Turkey’s Electoral Geography: Trends, Behaviors, and Identities 2020043674, 2020043675, 9780367632793, 9781003118626, 9781000351248, 9781000351262, 9781000351286, 9780367632830

Citation preview

Turkey’s Electoral Geography

Analyzing Turkey’s electoral geography, this volume evaluates the geographical repercussions of the elections in Turkey since the establishment of multiparty politics in 1950. The book focuses on the last two decades, examining the interaction between electoral behavior and regional dynamics. Various issues related to the geographical connotations of Turkish electoral politics are qualitatively and quantitatively addressed by scholars with diverse backgrounds in social sciences. The chapters herein examine how Turkey’s electoral geography has been shaped over the years to correspond with a certain aspect of multiparty politics, such as voting behaviors, political parties and party systems, nationalization and regionalization, redistricting, gender issues, identity dynamics, or ideological polarization. This comprehensive work contributes to the theoretical debates in electoral geography in general. Utilizing notions from electoral geography literature, this book develops new concepts through the Turkish case. Filling an important gap in the literature on Turkish politics, this contemporary analysis will be a key resource to policymakers, students, and scholars interested in political science, Turkey, and the Middle East. Edip Asaf Bekaroğlu earned his PhD in political science at Bilkent University, Turkey in 2010. Since then, he has worked as a full-​time faculty member at İstanbul University’s Department of Political Science and International Relations. His research interests are democratization, political parties and elections, secularism, and multiculturalism. Gülsen Kaya Osmanbaşoğlu earned her PhD in political science at Bilkent University in 2014. She is a faculty member at the Social Sciences University of Ankara, Department of Political Science and Public Administration, and the deputy director of Social Innovation Center (ASBU-​SIM). Her research interests are Turkish politics, political parties and elections, modernization, human rights, and social innovation.

Routledge Studies in Middle Eastern Politics

100 The Iraqi Kurds and the Cold War Regional Politics, 1958–​1975 Hawraman Ali 101 Islam, Jews and the Temple Mount The Rock of Our/​Their Existence Yitzhak Reiter and Dvir Dimant 102 Saudi Interventions in Yemen A Historical Comparison of Ontological Insecurity Caroline F. Tynan 103 Middle Eastern Monarchies Ingroup Identity and Foreign Policy Making Anna Sunik 104 Nationalism and Islamism in the Kurdistan Region of Iraq The Emergence of the Kurdistan Islamic Union Mohammad Salih Mustafa 105 Indigenous Land Rights in Israel A Comparative Study of the Bedouin Morad Elsana 106 Turkey’s Electoral Geography Trends, Behaviors, and Identities Edited by Edip Asaf Bekaroğlu and Gülsen Kaya Osmanbaşoğlu For a full list of titles in the series please visit: www.routledge.com/​middleeaststudies/​series/​SE0823

Turkey’s Electoral Geography Trends, Behaviors, and Identities Edited by Edip Asaf Bekaroğlu and Gülsen Kaya Osmanbaşoğlu

First published 2021 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 52 Vanderbilt Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2021 selection and editorial matter, Edip Asaf Bekaroğlu and Gülsen Kaya Osmanbaşoğlu; individual chapters, the contributors The right of Edip Asaf Bekaroğlu and Gülsen Kaya Osmanbaşoğlu to be identified as the authors of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing-​in-​Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-​in-​Publication Data Names: Bekaroğlu, Edip Asaf, editor. | Osmanbaşoğlu, Gülsen Kaya, editor. Title: Turkey’s electoral geography : trends, behaviors, and identities / edited by Edip Asaf Bekaroğlu and Gülsen Kaya Osmanbaşoğlu. Other titles: Routledge studies in Middle Eastern politics ; 106. Description: Abingdon, Oxon ; New York, NY : Routledge, 2021. | Series: Routledge studies in Middle Eastern politics ; 106 | Includes bibliographical references and index. Identifiers: LCCN 2020043674 (print) | LCCN 2020043675 (ebook) | ISBN 9780367632793 (hardback) | ISBN 9781003118626 (ebook) | ISBN 9781000351248 (adobe pdf) | ISBN 9781000351262 (mobi) | ISBN 9781000351286 (epub) Subjects: LCSH: Elections–Turkey–History–20th century. | Elections–Turkey– History–21st century. | Political parties–Turkey–History–20th century. | Political parties–Turkey–History–21st century. | Political geography. | Turkey–Politics and government–20th century. | Turkey–Politics and government–21st century. Classification: LCC JQ1809.A5 T875 2021 (print) | LCC JQ1809.A5 (ebook) | DDC 324.9561–dc23 LC record available at https://lccn.loc.gov/2020043674 LC ebook record available at https://lccn.loc.gov/2020043675 ISBN: 978-​0-​367-​63279-​3  (hbk) ISBN: 978-0-367-63283-0 (pbk) ISBN: 978-​1-​003-​11862-​6  (ebk) Typeset in Times New Roman by Newgen Publishing UK

Contents

List of figures  List of tables  Notes on contributors  Acknowledgments  1 Introduction to Turkey’s electoral geography: an overview since 1950 

vii viii x xii

1

E D I P AS AF B EK A RO Ğ LU A N D G Ü LSEN K AYA OS MANBA Ş O Ğ LU

2 Measuring party nationalization in Turkey: 1950–​2018 

24

ÖZ H AN D E MI R KO L A N D ED I P A SA F BEK A RO Ğ LU

3 Gerrymandering in Turkish elections since 1950: (re)setting the rules of the game? 

46

G Ü L S E N K AYA O SMA N BA Ş O Ğ LU

4 The dynamics of change and differentiation in voter preferences in the western coastal provinces of Turkey since the 1980s 

66

TAN J U T OSU N, BETÜ L AY D O Ğ A N Ü NA L A N D G Ü LG Ü N E RDO Ğ A N  TO SU N

5 The regional and national campaign discourse of the PDP and its predecessors in Turkey’s general elections, 2002–​2018 

85

HÜSEYIN ALPTEKIN

6 “Where are they doing politics?”: women’s cooperatives as sites of constellations of power  S E L I N AK YÜ Z A N D K Ü R Ş AT Ç I NA R

101

vi Contents

7 Election monitoring organizations in view of electoral geography: regional characteristics of volunteer networks in Turkey 

121

AY Ş E N U R  KI LI Ç

8 Engendering the parliament: a study of women deputies in Turkey, 1935–​2018 

152

MEZHER YÜKSEL

9 An analysis of Turkish overseas electoral districts: opportunities and challenges 

182

N E C AT I A N A Z A N D M E H M E T   K Ö S E

Index 

201

Figures

1 .1 2018 electoral outcomes map  1.2 Electoral outcomes in Erzurum, Gaziantep, Eskişehir, and Van, 1950–​2018  2.1 Party system nationalization and effective number of parties in Turkey (1950–​2018)  2.2 sPNS of the main Turkish parties (1950–​1977)  2.3 sPNS of the mainstream Turkish parties (1983–​2018)  2.4 sPNS of the Islamist, Nationalist and pro-​Kurdish parties (1950–​2018)  3.1 The RPP votes in Turkey and Malatya  3.2 Kırşehir general election results  3.3 Map of Evren and Ҫankaya’s location in Ankara  4.1 Comparison of lifestyle clusters defined by electors themselves in western coastal provinces and in Turkey in general  4.2 Comparison of average votes of major parties in 12 western coastal provinces with their average votes throughout Turkey (1983–​2018)  6.1 Fieldwork map: Provinces in which interviews were conducted  7.1 Distribution of newspapers and media platforms as sources of news followed  7.2 Percentage of trust in media, and trust in election results  7.3 Identities  7.4 Perceived similarity of neighbors, family, and friends’ views  8.1 Women MPs by election period, 1935–​2018  8.2 Women MPs by constituency and birthplace, 1935–​1946  8.3 Women MPs by constituency and birthplace, 1950–​1995  8.4 Women MPs by constituency and birthplace, 1999–​2018  9.1 Participation in Germany  9.2 Consulate base turnout in Germany 

2 6 34 35 37 38 52 54 59 70 75 108 127 129 130 135 153 161 170 172 192 193

Tables

2.1

Measures of stability in Party System Nationalization (1950–​2018)  32 4.1 Average votes of parties in 12 western coastal provinces (1983–​1999)  71 4.2 Average votes of parties in 12 western coastal provinces (2002–​2018)  73 4.3 Comparison of average votes of parties in 12 western coastal provinces with their average votes throughout Turkey (1983–​2018)  74 4.4 The number of provinces parties ranked first (1983–​2002)  77 4.5 The number of provinces parties ranked first (2002–​2018)  77 4.6 Provinces where parties ranked first according to elections in western coastal provinces  79 5.1 HDP’S predecessors in Turkish politics  86 5.2 The prevalent regional and national discourses of the HDP and its predecessors in Turkey’s general elections (2002–​2018)  97 7.1 Thematic classification of survey questions  143 8.1 Women MPs by political current  154 8.2 MPs and women MPs by political party, 1935–​2018  154 8.3 Women MPs by geographical region, 1935–​2018  155 8.4 Number and rate of women MPs by geographical region, 1935–​2018  156 8.5 Women MPs by profession, 1935–​2018  157 8.6 Women MPs by profession, 1935–​1946 (percent)  162 8.7 Women MPs by level of education and non-​Turkish languages they speak, 1935–​1946  163 8.8 Women deputies by profession, 1950–​1995  165 8.9 Women MPs by level of education and knowledge of non-​Turkish languages, 1950–​1995  168 8.10 Women MPs by level of education and knowledge of non-​Turkish languages, 1999–​2018  174

Tables  ix 8 .11 9.1 9.2 9.3 9.4

Women MPs by profession, 1999–​2018  2014 and 2018 presidential elections  2015 general elections  2017 referendum results  2018 general election results 

176 190 191 191 191

Notes on contributors

Selin Akyüz earned her PhD in Political Science at Bilkent University in 2012. She worked as a research fellow and instructor at Bilkent University. She is currently a member of the faculty at TED University, Ankara. Dr. Akyüz’s research interests are gender and politics, masculinity, and Turkish politics. Hüseyin Alptekin earned his PhD in Political Science at the University of Texas at Austin in 2014. Dr. Alptekin is working as Associate Professor of Political Science at İstanbul Medipol University. His research interests are elections, ethnic politics, nationalism, and political violence. Necati Anaz earned his PhD in Geography at University of Oklahoma in 2012. He is Associate Professor of International Relations at İstanbul University. Dr. Anaz’s research interests are political geography, international security, and diaspora politics. Betül Aydoğan Ünal earned her PhD in Public Administration at Dokuz Eylül Üniversity in 2016. She is currently working as a researcher at the Department of International Relations of Ege University. Dr.  Aydoğan Ünal’s research interests are electoral behavior, political campaigns, and electoral geography. Edip Asaf Bekaroğlu earned his PhD in Political Science at Bilkent University, Turkey in 2010. Since then, Dr. Bekaroğlu has worked as a full-​time faculty member at İstanbul University’s Department of Political Science and International Relations. His research interests are democratization, political parties, elections, secularism, and multiculturalism. Kürşat Çınar earned his PhD in Comparative Politics at Ohio State University in 2015. He is currently faculty member at Middle East Technical University. His research interests are political institutions, development, and women’s empowerment. Özhan Demirkol earned his PhD in Political Science at Bilkent University in 2014. Dr. Demirkol is currently working as a full-​time faculty member at Nigde University. His research interests are political parties, elections, and Turkish politics.

Notes on contributors  xi Gülgün Erdoğan Tosun earned her PhD in Public Administration at Dokuz Eylül University in 1998. Professor Erdoğan Tosun teaches at the Faculty of Communication of Ege University. Her research interests are political campaigns, elections, and mass communications. Gülsen Kaya Osmanbaşoğlu earned her PhD in Political Science at Bilkent University in 2014. She is a faculty member at Social Sciences University of Ankara, Department of Political Science and Public Administration, and the deputy director of Social Innovation Center (ASBU-​SIM). Dr. Kaya Osmanbaşoğlu’s research interests are Turkish politics, political parties and elections, modernization, human rights, and social innovation. Ayşenur Kılıç earned her doctoral degree from Bilkent University, Department of Political Science and Public Administration in 2016. She works at Social Sciences University of Ankara. Her research interests range from comparative media and cultural studies, political communication, ideational analysis, sociology of new media, social networking sites, Turkish political life, and political sociology. Mehmet Köse is a PhD candidate of International Relations at İstanbul University. Mr. Köse was the head of the Presidency of Turks and Related Communities (YTB). His research interests are diaspora politics and public diplomacy. Tanju Tosun is Professor of Political Science at Ege University. He earned his PhD at the Department of Public Administration at Dokuz Eylül University in 1997. Professor Tosun has extensively published on electoral behavior, electoral geography, and political parties in Turkey. Mezher Yüksel earned his PhD in Sociology at the Middle East Technical University in 2007. He is a faculty member at Kırıkkale University. Dr.  Yüksel’s research interests are development, political sociology, and gender.

newgenprepdf

Acknowledgments

This book is an outcome of a long journey that dates back to the 2016 MESA Meeting in Boston. In that meeting, we held a panel entitled “Geographical Connotations of Elections in Turkey” under the moderation of Professor Metin Heper. Our foremost thanks go to Professor Heper, of whom we are proud of being students from Bilkent University. We also benefited from detailed feedbacks from Michael Wuthrich, who acted as a discussant in the panel. Since then, many new colleagues have participated in our journey. For our collaborative research, we would like to express our sincere thanks to each of our contributors, who followed a hardworking and self-​disciplined process. They all took kindly to the idea of updating the book when Turkey had a snap election in 2018. We are also very grateful to Routledge’s editors, particularly Joe Whiting, for his support and understanding, and the feedback of two anonymous referees for their invaluable suggestions. Throughout this long process, we usually felt the support of our dearest families. We are very thankful to our families for their love, patience, and understanding. Our children Ayşe, Meryem, and Arın! Yes, we were writing a book! Our heartfelt thanks go to you with the hope of getting the most constructive feedback from you in the near future.

1  Introduction to Turkey’s electoral geography An overview since 1950 Edip Asaf Bekaroğlu and Gülsen Kaya Osmanbaşoğlu

Turkish politics have been characterized by “three maladies” since the 1970s: fragmentation, volatility, and polarization (Özbudun, 2000; 2013). In the 2000s, regionalization, as the fourth malady, became another property of multiparty politics in Turkey. Indeed, any map illustrating the outcomes of general elections since 2002 clearly manifests this trend. For example, the Turkish map in Figure  1.1, which displays the results of the 2018 general elections, is dominated by three regionally concentrated colors. The lightest grey is obviously the prevalent one, representing Adalet ve Kalkınma Partisi (the Justice and Development Party, the JDP), which has formed single-​ party governments in five consecutive elections and transformed Turkey into a “dominant party system”, according to some scholars (Sayarı, 2007; Çarkoğlu, 2011; Ayan Musil, 2015), or even into a “competitive authoritarianism” according to others (Özbudun, 2015; Sözen, 2019). The western coastal provinces in the same map are colored medium grey, representing Cumhuriyet Halk Partisi, (the Republican People’s Party, the RPP), which has been the main opposition party under the JDP rule. Finally, the southeast provinces, with Kurdish majority populations, are shown in dark grey, representing Halkların Demokratik Partisi (the People’s Democratic Party, the PDP), which speaks for the Kurdish political movement. What are the implications of this map? Apparently, a spatial pattern has been in charge among Turkish voters in the 2000s. Although such maps simplify the complex nature of voting behavior, and inferring causal relationships from these maps can be misleading, it is possible to argue that the voters in the western coastal districts are more likely to be motivated by secular sensitivities, the voters in the southeast provinces prioritize their Kurdish identity, and the majority of the voters in the remaining provinces have conservative, religious or nationalist political attitudes. What might explain such a pattern? Is it about regional differences in the associated levels of socio-​economic development? Or is it spatially concentrated identities that determine electoral behavior in Turkey? Do these patterns pertain to the 2000s? Do they have their roots in the multi-​ party politics that have been prevalent in Turkey since 1950? Or is it just a coincidence and temporary fashion? Such questions can easily be diversified for the Turkish case.

2 Bekaroğlu and Kaya Osmanbaşoğlu

Figure 1.1 2018 electoral outcomes map Note: The map was created by the authors based on the electoral outcomes that are available at the Supreme Election Council (Yüksek Seçim Kurulu, YSK)’s electronic archives at www.ysk.gov.tr/​tr/​24-​haziran-​2018-​secimleri/​77536 (2020).

Geography, in any case, is not a frozen and well-​defined territory in political means. Rather, it evolves, transforms, and carries out political meaning throughout time. Even geographical borders may change because of political competition and ambition. However, what does remain unchanged with regard to geography is its ongoing significance in terms of political repercussions. Although students of Turkish politics have accumulated very limited analyses on the geographical patterns of voting behavior, there is an established body of literature about electoral geography concerning American and west European politics. Whether human behavior is conditioned by geography or otherwise is certainly an old question, dating back to Ibn Khaldun’s Muqaddimah (1377) and Montesquieu’s De l’esprit des Lois (1748). Studies on electoral geography, on the other hand, started in the early 1900s and proliferated in the second half of the 20th century with the increasing use of sophisticated statistical analyses in the social sciences. As Forest (2018, p. 2) stated, Andre Siegfied (1913) and Edward Krehbiel (1916) are commonly accepted as pioneers of this field. Krehbiel (1916, p. 419), in his famous study on British parliamentary elections, linked natural and economic factors to geography, which carries an impact on voting behavior. Such earlier studies claimed to establish causality mostly between socio-​economic conditions and voting behavior in particular regions. Today, various versions of this claim are still common among electoral geography studies. They generally argue that the divergence of voter preferences in different regions can be explained by the distinctive composition or social environment of the voters

Introduction: Turkey’s electoral geography  3 in those regions. However, relying only on election outcomes and lacking individual-​level data about voters, these earlier studies remain unconvincing as to whether the relationships they propose were causal or just coincidental. With the proliferation of opinion surveys since the 1980s, proving or rejecting spatial arguments about voting behavior has, however, become increasingly plausible as an approach. Nevertheless, it is still an important question as to whether demography or geography is the more determining factor in voting behavior. Do people with similar demographic characteristics vote in the same way, regardless of regional differences? Or, on the contrary, do people with different demographic characteristics vote in the same way because of certain particularities about their region? With such significant questions in mind, studies into electoral geography are concerned with a broad range of topics such as the spatial leitmotifs, sociological repercussions of particular regions, demographic patterns, class relations, ethnic and religious affiliations, gender dynamics, and beyond. Still, the literature on electoral geography can be categorized into three groups regarding the associated level of analysis:  (1) studies focusing on voter behavior; (2)  studies focusing on political party behavior; and (3) studies focusing on incumbent behavior. As the individual-​ level data about voter behavior flourished with public opinion polls, the electoral geography literature has increasingly dominated studies focusing on the geographical patterns of voter behavior. For instance, Agnew (2014) defines three main trends for the scope of electoral geography: locale, sense of place, and location. In other words, subjective territorial identity cannot be completely separated from micro-​sociological patterns and place (Agnew, 2014, p. 28). So, according to this conceptualization, electoral geography is not a pure and dry concept of place, but instead covers a variety of sociological and psychological attachments attributed to the place. In other words, geography assuredly covers the place, but its political implications refer to more than the place itself. One may dwell on what territory suggests so as to grasp the reverberations of the territory. In his preliminary work, Johnston (1979, p. 36) looks at the issue from a different perspective and notes utilization of territory as a reflection of class conflict. Juxtaposing compositional and contextual approaches such as the influence of economic conditions and individual priorities, along with the neighboring effects of “talking together, voting together”, Johnston and Pattie (2006) developed a comprehensive empirical approach to electoral geography. Flint (2001) also combined economic factors with the political agency, which influenced the electoral geography in Nazi Germany from a world-​systems perspective. These voter-​oriented perspectives are also accompanied by studies analyzing the regionally varied campaigning behaviors of political parties. Besides its national and comprehensive campaign strategy, a political party not only can use its resources differently in different districts but may also change its discourse and methods to communicate with voters depending on the regional dynamics and electoral system (Pattie & Johnston,

4 Bekaroğlu and Kaya Osmanbaşoğlu 2003; Johnston et al., 2002; Forest, 2018). Finally, another important focus of electoral geography is how incumbents change the boundaries of electoral districts to their advantage, popularly known as gerrymandering (Forest, 2012; Webster, 2000). All these particular aspects of the aforementioned studies indicate just how wide the electoral geography issue actually is. The electoral geography literature is, to a large extent, dominated by empirical research and mostly focused on North American and west European cases. The literature suffers from a scarcity of studies dealing with non-​ western politics that may bring new insight into geographical dimensions of electoral behavior. The politics in non-​western democracies are usually based on different and more complicated cleavages than industrial societies. Hence, Turkey’s electoral geography is a compelling case in this regard.

Seventy years of Turkish electoral geography, 1950–​2020 In terms of the electoral geography in Turkish politics, comprehensive studies are still extremely scarce in the literature, apart from a few works that concentrate on electoral choices in particular regions or districts. This book, however, attempts to fill the vacuum in the literature by elucidating the electoral geography of Turkey with consideration for multiple dynamics. While transgressing the boundaries of prevailing cleavage models, this book also intends to augment its quantitative studies with qualitative analyses, adding a nationalization index, cartographic models, and considering issues such as gerrymandering, gender, regional campaigning, or voting from abroad to provide a vivid map of contemporary Turkish politics. In this regard, Turkey, being referred to as a “torn country” (Huntington, 1993, p.  42) with its vast diversity of spatial, ideological, and sociological patterns, constitutes a compelling case for electoral geography studies. Indeed, the politics in Turkey, since the introduction of free and fair elections in 1950, has displayed particular regional dynamics that have intermingled with the socio-​economic, ideological, and identity-​related components of Turkish society. While spatial dynamics appear to have played relatively marginal roles in the initial period of multiparty politics in Turkey, voter preferences have increasingly demonstrated regional patterns with the rise of identity politics, especially since the 1990s. Indeed, this “increasing geographical consistency” of political parties is pronounced by some scholars to be one of the major characteristics of today’s Turkish political party system and elections, together with high levels of electoral volatility and party fragmentation (Çarkoğlu & Avcı, 2002, p.119). However, there is no agreement about the dynamic and mechanisms behind the geographical patterns in political behavior. After Şerif Mardin’s (1973) seminal study, reading Turkish politics as a clash between the modern/​secular center and the traditional/​religious periphery has become widespread. According to this line of work, identity-​ based dynamics (e.g. secular-​religious, Turk-​Kurd, Sunni-​Alavi) overlap with Turkey’s electoral geography (Çarkoğlu & Avcı, 2002; Özbudun, 2006). An

Introduction: Turkey’s electoral geography  5 alternative explanation focuses on the political right and left orientations rather than center-​periphery cleavage (Secor, 2001). Wuthrich (2015), on the other hand, argues that neither the center-​periphery approach nor left-​right analyses, but rather voter pragmatism can explain the electoral and political party behavior in Turkey. These alternative explanations are actually complementary rather than competing. As with elsewhere in the world, different dynamics play a role in shaping electoral behavior in Turkey. For instance, ideologies might sometimes come to the front in a certain district, or at other times identities become prominent. In some cases, ideologies overlap with identities, and in other cases, ideologies or identities are moderated by political pragmatism and rational calculations. Nevertheless, generally speaking, the electoral districts in Turkey can be clustered into four categories in terms of dominant voter types, according to Çarkoğlu and Avcı (2002, pp. 128–​9). The first such cluster is characterized by the far right rather than center-​right and peripheral rather than central voting behavior. The voters in the second cluster, on the other hand, are more likely to support the center than peripheral political parties and are inclined to the center-​right rather than the center-​left. The third cluster is dominated by the center-​left and center-​right, where center-​left tendencies have been in ascendancy since 2002. And finally, the last cluster is dominated by pro-​Kurdish and religious parties. Figure 1.2 contains four graphs that analyze the electoral inclinations of Erzurum, Gaziantep, Eskişehir, and Van, each of which represents the first, second, third, and fourth type of cluster, respectively. For the sake of reducing the complexity of the graphs, we combined center-​right and far right voters into one line, but explain below the significant role of far right parties in their own right. As can be seen in the figure, the voting patterns are indeed clearly distinguishable in these four cities. While in Erzurum right-​wing tendencies have been sharply evident since the beginning of multiparty politics, in Gaziantep the gap between the left and the right opened in the 1980s to the benefit of the latter. In both cities, the presence of the left continuously dropped, but not as much in Gaziantep as in Erzurum. On the other hand, Eskişehir electorates seem largely stable in terms of ideological orientation. Also, in contrast to other cities, the left-​wing tendency is generally firm and consistent in Eskişehir. Finally, the voters in Van have frequently changed their preferences throughout the lifetime of multiparty elections. In spite of this, since the mid-​1990s, left-​wing voters have almost disappeared and right-​wing parties have suffered serious drops in support, while pro-​Kurdish parties have made a strong appearance. Pro-​Kurdish parties also made some appearances in Gaziantep and Erzurum in accordance with the national averages, but interestingly not in Eskişehir. Besides these general tendencies, a closer look at each city is necessary to understand how predominant voter types have evolved in a geographically sensible fashion. To begin with, Erzurum is well known for its conservative and nationalist identity whose right-​wing potential is explicitly superior. In

newgenrtpdf

PRO-KURDISH

57 19 61 19 65 19 69 19 73 19 77 19 83 19 87 19 91 19 95 19 99 20 02 20 0 20 15 2 7 0 20 (JU 11 15 NE (N 7) OV .1 ) 20 18

54

50

19

19

INDEPENDENT

ESKİŞEHİR

80%

19

99 0 20 2 0 20 15 2 7 0 20 (JU 11 15 NE (N 7) OV .1 ) 20 18 20

95

19

91

19

87

19

83

19

77

LEFT

19

73

19

69

65

19

19

61

57

19

19

54

19

50

19

19

RIGHT

GAZİANTEP

80% 70% 60% 50% 40% 30% 20% 10% 0%

RIGHT

PRO-KURDISH

VAN

80%

70%

LEFT

70%

60%

60% 40%

30%

30%

20%

20%

10%

10%

0%

0%

RIGHT

LEFT

PRO-KURDISH

19 5 19 0 54 19 57 19 61 19 6 19 5 69 19 7 19 3 77 19 83 19 8 19 7 91 19 9 19 5 99 20 02 20 20 0 15 2 7 20 (JU 011 15 NE (N 7) OV .1 20 ) 18

50%

40%

19 1950 54 19 57 19 61 19 65 19 69 19 73 19 77 19 83 19 87 19 91 19 95 19 9 20 9 0 20 2 0 20 15 2 7 0 20 (JU 11 15 NE (N 7) OV . 20 1) 18

50%

RIGHT

LEFT

PRO-KURDISH

INDEPENDENT

Figure 1.2 Electoral outcomes in Erzurum, Gaziantep, Eskişehir, and Van, 1950–​2018 Note: Electoral outcomes are available in the YSK’s electronic archives at www.ysk.gov.tr/​tr/​milletvekili-​ genel-​secim-​arsivi/​2644 (2020).

6 Bekaroğlu and Kaya Osmanbaşoğlu

ERZURUM

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

Introduction: Turkey’s electoral geography  7 this sense, the Erzurum electorate is similar in their preferences to a number of other districts in the eastern, northeastern, and central Anatolia and inner-​ Aegean regions. When we consider the Erzurum case in detail, one could argue that from the beginning of the 1960s, as a possible reaction to the relative rise of the leftist movements nationwide, nationalist parties began to grasp a considerable number of votes, as sustained by the rise of Islamist parties in the city. By the 1990s, Kurdish regions such as Karayazı, Karaçoban, and Tekman, located in the southern part of Erzurum, preferred to vote for the pro-​Kurdish parties which maintained a presence in the 2000s. On the other hand, in the northern part of the city, a noticeable pro-​Kurdish preference can be seen. In all, Erzurum, with its strong nationalist and conservative tenet, portrays a strong continuity in terms of “rightism”. However, that the left’s portion decreased as the pro-​Kurdish parties gained power created a particular division in the southern part of the city. This part no longer prefers to vote for the center-​left parties. Similar to Erzurum, the Gaziantep electorate is also more likely to support right-​wing political parties, but prefers the center-​right of the political spectrum to the religious or nationalist parties. The share of the vote gained by far right parties such as Refah Partisi (the Welfare Party, the WP) or Milliyetçi Hareket Partisi (the Nationalist Action Party, the NAP) have usually followed the national averages. Also, the percentages of left-​wing voters are considerably higher than in Erzurum. Between 1950–​1977 in particular, the average vote for the center-​left was 46.2 percent in Gaziantep. Although the amount of the vote received by center-​left parties significantly decreased after the 1980 military coup, which predominantly targeted leftist movements, it did not hit rock bottom as in Erzurum. The drop in the center-​left was also partially a result of the rise of pro-​Kurdish parties since the mid-​1990s. In a nutshell, Gaziantep voter behavior has not deviated from the general mood of the national political scene. Furthermore, the voter preferences of the Gaziantep electorate have been similar to those in İstanbul and Ankara. Eskişehir holds a center-​oriented voter type like Gaziantep, but as seen in the graph, the proportions of right-​wing and left-​wing votes has remained quite consistent compared to other cities. Another important attribute of the Eskişehir case is the underrepresentation of nationalist, religious, or pro-​ Kurdish parties. With such preferences, the electoral inclinations of Eskişehir are similar to those cities in relatively developed regions in the west. On the other side of Turkey, Van represents a quite different voter type, which is widespread in the least-​developed regions with Kurdish majority populations. The graph fairly displays the quite distinctive voter behavior in Van. First of all, until the 1990s, voter volatility was incredibly high in Van, where voters not only moved between parties in the same ideological family, but also from one ideological side to the extreme opposite. It seems that personal ties with the candidates in Van mattered more to voters than their ties with the political parties. It is obvious, when considering the fluctuations in the shares of the vote gained by the political parties, when an independent candidate entered

8 Bekaroğlu and Kaya Osmanbaşoğlu the race (see the ups and downs in the left-​wing votes between 1965 and 1977). Secondly, religious parties always performed much better than their national average in Van. Cumhuriyetçi Köylü Millet Partisi (the Republican Peasant Nation Party, the RPNP) in 1961 (21.4 percent), Milli Selamet Partisi (National Salvation Party, the NSP) in 1977 (20.5 percent), and the Welfare Party in 1987 (17.4  percent) are some significant examples of this trend. Finally, left-​wing voters have nearly vanished from the scene since the rise of pro-​Kurdish political parties in the mid-​1990s in Van. In addition to these four discernible voting patterns in certain district clusters, there have been some outliers as well. One of the most famous outliers, Tunceli, with a relatively marginal left-​wing electorate, was actually identified by Çarkoğlu and Avcı (2002, p. 129) as a cluster in itself. Such peculiar districts, especially during certain particular periods in time, have demonstrated distinguishable preferences. Tunceli, with its longstanding Alevi and Kurdish population, preferred to vote for leftist parties in the 2000s. Nevertheless, from the very beginning of the 1950 elections, center-​right parties gained a considerable amount of the vote in Tunceli. In other words, this current prolonged attitude was almost absent 70 years ago. On the other hand, at the very beginning of the multiparty era, Kırşehir, a small city in the middle of Anatolia, held a geographical representation pattern that showed support for far right individualistic political parties. Kırşehir maintained this pattern for decades, which resulted in a downgrading of Kırşehir’s status from a city to a town as a punishment for not supporting the Democrat Party in the 1954 general elections. Another interesting example is Fatsa, a small town in Ordu in the Black Sea region. What makes Fatsa an interesting outlier is not its center-​left orientation in the 1970s, but its election of an independent mayor who had ties with the radical left-​wing movement, Devrimci Yol (Revolutionary Path) in the 1979 local elections. Throughout this process, Mayor Fikri Sönmez, known as Terzi (Tailor) Fikri, ruled Fatsa through neighborhood commissions, a kind of bottom-​up governing strategy. His rule, however, was criminalized by the opposition, citing this experience as a call for socialism (Bora, 2017, p. 627). Interestingly, this leftist inclination did not continue in Fatsa after the 1980 military intervention. Fatsa began to vote for the center-​right parties in the 1980s and 1990s, with some minor exceptions. Apparently, the charismatic leadership of Terzi Fikri played a stronger role than any actual political ideologies in Fatsa.1 So far, we have provided an analysis of four major voting patterns in four representative cities to understand the general trends in Turkey’s electoral geography, and have also given three outlier examples that do not fit these overarching patterns. Both general trends and outliers suggest that the geographical tendencies may change in time, or some tendencies may become more apparent while some others may become more indistinct over the years. Nevertheless, in the Turkish case it is clear that the geographical tendencies have become increasingly discernible since the 1980s. In the remainder of this chapter, we will provide a historical analysis of Turkey’s electoral geography

Introduction: Turkey’s electoral geography  9 by dividing the multiparty election era into two distinct periods. The first period includes the elections between 1950 and 1977, where center-​right and center-​left parties were the predominant players in the democratic game, whose performances did not show significant regional particularities. The second period, on the other hand, starts with the 1983 elections and considers up to the present day. As explicitly displayed in Figure 1.2, with the rise of Kurdish and Islamist political parties, Turkish electoral geography started to show indications of regional dynamics in the latter period. Surely, the 1980 military coup and the 1982 Constitution thereafter also played a significant role in the transformation of Turkish politics through a broad change in its electoral system as well as policy-​making approaches.

1950–​1977: The era of political centrism From the beginning of multiparty politics in 1950 to the 1980 military coup, center-​right and center-​left political parties were dominant players, with no distinct variation in their regional performances. This domination, however, started to weaken in the 1970s when the RPP established coalition governments with the National Salvation Party, an Islamist political party. Indeed, far right political parties in particular became major players in the inner and eastern Anatolian regions from the 1970s onwards. Beyond any shadow of doubt, the prevailing electoral system(s) played a significant role in shaping the regional characteristics of election results.2 As a matter of fact, the elections in the 1950s did not exhibit a significant regional structure because of the fact that the simple majority block voting system very likely blurred regional lines. Nevertheless, some slight geographical patterns also began to arise from the very beginning of multiparty politics. For example, the Republican People’s Party, considered to represent the “center” against the “periphery”, was more successful than Demokrat Parti (the Democrat Party, the DP) in some of the less-​developed districts (Wuthrich, 2013, p.  759; Kudat, 1975). According to Wuthrich (2015, p.  129), both parties were able to successfully mobilize “different types of peripheral communities”. The RPP gained more votes in districts in the far eastern and certain Mediterranean regions where its pre-​ existing patronage networks from the single-​party period were still intact. Generally speaking, there was a significant negative correlation between the regional level of development and the RPP’s share of the vote in the initial multiparty elections. The DP, on the other hand, employed brokerage strategies where the RPP’s patronage networks were less influential. Here, Kudat (1975) argues that the villages who were eager to integrate with the rest of the world (namely getting daily newspapers, enjoying regular transportation to the city centers, and such) had the potential to vote for the DP, whilst the more isolated ones were more likely to vote for the RPP. In less developed regions in particular, local and ordinary cleavages (such as rivalries or competitions between influential families) played a greater role than ideological and identity-​based cleavages

10 Bekaroğlu and Kaya Osmanbaşoğlu in determining electoral behavior (Wuthrich, 2015, p.  132). This dynamic, however, varied with socio-​economic development. According to Özbudun (1976, p.  49), the less developed east enabled local notables to mobilize their supporters while relatively modernized regions created an opportunity for different campaign strategies based on the influence of merchants and professionals. In the 1950 elections, eight out of eleven cities that the DP lost were located in eastern or southeastern Anatolia.3 Similarly, in the 1954 elections, the RPP gained in only four cities, three of which were located in the Eastern Anatolia region. Even in the 1957 elections, when the DP continued to win the elections in the Aegean, Marmara, and mostly in the Black Sea and Mediterranean regions, the difference between the DP and RPP votes was not dramatic.4 Nonetheless, thanks to the simple majority block voting system, the number of DP seats in parliament was more than triple those of the RPP. On May 27th, 1960, Turkey’s inexperienced democracy was suspended by a military coup for 17 months. The military regime initiated the writing of a new constitution, which was approved by 61.7 percent of the people’s vote in a referendum on July 9th, 1961.5 Starting multiparty politics in 1946 under the course of the 1924 Constitution from the single-​party period, Turkey continued its democratic journey in the October 1961 general election with a new constitution as well as with the rising domination of civilian and military bureaucrats that dramatically changed the Turkish political system. With regard to representation and elections, the most important novelty was that of changing the electoral system from the previous simple majority block voting to proportional representation. Furthermore, the 1961 Constitution also introduced a bicameral legislative organ. The architects of the 1961 Constitution intended to enhance political representation in parliament, but at the same time wished to balance and control the people’s will in the legislation with an elite-​dominated republican senate. In the 1960s, under an electoral system that favored smaller political parties, the peripheral electorate seemed to head towards the newly established far right and far left political parties rather than the center-​right Adalet Partisi (the Justice Party, the JP), the DP’s successor, and the center-​left and the RPP (Özbudun & Tachau, 1975). Nevertheless, Islamist and nationalist parties had less geographical consistency compared to the center-​right and center-​left in the 1960s (Çarkoğlu & Avcı, 2002, p. 119). In the 1961 elections, it is possible to observe that pre-​coup political dynamics continued even though none of the DP’s successors could form the government, yet the RPP won the plurality of the votes with 36.7 percent . However, when combining the share of the votes of the DP’s larger successor, the JP, and smaller successor, Yeni Türkiye Partisi (the New Turkey Party, the NTP) (34.8  percent and 13.7  percent, respectively), and examining the districts where they attracted the electorate, it seems that the local patterns between political parties and voters that shaped the electoral geography in the 1950s continued into the 1961 general elections. Indeed, the districts where the JP or the NTP won in 1961 overlapped to a large extent with those of the DP-​dominated regions of the 1950s. Also, the

Introduction: Turkey’s electoral geography  11 DP’s successors, especially the JP, were still slightly more successful than the RPP in relatively developed provinces, particularly in the western coastal provinces. A new trend in terms of electoral geography, however, started to emerge with the 1965 elections with its introduction of a national remainder system. While economic development proposals and patronage links dominated the electoral politics without any significant geographical pattern, 1965 can be considered a “transitional election” where ideology with regard to the left and right axis started shaping political campaigns and electoral behavior (Wuthrich, 2015, p. 123). Before the 1965 elections, it was only the political parties led by idiosyncratic leader Osman Bölükbaşı that could be categorized as extremist since the political realm was, to a large extent, dominated by the center parties. In the 1961 Constitution’s relatively liberal framework for the mobilization of different ideas, far right, Islamist, or radical left political parties started entering parliament, and they became typical players in Turkish democracy starting with the 1965 elections. In parallel with the rise of ideological cleavages, the 1965 elections can also be considered a turning point for the Turkish political party system in terms of fragmentation, which became the major characteristic of Turkish politics between 1965 and 2002. The number of parties in the Turkish parliament rose to six in the 1965 elections, a number which never dropped in any subsequent election. The only exception is the parliament formed by the 1983 elections, which were closely monitored by the military regime. The ideological polarization and party fragmentation starting from the 1965 elections significantly influenced Turkey’s electoral geography. Two trends that appeared at this time are the most noteworthy. First of all, the RPP’s support base changed as its share of the vote increased in the more developed regions and dropped in the less developed ones (Özbudun, 2013, p. 44). This trend continued even through to the mid-​and late-​1970s when the RPP, under the new and energetic leadership of Bülent Ecevit, increased its votes and became the largest political party in parliament. Since then, there has been a clear positive correlation between the RPP’s vote and the indicators of socio-​economic development, such as urbanization, level of education, industrialization, etc. With the apparent incentive of defining the RPP’s ideology around the left of center, Ecevit’s RPP was capable of effectively communicating with the migrant workers living in the more industrialized cities. Secondly, the far right parties, both Islamist or Turkish nationalist, stabilized their electoral bases in the central Anatolia and inner Mediterranean regions at the expense of center-​right parties after the 1965 elections (Çarkoğlu & Avcı, 2002, p. 124). This trend not only caused fragmentation among right-​ wing political parties but also increased the overall regionalization of the Turkish political party system. The fragmentation within the right-​wing political parties and the rise of social justice concerns among urban voters ended up in the 1970s with the electoral victories of the RPP, which simultaneously transformed its political

12 Bekaroğlu and Kaya Osmanbaşoğlu discourse from one of a statist agenda to that of social democrat tones to attract the urban and rural poor. Starting from the 1969 elections, the regions where the RPP was weakest and strongest began to dramatically change. The RPP lost a substantial number of votes in the districts in the southeast region, where its votes used to outnumber the DP and JP in the 1950s thanks to the patronage relations generated during the one-​party era. On the other hand, industrial and urban centers, especially in the Marmara region, became the RPP’s strongest electoral bases in the 1970s (Wuthrich, 2015, 141). These changes in electoral geography can partially be explained by the significant transformations in the demography of the Turkish electorate in the 1960s and 1970s. Most importantly, the urban population had grown from 18.5  percent of the total population in 1950 to 38.2  percent in 1970 and 43.8  percent in 1980. While the total population has risen from 20.9 million in 1950 to 44.7 million in 1980, the urban population has grown by almost six-​fold from 3.8 million to 19.3 million. There are two characteristics of this rapid urbanization, which is a common phenomenon amongst most developing countries. First, urbanization was regionally uneven, concentrating in a few cities like İstanbul, Ankara, and İzmir. Secondly, the housing of newcomers in urban centers was highly unregulated and, indeed, unauthorized. These new inhabitants, predominantly poor people working in irregular jobs and living in shantytowns, constituted only 4.7 percent of the urban population in 1955. This number rose to 16.4 percent in 1960 and 23.6 percent in 1970. More dramatically, 65 percent of the population in Ankara and 45 percent in İstanbul were living in shanty towns in the late 1970s.6 According to Wuthrich (2015, p.  162), the JP and the RPP, as mainstream political parties, tried to appeal to these new urban-​poor communities through pragmatic approaches (e.g., promises of reward) directly addressing their survival needs rather than making ideological appeals. While the urban-​ poor initially voted for the JP, they mostly turned towards the RPP in the 1973 and 1977 elections, in which the RPP’s shares of the vote in İstanbul, Ankara, and İzmir were significantly higher than the national average. For example, in 1973, while the RPP’s national average was 33.3 percent, its share of the vote in İstanbul was 48.9 percent. Similarly, in 1977, the RPP gained 58.2 percent of the vote in İstanbul whereas its national share of the vote was 41.4 percent. Yet, despite the ultra-​nationalist tendencies beginning to become more salient in central Anatolia, the nationalists also found a place within the center-​right umbrella. In the 1965 elections, the center-​right JP lost only nine cities, eight of which were located in eastern or southeastern Anatolia. In the 1969 elections, the RPP gained five provinces, all of which constituted a block in the map from north to south. Similarly, independents won in eastern or southeastern cities while the JP continued its success in the rest of the country. Kırşehir remained an outlier district in the middle of Anatolia since Osman Bölükbaşı, an oratory opposition leader, was from that city. In 1977, with an apparent transformation of the RPP’s discourse from a statist agenda to social democrat tones, the

Introduction: Turkey’s electoral geography  13 electoral map was changed once more to the extent that the RPP’s representation increased in all regions. However, as a precursor to the rise of identity politics in the next period, the independent candidates, who were close to the socio-​cultural concerns of Kurdish citizens, continued to gain votes in the eastern and southeastern provinces.

