https://orcid.org/0000-0001-8246-739X
![Sample our Area Studies journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5922/TOC-Subject-Sample-2021-Area-Studies.jpg"})
ABSTRACT
To consolidate a predominant party system, an incumbent party will attempt to anchor voting behavior to social cleavages, a strategy called cleavage enclosure. However, does this strategy actually work? In Turkey, the incumbent AKP government has focused its campaigning on conservatives and nationalists. The analysis of the 2018 post-election survey reveals that the cleavage enclosure worked for conservatives but not for nationalists. Of the incumbent supporters in the previous election, conservatives replicated their support, whereas nationalists were less likely to support the incumbent than other identity holders. Nationalists tend to punish but not reward the incumbent party for its economic performance.
Disclosure statement
No potential conflict of interest was reported by the author.
Notes
1 A PPS is not rare in democracies. During the period from 1975 to 2010, 37 out of 75 consolidated democracies, or nearly half, have had a PPS. See Nwokora and Pelizzo, “Sartori Reconsidered.”
2 Sartori, Parties and Party Systems. We exclude the period between June and November 2015: although the incumbents lost their legislative majority in the June election, it was regained by the re-election in November.
3 Pempel, Conclusion: One-party Dominance, 344. In fact, Pempel analyzed one-party democracies, not predominant party systems. However, because the two concepts share many features, this paper discusses Pempel’s theory in the context of predominant party systems.
4 Pempel, Conclusion: One-party Dominance, 333–60.
5 Roberts, Changing Course in Latin America, 24.
6 We define conservatism in Turkey, as per Çarkoğlu and Toprak (Religion, Society and Politics) and Carkoglu and Kalaycıoğlu (The Rising Tide of Conservatism), as a moral framework that advocates Islamic and other traditional values.
7 While it would be appropriate to analyze the current Turkish political system in terms of authoritarianism, its systemic decay coincides with the dominant party’s attempt to consolidate the PPS by cleavage-based polarization. Keyman, “The AK Party”; Somer, “Turkey: The Slippery Slope”; and Wuthrich, National Elections in Turkey, 241–53. For the growing authoritarian tendency of the Turkish political system, see Sözen, “Competition in a Populist Authoritarian Regime”; Tansel, “Authoritarian Neoliberalism”; Esen and Gumuscu, “Rising Competitive Authoritarianism”; Özbudun, “Turkey’s Judiciary”; Kemahlıoğlu, “Winds of Change?”; Özbudun, “AKP at the Crossroads”; Erişen and Kubicek, Democratic Consolidation; and Somer, “Conquering versus Democratizing”. The Turkish political system was downgraded to the “not free” global freedom status in 2017 according to Freedom House.
8 Although the transition from the parliamentary system to the (nominally) presidential system in 2017 might have reduced incentive for the supporters of the incumbent (president) to vote for the incumbent party in the legislative election, Turkey’s new presidentialism, which lacks checks and balances, strongly favors the incumbent party in elections.
9 It could be possible to describe the current Turkey as a PPS under low electoral integrity if we apply the slightly expanded definition of PPS by Nwokora and Pelizzo, who replaced “one party” with “one party or the same coalition” in the PPS definition by Sartori. See Norris and Grömping, Electoral Integrity Worldwide, 6, ; Nwokora and Pelizzo, “Sartori Reconsidered,” 834.
10 Nwokora and Pelizzo, “Sartori Reconsidered,” 834.
11 Pempel, Conclusion: One-party Dominance, 333–60.
12 Muramatsu and Krauss, Dominant Party, 300.
13 Esping-Andersen, Single-party Dominance in Sweden, 48–55.
14 Aronoff, “Israel under Labor,” and Shalev, “The Political Economy,” 123.
15 Nwokora and Pelizzo, “Sartori Reconsidered,” 824–42. Only eight of the total 37 PPSs achieved six consecutive majority wins in legislative elections, which are defined as continuous dominance by Nwokora and Pelizzo.
16 Pempel, Conclusion: One-party Dominance, 333–60.
17 Lipset and Rokkan, Cleavage Structures. See also Roberts, Changing Course in Latin America, 24, and Bartolini and Mair, Identity, Competition.
18 Nannestad and Paldam, “The VP-Function,” 213–45; Lewis-Beck and Stegmaier, “The VP-Function Revisited,” 367–85.
19 Pempel, Conclusion: One-party Dominance, 333–60.
20 Bartolini and Mair, Identity, Competition.
21 Roberts, Changing Course in Latin America, 24.
22 Arceneaux, “Conditional Impact of Blame Attribution”; Weschle, “Two Types of Economic Voting”; and Tillman, “Economic Judgments.”
23 Tillman, “Economic Judgments,” and Weschle, “Two Types of Economic Voting.”
24 Çarkoğlu, “Economic Evaluations vs. Ideology”; Gidengil and Karakoç, “Which Matters More ”; Baslevent, Kirmanoglu, and Senatalar, “Party Preferences”; Kalaycıoğlu, “Justice and Development Party”; Çarkoğlu, “A New Electoral Victory”; and Öniş, “Conservative Globalism.”
25 Çarkoğlu, “Ideology or Economic Pragmatism?” and Baslevent, Kirmanoglu, and Senatalar, “Party Preferences.”
26 Kemahlıoğlu, “Winds of Change?” and Aytaç and Çarkoğlu, “Terror Attacks.”
27 Gumuscu, “Emerging Predominant Party System”; Ayan Musil, “Emergence of a Dominant Party Syste”; and Wuthrich, National Elections in Turkey, 246–7.
