google Scholar
14 Siyasi Partiler Kanunu, No. 648, 13 06 1965, Arts. 29–52.Google Scholar See also Dodd, Politics and Government in Turkey, p. 184, and Abadan, 1965 Seçimlerinin Tahlili, pp. 60–62 and 339–343. According to the original wording of Article 29 of this law, the party slate in each province was to be “established by means of a primary election in which all party members registered on the party electoral roll in each constituency may participate.” In practice, the parties could restrict this definition to local party functionaries. For example, the rules of the Reliance Party (1968) extend the right to vote in primaries to village headmen and local councillors who are members of the party, district party committeemen, and so on (Güven Partisi T.B.M.M. Üyeliäi Adaylik ye Ön Secim Yöneuneliği [Ankara: Reliance Party, 1968], pp. 4–6). In 1973 this practice was regularised by amending legislation (Law No. 1783 of 26 June 1973) which altered Article 29 so as to provide that the slate would be determined “according to principles laid down in the directions and standing orders of the political parties themselves.” During 1976, the R.P.P. decided to allow all party members to vote in primaries, but the necessary machinery to put this into effect was not established in time for the 1977 elections. (I am extremely grateful to Dr. Rona Aybay for advice on the legal aspects of this question, and to Professor Turan Güneş for information on the intended changes in the R.P.P.)Google Scholar
15 See Soysal, Anayasaya Giris, p. 260, and Cumhuriyet, 8 May 1968.Google Scholar
16 In the 1965 and 1969 elections the voter had been obliged by the ‘combined ballot paper’ to vote a straight party ticket, in the list order determined by the party primaries. Under the 1973 arrangements he could cast a ‘preferential vote’ in favour of one of the candidates in the list he preferred. Few voters appear to have made use of this refinement in 1973 (indeed, most of them were probably unaware of its existence). So far as I can judge, it had no significant effect on the 1973 election results.Google Scholar
17 Aygen, Cemal, ‘Memleketimizde Seçimler ve Neticeleri’, Ankara Üniversitesi Siyasal Bilgiler Fakültesi Dergisi,17, 1 (03 1962), 213.Google Scholar
18 Ibid., p. 206.
19 Abadan, 1965 Seçimlerinin Tahlili, p. 82.Google Scholar
20 In making this rough calculation, I have ignored the shorter intervals between the resignation of one government and the formation of its successor which are inevitable in any democracy.Google Scholar
21 The exception being that used for the 1961 senate elections described earlier.Google Scholar
22 See Karasüleymanog¯lu, Aydin, ‘Seçim Sistemi üzerinde Hikmet Sami Türk ile bir Konuşma’, Halkoyu, 10, 2 (05 1976), 56.Google Scholar
23 The cities of Ankara and Adana both contain ilçes with populations higher than those of many provinces. It was not possible to devise a hypothetical division of these ilçes without a detailed investigation on the spot, so they had to be left as single constituencies returning nine and seven members, respectively.Google Scholar
24 S.B.F. İdari İlimler Enstitüsünün Gerekçeli Anayasa Tasarisi ye Seçim Sistemi Hakkindaki Görüşü (Ankara University, Political Science Faculty, 1960), pp. 90–92.Google Scholar
25 Abadan, Nermin, ‘1960 S.B.F. Anayasa Seminerlerinde Beliren Esas Düşünceler,‘ Ankara Üniversitesi Siyasal Bilgiler Fakültesi Dergisi, 17, 2 (06 1962), 262–266.Google Scholar
26 Weiker, Walter F., The Turkish Revolution, 1960–1961 (Washington, D.C.: The Brookings Institution, 1963), pp. 163–164.Google Scholar
27 Once again, abnormally large ilçes could not be subdivided. In these cases, two or more hypothetical constituencies were lumped together for purposes of calculation, and the likely results estimated according to a variety of mathematical formulae. The results quoted must thus be regarded as no more than a reasonable estimate.Google Scholar
28 The problem would arise most acutely in the case of the Kurds who represent between 5 and 10 percent of Turkey's total population but are in a majority in some southeastern provinces. Under present circumstances, Kurdish separatism has little overt support in Turkey, but it might gain strength if a more localised system of election were introduced. For a discussion of this point see Mango, Andrew, Turkey: A Delicately Poised Ally, The Washington Papers, Vol. 1, no. 8 (Beverly Hills and London: Sage Publications 1957), pp. 51–56.Google Scholar