1983–​2018: The era of centrifugal politics and regionalization Turkey’s political party system was once again dramatically interrupted by the 1980 military coup, which led to the ban of all pre-​1980 political parties and leaders. This time, the tenure of the military rule was long and the intensity of the military domination was much stronger in comparison to the 1960 intervention. With dubious voting in the 1982 Constitution and the election of General Kenan Evren as the new president, Turkish politics fell into military-​ related trouble in return for stopping the civilian political turmoil that was the basis of left-​right distinction. Nevertheless, it was indeed a complex and chaotic stance, where multiple players involving sectarian, ethnic, and religious dimensions fragmented considerably in the pre-​1980 context. The new military rule transformed the perceived meaning of politics while cheering up the apolitical through the premise of combatting terrorism. The new system introduced a 10 percent national threshold for the legislative elections to moderate the extreme multipartyism that dominated Turkey from 1961 to 1980. New strict rules were introduced to monitor the activities of the political parties, along with particular restrictions about individual liberties, including a ban on speaking any language other than Turkish—​ which for certain will have had a particular influence on the Kurdish people. The military government reinitialized the multiparty elections with the “semi-​competitive elections” of 1983, which brought Anavatan Partisi (the Motherland Party, the MP) governments from 1983 to 1991 in two subsequent elections (Özbudun, 2013, p. 1). In the 1983 general election process, only three political parties who were approved by the military rule could join the Grand National Assembly, while the potential heirs of the previous political parties were all vetoed. The MP represented a center-​right position even though it was argued that the party juxtaposed four main tendencies consisting of nationalism, conservatism, liberalism, and the left. In Cizre’s terms, “political entrepreneurship” was the main political tool of the MP, other than previous ideological discourses, with its “moving from an abstract vocabulary to an issue-​based concrete, everyday language” (2002, p.  84). Tired of being exposed to political polarization in the pre-​1980 period with the governments’ poor economic management, Turkish voters in the post-​ 1980 period came close to Downsian7 voter choices, trusting Turgut Özal’s technocrat image that specialized in economics. By the 1983 elections, the MP portrayed a strong nationalization trend at the provincial level. Nevertheless, Milliyetçi Demokrasi Partisi (the Nationalist Democracy Party, the NDP), which was evidently supported by Kenan Evren and utilized a September 12

14 Bekaroğlu and Kaya Osmanbaşoğlu discourse throughout the election campaign, gained some cities only from the southeast Anatolian region, which was probably an outcome of enormous pressure over these cities due to the suppression of the Kurdish identity by introducing a new ban on speaking Kurdish. Furthermore, the allegiance of military rule with some of the local powers, the so-​called aşiret, might well have led to this result.8 On the other hand, the Populist Party, which can be perceived as a leftist alternative of the MP, gained the provinces on the borders of Turkey, but without any substantial penetration of the central Anatolian regions. In this regard, cities such as İzmir, Hatay, Kırklareli, and Edirne voted for the Populist Party with the common feature of being located either on the coastal regions or borders of the country. Even though these kinds of basic geographical patterns existed in these elections, when we analyze the numbers in detail, it is hard to identify a well-​established and deeply anchored regional alignment for the Populist Party due to the similarities between the shares of the vote gained by the MP and the Populist Party in the provinces that they won. On the other hand, the cities in which the military government promoted the MDP evidently won southeastern cities, which can be interpreted as being a result of the strict law relating to emergency actions, as well as the central military rule’s cooperation with local lords, to a greater extent than other factors. In the 1987 general elections, the MP won most of the cities around the country thanks to the fact that “the law was modified to raise the barrier for representation at the district level”, which converted the MP’s 36  percent vote into 65 percent in the Grand National Assembly (Waterbury, 1992, p. 130). However, certain geographical characteristics may still be observed. For instance, the TPP leader, Demirel’s Isparta, voted, expectedly, for the TPP whilst Kurdish dominated Diyarbakır, and Tunceli voted for the SDPP. Edirne and Kırklareli, two borderline cities in the Marmara region voted for the SDPP, who voted for the leftist Halkçı Parti (the Populist Party, the PP) in the 1983 elections similar to Diyarbakır and Tunceli. Apart from the geographical allocation of the votes, Özcan, studying İstanbulites’ voting preferences in the 1987 elections and 1991 elections, said that “It is interesting to note that region of birth still exerts an influence on party preference, reflecting differential political socialization among the regions” (2000, p. 512). Nevertheless, when the banned political leaders returned to politics in 1987, and the first Kurdish political party was founded in 1990,9 the MP’s electoral support gradually eroded and high levels of fragmentation, polarization, and volatility once again became the key features of Turkish politics. Since then, increasing levels of regionalization in the political party system and voter preferences as a new dimension of Turkish politics have also become salient along with the above fragmentation, polarization, and volatility (Özbudun, 2000; Tezcür, 2012). These features have displayed significant geographical patterns, especially since the mid-​1990s. To begin with, the Kurdish electorate increased the volatility in the southeastern regions by moving en masse from various

Introduction: Turkey’s electoral geography  15 political parties to the Kurdish nationalist Halkların Demokrasi Partisi (the People’s Democracy Party, the PDemP) in the 1995 elections. Since then, Kurdish voters in the southeast have consistently continued to support the PaDP and its successor parties. Even in the 2002 elections, when the Turkish party system experienced a significant realignment, volatility in the southeast was more than 10 percent lower than the Turkish average (Şekercioğlu & Arikan, 2008, p. 225). Besides the Kurdish nationalist parties, the rise of Islamist and Turkish nationalist parties in the 1990s showed regional patterns as well. While the political competition was shaped by the center-​right and center-​left in the 1980s and the early 1990s with a geographically balanced electoral distribution, Refah Partisi (the Welfare Party, the WP) and Milliyetçi Hareket Partisi (the Nationalist Action Party, the NAP) started dominating central Anatolian districts in the mid-​1990s. Furthermore, the WP increasingly drew electoral support in developed districts, most significantly in İstanbul and Ankara, which can be explained by “new geographies created by rural-​urban migration and shifting class orientations” (Secor, 2001, p. 546). The 1990s also saw the increased significance of local elections, after which gaining the metropole municipalities of İstanbul and Ankara began to be seen as the impetus to reach the electorate to gather votes in the general elections as well. In 1994, it was with the Islamist WP’s gain of the İstanbul and Ankara municipalities triggering voters’ support for that party in the forthcoming general elections that a semi-​geographical understanding began to take shape in Turkish politics, namely that the one who can successfully communicate with the masses in İstanbul and Ankara can, in turn, reach much more of the electorate and win Turkey in the end. These cities were uninterruptedly won by the JDP between 2004 and 2019, which was seen as significant grounds for the JDP’s success in the general elections as well. For this reason, the JDP’s arrow victory in the 2014 elections in Ankara was read as a considerable weakness of the governing party. The recent loss of these two metropoles was read by many students of Turkish politics as a major deficit in the governing party that in the future would probably make it hard to gain the majority of the votes in the national elections again.10 In 2002, Turkish politics experienced a major realignment, where the mainstream right-​and left-​wing parties (except for the RPP) were unable to achieve the 10  percent national threshold to gain seats in parliament. Since then, the JDP, whose founders were prominent politicians in the WP, became the representative of center-​right and dominated politics by winning all the national elections and forming the resultant governments. Other than the JDP’s domination in the elections, a significant characteristic of this new era can be observed in the geographical patterns of electoral behavior, which was shaped more or less by identity preferences. The coastal districts with secular preferences in western Turkey predominantly voted for the RPP as the main opposition party to the JDP government, while the electorate in the southeastern regions preferred to vote for the Kurdish political parties. On the other hand, the districts in central, eastern, and northern Anatolia

16 Bekaroğlu and Kaya Osmanbaşoğlu tended to vote for the conservative, nationalist, and Islamist parties, particularly for the JDP. Strikingly, the JDP has been either the winner or the first runner-​up party in the majority of districts. This spatial dependence of electoral behavior and political parties since 2002 has increasingly attracted scholarly attention about Turkey’s electoral geography, and indeed still requires further explanation. This volume, with its particular focus on the 2000s, adds new possible explanations to the electoral geography of Turkey. Along with the above-​mentioned metropolitan voting shift by the 2019 local elections, with its scope, region, and identity, the place has continued to hold ample significance while interpreting the political issues. In a nutshell, the end of the eloquent position of geography in Turkish politics is such a remote possibility in the near future that a considerable amount of prevailing and forthcoming students of political science will gradually continue to spend their time deliberating over the geographical repercussions of Turkish politics.

Organization of the book This edited volume aims to evaluate the prevailing geographical implications of the elections in Turkey following the establishment of multiparty politics in 1950, with a particular emphasis on the last two decades. To this end, a variety of issues related to the geographical connotations of Turkish electoral politics are addressed both qualitatively and quantitatively by bringing scholars together from various fields of the social sciences. Each chapter analyzes how Turkey’s electoral geography has been shaped to correspond to a certain aspect of multiparty politics such as voting behavior, political parties and party system, nationalization and regionalization, redistricting, gender issues, identity dynamics, or ideological polarization. This study, therefore, fills a considerable gap in the literature on Turkish politics by contributing a comprehensive analysis of Turkey’s electoral geography that will potentially soon gain an increased significance due to its fragile geographical dynamics. The most recent local elections in 2019 demonstrated that a collaboration of the RPP with the Kurdish political party PDP and the center-​right Good Party (İyi Parti, the GP) has the potential to change Turkey’s electoral geography. If this collaboration continues in the national elections, the RPP may likely break away from being a political party whose support base is mostly localized in the western coastal regions. Therefore, analyzing Turkey’s electoral geography not only sheds light on Turkish politics in the past but can also help us to forecast the potential directions taken by Turkish politics. Although there are books, edited volumes, journal articles, and special journal issues dealing with Turkish political parties, the electoral system, particular elections, or identity politics, a comprehensive study that emphasizes the geographical dynamics of Turkish political parties and elections is lacking. As this work covers not only elections, but also topics such as political parties, identity issues, political values, gerrymandering, gender, or voting from abroad, it will provide useful guidance to anyone interested in

Introduction: Turkey’s electoral geography  17 Turkish and Middle Eastern politics. This edited volume will also contribute to the theoretical debates in electoral geography literature in general. By utilizing and applying the concepts and hypotheses from electoral geography literature to the Turkish case, this study aims to improve the hypotheses in and develop new concepts for the existing literature about electoral dynamics in developing democracies. To this end, the present study starts with a chapter on the nationalization of the party system in Turkey. In this chapter, Demirkol and Bekaroğlu fill an important gap in the literature by measuring party nationalization at the national and party levels for all the parliamentary elections in Turkey between 1950 and 2018 that would help to explain certain particular characteristics of Turkish politics. The chapter first discusses the concept of nationalization, then reviews the literature on the regional dynamics of electoral behavior and political parties in Turkey. After identifying the fact that the literature mostly deals with regionalization rather than nationalization, Demirkol and Bekaroğlu measure the nationalization of the Turkish party system and political parties. Their analysis demonstrates that while Turkey has an intermediate level of party system nationalization, the average nationalization scores have continually dropped from 0.901 in the 1950s to 0.839 in the 2010s. According to Demirkol and Bekaroğlu, two factors seem to be responsible for the decreasing nationalization scores: the number of competent parties, and the rise of identity politics. The chapter argues that nationalization scores drop as the number of parties and identity politics increase. In the next chapter, Kaya Osmanbaşoğlu traces the attempts at gerrymandering in Turkish electoral politics from the very beginning of the multiparty era, which have gone hand in hand with clientelism. During the reign of the center-​right governments in particular, the ruling parties have been eager to toy with district borders and the electoral system, as well as the status of provinces or towns through upgrading or downgrading their categories to use them as “bargaining chips” in the elections. Here, one should note that the lack of democratic institutionalization has a direct relation with the gerrymandering application, and that such variety and prevalence of gerrymandering ultimately hinders the further consolidation of democracy. The third chapter deals with the electoral dynamics in the western coastal districts in Turkey that have exhibited a predominantly RPP inclination since the mid-​1990s. Tosun, Aydoğan Ünal and Erdoğan Tosun provide an analysis to understand the continuities and ruptures in the voting behavior in the western coastal districts that have the particular spatial characteristics of higher-​level socio-​economic status and secular values compared to other regions in Turkey. The provinces covered in this chapter range from Thracian to the Mediterranean coastal regions. Tosun et al. examine the political and sociological elements underlying the spatial differentiation in these provinces by looking at the election outcomes from 1983 to 2018. According to their analysis, while the voters in these provinces showed a preference for center-​right and center-​left parties in a relatively balanced manner until the mid-​1990s,

18 Bekaroğlu and Kaya Osmanbaşoğlu they have been inclined to vote for the center-​left, especially since the 1999 elections, as a response to the rise of Islamist and Kurdish political parties nationwide. They also question the volatility among the voters in the western coastal regions by asking where the electorate of declining mainstream right-​ wing parties has headed. The chapter argues that the voters previously preferred the mainstream right-​wing parties mostly directed towards the JDP and, to a lesser extent, to the NAP and RPP since 2002. The next chapter focuses on the southeastern provinces in Turkey that display completely different spatial characteristics to the western coastal regions. Alptekin, by analyzing the electoral campaigns of the Kurdish political parties since 2002, not only deals with the electoral dynamics in the southeastern provinces but also offers a temporal comparison of the campaign discourses of Kurdish political parties in the western and southeastern districts in the national elections. According to Alptekin, two significant observations can be made. First, the Kurdish political parties have shown longitudinal variation from a strictly class-​based socialist rhetoric intermingled with Kurdish nationalism to a new-​leftist tone with the greater inclusion of gender and environment-​related issues in their campaigns. Secondly, the Kurdish political parties, especially Halkların Demokratik Partisi (the People’s Democratic Party, the PDP), used a moderate, universalist, and “pro-​Turkey” (Türkiyeci) campaign discourse in western districts while their language turned into a more identity-​ based and radical tone in southeastern districts. Alptekin analyzes these inter-​ regional and longitudinal differences in the political campaigns of Kurdish parties by looking at the primary sources. Bringing an eminent interpretation of space, Akyüz and Çınar look at feminist political geography by displaying a wide array of women’s political engagement from national to municipal levels. With an ethnographic view of gender and gendered topography, authors circumscribe women’s cooperation in politics perceiving them both as patron and client. Here, based on their research conducted in 2017 with 23 out of 63 active women’s cooperatives, clientelism, which leaves room for exclusion and inclusion, is found to be related to feminist political geography. Utilizing “non-​random, purposive, and maximum variation sampling method so as to make sure the range of people and sites from which the sample is selected is representative of the larger population geographically, socioeconomically, and politically”, the study determines that at the local level, women’s cooperatives are generally under the patronage of male-​dominated political actors. Furthermore, given that the place of politics is of considerable importance, the study finds that even women politicians are still under patriarchal control and are maneuvering within the “proper” borders defined by masculinity. Tracing the geographical division lines of contemporary Turkey consisting of a secular coastal region in the western region of Turkey, the southeastern region, with its pro-​Kurdish tendency and with it being a predominantly conservative and nationalist region, Ayşenur Kılıç dwells on media use, trust, and democratic participation behaviors as well as political views and neighboring

Introduction: Turkey’s electoral geography  19 effects on two well-​known election monitoring organizations. Based on a survey and in-​depth interviews, she garnered the network characteristics, news consumption behaviors, media preferences, and political affiliations of the volunteers who took part in the activities of the two election monitoring organizations. Consequently, she clearly illustrates that far from being nationalized all around the country, the characteristics of volunteers to a large extent resemble the secular tendencies of voters in coastal regions in Turkey. As a complementary aspect of Akyüz and Çınar’s research, Mezher Yüksel brings women’s political representation to the fore with regard to the historical process of women’s representation in the Turkish parliament. Combining political geography patterns with other socio-​cultural aspects, he determines their social origins, place of birth, electoral area, political party, previous professions, level of education, and the foreign languages they know to portray their profiles. He notes considerable variation in the regional distribution of women deputies in Turkey, where the Marmara region gathers almost one-​third of all women deputies whilst the Black Sea and Mediterranean regions usually proportionately send far fewer women deputies. Furthermore, Yüksel finds that the status of electoral district is of particular importance in terms of metropoles eager to elect women candidates vis-​à-​vis the countryside, which has an opposing trend to the early Republican era. Forty percent of women deputies represent their place of birth, which is almost half the men’s representation of their place of birth. With these exciting blueprints, Mezher Yüksel also indicates the ideological impacts of women’s representation regarding their geographical constellations that pro-​Kurdish parties leave more room for women’s representation, which also compounds geographic concentration. Further, they enact symbolic meanings of women’s representation that fluctuates from modernization and secularization to ethnic and religious identity production. The final chapter broadens the scope of this book by drawing attention to the practices associated with voting from beyond Turkey’s internal electoral districts. Anaz and Köse tackle this understudied but increasingly important subject. Indeed, since the 2014 presidential elections in Turkey, voters abroad as ex-​pats have been allowed to cast votes in ballot boxes in their countries of residence rather than the customs stations. This practice continued in the following national elections and indeed has had some impact on the electoral outcomes in Turkey as it has significantly increased the foreign voter turnout to approximately 50 percent, which equates to about one and a half million voters. By analyzing the effects of voting from abroad on Turkey and comparing with other countries, Anaz and Köse suggest that Turkish lawmakers should consider taking a step further and establish diaspora electoral districts to elect their own representatives. Overall, the chapters in this book share the idea that Turkey’s electoral geography has become increasingly divided along the lines of identity issues rather than by socio-​economic ones. The rise of ideological and identity-​ based political parties such as the Welfare Party, the Nationalist Action Party,

20 Bekaroğlu and Kaya Osmanbaşoğlu and the Peoples’ Democratic Party has had various backlashes among the mainstream parties in Turkey, especially since the mid-​1990s. Nevertheless, with the continuum of the secular-​conservative divide along with the rise of Turkish nationalism vis-​à-​vis Kurdish nationalism, geographical dynamics have become one of the most explanatory sources by which to interpret the electoral alignments and behaviors in Turkey. Not only developed districts in the west and underdeveloped districts in the east but also the districts in the coastal regions, central Anatolia, and the Black Sea show specific characteristics of voting. Furthermore, some particular cities, such as Tunceli and Kırşehir, demonstrate distinct related and prominent voting patterns which have endured for decades. Given the growing significance of geographical aspects of elections in Turkey, this book represents a prominent source for students of Turkish politics who are interested in the dynamics of voting patterns in Turkey and various aspects of elections ranging from the nationalization of the political party system to gender representation in politics with a precise emphasis on electoral geography.

Notes 1 On June 18, 2020, an interview was conducted with Ahmet Becioğlu, who wrote a book on this issue. According to his observations, people were attracted by Fikri Sönmez’s charisma and leadership first, which then brought popular support for the far left. After the pressure of the junta regime, people were afraid of being accused of being terrorists and immediately, and drastically, changed their political stance. Nevertheless, he notes that this shift in political alignment did not take place from far left to far right at the first stance, but just “slipped” one layer to the right. For further details, see Becioğlu, A. (2019). 2 Electoral system shapes the party system and regional patterns in many cases. For instance, see: Dalton. R.J. (1996); Deschouwer, K. (2003); Birch, S. (2001); Jeffery, C. (2015). 3 Information about the election results are taken from Turkish Grand National Assembly official website. For further details, please visit TBMM (2019). www. tbmm.gov.tr/​develop/​owa/​genel_​secimler.secim_​cevreleri?p_​secim_​yili=1950 4 The DP obtained 47.8 percent of the votes while the CHP gained 41 percent. With these election scores, the DP gathered 424 seats in the parliament vis-​à-​vis 128 seats of the RPP. 5 For a detailed analysis of the referendums in Turkey, see Kaya Osmanbaşoğlu, G. & Bekaroğlu, E.A. (2019). 6 For a detailed analysis of urbanization in Turkey and relevant statistics, see Erman, T (2012). 7 In An Economic Theory of Democracy, which was followed and developed by many scholars later, Downs asserts that voter preferences largely focus on daily problems and economic conditions other than fixed ideologies. For details, see Downs, A. (1957), p. 98. 8 Devrim Ertürk describes two types of ashirets (local powers):  the ones who challenged the central authority and those who cooperated with the central authority to gain mutual benefits. For details, see Ertürk, D. (2015).

Introduction: Turkey’s electoral geography  21 9 Halkların Emek Partisi (the People’s Labour Party, the PLP) was founded in 1990 as the first political party to bid for the representation of the Kurds. Demokrasi Partisi (the Democracy Party, DemP), on the other hand, was founded in 1991 as a caution for a possible closure of PLP. After DemP’s closure in 1994, Halkların Demokrasi Partisi (the People’s Democracy Party, PDemP) was founded and entered the 1995 elections. Since then, PDemP and successor parties have become a regular political actor in Turkey. 10 Some of the works that dealt with 2019 local elections in this line are as below: Wuthrich, F.M., & Ingleby, M. (2020); Yavuz, M.H., & Ozcan, N.A. (2019); Esen, B., & Gumuscu, S. (2019); Genç, K. (2019).

References Agnew, J.A. (2014). Place and politics: The geographical mediation of state and society. London: Routledge. (First published in 1987.) Ayan Musil, P. (2015). Emergence of a dominant party system after multipartyism: Theoretical implications from the case of the AKP in Turkey. South European Society and Politics, 20(1),  71–​92. Becioğlu, A. (2019). Fikri Sönmez-​bilinmeyen yönleriyle. İstanbul: Su Yayınları. Birch, S. (2001). Electoral systems and party systems in Europe east and west. Perspectives on European Politics and Society, 2(3), 355–​77. Bora, T. (2017). Cereyanlar: Türkiye’de siyasi ideolojiler. İletişim Yayınları. Çarkoğlu, A.  (2011) Turkey’s 2011 general elections:  Towards a dominant party system? Insight Turkey, 13(3),  43–​62. Çarkoğlu, A., & Avcı, G. (2002). An analysis of the electorate from a geographical perspective. In Esmer, Y., and Sayarı, S. (Eds.) Politics, parties, and elections in Turkey (pp. 115–​36). London: Lynne Rienner Publishers. Cizre, Ü. (2002). From Ruler to Pariah:  the life and times of the True Path Party. Turkish Studies, 3(1), 82–​101. Dalton. R.J. (1996). Political cleavages, issues, and electoral change. In LeDuc, L., Niemi, R.G,. and Norris, P. (Eds.), Comparing democracies: Elections and voting in global perspective, 2, (pp. 319–​42). Thousand Oaks, CA: Sage Publications. Deschouwer, K. (2003). Political parties in multi-​layered systems. European Urban and Regional Studies, 10(3), 213–​26. Downs, A. (1957). An economic theory of democracy. New York: Harper and Row. Ertürk, D. (2015). Güneydoğuda CHP algısı (Kurumsal Bir Perspektiften CHP’nin Güneydoğu Anadolu Bölgesin’de Toplumsal Tabanını Kaybetmesinin Nedenleri 1990–​ 2011) (Vol. 3). Gündoğan Yayınları. Erman, T. (2012). Urbanization and Urbanism. In Metin Heper and Sabri Sayarı (Eds.), The Routledge handbook of modern Turkey (pp. 293–​ 302). London: Routledge. Esen, B., & Gumuscu, S. (2019). Killing competitive authoritarianism softly: The 2019 local elections in Turkey. South European Society and Politics, 24(3), 317–​42. Flint, C. (2001). A Time Space for electoral geography: Economic restructuring, political agency and the rise of the Nazi party. Political Geography, 20(3), 301–​29. Forest, B. (2012). Electoral redistricting and minority political representation in Canada and the United States. The Canadian Geographer/​Le Géographe canadien, 56(3), 318–​38.

22 Bekaroğlu and Kaya Osmanbaşoğlu —​—​—​. (2018). Electoral geography:  From mapping votes to representing power. Geography Compass, 12(1),  1–​17. Genç, K. (2019). Turkey’s rule of one: President Erdoğan’s government is challenging the result of İstanbul’s mayoral elections. This could test further whether separation of powers exists. Index on Censorship, 48(2),  8–​10. Jeffery, C. (Ed.). (2015). The regional dimension of the European Union: towards a third level in Europe? London: Routledge. Huntington, S.P. (1993). The clash of civilizations. Foreign Affairs, 72(3),  22–​49. Johnston, R. (1979). Class conflict and electoral geography. Antipode, 11(3),  36–​43. Johnston, R.J., & Pattie, C. (2006). Putting voters in their place: Geography and elections in Great Britain. Oxford: Oxford University Press. Johnston, R., Rossiter, D., Pattie, C., & Dorling, D. (2002). Labour electoral landslides and the changing efficiency of voting distributions. Transactions of the Institute of British Geographers, 27(3), 336–​61. Kaya Osmanbaşoğlu, G., & Bekaroğlu, E.A. (2019). Türkiye’de referandumlar. Ankara: Orion. Krehbiel, E. (1916). Geographic influences in British elections. Geographical Review, 2(6), 419–​32. Kudat, A. (1975). “Patron–​client relations: The state of the art and research in eastern Turkey.” In Akarlı, E.D. and Ben-​Dor, G. (Eds.), Political participation in Turkey (pp. 61–​87). İstanbul: Boğaziçi University Press. Mardin, Ş. (1973). Center-​periphery relations:  A key to Turkish politics. Daedalus, 102(1), 169–​90. Özbudun, E., & Tachau, F. (1975). Social change and electoral behavior in Turkey: Toward a “critical realignment”?. International Journal of Middle East Studies, 6(4), 460–​80. Özbudun, E. (1976). Social change and political participation in Turkey. Princeton, NJ: Princeton University Press. —​—​—​. (2000). Contemporary Turkish politics: Challenges to democratic consolidation. London: Lynne Rienner Publishers. —​—​—​. (2006). Changes and Continuities in the Turkish Party System. Representation, 42(2), 129–​137. —​—​—​. (2013). Party politics and social cleavages in Turkey. London: Lynne Rienner Publishers. —​—​—​. (2015). Social change and political participation in Turkey. Princeton, NJ: Princeton University Press. Özcan, Y.Z. (2000). Determinants of political behavior in İstanbul, Turkey. Party Politics, 6(4), 505–​18. Pattie, C.J., & Johnston, R.J. (2003). Local battles in a national landslide: Constituency campaigning at the 2001 British general election. Political Geography, 22(4), 381–​414. Sayarı, S.  (2007). Towards a new Turkish party system? Turkish Studies, 8(2), 197–​210. Secor, A. J. (2001). Ideologies in crisis:  Political cleavages and electoral politics in Turkey in the 1990s. Political Geography, 20(5), 539–​60. Sözen, Y. (2019). Competition in a populist authoritarian regime: The June 2018 dual elections in Turkey. South European Society and Politics, 24(3), 287–​315. Şekercioğlu, E., & Arikan, G. (2008). Trends in party system indicators for the July 2007 Turkish elections. Turkish Studies, 9(2), 213–​31.

Introduction: Turkey’s electoral geography  23 TBMM. (2019). Türkiye Cumhuriyeti Milletvekili Genel Seçimleri. www.tbmm.gov.tr/​ develop/​owa/​genel_​secimler.secim_​cevreleri?p_​secim_​yili=1950 Tezcür, G.M. (2012). Trends and characteristics of the Turkish party system in light of the 2011 elections. Turkish Studies, 13(2), 117–​34. Waterbury, J. (1992). Export-​led growth and the center-​right coalition in Turkey. Comparative Politics, 24(2), 127–​45. Webster, G.R. (2000). Playing a game with changing rules:  geography, politics and redistricting in the 1990s. Political Geography, 19(2), 141–​61. Wuthrich, F.M. (2013). An essential center-​ periphery electoral cleavage and the Turkish party system. International Journal of Middle East Studies, 45(4), 751–​73. —​—​—​. (2015). National elections in Turkey:  People, politics, and the party system. Syracuse, NY: Syracuse University Press. Wuthrich, F.M., & Ingleby, M. (2020). The pushback against populism: Running on “radical love” in Turkey. Journal of Democracy, 31(2),  24–​40. Yavuz, M.H., & Ozcan, N.A. (2019). 2019 elections in Turkey: End of the Erdoğan Era? Middle East Policy, 26(2),  85–​94. Yüksek Seçim Kurulu (YSK). (2020). Cumhurbaşkanlığı Seçimi ve 27. Dönem Milletvekili Genel Seçimi. www.ysk.gov.tr/​tr/​24-​haziran-​2018-​secimleri/​77536 —​—​—​. (2020) Milletvekili Genel Seçim Arşivi. www.ysk.gov.tr/​tr/​milletvekili-​genel-​ secim-​arsivi/​2644

2  Measuring party nationalization in Turkey 1950–​2018 Özhan Demirkol and Edip Asaf Bekaroğlu

Introduction The geographical distribution of electoral support for political parties in Turkey has attracted scholarly attention for more than four decades. This attention to geographical distribution reached its peak after the 2007 parliamentary elections with the rise of Adalet ve Kalkınma Partisi (the Justice and Development Party, the JDP) as a highly nationalized party (Tezcür, 2012), Cumhuriyet Halk Partisi (the Republican People’s Party, the RPP) whose support has effectively become entirely localized—​one might even say “stuck”—​in the western coastal cities (Gökmen & Tosun, 2015), and the consolidation of the Kurdish political party1 as regional to southeast Turkey (Alptekin, 2018). Since then, a three-​or four-​colored electoral map of Turkey has commonly been employed by media analysts.2 The extent to which the major political parties are represented throughout the country has also been a subject of debate among politicians since the 2011 elections. During the 2011 electoral campaigns, Recep Tayyip Erdoğan, the leader of the ruling JDP, argued that the major opposition parties were not even able to conduct meetings in provinces “east of Sivas”, but the JDP was either the first or second party in all the provinces in Turkey.3 Erdoğan has constantly repeated this argument in the subsequent elections. In short, there have been two obvious trends in post-​2002 political life in Turkey; while the JDP has obtained electoral ascendancy in most of the regions, the support bases of the opposition parties have become more or less regionally concentrated (Çarkoğlu, 2011; Kumbaracıbaşı, 2014; Özen & Kalkan, 2017). Rather than analyzing the regional electoral performances of the Turkish political parties, this study aims to measure the nationalization of political parties and the party system in Turkey. A  political party is defined as nationalized if its national vote share is dispersed along the electoral districts in an essentially balanced manner. When a party system is dominated by nationalized parties, it is defined as a nationalized party system. Our motivation to measure the nationalization of the political party system is threefold. First, except for relatively limited attempts by Özen and Kalkan (2017) and Kollman et al. (2019), there are no studies to our knowledge that have attempted to measure this relatively new indicator in the instance of Turkey.

Measuring party nationalization in Turkey  25 Therefore, this chapter is intended to fill an important gap in the literature by measuring party nationalization at the national and party levels for all the parliamentary elections in Turkey between 1950 and 2018. Secondly, measuring the party nationalization in Turkey may help to explain certain characteristics of Turkish politics. The literature suggests that political polarization increases nationalization (Abramowitz & Webster, 2016), where the legislators and candidates are more likely to be concerned about the national issues than regional ones in highly nationalized party systems (Hicken et al., 2008, p. 7; Jones, 2005, p. 14), the differences in party system nationalization may influence politicians’ strategies regarding the provision of public goods and spending (Canavan, 2010; Hicken et al., 2008; Jones, 2005), the growth rate of foreign direct investment may be negatively influenced by party system nationalization (Simmons & Hicken, 2011), and a nationalized party system is beneficial to democratic consolidation (Jones, 2005, p.  15). In short, the party nationalization may be a significant variable by which to understand various aspects of Turkish politics. Finally, measuring the nationalization of political parties has acquired a particular importance for contemporary Turkish politics since the 2017 referendum that resulted in the transition from parliamentarism to presidentialism. Under the amended constitution, Turkey experienced a presidential and general election on June 24, 2018. Although it is still quite early to observe how presidentialism has impacted party nationalization in Turkey, theoretically it might be expected to accommodate local voting in legislative elections as the electoral fates of the executive and legislative branches are independent (Morgenstern & Swindle, 2005). This study, therefore, is an attempt to measure nationalization at the national and party levels for the parliamentary elections in Turkey, which does not only help our understanding of the major characteristics of Turkish politics, but also enhances Turkey’s value for comparative studies. With these motivations, the chapter first discusses the concept of nationalization of political parties, then looks at the state of play in the literature that deals with the regional dimensions of the electoral behavior and political parties in Turkey. Finally, it measures the nationalization of the Turkish party system from 1950 to 2018 by using the Standardized Party Nationalization Score (sPNS) of Bochsler (2006; 2010b) and Party Nationalization (PNS) and Party System Nationalization Scores (PSNS) of Jones and Mainwaring (2003). In the conclusion, we discuss the implications of the ups and downs in the nationalization scores.

Measuring the nationalization of the party system Party and party system nationalization have been measured using four distinct indices:  Competition Indices, Indices of Variance, Distribution Coefficients, and Inflation Measures (Bochsler, 2010b).4 Caramani’s (2004) Territorial Coverage Index, Rose and Urwin’s (1975) Number of Uncontested Districts, and Lago and Montero’s (2014) and Lago-​Peñas and Lago-​Peñas’s

26  Özhan Demirkol and Edip Asaf Bekaroğlu (2009) Local Entrant Measure are some prominent examples of Competition Indices. These indices, capturing the districts that the parties run in elections, are not useful for Turkey because its Political Parties Law entails spatial requirements for political parties to appear on the ballot.5 Therefore, except for the period between the 1950 and 1961 elections, the competition indices are far from explaining the nationalization at the party and the national levels in Turkey. Another family of nationalization indices is based on the difference between the size of the regional and the national party system. The works of Cox (1999), Chhibber and Kollman (1998; 2004), Simmons and Hicken (2011), Hicken et al. (2008), and Cosano (2011) use the Inflation Indices, which are only suitable for measuring party nationalization at the national level, rather than at the party level. Since this study aims to capture party nationalization at both the party and the national level, we ultimately did not employ the Inflation Indices either. Indices of Variance, on the other hand, are based on the variance in the vote share of political parties across electoral districts and are used by Rose and Urwin (1975) and Vasselai (2009) in order to capture the deviation of regional electoral results from the national ones. Finally, the Distribution Coefficients are used to analyze uneven distributions of vote shares across electoral districts. Jones and Mainwaring (2003) and Jones (2005) used the Inverted Gini Coefficient to measure nationalization at the party and the national levels. Both the Indices of Variation and the Inverted Gini Coefficient provide information about nationalization at the party level; however, they do not take different sizes of territorial units into account, and both are insensitive to the number of territorial units (Bochsler, 2010b). Nevertheless, the PNS and the PSNS indices of Jones and Mainwaring (2003), which are based on the Gini-​coefficient, are useful to trace the changes in a party’s level of nationalization over time, and to compare the parties within the same country or across countries. This study employs two types of index to measure party nationalization. Firstly, we used the PNS and PSNS indices of Jones and Mainwaring (2003). Although these indices are insensitive to the size and number of territorial units, a number of studies have nevertheless used them to calculate party nationalization in various countries; therefore, the PSN and the PSNS indices should help us to compare Turkey with other countries. Secondly, we calculated the Standardized Party Nationalization Score (sPNS) proposed by Bochsler (2006; 2010b), since the sPNS is sensitive to the change in the number of electoral units and the population size of the electoral districts compared to the PNS and the PSNS.

Determinants of Party System Nationalization There are two main schools that analyze the causes of the PSN and the variance of PSN scores across nations:  the sociological approach, which

Measuring party nationalization in Turkey  27 centers on the structural explanations for party formation, and the institutional approach, which deals with the institutional incentives conditioning the nationalization of parties (Cosano, 2011, p. 8). The first line of scientific enquiry underlines the importance of the transformation of the social cleavages for the nationalization of parties; if the cleavages articulate territorial differences or the cleavage groups are geographically concentrated, a weakly nationalized party system is expected to emerge (Caramani, 2004; Hicken et al., 2008, p. 1; Morgenstern et al., 2009, p. 1328). Scholarly findings support the argument that territorial cleavages with geographically concentrated social groups erode party system institutionalization. There are robust findings that indicate that ethnic fractionalization, geographical concentration of ethnic minorities, and a politicized ethnic conflict are the primary factors hindering party system nationalization (Bochsler, 2006; Vasselai, 2009; Harbers, 2008). On the other hand, Wahman’s (2015) study of Africa found that social and institutional factors such as low average district magnitude, large territory size, high ethnic fractionalization, and low levels of urbanization were associated with lower party nationalization for the opposition parties in Africa but had no impact on the party nationalization scores of the incumbent parties. Tiemann (2012, p.  86), however, found no meaningful relationship between the level of ethnic fragmentation and the nationalization of post-​communist parties. The second major school of thought regarding nationalization underlines the importance of the institutional factors that condition political actors’ decisions to nationalize. Among the institutional factors, economic and political centralization are the most referred-​to reasons (Chhibber & Kollman, 1998; 2004). Accordingly, fiscal and political decentralization creates the opportunities for the formation of subnational party systems, which, in turn, leads to a decrease in party system nationalization (Golosov, 2016; Harbers, 2008; Lago & Montero, 2014). Cosano (2011), however, argues that decentralization decreases party system nationalization only in countries with candidate-​ oriented electoral laws. Another conditional support for this hypothesis comes from Schakel (2013) who finds that decentralization has a significant effect on the nationalization of regional elections rather than the nationalization at the party or the national levels. Yet there are other studies that challenge the relationship between decentralization and party system nationalization. Lago-​Peñas and Lago-​Peñas (2009) and Lago-​Peñas et  al. (2014) found that the impact of decentralization on party system nationalization was not automatic and universal. Finally, Bochsler (2006) questioned the proposed relationship between decentralization and lower party system nationalization by arguing that ethno-​religious parties transformed the state into a decentralized system in Central and Eastern Europe, which implies that decentralization is not the cause but the result of lower party system nationalization. In addition to political and economic centralization, the electoral system is also cited as a determinant for party system nationalization. Carey and

28  Özhan Demirkol and Edip Asaf Bekaroğlu Shugart (1995) suggested that close-​ list proportional representation with controlled party nominations will result with a more nationally focused policy strategy. The argument of Carey and Shugart (1995) was verified by many studies that single-​ member plurality systems decrease party nationalization (Golosov, 2014; Riera, 2018; Vasselai, 2009), or party-​list proportional representation systems led to higher levels of party system nationalization (Golosov, 2016, p. 257). In contrast, Morgenstern and Swindle (2005, p. 165) and Morgenstern et al. (2009, p. 1334) found that the electoral systems had only limited effect on local votes. Regarding the impact of the legal thresholds on party system nationalization, Bochsler (2006) found that the national legal thresholds reinforced party nationalization in Central and Eastern Europe. Similarly, Cosano (2011) found that the district magnitude and the national thresholds were positively associated with party system nationalization in western Europe. On the other hand, Lavery (2014) analyzed 27 African countries and found that using party laws to establish nationalized parties, which transcend ethno-​regional cleavages, resulted in greater party system nationalization. With regard to the prevalent governmental system, presidentialism is expected to encourage local voting in legislative elections. Nevertheless, the literature provides only partial support for this expectation. While Golosov (2016, p.  253) and Vasselai (2009) found that presidentialism had no significant impact on party system nationalization, Morgenstern and Swindle (2005, p. 164) found that executive type and federalism, combined with ethnic fractionalization, played the most dominant roles in increasing local voting in legislative elections. Morgenstern et al. (2009), on the other hand, found that “dynamic nationalization” significantly decreased under presidential systems while “static nationalization” was influenced mainly by the electoral system rather than the form of the executive. Although Turkey recently underwent a transition from parliamentary system to presidentialism in 2017, we did not expect a significant twist in nationalization at the party and the national levels in the 2018 elections since the Turkish presidential system is combined with unitarism, a proportional representation electoral system, and disciplined political parties. Other factors may have certain impacts on party system nationalization as well. Jurado and Leon (2017) found that lower economic growth prior to a set of elections can be associated with a decrease in party system nationalization. Lavery (2014, p. 208) underlined the impact of party system fragmentation during the transition to democracy on party system nationalization in 27 African countries, finding that those transitioning to democracy with high party system fragmentation were characterized by low party system nationalization. Similarly, Tiemann (2012, p. 88) noted the importance of the transitional context on party system nationalization. Lavery (2014) also found that nationalist independence leaders who implemented ethnic balancing strategies also contributed to party system nationalization. Borz and de Miguel (2017) analyzed the impact of organizational structures of political parties on party system nationalization and found

Measuring party nationalization in Turkey  29 that ideological diversity within a party and consensus amongst its leadership increased party nationalization.

The Turkish case While the recent literature has been increasingly occupied with the nationalization of political party systems, most studies analyzing the Turkish party system and electoral behavior focus on the domination of a party, the fragmentation and polarization of the system, or the volatility of voter preferences (Akgün, 2001; Çarkoğlu & Hinich, 2006; Sayarı, 2007; Tachau, 2000). Among the earlier works dealing with Turkey’s electoral geography, Özbudun and Tachau (1975) found that while Adalet Partisi (the Justice Party, JP) and the RPP, the mainstream parties of the period, received support from urban areas or the developed districts in Marmara and Aegean regions, the newer and radical parties, including Milli Selamet Partisi (the National Salvation Party, NSP), with its Islamist orientation, and Milliyetçi Hareket Partisi (the Nationalist Action Party, NAP), with its far right ideology, performed better in rural areas and underdeveloped districts in the southeast or northeast regions in the parliamentary elections between 1950 and 1973. According to Özbudun and Tachau (1975), the RPP increased its votes from the 1950s to the 1970s all across the country, with its more impressive outcomes observed in the Marmara, Mediterranean, and southeastern regions, while the JP lost some of its votes, especially among the urban poor and less developed regions, in favor of smaller right-​wing parties. More recently, Çarkoğlu (2002) argued that geographical regionalization had become a major characteristic of elections in Turkey. According to Çarkoğlu (2002, p. 33), Turkish provinces may be categorized into three groups for the period between the 1950 and 1999 elections: the east and southeastern provinces, the coastal provinces, and the remaining provinces. The least-​developed east and southeastern provinces reflect traditional identities through their support of Kurdish candidates or minor oppositional parties, while the coastal provinces prefer centrist parties and candidates. Finally, the central Anatolian provinces are more likely to support Turkish nationalist and Islamist candidates. The volume of such studies has grown since the 2000s when the JDP became the dominant political party in Turkey. Depicting the JDP as a peripheral force, some studies have analyzed Turkish electoral geography in terms of a center-​periphery cleavage. Özbudun (2006, p.  134), for example, argued that the “Turkish party system can still be accurately described as being based on a center-​periphery conflict”, where the JDP represent the periphery and stand as “a broadly-​based national party” that appeals to a wide segment of centrist votes together with Islamists. However, Wuthrich (2013) questioned the explanatory power of the center-​periphery cleavage in Turkish politics. By analyzing the electoral outcomes of the Turkey’s initial multiparty period, where the center-​periphery cleavage would be expected to be prevalent, Wuthrich (2013, pp. 757–​9) found that while the occupational difference between RPP and DP candidates was not significant, the RPP saw relatively

30  Özhan Demirkol and Edip Asaf Bekaroğlu greater success in less developed regions. Secor (2001), on the other hand, argued that rather than some mooted center-​periphery cleavage, the left-​right orientation, fed by a combination of identity issues and economic concerns, has been shaping the electoral geography in Turkey since the 1990s. Indeed, the impact of identity politics on Turkey’s electoral geography came to the fore more apparently than ever in the 2000s. Consequently, some regions increasingly coincided with certain identity preferences, which, in turn, triggered new kinds of studies that directly dealt with the electoral geography in Turkish politics. These relatively newer studies (Akarca & Başlevent, 2011; Çarkoğlu, 2000; Şekercioğlu & Arıkan, 2008; West, 2005) usually employed cluster analysis. In a pioneering analysis, Çarkoğlu (2000) grouped all the provinces for the 1999 general elections into seven clusters and found that Demokratik Sol Parti (the Democratic Left Party, DLP), the winner of the 1999 elections, relied heavily on a single cluster that mostly included the provinces in the Marmara and Aegean regions while the share of the votes for other parties were more evenly distributed across clusters, although each was better represented in one or two (Çarkoğlu, 2000, p. 159). For Çarkoğlu, there were also some sociocultural and ideological trends in the clusters, namely a significant “pro-​Islamist and nationalist” inclination in the clusters containing the eastern and Central Anatolian provinces, a pro-​ Kurdish inclination in the cluster formed from the southeastern provinces, and center-​left and center-​right inclinations in the western provinces. Compared to the study of Özbudun and Tachau (1975), the center-​left had moved its support base from less-​developed regions to more developed ones by the end of the 1990s. West (2005) argued that the issues of economic development and cultural identity have converged to create identifiable regional divisions and these regional patterns appear to persist despite the general volatility of the party system. For West (2005, p. 518), the 2002 election signaled a move from the fragmentation and mixed regional patterns of the 1980s and the 1990s to a concentration of voting behavior into three regions and one or two dominant parties. Among these three regions, the one that predominantly prefers pro-​Islamist and Turkish nationalist parties tended to expand from Central Anatolia in all directions and lay at the core of the JDP’s increasing domination and nationalization of the Turkish party system. Şekercioğlu and Arıkan (2008) and Akarca and Başlevent (2011) also employed cluster analysis to study electoral outcomes in Turkey. Both studies agreed with previous ones that while the JDP dominates the majority of the clusters without any particular serious challenge, the runner-​up was the RPP in the western regions, the NAP in central and northern regions, and the pro-​Kurdish parties/​candidates in the southeastern regions. Interestingly, Akarca and Başlevent (2011) argue that the cluster dynamics remained mostly unchanged despite Turkish politics experiencing a major realignment between 1999 and 2009. All these studies, with their various methodologies, point to the spatial dependency of electoral behavior and political parties in Turkey. Based on these studies, we may know which parties performed best in each province,

Measuring party nationalization in Turkey  31 or indeed might identify the clusters of provinces in which different party families performed better—​ or indeed worse—​ in any particular election. However, these indicators are far from giving us the picture of the other side of the coin, that is, the extent of the nationalization of political parties and the political party system in Turkey. It is not appropriate to conclude which party is the most nationalized merely by considering the number of the votes and the position each party achieved in each province. Two recent studies have attempted to overcome this problem by using the party nationalization indices. Based on the data provided by the Constituency-​ Level Elections Archive, Kollman et al. (2019) calculated the extent of party nationalization at the party, constituency, and national levels for 156 countries, including Turkey. This impressive project calculated the party nationalization scores for Turkey between 1950 and 2015 and enables researchers to conduct comparative studies. However, the data provided by the project on Turkey has three minor problems. So far, the project has not provided data on party nationalization in Turkey for the 2018 elections. More importantly, it did not calculate party nationalization scores for parties that received less than 5  percent of national vote in the elections. Finally, the data provided by the project does not reflect the actual number of electoral districts for the elections between 1991 and 2011.6 We have attempted to deal with these minor problems in order to reflect a more accurate picture of party nationalization in Turkey. Another attempt to measure the nationalization of political parties in Turkey came from Özen and Kalkan (2017), who studied the general elections from 2002 to 2015. Özen and Kalkan used the Effective Number of Parties index (ENPn) and party nationalization of the most geographically central party to measure the nationalization and regionalization dimensions of major political parties, and found that the major political party that had geographically expanded its electoral appeal between 2002 and 2015 was the JDP (Özen & Kalkan, 2017, p. 375). The strength of this study comes from the fact that it analyzed sub-​provincial results from parliamentary elections in order to avoid ecological fallacy (i.e., the bias to infer individual voting behaviors based on aggregated outcomes). Nevertheless, this study is limited in terms of its time span and due to leaving the smaller political parties out of analysis. It only covers five general elections from 2002 to 2015, and only provides findings for the JDP, RPP, NAP, and Halkların Demokratik Partisi (the People’s Democratic Party, PDP). Neither does it measure the nationalization of the political party system as a whole. For this reason, we believe that the analysis of party nationalization that covers the whole multiparty period in Turkey may contribute significantly to the literature.