28 Yeşilada and Noordijk, Religiosity and Political Values.
29 Saraçoğlu and Demirkol, “Nationalism and Foreign Policy”; Özpek and Yaşar, “Populism and Foreign Policy”; Başkan and Güney, “Turkey's June 2011 Parliamentary Elections”; and Cengiz and Hoffmann, “The 2011 General Elections in Turkey.”
30 Kesgin, “Turkey’s Erdoğan.”
31 Aytaç and Çarkoğlu, “Terror Attacks.”
33 It is very likely that a surge in nationalist identity and a temporary decline in nationalist support for the AKP, which took place between April and October 2017, stems from the establishment of the new nationalist party, Good Party (İYİ Parti, İP), in October 2017 by former MHP parliamentarians. It is more important, however, that the AKP quickly recovered nationalist support apparently at the expense of the MHP.
34 Keyman, “The AK Party”; Müftüler-Baç and Keyman, “Era of Dominant-Party Politics”; Çınar, “Electoral Success”; ans Wuthrich, National Elections in Turkey, 241–53. More generally, McCoy and Somer argue that polarization of society between “Us and Them” camps arises not from social structural or institutional characteristics but polarizing strategies such as divisive discourse by political entrepreneurs. McCoy and Somer, “Toward a Theory.”
35 The author would like to greatly thank Metropoll for sharing its valuable dataset.
36 For voting choices in the 2018 general election, Metropoll asked respondents which electoral alliance and/or political party they voted for. The electoral law, amended prior to the 2018 election, allows one to vote for an electoral alliance or a political party even if the party is in an electoral alliance. The dataset shows that all but two respondents who identified the electoral alliance they voted for indicated the party they voted for as well. This means that, although some respondents might have voted for an electoral alliance, they had a clear party preference in that alliance.
37 The weighted rates of abstention and non-response for the 2018 election are 13.6 and 0.9 percent, respectively. The actual abstention rate was 13.8 percent in the 2018 election according to the Supreme Election Council (http://www.ysk.gov.tr/tr/secim-arsivi/2612).
38 A more practical justification for the long-term models is the very strong party identity of the AKP supporters. Of the AKP supporters in the 2015 November election, 89.4 percent voted for the AKP in the 2018 election. Using the vote for the AKP in 2018 as the dependent variable, and the vote for the AKP in 2015 as a control variable, would seriously reduce the explanatory power of the major independent variables.
39 The “other” category consists of “Center-right/secular” and “Kurdish nationalist.”
40 In parentheses are sample weighted percentages.
41 Alvarez et al., “A Taxonomy of Protest Voting,” 144–7.
43 The weighted abstention rate is 14.2 percent, whereas the actual abstention (non-turnout) rate was 14.8 percent in the November 2015 election.
44 We used the Contrast command (for orthogonal polynomial contrasts) of STATA, as was suggested by Jeph Herrin April 2014 Posts: 191, http://www.statalist.org/forums/forum/general-stata-discussion/general/1335688-regression-with-continuous-dependent-variable-with-ordinal-independent-variables.
45 Paolini and McIntyre, “Bad Is Stronger Than Good”; Frijda, The Emotions; and Fazio, Eiser, and Shook, “Attitude Formation.” Valence asymmetry also accounts for the prospect theory that individuals are more loss-averse than gain-seeking, as well as the grievance asymmetry hypothesis, that voters are more likely to attribute economic failure than success to the incumbent (Kahneman and Tversky, “Prospect Theory”; Stevenson, The Economy as Context; Nannestad and Paldam, “Grievance Asymmetry Revisited”; and Park, “Punishing Without Rewards?”).
46 Stevenson, The Economy as Context; Nannestad and Paldam, “Grievance Asymmetry Revisited”; Park, “Punishing without Rewards?”; Reidy, Suiter, and Breen, “Boom and Bust”; and Kappe, “Asymmetric Evaluations,” 133–8.
47 The left-right index for electoral manifestos (with negative/positive values indicating left/right) was obtained from Volkens et al., “The Manifesto Data Collection.”
48 Agnew, Place and Politics.
49 Linke and O’Loughlin, Spatial Analysis, and Çarkoğlu and Kalaycıoğlu, Turkish Democracy Today, 177–87.
50 We thus followed the estimation model by Çarkoğlu and Kalaycıoğlu that used NUTS1 as the spatial unit. Çarkoğlu and Kalaycıoğlu, Turkish Democracy Today, 177–87.
51 The full estimation results of the logit model are presented in Appendix 3.
52 The estimated probability for conservatives is larger, although not significantly, than that for other identities, as indicated by the positive but insignificant coefficient for conservatives (0.155) in Model 2.
53 In and , previous incumbent support is indicated by ‘Incumbent’ in ‘Vote choice 2015.’
54 Overlapping confidence intervals do not necessarily indicate that the difference between the two estimates is not statistically significant.
55 This was against the backdrop of the eruption of terror attacks in mid-2015, as stated earlier. See Aytaç and Çarkoğlu, “Terror Attacks.”
56 Stanig, “Political Polarization.”
57 The author thanks one of the referees for this perspective.
Additional information
Funding
Notes on contributors
Yasushi Hazama
Yasushi Hazama is a senior research fellow at the Institute of Developing Economies, Japan External Trade Organization. He received his MA in Public Administration from Middle East Technical University and Ph.D. in Political Science from Bilkent University, Ankara, Turkey. His research interests include political behavior and public opinion. He published in International Political Science Review, Acta Politica, Turkish Studies, and Developing Economies.