Analysis and findings In this study, we intended to measure the extent to which the Turkish parties and party system have been nationalized. We look at all parliamentary elections starting from the 1950 elections, which means we will also analyze

32  Özhan Demirkol and Edip Asaf Bekaroğlu any fluctuations in nationalization. For the parliamentary elections between 1950 and 1957, we used data provided by the TÜİK (2008). For the elections between 1961 and 2007, we used data provided by the “Global Elections Database” (Brancati, 2016) and verified the electoral results using the official provincial results provided by the TÜİK (2008). For the 2011, 2015,7 and 2018 elections, we used the official electoral results published in the Official Gazette. While collecting data, we took the mergers, party splits, changes in the names of the parties, and the electoral coalitions into account. In case when a party was outlawed, we calculated the sPNS, PNS, and PSNS scores for the successor party. Data on mergers, splits, name changes, and electoral coalitions for the period between 1950 and 2015 was obtained from the work of Kaynar (2007) and Demirkol (2015). Data for the post-​2015 period was obtained primarily by conducting a media survey. Given that all Turkish parties were closed in 1981 after the 1980 military coup, we analyzed the Turkish parties over two periods, namely 1950–​1977 and 1983–​2018. To calculate party nationalization and party system nationalization, we used the template formed by Bochsler (2010a). This template calculates Bochsler’s (2010b) sPNS, Laakso and Taagepera’s (1979) index of effective number of parties by their votes at the national level and at the district level, Caramani’s (2004) Territorial Coverage Index, Chhibber and Kollman’s (2004) Inflation Score, Cox’s (1999) Inflation Index, and Rose and Urwin’s (1975) Number of Uncontested Districts. Additionally, we prepared an excel template in order to calculate the PNS and PSNS of Jones and Mainwaring (2003). Table  2.1 presents the indicators to evaluate the characteristics of party system nationalization in Turkey. Turkey has intermediate levels of party system nationalization, with an average sPNS of 0.857 and PSNS of 0.755. Turkey’s average PSNS score is similar to those of Guatemala (0.76), Venezuela (0.76) and Bolivia (0.76) in the Americas (Harbers, 2008; Jones & Mainwaring 2003); Mozambique (0.81) and South Africa (0,78) in Africa (Wahman, 2015); Australia (0.75) (Canavan, 2010); Estonia (0.77), Poland (0.78), and Romania (0.78) in Central and Eastern Europe (Harbers, 2008); and Ireland (0.75), Norway (0.76), and Portugal (0.77) in Europe (Lago-​Peñas et  al., 2014). If we take into account the average inflation score of Chhibber and Kollman

Table 2.1 Measures of stability in Party System Nationalization (1950–​2018) Average sPNS (Bochsler, 2010b) Average PSNS (Jones & Mainwaring, 2003) Average inflation score (Chhibber & Kollman, 1998) Average inflation index (Cox, 1999) Computed by the authors. *  The larger the figure, the higher the nationalization **  The larger the figure, the poorer the nationalization

0.857* 0.755* 0.37** 0.09**

Measuring party nationalization in Turkey  33 to compare Turkey and western European countries, Turkey’s party system nationalization score of 0.37 again falls into the intermediate levels together with the UK (0.30) and Portugal (0.42) (Cosano, 2011) and resembles those of Denmark (0.35) and Portugal (0.43) (Lago & Montero, 2014). Finally, if we compare the average inflation index of Turkey with the findings of Hicken et al. (2008), who used the same index, Turkish party system nationalization (0.09) is similar to the United States (0.09) and South Africa (0.09).8 Our finding of the intermediate level of party system nationalization in Turkey is also corroborated by Vasselai (2009), who uses a modified version of Rose and Urwin’s (1975) Cumulative Regional Inequality index. It should be noted that the development of party system nationalization in Turkey has been interrupted by military coups. Figure 2,1 shows the destabilizing effects of these military interventions on the evolution of the party system. The peak and the lowest sPNS and PSNS occurred during the 1983 and 1961 elections that took place after the coups. A decrease of the sPNS and the PSNS in the 1961 elections is primarily the result of the military regime banning the DP, the governing party with high levels of nationalization. Although the JP and Yeni Türkiye Partisi (the New Turkey Party, NTP) were founded to compete for the votes of the outlawed DP, the short period between their formation and the initial elections may have prevented them appealing to the DP base. In contrast, the PSNS and the sPNS reached their peak in the first elections after the 1980 coup. The military regime outlawed all the political parties and leaders, and permitted only three new political parties to compete in the 1983 elections, where also a 10 percent national threshold was in effect. The sPNS and PSNS scores indicate that Turkish voters had an even preference for the permitted parties across the districts. Consequently, the sPNS and the PSNS reached their peaks. Therefore, we may conclude that the military interventions in 1960 and 1980 disturbed the natural evolution of party system nationalization in Turkey. Despite the effects of the military intervention on the stability of party system nationalization, Turkey shows intermediate levels of party nationalization. However, Figure 2.1 shows that the Turkish party system has gradually evolved through a less nationalized system, indicating greater regional concentration of the voter base for all parties. The average sPNS, which was 0.901 in the 1950s, decreased to 0.835 between the 1961 and 1977 elections. This is contrary to the expectations of Carey and Shugart (1995) and the findings of Golosov (2016), who argued that proportional representation electoral systems tend to encourage party nationalization. Although the Turkish electoral system was transformed from a multi-​member district plurality system in the 1950s to one of proportional representation in the 1960s, the PSNS and the sPNS scores decreased slightly in the 1960s. Furthermore, the introduction of spatial requirements did not increase party system nationalization in Turkey. Although parties had no spatial requirements to appear on the ballot in the 1950s, they had to organize and nominate at least 15 of the 67 electoral districts in the 1960s and the 1970s. This spatial requirement,

0.950

8.00

0.900

7.00

0.850

6.00

0.800

5.00

0.750

4.00

0.700

3.00

0.650

2.00

ENEP

sPNS and PSNS

34  Özhan Demirkol and Edip Asaf Bekaroğlu

sPNS

PSNS

ENEP

Figure 2.1 Party system nationalization and effective number of parties in Turkey (1950–​2018) This graph and the following graphs in Figures 2.2–​2.4 were created by the authors using the electoral outcomes obtained from the electronic archives of Supreme Election Council (Yüksek Seçim Kurulu, YSK, 2020) and Turkish Statistical Institute (Türkiye İstatistik Kurumu, TÜİK, 2008).

which was expected to force parties to organize in all electoral districts in order to appear on the ballot, unexpectedly resulted in a slight decrease in party nationalization. However, more importantly, the figure shows that party system nationalization was negatively correlated with the effective number of parties. As the number of the competing parties increased, party system nationalization decreased in Turkey. The finding confirms the studies of Jones and Mainwaring (2003), Vasselai (2009) and Golosov (2016, p.  257), who found a negative correlation between party system fragmentation and party system nationalization. The average sPNS between the 1983 and 2018 elections was 0.854, which is lower than the 1950s and slightly higher than the period between 1961 and 1977. This slight increase may be explained by reference to institutional factors. The spatial requirements to appear on the ballot increased from organizing and nominating candidates of at least 15 of the 67 electoral districts to 50 percent of the electoral districts. Besides, as proposed by Bochsler (2006) and Cosano (2011), the introduction of the 10 percent national threshold in 1983 seems to have increased party system nationalization to a small extent. However, the figure illustrates a negative correlation between the effective number of parties and party system nationalization between the 1983 and June 2015 elections. Furthermore, the figure shows a decreasing trend in party system nationalization from the 1980s to the 1990s and the 2000s. This, as

Measuring party nationalization in Turkey  35 will be explained below, is primarily related to the politicization of the ethnic cleavages and the geographical concentration of Turkey’s ethnic minorities.

Nationalization at the party level Figure 2.2 shows the sPNS scores of the major Turkish parties between the 1950 and 1977 elections. It is clear that the DP was a highly nationalized party with an average sPNS of 0.951. The RPP had an average sPNS of 0.932 between the 1950 and 1957 elections, and this average dropped down to 0.914 between the 1961 and 1977 elections. Nevertheless, the RPP was also a highly nationalized party with an average sPNS of 0.921 between 1950 and 1977. The JP was slightly less nationalized than the RPP with an average sPNS of 0.878 between 1961 and 1977. If we exclude the 1961 elections, the average sPNS of the JP increased to 0.900. In both cases, the JP was a slightly less nationalized party than the RPP. Figure 2.2 also demonstrates the nationalization of the main competitors of the major Turkish parties. The NTP, which competed with the JP for the votes of the outlawed DP, never became a nationalized party, with an average sPNS of 0.495. The Demokratik Parti (the Democratic Party, DkP), which split from the JP in 1970 and argued to be the heir to the outlawed DP, also had a slightly lower average sPNS of 0.801. On the other hand, the Güven Partisi (the Reliance Party, RP), which split from the RPP in 1967 and changed its name to Cumhuriyetçi Güven Partisi (the Republican Reliance Party, RRP) in 1973, was again less nationalized than its main competitor, with an average sPNS of 0.728. In this sense, the competitors of the JP and

1.000 0.900 0.800 0.700 0.600 0.500 0.400

1950

1954

1957

1961

1965

1969

1973

DP (1950-57)

RPP (1950-77)

JP (1961-1977)

NTP (1961-69)

RP-RRP (1969-77)

DkP (1973-77)

Figure 2.2 sPNS of the main Turkish parties (1950–​1977)

1977

36  Özhan Demirkol and Edip Asaf Bekaroğlu the RPP never became highly nationalized parties. Indeed, their impact on party system nationalization was ultimately achieved through their impact on party system fragmentation. The 1980 military regime reinitialized the multiparty elections in 1983, but did not permit the pre-​1980 parties to run. It also wanted to reduce the fragmentation and instability of the system. For that reason, a 10 percent national threshold was introduced, which was intended to create a two-​party system where the far right and far left would become insignificant players. In the 1983 elections, only three parties were allowed to compete: Anavatan Partisi (the Motherland Party, MP) representing the center-​right, Halkçı Parti (the Populist Party, PP) representing the center-​left, and Milliyetçi Demokrasi Partisi (the Nationalist Democracy Party, NDP), which was established by retired officers. Nevertheless, with the gradual removal of the ban on the pre-​ 1980 leaders and parties, the party system became increasingly fragmented until 2002. Between the 1987 and 2002 general elections, the center-​right was represented by the MP and Doğru Yol Partisi (the True Path Party, TPP), the successor of the pre-​intervention JP, whereas the DLP and the PP, which first changed its name to Sosyal Demokrasi Partisi (the Social Democracy Party, SDP) in 1985 then merged with the re-​established RPP in 1995, competed for the center-​left votes. However, a major realignment took place with the 2002 elections and the JDP and the RPP became the new major parties in the period after 2002. Therefore, it is reasonable to analyze the post-​1983 period in terms of before and after 2002. Figure 2.3 shows that during the initial period of the noninstitutionalized two-​party system, the MP and the PP had high nationalization scores, where the MP (0.932) was slightly better off than the PP (0.917). The nationalization score of the MP reached its peak in 1987 at 0.942, but continuously dropped to 0.899 by 1999, together with its vote share. The same trend is true for the RPP and the TPP. Still, the MP scores were higher than all the other mainstream parties until 1999. The nationalization scores of the PP and its successors, namely the SDP and the RPP, dropped from 0.917 in 1983 to 0.844 in 1995, yet increased to 0.854 in 1999 when the RPP’s vote share dropped to 8.7 percent. The DLP saw a steadily increasing vote share from 1987 to 1999, while its nationalization score first dropped from 0.844 in 1987 to 0.810 in 1995, and then increased to 0.825 in 1999 when its votes reached their peak of 22.2 percent. Nevertheless, the sPNS of the RPP is higher in each election than the DLP between 1987 and 1999. Finally, the TPP kept its nationalization score at the same level to a large extent. Overall, the center-​right parties scored higher than center-​left parties in terms of the average nationalization scores from 1983 to 1999. While the MP and the TPP’s average nationalization scores were 0.927 and 0.878, the RPP and the DLP had a sPNS of 0.878 and 0.825, respectively. Turkish politics were faced with a major realignment during the 2002 elections. With the exception of the RPP, the parties that represented the center-​right and the center-​left could not win enough votes to obtain seats in

Measuring party nationalization in Turkey  37 1.000 0.950 0.900 0.850 0.800 0.750 0.700

1983

1987

1991 MP

1995 TPP

1999

2002

2007

PP-SDP-RPP

2011

2015 7 2015 1 JUNE NOV DLP JDP

2018

Figure 2.3 sPNS of the mainstream Turkish parties (1983–​2018)

parliament. Furthermore, the JDP, which split from the Islamist Refah Partisi (Welfare Party, WP), the successor of the NSP of the 1970s, has since been the primary representative of the center-​right. Although the MP and the TPP merged in 2007 under the name DP in order to compete with the JDP for center-​right votes, it has not since succeeded in this attempt. After 2002, there were two significant trends concerning nationalization of the major parties. First, the JDP increased its sPNS together with its vote share. However, the sPNS of the JDP, which was 0.903, was slightly lower than the average sPNS of the DP (0.951) in the 1950s and the MP (0.927) in the 1980s and the 1990s, but higher than the JP (0.878) of the 1960s and the 1970s and the TPP (0.829) in the post-​1983 period. Compared to the RPP, which has an sPNS of 0.853 for the post-​1983 period, the JDP is a more nationalized party. Secondly, the RPP turned into a less nationalized party. Although the average sPNSs of the RPP were 0.932 and 0.914 for the 1950–​ 57 and the 1961–​77 periods, the party became nationalized with averages of 0.878 and 0.854 for the 1983–​99 and 2002–​18 periods, respectively. Finally, the Islamist, nationalist, and pro-​Kurdish parties, which have been important political actors since the 1970s and the 1990s, deserve attention. Until the mid-​1960s and early 1970s, nationalist and conservative voters were represented primarily by the DP and the JP. However, Millet Partisi (the Nation Party, NP), and its successors, namely Cumhuriyetçi Millet Partisi (the Republican Nation Party, RNP) and Cumhuriyetçi Köylü Millet Partisi (the Republican Peasant Nation Party, RPNP), competed for a share of conservative voters. Following the leadership of Alparslan Türkeş in 1965, the RPNP changed its name to NAP in 1969 and adapted a more nationalist

38  Özhan Demirkol and Edip Asaf Bekaroğlu 0.900 0.850 0.800 0.750 0.700 0.650 0.600 0.550 0.500 0.450 0.400

1950 54

57 1961 65

69

NAP tradion

73

77 1983 87

91

NOM Tradion

95

99 2002 7

11 2015 15 2018 JUNE NOV

Pro-Kurdish Pares

Figure 2.4 sPNS of the Islamist, Nationalist, and pro-​Kurdish parties (1950–​2018)

outlook. Although the party was outlawed in 1981, it was revitalized as the Milliyetçi Çalışma Partisi (the Nationalist Working Party, NWP), which then changed its name to NAP in 1993. On the other hand, Milli Görüş Hareketi (the National Outlook Movement, NOM), which formed the NSP in 1972, was succeeded by the WP, Fazilet Partisi (Virtue Party, VP) and Saadet Partisi (the Felicity Party, FP) after each party was outlawed. Finally, the politicization of ethnic cleavages resulted in the formation of a number of successor pro-​Kurdish parties, which are currently organized under the PDP.9 Figure  2.4 shows that the Islamist and the nationalist parties remained relatively less nationalized between the 1950 and 1977 elections. The average sPNS of the Islamist NSP was 0.778, whereas the NAP scored 0.715 for the period between the 1961 and 1977 elections. Starting with the post-​ 1983 period, the sPNS of the Islamist and the nationalist parties gradually increased. The figure shows that nationalist parties have remained more nationalized than the Islamist parties since the 1980s. While the average sPNS of the NOM parties was 0.806, the major representative of the nationalist voters had an average of 0.853 for the post-​1983 period. However, compared to their competitors in the center-​right, the Islamist and the nationalist parties remained relatively less nationalized. On the other hand, the pro-​Kurdish parties remained highly regionalized with an average sPNS of 0.572 between the 1991 and 2018 elections. Therefore, we may conclude that the Islamist and the nationalist parties have increased their nationalization levels, especially since the 1980s, whereas the pro-​ Kurdish parties have effectively maintained their regional character.

Measuring party nationalization in Turkey  39

Discussion and implications This analysis of the nationalization of Turkish political parties and party system from 1950 to 2018 demonstrates that Turkey has an intermediate level of party system nationalization with an average Standardized Party Nationalization Score (sPNS) of 0.857 and Party System Nationalization Score (PSNS) of 0.755. On the other hand, we also found that the levels of the party system nationalization in Turkey have continuously dropped from an average sPNS of 0.901 in the 1950s to 0.839 in the 2010s. Although the development of the political parties in Turkey has been interrupted by two military coups in 1960 and 1980, and the electoral system has experienced engineering attempts such as prescribing a 10 percent national threshold to enter parliament or the spatial requirements for political parties to run in the elections, the drop in the party system nationalization has been one of the most important characteristics of the Turkish party system. These findings are interesting because, according to the literature, proportional elections with high national thresholds and spatial requirements tend to increase party nationalization. Therefore, it is legitimate to ask why the nationalization of the party system in Turkey has decreased overall. To answer this question, a closer look at the ups and downs of party nationalization is necessary. To begin with, the lowest and the highest nationalization scores, as recorded in the 1961 and 1983 elections, respectively, took place after the military rules. The dramatic decrease in the nationalization in the 1961 elections can be accounted for through the banning the DP, the most nationalized party of the 1950s. The parties claiming to be the heirs of DP had lower nationalization scores, as displayed in Figure 2.2. Nevertheless, banning the political parties did not result in a decrease in nationalization in the 1983 elections—​ rather it reached a peak. Therefore, arguing that banning parties should decrease nationalization is a premature conclusion. The institutional design may be a better explanatory factor. Indeed, the introduction of a proportional electoral system with D’hondt and national remainder methods encourages the representation of smaller parties. This system increased the number of effective parties, which in turn decreased the nationalization of the party system. In the 1983 elections, on the other hand, the military regime started the proportional electoral system with D’hondt and a 10  percent national threshold to decrease the number of parties in the Turkish parliament if not to create a two-​party system. To ensure this, the military regime allowed only three new parties to run in the 1983 elections. Although the Motherland Party won the majority of seats in parliament albeit the expectations of officers, the outcome was the highest party system nationalization score in Turkey’s multiparty history. This negative relationship between the number of effective parties and nationalization holds true when looking at the particular periods in Turkey’s electoral history. Figure  2.1 clearly displays this relationship. The higher sPNS and PSNS scores in Turkey are found when the number of effective

40  Özhan Demirkol and Edip Asaf Bekaroğlu parties (ENEP) was fewest. Indeed, as Turkey came closest to looking like a two-​party system in the 1950s and 1983, the nationalization scores reached their peaks of 0.901 and 0.917, respectively. On the contrary, the Turkish party system became less nationalized in the 1960s, 1970s, and 1990s when the party system was fragmented, and the number of effective parties increased. The average sPNS dropped to 0.835 between the 1961 and 1977 elections when the average number of effective parties increased from 2.32 in the 1950s to 3.36 in the 1960s and 1970s combined. Similarly, when the average ENEP increased to 5.86 in the 1990s, the 0.917 sPNS of 1983 sharply decreased to an average of 0.855. The only exception to this relationship was the recent election in 2018, which was the first national election under Turkey’s new presidential system. Contrary to our expectations, both the ENEP and the nationalization scores slightly increased in 2018. This is also contrary to the literature on government systems and party nationalization, where presidential systems are expected to encourage local voting—​in other words, to decrease party nationalization. A possible explanation for this lies behind the electoral regulations that necessitate holding the presidential and legislative elections on the same day and at the same ballot box. Combining this regulation with Turkey’s disciplined political parties and unitary system, it is actually not surprising that the presidential campaign took precedence over the legislative elections and uplifted the nationalization of political parties. Beside these general patterns of party system nationalization, our findings with regard to the nationalization scores of some particular parties are also significant. First of all, the nationalization levels of the mainstream political parties have decreased as well. Most significantly, the RPP has gradually lost its nationalized character since the 1980s (see Figure  2.3). Having an average sPNS of 0.932 in the 1950s and of 0.921 between 1961 and 1977, the RPP’s nationalization scores dropped to an average 0.862 in the 1990s and 0.831 since the 2002 elections. It is especially noteworthy that while the RPP’s sPNS was slightly higher than the JP, its main rival on the right, in the 1960s and 1970s, the mainstream right-​wing parties (the MP in the 1980s, the MP and TPP in the 1990s, and the JDP since 2002)  have appeared to be more nationalized than the PP, SDP, and RPP. Interestingly, even the NAP, a far right nationalist party, has become more nationalized with an average sPNS of 0.847 compared to the RPP since the 2002 elections. The RPP’s decreasing nationalization levels do not only coincide with an increase in ENEP, but also with the rise of identity politics since 1987 (see Figure 2.4). The last relatively high nationalization score recorded by the RPP (at that time, the SDP) was in the 1991 elections when the newly established Kurdish political party (at that time, the PLP) candidates appeared in the ballot under the SDP banner. However, after that, the Kurdish political parties have run in the elections either with independent candidates or as a party. This was the most important factor that decreased the RPP’s nationalization scores. In the southeast regions, where Kurdish voters are concentrated, the electorate

Measuring party nationalization in Turkey  41 either voted for the Kurdish political party or the JDP. As mentioned above, the central regions predominantly voted for conservative, Islamist, or nationalist parties. This left the RPP only the coastal provinces in the western part of Turkey. The higher nationalization scores of the far right NAP compared to the RPP can also be explained by the rise of the Kurdish political party as a regionalized actor, which was probably caused by a backlash among non-​ Kurdish voters in all regions to vote for NAP at certain levels. Overall, this study fills a gap in the literature by measuring the nationalization of the parties and the party system in Turkey, and by providing possible explanations for the ups and downs in the nationalization scores. There is a clear relationship between party system fragmentation and nationalization. In other words, the higher the effective number of political parties, the lower the nationalization levels. Secondly, the rise of identity politics also negatively influenced the nationalization of the party system and certain political parties, most notably the RPP, the oldest mainstream left-​wing party in Turkey. Nevertheless, further studies are needed to determine how some other political and economic dynamics affected, or indeed were affected by, the decreasing nationalization levels. Party system nationalization may be correlated with polarization, volatility, economic development, party splits and merges, public spending, clientelism, disciplined political parties, candidate/​ legislative behavior, or democratic consolidation. This study only constitutes a step towards new studies to understand other dimensions of the issue.

Notes 1 Then Demokratik Toplum Partisi (the Democratic Society Party, DSP), now Halkların Demokratik Partisi (the Peoples’ Democratic Party, PDP). 2 For an example, see Figure 1.1 of the Introduction. 3 For an example, see “Erdoğan, Çorum’da MHP’ye yüklendi”, Sabah, 18 May 2011. www.sabah.com.tr/​gundem/​2011/​05/​18/​biri-​ulkeyi-​biri-​sskyi-​batirdi 4 For a bibliography of the studies on party nationalization, see Kollman et al. (2019). 5 According to Article 36 of the Political Parties Law, a political party must establish its organizations in at least half of the districts in Turkey to enter the elections. 6 Until the 1991 elections, provinces were taken as electoral units in Turkey. However, starting with the 1991 elections, major provinces were divided into minor electoral units. Consequently, although there were 74 provinces, the number of electoral units in 1991 was 107. Between 1995 and 2015, provinces of Ankara and İzmir were divided into two electoral units and İstanbul had three electoral units. In the 2018 elections, the number of electoral units increased to 87 despite no change in the number of provinces. 7 There were two parliamentary elections in 2015. Since Turkey’s Grand National Assembly could not establish a government after the June 7 elections, President Erdoğan used his constitutional power and called for a snap election, which took place on November 1. 8 Studies by Hicken et al. (2008) and Lago-​Peñas et al. (2014) measured the party system nationalization for Turkey and found the inflation index and the PSNS as

42  Özhan Demirkol and Edip Asaf Bekaroğlu 0.17 and 0.703, respectively. However, none of the studies provide information about the number of observations made for Turkey. The disparities between the findings of our study and of these studies may be related with measuring different elections. 9 Although their roots can be traced back to the socialist parties in the 1960s, the first pro-​Kurdish party was the People’s Labor Party, which was succeed by the People’s Democracy Party, the Democratic People’s Party, the Peace and Democracy Party, and the Democratic Society Party after each party was outlawed by the Turkish Constitutional Court. Pro-​Kurdish parties nominated independent candidates in the 2007 and 2011 elections in order to overcome the 10 percent national threshold. Due to this reason, we considered the sPNS scores of the independent deputies as the sPNS scores of the pro-​Kurdish parties for the 2007 and 2011 elections.

References Abramowitz, A.I., & Webster, S. (2016). The rise of negative partisanship and the nationalization of US elections in the 21st century. Electoral Studies, 41,  12–​22. Akarca, A.T. & Başlevent, C. (2011). Persistence in regional voting patterns in Turkey during a period of major political realignment. European Urban and Regional Studies, 18(2), 184–​202. Akgün, B. (2001). Aspects of party system development system in Turkey. Turkish Studies, 2(1),  71–​92. Alptekin, H. (2018). Kurdish votes in the June 24, 2018 elections: An analysis of electoral results in Turkey’s eastern cities. Insight Turkey, 20(4), 211–​30. Bochsler, D. (2006). The nationalization of political parties:  A triangle model, applied on the central and eastern European countries. Political Science Journal, 4, 6–​37. —​—​—​. (2010a). Calculating Party Nationalisation. www.bochsler.eu/​pns/​ —​—​—​. (2010b). Measuring party nationalisation:  A new Gini-​based indicator that corrects for the number of units. Electoral Studies, 29(1), 155–​68. Borz, G., & de Miguel, C. (2017). Organizational and ideological strategies for nationalization:  Evidence from European parties. British Journal of Political Science, 49(4), 1499–​526. Brancati, D. (2016). Global Elections Database. www.globalelectionsdatabase.com Canavan, M. (2010). Party nationalization and the regional allocation of public infrastructure investment [Working Paper] www.psa.ac.uk/​sites/​default/​files/​1008_ 529.pdf Caramani, D. (2004). The nationalization of politics:  The formation of national electorates and party systems in western Europe. Cambridge: Cambridge University Press. Carey, J.M., & Shugart, M.S. (1995). Incentives to cultivate a personal vote: A rank ordering of electoral formulas. Electoral Studies, 14(4), 417–​39. Çarkoğlu, A. (2000). The geography of the April 1999 Turkish elections. Turkish Studies, 1(1), 149–​71. —​—​—​. (2002). Turkey’s November 2002 elections:  A new beginning? Middle East Review of International Affairs, 6(4),  30–​41. —​—​—​. (2011) Turkey’s 2011 general elections:  Towards a dominant party system?. Insight Turkey, 13(3),  43–​62.

Measuring party nationalization in Turkey  43 Çarkoğlu, A., & Hinich, M.J. (2006). A spatial analysis of Turkish party preferences. Electoral Studies, 25(2), 369–​92. Chhibber, P., & Kollman, K. (1998). Party aggregation and the number of parties in India and the United States. The American Political Science Review, 92(2), 329–​42. —​—​—​. (2004). The formation of national party systems: Federalism and party competition in Canada, Great Britain, India and the United States. Princeton, NJ: Princeton University Press. Cosano, P.S. (2011). Understanding the nationalization of party systems in western Europe [Unpublished PhD Dissertation]. Barcelona: Universitat Pompeu Fabra. Cox, G.W. (1999). Electoral rules and electoral coordination. Annual Review of Political Science, 2, 145–​61. Demirkol, Ö. (2015). Party splits in Turkish party system:  The case of centre-​right parties. Turkish Studies, 16(1), 97–​114. Gökmen, Y.C., & Tosun, T. (2015). The CHP in the June and November 2015 elections: An evaluation on political impasse. Insight Turkey, 17(4), 193–​208. Golosov, G.V. (2014). The idiosyncratic dynamics of party system nationalization in Russia. Post-​Soviet Affairs, 31(5), 397–​419. —​—​—​. (2016) Factors of party system nationalization. International Political Science Review, 37(2), 246–​60. Harbers, I. (2008). Decentralization as a condition of party system nationalization: Evidence from Latin America and Central and Eastern Europe [Paper presentation, April 11–​16, 2008]. 2008 ECPR Joint Sessions of Workshops, Rennes. Hicken, A., Kollman, K., & Simmons, J.W. (2008). Party System Nationalization and the provision of public health services [Paper presentation, April 3–​6, 2008]. Annual Meeting of Midwest Political Science Association, Chicago, IL. Jones, M.P. (2005). The role of parties and party systems in the policy making process [Paper presentation, February 28–​March 2,  2005]. Inter-​American Development Bank Workshop on State Reform, Public Policies, and Policy-​Making Processes, Washington, DC. Jones, M.P, & Mainwaring S. (2003) The nationalization of parties and party systems: An empirical measure and application to the Americas [Working Paper]. The Helen Kellogg Institute for International Studies. Jurado, I., & Leon, S. (2017). Economics crises and the nationalization of politics. European Journal of Political Research, 56(4), 777–​800. Kaynar, M.K. (Ed.). (2007). Cumhuriyet Dönemi Siyasi Partileri, 1923–​2006. İmge Kitapevi. Kollman, K., Hicken, A., Caramani, D., Backer, D., & Lublin, D. (2019). Constituency-​ level elections archive. Center for Political Studies, University of Michigan. www. electiondataarchive.org/​index.html Kumbaracıbaşı, A.C. (2014). Cumhuriyet halk partisi oylarının birikimli düzey analizi ve Türkiye’ de merkez solun seçim coğrafyasının İncelemesi. Amme İdaresi Dergisi, 47(4),  59–​92. Laakso, M., & Taagepera, R. (1979). The “effective” number of parties: A measure with application to west Europe. Comparative Political Studies, 12(1),  3–​27. Lago-​Peñas, I., & Lago-​Peñas, S. (2009). Decentralization and the nationalization of party systems [Paper presentation, November 5–​6, 2009]. Conference on “The Political and Economic Consequences of Decentralization, Galicia, Spain. Lago-​Peñas, S., Sacchi, A., & Simón-​Cosano, P. (2014). Who honors the rules of federalism? Party system nationalization and fiscal performance. In Working Papers.

44  Özhan Demirkol and Edip Asaf Bekaroğlu Collection A: Public Economics, Governance and Decentralization 1409, Universidade de Vigo, GEN –​Governance and Economics Research Network. Lago, I., & Montero, J.R. (2014). Defining and measuring party system nationalization. European Political Science Review, 6(02), 191–​211. Lavery, G.D. (2014). African political party development and the limits of institutional engineering [PhD Dissertation]. Michigan State University. d.lib.msu.edu/​ etd/​2576 Morgenstern, S., & Swindle, S.M. (2005). Are politics local? An analysis of voting patterns in 23 democracies. Comparative Political Studies, 38(2), 143–​70. Morgenstern, S., Swindle, S.M., & Castagnola, A. (2009). Party nationalization and institutions. The Journal of Politics, 71(4), 1322–​41. Özbudun, E. (2006). Changes and continuities in the Turkish party system. Representation, 42(2), 129–​37. Özbudun, E., & Tachau, F. (1975). Social change and electoral behavior in Turkey: Toward a critical realignment? Middle East, 6(4), 460–​80. Özen, I.C., & Kalkan, K.O. (2017). Spatial analysis of contemporary Turkish elections: A comprehensive approach. Turkish Studies, 18(2), 358–​77. Riera, P. (2018). The impact of intraparty electoral reforms on party system nationalization. Publius: The Journal of Federalism, 48(2), 191–​216. Rose, R., & Urwin, D.W. (1975). Regional differentiation and political utility in western nations. London: Sage Publications. Sayarı, S. (2007). Towards a new Turkish party system? Turkish Studies, 8(2), 197–​210. Schakel, A.H. (2013). Nationalisation of multilevel party systems: A conceptual and empirical analysis. European Journal of Political Research, 52(2), 212–​36. Secor, A.J. (2001). Ideologies in crisis:  Political cleavages and electoral politics in Turkey in the 1990s. Political Geography, 20(5), 539–​60. Şekercioğlu, E., & Arıkan, G. (2008). Trends in party system indicators for the July 2007 Turkish elections. Turkish Studies, 9(2), 213–​31. Simmons, J., & Hicken, A. (2011). Dividing the spoils:  Party nationalization, credibility, and foreign direct investment [Paper presentation]. Annual Meeting of the Midwest Political Science Association 67th Annual National Conference, The Palmer House Hilton, Chicago, IL. Tachau, F. (2000) Turkish political parties and elections: Half a century of multiparty democracy, Turkish Studies, 1(1), pp. 128–​48. Tezcür, G.M. (2012). Trends and characteristics of the Turkish party system in light of the 2011 elections. Turkish Studies, 13(2), 117–​34. Tiemann, G. (2012). The nationalization of political parties and party systems in post-​ communist Eastern Europe. Communist and Post-​ Communist Studies, 45(1–​2),  77–​89. TÜİK. (2008). Milletvekili Genel Seçimleri 1923–​2007. Ankara:  Türkiye İstatistik Kurumu Matbaası. Vasselai, F. (2009). Comparing the nationalization of party systems in 43 democracies [Paper presentation]. 33 Encuentro Anual de Anpocs, San Pablo. Wahman, M. (2015). Nationalized Incumbents and Regional Challengers: Opposition-​ and Incumbent-​ Party Nationalization in Africa. ETH Zurich. www.isn.ethz. ch/​D igital-​L ibrary/ ​ P ublications/ ​ D etail/ ​ ? ots591=0c54e3b3-​1 e9c-​b e1e-​2 c24a6a8c7060233&lng=en&id=190033

Measuring party nationalization in Turkey  45 West, W.J. (2005). Regional cleavages in Turkish politics: An electoral geography of the 1999 and 2002 national elections. Political Geography, 24(4), 499–​523. Wuthrich, F.M. (2013). An essential center–​ periphery electoral cleavage and the Turkish party system. International Journal of Middle East Studies, 45(04), 751–​73. Yüksek Seçim Kurulu, YSK. (2020). Milletvekili Genel Seçim Arşivi. www.ysk.gov.tr/​ tr/​milletvekili-​genel-​secim-​arsivi/​2644

3  Gerrymandering in Turkish elections since 1950 (Re)setting the rules of the game? Gülsen Kaya Osmanbaşoğlu

Introduction Apart from two minor steps to establish the multiple party system in the early years of the Turkish Republic, which, in any case, ceased within a short time, the real political competition began with the 1950 elections. Demokrat Parti, (the Democrat Party, the DP) was founded in 1945 and gained victory in the 1950 elections. Even though the RPP has dealt with some gerrymandering attempts before the multiparty period, following the 1954 general election the governing DP drastically began to redraw election districts to gain political advantage. The manipulations included downgrading the status of provinces to towns for political retribution and upgrading the status of towns to provinces as political reward. Additionally, the DP divided some larger provinces into smaller ones, creating new electoral populations in a way that was beneficial to the DP. After the 1960 coup, the general election system was transformed from a majority vote system to a proportional representation system, decreasing the potential for trouble-​free abuse of political districts, though not completely eradicating gerrymandering. Nevertheless, gerrymanderers diversified their strategies, a practice that has been ongoing since the 1980s, with a special focus on local elections as well as on metropolitan areas. In general, it is possible to argue that center-​right parties engaged in gerrymandering activities in line with their clientelist policymaking style. Following a brief overview to illustrate the concepts of gerrymandering and clientelism, I will focus on various types of gerrymandering in Turkey and their effects on electoral results.

The concept and connotations Historically, gerrymandering, to a large extent, flourished in the US, which is the very basis of the term itself; it is derived from Massachusetts’s Governor Elbridge Gerry’s name for his intentional distriction. Indeed, the word is actually a combination of his name and the word “salamander”, as his reshaping of electoral districts in favor of the Democratic-​Republican Party had the appearance of such; the term was first used in the Boston Gazette on March

Gerrymandering in Turkish elections  47 26, 1912 (Bickerstaff, 2020, p.  10). Gerrymandering refers to the deliberate manipulation of electoral districts and borders for political gain on the basis of division, concentration, malapportionment, and such calculations based on demographic presumptions. Gerrymandering turns the one person, one vote principle into “one person, one vote, one mess”, creating a kind of electoral bias in which a small percentage of a particular voting population can significantly affect the election results (Niemi & Deegan Jr., 1978; Grofman, 1983). Yet, it remains difficult to unequivocally determine when gerrymandering has occurred. Current election laws and judicial standards are not sufficiently specific or uniform to detect gerrymandering in most cases. As Browning and King note, gerrymandering became a justiciable issue in 1986 in the US, as per the Bandemer case in the literature (1987, p. 305). A decision issued in 2004 by the US Supreme Court attempted to curtail the partisan gerrymandering activities in the US that began with the 2010 redistricting process (McGann et al., 2016). Although gerrymandering is not defended by the courts in general, well-​defined standards for detecting gerrymandering in the US remain undefined. Gerrymandering is seen a barrier to equal representation of the citizenry with the potential to damage the popular representation aspect of a democratic system. Even in consolidated democracies, gerrymandering continues to exist as a method used to manipulate election results by a governing party or opposition, as Balinski (2008, p. 98) observes: “Gerrymandering is widespread and decidedly ecumenical:  both parties indulge”. Sometimes the motives behind gerrymandering can be based on the mode of economic production, as Malesky (2009) observed for the Venezuelan subnational division case. On the other hand, in the Slovakian case, nationalism that stemmed from ethnic and religious identity shaped the gerrymandering activities rather than using districting as a reflection of functional economic means (Halás & Klapka, 2017, p. 1572). Here, one should also remember that the Congress Party of India also utilized gerrymandering as a unifying strategy to maintain its majority power in religiously and linguistically divided India (de Mesquita, 1978). Researchers have observed that the effectiveness of gerrymandering varies according to the type of electoral system; however, gerrymandering is more effective in a simple majority system. Gerrymandering can be accomplished via concentration of votes, allocation of opposition votes, and incumbent displacement, such that opposition voters are placed in as few districts as possible. Allocation gerrymandering is characterized by the redistribution of opposition votes in order to limit their representation and hinder their ability to affect election results. Incumbent displacement gerrymandering is characterized by “eliminating the seats held by the members of the opposing party and combining the homes of opposing party incumbents into a single district” (Owen & Grofman, 1988, p. 6). So, gerrymandering is far from being a monolithic and standardized strategy to gain benefit, but its connotations may depend on the context and legislation.

48  Gülsen Kaya Osmanbaşoğlu The effects of gerrymandering vary as well. Some researchers think that gerrymandering can directly shape election results. For instance, Owen and Grofman stated that, By a combination of the dispersal and concentration gerrymandering techniques a party which expects to have only a minority share of the popular vote (greater than 25  percent), but which is somehow in control of the districting process, can still win a majority of seats by simply allowing its opponent overwhelming control of some districts in which much or all of the opponent’s strength will be wasted. (1988, p. 6) This suggests that it is possible to transform a party that is expected to lose into a winning party. Nowadays, some researchers no longer consider gerrymandering to be a risky undertaking; rather, it is viewed as a mathematical strategy with high possibility of obtaining the expected results. Balinski (2008) reported that computer-​based statistical analysis has fundamentally changed gerrymandering from an art into a science. Starting from this point of view, Altman and McDonald also underline the complexity of the issue, stating that computer programs might bring more equal representation unless they are abused (2010). So, scholars have developed different points of view as to the extent that computerized systems leave room for attempts at gerrymandering. On the other hand, the effects of gerrymandering are not consistent, and it does not always return the expected result. McCarty et al. (2009, p. 667) stated that, “manipulation of districting exacerbates polarization”; as such, it is not always possible to predict voter behavior. Carson et al. (2007) observed that in the US, voting is more polarized—​according to party affiliation—​in newly created electoral districts as compared to well-​established districts. In this regard, Bickerstaff also adds the French case, which indicates an unfruitful ecosystem for a gerrymander (2020, p. 214). Voter volatility makes accurate measurement of voter behavior difficult before an election, so gerrymandering, through redistricting, does not always work accordingly (Abramowitz et al., 2006). Theriault (2008) also thinks that redistricting is a factor that intensifies polarization, as do social and economic structures, legislative issues, and leadership strategies. As gerrymandering can have partisan effects (Bartels, 2000), McCarty et al. (2009, p. 668) suggest that, “aggressive gerrymandering makes majority party seats less safe”, which further indicates that gerrymandering does not always result in the expected outcome for majority power holders. The above-​referenced studies clearly show that multiple factors play crucial roles regarding the possible outcomes of gerrymandering. Although some researchers suggest that the use of gerrymandering should be eliminated or at least minimized, others continue to investigate how to optimize the process. Redistricting to minimize the effect of partisan voters and accentuate the effect of median voters is suggested for optimal

Gerrymandering in Turkish elections  49 gerrymandering (Friedman & Holden, 2008). Coate and Knight (2007, p. 1) introduced the concept of the “socially optimal” gerrymander, which creates more responsive results in terms of “how the composition of the legislature changes in response to changes in citizens’ voting behavior”. Owen coined the term “sophisticated optimal gerrymander” to describe gerrymandering that maximizes the number of seats for a given party or the potential for majority control of a decade or more,1 and ensures large voter turnout. In a nutshell, the findings reported in the gerrymandering literature concerning its effectiveness and outcomes are inconsistent. Whereas recent research has indicated that socially optimal gerrymandering can be used with greater predictability, gerrymandering still remains a largely unpredictable endeavor. This literature can be referenced in an effort to understand the use of gerrymandering in Turkey. Apart from its potential to provide an advantage to gerrymanderers, another aspect of the issue can be related to political bargaining. With regard to the Turkish case, I presented a paper in MESA 2016 that is the very basis of this chapter, namely that of dealing with the issue within the framework of clientelism. Notably, Tunç and colleagues published a book on gerrymandering practices in Turkey with particular reference to redistricting through the introduction of the 2013 metropolitan law, Law No:  6360 (2014)2. Bickerstaff also considered the Turkish gerrymandering experience on the basis of pro-​Kurdish regions and beyond (2020, p. 209). Çarkoğlu and Aksen published a paper on partisan bias and malapportionment, noting how the system allows for the overrepresentation of the dominant parties (2019). Nevertheless, the vast use of gerrymandering in Turkey is not proportionally reflected in the associated literature, however, remaining to date an undervalued academic issue. Since the literature on Turkey’s gerrymandering practices is still far from being either satisfactory or broad, this chapter attempts to add a humble perspective to help fill the existing gap.

The Turkish case Upgrading and downgrading the status of towns and provinces, manipulation of voter eligibility via redistricting, as well as retributive penalization or clientelism/​patronage based on election results are all characteristic of gerrymandering in Turkey. In the very first elections that opposition parties could enter in the Republican Period, the first indications of gerrymandering in Turkish politics began to appear in the form of changing districts in the post-​election period. Serbest Cumhuriyet Fırkası (the Free Republican Party, the FRP) gained the 1930 mayoral elections in 31 districts of which only two were provinces, namely Aydın and Silifke.3 After the 1930 elections, Silifke’s support for the Free Republican Party was penalized and the Cumhuriyet Halk Partisi (the Republican Peoples Party, the RPP) degraded the city to a town by moving the city center to İçel and thus, Mersin (İçel) became a new province replaced by Silifke (Turan, 2008, p. 97). Granted, the 1930 elections were not

50  Gülsen Kaya Osmanbaşoğlu an example of free and fair elections as the governing party, enjoying its many privileges, could not even tolerate the FRP’s gain in Silifke district.4 This short period was ended by the closure of the FRP and the RPP maintained its rule of the country in a single party regime until 1945. Establishing a multiparty system was on the mid-​1940s political agenda in Turkey, which was defined as an endeavor for “the establishment of political freedom” by Karpat (2016 [1959], p. x). The Democrat Party (the DP), as one of the first5 opposition parties of the multiparty system, emphasized free and fair elections because of the doubts about the fairness of the electoral system6. Here, the political alignment of the DP should not be exaggerated in terms of regional representation since the party stemmed from the RPP and, to some extent, represented the center-​right in comparison to the slightly center-​left positioning of the RPP. Karpat evaluates this difference on the basis of the power-​holding status of the RPP with romantic historical conditions and the deprivation of power for the DP for the “well-​defined interests” (2016 [1959], p. 417). However, the confrontation between the governing parties and opposition, to a large extent, went beyond the ideological dispute (Ahmad, 1977, p. 36), and intolerance towards the opposition, especially in the last years of the DP rule, almost became a norm of Turkish politics. At the beginning of the multiparty era, the governing RPP, as the state party, made use of broad government privileges during the general elections of 1946, which was harshly criticized by the DP. In an effort to limit the advance of the newly emerged opposition party (Bayar, 1969, p.  56), just before the elections, on June 5th, the Election Law was changed and the date of the election was declared just 18  days before voting. Thus, single-​stage elections with an open vote and secret counting principles were carried out without any judicial oversight, while election councils, composed of people close to the RPP, were authorized to announce the results (Albayrak, 2004, pp.  84-​2). Thus, the ruling party’s unfair superiority in the election process led, in turn, to widespread suspicions about the fairness of the elections. The DP’s insistence on the establishment of a fairer and freer election system fervently continued, and the Grand National Assembly approved a new law on February 16, 1950, Law Number 5545, introducing a new election system that relied on a single, equal, and secret ballot majority system, as well as on the principle of open classification. As the victim of an unfairly managed election system, the DP initially tended to advocate for the significance of elections to ensure a well-​operating democratic system, which was not maintained over time. Given these circumstances, it was expected that the DP would respect the election results in all cases. However, after the DP entrenched its power in the 1954 elections, the party opened the door to altering the status of the districts.7 This would commonly be utilized in Turkey to achieve political advantage, sometimes through clientelism. Here, it is necessary to note that clientelism, as a political cultural element, also flourished within particular institutional and structural values embedded in the Turkish political climate (Wuthrich, 2015, p. 89).

Gerrymandering in Turkish elections  51 Wuthrich notes the elite domination of politics distanced from society and the limitations of the ideological politics, where such structural factors gave rise to a kind of client–​patron relationship from the beginning of the multiparty regime (2015, p. 78). Heper and Keyman also state that patronage in Turkish politics, as carried out by the DP, had its basis in the duality of the state’s interest vis-​à-​vis the nation’s will (1998, p.  261). The state’s interest, which was represented by the bureaucrats and the RPP, now began to be challenged by the DP with the excuse to protect the people’s interests (Sayarı, 1977, pp. 105-​2). So, the DP, with the claim of representing the masses while actually representing the elite in many ways, toyed with clientelism in the subsequent government periods. Thanks to the simple majority system, whereby the majority in a province wins all its seats, the DP garnered 58.4 percent of the votes in the elections, corresponding to an outright majority in the Turkish parliament of 503 seats out of 542 in the 1954 elections (TÜİK, 2012, p.  12). Except for just five districts, the DP won the elections everywhere.8 The major opposition party, the RPP, won the elections in four of these districts, obtaining most of its seats from Malatya. Immediately after the elections, the DP began its gerrymandering activities by dividing Kocaeli and Malatya into two cities. It created two new provinces called Adapazarı (Sakarya) and Adıyaman, which helped to create new DP districts as well as to distribute the opposition party’s power. Here, Malatya’s division into two cities involved more partisan tenets than Kocaeli. Electing 12 deputies to the parliament where the RPP already had 31 seats in total, Malatya was the strongest district of the RPP, and this was where the RPP’s leader, İnönü, came from. Thus, on the one hand, the DP punished İnönü’s hometown and also decreased the number of seats in Malatya from 12 to 9, where the RPP would probably win in the next elections; on the other, the DP promoted Adıyaman by raising its status to that of a province. When we examine the election results for Adıyaman, we see popular support for the DP in the first general election in 1957, 1.2 percent above the DP’s national share (TÜİK, 2012, pp. 22-​2). For the DP, it is possible to argue that the voters of Adıyaman in a way indicated their gratitude to the governing party, whereas in the following elections of 1961, despite the relative success of the Justice Party (Adalet Partisi, the JP) at the national level, the DP’s successor, the JP, with 2.8 percent of the vote, could not even get a single deputy in Adıyaman, but the New Turkey Party (Yeni Türkiye Partisi, the NTP), another successor of the DP, and the RPP obtained two deputies each.9 Ironically, in the 1961 Turkish general elections, the RPP garnered 28.8  percent of the vote and was the winning party. Taking second place, the JP gained 27.3  percent of the vote, whilst Cumhuriyetçi Köylü Millet Partisi (the Republican Peasant Nation Party. the RPNP), an amalgamation of the Republican Nation Party and Republican Peasant Party, took third place, achieving 11 percent of the vote. Finally, the New Turkey Party gained 10.8 percent and was fourth. But, as stated above, Adıyaman showed a

52  Gülsen Kaya Osmanbaşoğlu The RPP (Malatya/Turkey Comparison)

90% 80%

78%

70%

67%

60% 50% 40% 30%

54% 35%

41%

36%

20% 10% 0%

1954

1957

1961

Figure 3.1 The RPP votes in Turkey and Malatya Source:  The graph was created by the author based upon the data from the YSK (Supreme Electoral Council) results.

preference to vote for the NTP instead of the AP of the DP’s two successors. So, despite the JP’s relative success in the country as a whole, the party was unable to collect many votes in Adıyaman. However, Adıyaman’s support for the DP’s successors continued until the 1969 elections. Furthermore, the DP’s punishment did not lead to a decrease of the RPP votes in Malatya. Instead, motivation to vote for the RPP rose in Malatya in the following elections from 54 percent to 78.1 percent (TÜİK, 2012). As Figure 3.1 illustrates, in the 1957 elections, following Adıyaman’s separation from Malatya, the gap between the RPP’s nationwide votes and Malatya votes grew, and the RPP gained a 37 percent greater vote in Malatya than its national score. With regard to Kocaeli, the province was divided into two and Sakarya was made a new province and election district. Dividing Kocaeli did not generate much opposition from the Kocaeli side, however, the total number of seats in Kocaeli and Sakarya, increased from 12 to 13. Kocaeli voted for the DP in the 1950 elections with 59.2 percent of the vote. This reached 63.2 percent in the 1954 elections. After the separation of Sakarya from Kocaeli, similar to the national trend, the DP votes decreased in Kocaeli to 50.7 percent in the 1957 elections, which was still higher than the DP’s national average of 47.8 percent. Also, the DP continued to be the winning party in Kocaeli. However, making Sakarya a new city also helped to increase the DP votes to 58.9 percent in 1957 in this newly emerging city, which was drastically greater than the DP’s national share. Sakarya also continued to be an ardent supporter of the DP and its successors in the following elections.

Gerrymandering in Turkish elections  53 Nevertheless, degrading Kırşehir’s status to a town and linking it to Nevşehir as a newly created province while allocating its towns to other neighboring cities was a more partisan gerrymandering strategy, which resulted in a “revenge vote” from the city for the Republican Nationalist Party. In the May 2, 1954 election, under the leadership of Osman Bölükbaşı, the Republican Nation Party obtained five deputies in Kırşehir while losing elections in all other districts. Furthermore, the leader of the RNP, Osman Bölükbaşı, held a disproportionate share of power in Parliament, which meant that in spite of representing his party with a couple of deputies, his oratory speeches in some cases became more efficient and challenging than those of the main opposition party, the RPP. As a person who represented the periphery, which the DP also claimed to do, Bölükbaşı thus became a strong rival to the DP. Indeed, the DP’s position towards the opposition can be described, in Ahmad’s words: “Historically, opposition of any sort had been equated with hostility, in the minds of both rulers and ruled” (1977, p. 38). Therefore, the DP government wanted to eliminate its opposition by keeping it from parliament in future elections. As a means of gerrymandering, his election district, Kırşehir, was downgraded to a town and linked to Nevşehir, a town which was upgraded to a province to effectively eliminate Kırşehir as an election district. Beyond any shadow of doubt, the gerrymanders did find excuses to justify their intentional distriction. In Assembly meetings, Minister of Interior Namık Gedik claimed that making Nevşehir a province was an absolute necessity because its central population was greater than Kırşehir’s (TBMM Tutanak Dergisi, 1954, p.  355). Likewise, some DP members advocated Kırşehir’s downgrade to a town under the pretext of the so-​called reactionary character of the city.10 In Assembly meetings, the Kırşehir issue was discussed extensively, and not only the Republican Nation Party but also the RPP reacted against Kırşehir’s redesignation as a town. In these meetings, Sırrı Atalay of the RPP stated that the DP punished not only Kırşehir, but also all opposition parties, and that the insight and free will of the citizens was being obstructed (TBMM Tutanak Dergisi, 1954, p. 344). Even some DP deputies, such as İsmail Hakkı Akyüz, Mustafa Ekinci, and Zihni Ural (at that time, these names formed Hürriyet Partisi, the Liberty Party) also criticized their own party’s decision in this regard.11 Despite these internal and external criticisms, on the 20th of June, with the parliamentary majority of the DP, the law was passed. Given that the DP won 74 out of 76 villages in Kırşehir in the village headman (muhtar) elections, some of the DP members described this incident as a victory for the DP in Kırşehir (Burçak, 1998, p. 231), but it was not a longstanding and entrenched success. The possible instigator of this shift was the lack of election contests, since the 1955 local elections were protested by the main opposition parties, namely the RPP and the RNP.12 In the local 1955 elections, Kırşehir was among one of ten cities where the DP was defeated by an independent or Peasant Party candidate who won in the municipal council elections (Ahmad & Ahmad, 1976, p. 143; Turan, 2008, p. 97).13 In

54  Gülsen Kaya Osmanbaşoğlu

Kırşehir General Election Results 63

43.5 36 33

30

25

25

23

11

1950

1954 DP

CMP/MP

1957 CHP

Figure 3.2 Kırşehir general election results Source:  The graph was created by the author based upon the data from the YSK (Supreme Election Council) results.

the following years, Kırşehir’s outlier position was noted and the JP, as the successor of the DP, found a pragmatic formula to maintain or better their political position on the basis of compromise with the RPP and RPNP (an alliance of the PP and RNP) for the 1963 mayoral elections and supported an independent candidate called Refik Soykut (Turan, 2008, p. 108; Akbulut, 2001, p. 56). So, Kırşehir’s case, as seen in Figure 3.2 suggests that the negative impact of gerrymandering towards the gerrymanders can last for a number of subsequent elections and might result in collaboration among the opposition parties to break an existing alignment. The ruling DP then tried to find another formula while replacing the gerrymandering of electoral districts with the gerrymandering of incumbents, with a particular focus on the opposition of Bölükbaşı. So, Kırşehir became a province again before the 1957 elections, but Osman Bölükbaşı lost his privilege of immunity, which led to his subsequent arrest (Eroğul, 2003, p. 198; Öymen, 2013, pp. 72-​2). This might suggest a tragic version of an incumbent-​ displacement form of gerrymandering. Nevertheless, when we look at the following general elections, we see that support for the DP could not be generated again. Despite the DP initially having 36 percent of the city’s vote in 1950, that support gradually decreased. In contrast, the support for the Republican Nation Party and it successors, as the victims of gerrymandering, was considerably increased. As witnessed in the case of Kırşehir, those who did not support the DP received certain punishments along with deprivation of public services, while

Gerrymandering in Turkish elections  55 DP supporters were rewarded. Beginning with the DP governments and as carried on by its successors, center-​right parties in Turkey developed clientelist networks with high levels of party patronage.14 As Sunar notes, The clientelist incorporation of the rural population, the patronage-​ induced private initiative, and the great but haphazard social dynamism fueled by populism—​all of these have not only outlived the DP, but have become permanent features of center-​right politics, dominant in Turkey since 1950. (1990, p. 752) So, gerrymandering in the Turkish context began to go hand in hand with political patronage by the DP, which was seen as normal to pre-​election campaign promotions, such as through distribution of “funds for schools and mosques” (Ahmad, 1977, p.  57) or “by paying higher price for produce” (Ahmad, 1977, p. 56). Even without an assessment of these actions from an ethical perspective, the associated pragmatism to a large extent dominates the political climate throughout center-​right governments. However, the issue is not limited to center-​right governments per se. After the 1960 military intervention, the 1961 Constitution was introduced and a proportional representation system was endorsed, which required more comprehensive strategies to get concrete results from such gerrymandering manipulations. However, Wuthrich notes the fragmentation effect of the coup in that DP votes were divided among its various successors (2015, p. 97). So, it is also possible to argue that the closure of the DP, which governed the country for three terms, led to a kind of gerrymandering in a different type of incumbent replacement. Closure of political parties, which was maintained by the 1980 coup and targeted the pro-​Kurdish and pro-​Islamist parties much more, may also have an indirect gerrymandering effect, too. On the other hand, under the political instability of coalition governments as well as the strict supervision of the political sphere by the civilian and military bureaucracy, gerrymandering seemed to decline in the 1960s and 1970s, becoming far more blatant again in the 1980s at the hands of the Anavatan Partisi (the Motherland Party, the MP) through the creation of new metropolitan municipalities and the repeated adjusting of election laws. Nevertheless, it should be noted here that the national election system turned to the D’Hondt method, with 10 percent of the national threshold, which complicated the picture further. Since gerrymandering for political gain in general elections could not bring well-​predicted concrete gains, gerrymanderers began to focus more on local elections. With the 1982 Constitution, a legal basis for the establishment of metropolitan municipalities was posited and İstanbul, Ankara, and İzmir became the first three metropolitan municipalities in 1984.15 This was also related to gerrymandering since new electoral districts, such as towns central to the metropoles, were created, and upgrading the level of a city could also

56  Gülsen Kaya Osmanbaşoğlu be read as an election deposit to please the residents. Furthermore, metropolitan municipalities began to enjoy a certain financial and administrative autonomy, which was supported by the minor decentralization steps of the central government (Heper, 1986). Newly established metropolitan municipalities16 entered the 1984 mayoral elections, where the MP trenchantly accumulated votes. Obtaining 50.2 percent of the votes in metropoles while getting 43.2  percent nationwide, the MP’s voting base was considerably stronger in the metropoles in the 1984 municipal elections. This also paved the way for a service-​oriented Özal government collaboration with the metropolitan municipalities (Kalaycıoğlu, 1994), which offered a more advantageous position to these gerrymandered places. It was speculated that transforming a bigger city into a metropole would increase the MP’s votes in that city in such a clientelist policymaking strategy.17 The number of the metropolitan municipalities was increased from three to eight by upgrading the status of Adana, Bursa, Gaziantep, Kayseri, and Konya to that of metropoles on June 20, 1987. Ironically, gerrymanders faced considerable frustration in the 1989 local elections, since the MP was unable to gain any of these metropoles.18 Furthermore, its voting share decreased to 23.6 percent in metropoles, which was slightly lower than its nationwide total of 23.7 percent. Even the MP’s share of the vote in Sincan, which was bound to the capital city Ankara as a promotional reward just before the elections, dramatically decreased from 41.4 percent to 17.6 percent. One should note here that there were other significant factors contributing to the decrease in the MP’s share of the vote in the 1989 mayoral elections:  Sosyal Demokrat Halkçı Parti (the Social Democratic Populist Party, the SDPP) began to communicate with the masses through the social problems in an organized way; Doğruyol Partisi (the True Path Party, the TPP) leader Süleyman Demirel, who was banned from politics after the 1980 military coup, turned back to the political race; and Islamic politics began to gain ground, especially in the metropoles.19 In all, the 1989 local elections frankly suggested that gerrymandering by itself was not a safe instrument for the manipulation of votes in Turkey. However, the MP government pursued further manipulations and sought new ways to gain partisan interest. On June 21, 1989, by law number 3578, Bayburt, Kırıkkale, Aksaray, and Karaman were turned into provinces as a kind of patronage relationship before the elections. Throughout this year, upgrading a city’s status was clearly related to gerrymandering as well:  first of all, it was a way of pleasing residents living there by providing direct services in the city centers that they were previously deprived access to while classified as towns. Further, a new city meant new election districts, which would probably receive higher representation.20 On the anniversary of Bayburt’s provincial celebrations,21 Özal aggrievedly declared that the MP could win only one of the four newly established cities. While expressing his disappointment, Özal stated that they transformed the status of approximately 120 kasabas (small towns) to that of

Gerrymandering in Turkish elections  57 towns, which indeed seemed intended to transfer public services more easily rather than to gather votes. Despite his service-​oriented discourse, one may question why these cases were chosen for an upgrade to their official status. On the eve of the elections, his speeches indicated a certain expectation of a flood of votes, to the benefit of his party. Even through Özal was not satisfied with the electoral results of his party’s gerrymandering, when we consider the picture in detail, gerrymandering in some cases worked somewhat well. For example, in the 1989 local elections, while the MP gained Bartın, a town in Zonguldak, Zonguldak as a whole voted for the SDPP. On September 7, 1991, immediately before the national elections, Bartın was separated from Zonguldak under the prime ministry of Mesut Yılmaz, the head of the MP, when Turgut Özal, the leader of the MP, became the president. Somewhat similar to the Malatya case, Zonguldak was center-​left opposition leader Bülent Ecevit’s election district. When we consider the 1987 elections, the MP won four, vis-​à-​vis five seats of the TPP in Zonguldak. In the 1991 elections, however, the TPP won six seats while the RPP gained only one, and the MP were unable to gain any. Bartın, the city of Köksal Toptan,22 voted 37 percent for the TPP, which corresponded to one deputy. On the other hand, the DLP also obtained one seat with 25  percent, but the MP, with 21.9  percent of the votes, could not get any seats from that newly established province. In line with the MP’s diminishing power all around the country, gerrymandering did not work in the case of Bartın in general elections. Nevertheless, in the 1994 mayoral elections the MP won both in Bartın and Zonguldak, but this victory streak was broken by the subsequent 1999 local elections, with the DLP defeating the MP in both cities. In 1980, the number of towns in Turkey was 572, a figure which increased to 829 within a decade with the MP’s clientelist gerrymandering policies. Throughout the MP era, the party not only tried to motivate voters with promises to upgrade their districts’ status, but also threatened electorates, saying that if they voted for other parties, the MP would not bring effective public services. In the same fashion, in 1980 the number of provinces was 67, but rose to 81 within two decades. These were seen as instruments to gain partisan advantage, but in many cases they did not bring long-​standing expected results. Even if the parties change, the patronage networks remain the same for the Turkish center-​right. The rival, and at the same time more or less ideological substitute of the MP, the TPP government, under Süleyman Demirel’s leadership, upgraded the status of the town of Ardahan and Iğdır to province via Law Number 3806 in 1992, which helped the TPP win both in the 1994 local and 1995 general elections in Ardahan. However, while wining Iğdır in the 1994 local elections, the TPP could only take second place in the 1995 general elections held in the city. Without being able to provide long-​term benefits, in the 1999 general elections the TPP was defeated in both cities and only gained Ardahan in the 1999 mayoral elections.

58  Gülsen Kaya Osmanbaşoğlu Despite the long-​term ineffectiveness of such attempts at gerrymandering through creating new election districts, the promises to make bigger towns into provinces continued in the mid-​1990s with Tansu Çiller, as the successor to Süleyman Demirel of the TPP. However, her policymaking style to a large extent resembled MP leader Turgut Özal’s. Before the mayoral elections of 1994, Tansu Çiller promised Yalova that it would become the 77th city of Turkey if it elected the TPP candidate, İbrahim Uzun. As a possible outcome of that promise, the TPP won the Yalova municipality with 43.9 percent of the vote, fulfilling the stipulation to become a province. In the following general elections after becoming a province, however, the MP defeated the TPP in Yalova.23 The TPP was now only the second party in the city, with 23 percent of the votes. So, regarding the Yalova case, despite gerrymandering working for the municipal elections, the voters’ behavior immediately changed and it ultimately did not give any long-​lasting results in terms of the national elections. In 1993, under Tansu Çiller’s prime ministry of the TPP and the SDPP coalition government, law number 504—​a decree law—​formed new metropoles such as Mersin, Eskişehir, Antalya, Diyarbakır, Samsun, İzmit (Kocaeli), and Erzurum. In the 1994 mayoral elections, the TPP gained Antalya and Eskişehir whilst the SDPP won İzmit. Subsequently, the TPP only won elections in Antalya while the SDPP could just pass the national threshold along with the RPP coalition in the 1995 national elections. Nevertheless, it is hard to consider the TPP’s gain in Antalya as a victory since its votes dramatically decreased from 43.2 percent to 26.5 percent between 1991 and 1995. Very fragmented results in the 1995 general elections were observed all around the country. This highly polarized and volatile political context (Özbudun, 2000) demonstrated that gerrymandering in itself did not lead to effective results. On October 24, 1996, before going to mid-​elections following the death of its mayor,24 under the Refahyol government, which consisted of the TPP and Refah Partisi (the Welfare Party –​the WP), Osmaniye was upgraded to a province.25 The WP gained Osmaniye municipality in the 1996 mid-​elections, but this success did not in any case succeed because Milliyetçi Hareket Partisi (the Nationalist Action Party, the NAP) won the general elections in 1999 as well as the very next mayoral elections in 1999 in Osmaniye.26 Given Osmaniye is the hometown of Devlet Bahçeli, the leader of the NAP, this case suggests that the scope of gerrymandering does not always extend beyond the boundaries of leaders’ domination in their hometowns. As a reaction to the continuous tailoring of regions, in 2001, with law number 4709, a constitutional amendment was enacted stating that law pertaining to elections could not be applied to an election less than one year from its enactment. This law was designed to reduce last-​minute manipulation of elections. However, other large-​scale gerrymandering took place through law number 6360 in 2013.27 Before the 2014 mayoral elections, by law number 6360, the Justice and Development Party increased the number of metropoles from 18 to 30.28

Gerrymandering in Turkish elections  59

Figure 3.3 Map of Evren and Ҫankaya’s location in Ankara Source: The map was created by the author.

Furthermore, the electorates of the metropolitan districts were redefined by expanding the borders of the metropoles to cover all the towns as well, which were previously limited to the residents of central towns. Since villages and small towns (kasaba) became neighborhoods in the metropolitan borders, villages and kasabas were also linked to metropoles. These alterations also aimed to influence not only the metropolitan mayoral elections but also the municipal council allocations by including the periphery votes of the metropoles. For example, in the case of Ankara as illustrated in Figure 3.3, in terms of municipal councilor elections, Aygül notes that a vote in Evren, a newly formed right-​wing peripheral town of Ankara, is almost equal to 61 votes in the Çankaya29 district, which is well known for its resolute center-​left tendencies. Thus, even if the governing party lost metropolitan mayors, this gerrymandering enabled it to maintain its “weight” in the municipal councils, as was realized in Eskişehir and Hatay in the 2015 elections (Aygül, 2016, p. 182). Before the 2015 local elections, in terms of concentration of votes, opposition voters were curtailed in the Çankaya region through binding some of the neighborhoods of Yenimahalle, such as Ümitköy, Yaşamkent, and Çayyolu to Çankaya. It was posited that in Yenimahalle, the JDP had the potential to defeat the RPP, but in Çankaya this was perceived to be impossible, which is known as the “castle” of the RPP. However, according to the election results, the JDP could not even win in Yenimahalle, which was probably a result of crossover votes; people who usually voted for the NAP began to vote for the RPP in Yenimahalle to defeat the JDP there.30 Similar districting was pursued in Maslak and Ayazağa, which also resulted in similar outcomes. Furthermore, the composition of the electorates of the metropoles changed with this law, so that all voters living within the province henceforth

60  Gülsen Kaya Osmanbaşoğlu became eligible to vote for the metropolitan mayors. To a large extent, this influenced the local election results to the benefit of the governing party. The JDP would have won two of the cities (Mardin and Van) while losing two others (Ankara and Antalya) if the elections were simulated according to the previous metropolitan districts. On the other hand, the RPP would have won Antalya and Ankara. Halkların Demokrasi Partisi (the People’s Democracy Party, the PDemP) and the NAP also benefited from the alteration to the law while the RPP seemed to be the most disadvantaged by it. In terms of numbers, even the JDP received two of the cities while losing two others. However, the strategic importance of Ankara and Antalya brought a considerable advantage for the governing party, which was probably still below the expectations redistricting might otherwise have suggested. Furthermore, the JDP’s advantage on the municipal councils was much more evident. Yet any long-​term effects of this gerrymandering were not apparent in the 2019 local elections, where the JDP lost a large number of metropoles, including Ankara, İstanbul, and Antalya. The governing party also faced frustration in the newly established metropoles of Hatay, Tekirdağ, and Manisa. In line with the previous gerrymandering experiences, it is still possible to argue that this gerrymandering will also not provide enduring benefits for the governing party in the near future.

Conclusion In conclusion, one may argue that in the case of Turkey, gerrymandering is more widespread than it might otherwise have seemed.31 Since no parties admit that they intend to gerrymander, they offer various excuses for their initiations. Regarding the impacts of gerrymandering in the Turkish context, it is possible to argue that the process did not offer any enduring benefits to the governing party. Gerrymandering in the DP era was more fruitful for the gerrymanderers, probably because of the lack of political alternatives and simple majority system; afterwards, it became harder to get concrete results from gerrymandering, which is probably an outcome of the proportional representation system with its various alternative parties, even in the same part of the political spectrum (e.g., the MP and the TPP). Either volatility or the electorate’s strategic protest or crossover voting has hindered the realization of the gerrymanderers’ expectations in many cases. Even when parties initially get the expected results via gerrymandering, the ongoing impact of such to any given party never seems to last very long. When the voters strategically satisfy the expectations of the gerrymanderers for one or two terms due to the clientelist promotions, without a sense of devotion they do not continue to vote in the expected direction. So, an optimal partisan gerrymandering that maintains its effects for more than a decade could not be achieved in most of the tested cases in the Turkish context. In all, with regard to policymaking, gerrymandering still hovers around clientelism and political retribution, which largely makes gerrymandering

Gerrymandering in Turkish elections  61 a center-​right party activity. Since center-​right parties have governed the country more than their center-​left or far right rivals, as well as parties dealing with clientelist politics more extensively, they saw gerrymandering as a feasible strategy to reach the electorate. Utilization of gerrymandering sometimes appears in the form of a bargaining chip to motivate voters to support ruling parties. Sometimes, it becomes a punishment in the pre-​or post-​election periods through dividing or redefining electoral districts. Elimination of particular candidates in the form of incumbent displacement can also be observed. When we look at the process, it is possible to argue that with the domination of the rise of clientelism in Turkish politics, gerrymandering attempts also rose in the same fashion. Here, one should also note that the opposition leaders’ hometowns or electoral districts have a significant place in Turkish gerrymandering practices. Sometimes, these cities are divided to pack votes in newly founded cities as well as to promote the divided parts while upgrading their status. Nevertheless, in many cases, it is also observed that the leaders’ hometowns continued to vote for their own sons, namely the opposition leaders’ parties, which bears the maintenance of nepotistic links in Turkish politics as well. In all, gerrymandering has seen widespread use as a tool for clientelist policymakers, but it does not provide longstanding results in the Turkish context, probably because of voters’ volatility and pragmatism. So, in such a context, not only do gerrymanders play with the electoral geography, but the voters also have the potential to toy with politicians’ machinations in this regard.

Notes 1 The redrawing of district lines in the US after decennial census might lead one to take a decade as an indicator for the term. 2 In this book, many gerrymandering experiences were disregarded and only two cases, namely Kırşehir and Abana, were recorded in addition to the 2013 Metropolitan Law’s redistricting account. 3 For details, see Çolak, H. (2007), pp. 158-​63; Günver, G. (2007). 4 Before the multiparty regime, a couple of provinces were also degraded to towns, such as Genç and Doğubeyazıt, but their motivation to a large extent related to security issues rather than gathering votes for the benefit of the ruling party. 5 Chronologically, Milli Kalkınma Partisi (the National Development Party) was founded before the DP. 6 For details, Articles 9 and 10 of the DP program can be considered. See Demokrat Parti Programı. (1946), p. 3. 7 Even in 1953, the DP carried Abana—​ a town that supported the RPP—​ to Pazaryeri—​a town that supported the DP—​but this decision was cancelled by the Constitutional Court in 1967. For details, see Tunç et al. (2014, p. 125-​2). 8 These cities are Kırşehir (RNP), Malatya, Kars, Sinop, and Tunceli (RPP). 9 In the 1961 elections, as the other successor of the DP, the NTP’s share of the vote was greater than that of the RPP in Adıyaman. See TBMM. (2019). 10 See Burçak, R.S. (1998), p. 224.

62  Gülsen Kaya Osmanbaşoğlu 11 For further details, see TBMM Tutanak Dergisi. (1954). 12 Apart from the DP, another minor party called the Köylü Partisi (the Peasant Party) joined that election. 13 Other cities were Edirne, Maraş (from the PP), Burdur, Çorum, Diyarbakır, Kırşehir, Muğla, Muş, Sivas, and Siirt (independent candidates). 14 For a broad evaluation of center-​right politics in Turkey, see Kaya Osmanbaşoğlu, G. (2014). 15 See Resmi Gazete (2008). Metropolitan Municipalities Act, Law No:3030. www. resmigazete.gov.tr/​eskiler/​2008/​08/​20080819-​7.htm 16 For a brief analysis of the duties of metropolitan municipalities, see Metin Heper 1987. “Introduction” in Democracy and Local Government: İstanbul in the 1980s. Walkington: The Eothen Press, pp. 4–​5. 17 The MP was so sure that the metropolitan municipalities were reserved for the MP that the party tried to make an alteration that central town mayors would be elected by all the residents living in the metropoles in order to win all the central towns, which was not approved by the Constitutional Court (Turan, 2008, p. 217). 18 At the end of the elections, the SDPP won six metropoles, the TPP gained one, and the WP got one while the MP could not get any of them. 19 The Welfare Party’s share of the vote increased from 3.7 percent to 9 percent in metropoles in the 1989 elections (Turan, 2008, p. 22). It should be noted here that the electorate composition of the metropoles also changed during this period because of the new metropoles. 20 According to this election system, smaller cities, including the newly established ones, hold considerable power since one vote in a small city is more valuable in relation to one vote in metropoles due to the allocation of parliamentary seats. 21 In fact, Bayburt’s status of province became an instrument for political gain to such an extent that even Menderes, before the May 2, 1954 elections during his Erzurum visit, promised to make this small town a province (TBMM Tutanak Dergisi, 1954, p. 350). 22 A politician from the TPP who also became the head of the party for a short while. 23 For details, see:  www.tbmm.gov.tr/​develop/​owa/​secim_​sorgu.secim_​cevresi_​ partiler?p_​secim_​yili=1995&p_​il_​kodu=77. 24 Musa Şahin, the mayor of Osmaniye between 1994–​1996 was from the Nationalist Action Party (the NAP). 25 Osmaniye is also hometown of the NAP leader Devlet Bahçeli who became the head of the party in 1997. For details, see the Nationalist Action Party. (2016). 26 For details, visit YSK (2019); Osmaniye Municipality (2019). 27 For details, see Resmi Gazete (2012). 28 In fact, the Justice and Development Party initiated to make a law (Law No:5019) to expand the borders of the metropolitan municipalities, which was not approved by the President Ahmet Necdet Sezer in 2003. 29 Aygül delineates that “14 votes cast for the JDP for each 1000 votes cast for the RPP” (2016, p.186). 30 When we look at the results of the mayoral elections in Yenimahalle, the RPP garnered 50.8 percent, the JDP 37.9 percent and the NAP 7.8 percent, whereas in the municipal council elections, the JDP gained 36 percent, the RPP 34 percent, and the NAP 17.4 percent. For details, see the YSK Website at www.ysk.gov.tr.

Gerrymandering in Turkish elections  63 31 Nevertheless, it is hard to argue that all actions upgrading a town to a province or such electoral alterations involved pure gerrymandering tactics. Sometimes, security issues and other factors related to development may play a role as well.

References Abramowitz, A.I., Alexander, B., & Gunning, M. (2006). Incumbency, redistricting, and the decline of competition in US House elections. The Journal of Politics, 68(1),  75–​88. Ahmad, F. (1977). The Turkish experiment in democracy, 1950–​1975. London: Royal Institute of International Affairs. Ahmad, F., & Ahmad, B.T. (1976). Türkiye’de çok partili politikanın açıklamalı kronolojisi, 1945–​ 1971. (The Explanatory Chronology of Turkish Multiparty Politics: 1945–​1971). Vol. 32). İstanbul: Bilgi Yayınevi. Akbulut, Ö.Ö. (2001). Ulusal Siyaset-​Yerel Siyaset İlişkisi Bağlamında 1963 Yerel Seçimleri. Çağdaş Yerel Yönetimler, 10(4),  36–​60. Albayrak, M. (2004). Türk siyasi tarihinde Demokrat Parti:(1946–​1960). Ankara: Phoenix Yayınevi. Altman, M., & McDonald, M. (2010). The promise and perils of computers in redistricting. Duke Journal of Constitutional Law & Public Policy, 5(69), 70–​111. Aygül, C. (2016). Electoral manipulation in March 30, 2014 Turkish local elections. Turkish Studies, 17(1), 181–​201. Balinski, M. (2008). Fair majority voting (or how to eliminate gerrymandering). The American Mathematical Monthly, 115(2), 97–​113. Bartels, L.M. (2000). Partisanship and voting behavior, 1952–​1996. American Journal of Political Science, 44(1),  35–​50. Bayar, C. (1969). Basvekilim Adnan Menderes. Ed. Ismet Bozdag. İstanbul:  Baha Matbaasi. Bickerstaff, S. (2020). Election systems and gerrymandering worldwide: Studies in choice and welfare. Cham: Springer. Browning, R.X., & King, G. (1987). Seats, votes, and gerrymandering: Estimating representation and bias in state legislative redistricting. Law & Policy, 9(3), 305–​22. Burçak, R.S. (1998). On yılın anıları (1950–​1960). (Memories of ten year (1950–​ 1960)). Ankara: Nurol Matbaacılık. Carson, J. L., Crespin, M. H., Finocchiaro, C. J., & Rohde, D. W. (2007). Redistricting and party polarization in the US House of Representatives.  American Politics Research, 35(6), 878–​904. Coate, S., & Knight, B. (2007). Socially optimal districting: A theoretical and empirical exploration. The Quarterly Journal of Economics, 122(4), 1409–​71. Çarkoğlu, A., & Aksen, D. (2019). Partisan and apportionment bias in creating a predominant party system. Political Geography, 69,  43–​53. Çolak, H.Y. (2007). 1930 Belediye Seçimleri (PhD dissertation, Ankara Üniversitesi Türk İnkılap Tarihi Enstitüsü). Demokrat Parti Programı. (1946). www.tbmm.gov.tr/​eyayin/​gazeteler/​web/​ kutuphanede%20bulunan%20dijital%20kaynaklar/​kitaplar/​siyasi%20parti%20 yayinlari/​200805461%20demokrat%20parti%20programi%201946/​200805461%20 demokrat%20parti%20programi%201946.pdf

64  Gülsen Kaya Osmanbaşoğlu Eisenstadt, S.N., & Roniger, L. (1984). Patrons, clients and friends: Interpersonal relations and the structure of trust in society. Melbourne: Cambridge University Press. Eroğul, C. (2003). Demokrat Parti: tarihi ve ideolojisi. Ankara: İmge Kitabevi. Friedman, J.N., & Holden, R.T. (2008). Optimal gerrymandering:  Sometimes pack, but never crack. American Economic Review, 98(1), 113–​44. Grofman, B. (1983). Measures of bias and proportionality in seats-​votes relationships. Political Methodology, 9(3), 295–​327. Günver, G. (2007). Serbest Cumhuriyet Fırkası’nın Aydın’da teşkilatlanması ve 1930 belediye seçimleri üzerinde oluşan tartışmalar. AÜDTCF Tarih Araştırmaları Dergisi, 25(39), 117–​47. Halás, M., & Klapka, P. (2017). Functionality versus gerrymandering and nationalism in administrative geography:  Lessons from Slovakia. Regional Studies, 51(10), 1568–​79. Heper, M. (1986). Local government in Turkey:  An overview with special reference to the municipalities, 1923–​ 1980. In Heper, M. (Ed.), Dilemmas of Decentralization: Municipal Government in Turkey (pp. 7–​36), Bonn:  FriedrichEbert-​Stiftung,  Bonn. —​—​—​. (Ed.). (1987). Democracy and local government: İstanbul in the 1980s. Beverly: Eothen Press. Heper, M., & Keyman, E.F. (1998). Double-​faced state: Political patronage and the consolidation of democracy in Turkey. Middle Eastern Studies, 34(4), 259–​77. Kalaycıoğlu, E. (1994). Decentralization of government. In Heper, M., & Evin, A. (Eds.), Politics in the Third Turkish Republic (pp. 86–​100), Boulder, CO: Westview Press. Karpat, K.H. (2016 [1959]). Turkey’s politics: The transition to a multi-​party system. Princeton, NJ: Princeton University Press. Kaya Osmanbaşoğlu, G. (2014). Typology of the center-​right in Turkey. Unpublished PhD Dissertation. Ankara: Bilkent University Institute of Economics and Social Sciences. repository.bilkent.edu.tr/​bitstream/​handle/​11693/​18499/​10046129.pdf ? sequence=1&isAllowed=y Malesky, E. (2009). Gerrymandering—​Vietnamese style: Escaping the partial reform equilibrium in a nondemocratic regime. The Journal of Politics, 71(1), 132–​59. McCarty, N., Poole, K.T., & Rosenthal, H. (2009). Does gerrymandering cause polarization?. American Journal of Political Science, 53(3), 666–​80. McGann, A.J., Smith, C.A., Latner, M., & Keena, A. (2016). Gerrymandering in America:  The House of Representatives, the Supreme Court, and the future of popular sovereignty. Cambridge: Cambridge University Press. de Mesquita, B.B. (1978). Redistricting and political integration in India. Comparative Political Studies, 11(2), 279–​88. Nationalist Action Party. (2016). Devlet Bahçeli’s lifestory. www.mhp.org.tr/​htmldocs/​ genel_​baskan/​hayati/​mhp/​Devlet_​Bahceli.html Niemi, R.G., & Deegan, Jr, J. (1978). A theory of political districting. The American Political Science Review, 72(4), 1304–​23. Osmaniye Municipality. (2019). Osmaniye Municipality website. osmaniye-​bld.gov.tr/​ baskan/​eski-​baskanlarimiz Owen, G., & Grofman, B. (1988). Optimal partisan gerrymandering. Political Geography Quarterly, 7(1),  5–​22. Öncü, A. (1986). The potential and limitations of local government reform in solving urban problems:  The case of İstanbul. In Heper, M. (Ed.), Dilemmas of Decentralization:  Municipal Government in Turkey (pp. 57–​92). Friedrich-​EbertStiftung, Bonn.

Gerrymandering in Turkish elections  65 Öymen, A. (2013). …Ve ihtilal. İstanbul: Doğan Kitap. Özbudun, E. (2000). Contemporary Turkish politics: Challenges to democratic consolidation. London: Lynne Rienner Publishers. Resmi Gazete. (2008). Metropolitan Municipalities Act, Law No:3030. August 19. www.resmigazete.gov.tr/​eskiler/​2008/​08/​20080819-​7.htm —​—​—​. (2012). Law for new metropoles and towns. December 6. www.resmigazete.gov. tr/​eskiler/​2012/​12/​20121206-​1.htm Sayarı, S. (1977). Political patronage in Turkey. In Gellner, E., & Waterbury, J. (Eds.), Patrons and clients in Mediterranean societies (pp. 103–​ 13). London:  Gerald Duckworth & Company. Sunar, İ. (1990). Populism and patronage:  The Demokrat Party and its legacy in Turkey. Il Politico, 55(4), 745–​57. TBMM Tutanak Dergisi. (1954). June 30. www.tbmm.gov.tr/​tutanaklar/​tutanak/​ tbmm/​d10/​c001/​tbmm10001015.pdf TBMM. (2019). Türkiye Cumhuriyeti milletvekili genel seçimleri. www.tbmm.gov.tr/​ develop/​owa/​secim_​sorgu.genel_​secimler Theriault, S. (2008). Party polarization in Congress. New York: Cambridge University Press. Tunç, H., Erdoğan, M., Yurtlu, F., Ölmez, G., & Çıtak, H.A. (2014). Seçim sistemleri ve Türkiye’de (gerrymandering) seçim hilesi uygulamaları. Ankara: Legal Kitabevi. Turan, A.E. (2008). Türkiye’de yerel seçimler. İstanbul:  İstanbul Bilgi Üniversitesi Yayınları. TÜİK. (2012). Milletvekili genel seçimleri 1923–​2011. Ankara: TÜİK Yayınları. Wuthrich, F.  M. (2015).  National elections in Turkey:  People, politics, and the party system. Syracuse, NY: Syracuse University Press. YSK. (2019). Higher Election Board website. www.ysk.gov.tr

4  The dynamics of change and differentiation in voter preferences in the western coastal provinces of Turkey since the 1980s Tanju Tosun, Betül Aydoğan Ünal and Gülgün Erdoğan Tosun

Introduction Examining the geographical spread of political parties on the electoral map of Turkey indicates that there is a spatial differentiation that has become more apparent in recent years. In other words, spatial voter preferences are highly determined according to the ideological positions of political parties. Rather than there being a balanced distribution among parties over the electoral geography, this differentiation manifests itself as imbalanced on the regional, provincial, and district levels. Although electoral maps for the 1950–​1980 period are more evenly spread, spatial differences became evident with the 1983 elections in the re-​democratization process following the September 12 coup d’état. This has been rapidly increasing since Adalet ve Kalkınma Partisi (Justice and Development Party, the JDP) came to power in the 2002 elections, and became clearly evident after the 2011 elections. Today’s electoral map has a distinguishing feature, as parties with different ideological orientations have a very strong dominance over certain regions. Those places differ not only in terms of political preferences but also in terms of their socio-​economic profiles. The outlook of this differentiation on the electoral map is the fact that the ruling JDP performs below its overall vote average in the region from the Trace in the northwest to the coastal provinces of the Aegean-​Marmara and Mediterranean regions. In return to this, the main opposition party, Cumhuriyet Halk Partisi (Republican People’s Party, the RPP), enjoys above its overall vote average in the same region. On the other hand, the Kurdish parties have a clear dominance in southeastern Anatolia, which has a large Kurdish population. Therefore, it would not be wrong to state that today’s electoral map of Turkey has a tripartite structure. Although the JDP receives considerable votes in the provinces from the Trace in the northwest to the coastal provinces of the Aegean-​Marmara and Mediterranean regions, the party’s share still remains below its national

Differentiation in voter preferences  67 average. The JDP’s main competitor RPP, on the other hand, emerges as the top party in most provinces of the mentioned electoral geography and its average share is much higher in these provinces when compared to its national average. The more interesting point is that the JDP, the RPP, and Milliyetci Hareket Partisi (Nationalist Action Party, the NAP), have increased their votes since 2002 in a context where mainstream right-​wing parties have lost their electoral bases. This study will proceed in four sections. Firstly, the spatial differentiation in the western coastal provinces of Turkey, which is clearly seen on the electoral map, will be examined. Second, the political and sociological dynamics and elements underlying this spatial differentiation will be evaluated in terms of election results. The socioeconomic changes and political developments that the region has undergone will be examined closely. Following, factors that determine voter preferences in this geographical area and in Turkey in general will be discussed with regard to elections held since 1983. This is because voter preferences in the western coastal provinces have been highly different from the rest of the country for a long time. The RPP currently dominates most of this geographical area as a successor of historically dominating the center-​ left and social democrat parties. Provinces in this region have also distinctive characteristics such as higher education and income levels, and cultural/​ethnic diversity. The RPP appeals to relatively better-​educated, middle-​and upper-​income groups with a western lifestyle. The defining characteristics of RPP supporters appear to be matched by many that live on the western coast. However, recent election results illustrate the fact that the JDP has also been rising in this region by mostly receiving the support of center-​right voters, who are generally religious conservatives.

The electoral map of Turkey One of the key themes in electoral geography literature is the effect of geography on voting. What factors motivate people living in different areas to vote differently? In other words, the question of “do similar-​minded people really vote for the same party everywhere regardless of where they live?” has become increasingly important. Rather than focusing on the individual causal relation, focusing on how political institutions are shaped by spatial relations in an attempt to answer this question will deepen the studies in the field of electoral geography.1 The current electoral map of Turkey reveals that the country is sharply divided geographically: the center-​left in the western coast, the conservative masses in central, eastern, and northern Anatolia, and the pro-​Kurdish voters in the southeast (See Figure  1.1 in the Introduction). Whether this reflects any underlying divisions in the Turkish electorate has been widely considered to be one of the critical questions in the literature. A  commonly accepted

68  Tosun, Aydoğan Ünal and Erdoğan Tosun explanation to this question is offered by Mardin (1973), who argues that Turkish voting behavior has been mainly shaped by socially transforming events that date back to the late Ottoman era. According to Mardin, this has resulted in the “center-​periphery” divide between a modernizing center consisting of bureaucratic and military elites and a periphery excluded from state institutions. In the center-​periphery framework, while the center is characterized by a secular, urban, and western way of life, the periphery has a rural, religious, and traditional life-​style. Mardin’s formulation of the center-​periphery has a similar approach with the concept of “social cleavage” introduced by Lipset and Rokkan (1967) to enlighten the mechanism of sociocultural division. This concept has been viewed as a causal factor in explaining the formation of party systems and voting behavior in western democracies for a long time. In a broad sense, this concept can be defined as a deep and lasting division of society into blocs as a consequence of extreme social structural transformations, such as nation building and industrialization, and each bloc develops its own practices in almost every aspect of life. Therefore, party divisions are a fairly determined reflection of the main lines of social cleavages. In spite of the fact that numerous research and empirical data from pre-​ 1980 Turkey displays a solid challenge to this explanation, many researchers of Turkish voting behavior have taken this division as an explanatory of a cleavage inside the party system especially after the military intervention in 1980. Accordingly, center-​left parties represent the views and values of the center, and the right-​wing parties defend the political and social agenda of the periphery (Wuthrich, 2015, p. 55). The western coastal provinces were more industrialized compared to the rest of the country, and as a result, the population residing there were people with higher education and income levels, and with a western understanding of life. Therefore, as expected, they mostly belonged to the center, and tended to vote for center-​left parties. Political scientists, sociologists, and political analysts have come to share a widespread common opinion that the electoral map of Turkey, especially in recent years, poses a tripartite outlook (see Figure 1.1 in the Introduction). This opinion is supported by the fact that the main opposition party, the RPP, has higher voting power in the Trace Region, Aegean, and coastal Mediterranean provinces in comparison to other constituencies. On the other hand, the ruling JDP has a wider support base when one moves from the coastal provinces to the central regions.2 Parties representing the Kurds rank first in constituencies in the southeast. In order to understand whether this proposition has strong foundations and whether it reflects reality, we will first examine the election results from 1983 to 2018 in the western coastal provinces and then in Turkey in general. In addition to this, the voting powers of parties that belonged to the same ideological blocks will be compared. After taking a snapshot of the electorate’s party preferences, the underlying dynamics of these preferences will be evaluated.

Differentiation in voter preferences  69

An indication of new sociological divisions and political polarizations? While Lipset and Rokkan’s (1967) analysis still largely upholds its validity in the literature on the relation between social cleavages and party divisions, it is argued that the new parties are related to new and different cleavages. In other words, they come to existence by still responding to the social cleavages in a country. When revising the relevant literature on social cleavages, Lawson (2004, pp. 252–​253) states that Ronald Inglehart explains the new cleavages with post-​ material value orientations and value change, while Herbert Kitschelt explains them with lifestyle issues of the environment, racism, peace, and gender. Lawson also emphasizes Richard Katz’s assertion that the power of primal cleavages (e.g. religion, language, origin, or location) should not be overlooked. With this regard, it is important to reveal whether the pattern of the electoral map in the western coastal provinces of Turkey corresponds to new social cleavages shaped on the basis of lifestyle. The provinces in this region are Edirne, Kırklareli, Tekirdağ in Thrace, Çanakkale, Balıkesir, İzmir, Aydın, Muğla3 in coastal Marmara and coastal Aegean, Antalya, Mersin, Adana, and Antakya in coastal Mediterranean. Consisting of 12 provinces, the western coastal provinces of Turkey have 21.8  percent (12,977,025)4 of all the total registered electorate of Turkey (59,367,469) as of the last general elections (June 24, 2018). In other words, this area consists of one-​fifth of the total electorate in Turkey. Another characteristic of these provinces is that they have some of the highest average scores in the socioeconomic ranking of provinces and regions (SEGE, 2013, p.  50) according to the Ministry of Development. İzmir is ranked third, Muğla eighth, Tekirdağ ninth, Edirne 12th, Çanakkale 14th, Kırklareli 15th, Antalya fifth, Adana 16th, Mersin 24th, Aydın 19th, Balıkesir 22nd, and Hatay 46th. These provinces draw our attention not only for being above the average of Turkey in terms of socioeconomic development, but also in terms of other indicators.5 As a result of migration from underdeveloped regions, the population in these provinces is more heterogeneous in terms of social life and culture, more pluralistic, more western-​oriented, and more embracive of modern social values compared with other regions. Nonetheless, there is no significant differentiation between the electorate of these provinces and the rest of Turkey in terms of lifestyle clusters determined by the lifestyles in which they define themselves (See Figure 4.1). The most striking finding of Figure 4.1 is that the rate of the electorate defining their lifestyles as “modern” and living in coastal provinces is higher than the average of Turkey. On the other hand, the rate of those defining their lifestyles as “religious conservative” is less when compared to the average of Turkey, while those defining themselves as “traditional conservative” is higher in coastal provinces, although the margin is narrow. This data is remarkable in the sense that almost half of the electorate in western coastal provinces and of Turkey as a whole define themselves as “traditional conservative”. Likewise,

70  Tosun, Aydoğan Ünal and Erdoğan Tosun 50.0

47.5

45.0

46.6

40.0 35.0 30.0

32.9 26.7

25.0

26.7 19.6

20.0 15.0 10.0 5.0 0.0

Modern

Traditional Conservative Religious Conservative West Coast Provinces Turkey Average

Figure 4.1 Comparison of lifestyle clusters defined by electors themselves in western coastal provinces and in Turkey in general Note: The graph was prepared by using Konda Barometre (2010–​2017) data obtained from the Data Warehouse of Konda Research. Although the analyses in this study cover 12 coastal provinces, as Konda data covers nine provinces (Adana, Antalya, Balıkesir, Çanakkale, Edirne, Hatay, İzmir, Mersin, Tekirdağ), the lifestyle clusters of western provinces are presented as proportioned by calculating the average of nine provinces. The average of Turkey was calculated using the data of 42 provinces representing the country.

it is notable that one out of every five electors in the coastal west define their lifestyle as “religious conservative”.

Voter preferences in Turkey’s western coastal provinces from 1983 to 2018 Party preferences in the general elections from 1983 to 2018, which we will elaborate on below, will give us an idea in this regard. In this study, we will examine voter preferences between 1983 and 2018 as two separate periods: 1983–​1999 and 2002–​2018. The reason for this is that the two periods are different in terms of parties, leaders, and electoral behaviors. Examining Table 4.1 in terms of parties and ideological blocks, it is clear that the vote averages of parties in western coastal provinces vary from one election to the other.6 In other words, there is a high level of volatility in western coastal provinces from 1983 to 2018. For example, Refah Partisi (Welfare Party, the WP) votes increased from 3.1  percent in 1987 to 10.3  percent in 1995, whereas Fazilet Partisi (Virtue Party, the VP), the party founded to replace the WP, received 6.3 percent of the vote in 1999. Halkçı Parti (Populist Party,

Differentiation in voter preferences  71 Table 4.1 Average votes of parties in 12 western coastal provinces (1983–​1999) Party/​Election

6-​Nov-​83

WP–​VP* -​ PP–​SDPP–​RPP** 33.4 DLP -​ MP 40.4 TPP -​ NLP–​NAP*** -​ NDP 23.1 YP -​ VP -​ GUP -​ PDemP -​

29-​Nov-​87

20-​Oct-​91

24-​Dec-​95

18-​Apr-​99

3.1 29.4 9.0 34.6 21.3 1.8 -​ -​ -​ -​ -​

7.7 23.2 12.3 23.7 32.7 -​ -​ -​ -​ -​ -​

10.3 13.5 19.4 18.3 25.7 8.6 -​ -​ -​ -​ 2.7

6.3 10.4 30.5 14.8 13.4 17.4 -​ -​ -​ -​ 3.1

* WPVP line: Based on the votes of the Welfare Party (the WP) in the 1983, 1987, 1991, and 1995 elections. The votes of this party for the 1999 elections were considered to be the votes of the Virtue Party (the VP), which was established to replace this party after it was shut down with the decision of the Constitutional Court following the post-​modern coup in February 28, 1997. ** PP–​SDPP–​RPP line:  Votes of the Populist Party (the PP) were considered in the 1983 elections, the Social Democratic Populist Party (the SDPP) in the 1987 and 1991 elections and the RPP in the 1995 and 1999 elections. *** NLP–​NAP line: It participated as Nationalist Labor Party (the NLP) in the 1987 elections and as the NAP in the later elections. Source:  The data in this table was compiled from the following website:  biruni.tuik.gov.tr/​ secimdagitimapp/​secim.zul

the PP) from the PP-​SDPP-​RPP line received 33.4 percent of the vote in 1983, whereas Sosyaldemokrat Halkçı Parti (Social Democratic Populist Party, the SDPP) replacing it received 29.4  percent in 1987 and 23.2  percent in 1991, and later continuing as the RPP following the unification of the SDPP and the RPP, it received 13.5 percent of the vote in 1995 and dropped to 10.4 percent in 1999. Demokratik Sol Parti (Democratic Left Party, the DLP), in the same ideological block with the RPP, received 9 percent of the average vote in 1987, which increased to 12.3 percent in 1991, 19.4 percent in 1995, and 30.5  percent in 1999. This data indicates the fact that the electorate with a social-​democratic tendency headed from the RPP to DLP. Anavatan Partisi (Motherland Party, the MP) at the center-​right regularly lost votes in these provinces after the 40.4  percent vote it received in 1983. Hence, its average decreased from 34.6 percent in 1987 to 23.7 percent in 1991, to 18.3 percent in 1995 and to 14.8 percent in 1999. Meanwhile, the votes of Doğru Yol Partisi (True Path Party, the TPP), its competitor in the same block, showed a fluctuating course. The TPP’s 21.3 percent vote in 1987 increased to 32.7 percent in 1991 and then dropped to 25.7 percent in 1995 and to 13.4 percent in 1999. NAP, at the extreme nationalist right gradually increased its electoral base. Its 1.8 percent average in 1987 increased to 8.6 percent in 1995 following its election alliance with the WP in 1991 and then

72  Tosun, Aydoğan Ünal and Erdoğan Tosun increased to 17.4  percent in 1999. Milliyetçi Demokrasi Partisi (Nationalist Democracy Party, the NDP) supported by the generals of the September 12 coup reached a vote of 23.1 percent in the western coastal provinces in 1983, when it participated in the elections for the first and last time. On the other hand, Halkın Demokrasi Partisi (People’s Democracy Party, the PDemP), the representative of the Kurdish electorate had a vote of 2.7 percent in 1995 and 3.1 percent in 1999. It would be correct to take the 1995 election as a basis in order to see the vote change in ideological party blocks during this period. Let us first examine the issue in terms of the center-​right. This is because the MP and TPP votes, which had a total of 51  percent countrywide and 56.4  percent in the 12 western coastal provinces in the first half of the ’90s (in the 1991 elections), decreased to a total of 38.8 percent countrywide and to 44 percent in the coastal provinces in the 1995 elections in connection with their collapse. The 12 percent loss across the country and in the coastal provinces from 1991 to 1995 increasingly continued in the subsequent elections. The important issue here was the parties to which electors withdrawing from the center-​right were inclined to turn to. Considering the fact that the WP’s coastal province average dropped from 10.3 percent in the 1995 elections to 6.3 percent in the 1999 elections, it is understood that a considerable portion of the electors was beginning to detach from center-​right parties—​in fact there was a 13.2 point loss in the total votes of these two parties from 1995 to 1999 (the MP and the TPP together had 44 percent in 1995 and 30.8 percent in 1999)—​and was headed towards the DLP on the center-​left and to the NAP on the extreme nationalist right. Hence, this judgment seems to be correct when we consider the fact that the RPP and the DLP’s total vote increased from 32.9 percent in 1995 to 40.9 percent in 1999, and the NAP’s total vote increased from 8.6 percent in 1995 to 17.4 percent in 1999. Table 4.2 displays some striking trends in terms of voter volatility in the region. It can be stated that on the center-​right, the MP and the TPP were virtually wiped from the electoral geography in western coastal provinces and across the country as of the 2007 elections. The MP’s 6.9  percent and the TPP’s 14.5 percent vote shares in these provinces in 2002 vanished in the subsequent elections. The same is true for their nationwide average. Their 20.2 percent of the total vote fully depleted. On the other hand, a new party founded by a businessman, Genç Parti (Young Party, the YP), became visible with a considerable vote for a couple of elections and then disappeared, receiving 10.6 percent in 2002 and 6.2 percent in 2007. In order to understand where the electoral base of these parties headed towards, it is necessary to examine the electoral performances of the JDP, the RPP, and the NAP since the 2002 elections. To begin with, the JDP’s vote in the western coastal provinces reached 20.2  percent in 2002. Considering the fact that its predecessor, the Virtue Party, only received 6.3 percent of the vote in the 1999 elections, it is legitimate to state that the JDP began to attract the center-​right electorate even in the

Differentiation in voter preferences  73 Table 4.2 Average votes of parties in 12 western coastal provinces (2002–​2018) Party/​Election

3-​Nov-​02

22-​Jul-​07 12-​Jun-​11 7-​Jun-​15

1-​Nov-​15 24-J​ un-1​ 8

JDP RPP DLP MP TPP NAP NDP YP VP GUP DPP–​PDP

20.2 26.0 1.9 6.0 14.2 10.5 -​ 10.6 1.1 0.6 3.6

31.0 30.2 -​ -​ -​ 19.4 -​ 6.2 1.2 -​ -​

35.9 40.8 0.1 -​ 0.0 14.5 -​ -​ 0.5 0.4 6.3

36.7 40.3 0.3 -​ 0.2 16.2 -​ -​ 0.7 0.4 -​

30.6 39.6 0.2 -​ 0.1 18.7 -​ -​ 1.2 -​ 7.5

31.6 35.7 -​ -​ -​ 8.3 -​ -​ 0.8 -​ 7.8

* Parties that politically represent Kurds participated as the DPP in the 2002 elections, with independent candidates in the 2007 and 2011 elections and as the PDP in the June 7 and November 1, 2015 and 2018 elections. Source:  The data in this table was compiled from the following websites:  biruni.tuik.gov.tr/​ secimdagitimapp/​secim.zul, www.ysk.gov.tr

very first election it appeared in the ballot. Furthermore, the JDP managed to increase its votes to 31 percent, 36.7 percent, 30.6 percent, and 35.9 percent in the 2007, 2011, 2015, and 2018 elections respectively, which means that it consolidated its center-​right electorate in the western coastal districts. On the center-​left, the RPP increased its vote of 10.4 percent in 1999 to 26 percent, 30.2 percent, 40.3 percent, 39.6 percent, and 40.8 percent in the 2002, 2007, 2011, June 7, 2015, and November 1, 2015 elections respectively, while the DLP’s 30.5 percent vote in 1999 was almost fully wiped out in the subsequent elections. Although the loss of the DLP may be the main factor to explain the RPP’s rise, up to 40 percent of votes after the 2011 elections indicate that the party also received some support from electors who previously voted for center-​right parties. Indeed, the RPP has received approximately 50  percent of the vote share since 2007 in provinces such as Aydın, Muğla, and Tekirdağ, where the center-​right MP and TPP were strong in the past. Still, it should also be emphasized that the JDP’s 20.2 percent vote in western coastal provinces in 2002 reached 35 percent in 2015 and 31.6 percent in recent elections, which suggests that a considerable part of the center-​right electorate has headed towards the JDP. On the other hand, the NAP, which only managed to receive 10.5 percent of the vote in 2002, increased its share to 19.4 percent in 2007. This indicates that, depending on the mode of nationalist inclinations of the electorate, a considerable section of the center-​right electorate headed towards the NAP in the 2000s. The rise of Turkish nationalism can be the outcome of 6 to 7 percent vote share of the Kurdish political parties (Democratic People’s Party, the DPP and People’s Democratic Party, the PDP) in the same provinces.

74  Tosun, Aydoğan Ünal and Erdoğan Tosun Table 4.3 Comparison of average votes of parties in 12 western coastal provinces with their average votes throughout Turkey (1983–​2018) Party/​Election

1983–​2002 (Turkey)

1983–​2002 (Western Coast)

2002–​2018 (Turkey)

2002–​2018 (Western Coast)

WP-​VP-​JDP PP-​SDPP-​RPP DLP MP TPP NLP-​NAP NDP YP FP GUP PDemP-​DPP-​PDP Good Party

15.2 19.1 14 27.6 19.3 9.7 23.3 -​ -​ 1.5 4.4 -​

6.9 22 17.8 26.3 23.2 9.3 23.1 -​ -​ 0.6 2.9 -​

43.9 22.7 0.4 5.1 2.4 12.5 -​ 5.1 1.5 0.7 10.3 10

31.1 35.5 0.6 6 3.6 14.6 -​ 8.4 0.9 0.5 6.1 14.1

Source: The numerical data in this table was compiled from the following websites: biruni.tuik. gov.tr/​secimdagitimapp/​secim.zul; www.ysk.gov.tr

Table  4.3 illustrates the vote averages of political parties and blocks in the western coastal provinces between 1983–​2002 and 2002–​2018 alongside their vote averages on the national level. This comparison indicates which direction the electorate, especially the center-​right, headed to. The first issue that draws attention in the table is that the average vote share of the MP and the TPP on the center-​right in the 1983–​2002 period (46.9 percent) is slightly lower than their vote averages in western coastal provinces (49.5  percent). This changed in the 2002–​2018 period. The two dissolving parties had a total vote average of 7.5  percent countrywide and 4.2  percent in the mentioned provinces. It clearly indicates that the voting base of the center-​right block further dissolved in these provinces compared to the nationwide vote share. Nevertheless, the question of where the center-​right electorate moved after distancing themselves from their parties in western coastal provinces still stands. As shown in Figure 4.2., the WP–​VP line received approximately ten points less than its country average in western coastal provinces (6.9 percent). Similarly, the JDP had a lower voting average in western coastal provinces (31.1  percent) compared to its country average (43.9  percent) in the 2002–​ 2018 period. Yet, the differences between the vote shares of the WP–​VP line and the JDP in the western coastal provinces is remarkable when comparing the 1983–​2002 period with the 2002–​2018 period. It is again clear that that the electorate heading from the center-​right to the JDP had a significant contribution in its vote share. In any case, the 2002 elections were a remarkable development where pro-​Islamic and nationalist parties received a vote of approximately 53 percent, which was their all-​time highest vote rate compared to the pre-​2002 era (Çarkoğlu, 2002).

Differentiation in voter preferences  75

rty

P

Pa

D

G

oo d

-P PP D PD

em

P-

N

LP -

N

A

P

PP PP D -S PP

W

P-

V

P-

-R

JD

P

50 45 40 35 30 25 20 15 10 5 0

1983-2002 (Turkey)

1983-2002 (Western Coast)

2002-2018 (Turkey)

2002-2018 (Western Coast)

Figure 4.2 Comparison of average votes of major parties in 12 western coastal provinces with their average votes throughout Turkey (1983–​2018)

A similar increase is also true for the RPP to a certain extent.7 While the RPP had a vote rate of 22 percent in the western coastal provinces that was only three points higher than its Turkey average (19.1 percent) between 1983 and 2002, in the subsequent period, an average of 35.5  percent against a country average of 22.7  percent is remarkable.8 The most concrete conclusion that can be drawn from these electorate inclinations is that although the RPP’s electoral base in the western coastal provinces (35.5 percent) is wider compared to that of the JDP (31.1 percent), the JDP managed to enlarge its electorate compared to the RPP. The JDP’s vote average in these provinces is 24 points higher than WP–​VP’s average in the1983–​2002 period, and the RPP increased its share by 13.5 points. However, in terms of the aggregate average of center-​left votes, the 39.8 percent vote share of the previous period decreased to 36  percent. This means that although the RPP was stronger than the JDP in terms of vote average, its vote share could not grow as much as the JDP in this region. Moreover, the RPP seems to fail to receive considerable support from the former center-​right electorate, and has only managed to become the first party in this region by attracting the former DLP votes.9 The JDP also increased its votes in the November 1, 2015 elections compared to the June 7, 2015 elections (Çarkoğlu & Yıldırım, 2015). Similar to June 7, the RPP draws attention as the strongest party in terms of sociological base, especially in coastal provinces (Gökmen & Tosun, 2015). The country average of the NLP–​NAP line was 0.4 points higher than western coastal provinces in the 1983–​2002 period. However, in the 2002–​ 2018 period, the NAP’s 14.6 percent vote share in western coastal provinces is 2.1 points higher than its 12.5  percent country average. Also, there is a growing NAP presence in western coastal provinces in recent years. Its

76  Tosun, Aydoğan Ünal and Erdoğan Tosun average of 9.3  percent in the first period increased to 14.6  percent. It can be suggested that this rise of the NAP is mostly related to the conjectural increase in nationalist sensitivities of the electorate who previously voted for center-​right parties. The reality proclaimed by the aforementioned situation is that the center-​ right electorate, which first dissolved and then detached from these parties in relation to the representation and legitimacy crisis encountered as of the second half of the 1990s, searched for parties and distributed their preferences among the JDP, the RPP, and the NAP in western coastal provinces as they did throughout Turkey. Although the RPP has a higher vote average compared to the JDP in the mentioned provinces, it can be stated that the JDP and the NAP are the parties that are actually growing in this electoral geography. In the present situation, it is a reality that the JDP does not have the same strength in western coastal provinces as it does countrywide, and is roughly 13 points behind. The RPP receives 13 points higher than its country average, and although the NAP’s vote was around 9 percent countrywide and in these provinces during the previous period, it began to receive 5.3 points higher in these provinces compared to its country average in recent years. Parties that are political representatives of Kurds (PDemP–​DPP–​PDP) only receive 6.1 percent in western coastal provinces, which is four points less than their county average. At this point, there are fundamental questions that need to be answered. What are the reasons behind the fact that the JDP does not receive its countrywide vote in western coastal provinces? Despite its low countrywide performance, why is the RPP more powerful in these regions compared to its strongest competitor? Finally, why is the NAP’s electoral performance in these provinces stronger than its overall national vote? After examining the vote averages of parties participating in the five general elections held between 1983 and 2002, we consider the dominance of these parties in the electoral competition in 12 provinces. As Table  4.4 demonstrates, the two parties in the WP–​VP line on which the JDP tradition is based have no provinces where they ranked as the first party in the coastal west in the five elections held from 1983 to 2002. Parties in the center-​left, consisting of the PP–SDPP–​RPP ranked first in four provinces in the 1983 and 1987 elections and in one province in the 1991 and 1995 elections. On the other hand, in 1999, the RPP failed to rank first in any of these provinces. The center-​left DLP in the same ideological line ranked as the first party in five of these provinces in the 1995 elections and in nine provinces in the 1999 elections. The center-​right MP ranked as the first party in eight provinces in the 1983 and 1987 elections, but failed to come first in any province in the 1991 and 1999 elections, and ranked first only in one province in the 1995 elections. In the same ideological line, the TPP ranked as the first party in nine provinces in the 1991 elections and in five provinces in the 1995 elections. In the extreme nationalist right, the NAP ranked as the first party in three provinces in the 1999 elections.

Differentiation in voter preferences  77 Table 4.4 The number of provinces parties ranked first (1983–​2002) Party/​Election

6-​Nov-​83

29-​Nov-​87

20-​Oct-​91

24-​Dec-​95

18-​Apr-​99

WP–​VP PP–​SDPP–​RPP DLP MP TPP NLP–​NAP NDP YP FP GUP PDemP

-​ 4 -​ 8 -​ -​ -​ -​ -​ -​ -​

-​ 4 -​ 8 -​ -​ -​ -​ -​ -​ -​

-​ 1 -​ -​ 11 -​ -​ -​ -​ -​ -​

-​ 1 5 1 5 -​ -​ -​ -​ -​ -​

-​ -​ 9 -​ -​ 3 -​ -​ -​ -​ -​

Source: The data used in this table was compiled from the following websites: biruni.tuik.gov.tr/​ secimdagitimapp/​secim.zul, www.ysk.gov.tr

Table 4.5 The number of provinces parties ranked first (2002–​2018) Party/​Election

3-​Nov-​02

22-​Jul-​07 12-​Jun-​11 7-​Jun-​15

1-​Nov-​15 24-​Jun-​18

JDP RPP DLP MP TPP NAP NDP YP FP GUP DPP–​PDP

4 8 -​ -​ -​ -​ -​ -​ -​ -​ -​

6 5 -​ -​ -​ 1 -​ -​ -​ -​ -​

6 6 -​ -​ -​ -​ -​ -​ -​ -​ -​

6 6 -​ -​ -​ -​ -​ -​ -​ -​ -​

4 8 -​ -​ -​ -​ -​ -​ -​ -​ -​

6 6 -​ -​ -​ -​ -​ -​ -​ -​ -​

Source: The data used in this table was compiled from the following websites: biruni.tuik.gov.tr/​ secimdagitimapp/​secim.zul; www.ysk.gov.tr

With the inclusion of the JDP into the system, in the six elections held during the 2002–​2018 period, it is remarkable to note an evident change in parties that ranked first in the western coastal provinces. While the WP–​VP, whose heritage was assumed by the JDP, could not rank first in any of the 12 coastal provinces before 2002, the JDP succeeded in ranking as the first party in four provinces in 2002, in six provinces in 2007 and 2011, on November 1, 2015, and in 2018, and in four provinces on June 7, 2015. Examining the success of the RPP, it is observed that the party received the highest vote in eight provinces in 2002 and on June 7, 2015, in six provinces in 2018, on November 1, 2015, and in 2011, and in five provinces in 2007. The NAP ranked as the first party in one province only in the 2007 elections, and

78  Tosun, Aydoğan Ünal and Erdoğan Tosun the MP and the TPP, the mainstream right-​wing parties of pre-​2002 Turkey, failed to rank as the first party in any province. A similar pattern is also valid for the DLP, the party that won the national elections and formed a coalition government in 1999. It is clear that the electorate in the western coastal provinces has headed towards the RPP and the JDP since 2002. In order to evaluate the party whose electorate constituted the source of this gathering, it could be useful to examine the change throughout the elections in the provinces where the parties ranked first. According to Table  4.6, when a right-​wing party lost its first rank in a province, it was usually replaced by another right-​wing party in the general elections from 1983 to 2002. For example, the TPP ranked first in the 1991 elections in the provinces that were previously won by the MP. Interestingly, from 1991 to 1995, the TPP lost the provinces it ranked first to the DLP and continued its dominance in certain provinces. In the change from 1995 to 1999, it is remarkable that the TPP lost many provinces to the DLP and one province to the NAP. In terms of the basic parameters of the change in the 2002–​2018 period, the RPP and the JDP mostly maintained the provinces they ranked first from one election to the other. Hence, the RPP ranked the first party in Edirne, Kırklareli, Tekirdağ, İzmir, Muğla, and Aydın, except in 2007. The same is mostly valid for the JDP in Çanakkale, Balıkesir, Adana, and Hatay. This pattern reveals that the RPP and the JDP have been consolidating their electorates in western coastal provinces as they did throughout the country. In addition to the RPP and the NAP enlarging their electoral base to a limited extent with incoming support from the center-​right in western coastal provinces, especially after 2002, the JDP has gradually enlarged its base. Hence, these provinces have today very evidently assumed a tripartite structure controlled by the JDP, the RPP, and to a lesser extent by the NAP. In other words, the electoral geography left by the center-​right in western coastal provinces has today been filled by these three parties. Findings reflect a change in this pattern as of the 2017 referendum, which also suggests some changes are likely to occur in voter preferences in this geography in the forthcoming period. The fact that “No” votes were considerably higher than the country average in western coastal provinces in the constitutional referendum is a precursor of this situation. In the constitutional referendum held on April 16, 2017, which was a vote on whether Turkey should have a transition from the parliamentary system to a presidential one, the “Yes” vote marginally won with 51.2 percent. Nevertheless, the average of “Yes” and “No” votes in the 12 western coastal provinces were very different from the country average. “Yes” votes supporting the constitutional amendment remained at 36.9  percent whereas “No” votes had an average of 63.1  percent. In these 12 provinces “Yes” votes ranged between 28.7  percent (Kırklareli) and 45.6  percent (Hatay), while “No” votes ranged among provinces between 54.4 percent (Hatay) and 71.3 percent (Kırklareli).

newgenrtpdf

Table 4.6 Provinces where parties ranked first according to elections in western coastal provinces 1983

1987

1991

1995

1999

2002

2007

2011

7 Jun 2015

1 Nov 2015

24 Jun 2018

Edi̇ rne Kirklareli̇ Teki̇ rdağ Çanakkale Balikesi̇ r Ïzmir Aydın Muğla Antalya Mersi̇ n Adana Hatay

PP PP MP MP MP HP MP MP MP MP MP PP

SDPP SDPP MP MP MP MP MP MP MP SDPP MP SDPP

TPP TPP TPP TPP TPP TPP TPP TPP TPP TPP TPP SDPP

DLP DLP DLP TPP TPP DLP TPP TPP TPP MP DLP RPP

DLP DLP DLP DLP DLP DLP DLP DLP NAP NAP DLP NAP

RPP RPP RPP JDP JDP RPP RPP RPP RPP RPP JDP JDP

RPP RPP RPP JDP JDP RPP JDP RPP JDP NAP JDP JDP

RPP RPP RPP JDP JDP RPP RPP RPP JDP JDP JDP JDP

RPP RPP RPP RPP JDP RPP RPP RPP JDP RPP JDP JDP

RPP RPP RPP JDP JDP RPP RPP RPP JDP JDP JDP JDP

RPP RPP RPP JDP JDP RPP RPP RPP JDP JDP JDP JDP

Notes: Parties: Populist Party (PP), Motherland Party (MP), Social Democratic Populist Party (SDPP), True Path Party (TPP), Democratic Left Party (DLP), Republican People’s Party (RPP), National Action Party (NAP), Justice and Development Party (JDP).

Differentiation in voter preferences  79

Provinces/​Parties

80  Tosun, Aydoğan Ünal and Erdoğan Tosun Considered as a whole, the following section will examine the underlying dynamics of center-​right voters with liberal tendencies heading towards the DLP on the center-​left and to the NAP on the extreme nationalist right in the second half of the 1990s, to the JDP, the RPP, and to the NAP as of the early 2000s.

Dynamics of voter preferences in western coastal provinces The neo-​liberal economic policies implemented in the 1980s triggered a wave of mass migration to the rural areas of relatively industrialized cities, and this exacerbated income inequality. Newly settled dwellers of big cities could find their voice in the pro-​Islamist parties (Öniş, 1997). The fact that pro-​ Islamist parties were able to extend their electoral base with the support of the urban poor increased political polarization significantly along secular and Islamist lines (Toprak, 2005). Çarkoğlu (2012, p. 162) argues that despite urbanization and the implementation of market-​led economic policies, the cultural divide between the center and the periphery has continued to exist in society. The modernization project, which has existed since the creation of the Republic, has evolved into unforeseen and unplanned formations. The period between 1995 and 1999 is regarded as the process of the periphery’s radicalization. Those from the periphery who supported the MP in the 1980s started to support the WP during this period. The years between 1997 and 2002 are considered as a restoration and renewal period of the periphery. This can be interpreted as an important shift in the political preferences of the periphery towards conservatism (Akyol, 2013, pp. 439-​40). Short-​living coalition governments and economic turmoil have triggered a search for an alternative in the center-​right. As the JDP remained as the dominant party for over a decade, the exclusion of the periphery from power relations lost its reality and this allowed for the periphery to decide who the winner is. The distance between the cultural center and the periphery gradually shrank at the end of every election since 2002. When compared to the politics of the 1990s, as of 2010, the political arena in Turkey has become much more stable and less volatile. As a natural consequence of this, all major political parties switched their campaign focuses from the issues of culture and identity to more pragmatic issues, which were concerned in improving the standards of daily life (Wuthrich, 2013). We should also consider the evaluation of the economic performance of the incumbency from the individual’s viewpoint, as this is another important factor that determines voter preferences. As long as the JDP has a satisfactory economic performance, their electoral base expands. The same also applies to the western coastal provinces. The tension between secular and religious camps peaked during the republican rallies that took place several months before the presidential elections of 2007 in Ankara, İstanbul, İzmir, Manisa, and Çanakkale. Reportedly,

Differentiation in voter preferences  81 millions of people10 marched in a series of peaceful protests initiated by Atatürkçü Düşünce Derneği (Association for Atatürkist Thought) in support of the secular nature of the state. It was basically a reaction against the probability of the election of a religious conservative candidate as president. Chanting the slogan “Turkey is secular and will remain secular”, this was the last resistance point of the secular camp (Aslan, 2017, p. 93). The leadership of Erdoğan as the representative of the pious periphery has an enormous and facilitating impact on the periphery’s march to the center by struggling to survive against the principle of the separation of religion and state. The periphery has become a successful social, economic, and political rival to the center, and the hegemony of the republican establishment has ended on that account (Heper, 2013). The traditionalist and religious periphery have started to dominate the center, which was previously occupied by the secular and modern camp. Since then, the Kulturkampf between the secularists and religious conservatives has prevailed due to the deteriorating relationship between the JDP and main opposition parties in this process (Kalaycıoğlu, 2011; 2012, p. 174). This transformation revealed in the campaign speeches of party leaders can be seen as a natural consequence of the gradual elimination of the military, a strong defender of the secular state, as an extra-​parliamentary actor since the second half of the 2000s. While the JDP could make references to religious freedom without seeing any need to clarify its position, secularist parties gradually changed their attitudes from statist and restrictive secularism to a more pluralist and passive secularism (Dikici Bilgin, 2018, p. 189). Especially after 2011, it became much clearer that the JDP, under the strong leadership of Erdoğan, can powerfully consolidate its electorate by using polarizing language. As the main parties and their electoral bases have become more stable, the main divide separating society into two has changed, whether being from the center or the periphery, to either supporting President Erdoğan’s positions or opposing them. The Turkish public has become more polarized on numerous issues, and eventually, two main camps formed in the country: one for President Erdoğan and one against him (Erişen, 2018, p. 25). According to a survey called the “Dimensions of Polarization in Turkey”, which was carried out in 2017, political polarization among Turkish society is growing (Erdoğan, 2018). The survey reveals that two camps live in their own “echo-​chambers” in which they can only exchange affirmative views among them and shut themselves off from the other camp by filtering out opinions and news that may affect their beliefs. Social media platforms are also included within this, as people tend to be friends (or connect with people) with similar political leanings.

Conclusion As illustrated in this study, although the electoral map clearly shows that Turkey is divided into three main geographical and social blocks, the

82  Tosun, Aydoğan Ünal and Erdoğan Tosun question as to what divides society does not have a simple and an all-​time-​ valid answer. As society transforms itself, the dynamics of voter preferences change. Due to the exclusion of the periphery from the political power struggle, the center-​periphery framework is key to understanding Turkish voting behavior in the 1980s. When groups of diverse people began moving to cities as a result of neo-​liberal economic policies, the process of urbanization began for this group in society. Rather than fully adopting secular worldviews in modern cities, those people continued to hold on to their religion as part of their political and cultural identity. They were thus able to find their voice in the pro-​Islamist WP, which came to power both at the major city and national level in the mid-​1990s. As a result, the periphery’s distance to the political power center shrank to a greater extent, and the main separating element of the center-​periphery divide has been resolved. Yet then, with the clash between the WP and secular institutions, the polarization of society between the secularists and religious conservatives became much more prominent. Following years of unstable coalition governments and economic turmoil, almost all political parties lost their credibility in the eyes of voters, and a search for an alternative party began. With the support of religious and traditional conservatives, the JDP gained the majority in parliament even in its first general elections in 2002. The JDP’s success story continued after every local and general election with an exception of 2019 local elections where it lost some of the major cities like İstanbul, Ankara, and Antalya to the RPP. Still, the more the party stayed in power, the more they found the courage to eliminate the check-​and-​balance functions of state institutions over the government. In the end, the religious masses had no vital reason left to clash with the secularists. However, under the strong leadership of Erdoğan, the JDP has increasingly utilized polarizing rhetoric to create a new identity in the “new Turkey”. This polarization is evident both at the public and the party level. Attempts to form a coalition government after the June 2015 general elections failed in a short space of time due to the irreconcilable differences among political parties. Society is now highly polarized around being for or against Erdoğan. The current Turkish electoral landscape tells us that the RPP appeals to inhabitants of the western coastal region who are relatively better-​educated, and who are from middle-​and upper-​income groups. Although the RPP aims to be a representative of all classes and groups as the founding party of the Republic, it is almost non-​existent in eastern and southeastern Turkey and seems to be stuck in the western coastal provinces where people have higher levels of socio-​economic status. This study shows that the results of the last general elections indicate that the RPP has reached its natural limits of electoral base to a large extent in that area while the JDP has been gradually increasing its electoral power through solidifying the pro-​Erdoğan camp and including the center-​right.

Differentiation in voter preferences  83

Notes 1 For a similar approach, see Clem (2006), Martis et al. (1992), and Walks (2006). 2 This situation is evaluated as a “duality” by some authors. See Bumin (2009). 3 For a study analyzing the electorate preferences in Muğla, see Kiriş and Köklü (2012). 4 www.ysk.gov.tr. 5 For a study examining geography, socio-​economic factors, and party preferences, see Kaya (2008). 6 Comparison of average votes of parties in 12 western coastal provinces with their average votes throughout Turkey (1983–​2018) will be discussed in detail later. 7 For a study evaluating the results of the June 7 elections in the case of RPP, see Tosun and Erdoğan Tosun (2015). 8 For an evaluation on this topic, see Tosun (2010). 9 The fact that the RPP achieved growth in the 2011 elections with the DLP’s votes was clearly observed in the 2011 elections. For an evaluation on this topic, see Çarkoğlu (2011, pp. 48–​49). 10 “Turks protest ahead of early elections”. The Sydney Morning Herald, May 14, 2007, www.smh.com.au/​world/​turks-​protest-​ahead-​of-​early-​elections-​20070514-​ cjy.html

References Akyol, E. (2013). Merkez Çevre İlişkileri ve Modelin Geçerliliği. In A. Karadağ (Ed.), Osmanlı’dan Cumhuriyet’e Türkiye’de Siyasal Hayat (pp. 393-​451). Ankara: Orion Kitabevi. Aslan, A. (2017). Yerli ve Milli Siyaset. In İ. Çağlar & A. Aslan (Eds.), AK Parti’nin 15 Yılı Toplum (pp. 77–​104). SETA. Bumin, K. (2009). Bu ‘ikiciliği’ Türkiye siyaseti de aşmalıdır. Haber 7. www.haber7. com/​yazarlar/​kursat-​bumin/​392798-​bu-​ikiciligi-​turkiye-​siyaseti-​de-​asmalidir Clem, R.S. (2006). Russia’s electoral geography:  A review. Eurasian Geography and Economics, 47(4), 381–​406. Çarkoğlu, A. (2002). Turkey’s November 2002 elections:  A new beginning?. Middle East Review of International Affairs, 6(4),  30–​41. —​—​—​. (2011). Turkey’s 2011 General Elections: Towards a Dominant Party System. Insight Turkey, 13(3),  43–​62. —​—​—​. (2012). Voting behavior in Turkey. In M. Heper & S. Sayarı (Eds.), The Routledge handbook of modern Turkey (pp. 160–​70). London: Routledge. Çarkoğlu, A., & Yıldırım, K. (2015). Election storm in Turkey: What do the results of June and November 2015 elections tell us? Insight Turkey, 17(4),  57–​79. Dikici Bilgin, H. (2018). Social conflicts and politicised cleavages in Turkey. In S. Sayarı, P. Ayan Musil, & Ö Demirkol (Eds.), Party politics in Turkey: A comparative perspective (pp. 179–​95). London: Routledge. Erdoğan, E. (2018). Dimensions of polarization in Turkey: Social distance, perceived moral superiority, and political intolerance. The German Marshall Fund of the United States. www.gmfus.org/​publications/​dimensions-​polarization-​turkey Erişen, C. (2018). Political behavior and the emotional citizen. London:  Palgrave Macmillan.

84  Tosun, Aydoğan Ünal and Erdoğan Tosun Gökmen, Y.C., & Tosun, T. (2015). The RPP in the June and November 2015 elections: An evaluation on political impasse. Insight Turkey, 17(4), 193–​208. Heper, M. (2013). Islam, conservatism, and democracy in Turkey: Comparing Turgut Ozal and Recep Tayyip Erdoğan. Insight Turkey, 15(2), 141–​56. Kalaycıoğlu, E. (2011). Kulturkampf in Turkey: The constitutional referendum of 12 September 2010. South European Society and Politics, 17(1),  1–​22. —​—​—​. (2012). Political culture. In M. Heper & S. Sayarı (Eds.), The Routledge handbook of modern Turkey (pp. 171–​81). London: Routledge. Kaya, İ. (2008). Coğrafi Bağlam, Siyasal Katılım ve Parti Tercihleri:  22 Temmuz 2007 Türkiye Genel Seçimlerine Coğrafi Bir Bakış [Paper presentation]. Türkiye Coğrafyası Araştırma ve Uygulama Merkezi (TÜCAUM) V. Coğrafya Sempozyumu, Ankara (pp.  199–​212). tucaum.ankara.edu.tr/​wp-​content/​uploads/​sites/​280/​2015/​ 08/​semp5_​20.pdf Kiriş, H.M., & Köklü, T.E. (2012). Bir ‘Kıyı Şeridi’ Seçim Bölgesi Olarak Muğla’da Siyasal Gelenekler ve Yeni Yönelimler. Hukuk ve İktisat Araştırmaları Dergisi, 4(2),  31–​40. Lawson, K. (2004). Five variations on a theme: Interest aggregation by party today. In K. Lawson & T. Poguntke (Eds.) How political parties respond: Interest aggregation revisited (pp. 250–​66). London: Routledge. Lipset, S.M., & Rokkan, S. (1967). Cleavage structures, party systems and voter alignments:  An introduction. In S.M. Lipset & S. Rokkan (Eds.). Party Systems and Voter Alignments:  Cross-​National Perspectives (pp. 1–​64). New  York:  The Free Press. Mardin, Ş. (1973). Center-​periphery relations:  A key to Turkish politics? Daedalus, 102(1), 169–​90. Martis, K.C., Kovacs, Z., Kovacs, D., & Peter, S. (1992). The geography of the 1990 Hungarian parliamentary elections. Political Geography, 11(3), 283–​305. Ӧniş, Z. (1997). The political economy of Islamic resurgence in Turkey: The rise of the welfare party in perspective. Third World Quarterly, 18(4), 743–​66. Socio-​ Economic Development Ranking of Provinces and Regions 2011 (SEGE). (2013). Ankara: Ministry of Development, General Directorate of Regional Development and Structural Compliance. Toprak, B. (2005). Islam and democracy in Turkey. Turkish Studies, 6(2), 167–​86. Tosun, T. (2010). Hayırcı Sahil Şeridinin Siyasal Dinamikleri. Sabah. www.sabah.com. tr/​perspektif/​2010/​10/​09/​hayirci_​sahil_​seridinin_​siyasal_​dinamikleri Tosun, T., & Erdoğan Tosun, G. (2015). 7 Haziran ve RPP: Olanaklar ve Olasılıklar Üzerine Bir Değerlendirme. Sosyal Demokrat Dergi. www.sosyaldemokratdergi. org/​tanju-​tosun-​gulgun-​tosun-​7-​haziran-​ve-​RPP-​olanaklar-​ve-​olasiliklar-​uzerine-​ bir-​degerlendirme/​ Walks, R.A. (2006). The causes of city-​suburban political polarization? A Canadian case study. Annals of the Association of American Geographers, 96(2), 390–​414. Wuthrich, F.M. (2013). An essential center–​ periphery electoral cleavage and the Turkish party system. International Journal of Middle East Studies, 45(4), 751–​73. —​—​—​. (2015). National Elections in Turkey: People, Politics, and the Party System. Syracuse, NY: Syracuse University Press.

5  The regional and national campaign discourse of the PDP and its predecessors in Turkey’s general elections, 2002–​2018 Hüseyin Alptekin

Introduction Halkların Demokratik Partisi (The People’s Democratic Party, the PDP) is not only one of the major political parties in Turkey’s parliament but it also stands at the center of controversial debates on Turkey’s Kurdish question. This chapter analyzes the PDP and its predecessors’ discourses in their electoral campaigns since 2002. It is natural for political parties to show variation in their campaign messages in different regions and at different periods. This chapter examines such variation in the case of the PDP. Overall, this chapter argues that the PDP and its predecessors have embraced a moderate, universal rights-​based, pro-​Turkey (Türkiyeci) discourse in their campaigns in western Turkish districts, while at the same time embracing an identity-​based and, at times, radical discourse in their campaigns in eastern Turkish districts. Hence, the PDP has mostly performed as a social democratic party in its campaigns in western Turkey but almost exclusively as an ethnic and, at times, pro-​PKK,1 party in its campaigns in the east. Besides this interregional difference, the movement has also shown longitudinal variation as its strictly class-​ based socialist rhetoric has increasingly embraced a new-​leftist tone with the greater inclusion of gender-​and environment-​related issues. Such supplemental issues have occasionally received greater emphasis than the ethnic character of the party itself, especially in western Turkey. Another temporal change in the party’s discourse has been its increasingly anti-​Erdoğan rhetoric in parallel to the rising influence of Erdoğan (as the PM from 2003 to 2014 and president since then) in Turkish politics. The PDP, unlike other major parties in contemporary Turkish politics, has passed through multiple political stances under different labels. The party continuously rebranded itself between the 1990s and 2010s because of the various court cases and bans it faced. While this rebranding process was initially out of necessity due to the party closures, the movement has begun to change its party label once again in the years after its last judicial closure in 2009 (see Table 5.1).

86  Hüseyin Alptekin Table 5.1 HDP’S predecessors in Turkish politics Party

Opening date

Closure date

Method of closure

Halkın Emek Partisi June 7, 1990 (People’s Labor Party, the PLP) Özgürlük ve Eşitlik June 25, 1992 Partisi (Freedom and Equality Party, the FEP) Özgürlük ve Demokrasi October 19, Partisi (Freedom 1992 and Democracy Party, the FDP)

July 14, 1993

Constitutional court ban

July 2, 1992

Self-​termination and joining to the PLP

April 30, 1993 Self-​termination with the expectation of banning in the ongoing constitutional court trial June 16, 1994 Constitutional court ban

Demokrasi Partisi (Democracy Party, the DeP) Halkın Demokrasi Partisi (People’s Democracy Party, the PDemP)

May 7, 1993

Demokratik Halk Partisi (Democratic People’s Party, the DPP)

October 24, 1997

Demokratik Toplum Partisi (Democratic Society Party, the DSP) Barış ve Demokrasi Partisi (Peace and Democracy Party, the PaDP) Halkların Demokratik Partisi (Peoples’ Democratic Party, the PDP)

November 9, 2005

Constitutional court ban and a 5-​year long ban from politics for 46 people including the party leader Murat Bozlak November 19, Self-​termination and 2005 joining to the DSP with the expectation of banning in the ongoing constitutional court trial December 11, Constitutional court ban 2009

May 2, 2008

July 11, 2014

October 15, 2012

Functioning as of January 2021

May 11, 1994

March 13, 2003

Self-​termination and joining to the PDP

Notes: The electoral results for these parties in the general elections are as follows: PDeP 4.2 percent in 1995, PDemP 4.8 percent in 1999, DPP 6.2 percent in 2002, DSP (as a bloc of independent candidates) 3.8 percent in 2007, PaDP (as a bloc of independent candidates) 5.7 percent in 2011, PDP 13.1 percent in 2015 (June 7), PDP 10.8 percent in 2015 (November 1), and PDP 11.7 percent in 2018. Source: Adapted and revised from Alptekin (2014).

Regional and national campaign discourse  87 The party has further diversified its organizational structure under different names in different regions of the country. It ran under the banner of the PDP in the western districts of Turkey and as Barış ve Demokrasi Partisi (the Peace and Democracy Party, the PaDP) in the eastern parts in the 2014 local elections. Later on, the municipalities won by the PaDP embraced a new title, Demokratik Bölgeler Partisi (the Democratic Regions Party, DRP) in the four months following the 2014 local elections. On top of name changes at the national scale and regional rebranding, the movement also ran for a number of elections not under its party banner but as a bloc of independent candidates. The rationale for this decision was to get around the 10 percent electoral threshold, which is applied to political parties, but not to independent candidates, in Turkey’s general elections. All these name changes and tactical choices to run with independent candidates make it difficult to trace the party’s electoral history in a linear sense. Hence, we are forced, in a way, to refer to the movement using different names in this chapter as it has changed many times. This chapter traces the movement’s trajectory of electoral campaigns for parliamentary elections by examining the individual political parties belonging to the movement. Hence, local elections do not form part of our analysis. The reason for this choice is that the party’s discourse on national politics can be seen in full force in the parliamentary elections rather than the local ones. As for each parliamentary election, the chapter examines the candidate profiles, the lyrics of the marches written for the campaigns, and the speeches delivered in the two biggest rallies of the movement, respectively, in İstanbul and Diyarbakır. While the İstanbul rallies represent the movement’s campaign discourse for the entire country, the Diyarbakır rallies reflect the regional (heavily Kurdish-​populated eastern and southeastern districts) tone of the campaigns. Once enough data is presented for each individual election, then the overall tone of the campaign is analyzed based on the content of such electoral speeches, banners, and marches. This chapter traces the movement’s trajectory from the 2002 parliamentary elections to 2018 elections, and hence starts with Demokratik Halk Partisi (the Democratic People’s Party, DPP). Then, it analyzes Demokratik Toplum Partisi (the Democratic Society Party, the DSP) in the 2007 elections. The next elections to be examined are those in 2011, alongside the Peace and Democracy Party’s electoral strategy and discourse in this same election. The study tackles two general elections in 2015, the June 7 general election and the November 1 snap election. The last election to be analyzed in the chapter is the June 24 election of 2018. The party banner under analysis from the June 2015 to June 2018 elections is the PDP. Hence, the chapter will, at times, simply refer to the movement as the PDP and/​or the PDP’s predecessors. This chapter is concluded with a brief summary of and future projections for the movement.

88  Hüseyin Alptekin

The Democratic People’s Party’s November 3, 2002 election discourse and message content Halkların Demokrasi Partisi (the People’s Democracy Party, the PDemP), Emek Partisi (the Labor Party, the LP), and Sosyalist Demokrasi Parti (the Socialist Democracy Party) decided to run for the 2002 elections under the umbrella of Demokratik Halklar Partisi (the Democratic People’s Party, the DPP) banner. The party was successful in the sense that it increased the 4.75  percent share of the vote of its predecessor (PDemP) in the 1999 elections to 6.22 percent in 2002, but this was still not enough to carry the party to the parliament due to the 10 percent electoral threshold. Since the party’s vote share was below the electoral threshold required for representation in the national parliament, the movement did not win any seats despite leading the elections in the southeastern region of Turkey. The DemPP was quite successful not only in the deep southeast but it also showed a significant increase in its number of votes as well as its spread across the provinces of the east and southeast (Çarkoğlu, 2002). The DemPP came first in 13 cities in the eastern and southeastern regions of Turkey. The party’s main message in the western districts was democracy and working class rights, while it stressed a discourse of peace and the chosen trauma (a selective narrative of the 1990s based on the injustices committed by state authorities) in the eastern districts.2 The DemPP’s 2002 electoral campaign was unprecedentedly live and even aggressive in its campaign for some analysts (Kaya, 2002). The party increased its tone with more ambitious electoral goals and an increased visibility in the electoral campaign arena. While the DemPP deputy candidate Mustafa Yalçıner claimed that the party’s İstanbul rally hosted tens of thousands (Evrensel, 2002a), its most significant rally at the national scale took place in Adana on September 27, 2002 just a week before the elections. It was, for many, the most crowded DemPP rally during the campaign, with around 60 thousand participants (Evrensel, 2002a). Due to the party’s electoral alliance with smaller left-​wing parties, the party discourse was filled with class-​based themes with a strong emphasis on working class rights. Another dominant theme was “peace”, with a particular reference to the then-​frozen clashes between the PKK and Turkey. Amongst the banners that decorated the rally square in Adana, the more noteworthy included “DemPP for improving agricultural production and cattle farming”, “DemPP for full democracy and rule of law”, and “Budget for education, not for war” (Evrensel, 2002a).3 Besides the banners that were placed by the party, the noteworthy banners raised by its supporters were “Votes for DemPP for freedom and brotherhood, for the brotherhood of Turkish, Kurdish, Arab, Circassian peoples”, “Votes for DemPP to say no more torture”, “No more war, from Edirne to Batman”, “Workers, public officers, small shop owners, votes for DemPP for peace”.4 The PDemP (the key constituent of the DemPP

Regional and national campaign discourse  89 bloc) leader, Murat Bozlak, and other speakers also underlined the theme of “peace” and empowering women and workers in the Adana rally. The party held many other rallies during the campaign process. The rallies in predominantly Kurdish-​populated districts prioritized “peace” and the chosen trauma. The DemPP chairman, Mehmet Abbasoğlu, also gave many positive messages in his speech during the Diyarbakır rally ranging from promises to join the EU to the improvements in his party’s campaign process compared to the previous election (Hürriyet, 2002). The marches that were sang in these rallies were the ones that narrated the chosen trauma, which was based on the events of the 1980s and 1990s (e.g., forced village evacuations, torture, unsolved murders). One such song was called “We were burnt in flames” (alev alev yandık), or with its popularly known name, “We resisted under torture” (işkencelerde direndik), and was repeatedly sang in a number of marches and rallies in Turkey’s southeastern districts. The march narrates the chosen trauma with lines such as the following: Some of us were hurt, some died Some were buried in unknown places Some did not give in to fascism Some submitted to fascism and joy5 The above march not only attempted to mobilize the ethnic base around the theme of chosen trauma, but also put forward a dichotomy: the Kurds who “suffered” for the right cause, and the Kurds who sold their cause in exchange for joy and pleasure. This was a clear message from the so-​called “good” Kurds to the “bad” Kurds, namely those who did not vote for the movement. While the rallies in southeast Turkey stressed the narrative of chosen trauma, a class-​based leftist jargon was used in more multiethnic constituencies. One reason for this difference was the effect of workers’ unions in the relatively more industrialized western districts. The workers’ unions mobilized their masses in multiethnic constituencies where the DemPP would have had a shortage of supporters to gather for its rallies.6 Kamu Emekçileri Sendikaları Konfederasyonu (the Confederation of Public Employees’ Unions, the KESK) was the leading syndicate in the campaign process. Besides, the prevailing messages were aimed at erasing the separatist image of the party in the western districts. The banner “Long live the democratic republic” (Yaşasın demokratik cumhuriyet) (Evrensel, 2002c) in the Aydın rally was quite striking in this regard. Osman Özçelik, in his speech during the rally, even claimed that his party had gathered a bigger crowd than so-​called Turkish parties in the city of Aydın, which is only sparsely populated by Kurds. The overall campaign in the western cities was held with the stated goal of “democratizing Turkey”. Hence, we saw the starting of a pattern with the 2002 elections: class-​based discourse and pro-​integration remarks in the western districts versus the ethno-​political remarks in the east.

90  Hüseyin Alptekin

The Democratic Society Party’s July 22, 2007 election discourse and message content After the success of its electoral alliance with marginal leftist parties in the 2002 general election, the DPP attempted a similar strategy in the 2004 municipal elections but failed to reach the broader leftist electorate. After its earlier party organizations were banned by the Constitutional Court or abolished themselves due to fear of expected legal sanctions, on top of seeing that it did not have any realistic expectation of passing the 10 percent national threshold, the newly found DSP came up with a new electoral strategy. The DSP decided to run for the 2007 elections with a list of independent candidates, again in collaboration with leftist political organizations including the LP, the SDP, and the Özgürlük ve Dayanışma Partisi (Freedom and Solidarity Party, the FSP). The bloc was named the Thousand Hope Candidates (Bin Umut Adayları). The list of nominated individuals included not only the famous leftist figures but also some names that were familiar in Islamist circles: Ayhan Bilgen, the president of the Islamist Association for Human Rights and Solidarity for the Oppressed (İnsan Hakları ve Mazlumlar İçin Dayanışma Derneği, Mazlum-​Der), and Ahmet İnan, professor of theology at Dicle University, were two such names. This was a signal that the movement was no longer going to seek to receive pro-​PKK and secular leftist votes alone but was also attempting to expand its support base. The party’s election song was sang both in Turkish and Kurdish, not surprisingly more in Turkish in the western regions and Kurdish in the eastern. The song’s main theme was democracy, and it called on the electorate to “choose independents for democracy”. Democracy was the most frequently repeated word throughout the song, while it also made direct references to equality and freedom. Another striking feature of the song was the attempt to target the broader masses that were not traditionally targeted by the movement. Its lyrics included lines such as “in order to be united”, “we have an oath to Turkey”, “in İzmir, İstanbul, Kars with Ahmet Türk”.7 All these excerpts indicated a departure from the earlier separatist discourse of the movement. The party held its İstanbul rally in Kağıthane Hasbahçe Square on July 15 and named it the “we have an oath to Turkey” (Türkiye’ye sözümüz var) rally (Sabah, 2007). One of the symbolic names of the movement, Leyla Zana, said that “This city is the address for diversity and freedom” in her rally speech. The rally was also accompanied by a number of singers appealing to the broader Turkish audience. The rally, the name of which speaks for itself, focused on Turkey’s democracy problem as a whole and stressed that the DSP-​ backed independents were a guarantee of Turkey’s territorial integrity. The bloc’s other significant event, the Diyarbakır meeting, took place in Istasyon Square on July 19, four days after the İstanbul rally. The Diyarbakır rally focused more on the Kurdish question and naturally used themes and songs in the Kurdish language. While the İstanbul rally focused on the bloc’s oath to Turkey, the Diyarbakır rally placed a much greater emphasis on the Kurdish

Regional and national campaign discourse  91 question. The noteworthy banners at the latter included messages such as “We are going to the parliament for our oath to the people” and “Kurds will bring the true democracy to this country”.8 Despite this well-​crafted strategy and these carefully picked and geographically sensitive messages, the bloc came first only in six cities, seven fewer than the 2002 results. Nevertheless, while the party lost votes, it gained 22 seats in parliament with its strategy of running with independent candidates. The strategy of running with a list of independent candidates rather than the party banner freed the movement from the restrictions of the 10 percent national threshold.

The Peace and Democracy Party’s June 12, 2011 election discourse and message content After the DSP was banned by the constitutional court in 2009, the movement’s new address became the PaDP. The party was led by a younger figure, Selahattin Demirtaş, who is the brother of Nurettin Demirtaş, the former chairman of the DSP. The PaDP decided to follow the DSP’s strategy of running with independent candidates for the parliamentary elections. This time the bloc was called the Labor, Freedom, and Democracy Bloc (Emek, Özgürlük ve Demokrasi Bloğu) and nominated 61 candidates in 39 cities. The bloc did not nominate anyone in cities such as Eskişehir, Hatay, Ordu, Rize, Sakarya, Samsun, Tekirdağ, and Yalova, namely where it did not gain any significant proportion of the vote in the 2007 election. The movement’s list broadened its spectrum considerably and included more people from mainstream, socialist, and Islamist circles. While the campaign in the west stressed democracy, it, once again, stressed identity in the east. The campaign also had an overarching theme this time, that of “democratic autonomy”. Democratic autonomy was proposed as a form of territorial autonomy for Kurds, although it was claimed that the PaDP wanted it for everybody in the entire country. The Democratic Society Congress, an offshoot of the PaDP led by Ahmet Türk and Aysel Tuğluk, also self-​declared democratic autonomy on July 14, about one month after the elections. The campaign song of the PaDP ended with the line “to democratic autonomous, free tomorrows”.9 The idea was not new to the movement and had been first devised by the PKK’s jailed leader, Abdullah Öcalan, who was inspired by the writings of American political philosopher and social theorist, Murray Bookchin.10 Soon after the elections, the party’s co-​chair Selahattin Demirtaş’s remarks about Abdullah Öcalan started a year-​long discussion and became a key element of his trial and imprisonment. In the southeastern district of Kızıltepe, Demirtaş said to the crowd in the rally that “They will erect a statue of chief Apo”, referring to Abdullah Öcalan. He further referred to Öcalan as the “Kurdish people’s leader” and claimed that Öcalan’s poster should be hung in Kurdistan (Hürriyet Daily News, 2012). The bloc organized its İstanbul rally in Kağıthane a week before the June elections. While the messages in the rally were quite diverse, ranging from

92  Hüseyin Alptekin LGBT rights to the empowerment of women and the working class, democratic autonomy was also emphasized in this rally. Amongst the speakers, the İstanbul candidate Sırrı Süreya Önder’s speech was quite bold as it threatened the police with “breaking their hands” if they touched the donations or the posters in the campaign offices of the bloc during their raids (UİD-​DER,  2011). The bloc held its Diyarbakır rally on May 14, one month before the elections. The prominent speakers at the rally were Altan Tan, a prominent Islamist, Şerafettin Elçi, a former minister of Kurdish origin from Cumhuriyet Halk Partisi (the Republican People’s Party, the RPP), and Leyla Zana, an iconic name for the movement due to her long years in prison, and whose imprisonment is still thought to be unfair in Turkish public opinion. While the Kurdish character of the rally was naturally more dominant than it was at the İstanbul rally (to be held three weeks later), the rally paid particular attention to religious discourse. Altan Tan spoke of the continuing problems with the legacy of headscarf bans, which had already been removed by the ruling Adalet ve Kalkınma Partisi (Justice and Development Party, the JDP). Leyla Zana made references to Sheikh Said, the leader of the religious order that started the first widespread Kurdish rebellion of the republican period. Other speakers, such as Şerafettin Elçi and Emine Ayna, also made strong references to Kurdish identity. Hence, as the Diyarbakır rally showed, with the 2011 elections, Islam and Kurdishness were no longer mutually exclusive but complementary. From now on, the movement raised its Islam-​friendly tone quite strongly until the PKK found itself clashing with fundamentalist terrorist organizations such as the Islamic State of Iraq and Syria (ISIS) and other moderate Islamist rebel groups (including the groups affiliated with the Turkey-​backed Free Syrian Army) in Syria in 2014.

The People’s Democratic Party’s June 7, 2015 election discourse and message content The new name of the party in the 2015 elections was the PDP. This time the reason for rebranding was not a constitutional court ban but was instead purely the party’s own choice. The movement abolished the PaDP and united under the banner of the PDP following the 2014 local elections, where the movement had run under the PDP banner in the western cities and PaDP banner in the eastern cities. The municipalities gained by the PaDP in the 2014 local elections later embraced a new title, Demokratik Bölgeler Partisi (the Democratic Regions Party, the DRP). In fact, a lot had changed between the 2011 election and the one in 2015. Practically in late 2012, and officially starting in early 2013, Turkey entered a new conflict resolution process with regard to its decades-​long terror problem with the PKK. The process lasted about three years until the summer of 2015. The parliamentary elections of June 2015 were held when the indirect negotiations between the PKK and the Turkish government had just ended.11 Nevertheless, despite the failure of the negotiations to bring a final resolution,

Regional and national campaign discourse  93 the PKK’s ceasefire (declared in March 2013) was still in place and Turkey was not carrying out counterterrorism operations against the PKK either. While the PDP organized multiple rallies in İstanbul, the big event was hosted on May 30 in Kazlıçeşme Square (Sendika, 2015). The rally was organized on the same day with the JDP’s İstanbul rally. The PDP’s event was named “The Great Humanity Rally”.12 The paramount slogan of the rally was “We will demolish the thresholds and the palaces”, in which the term “palace” was a reference to the recently elected (in 2014) President Erdoğan’s new residence and office, the Beştepe Presidential Complex. Besides the old slogans and messages, there were two new points in this rally: glorification of the “Kobane resistance”,13 and anti-​Erdoğanism. The PKK’s Syrian offshoot, PYD,14 gained unprecedented popularity during the ISIS siege of the Syrian city of Kobane (Ayn al Arab). The PDP, in its electoral campaign, benefited from the increasing global popularity of the PYD and showed its solidarity with the organization. The second theme of the campaign in general, and the İstanbul rally in particular, was anti-​Erdoğanism. The content and tone of the speeches was more about how to stop President Erdoğan rather than promoting the PDP’s own agenda. The speakers argued that Erdoğan wanted to end the resolution process (known as Çözüm Süreci) with the PKK. Another feature of this election was that of the rising popularity of the party leader, Selahattin Demirtaş. Demirtaş was celebrated not only by his mass base but also by mainstream media outlets. His rise later became a concern for PKK leadership in Qandil, Iraq, who thought that Demirtaş’s increasing popularity was undeserved as they were the ones who had paid—​and indeed were still paying—​the price for the Kurdish cause. While Demirtaş gained popularity in different segments in Turkey, the PKK leadership watched his rise with caution and at times with critical responses. Duran Kalkan, one of the members of the PKK’s executive committee, once said that Demirtaş’s calls for the PKK to end its violence was baseless. He said that “What have they [the PDP] achieved so far?” (Cumhuriyet, 2015), “The PDP was not creative and successful in politics”, and hence implied that the PDP had no right to make such calls on them (Sabah, 2015). While the party’s Diyarbakır rally was expected to have the same pro-​PYD and anti-​Erdoğan discourse, together with an emphasis on how to “demolish” the electoral threshold, a terror attack hit the event. The rally was organized on June 5 and started as planned, but ISIS militants targeted the meeting, with the incident costing five lives and seeing many more injured.15 Overall, the PDP’s primary electoral strategy for the June elections was embodied in the phrase strategic voting.16 The party called on the people to vote for the PDP even if it was not their first choice. According to this strategy, it was more rational for opposition voters to vote for the PDP because it was the only way to prevent the JDP from gaining another electoral victory and thus gaining a majority in parliament. The party advocated the idea that its failure to pass the 10  percent electoral threshold would guarantee another term for the JDP’s single party rule. The campaign was quite successful as

94  Hüseyin Alptekin the party actually did subsequently pass the 10 percent threshold for the first time in movement’s history. In this historical victory for the movement, the PDP won 13.1 percent of the vote and 80 seats in the 550-​seat parliament. This outcome showed that the PDP was evolving toward becoming a party for Turkey in its entirety. If the evolution had more time, it would probably have led to the PDP becoming an autonomous (from the PKK), effective party, appealing to the entire country. Nevertheless, such a scenario would upset the PKK leadership in Qandil, which saw the PDP as nothing more than a necessary instrument to infiltrate Turkey’s legal political landscape. This was the main reason why the PKK leadership embraced an increasingly critical tone toward the PDP and subsequently tried to register the PDP’s electoral success as its own.17 Since this election, the PDP has found itself quite restrained, if not paralyzed, in the midst of the PKK’s new terror campaign and Turkey’s tightening counterterrorism measures, which also targeted PDP politicians.

The PDP’s November 1, 2015 election discourse and message content The June 7 elections did not lead to the formation of an enduring government, either in the form of a coalition or a single party government. As the parliamentary parties could not agree on the formation of a coalition government, the president called for a snap election. The mood was considerably different this time, just five months after the June election. The country, which had become used to having long-​lasting stable governments during the past 13 years, found itself in turmoil. The PKK blamed the Turkish government for the ISIS terror attack in Suruç, Şanlıurfa on July 20 and started its attacks in Turkey once again. After the organization killed a soldier and three police officers in three attacks in the timespan between July 20 and 23, the government ordered widespread counterterrorism operations.18 In the midst of the terror attacks and Turkey’s counterterrorism operations in the summer of 2015, it became clear that the November elections would be held under the shadow of terrorism and stiff security measures. The PDP had won an unprecedented victory in the June elections with over 13 percent of the votes and 80 seats in parliament. Nevertheless, the spiral of violence in the summer of 2015 diminished the PDP’s role and reminded the resentful voters of the JDP that the PKK was the real actor of the movement and it had no intention of losing its central position to the PDP. Accordingly, the PDP lost 2.5 points and thus only just met the parliamentary threshold with its 10.56 percent share of the vote.19 Its number of seats also decreased from 80 to 59. In a nutshell, the PDP lost one million votes and 21 seats in the five-​month period between the June to November elections. During the campaign period, the PDP-​led Peace Bloc organized a peace rally on August 9.  The rally also included some Republican People’s Party (RPP) deputies, but they left when Abdullah Öcalan posters were opened during the rally (T24, 2015). The party’s co-​chairwoman Figen Yüksekdağ’s

Regional and national campaign discourse  95 statement “We lean our back on YPJ, YPG, and PYD and we do not see any harm in mentioning and defending it” (Sabah, 2015) became another controversial remark given that the PYD is a designated terrorist organization in Turkey. The real blow to the campaign process came with an ISIS attack, which killed 109 participants in the PDP’s rally in Ankara on October 10. Following this, the PDP cancelled ten major rallies including the ones in İstanbul and Diyarbakır. The party continued its campaign for the November elections mostly through social media instruments as it had lost its ears in the mainstream media, which had covered the PDP quite extensively before the June 7 elections. The overall theme of the party both at the national and regional (southeastern Turkey) scale was to blame the Turkish government for the end of the ceasefire process. In a nutshell, the PDP’s messages in its November elections campaign were lost under the noise of bullets. The loss of support for the PDP, compared to the June elections, led political analysts to question the party’s ability to maintain its increased popularity of the preceding couple of years.20

The PDP’s June 24, 2018 election discourse and message content The June 24 elections had two ballots, one for the parliamentary and the other for the presidential elections. The party nominated its jailed co-​chairman, Selahattin Demirtaş, as its presidential candidate. The PDP’s election campaign for the June 24 elections adopted a negative agenda, centering on the incumbent JDP’s policies targeting the PDP. While this was a new theme for the party, separating this campaign from those of the 2015 elections, the party continued its 2015 discourse regarding strategic voting. The PDP stressed that (a) if the party receives over 10 percent of the vote and hence enters parliament, the JDP would automatically lose its majority of the seats; and (b) if the presidential elections do not have a winner in the first round, Erdoğan would lose it to the contestant in the run-​off stage. The projection claimed that the JDP would lose both in the parliamentary and presidential elections only if the PDP received enough support for the parliamentary elections and, in return, supported the contestant against Erdoğan in the presidential elections. This strategic voting-​centered campaign was a revised replication of the 2015 campaigns, but the PDP achieved a certain novelty through its increasing tone about the pressure it faced. The party officials, in their speeches, stressed the arrested members21 of the party, including the former co-​chairs Selahattin Demirtaş and Figen Yüksekdağ along with seven other MPs and many mayors. Demirtaş was the key figure amongst these, who continued to tweet from behind bars and twice appeared on national TV channels to deliver his campaign speeches. Demirtaş was arrested for his speeches, allegedly, in favor of the PKK and its jailed leader, Öcalan, as well as calling for “resistance” during the Kobane events in Turkey, in which the participants lynched people for being tied to certain religious groups.22

96  Hüseyin Alptekin Demirtaş ran for the presidential spot from behind bars as the PDP’s candidate and appeared twice on the state-​owned television TRT channel as a legal right offered to all presidential candidates. His first appearance adopted a theme of working for peace and justice. Demirtaş’s speech was powerful in terms of delivering its message to those who already supported him, but not impressive enough to sway those who were critical of him. He began his speech by questioning the rationale and fairness of his imprisonment. Nevertheless, among the 15 people Demirtaş cited in his speech, nine were figures who had instigated armed campaigns in one way or another. Amongst those, the most prominent was Sheikh Said, who started an ethnic separatist rebellion against Turkey in 1925. He was the most prominent Kurdish rebellion leader of the early republican era. The next person in the list was Seyit Rıza, a leader of another rebellion in the early republican era. He also rebelled, failed, and was executed. Then Demirtaş paid tribute to Deniz Gezmiş, Hüseyin İnan, and Yusuf Aslan, three revolutionary figures from the late 1960s and early 1970s. They instigated armed attempts to start revolutionary guerrilla warfare against the Turkish state and were themselves captured and executed in 1972. The next person Demirtaş cited was the most troublesome for many. Mahir Çayan, another leader of a radical leftist organization, is also a documented killer who died in combat with Turkish security forces. Perhaps the most ironic part of Demirtaş’s speech was about Yılmaz Güney, a famous actor and director. Güney injured one man and killed another. What makes this comparison more tragic is that Güney then fled the country. In a nutshell, the PDP’s most important campaign moment was Demirtaş’s televised speech, but it was not convincing enough in terms of the message delivered to say, “I am a man of peace”.23 The PDP’s rallies in İstanbul and Diyarbakır were quite crowded. Unlike in earlier rallies, the gap between the national and regional discourse was narrower. The significant themes of the İstanbul rally included imprisoned party officials, the electoral threshold and strategic voting to end the JDP’s rule, and the authoritarianism charges against Erdoğan’s rule (Evrensel, 2018). The Diyarbakır rally repeated these themes but also placed a stronger emphasis on the Kurdish question. In particular, it focused on the incumbent JDP’s switch toward nationalism and the speakers made an association between this switch and the violations of human rights in the 1990s under the disguise of counterterrorism and nationalism (Hürriyet Daily News, 2018a). A  summary of the similarities and differences in the PDP’s electoral campaign discourses from one election to another and from the national scale to the regional can be seen in Table 5.2.

Conclusions This chapter has discussed the PDP and its predecessors’ discourses in the electoral campaigns in six parliamentary elections from 2002 to 2018. The analysis shows that on the national scale, the movement had a looser grip of

Regional and national campaign discourse  97 Table 5.2 The prevalent regional and national discourses of the HDP and its predecessors in Turkey’s general elections (2002–​2018) Election

HDP’s regional discourse

HDP’s national discourse

2002

Identity politics; Chosen trauma Embrace of Islam

Leftist (class-​based) discourse

2007 2011 2015, June

Stronger stress on Islam and its compatibility with Kurdish nationalism Anti-​ISIS and pro-​PYD with a stress on “the Rojava revolution”

2015, Nov

Equating Erdoğan with ISIS

2018

The JDP’s switch toward nationalism and to the ‘old Turkey’

Stress on Turkey’s unity and democracy Democratic autonomy Pro-​PYD discourse with a focus on Kobane; Demirtaş’s personality traits; Strategic voting to counterbalance the JDP Blaming Erdogan for the rise of violence between Turkey and the PKK; Blaming Erdogan for ISIS attacks Strategic voting against Erdoğan and the JDP

Note: Each election campaign also preserved the themes of the previous election but once these themes entered the party’s discourse, they are not repeatedly shown on the table.

the socialist ideology and has gradually embraced “the new leftist” tone. With this shift, we saw an increasingly looser stress on the class-​based analysis of Turkish society and greater stress on its marginalized identity groups. The specific arguments for such groups were never the main theme of any electoral campaign, but there were many references to such groups, from religious minorities to the LGBT community, that could be read between the lines of the campaign speeches and messages. The overarching theme of the electoral campaign discourse in the national scale also embraced an increasingly anti-​ Erdoğan rhetoric in parallel to Erdoğan’s increasing grip on power in national politics. This discourse served the party’s main strategy of appealing to the left-​inclining Erdoğan critics within the mainstream Turkish electorate. The movement’s regional discourse, on the other hand, has never lost its focus on the Kurdish question. The party was less of a leftist but more of an ethnic party in eastern Turkey. Overall, the campaigns in the eastern districts did not make much effort to distinguish the party from the PKK. That being said, there was still longitudinal variation in the party’s electoral discourse at the regional level. The movement, at times (particularly from the 2007 elections to the onset of the June 2015 elections), initiated a religious opening to appeal to conservative Kurdish voters. Nevertheless, a tone of secular tendency has been on the rise as Islamist groups in Syria have been portrayed as the “enemy

98  Hüseyin Alptekin of Kurds”. While ISIS was an obvious case in this point, other Islamist groups and the Turkish-​backed Free Syrian Army were also portrayed in the same manner. In a nutshell, we can conclude that the movement is a left-​inclining party on the national scale and a purely ethnic party on the regional scale, although some variation can be observed over time.

Notes 1 The Kurdistan Workers’ Party (Partiya Karkaren Kurdistan, the PKK) is a designated terrorist organization listed as such by several states and organizations including Turkey, the United States, the United Kingdom, Germany, the European Union, the North Atlantic Treaty Organization (NATO), etc. The PDP’s pro-​ PKK discourse in eastern Turkey is discussed further in the following pages with a number of direct quotes from the party’s leading figures. 2 For a discussion of chosen trauma, see Volkan (2001). 3 In original Turkish, respectively: “Tarımsal üretim ve hayvancılığı geliştirmek için DEHAP”, “Tam demokrasi ve hukuk devleti için DEHAP”, “Savaşa değil eğitime bütçe”. 4 “Özgürlük ve kardeşlik için Türk, Kürt, Arap, Çerkez, halklarının kardeşliği için, oylar DEHAP’a”, “İşkenceye son demek için oylar DEHAP’a”, “Edirne’den Batman’a, savaşa son”, “İşçiler, memurlar, esnaflar barış için oylar DEHAP’a”. See Evrensel (2002a). 5 Kimimiz yaralı, kimimiz öldük, Kimimiz bilinmez yere gömüldük, Kimimiz faşizme boyun eğmedik, Kimimiz faşizme zevke kul olduk, Tuzaklara kandık işkencelerde. For a clip of this song being sung at the Şırnak rally, see the music video on YouTube (Rustemcudi, 2012). 6 For a comparison of Gaziantep and Viranşehir rallies, see Evrensel (2002c). 7 In Turkish: “Birlik olmak için”, “Türkiye’ye sözümüz var”, “Ahmet Türk ile İzmir’de, İstanbul’da, Kars’ta”. Ahmet Türk was the chairman of the DSP, the last banned party of the movement. 8 In Turkish: “Halka sözümüz için Meclis’e gidiyoruz”, “Bu ülkeye gerçek demokrasiyi Kürtler getirecek”. For details of the rally, see Evrensel (2007). 9 In Turkish original: “Demokratik özerk, özgür yarınlara”. 10 Öcalan also wrote a book named Democratic Autonomy. For a discussion on the transition from Leninist centralism to democratic confederalism and the debates on the ambiguities in this transition, see Leezenberg (2016). 11 For a discussion of the negotiations and its aftermath, see Alptekin and Köse (2018). 12 In Turkish, “Büyük İnsanlık Mitingi”. 13 The YPG-​led forces, the PKK’s Syrian offshoot, was the main ground force against the siege carried out by ISIS, which lasted from September 2014 to March 2015. While the US-​led international coalition airstrikes targeted the ISIS forces during the siege, Turkey also allowed reinforcements from the Free Syrian Army and Iraqi Kurdistan to enter Kobane to defend the city. 14 The PYD was shown to be the PKK’s Syrian offshoot by official American documents but reference to such had been removed from all such documentation by 2017 (Hürriyet Daily News, 2018b).

Regional and national campaign discourse  99 15 The number of casualties were two at first but three more passed away amongst the wounded. For an initial report on the explosion, see Al Jazeera (2015). 16 For a brief discussion of strategic voting in the context of the PDP, see Alptekin (2018b, p. 224). 17 Duran Kalkan, a prominent leader of the PKK in Qandil, stated that the PDP was not creative and successful. He also disparaged the PDP’s calls to the PKK with his sarcastic question “what have they [the PDP] achieved to make calls on us?” (Cumhuriyet, 2015). 18 The killings took place on three separate occasions. A  soldier was killed in Adıyaman on July 20, two police officers were killed in Ceylanpınar, Şanlıurfa on July 22, and a traffic officer was killed in Yenişehir, Diyarbakır on July 23. 19 For the official results of the elections, see YSK (2015). 20 For examples, see Grigoriadis (2016) and Köse (2015). 21 For a list of the arrested MPs, see T24 (2017). 22 Amongst such speeches, the most striking was “Resistance is writing a heroic history” (direniş kahramanca bir tarih yazıyor) with reference to the rebellion and counterterrorism operations in a number of districts in southeastern Turkey, in which 249 state security personnel lost their lives in combat. The quote was the title of Demirtaş’s interview given to Özgür Gündem on December 30, 2015 (Özgür Gündem, 2015). The newspaper was also closed by a court decision on August 16, 2016. 23 For a detailed analysis of the speech, see Alptekin (2018a).

Bibliography Al Jazeera. (2015). HDP’nin Diyarbakır miting alanında patlama:  2 ölü 100 yaralı. www.aljazeera.com.tr/​haber/​HDPnin-​diyarbakir-​miting-​alaninda-patlama-​2olu-​100-​yarali Alptekin, H. (2014). Explaining ethnopolitical mobilization:  ethnic incorporation and mobilization patterns in Bulgaria, Cyprus, Turkey, and beyond [Unpublished Dissertation]. University of Texas at Austin. —​—​—​. (2018a). Demirtaş’s Armed Heroes for in his Speech for Peace. The New Turkey. thenewturkey.org/​demirtass-​armed-​heroes-​in-​his-​speech-​for-​peace/​ —​—​—​. (2018b). Kurdish votes in the June 24, 2018 elections: An analysis of electoral results in Turkey’s eastern cities. Insight Turkey, 20(4), 211–​29. Alptekin, H., & Köse, T. (2018). AK Parti’nin Kürt politikası. In N. Miş, & A. Aslan, AK Parti’nin 15 Yılı: Siyaset (pp. 315–​38). SETA. Çarkoğlu, A. (2002). Turkey’s November 2002 elections:  A new beginning? Middle East Review of International Affairs, 6(4),  30–​41. Cumhuriyet. (2015). PKK’li Duran Kalkan’dan Demirtaş’a:  Neyi başardın da çağrı yapıyorsun? www.cumhuriyet.com.tr/​h aber/​dunya/​352435/​P KK_​li_​D uran_​ Kalkan_dan_Demirtas_​a_​_​Neyi_​basardin_​da_​cagri_​yapiyorsun_​.html Daily Sabah. (2015). HDP co-​chair Figen Yüksekdağ being investigated for terror propaganda. www.dailysabah.com/​politics/​2015/​07/​31/​PDP-​co-​chair-​figen-yukse kdag-​being-​investigated-​for-​terror-​propaganda Evrensel. (2002a). Adana’daki en görkemli miting. www.evrensel.net/​haber/​135909/​ adana-​daki-​en-​gorkemli-​miting —​—​—​. (2002b). DEHAP binlerle miting yaptı. www.evrensel.net/​haber/​136042/​ DPP-​binlerle-​miting-​yapti

100  Hüseyin Alptekin —​—​—​. (2007). Diyarbakır Bin Umut dedi. www.evrensel.net/​haber/​240812/​diyarbakirbin-​umut-​dedi —​—​—​. (2018). HDP’den ‘Büyük İstanbul Mitingi’:  Barajlar bizi engelleyemez. www.evrensel.net/​haber/​355086/​PDPden-​buyuk-​istanbul-​mitingi-​barajlar-​bizi-​ engelleyemez Grigoriadis, I.N. (2016). The peoples’ democratic party (HDP) and the 2015 elections. Turkish Studies, 17(1),  39–​46. Hürriyet. (2002). DEHAP mitinginde İmralı’ya selam. www.hurriyet.com.tr: https://​ www.hurriyet.com.tr/​gundem/​dehap-​mitinginde-​imraliya-​selam-​107160 —​ —​ —​ . (2012). BDP wants Öcalan statue. www.hurriyetdailynews.com:  www. hurriyetdailynews.com/​PaDP-​wants-​ocalan-​statue-​34609 —​ —​ —​ . (2018a). Pervin Buldan:  Kayyumları, Erdoğan’ın Kıraathanelerine Göndereceğiz. www.hurriyet.com.tr/​pervin-​buldan-​kayyumlari-erdoganinkiraathan-40873186 —​—​—​. (2018b). US removes PYD/​ YPG from annual terrorism report. www. hurriyetdailynews.com/​us-​removes-​pyd-​ypg-​from-​annual-​terrorism-​report-​137035 Kaya, K. (2002). Turkey’s elections: What impact for Eurasia? Biweekly briefing from Central Asia–​Caucasus Analyst, 6. Köse, T. (2015). Identity dynamics of the June and November 2015 elections of Turkey: Kurds, Alevis and conservative nationalists. Insight Turkey, 17(4), 105–​23. Leezenberg, M. (2016). The ambiguities of democratic autonomy:  the Kurdish movement in Turkey and Rojava. Southeast European and Black Sea Studies, 16(4), 671–​90. Özgür Gündem. (2015). Direniş Kahramanca Bir Tarih Yazıyor. (S. Demirtaş, Interviewer). Rustemcudi. (2012). Yusuf Uğur-​Direndik İşkencelerde (Şırnak 2002 DEHAP Mitingi). www.youtube.com/​watch?v= g5N211NImVE Sabah. (2007). Zana’dan İstanbul’un Bağımsızlarına Destek. arsiv.sabah.com.tr/​2007/​ 07/​16/​haber,0820C699E15B4A0093E32A4BDDB0C461.html —​—​—​. (2015). PKK’dan Demirtaş’a Çok Sert Yanıt. www.sabah.com.tr/​gundem/​ 2015/​08/​25/​pkkdan-​demirtasa-​cok-​sert-​yanit Sendika. (2015). HDP’nin İstanbul mitinginde yüz binler buluştu: “Barajları da sarayları da yıkarız”. sendika62.org:  sendika62.org/​2015/​05/​HDPnin-​istanbul-​mitinginde-​ yuz-​binler-​bulustu-​barajlari-​da-​saraylari-​da-​yikariz-​267473/​ T24. (2015). Barış mitingi konuşulduğu gibi gerçekleşmedi, Öcalan posteri çıkınca CHP’li vekiller alanı terk etti. t24.com.tr:  t24.com.tr/​haber/​baris-​mitingi-​konusuldugu-​ gibi-​gerceklesmedi-​ocalan-​posteri-​cikinca-​chpli-​vekiller-​alani-​terk-​etti,306069 —​—​—​. (2017). 9 vekil, 68 belediye başkanı tutuklu; ‘HDP’ operasyonu nasıl başladı, bugüne dek neler yaşandı? t24.com.tr:  t24.com.tr/​haber/​9-​vekil-​68-​belediye-​ baskani-​tutuklu-​PDP-​operasyonu-​nasil-​basladi-​bugune-​dek-​neler-​yasandi,510710 UİD-​DER. (2011). On Binler İstanbul’da Emek, Demokrasi ve Özgürlük İçin Buluştu. uidder.org:  uidder.org/​binler_​istanbul%E2%80%99da_​emek_​demokrasi_​ve_​ ozgurluk_​icin_​bulustu.htm Volkan, V.D. (2001). Transgenerational transmissions and chosen traumas: An aspect of large-​group identity. Group Analysis, 34(1),  79–​97. YSK-​Yüksek Seçim Kurulu. (2015). 1 Kasım 2015 Genel Seçimleri Kesin Sonuçları. ysk.gov.tr:  https://​www.ysk.gov.tr/​doc/​dosyalar/​docs/​Milletvekili/​1Kasim2015/​ KesinSecimSonuclari/​96-​A.pdf

6  “Where are they doing politics?” Women’s cooperatives as sites of constellations of power Selin Akyüz and Kürşat Çınar

Introduction Doing politics is a contested term. The practices within the act of politics have been deeply analyzed by scholars of political science, sociology, psychology, history, and anthropology. Especially when one focuses on the questions of who is doing politics and where and when s/​he is doing politics, the potential answers closely concern feminist scholars. The discussions may extend from the subject of doing politics and the spatial construction of the politics to the issue of performativity. Butler’s canonical term of performativity links performances, gender, and “doing”. For Butler, while gender is a “repetition of acts (…) which are discontinuous”, the audience and the actors themselves “come to believe and to perform in the mode of belief ” (1990, p. 4). Masculinity and femininity and all related binary depictions are also associated with certain practices, discourses, and spaces. Papart and Zalewski (2008, p. 10) claim that some performances of gender produce and also are produced, and that “metaphors act as motors of discourse which work to frame and naturalize masculinist assumptions”. That is, metaphors and assumptions reinforce the relation between masculinity and power, and hence naturalize masculine domination. This naturalization also contributes “to the ongoing struggle to maintain gender hierarchies” (Zalewski & Parpart, 1998, p.  203) by which certain qualities, such as aggression, rationality, and bravery are valued as power enhancing and are defined as masculine. This understanding offers a very crystallized narrative of masculinity, or in other words, the rhetorical work of exclusion. While men are identified as the dominating one, women are recognized as the other. A discursive strategy operates in which the masculine/​feminine dichotomy is constitutive of the public/​private dichotomy (Zalewski & Parpart, 1998; Papart & Zalewski, 2008). At this point, what is critical is analyzing the practices and discourses that are circulated across multiple scales. In line with Kofman and Peake’s (1990, p. 315) apt argument “that the political should not be located in specific institutions and that it can take place at any level”, it is important to analyze networks, processes, and overarching structures that shape women’s representation in different spheres of life. Through a lens of feminist critiques, public and private spheres blur

102  Selin Akyüz and Kürşat Çınar and the multifaceted dimensions of (in)formal networks, or in a broader sense, the inhabited spaces of our lives, especially “local action gain salience” (Secor, 2001, p. 193). Women’s representation in politics, at national and at local levels, has been mostly examined in literature. Studies on women’s representation in national politics has mainly analyzed their conduct of politics in political parties and parliament, and has aimed to reveal the (in)visible obstacles that they have been facing (Tekeli, 1979; Arat, 1985; Çakır, 2005; Kasapoğlu & Özerkmen, 2013). Very important studies on the absence of women in local politics have also touched upon the “barriers” in Turkey (Alkan, 2005; Cindoglu, 2011). The works that focus on female mayors well map different gendered experiences of women at local level (Alkan, 2006; 2009; Koyuncu & Sumbas, 2016). There are also researches that question different discourses and networks that hinder women’s political agency with respect to masculinist assumptions (Alkan, 2009; Yaras, 2014; Yıldırım et  al., 2017). The patriarchal gaze has marked women in politics as wives/​mothers/​daughters and, thus, limits their agency. Women have faced a more direct patriarchal control in cities, acting as representatives of political parties’ local branches. The conceptualization of “female local politician” is actively created through traditional values and expected gendered roles, hence constructing power relationships and negotiating the boundaries of inclusion/​exclusion. According to Cindoglu and Unal (2014), women’s public credibility has been communicated through their sexuality and renders them “fragile”. Masculinist assumptions intervene to their bodies and construct “suitable” places, dictating the boundaries of where they can act within different layers of political action. As stated by feminist scholars, actions limiting mobility of women and their marginalized status “have had a reproducing impact on gender asymmetry via the social space of localities, are obstacles for, not only a more democratic society and realizing the principle values of local self-​government, but also women’s empowerment and full citizenship” (Alkan, 2009, pp. 48–​9). Other than thinking within the fixed boundaries and/​ or dualities, questioning alternative sites has the potential to open new avenues of discussions. Hence, women’s cooperatives not being commercial business model per se—​operating on principles intrinsic to conducting business such as negotiation, bargaining, book keeping, contract signing, etc.—​provide a fertile ground to analyze women’s engagement with local authorities as both recipient and provider. Especially clientelist relationships that they have been developing within local networks and authorities and their decisive role in structuring informal politics provide venues of political participation beyond voting or representation in formal politics. Feminist political geography in general, and electoral geography in particular, interprets these sites and analyzes how women spatialize different forms of being political. Mapping different layers of women’s political engagement through the “conduct” of cooperatives vis-​à-​vis local authorities is important at this point. How do the members of cooperatives negotiate with local actors? How do “party

Women’s cooperatives as sites of power  103 support” and “political capital” interact to “gain salience”? By scrutinizing these alternative venues as spaces of conducting politics, this work makes a humble step towards transgressing fixed dualities and formal/​institutional places of doing politics. In addition, so as to answer Kandiyoti’s call for “ethnographic specificity” for developing the literature of feminist electoral geography, it is also valuable to map gendered social bases of geographies of party support in Turkey (2011). Turkey offers an interesting case not only because of its low level of women’s political participation at local level but also its political space as a venue for conducting politics in its polarized atmosphere, especially the schism between the secular vs. the conservative segments of society. Due to increasing social interactions, women have been involved with different actors—​not only new spaces for doing politics but also new forms of presence constructed. Transgressing place-​based politics offer us new “topographies” (Sharp, 2007) of women where processes and networks become more crystallized. Other than being in open resistance to public and private dualism, this research focuses on spaces within, behind, and above the “designated” spaces for conducting politics. Taking the categories of public and private, formal and informal as not distinctively delineated but rather blurred into and/​ or overlapping one another would widen the feminist lens to understand masculinized local politics. So as not to limit the understanding of women’s political engagement, such a perspective would enlighten “new” dimensions for women’s participation in local politics and, also in gendered performances of politics. Upon this rationale, this work analyzes women’s cooperatives in Turkey as channeling women’s voices into municipal and national authorities. In this framework, this research will first describe the milestones of feminist electoral geography and map women’s cooperatives in Turkey. After a brief discussion on the methodology of the research, the study will discuss women as patrons and women as clients within the conduct of local politics at different levels in contemporary Turkey.

Feminist electoral geography Kofman and Peake’s (1990) argument on the absence of gender in the understanding of spatialized forms of politics and Bell’s (1995) call for integrating sexuality in political geography reformed the foci of the electoral geography discipline. The integration of different sites of power relations, oppression, and domination, has also challenged political geographers’ understanding of territoriality and provided new discussions around multiple understandings of power and the political. A diverse array of topics, namely democracy, nationalism, borders, sexuality, cultural politics, warfare, and citizenship, were encompassed by scholars, but this diversity made it difficult for the development of the discipline as it led to fragmentation (Brown & Staeheli, 2003, p. 248). Brown and Staeheli, in their important work categorize the literature around two main questions: what is political and where is it located

104  Selin Akyüz and Kürşat Çınar (2003, p. 248)? They describe three approaches to answer these questions and these three interrelated framings mark a “nuanced attention to the social construction of the spatial scale” (2003, p. 253). The first is distributional. This approach mainly evolves around the understanding of power to and power over. Distributional political geography addresses patterns and relations of inequality on the basis of “who gets what, where, when, why, and how” (2003, p. 249). While drawing attention to the unequal distribution of power, they also highlight the significant role of scale, or in other words, with a feminist lens, patriarchal spatial structures. The second approach is the antagonistic one. This perspective focuses on the conflictual dimensions of any interactions that characterize the political (2003, p. 251). The exclusive places that are tactically used by politicians in particular are in the scope of this perspective, especially in the framework of how women construct “their” politics spatially on mobilizing their “constituencies” by using different “sites” (Fincher & Panelli, 2001). The third approach is the constitutive one. This emphasizes processes within the act of politics (Brown & Staeheli, 2003, p. 252). This perspective marks the significance of “spaces of home, neighborhood, work, religious observance, and community, (…) in these spaces that political subjects may be formed and new politics created” (2003, p. 252). While the antagonistic approach develops an exclusive “we”, this one attempts to understand the ways in which an inclusive “we” can be formed. The authors give the example of Secor’s work (2003) on the politics of Islamist parties in Turkey. In her work, Secor (2003) analyzes the ways in which women claim certain spaces and, for the author, this is an ongoing statement to the city that has become visible during women’s daily lives. According to Brown and Staeheli, in this perspective, “politics is constitutive, rather than constituted” (2003, p. 253). Upon this outline, it is important to discuss power relations and spatialized forms of politics from a feminist standpoint. Political scientists, sociologists, and anthropologists have well argued that networks within institutions in and the general discourse on local politics is masculinist, and they have analyzed different constructions of those gendered networks, institutions, and the resultant discourse. However, an engagement with feminist geography is missing in the understanding of local politics. While doing politics exceeds designated spaces, the political self with her/​his gendered characteristics is actively produced in the relations. Hence, it is very important to read spatial power relations and the “politics of location” (Mountz & Hyndman, 2006, p. 455) from a feminist perspective. Here, at this point, there is an important nuance that one shall be careful about. Feminists have well documented the “less valued” and relatively invisible side of (in)formal politics:  “Women have been shown to be strongly involved in local, community-​based activities, campaigning for example around issues to do with housing, education, food, safe play areas for children, and so on and so forth” (Laurie et al., 1997, p. 132). While the gendered processes within the act of politics, as discussed by the literature, are in parallel with domestic roles of women, this understanding reproduces the public and private binary. As discussed by

Women’s cooperatives as sites of power  105 Yaras, there is a need for “different analytical tools to analyze the interaction between the gender roles in the institutions of politics and in other social institutions” (2014, p. 111). Feminist political geography has the potential to offer such tools as it focuses on the intersecting sites of informal activities and gendered political engagements. In addition, feminist electoral geography opens wider venues as it questions the interactions between political cleavages and geographies of party support. Electoral geography has traditionally focused on voting behaviors, party systems, and regionalization, or in other words, institutionalized politics (Woolstencroft, 1980; De Miguel, 2017; Kavianirad & Rasouli, 2014). Studies on electoral geography revealed the significance of scale while analyzing geographies of party support patterns, voter turnout trends, voter cleavages, and regionalization (Meleshevich, 2006; Sui & Hugill, 2004; West, 2005; Agnew, 1996; Rohla et al., 2018). However, there is lack of research that integrates the feminist lens into the analysis of these scales. Especially, other than the works on gendered electoral behavior (Secor, 2004; McClurg et  al., 2012), studies on ideological preferences and processes that manifest social cleavages may evince channels and sites that link “political culture, patriarchal structures, and patterns of women’s representation in particular settings” (Secor, 2004, p. 265). McGing (2014, p. 9) marks an important point when she writes “numerous empirical studies, mostly in political science, have tested the extent to which women ‘make a difference’ to politics. The work of electoral geographers provides insight here: if women representatives are making a difference, where are they doing?” She asks an important question as most researches focus on parliament and other institutional spaces: “are feminist political scientists exploring the wrong places?” (McGing, 2014, p.  9). The convenience of “working at home” facilitates access to networks, working with interest groups to gain salience and also funding for the local area (Cain et al., 1984 cited in McGing, 2014, p. 9). While there is research on women as voter and voted (Richardson & Freedman, 1995; Norris, 1996; Childs, 2004), there is a lack of studies on the role of party ideologies and political cleavages. In this framework, Secor’s (2004, p. 261) apt question has the potential to add significant insights: “how [do] social cleavages and geographies of party support interact to provide particular constellations of power and support in a society?”

Patron-​client relations and their significance in spatialized constructions of politics According to Cinar, clientelism is a selective, particularistic, and hierarchical mutual support system in which politicians and political parties offer material favors (such as jobs and goods) to electors in exchange for political support. It is selective and particularistic because it is targeted toward an exclusive

106  Selin Akyüz and Kürşat Çınar group of voters based on their political preferences, thereby discriminating against the remainder of the society. (2017, p. 109) Clientelism is hierarchical as it entails an asymmetrical relationship between a powerful politician (be it national or local) and his/​her “client” of material and social backing (see also, Stokes et  al., 2013). Clientelism by its nature is based on inclusion and exclusion of certain segments of society based on political inclinations and preferences. It incorporates a stick-​and-​carrot approach in which it entails both rewarding mechanisms for loyal clients (in the form of material support as well as provision of social networks at the local and national level) and punishing those who are disloyal to the political patron (in the form of obstructing business contracts or firing or eliminating “unruly” mavericks, etc.). Clientelism is prevalent in societies in which the state fails to provide citizens with even basic services (such as security, infrastructure, and health). This is especially the case for poorer citizens who cannot afford these services on their own and people who live in isolated localities beyond the state’s full coverage of basic public goods (Kitschelt & Wilkinson, 2007). To this regard, clientelism stands as a substitute for such citizens, whereby they can get access to material and social benefits from their political patrons in exchange for their political support. Linking clientelism to the realms of political geography and electoral geography, we observe that clientelism comes in different forms in different spatial settings. On the one hand, rural clientelism, which entails normative bonds of deference and loyalty to the local patron and Weberian notions of traditional authority, is usually observed in underdeveloped and secluded localities, such as Andalucía and Galicia in Spain, Eastern and Southeastern Anatolia in Turkey, and rural sections of Thailand (Cinar, 2016; Hicken, 2007). On the other hand, in urban forms of clientelism (which is more prevalently observed in the urban periphery), there is competition between different providers (aka political machines) for the provision of selective and particularistic incentives in a Downsian framework of politics in which political actors are motivated by desire for power, income, and prestige (Cinar, 2017). Urban clientelism is salient in many parts of the world, including Latin American countries such as Argentina, Brazil, and Mexico (Stokes, 2005; Magaloni, 2006), as well as in Turkey (Bayraktar & Altan, 2013; Sayarı, 2014; White, 2002). Within the realm of political and electoral geography, patron-​ client relationships act as linkages between the local and national political arenas (Sayarı, 2014, p. 657). In many settings, clientelism incorporates the horizontal social networks of mutual help and assistance as well as vertical networks of patronage relations (White, 2002). To this end, as is the case for the women in women’s cooperatives (the focus group of this chapter), people may become both the “patron” and the “client” in those relations in which a person explores clientelistic relations with local and national politicians (Bayraktar & Altan, 2013), while the same person offers clientelistic services to the members of her cooperative in the form of jobs.

Women’s cooperatives as sites of power  107 The study of clientelism at the intersection of feminist studies and political and electoral geography offers us a fertile ground for research. In light of an extensive fieldwork based on women’s cooperatives in Turkey, this chapter aims to shed light on these linkages and how women aspire to increase their share in “distributional” socioeconomics and politics, look for ways to cope with “antagonistic” interactions in public life, and “constitute” new forms of presence in expanding spatial construction of politics. While the major focus of our analysis will be the role of patron-​client relationships in shaping women’s interactions in socioeconomic and political life, we offer several other insights regarding the functions of women’s cooperatives in new dimensions of women’s presence in the public realm.

Feminist political geography at work: The case of women’s cooperatives in Turkey Women’s cooperatives are unique organizations thanks to their economic as well as social targets of attaining affluence, equality, and social progress while also (mostly) preserving a horizontal, non-​hierarchical organizational structure (Datta & Gailey, 2012; Eccarius-​Kelly, 2006; Ferguson & Kepe, 2011; Jones et  al., 2012; Majee & Hoyt, 2011). Both throughout the world and in Turkey, the main case study of this research, the scope of operations of women’s cooperatives, spans various sectors such as agriculture, handicrafts, food, and child and elderly care (Bacon, 2010; Cinar et al., 2019; Datta & Gailey, 2012; Ferguson & Kepe, 2011), According to a recent survey of women’s cooperatives in Turkey, out of 63 active cooperatives, 42 are identified as enterprise cooperatives (67 percent ) and 18 are agricultural cooperatives (29 percent), while the remaining cooperatives operate in small arts, consumers, and manufacturing (Duguid et al., 2015). This chapter will focus on the varying functions of women’s cooperatives in the creation and expansion of the spatial construction of politics at the individual and group levels, exploring the ways women engage in interactions with several political actors at the local and national levels and how this relates to their presence in the public realm. In our research, we conducted an extensive fieldwork in 2017 on women’s cooperatives throughout Turkey. A  recent World Bank report on women’s cooperatives states that there were 63 active cooperatives (Duguid et  al., 2015), though this number reduced significantly due to the emergency rule in Turkey between July 2016 and July 2018 (several women’s cooperatives were closed down, especially in Eastern and Southeastern Turkey during the emergency rule). In our fieldwork, we contacted 23 women’s cooperatives throughout Turkey, which corresponds to nearly half of the currently active cooperatives for the time being (see Figure  6.1 for our fieldwork map). In our study, we employed a non-​random, purposive, and maximum variation sampling method so as to make sure the range of people and sites from which the sample is selected is representative of the larger population geographically, socioeconomically, and politically. In our research, we conducted

newgenrtpdf

Kütahya

Burdur

ma

Nig˘de Adana

e niy

Karaman

ma

Antalya

Mersin

Bingöl

¸

Diyarbakır

ara

Aksaray

nm

Konya

Mus¸

Adıyaman

hra

Isparta

Os

Mug˘la

Denizli

Ag˘rı

Elâzig˘

Malatya

Bitlis

Van

Siirt S¸ırnak

Hakkâri

Mardin

Ka

Aydın

Tunceli

Kayseri

Ig˘dır

Erzurum

Erzincan

Sivas

s¸

Nevs¸ehir

Us¸ak Afyonkarahisar

. Izmir

Yozgat

Kırs¸ehir

Manisa

Kars Bayburt

Batman

Eskis¸ehir

Ankara

Tokat

Trabzon

Ardahan

Rize

ne

Balıkesir

Ordu

Amasya

sha

Bilecik

C¸orum

Giresun

Sak

C¸ankırı

Bolu

Artvin Samsun

mü

Bursa

C¸anakkale

Düzce

Sinop

Gü

Yalova

arya

Kocaeli

n r tı Ba k Kastamonu bü ra Ka

Kırıkkale

e irn Ed

Zonguldak

Tekirdag˘ Istanbul

Gaziantep

S¸anlıurfa

Kilis Hatay

Figure 6.1 Fieldwork map: Provinces in which interviews were conducted Note: The number of interviews conducted in each province are shown in parentheses: Ankara (5), İstanbul (3), İzmir (4), Konya (2), Eskişehir (1), Zonguldak (1), Trabzon (1), Artvin (1), Adıyaman (2), Gaziantep (1), Mardin (1), and Hatay (1). Source: Map adapted from D-​Maps.

108  Selin Akyüz and Kürşat Çınar

Kırklareli

Women’s cooperatives as sites of power  109 semi-​ structured in-​ depth interviews with several members of women’s cooperatives. In nine of the women’s cooperatives, we were able to conduct group interviews with multiple members of women’s cooperatives assuming different positions in the cooperatives. Individual in-​depth interviews lasted around an hour, while group interviews in each cooperative lasted around two to three hours. To ensure the privacy and anonymity of our respondents, we use pseudonyms throughout the article.

Women in the distributional networks of politics and socioeconomics The questions of “who gets what, where, when, why, and how” in the distributional aspects of social constructions and divisions of power (both political and resultant social and economic powers) shed light on the patriarchal spatial structures and how women deal, cope, and bargain with these structures. In our research, we have found that some women’s cooperatives aspire to create their own spaces in the public realm so as to be more independent from the constraining forces of patriarchal and political structures. We do not want a political actor or agency above us. The municipality offered us help but we did not want it. We do not want to be one of the propaganda tools of the machine politics. Some of the cooperatives work under the patronage of mayors—​the provision of subsidies, workplace, money, etc. Whenever the mayor changes, the cooperatives are affected very badly. They would not have been affected this way if they had been more impartial to politics. Same goes for the cooperatives in the Southeast Anatolia region—​they are also dependent on municipalities. They have been closed down after the emergency rule. (Solmaz, Ankara) Solmaz’s narrative well exemplifies how women are well aware of the masculinist networks that function within local politics. This also marks how women explicitly estimate the cost-​benefit of political support and also implicitly cope with antagonistic interactions between “other” cooperatives and the municipality. Our membership has reached to 150 during our establishment. We all know each other in the neighborhood. We stand together to solve our problems, on our own. We do not work simply for the sake of our members. We do whatever we can for every one of us in our vicinity. (Pınar, İstanbul) We have to be active, we have to be everywhere. You cannot do our stuff with the mindset of a civil servant. We have used our local networks here, we have done everything through cooperation [imece]. (Hülya, Artvin) Two quotes by the representatives of women’s cooperatives highlight the significance of women solidarity, especially when the need to create their own

110  Selin Akyüz and Kürşat Çınar spaces within the parameters of local politics is apparent. So as to increase their share in distributional politics, women expand their presence and make strategic use of their networks. Furthermore, other women’s cooperatives have aspired to create and use networks with national and international non-​governmental organizations (NGOs) and private firms so as to have a higher say and higher share in distributional issues. We have prepared cloth carry bags for The Union of Chambers of Turkish Engineers and Architects (TMMOB) and The Confederation of Public Employees’ Trade Unions (KESK). They supported us via these transactions. (Emek, Ankara) We have got in touch with the Foundation for the Support of Women’s Work (KEDV). They have offered us seminars and training programs. (Figen, Adıyaman) We have got in contact with universities. We have prepared for the United Nations (UN) support programs. United Nations Development Programme (UNDP) has paid for the establishment costs. We have cooperated with KEDV. Petrol Ofisi—​a private gas station company—​ has given us space for our products. (Yıldız, Gaziantep) All three examples, especially the last one, reveal how the widening of scale attributes to the growing of the women’s cooperatives. Their contact and collaboration with other associations, unions, and universities might have helped women to gain salience within their local settings. Yet, for many women’s cooperatives, the patronage networks with local and national political actors were crucial to continue their operations. I have talked to the governor and mayor regarding the production of buckwheat [a produce which this cooperative tries to process and sell]. I also know the chair of the Chamber of Commerce—​maybe he can help us buy the machinery to process buckwheat. Everything will be figured out once I complete these meetings. (Nilgün, Ankara) This narrative illustrates how some women feel the urge and need to be incorporated into patron–​client relationships so as to have a presence in distributional issues. Although the gendered networks of governor, mayor, and chair of Chamber have initially hindered women’s active participation, the way women negotiated with those local actors challenges masculinist constellations of power.

Women’s cooperatives as sites of power  111 Our mayor took the initiative in our cooperative. She chose me as the president. Thanks to her support, we finished off our debts. We have remained afloat with the help of our mayor. (Filiz, Konya) It is very hard to do and continue the business without the support of municipalities. Local politicians already support those who are close to themselves. This is a bitter reality. (Seval, Trabzon) I have talked to the mayor. The MP from Cumhuriyet Halk Partisi (the Republican People’s Party, the RPP) also came to talk to us. I have got in touch with everyone. We have talked about the importance of women’s cooperatives. Then, thanks to the support of the mayor, we have created our founding committee. It has been a great advantage for us that we have a woman mayor here. (Nilay, İzmir) The term “support” is very critical at this point, as all three narratives of representatives of cooperatives illustrate. While the financial “aid” may create another level of hierarchy between the cooperative and governing authorities, the emotional “encouragement” may turn into a distributional gain. Hence, the nuance between different forms of supports endanger not the output but the processes that shape women’s presence. Every women’s cooperative under the patronage of municipalities advance considerably. (Esin, İstanbul) The mayor always says “Elder-​sister [abla], you just say the word”. I know how things work here. The most important things are social networks, dialogues within. (Iclal, İzmir) It is important to note that many women’s cooperatives choose to be (or in many cases are obliged to be) under the tutelage and patronage of political actors, especially at the local level. The patron–​client networks that they engage with and that they create help them construct and/​or build upon the avenues of distributional domains. This is both the case for the localities ruled by the governing Adalet ve Kalkınma Partisi (the Justice and Development Party, the JDP) (as is the case for Filiz in Konya) as well as localities governed by opposition mayors (as is the case for Iclal and Nilay in İzmir). Finding presence in the distributional networks through patron–​client relationships help these women get their share in the distribution of power and economic resources. However, one should note that these patronage networks are contingent upon

112  Selin Akyüz and Kürşat Çınar the continuation of the roles of both the “patron” as well as the “client”. The mayor should continue to be present in his/​her office to provide his/​her clients with these services. The clients should also continue to “support” the standing mayor so as to have access to patronage resources and (for the presidents of women’s cooperatives) to create their own patronage networks within their cooperatives (the presidents of women’s cooperatives in a way also become “patrons” in distributional networks within the cooperatives, extending services and membership to loyal members). The issue of contingency also points out the fragile nature of these patronage networks. While these patronage networks are the life source for many women’s cooperatives, their fragility leaves the cooperatives open to potential financial stress in the cases of the removal of these clientelistic networks (e.g., due to changes in the electoral office, rifts between politicians and the women, etc.). For instance, in one of our interviews, we have been told that a seemingly very successful women’s cooperative in Nevşehir in Central Anatolia (who got the backing of the JDP mayor) is faced with the removal of their offices and work stations (all of which were provided by the municipality) when the new mayor from the opposing RPP got into the office and stated that he cannot support a cooperative once under the patronage of his biggest rival political party. As this example presents strikingly, the lack of institutionalization and financial independence and the dependency on patron–​client relationships leaves many women’s cooperatives prone to existential crisis. This also shows us the hardships members of women’s cooperatives face in their efforts to partake in distributional networks in the public realm and also “documents patterns and relations of inequality” as the distributional perspective has raised (Brown & Staeheli, 2003, p. 249).

Cooperatives as “topographies” of women’s empowerment As politics, especially at the local level, have spatialized in male dominated spaces, this has reproduced the private/​woman and public/​man dichotomy. This process has also led women to make spatial tactics to maneuver for masculinized networks and institutions. For not only attempting to position themselves to utilize resources but also for creating a “we” of effective agency, women formed groups that exclude some others. We have first accepted those from Meram (a town of Konya). Now we are also accepting those coming from out of Meram. We are accepting newcomers from towns of Konya. But we still want that our villagers produce the jam. (Filiz, Konya) This is our home. We have learnt cooperation, unification and act together here, in this place. We have learnt to say “we” not “me”. (Selma, Eskişehir)

Women’s cooperatives as sites of power  113 I am the president of the cooperative. The members are all locals. They are all living here. My network is my biggest support mechanism. I just give a call and ask help or support. It happens. We are like a family but I am their elder sister [abla]. They say me that without me they would fall apart. (Iclal, İzmir) Creating a space for certain groups of women as a strategy to protect group identity is very apparent in the narratives given above. Most of the participants of the research positioned the cooperative as a place to constitute a “we” for women, especially those who are with them, are involved in activities or are living in the same city, village, or neighborhood. This is a supportive mechanism as Iclal mentioned, especially when they engage with political actors. There are women who work under the counter as opponents. Our work is harder, we have more expenses than them. (Hulya, Artvin) On the other hand, the other dimension of creating a form of us identity has the risk to bring agonism and, implicitly, to promote hierarchical power relations among the members of the cooperative. Friends who are working here all already inclined to define themselves as someone. They are telling I have this responsibility in this cooperative or I am the president of it and this and that… When one becomes the president, she is bragging about being the president… (Nilgun, Ankara) There is no class difference in cooperatives. There shall be no boss. (Alev, Ankara) That spirit of togetherness has lost when she had joined our group. The interests were on the table. It was about sharing the cake. The fights began. (…) There were sides and polarizations. I have felt myself as hypocrite. It was like a boss who was shouting slogans on Labor Day. We were asked to leave. They started to foist. It was like a divorce between partners, two different camps. The name of the cooperative was like our child that we did not want to give up. It was like the fight over joint child custody. We failed. (Emek, Ankara) While women are aspiring to increase their share, they create relations based on patronage. The vertical networks of patronage relations, as a “political site”, trigger unequal distribution of power among women in the cooperative.

114  Selin Akyüz and Kürşat Çınar Especially the case of the cooperative in Ankara well exemplifies this. The distribution of power triggers agonism and antagonistic relations in return. I asked them not to use the title of president. It was against the feminist spirit, movement. But when woman gains such status, she wants to use all those powerful titles. They do not want to lose the chance to be called as such. I am the president, I am a board member… They are using them. (Seval, Trabzon) All of a sudden, they started to call me president. (Figen, Adiyaman) The last two quotes by the representatives of women’s cooperatives mark how a cooperative creates a space for widening woman’s engagement while it becomes a site for contestation. Women may (un)consciously reproduce the traditional masculinist hierarchies that have historically excluded them. This also marks a reproduction of an unconventional form of presence for women in a cooperative. In addition, while some participants placed women’s cooperatives as opposed to masculine groups and masculinized spatial political engagement, such examples allow us to re-​conceptualize cooperatives as configurations of different social practices. As opposed to these quotes, the narratives shared below mark another angle of the forms of political interaction. They highlight multivocality as a way of doing politics. This constitutes another form of presence for women and also a way to cope with hierarchical overarching structures. There must be common consciousness. There must be collective consciousness. We have attempted to internalize horizontal institutionalization. We are also experiencing some problems as all the other organizations are facing with, but we always find a way to solve them. (Nilay, İzmir) There is no competition, the aim is same. You are not the target of this, all of the women are. (Gunay, İzmir) We have multivocality. We are not multi-​headed. There is no hierarchy. We have already created harmony among each other. (Kevser, İzmir) Borrowing Sharp’s (2007) evocative notion on understanding the importance of connecting differences, “topographies”, at this point, it is important to evaluate a form of feminist politics that transgresses place-​based macro-​ politics. The micro-​politics of everyday life is also very important as “the webs of power and social relationships that are the basis of connection” are

Women’s cooperatives as sites of power  115 constructive for women’s political engagement, which triggers empowerment (Staeheli & Kofman, 2004, p. 6). Beyond place-​based politics and the conventional understanding of doing politics, women who are mobilizing networks and connections with varying forms encourage bonding and, in return, point to the need to develop a kind of “feminist politics that celebrates jumping scales” (Sharp, 2007, p. 382).

Women’s presence as the patron and the client The constitutive approach to the political in the literature of feminist political geography marks a critical point that underlines the processes. Politics is not about the formal settings, rather political subjects are formed and politics are created throughout the ongoing processes within the act of local politics. In parallel to the discussions given in the previous part on topographies of women’s empowerment, understanding “the ways in which an inclusive ‘we’ can be created” (Brown & Staeheli, 2003, p. 252) is significant. The participants were generally talking as “we” while they were narrating about their accomplishments. We were on TV. We are very active on social media. (Ayse, Ankara) Our friends acted in Şekerbank (a Turkish bank) commercials. (Sıdıka, Zonguldak) Women need to be more visible. We are a means to achieve this. We are agents, actors… We are working together. (Gonul, Hatay) The presence/​visibility of women and evidences of changed attitudes can also be read as spatial strategies of publicity. It is not only asserting “the personhood of women” but also doubling her roles as the patron and the client (Brown & Staeheli, 2003, p. 253). She is becoming the provider of the venues that enable women to reach resources and also the user of those resources through channels created through the conduct of local politics. Some of us consider to be a mukhtar. We are encouraging her. The people around us are noticing the change in us. (Ayse, Ankara) We are doing handcrafts to our prime minister. We gave a table cloth as a present to the president, Mr. Erdoğan. I also told to our governor that her office desk is not fancy enough for a woman, then we made one for her as well. (Filiz, Konya)

116  Selin Akyüz and Kürşat Çınar Mutually constructive relationships between women and the actors in local—​ even national—​politics uphold rules and patterns of expectations for women’s representation. In particular, the potential answers to the question of “where are women doing politics?” expand. The acts of women at cooperatives, in addition to women at home and/​or women at municipalities, offer a space for analyzing different constellations of power. The presence of women in varying social spaces emboldens her to gain salience. Women do not know how to walk, so cooperatives shall make them learnt. After being a member of the cooperative, I see some of them in Samanpazari (a bazaar in old town of Ankara), I can notice how they have changed. (Solmaz, Ankara) I have thought I have not had shells, but I had. I was always socializing with same women. I have overcome this with the help of the cooperative. (Emek, Ankara) Thanks to the cooperative, I raised my status within the family. (Derya, Mardin) The quotes by the representatives of women’s cooperatives hint some nuances of female empowerment as well. Although discussions around female empowerment are not within the scope of this research, it is important to note that women in cooperatives not only acquire some material benefits but they also use this ability effectively for positive change. Higher self-​confidence and improved status within the society and family are some of these gains shared by participants. I used to have difficulty talking. Now I give interviews on TV. (Günay, İzmir) A friend of ours told me I  am not whom I  used to be. I  can answer questions easier than before. I am more confident. I don’t need to lean on anyone now. (Nergis, İstanbul) Our goal is to reach out to women who have fears. Let her come out of her household, greet people, see a flower, keep her spirit up. Our goal is to make her chat with other people. For her to breathe… (Pinar, İstanbul) Women who constitute new ways of presence in expanding spatial constructions of politics transgress conventional understanding of not only being visible in public life but also transgressing boundaries of “doing local politics” as active agents within it.

Women’s cooperatives as sites of power  117

Conclusion This research has attempted to analyze women’s cooperatives as sites where different constellations of power are observed. Analyzing different practices and discourses throughout processes within the “management” of cooperatives offers a fertile ground to spatialize women’s political engagement in local contexts in Turkey. While feminist literature has well discussed women’s “proper” place and the resultant discourses, it has highlighted the significance of masculinist assumptions that have had an impact on the (re) construction of gender asymmetry. The characteristics of the woman politician are expected to be in tune with patriarchal control while they are acting as the representative of political parties in public life. However, other than doing politics as mayors, members of city councils, or women’s branches at the local level, scrutinizing different venues has the potential to offer “new” dimensions for gendered performances of politics. The literature on feminist geography with three interrelated perspectives—​ distributional, antagonistic, constitutive—​provides a broad analytical framework, while women’s cooperatives as sites of power relations among women and/​or vis-​à-​vis men in politics offers an alternative room. In our research, we have discussed women’s cooperatives not only as sites of gathering women but also channeling their voices in varying terms. Reading different spatial arrangements revealed women’s roles in distributional networks of politics as active agents. Cooperatives as “topographies” of a woman’s channeling different needs and aspirations for sorting “things” out embolden her to have different roles. Based on the narratives, it is observed that women have maneuvered between being the client of the service(s) or patron of the whole process(es). In addition, borrowing from the work of electoral geographers, discussing where women are doing politics is critical, especially when integrating the role of party support in the form of close relations with the powerful actors discloses horizontal and vertical networks in local politics. Women in cooperatives engage with different issues ranging from manufacturing, management of money, human resources, and marketing. Hence, they have expanded spatial constructions of local politics while masculinist assumptions have already designed her “proper” place. This work is a humble step that sheds light on unconventional spaces where women negotiate the boundaries of inclusion/​exclusion. It is important to analyze blurred and/​or overlapped spaces with a critical feminist lens to reveal different ways in which performances are experienced and negotiated so as to map “topographies” of women not devoid of political agency.

References Agnew, J. (1996). Mapping politics:  How context counts in electoral geography. Political Geography, 15(2), 129–​46. Alkan, A. (2005). Yerel Yönetimler ve Cinsiyet: Kadınların Kentte Görünmez Varlığı. Dipnot.

118  Selin Akyüz and Kürşat Çınar —​—​—​. (2006). Belediye Kadinlara da Hizmet Eder!.. Kadin Dostu Belediye Hizmetleri: Neden, Nnasil? 2nd Ed. Ankara: Ka-​der. —​—​—​. (2009). Gendered structures of local politics in Turkey. Digest of Middle East Studies (DOMES), 18(1),  31–​57. Arat, Y. (1985). Obstacles to political careers:  Perceptions of Turkish women. International Political Science Review, 6(3), 355–​66. Bacon, C. (2010). A spot of coffee in crisis: Nicaraguan smallholder cooperatives, fair trade networks, and gendered empowerment. Latin American Perspectives, 171(37/​ 2),  50–​71. Bayraktar, U., & Altan, C. (2013). Explaining Turkish party centralism: Traditions and trends in the exclusion of local party offices in Mersin and beyond. In E. Massicard & N.F. Watts (Eds.), Negotiating Political Power in Turkey: Breaking Up the Party (pp. 17–​36). London: Routledge. Bell, D. (1995). [Screw]ing geography (censor’s version). Environment and Planning D: Society and Space, 13,  127–​31 Brown, M., & Staeheli, L. (2003). “Are we there yet?” Feminist political geographies. Gender, Place and Culture: A Journal of Feminist Geography, 10, 247–​55. Butler, Judith P.  (1990). Gender trouble:  Feminism and the subversion of identity. New York: Routledge. Çakır, S. (2005). Eşitsizliğin Parlamentodaki Tezahürü:  Türkiye’de Parlamentoda Kadın. (The Appearance of Inequality in the Parliament:  Women in Turkish Parliament). Bilim, Egitim Toplum, 3(10),  40–​9. Childs, S. (2004). A feminized style of politics? Women MPs in the House of Commons. British Journal of Politics and International Relations, 6(1),  3–​19. Cinar, K. (2016). A comparative analysis of clientelism in Greece, Spain, and Turkey: The rural–​urban divide. Contemporary Politics, 22(1),  77–​94. —​—​—​. (2017). Clientelism” in Fathali M. Moghaddam (Ed.) The Sage encyclopedia of political behavior (pp. 109–​11). Thousand Oaks, CA: Sage Publications. Cinar, K., Akyuz, S, Ugur-​Cinar, M., & Onculer-​Yayalar, E. (2019). Faces and phases of women’s empowerment:  The case of women’s cooperatives in Turkey. Social Politics. doi.org/​10.1093/​sp/​jxz032 Cindoglu, D. (2011). Headscarf ban and discrimination:  Professional headscarved women in the labor market. TESEV. Cindoglu, D., & Unal, D. (2015). Gender in political sex scandals in contemporary Turkey:  Women’s agency and the public sphere. Journal of Women, Politics & Policy, 36(4), 464–​85. Datta, P.B., & Gailey, R. (2012). Empowering women through social entrepreneurship: Case study of a women’s cooperative in India. Entrepreneurship: Theory and Practice, 36(3), 569–​87. De Miguel, C. (2017). The role of electoral geography in the territorialization of party systems. Electoral Studies, p. 67. Duguid, F., Durutaş, G., and Wodzicki, M. (2015). The current state of women’s co-​ operatives in Turkey. Washington, DC: World Bank. Eccarius-​Kelly, V. (2006). Guatemalan women’s cooperatives and state neglect. Peace Review, 18(1),  37–​43. Ferguson, H., & Kepe, T. (2011). Agricultural cooperatives and social empowerment of women: a Ugandan case study. Development in Practice, 21(3),  421–​9. Fincher, R., & Panelli R. (2001). Making space: Women’s urban and rural activism and the Australian state. Gender, Place and Culture, 8(2), 129–​48.

Women’s cooperatives as sites of power  119 Hicken, A. (2007). How effective are institutional reforms? in F.C. Schaffer (Ed.), Elections for sale:  The causes and consequences of vote buying (pp. 145–​ 60). London: Lynne Rienner. Jones, E., Smith, S., & Wills, C. (2012). Women producers and the benefits of collective forms of enterprise, Gender & Development, 20(1),  13–​32. Kasapoğlu, A., & Özerkmen, N. (2011). Gender imbalance: Women political participation in Turkey. International Journal of Women’s Study, 12, 97–​107. Kavianirad, M., & Rasouli, M. (2014). Explanation of relationship between geography and elections (electoral geography). Geopolitics Quarterly, 10(4), 94–​106. Kitschelt, H., & Wilkinson, S. (2007). Patrons, clients, and policies:  Patterns of democratic accountability and political competition. Cambridge:  Cambridge University Press. Kofman, E., & Peake, L. (1990). Into the 1990s: A gendered agenda for political geography. Political Geography Quarterly, 9(4), 313–​36. Koyuncu, L.B., & Sumbas, A. (2016). Discussing women’s representation in local politics in Turkey:  The case of female mayorship. Women’s Studies International Forum, 58,  41–​50. Laurie, N. et al. (1997). In and out of bounds and resisting boundaries: feminist geographies of space and place. In: Feminist geographies: explorations in diversity and difference (pp.112–​45). Harlow: Longman. Magaloni, B. (2006). Voting for autocracy: Hegemonic party survival and its demise in Mexico. Cambridge: Cambridge University Press. Majee, W., & Hoyt, A. (2011). Cooperatives and community development: A perspective on the use of cooperatives in development, Journal of Community Practice, 19(1),  48–​61. McClurg, S.D., Wade, M.L., & Wright-​Phillips, M.V. (2012). Gender, social networks, and voting behavior. Working Paper 64. McGing, C. (2014). Towards a feminist electoral geography. https://​dx.doi.org/​10.2139/​ ssrn.2457175 Meleshevich, A. (2006). Geographical patterns of party support in the Baltic States, Russia, and Ukraine. European Urban and Regional Studies, 13, 113–​29. Mountz, A., & Hyndman, J. (2006). Feminist approaches to the global intimate. Women’s Studies Quarterly, 34(1/​2), 446–​63. Norris, P. (1996). Mobilising the “women’s vote”: The gender-​generation gap in voting behaviour. Parliamentary Affairs, 49(2), 333–​42. Parpart, J.L., & Zalewski, M. (2008). Rethinking the man question:  Sex, gender and violence in international relations. London:  Zed Books; New  York:  Palgrave Macmillan. Richardson, L.E., Jr., & Freeman, P.K. (1995). Gender differences in constituency service among state legislators. Political Research Quarterly, 48(1), 169–​79. Rohla, R., Johnston, R., Jones, K., & Manley, D. (2018). Spatial scale and the geographical polarization of the American electorate, Political Geography, 65, 117–​22. Sayarı, S. (2014). Interdisciplinary approaches to political clientelism and patronage in Turkey. Turkish Studies, 15(4), pp. 655–​70. Secor, A. (2001). Toward a feminist counter-​geopolitics:  gender, space and Islamist politics in İstanbul. Space and Polity, 5(3), 191–​211. —​—​—​. (2003). Citizenship in the city: identity, community and rights among women migrants to İstanbul. Urban Geography, 24(2), 147–​68.

120  Selin Akyüz and Kürşat Çınar —​—​—​. (2004). Mapping women, making politics: Feminist perspectives on political geography. In: Mapping Women, Making Politics: Feminist Perspectives on Political Geography, 1st ed. New York: Routledge, 261–​72. Sharp, J. (2007). Geography and gender:  Finding feminist political geographies. Progress in Human Geography, 31(3),  381–​7. Staeheli, L., & Kofman, E. (2004). Mapping women, making politics:  Feminist perspectives on political geography. In: Mapping women, making politics: Feminist perspectives on political geography (pp. 1–​15). New York: Routledge. Stokes, S.C. (2005). Perverse accountability: A formal model of machine politics with evidence from Argentina. American Political Science Review, 99(3), 315–​25. Stokes, S.C, Dunning, T., Nazareno, M., & Busco, V. (2013). Brokers, voters, and clientelism: The puzzle of distributive politics. Cambridge: Cambridge University Press. Sui, D.Z., & Hugill P.J. (2002). A GIS-​based spatial analysis on neighborhood effects and voter turn-​out: a case study in College Station, Texas. Political Geography, 21, 159–​73. Tekeli, Ş. (1979). Türkiye’de Kadının Siyasal Hayattaki Yeri. In N. Abadan Unat (Ed.), Türk Toplumunda Kadın. Türk Sosyal Bilimler Derneği Yayınları. West, J.W., II. (2005). Regional cleavages in Turkish politics: An electoral geography of the 1999 and 2002 national elections. Political Geography, 24(4), 499–​523. White, J. (2002). Islamist mobilization in Turkey. Seattle: University of Washington Press. Woolstencroft, R. (1980). Electoral geography. International Political Science Review, 1(4), 540–​60. Yaras, S. (2014). A historical synopsis of over-​feminized local governance in neoliberal Turkey. Turkish Studies, 15(1), 100–​16. Yıldırım, S., Ucaray-​Mangitli, B., & Tas, H. (2018). Intimate politics: Strategies and practices of female Mukhtars in Turkey. British Journal of Middle Eastern Studies, 45(5), 661–​77. Zalewski, M., & Parpart, JL. (1998). The “man question” in international relations. Boulder, CO: Westview Press.

7  Election monitoring organizations in view of electoral geography Regional characteristics of volunteer networks in Turkey Ayşenur Kılıç

Introduction: “Cats in the substation” and revisiting electoral integrity? Electoral geography mainly focuses on the interaction of voting behavior, electoral systems, and geographical/​regional features. Existing studies often take a special interest on spatial-​ analytic, political-​ economic, and post-​ modern approaches to political geography (Agnew et al., 2003, p. 4). Similarly, a considerable number of electoral geography studies on Turkey (Özbudun & Tachau, 1975; Özbudun, 1976; Özbudun, 1980; Çarkoğlu, 2000; West, 2005) have adopted the spatial analytic approach while examining interactions between voting patterns and geography. Examining electoral geography within the scope of election monitoring organizations (EMOs) in this chapter, I  aim to contribute to the literature of electoral geography from an alternative angle. Focusing on the first two EMOs in Turkey—​the first nationwide association, Vote and Beyond (Oy ve Ötesi Derneği –​VB hereafter), and the first local (Ankara-​based) election monitoring social media platform, Ankara’s Votes (Ankara’nın Oyları –​AV), both of which were founded in 2014—​I mainly ask how VB, AV, and their volunteer networks represent three main regional characteristics of the entire electoral geography of the country in the 2014–​2018 time period—​ the secular west, the predominantly pro-​Kurdish southeast, and conservative nationalists. This research question is significant mainly in three respects. First, it focuses on the issue of electoral geography through the lenses of domestic election monitoring initiatives: that is to say, it makes a contribution to the existing literature on political geography. Second, in comparison to the international election monitors, less is known about the role of domestic monitors (Grömping, 2017, p. 167). This research attempts to focus on this neglected area by considering two different levels: i) at the national level, by looking at a nationwide EMO, and ii) at the local level, by looking at a metropolitan city organization. The underlying idea in the existing literature is that “a strong ‘ecology’ of domestic observers will increase the transparency of the electoral

122 Ayşenur Kılıç process and downward accountability. Thus, it is hoped to eventually help strengthen electoral integrity” (Grömping, 2017, p.  167). Finally, the comparison of the two EMOs—​the first nationwide EMO (an association), VB; and the first local and Ankara-​based EMO (a social media platform), AV—​ is significant for seeing organizational, behavioral, and financial differences between two civil societal organizations in a hybrid democracy environment. In the given electoral map of Turkey (a trio of electoral regions), the research question is significant for two reasons: first, it allows researchers to see which societal group(s) these kinds of democratic demands (of electoral integrity and election monitoring) come(s) from initially, and, possibly, why. Grasping underlying or leading factors of this kind of civil societal demand (of election screening) may provide some clues in interpreting the sociological and political factors that led to it. Secondly, existing research on electoral geography addresses the issue using largely a spatial-​analysis approach. However, political dealignment, fragmentation, volatility, or regionalization of votes cannot solely be explained through ideological reasons, group identity, or geography; newly existing identities that crosscut group identities need also to be taken into account. For example, voters’ media preferences can be considered one of these new identities that crosscut traditional group identities. Since the transition to a multiparty system in 1946, elections in Turkey have not been free from rigging, irregularities, and “discursive manipulation” (Toros & Birch, 2019) of electoral integrity by different political actors during or just after elections. However, the issue of electoral integrity did not become national in scope and rouse emotions to its current degree until 2014. Turkey had neither invited international election observers up until 20021 nor had domestic observers to screen the elections. The Freedom House scored Turkey’s electoral integrity record as relatively high (“partly free”) up to 2014 as well (Toros & Birch, 2019, p.  8). However, four electoral incidents since 2014 have cast a long shadow over electoral integrity in Turkey. An abrupt nationwide cut in power supply during the local elections of March 30, 2014 became the first incident raising doubts. Then-​Minister of Energy and Natural Resources, Taner Yıldız, announced that the power cuts in the country had stemmed from “the cats in the main electrical substation”.2 A second public debate on Turkey’s electoral integrity—​in the framework of vote counting—​resurfaced subsequent to the April 16, 2017 constitutional referendum in which the then-​head of the Supreme Election Council of Turkey (YSK, hereafter), Sadi Güven, announced on the evening of the referendum—​ while the vote counting was in process—​that unsealed and unstamped votes would be accepted as valid. This debate on electoral integrity has remained in the spotlight during the most recent legislative and presidential elections (held on June 24, 2018). Political party alignments prior to the elections also made the issue of vote counting more complex in the eyes of the electorate and new arrangements, as expressed by the YSK, left the new methods of vote counting murky and open to irregularities. Finally, inexplicable delays in the vote-​counting process after March 31, 2019, the reluctance of the ruling party

Election monitoring organizations  123 candidates to concede defeat, and the eventual cancellation of local elections in İstanbul became the last straw. Partly owing to this debate, the issue of election monitoring has emerged as a new area of civic participation in Turkey. It is important to note that the initial monitoring platforms (they were not originally associations, but social media platforms) appeared subsequent to the Gezi Park protests. These monitoring platforms explained that their foundations were based upon what they called the “spirit of Gezi”, which is interpreted as a call for participatory democracy, equality, justice, transparency, and accountability (Çelebi, 2015, p. 73; Bakıner, 2014, pp. 71–​4; Ugur-​Cinar & Gunduz-​Arabaci, 2018, p. 22). The post-​Gezi environment has carried several forums, blogs, young generation groups, social media platforms—​overall, civil societal entities—​onto the Turkish political scene. The VB was founded first as a social media platform in the months following the Gezi Protests, and it officially turned into an election monitoring association on April 24, 2014. In almost all of their media appearances, the underlying story of their foundation was that it was grounded in the “Gezi spirit”: a young group of mostly well-​educated friends who had gathered and decided to monitor the elections in Turkey.3 The first chair and spokesperson of the group was Sercan Çelebi, a foreign-​educated, young, and energetic face of the organization.4 Initially, they functioned as an İstanbul-​based organization and monitored the 2014 local elections; and since then they have continued election monitoring in subsequent elections and referendums with a growing number of volunteers across Turkey. The VB have steered the monitoring process via a free verification application, T3, where people can easily upload photographs of the official polling results and cross-​check and confirm the validity of the vote-​counting process. The AV was founded concurrently with the VB.5 Initially, the AV remained as an Ankara-​based social media platform compared to the VB’s nationwide monitoring activities, and they screened local and national elections only in Ankara province. They later enhanced their volunteer networks and began to provide election-​monitoring training not only in Ankara but also in other cities over time (in Turkey under the name of Türkiye’nin Oyları—​Votes of Turkey—​ and outside Turkey under the name of Gurbetin Oyları—​Votes Abroad). In the 2015 elections, VB and AV collaborated in monitoring elections across the entire country. The 2017 referendum was the last public vote in which the AV actively conducted election-​monitoring activity. After this election, some volunteers from the platform began participating in monitoring on behalf of political parties, while others volunteered for other EMOs.

Methodology How do the volunteer networks of the first national (VB) and of the first local (AV) EMOs in Turkey represent the three main regional characteristics of Turkey’s electoral geography? Since electoral geography and voting

124 Ayşenur Kılıç behavior are complex issues that cannot be explained solely through causal relations or via a few variables, I tried to examine this research question by focusing on media use (news preferences, media literacy, and fact-​checking behaviors), trust (in media and in election results), socio-​economic status (monthly income), volunteer networking behaviors and characteristics (such as the length of voluntary activity, the rationale behind the first involvement with the voluntary activity), information channels in volunteer networking, and the perceived political identity of the informants’ neighbors, family, and friends.6 In this research, I mainly applied semi-​structured interviewing and online survey techniques. I also analyzed reports and documents from the VB and AV as published online on their webpages, www.oyveotesi.org and www. ankaraninoylari.com. All public appearances in between the 2014–​2018 time period by the two groups on TV channels, online platforms, or in newspapers were also analyzed systematically. Complementary to the semi-​structured face-​ to-​ face interview with the founding team member of AV were two groups of online surveys conducted with a total amount of 42 volunteers from both associations. The sample selection of the survey was carried out using a snowballing technique: the researcher’s extended network (friends of friends, and their acquaintances, and so on) was used as gatekeepers to reach the anonymous participants in the survey. The survey participants mostly reside in Ankara or İstanbul, the two most populous and cosmopolitan cities in Turkey, and both of them are representative in terms of containing diverse socio-​political segments of the country.7 Data were collected from March to October 2018 through interviews, online surveys, and the two organizations’ online reports and media appearances. A face-​to-​face semi-​structured and in-​depth interview was carried out with a representative from the founding members of AV—​Burcu Akçaru. Two online surveys were designed via the online surveying platform Surveymonkey.com. The first group of surveys was filled out by volunteers for VB in the July 30–​September 30 period; while the second was conducted with AV volunteers between September 13 and October 2, 2018. Though the questions in the two surveys were almost identical, some adjustments were made due to the differences between the two organizations. Some structural errors were also corrected in the design of the second survey. Surveys were conducted on the principle of full anonymity. The survey sample was constructed based on the snowballing technique in which the networks of the researcher’s acquaintances or of their acquaintances became anonymous informants for the survey. The researcher never allowed the sharing of survey’s web links on online media platforms in order to avoid any kind of possible manipulation or misconduct. This partly explains why survey turnout was quite low.8 Another likely explanation for the low turnout is that first, the survey was conducted during the restricted political atmosphere of the “state of emergency” (2016–​2018) in Turkey. Therefore, people may understandably have been reluctant to talk about their activities,

Election monitoring organizations  125 political ideas, and the monitoring experience. Second, concurrently with the state of emergency atmosphere, an election build-​up suddenly appeared coincidentally with the conduct of this survey, and thus, some people mistakenly considered this survey as public opinion polling for the elections. The total number of surveys is 42 (29 volunteers from VB, and 13 from AV). Although the number is low in terms of the representativeness of the sample, the survey outcome provided crystal clear conclusions and answers to the main research questions and the hypotheses of the study. That is to say, despite its limitations, this study could pave the way for future studies on EMOs in Turkey. Besides, this research has potential for reliability and validity tests in further inquiries. The research started out with three main hypotheses. H1:  The predominant majority of the people who perform election-​ monitoring activity in Turkey adopt political identities or affiliations that reflect the worldviews of political opposition parties in the country [considering the argument that election-​monitoring activity per se is (politically) oppositional in character]. H2:  The politically oppositional character of the volunteers of VB and AV reflect/​represent one of the three electoral regions (the secular-​coastal west) preponderantly, and yet, this character is not heterogeneous across other alternative opposition parties/​groups. H3:  There is a negative correlation between trust in election results and the “desire” for monitoring elections. Lower trust pushes citizens to monitor elections more.

Data analysis Among the different variables the surveys attempted to measure in relation to the electoral space, some of them—​trust, media use, and political identity—​provided a picture of the informants, which substantiated the interpretation and discussion of our hypotheses. The volunteers’ preferences in media platforms and newspapers, their level of trust in election results and the news, and their political identity and place on the political spectrum were all areas where the survey sample projected clear (but not unanimous) characteristics predominantly associated with one of the three regions of Turkey’s electoral map—​that is, the western, coastal, secular region. In the following parts of the data analysis, I first explain the dimensions that reflect the most salient characteristics associated with the secular-​oriented electoral region in Turkey, and secondly, I focus on the other factors—​such as the characteristics of the volunteer networks, demographic characteristics, and perception of the political character of their social environment—​that have secondary importance in explaining the relationship between the sample and Turkey’s electoral geography.

126 Ayşenur Kılıç Media use and the news preferences of election monitoring networks Our survey sought to illustrate the links between electoral geography and media through four main questions. First, given the three main options (conventional media—​TV, radio, and the printed press; online news platforms on social media; and newsfeeds provided by the social networks of the user) the informants were asked their preferences in terms of the channels of information from which they learned the daily news. Secondly, informants were asked their preferences among both online and offline (print) media platforms and/​or national newspapers while following the daily news. Trust in news was a third question in the field of participants’ media use. Fourth, participants were asked about their fact-​checking behavior—​a follow-​up question relating to trust in the news media. The informants were also asked the level of their computer and Internet technology skills. The data indicated that the participants were prone to follow daily news via social media more than conventional media. VB volunteers and AV volunteers reflected similar patterns in news consumption behaviors: 14.2 percent of the entire sample reported that they generally followed the news through conventional media (TV, radio, or printed newspapers); 47.6 percent via online news platforms’ websites on social media; and 26.1 percent through their newsfeed on social media; while 11.9  percent said that they did not follow the news regularly. On the one hand, and in contrast to previous research data,9 which has indicated that a considerable proportion of the Turkish public largely follows the news via the conventional media (particularly TV) instead of social media,10 our research displayed that 73.7 percent of all participants get their news by social media. On the other hand, our data overlaps with previous surveys that show that, unlike the general public in Turkey, the western, coastal electorate seems to use social media more as channels of information and to learn the news.11 As Figure  7.1 indicates, when asked to choose among 25 national newspapers12 and alternative online news platforms, the volunteers from both EMOs concentrated around certain newspapers or online news platforms as their main sources of daily news. Alternative online news platforms (including those involved in citizen journalism—​such as Dokuz8haber and Ötekilerin Postası) were included among the options in order to see whether or not these platforms were among the news preferences of the informants, since most of these platforms were either established concurrently with the Gezi Park protests, while some of the already existing ones gained prominence during the protests,13 and the flow of information among protesters was carried out intensively via these channels of oppositional and alternative media platforms during the protests. The results of the surveys in this research indicated that the informants opted for the alternative and oppositional media platforms rather than the conventional national news media. They selected the mainstream news media as their sources of information only if they represented oppositional views. Among the many options, the most preferred mainstream

Election monitoring organizations  127

Akşam Bianet BirGün Cumhuriyet Diken Dokuz8haber Dünya Evrensel Gazete Duvar Haber Sol Habertürk Hürriyet Karar Medyascope.tv Milliyet Ötekilerin Postası Posta Sabah Sözcü Star Türkiye T24 Yeni Akit Yeni Çağ Yeni Şafak Diğer I dont read newspapers No answer

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

Which of the following naonal newspapers and online media plaorms would you say you follow more than others to receive news or read opinion columns, even if you follow more than one source? Please select one or more.

AV

VB

Figure 7.1 Distribution of newspapers and media platforms as sources of news followed

news media were Cumhuriyet (46.15 percent of AV volunteers; 41.38 percent of VB volunteers; 42.85  percent of the total), and Sözcü (61.54  percent of AV volunteers; 34.48  percent of VB volunteers; 42.85  percent of the total) newspapers, which predominantly act as the mouthpiece of the main oppositional party (RPP), and are political opponents of the governmental party. Among the other options, the informants never read the alleged partisan press or news platforms, which represent politically more conservative or right-​ oriented media as their sources of news. Comparing the AV volunteers’ selections with those of the VB, Yeni Çağ makes a difference, albeit to a limited extent, in terms of the informants’ rare selection of news platforms from politically right-​wing media channels. Almost all of the informants (both from AV and VB) generally prefer more critical, oppositional, and mostly leftist, libertarian, and unorthodox (non-​mainstream) media such as BirGün, Diken, Bianet, and T24; a few AV volunteers (15.38  percent) selected Yeni Çağ—​a nationalist, right-​wing newspaper—​which makes this one of the minor differences between VB and AV volunteers’ news sources preferences. Nevertheless, the number of those who selected Yeni Çağ is not representative enough to generalize for the AV volunteers.14 The news preferences of EMOs are also prominent in explaining a relatively homogeneous group identity. The news choices of AV and VB volunteers are diverse and dispersed only within mostly oppositional, critical, and left-​ oriented media platforms, with some minor exceptions. Overall, the media

128 Ayşenur Kılıç preferences of the informants accentuate mostly a homogeneous character in line with those of the western, coastal, secular-​oriented electoral region. Previous reports point at a similar panorama of the news media (newspapers and TV channels) preferences of two main groups amongst the Turkish electorate. Certain electoral groups accumulate more homogeneous preferences of newspapers and TV channels. According to data collected in 2017, a report showed that while the ruling party (JDP) electorate follow news mostly via pro-​government TV channels15—​such as ATV, AHaber, and state broadcaster TRT—​supporters of the main oppositional party (RPP) tended to watch Fox TV, Kanal D, and Halk TV, and chose Sözcü, Hürriyet, Posta, and Cumhuriyet as their newspapers.16 RPP electors select Sözcü and Cumhuriyet as their newspaper preferences at much higher rates than the Turkish electorate as a whole.17 Based on the given analysis of both previous reports, and of our data, the news preferences of our sample seem to be consistent with the preferences of Turkey’s western, coastal region electorates. Interview data supports the arguments above in that the dialogue between the media and EMOs is generally monolithic and homogeneous in character. Relations between pro-​government mainstream media and these organizations do not have a chance to flourish efficiently, but oppositional or critical media channels and actors have relatively better relations in making the voice of these organizations heard during their pre-​election public awareness campaigns in voting and vote-​counting processes. The interviewee pointed at how reluctant pro-​government media had been to host them on their TV programs or newspapers. Therefore, the interviewee stated that they used their personal networks to help them in providing access to media and gaining public visibility.18 Concordantly, survey data refers to a possible reciprocity between the oppositional media and EMOs, and in return, the volunteers’ priorities in their media preferences, and/​or vice versa. Trust in elections, and in news media The other area in which the survey results revealed cogent divisions is trust. The results indicate that trust in elections functions in tandem with trust in news media, and with the predominant rationale of voluntary activity. With a total of 42 survey participants from both organizations, only 4.7 percent (2/​ 42) reported trust in election results in Turkey vis-​à-​vis whopping percentages of 59.5 percent (25/​42) of “no trust”, and 35.7 percent (15/​42) “partial trust” in the election results in Turkey. AV and VB volunteers responded identically to the issue of trust in the news media and in elections. The results are supported by data provided by Çarkoğlu and Aytaç (2015), indicating an increasing ratio of public distrust in elections in the last decade. Comparing 2007, 2011, and 2015 polls, their data displayed a decreasing trend in trust. (Çarkoğlu & Aytaç, 2015; also see Yardımcı-​Geyikçi, 2015, p. 534; Toros & Birch, 2019, pp. 15–​6) Likewise, Uncu also notes an increasing percentage—​in

Election monitoring organizations  129

Trust in Media and Elecon Results

80.00% 70.00% 60.00% 50.00% 40.00% 30.00% 20.00% 10.00% 0.00%

No Trust

Very Lile Trust

Generally Trust

It depends

AV-Trust in Media

AV-Trust in Elecon Results

VB-Trust in Media

VB-Trust in Elecon Results

Figure 7.2 Percentage of trust in media, and trust in election results

two years, from 2015 to 2017—​in the number of those who distrust election results both countrywide and among the western coastal electorate (2018).19 Our research has two supplementary explanations on previous data. The data analysis displays a strong correlation between the level of trust in elections and the level of trust in the news as stated in Figure  7.2. Only 4.7  percent in total trust in elections and 7.1 percent express “trust in news generally”; 95.2 percent and 92.7 percent indicated some level of distrust in elections and in the news media respectively. The data illustrate a peak amount of distrust (albeit at different levels). The results also point at a correlation between distrust in elections and rationale of voluntary activity for monitoring elections. 47.6 percent of participants, which is by far the highest ratio amongst other factors explaining the rationale of voluntary activity, stated their distrust in elections as the initial reason for their decision to volunteer for election monitoring. The volunteers of both groups reported considerably lower satisfaction with the functioning of democracy: another possible explanation for high level of distrust in elections and the news.20 Political identities Our surveys display a graphic account of a relative homogeneity within both volunteer groups in terms of political identities—​as the Figure 7.3 indicates.. Among ten choices for political or religious identities in the first survey (for VB volunteers), and 12 choices in the second one (for AV volunteers),21

130 Ayşenur Kılıç

70%

Which of the idenes below best describes you?

60% 50% 40% 30% 20% 10% 0%

AV

VB

Figure 7.3 Identities

41.3  percent of VB informants and 69.2  percent of AV informants chose “Secular and Atatürkist” as the identity that represented them the best. Here, the options given also involved only “Atatürkist” (without signifying “secularism”). When the amount of those who selected “Atatürkist”, and “Secular and Atatürkist” are merged together,22 48.2  percent of VB, and 69.2  percent of AV volunteers considered Atatürkism and secularism as the identity best representative of themselves. Overall, 54.7 percent of the 42 informants reported having a secular Atatürkist identity. The political identities of the informants as expressed in the surveys are also remarkable: not even one participant selected any of the options among right-​wing political identities.23 In neither group did any participant signify him/​herself as feeling close to or having a religious identity, or otherwise being atheists. The given amounts and features illustrate propinquity between the expressed identities of VB and AV volunteers, and the coastal region electorate of Turkey.24 The other survey question to measure political affiliations asked participants to rate where they stood on a left–​right political spectrum. Since there were some structural errors in the initial survey,25 18 individual answers were not applicable, 20 participants (47.6 percent) were marked as left or center-​left, and four (9.5 percent) chose right or center-​right identities. Despite the structural error, the results imply still that both voluntary groups chiefly locate themselves on the left or center-​left of politics. Characteristics of democratic participation behaviors To measure some characteristics of the democratic participation behaviors of the volunteers, the surveys consisted of three questions on, respectively:

Election monitoring organizations  131 political party membership, civil society organization membership (other than the EMO), and the level of satisfaction in the functioning of democracy in Turkey. 14.2 percent of the entire group of informants were members of political parties as opposed to 85.7  percent with no party membership. Regarding the membership of civil society organizations—​apart from their existing membership in EMOs—​42.8  percent reported active membership in at least one group while 57.1 percent of the entire group of participants indicated no civil society membership. According to the results of this data, the volunteers of EMOs participate in civil society organizations more than they do in political party organizations. The relative dearth of membership in political party organizations may shed light on how the political sphere has interacted with the civil societal sphere in Turkey. The AV representative, Akçaru, underscored in our interview the importance of ties between civil society and the political sphere to function more effectively to observe electoral integrity.26 When asked why the AV did not monitor the last two elections (the legislative and presidential elections held on June 24, 2018)  in Turkey, Akçaru stated that the AV had completed its function and mission, and that it was now better to function within political party organizations in order to be more active and effective in bringing the outcomes of election monitoring into action because the current regulations allow only political parties to legally appeal to courts in the case of alleged misconduct in the election processes.27 As Chernykh and Svolik (2015, p. 407) pointed out, it is true that electoral observers “have a few de jure powers and almost never any de facto powers: they do not command police forces, armies…”, yet in Turkey, domestic electoral observers are dependent on political parties to even have a simple badge and observe elections. Collaboration of political parties and EMOs is also new in the Turkish civil society sphere. Although EMOs were initially established in 2014, political parties have become interested in these organizations only recently. One of our interviewees explained how the RPP had previously paid no attention to these organizations and did not cooperate because they already had their own party network to monitor elections.28 The interviewee concluded that the RPP have only recently comprehended the importance of these organizations. In the June 24, 2018 elections, the RPP not only enhanced their relations with the existing EMOs, but it also brought its own efforts onto the agenda by constructing the adilsecim.net platform, which aimed at providing an alternative channel of information to the state news agency Anadolu Ajansı (Anatolian Agency, AA) by compiling first-​hand unofficial election results through their own collective efforts. Though the RPP’s new platform proved technically inadequate to meet expectations, nascent cooperation between political parties and EMOs, and public debate on electoral integrity exhibit a widening interest in election monitoring, as also evidenced by the June 24, 2018 and March 2019 local elections. Nonetheless, both political parties’ cooperation with EMOs as well as fierce public discussions—​mostly on social media platforms—​on electoral integrity in the pre-​and post-​election time

132 Ayşenur Kılıç periods in June 24, 2018 elections exhibit increasing interest (both of the public and of political parties) in election screening processes. Having mentioned interactions between civil society organizations and political parties, the Liberal Democrat Party (LDP)—​a very small party which lacks both the membership and votes to be able to enter the national assembly29—​makes an interesting example in terms of indicating how even a small political party can serve a critical democratic function. From my own interviews and analysis of social media posts of the election monitoring, volunteers showed that the LDP was one of the principal parties in the 2014 and 2015 elections to provide the obligatory badges for election monitoring volunteers. At a time when the ruling political party was very reluctant to provide badges for the AV and VB volunteers and the main opposition party had not yet comprehended the potential significance of collaborating with these organizations, the amount of badges the LDP provided were far above the amount of votes the party has received at any time—​a source of black humor for the political party. The other and equally interesting element in terms of the development of the civil societal sphere in Turkey is a new form of treating civil society organizations as similar to commercial companies. The data show that the VB reflects a more liberal–​capitalistic understanding of their civil society activities, which appears to be a new interpretation in the Turkish civil society sphere. In contrast to conventional interpretations of civil society as acting on a voluntary basis, in this new form of civil society understanding, it is possible to argue that the VB consider themselves social entrepreneurs, and proceed with a more economically liberal approach to civil society organizations.30 Considering the two paradigms of civic participation (Wells, 2013, pp. 4–​ 5; Bennett et al., 2009, p. 106), the dutiful citizen and the actualizing citizen, the actualizing citizen model might be explicative of the new approach of the newly emerging EMOs (particularly of the VB’s) in Turkey. In comparison to the previous decades’ dutiful citizenship model, the actualizing citizen type is less guided by the authoritative form of membership (of civil society organizations) or identification with the social group, but is more driven by individual interests and trusted networks. (Wells, 2013, p.  5) In the former model (the dutiful citizen), the citizen appears to be a consumer of information while the people receive the news as provided by mass media and political leaders. In the actualizing era, however, the citizen does not have a purely consumerist orientation to information, as the digital era allows them to find alternative sources of information and further participation opportunities (Wells, 2013, p. 4). With its volunteers mostly hailing from the younger generation, its approach as content-​provider rather than merely as consumers of information, and the organization’s peer-​to-​peer relationships with citizens through digital media instead of a more authoritative style, the VB makes a fitting example of changing forms of civic participation, thus differing from the previous dutiful type.31

Election monitoring organizations  133 The VB’s liberal–​capitalistic style as a new civil society organization is, therefore, compatible with an actualizing citizen participation model since the initial rationale of the VB volunteers and members are more determined by their lifestyles, and less determined by “a sense of duty to participate politically in conventional ways such as voting or following issues in the news, while displaying a greater inclination to embrace issues that connect to lifestyle values, ranging from moral concerns to environmental quality” (Bennett, 1998 as cited in Bennett et al., 2009, p. 106). Characteristics of volunteering networks and activities To characterize some features of voluntary networks among EMOs, these surveys explored the duration of voluntary activity, party selection behaviors at the time of monitoring activity, the adequacy of the volunteer training by the organization, their literacy in the T3 application for validating monitoring results, their first involvement with the organization, the main rationale (initial reason) for engaging in volunteer activity, the spread of voluntary membership within the social environment of the informant, the number of elections in which they have participated in monitoring activity (election-​based), the belief in the effectiveness of voluntary activity, and their social bonds with other members and founders of the organization. A significant result, which is compatible with the results on media use, is about the first encounter of the volunteer with the EMO (Where did you hear about this organization first?) Among 42 participants, 57.1 percent said that they initially heard of it through their social media networks, 21.4 percent in offline (everyday, not-​computer-​mediated) life, 11.9  percent said they heard of it on the Internet by coincidence, and only 2.3 percent from the conventional media. The results are explanatory in terms of how influential social media networks are in the spreading of civil society activity and participatory behavior. The percentages are also compatible with previous research on the relatively higher social media usage of the electorate of the western, coastal regions of Turkey in comparison to that of the voters of the central Anatolian regions. In terms of how the voluntary membership of the EMOs spread, the data provides significant details. In total, 54.7  percent of volunteers stated that they led 1–​6 persons from their social networks to become volunteers for the organization, 16.6 percent said more than 20, and 11.9 percent said 7–​20 persons. A characteristic difference between VB and AV here is that the larger organization (VB) volunteers mostly recruited between 1–​6, while the smaller organization (AV) volunteers functioned more effectively, with more of its respondents recruiting more than 20 people. (VB/​AV:  18/​5 for 1–​6 persons, and 1/​6 for more than 20 persons). Thus, smaller groups seem to recruit greater numbers of members. In terms of technological “literacy”, 41.3 percent of VB volunteers stated that they had been able to learn the use of the T3 application by their own

134 Ayşenur Kılıç effort, and 51.7 percent said they had learnt through VB trainings. The survey data show that 31 percent of VB volunteers stated that they came across a situation on the election day that required information outside the scope of what the VB trainings had provided them. Despite these partial structural limitations, 78.5 percent believed in the effectiveness of their voluntary activity for vote counting processes, while 21.4 percent were more pessimistic about their “added value” for elections processes and said they did not believe they had made a contribution. Regional (demographic) characteristics Based on questions regarding the place of residence, place of birth, the features of the place of residence (either metropolitan city, city, or town), and the socio-​economic status (monthly individual income) of the respondents, two issues came to the fore. Firstly, a large majority of the participants lived in metropolitan cities (89.9  percent of VB participants, 100  percent of AV participants). For place of birth, the data was representative of the seven main regions of Turkey. Secondly, in terms of socio-​economic status, they seem to represent the urban middle or upper-​middle class of Turkish society based on the information they provided on their monthly income. Although these results cannot be generalized specifically to the western coastal electorate, previous survey data illustrates that in comparison to the entire society, the secular western-​coastal regions have a higher ratio in terms of: i) people who live in metropolitan cities, and ii) their proportion of middle-​class people within the general distribution of diverse social classes.32 Participants’ perceptions of political behavior in their social environment Even though the neighborhood effect is a contentious area within the field of electoral geography (Johnston et  al., 2004, pp.  369–​71), the researcher included some questions that were not directly related to neighborhood effects, but rather focused on the participants’ perceptions about the political party preferences of their social environments. The questions inquired about whether or not the participants’ perceived neighbors’ and relatives’ voting behaviors were considered close to their own. The questions were deemed important since, as Ellis argued, not only geographical but also “contextual effects are significant and that the process of day-​to-​day personal communication and social interaction in the community—​or in a variety of communities—​influences voting patterns” (2011, p. 747). The perception of a neighborhood’s political preferences may be one of these contextual effects. The Figure 7.4 below indicates that according to data obtained, 40.4 percent of informants (of a total of 42 participants)33 thought that most of their neighbors held similar political views to their own, while 28.5 percent thought that very few of their neighbors had similar political views, and 16.6 percent reported that some of their neighbors entertain similar political views. Only

Election monitoring organizations  135

AV-Predic­ons of VB-Predic­ons of Neighbors' Neighbors' Poli­cal Opinions Poli­cal Opinions

AV-Predic­ons of VB-Predic­ons of How Family, How Family, Rela­ves, and Rela­ves, and Friends Vote Friends Vote

Which of the statements best describes your predic­ons about your neighbors/family and close friends? No answer Almost all similar Majority Some Very lile No answer Almost all Majority Some Very lile No answer Majority Some Very Few Lile No answer Majority Some Very Few Lile 0.00% 10.00% 20.00% 30.00% 40.00% 50.00% 60.00% 70.00% 80.00% 90.00%

Similarity Figure 7.4 Perceived similarity of neighbors, family, and friends’ views

4.7 percent of the informants said that none of their neighbors had similar views to them. Nine percent said that they had no idea or don’t know about the situation. When asked a similar question about their close social circle (family, relatives, and close friends) and this circle’s political party choices, 73.8 percent of total participants stated that they know/​think that the closest people to them voted for the same party as themselves. Taking this quite high rate into consideration, we can ask if there could be a possible relation between the closed circuit of news sources one gets, and the belief in the homogeneity of political ideas or choices around them. As we discussed previously, the data results on the media use characteristics indicated that 74.7 percent of all participants receive news by social media—​either from their close circle’s newsfeeds or from other various channels on the Internet. Since there were many that expressed a belief in homogeneous political views within their social environment, and this environment is one of the main sources of getting news and information, a further inquiry might reveal further interesting connections. Likewise, Ugur-​Cinar and Gunduz-​Arabaci provide similar results in terms of homogeneity and the like-​mindedness of these

136 Ayşenur Kılıç people in their research on public forums in the western coastal region, which most conspicuously appeared in the post-​Gezi environment, that these public forums “encourage more socialization with like-​minded people’ in terms of their ideological leanings” (2018, pp. 20–​2).

Conclusion The present research attempted to test three interacting hypotheses about the nature of the first EMOs in Turkey. The data on these organizations and their volunteers point out that the sample in this study represented identical characteristics and behaviors with the western coastal region’s electorate (generally speaking, RPP voters). In terms of H1, that is election-​monitoring activity per se is—​politically—​ oppositional in character, this is not a conclusion that we can derive from the data, but from the existing literature. Even though we do not directly test H1 by measuring different components of being “politically oppositional”, the peculiarities of the media preferences of the sample have provided considerable explanations which show how the pro-​government partisan news media preferences of the informants remain quite low in comparison to their choices of critical, oppositional, and alternative media channels. The data is explanative for our first two hypotheses, respectively that predominant amount of the people who perform election-​monitoring activity in Turkey adopt the political identities or affiliations that reflect the politically oppositional parties’ worldviews in the country, and the politically oppositional character of the volunteers of VB and AV mostly reflect/​represent one of the three electoral regions, that is, the secular-​coastal west, preponderantly, and this character does not apply to all alternative oppositional parties/​groups. With regard to the second hypothesis, the collected data does not provide adequate information to signify particularistic differences among different oppositional political parties. The results on political identities display that the sample is more likely to be made up of RPP voters, and less likely to be NAP,34 GP, or PDP voters, for instance, since they identify themselves through predominantly Atatürkist, secular, and leftist identities rather than by right-​wing, nationalist, or conservative characteristics in both survey questions about political identities and on a left–​right scale. Admittedly, it would not be surprising to find supporters of EMOs largely align with the largest opposition party. However, as Chernykh and Svolik (2015) argued, a ruling party may have strong incentive to support impartial third parties, such as electoral observers, in order to foreclose possibility of post-​election violence when it estimates that an impending election will be a close one. Besides, advancing electoral integrity may not be the only motivation for political parties to support election monitoring civil society organizations. A  ruling party in particular, as President Hugo Chavez is claimed to have done in Venezuela through “units for electoral battle” with the help of his Circulos Bolivarianos (Lean, 2012, p.  99), may seemingly support domestic

Election monitoring organizations  137 electoral observation by establishing or helping to establish outright partisan or allegedly nonpartisan electoral observation groups in order to emasculate existing election monitoring groups. In brief, whether to increase the legitimacy of its claims to victory in a close election or to provide a populist counterweight to nonpartisan civil election monitoring groups, ruling parties may at times support domestic electoral observation as well. It may therefore still be surprising that VB and AV did not draw from JDP supporters. It is surprising, however, that VB and AV seem to have failed to draw volunteers/​members from other oppositional parties such as the PDP, NAP, and GP. Though our data can neither provide an answer to this question nor claim being representative enough to generalize, some possible explanations can still be offered. The absence of volunteers/​members from opposition parties outside the RPP could initially result from the organizational incapacity of VB and AV. As one VB survey informant stated, when VB was first established it was unable to organize itself well in regions most in need of electoral observation. After all, unlike Alianza Civica (National Alliance), a conglomeration of various civil society groups coming together for the purpose monitoring elections in Mexico during a transition period in the mid-​ 1990s, which had been able to organize in 31 states and the federal district in Mexico in the 1994 elections, (Lean, 2012, pp. 66–​7); VB and AV had not become “national” in the true sense of the word. VB and AV’s networks in the east and southeastern regions remained far behind their well-​organized networks in İstanbul, Ankara, and nearby places. The second explanation could be related to the supply side:  differing manners and practices of civic participation awareness in different regions of the country may have a direct influence over the volume and coloring in the supply of members/​volunteers. Could it be that civic participation awareness is not as high among nationalists, conservatives, pro-​Kurdish, or the rural area electorate in comparison to the politically oppositional, principally urban, and upper-​middle classes? Or is it possible that conservative nationalist voters interpret the practice of voting in elections more as a form of “duty of citizens to their state” than a democratic participatory act? These are questions that require further studies. Last but not least, unwarranted trust by political parties in the ability of their own organizational capacity may be the culprit. If the PDP’s party organization, for instance, feels strong enough in the southeast to conduct election observation on their own, the PDP-​electorate may not have felt any need to volunteer to screen ballot boxes and monitor vote counting processes outside the party organization. More, further research may reveal ideational distance and/​or (intended or unintended) communication barriers between some of these political parties and these EMOs. Finally, the third hypothesis (H3:  There is a negative correlation between trust in election results and the “desire” for monitoring elections. Lower trust urges citizens to monitor elections more) is compatible with the research results. The data identifies a considerable correlation between the level of distrust

138 Ayşenur Kılıç in elections and the level of distrust in the news. Both levels of distrust were found to be high among electoral monitoring volunteers. One of the basic reasons for many of the sample to choose to act as watchdog in the elections is related to their distrust in the elections.

Notes 1 The first invitation to OSCE for monitoring parliamentary elections in Turkey was done by Turkey’s Foreign Ministry on September 18, 2002. The Ministry invited ODIHR to consider sending an observation mission to the November 3, 2002 elections. Please see: Parliamentary elections, November 3, 2002. (2002). OSCE. www.osce.org/​odihr/​elections/​turkey/​115703. 2 Enerji Bakanı Taner Yıldız: trafoya kedi girdi. (2014). Hürriyet. www.hurriyet.com. tr/​gundem/​enerji-​bakani-​taner-​yildiz-​trafoya-​kedi-​girdi-​26128954; Seçim karanlığı (2014). Hürriyet. www.hurriyet.com.tr/​gundem/​secim-​karanligi-​26121097. 3 For some of the media appearances: Arman, A.  (2015). Turkey’s Vote and Beyond attracts even more volunteers for November 1 election. Hürriyet Daily News. www.hurriyetdailynews.com/​ turkeys-​vote-​and-​beyond-​attracts-​even-​more-​volunteers-​for-​nov-​1-​election-​ 88939. Guarding the Turkish Election. (2015). DW. www.dw.com/​en/​guarding-​the-​ turkish-​election/​a-​18809209. Bozan, İ. (2018). Oy ve Ötesi’nden Gözde Elif Soytürk: “kolluk kuvvetinin sandık başında durması bir norm haline gelirse, seçimin uygulanmasında baskı oluşturma ihtimali doğar”. medyascope.tv/​2018/​03/​31/​oy-​ve-​otesinden-​gozde-​ elif-​soyturk-​kolluk-​kuvvetinin-​sandik-​basinda-​durmasi-​bir-​norm-​haline-​ gelirse-​secimin-​uygulanmasinda-​baski-​olusturma-​ihtimali-​dogar/​. Arslan, R. (2014). Gönüllü müşahitler sandık başına gidiyor. BBC. www.bbc.com/​ turkce/​haberler/​2014/​03/​140315_​turkiye_​secim. 4 Our initial attempts to contact the current (2018) administration of the VB failed although they received our survey and interview questions via the Internet upon their request. 5 Both VB and AV have their foundational basis at an online platform, Ortak Nokta Derneği (ON—​onokta.org, www.facebook.com/​onokta.org/​ and twitter.com/​on_​ org) that was founded during the Gezi Park protests. 6 Please see Table  7.1 in Appendix I  for the themes and subthemes of the entire survey questions. 7 In order to keep the online survey length concise and to obtain a better survey turnout, the surveys did not include questions about demographics such as gender, age, or education. Nonetheless, as a result of the EMOs’ reports, our interview, and the information gatekeepers provided, the survey participants were found to be predominantly university or higher educated, middle or upper-​middle class people, metropolitan, mostly aged between 24–​39, and more than half of them were women. 8 The problem of low turnout in the online survey method is not peculiar to this study. As Brenkert-​Smith et al. (2018, p. 2) indicated in their test on online surveys, response rates varied from 3 percent to 30 percent (for pretesting) and reached at most 35 percent. (Please note that these rates are valid under the circumstance that the respondents were given two dollars as an incentive.)

Election monitoring organizations  139 9 The World Values Survey (2012) results showed that people in Turkey largely prefer TV and then newspapers as the initial sources to obtain information, while the Internet and cellular phones have a relatively lower ratio. www.worldvaluessurvey. org/​WVSDocumentationWV6.jsp. 10 Uncu, 2018, p. 60. 11 Doğan, 2018, p. 19. 12 The ratings of these mainstream media newspapers are available at www. medyatava.com/​tiraj. Although these ratings were taken into consideration while designing the relevant survey questions, ideational/​ political diversity, oppositional or government supporter characteristics of the media organ, and the newly emerged but mostly preferred online news media platforms also located among the compulsorily limited options of the questions. 13 Yeşil, B. (2016). 14 The selection of Yeni Çağ by some respondents might be explained by the additional fact that at least one of the founding members of AV is now also a member of a recently-​founded political party, namely the “Good Party” led by Meral Akşener, with which Yeni Çağ has aligned. Therefore, this particular result is most likely under-​representative of the AV volunteers. 15 The report does not have a reference for the particular newspaper preferences of the JDP electorate. Nevertheless, compared to other parties’ (NAP, RPP, and PDP) electorates, the ruling party electorates follow the news predominantly via the TV instead of newspapers or social media. (Uncu, 2018, p. 24, p. 60.) 16 Doğan (2018, pp. 17–​18). It is worthwhile to note that unlike our analysis, which includes alternative or online news platforms, the KONDA report provides only mainstream media among newspaper choices in their survey questions. 17 23 percent of RPP voters read Sözcü vs. 7 percent of Turkey; and 8 percent of RPP voters vs. 2 percent of Turkey overall for Cumhuriyet. (Doğan, 2018, p. 18) 18 Kılıç, interview with B. Akçaru, September 25, 2018. 19 Doğan, 2018, p. 25. 20 With a 1–​5 scale of satisfaction with democracy in Turkey (1 being the least satisfactory, 5 being the most), 33 participants chose ‘1’ (78.5 percent). 21 The options “social democracy” and “Christian” were not included in the first survey but were added to the second. 22 It is possible that the informants who selected “Atatürkism” might have assumed that secularism is already inherent in Atatürkism, because after all it is one of the six principles of Atatürkism. 23 Nationalist, conservative, and nationalist and conservative were taken to imply right-​wing political identities. “Muslim” was taken to show a traditional religious identity adopted by much of Turkish society. However, none of these identities were selected by the informants. 24 Our results are compatible with other recent surveys. Higher rates of Atatürkism as a signifier of political identity of RPP informants are apparent in a KONDA survey done in 2018. (Doğan, 2018, p. 21) 25 The question asked participants where they located themselves in a 1–​10 left–​ right spectrum (1 being furthest left and 10 being furthest right). Unfortunately, a second scaling was mistakenly added for the individual options, which complicated the results achieved for that question. For instance, when a respondent chose ‘3’ in the original and intended scale of 1 to 10, the system asked the respondent to also do another 1 to 10 rating within “3”. Even though this structural error

140 Ayşenur Kılıç was recognized early on, and the researcher added to the system a warning for participants, it was still confusing for most of the participants since the system did not allow adjusting the options accordingly. The researcher corrected this error in the second survey by putting a radio button for rating. 26 Kılıç, interview with B. Akçaru, September 25, 2018. 27 Akçaru also stated that most of the previous volunteers and founders of AV have joined political parties, and they continue monitoring activities under the roof of these party organizations. 28 Kılıç, interview with B. Akçaru, September 25, 2018. 29 For almost the last three decades, the percentage of votes that the LDP has received in general elections has never exceeded 0.5 percent. 30 However, the researcher did not come across these types of rhetoric from AV in the interview. The difference in the financial sources of each group might be an explanation for this difference between two monitoring organizations. While VB is an official organization with financial sponsors, AV has preferred to remain a social media platform, refused sponsorships, and financed themselves on a voluntary basis. Akçaru also pointed out that in the restrictive atmosphere of the state of emergency in Turkey (2016–​2018), it would have been difficult to account for their financial resources if they had been transformed into an official organization. (Kılıç, interview with B. Akçaru, September 25, 2018) 31 Nonetheless, the VB’s vertical membership structure is an exception to this actualizing model. The membership (not the voluntary membership, but the organizational membership) is based on an authoritative and vertical form rather than a trusted network form. 32 Doğan, 2018 May, p. 11, p. 16. 33 Here, the percentages point at the total number of participants from both organizations. In Figure  7.4 below, however, the “similarity” percentages are shown for VB and AV volunteers separately. 34 Those who identified themselves as Atatürkist or even secular in the survey could also be interpreted as supporting the NAP. However, no respondent described him/​herself as “nationalist” or “nationalist-​conservative”, which, in the case of someone orienting towards the NAP, would be odd.

References Agnew, J.A., Mitchell, K., & Ó Tuathail, G. (2003). A companion to political geography. Hoboken, NJ: Wiley-Blackwell. Arman, A. (2015). Turkey’s Vote and Beyond attracts even more volunteers for November 1 election. Hürriyet Daily News. www.hurriyetdailynews.com/​ turkeys-​vote-​and-​beyond-​attracts-​even-​more-​volunteers-​for-​nov-​1-​election-​88939. Arslan, R. (2014). Gönüllü müşahitler sandık başına gidiyor. BBC. www.bbc.com/​ turkce/​haberler/​2014/​03/​140315_​turkiye_​secim. Bakıner, O. (2014). Can the spirit of Gezi transform progressive politics in Turkey? In U. Özkırımlı (Ed.), The making of a protest movement in Turkey: #occupygezi. London:  Palgrave Macmillan (Palgrave pivot). pp. 65–​ 76. doi:  10.1057/​ 9781137413789.0008. Bennett, W.L., Wells, C., & Rank, A. (2009). Young citizens and civic learning: two paradigms of citizenship in the digital age. Citizenship Studies, 13(2), pp. 105–​20.

Election monitoring organizations  141 Bozan, İ. (2018). Oy ve Ötesi’nden Gözde Elif Soytürk:  “kolluk kuvvetinin sandık başında durması bir norm haline gelirse, seçimin uygulanmasında baskı oluşturma ihtimali doğar”. medyascope.tv/​2018/​03/​31/​oy-​ve-​otesinden-​gozde-​elif-​soyturk-​ kolluk- ​ k uvvetinin-​ s andik-​ b asinda-​ d urmasi-​ b ir-​ n orm-​ h aline-​ g elirse-​ s ecimin-​ uygulanmasinda-​baski-​olusturma-​ihtimali-​dogar/​. Brenkert-​ Smith, H., Dickinson, K., & Flores, N. (2018). Relevant recruiting for online survey participation. SAGE Research Methods Cases. dx.doi.org/​10.4135/​ 9781526438027. Çarkoğlu, A. (2000). The geography of the April 1999 Turkish elections. Turkish Studies, 1(1), pp. 149–​71. Çarkoğlu, A., & Aytaç, E.S. (2015). Public opinion dynamics towards June 2015 elections in Turkey. aciktoplumvakfi.org.tr/​medya/​presentation.pptx. Chernykh, S., & Svolik, M.V. (2015). Third-​party actors and the success of democracy: how electoral commissions, courts, and observers shape incentives for electoral manipulation and post-​election protests. The Journal of Politics, 27(2), pp. 407–​20. Çelebi, S. (2015). Civic engagement in Turkey’s democracy: The case of “Oy ve Ötesi”. Turkish Policy Quarterly, pp. 71–​8. Doğan, S. (2018). KONDA seçmen kümeleri –​Ak Parti seçmenleri. İstanbul: KONDA. konda.com.tr/​wp-​content/​uploads/​2018/​05/​KONDA_​SecmenKumeleri_​AkParti_ ​ Secmenleri_​Mayis2018.pdf. Ellis, A. (2011). Electoral geography. In B. Badie, D. Berg-​Schlosser, and L. Morlino, (Eds.), International encyclopedia of political science (pp. 747–​49). Vol. 1. Thousand Oaks, CA: Sage Publications. Enerji Bakanı Taner Yıldız: trafoya kedi girdi. (2014). Hürriyet. www.hurriyet.com.tr/​ gundem/​enerji-​bakani-​taner-​yildiz-​trafoya-​kedi-​girdi-​26128954. Guarding the Turkish Election. (2015). DW. www.dw.com/​en/​guarding-​the-​turkish-​ election/​a-​18809209. Grömping, M. (2017). Domestic monitors. In P. Norris, and A. Nai (Eds.), Election watchdogs:  transparency, accountability and integrity (pp. 167–​ 90). New  York: Oxford University Press. Johnston, R. et al. (2004). Party support and the neighbourhood effect: spatial polarisation of the British electorate, 1991–​2001. Political Geography, 23, 367–​402. doi: 10.1016/​j.polgeo.2003.12.008. Lean, S.F. (2012). Civil society and electoral accountability in Latin America. London: Palgrave MacMillan. Parliamentary elections, November 3, 2002. (2002). OSCE. www.osce.org/​odihr/​ elections/​turkey/​115703. Özbudun, E., & Tachau, F. (1975). Social change and electoral behavior in Turkey: Toward a “critical realignment”? International Journal of Middle East Studies, 6(4), . 460–​80. doi:10.1017/​s002074380002537x. Özbudun, E. (1976). Social change and political participation in Turkey. Princeton, NJ: Princeton University Press. Özbudun, E. (1980). Voting behaviour:  Turkey. In J.M. Landau, E. Özbudun, & F. Tachau (Eds.), Electoral politics in the Middle East: Issues, voters, and elites (pp. 107–​43). Croom Helm. Seçim karanlığı. (2014). Hürriyet. www.hurriyet.com.tr/​gundem/​secim-karanligi26121097.

142 Ayşenur Kılıç Toros, E., & Birch, S. (2019). Framing electoral impropriety:  the strategic use of allegations of wrong-​doing in election campaigns. British Journal of Middle Eastern Studies, 23, 367–​402. doi: 10.1080/​13530194.2019.1566694. Ugur-​Cinar, M., & Gunduz-​Arabaci, C. (2018). Deliberating in difficult times: lessons from public forums in Turkey in the aftermath of the Gezi protests. British Journal of Middle Eastern Studies, 23, 1–​23. doi: 10.1080/​13530194.2018.1491294. Uncu, B.A. (2018). KONDA seçmen kümeleri  –​Ak Parti seçmenleri [pdf]. İstanbul: KONDA. konda.com.tr/​wp-​content/​uploads/​2018/​05/​KONDA_​SecmenKumeleri_​ AkParti_​Secmenleri_​Mayis2018.pdf. Wells, C. (2013). Two eras of civic information and the evolving relationship between civil society organizations and young citizens. New Media & Society, 16(4), pp. 615–​36. doi: 10.1177/​1461444813487962. West, W.J., II. (2005). Regional cleavages in Turkish politics: An electoral geography of the 1999 and 2002 national elections. Political Geography, 24, pp. 499–​523. doi: 10.1016/​j.polgeo.2005.01.003. WVS Wave 6 (2010–​2014). (2012). World Values Survey. www.worldvaluessurvey.org/​ WVSDocumentationWV6.jsp. Yardımcı-​Geyikçi, Ş. (2015). Party institutionalization and democratic consolidation: Turkey and southern Europe in comparative perspective. Party Politics, 21(4), 527–​38. doi: 10.1177/​1354068813487110. Yeşil, B. (2016). Social media use and political activism in Turkey:  140journos, the post of others, and Vote and Beyond. Medium. medium.com/​civic-​media-​project/​ social-​media-​use-​and-​political-​activism-​in-​turkey-​140journos-​the-​post-​of-​others-​ and-​vote-​and-​e5bfe82fad06.

Acknowledgments I would like to thank all people who kindly contributed to this research by bringing the online surveys to the attention of their networks, friends, or acquaintances. I also thank to Ömer Aslan and anonymous reviewers for their constructive feedback; to the interviewee and survey informants for taking time off from their busy schedules; and to Gülsen and Edip, the editors of this book, who have been very patient and helpful throughout the long journey of this book.

Appendix 7.1: Surveys Survey Questions for VB (list) While answering the questions about “elections”, please answer them in accordance with June 24, 2018 elections. If your volunteering for Vote and Beyond is before the given date and you did not make election observation, please answer them in reference to either November 1, 2015, or June 7, 2015 general elections. If any of the circumstances do not fit to your volunteering experience, then choose the specific answer (such as I did not vote at all /​ I did not observe the elections /​ I did not participate in any of the election monitoring tasks given by the VB /​ I have not been in the field during the elections, etc.) in each question that is the most suitable for your situation.

Election monitoring organizations  143 Table 7.1 Thematic classification of survey questions Characteristics of volunteering

Duration (Q1) Number of election observation for the monitoring organization (Q12) Reason of voluntary action (Q8) Other civic societal and participatory actions and behaviors (Q23, Q24) Belief in the efficacy of voluntary actions (Q13) Voluntary Network Features (Q7, Q10, Q26, Q27, Q28)

Use/​knowledge of technology (Q5, Q9) Media use

Channels of information (news) (Q18, Q19) Media literacy (fact-​check behavior) (Q21)

Regional characteristics

Place of Residence (Q14, Q15) Place of Birth (Q16) Socio-​Economic Status (Q17)

Characteristics of political identity

Political ideology and identity (Q10, Q22) Perception of neighborhood’s political identity (Q29) Perception of family, relatives, and close friends' political identity (Q30)

Trust

Trust to elections (Q6) Trust to media (Q20) Satisfaction from the function of democracy in the country (Q25)

Open-​ended comment box (Q31)

1. I have been a member of VB since … 0–​12 months More than a year –​less than 2 years More than 2 years –​less than 3 years More than 3 years 2. If I was given the chance, I could have chosen a different political party batch for election observation (other than the one I had already obtained) Yes No It does not matter I did not do election monitoring

144 Ayşenur Kılıç 3. In the elections, I voted for the same party that I have been keeping the badge of. Yes No I did not vote in the general elections I do not want to answer this question I did not monitor the elections 4. On election day (during and/​or after the voting process), I came across a situation that was not covered in the briefings (education) given by the VB to the volunteers in pre-​elections period. Yes No 5. I have learnt the use of T3 application By myself (either searching through the Internet or by trial and error method) With the help of a friend, acquaintance, or a family member As a result of VB’s instructions given in the pre-​election briefs Other 6. In general, I trust in the accuracy of elections results conducted up until now in Turkey.* Yes No I trust some of them and do not trust others. (It depends on the election.) I do not want to answer this question 7. I first heard of VB … (Please fill in the blanks accordingly with your situation.)* Thanks to my family, relatives, friends, or acquaintances (in everyday— face-​to-​face—​life) Through social networking sites of family, friends, acquaintances, or online “friends”/​websites that I follow (online) On the Internet, by coincidence Through the TV, radio, or printed press Other I do not remember 8. I decided to be a VB member first … (Please fill in the blanks accordingly with your situation and choose only one answer.)* With the suggestion of a friend/​acquaintance/​etc. As a result of one or a few political event(s) Due to my political opinions Since I have been skeptical about the reliability of election results With an aspiration of participating in and contributing to the electoral process Other

Election monitoring organizations  145 9. My Internet and computer technologies skills are* Too restricted/​I need help At an alright level that I can take care of my needs without getting help from others At a good level At a very good level (that I can code, create my own website, and so on) 10. Which of the identities below describes you the best? (Please choose only one answer that you think it represents you best)* Liberal Socialist Conservative Nationalist Nationalist and conservative Secular Atatürkist Secular and Atatürkist Muslim Atheist None of the above I do not want to reply 11. During my volunteering, I  directly asked or encouraged … number of people around me to volunteer for VB (such as friends, family members, relatives, business friends, etc.) (Please choose the proper interval for your case.) 0 (zero) 1–​6 7–​20 20+ 12. I participated in … elections and took part in ... elections as an observer. (Please choose the number of elections that you were involved with as a VB volunteer or choose among the other suitable options accordingly if you did not take part in election monitoring processes.) 1 2 3 4 I did not take part in any of the election monitoring processes I did not actively observe elections in the field, but I helped VB in other works on the “factory floor” 13. As a VB volunteer, I  think I  made an affirmative difference in the vote count process. Yes No

146 Ayşenur Kılıç 14. Which geographical region do you reside in? Marmara Eastern Anatolia Southeast Anatolia Central Anatolia Black Sea region Mediterranean Aegean 15. The place of my residence is A metropolitan city A city A sub-​province A countryside 16. Where is your place of birth? Marmara Eastern Anatolia Southeast Anatolia Central Anatolia Black Sea region Mediterranean Aegean I was not born in Turkey I do not want to answer 17. Your monthly level of income (in Turkish Lira) 1–​1065 1066–​2029 2030–​5000 5001 or above I do not want to answer I do not have a regular income/​I am unemployed 18. I generally follow the daily news in this way: Through TV, radio, printed press Via the online news platforms on social media Via the people I follow and their posts about news on social media I do not follow daily news regularly 19. We can receive the news from more than one source today. Which of the following national newspapers and online media platforms would you say you follow more than others to receive news or read opinion columns, even if you follow more than one source? Please select one or more among the following options. (Newspapers listed alphabetically.)

Election monitoring organizations  147 Akşam

Bianet

Evrensel Gazete Duvar

Bir Gün

Cumhuriyet Diken

Haber Sol Habertürk

Dokuz8haber Dünya

Hürriyet Karar

Medyascope.tv

Milliyet Ötekilerin Posta Postası

Sabah

Sözcü

Star

Türkiye

T24

Yeni Şafak

Other

I don’t want to reply

I don’t read newspapers

Yeni Akit Yeni Çağ

20. “I trust in the accuracy of the news about Turkey that I followed in the Turkish media (printed press, TV, or online media) last week”. Which of the below is the closest to what this sentence means for you? I trust all of them I trust very few of them I generally trust despite some exceptions I do not trust the news on the Turkish media, other than the ones that appear on the media channels I follow 21. Last week, I  checked (followed) the fact-​checking news platforms for the news that I am doubtful of. Yes No I don’t know what a fact-​checking platform is 22. We often hear that people talk about political “left” and “right”. Below, there is 10-​point left–​right scale. 1 represents the furthest left, and 10 the furthest right. Where do your views stand in this scaling? 1–2–3–4–5–6–7–8–9–10 23. I am a member of a political party. Yes No I do not want to answer 24. Apart from the VB, I am a member of a civil societal organization.* Yes No I do not want to answer 25. To what extent are you content with the functioning of Turkish democracy in general? (1—​the least, 5—​the most) 1–2–3–4–5 26. I meet with other volunteers of the VB or with the people in the VB administration outside the activities of the association. (0—I don’t meet at all. 5—I frequently meet, I have friends in that group.) (The question refers to face-​to-​face meeting.) 0–1–2–3–4–5

148 Ayşenur Kılıç 27. I interact with other volunteers of the VB or with the people in the VB administration outside the activities of the association. (0—I don’t contact at all. 5—I frequently contact with, I have friends in that group.) (The question refers to online, phone, or other types of mediated contact.) 0–1–2–3–4–5 28. I had known some people from the VB volunteers or from the VB administration before I became a voluntary member of VB. Yes No I do not want to answer 29. Which of the below statements best describes your predictions about your neighbors’ political views? I​ think most of my neighbors have completely different political views from mine I think very few of my neighbors have similar political views to mine I think some of my neighbors have similar views as mine, except some minor differences I think most of my neighbors have similar political views to mine I do not know/​I do not want to answer 30. Which of the below best describes your views about your family, close friends, and close relatives? I think/​know very few of them voted for the same party as I did in the elections I think/​know some of them voted for the same party as I did in the elections I think/​know most of them voted for the same party as I did in the elections I think/​know that almost all of them voted for the same party as I  did in the elections I do not want to answer/​I do not have any idea 31. [Open-​ended comment box] Please insert your ideas that you would like to share either about the VB, your voluntary experience of VB, or anything that you think that might help for this study.

Appendix 7.2: Interview Burçu Akçaru. 2018. An interview on the Ankara’s Votes. Interviewed by Ayşenur Kılıç [face-​to-​face interview] Ankara: September 25, 2018.

The list of interview questions (with adjusted follow up questions) 1. Can you please tell me about your position in AV and the time period you have worked/​volunteered for it? (Follow up question: Do you still work or volunteer for AV? If not, do you still keep contact with the AV team?) 2. How did your path cross with the AV’s? Had you known the founding members before the AV? (Follow up question: AV’s founding date is very close to that of the

Election monitoring organizations  149 VB. Are you inspired by them? How did this process develop? Was it influenced by the Gezi protests?) 3. I always wonder how do AV-​type organizations—​as an idea—​germinate? A group of friends gathered, talked about the country issues, saw some problems about those issues, and realized some necessities. And then? How does it bloom, how does it happen and spread? What can you tell us about these kind of processes thanks to your AV experience? Why did AV not become an association and was it disadvantageous for the AV social platform? 4. How and by which criteria is the AV’s administrative body designed or assigned? Can you tell me more about the administrative details? 5. How many elections have you screened and volunteered for the AV? Which ones? 6. Do you have any research data or information available about the demographic and social profiles of your volunteers? Could you please, one more time now, inform us about the volunteer profile of the AV under your term?   Follow up questions: • How many volunteers did you have at that time? • The reports state that your volunteer network generally consists of young generations. What can you tell us about their socio-​economic status? For instance, they are generally named as “white collars”. Can you be more specific, please? 7. Is Turkish citizenship a pre-​requisite for being a AV volunteer? 8. From the establishment up until your term, how has the number of volunteers increased and spread? What kind of strategies or methods have you followed in order to increase the number and scope of your volunteer network? 9. I have already asked about your platform’s strategies in order to accelerate the spread of your volunteer network. But there is also the volunteers’ side (in realizing your platform). Have you had any research or personal impressions such that how have your volunteers heard about your organization initially, or how have they decided to be a volunteer, etc.? 10. Beside its volunteer network, the VB also has a vertical organizational structure. It has an individual responsible from each electoral district unit (provinces, and polling place buildings). How have these individuals being assigned? How often do they change? Does AV have any similarities with the VB in this sense? • Follow up question:  I have already asked a similar question in terms of the founding group. What about the relations between the vertical organizational structure and the founding cadre? How have these relations flourished? Have they known each other before the AV? 11. Is there anything about your volunteer profile that you have not been content with? For instance, have you ever thought about enhancing your target group as your volunteer network? 12. Have you experienced a problem or an issue with any of your volunteers? Or has there ever been a situation where you did not accept someone as your volunteer? 13. What do you think about your volunteers’ representativeness of Ankara? Which geographical regions and political or social groups do you think your volunteers represent the most in Turkey? 14. Do you know what kind of changes your volunteer network has had from 2015 elections up to this day?

150 Ayşenur Kılıç 15. Did your volunteer network change between the June 7, 2015 elections and the November 1, 2015 snap elections (either in terms of number or the amount of participation to election observations) 16. How many active volunteers do you have today? 17. During your post with the AV, have you ever considered any change or reform in your volunteer profiles? Did you want to enhance this profile, and if yes, in terms of what? 18. Do you have any research on the sustainability of volunteer activity of your volunteers? 19. The VB developed the T3 application. How has the AV counted/confirmed the voting reports? 20. On the VB website, it says that the membership has a special system in which they only accept members with endorsement by a reference. However, you have a different organizational structure for the AV that you call the skeletal system, the musculature, etc. Can you please tell me more about the details of this? 21. As an organization, how did you decide on the electoral districts where you would observe the elections? For instance, you included fewer eastern districts in comparison to the other geographical regions. What were the reasons for this choice—​ insufficient number of volunteers in this region, insufficient number of volunteers in general, or some other factors? 22. You stated that AV did not observe the June 24, 2018 elections due to the state of emergency, but there was a similar situation in 2017 as well, yet then AV observed the elections. Why? 23. How did you organize the volunteers in accordance with the electoral districts? 24. There is a partisan system of election observation in Turkey’s electoral system. Meaning that only political party representatives can become official electoral observers as long as the person has a party batch. Does this partisan system challenge the AV’s claim of being a supra-​political party electoral observer since you have to have a party batch (and the governmental party does not provide this batch to your organization)? • Follow up question:  How was the distribution of party batches in the 2015 elections? 25. If you compare and contrast the application of SEÇSİS (the vote-​counting system of the Supreme Election Council of Turkey) and T3 (the VB’s verification system of the counted votes), what can you tell us? 26. In recent years, there are some studies which argue that the constituents’ behavioral patterns in Turkey can be explained through the electoral geography of three main categories. For instance, the western coastal line generally votes for secular oppositional parties; middle Anatolia mostly vote for nationalist-​conservative political parties; while eastern and southeastern parts mostly vote for pro-​Kurdish parties. As an election observer, what are your impressions about this subject? If you were to reflect Turkey’s electoral geography, how would you depict it? 27. Do you have information about the demographic distribution of your volunteer network? Which regions are they from? How do you think your volunteers represent Turkey’s electoral geography? 28. Now I will provide you some details from a public survey result. In the survey conducted by the academics, Ali Çarkoğlu and Erdem Aytaç, prior to the 2015 elections, there are two questions about electoral integrity. Among participants asked if the coming elections would be fair, 40 percent of the oppositional parties’

Election monitoring organizations  151 constituents say “No, not at all”, and 29 percent say “it will be quite unfair”; vis-​ à-​vis 3 percent and 8 percent of the government party electorates who gave the same answers respectively. Similarly, “To what extent do you believe in the accuracy of the vote counting of the coming elections?”, 38 percent and 34 percent of the oppositional party electorates say “never, or very little”, vis-​à-​vis 4  percent and 11  percent of the government party electorates. In this context, the results illustrate that trust in elections seems to be correlated with the electorate’s political standing point—​whether he votes for the oppositional parties, or the governmental party. From this point of view, can we say that the AV volunteers have a sense of distrust in election processes (voting and counting), and they volunteer for elections because of this feeling? (You should not consider this question only specific to the current government–​opposition parties.) 29. Based on your AV experience, if I request a little sociological evaluation, what can you tell us about the AV’s administration and volunteers? Compared to Turkey’s socio-​political atmosphere in general, do you think the AV reflects a similar diversity of views? 30. Do you have contact with international election observation organizations such as the OSCE?

8  Engendering the parliament A study of women deputies in Turkey, 1935–​20181 Mezher Yüksel

Introduction Women in Turkey got the right to elect and be elected at a relatively early point due both to the women’s movement in the late Ottoman Empire and the political-​ideological choices of the founding Republican elites. Prior to the foundation of the Turkish Republic, women even attempted to establish a political party by the name of the People’s Party of Women. However, as they were not able to get the required permission, they continued their struggle under the umbrella of an association2 (Toprak, 1988; Zihnioğlu, 2003). Although they prevented women from establishing an independent political party, the founding Republican elites viewed women’s rights as a symbol of achieving the same level of progress with the civilized world. Moreover, it was a debt to be paid back to the women who suffered by taking on the burden of wars. Yet most important of all, it was the most effective political and ideological step that would make return to the theocratic Ottoman state impossible while eradicating the hegemony of religion (Tekeli, 1982, pp.  208–​9). Therefore, as a part of building a contemporary and western-​style society, a number of regulations concerning women were put into practice in the early Republican period (Tekeli, 1982, p. 218; Kadıoğlu, 1999, p. 107; Arat, 1997). Undoubtedly, the most significant of all was granting women the right to vote and the right to be elected. While in 1930 women were granted the right to participate in local elections, in 1934 they were granted the right to participate in national elections. Thus, Turkey was one of the first countries in the world that recognized women’s right to elect and be elected (Güneş-​Ayata, 1998, p. 237). Moreover, viewed as symbolic signs of contemporariness during the single-​party period, women took their seats in the Parliament with relatively high proportions of representation. Following the transition to the competitive multi-​party system, women lost their symbolic significance and for several decades to follow, they were represented in very low proportions in the national assembly. From the 1990s on, thanks both to the diversified and empowered women’s movement on the one hand, and to the electoral achievements of the political parties that reserved more seats to women candidates on the other,

Engendering the parliament  153 20.0 18.0 16.0 14.0 12.0 10.0 8.0 6.0 4.0 2.0 0.0

17.8 14.4

17.3 14.7

9.1 4.5 3.7 3.5

1.9

1.8 1.1 1.3 0.9 0.6 0.7 1.3 0.7

3.0

2.4 1.3 1.8

4.2 4.4

Figure 8.1 Women MPs by election period, 1935–​2018 (percent)

women found relatively more opportunities for political representation in nation-​wide political life. Based on social origins of women deputies in Turkey, this study goes into a discussion of the historical development of women’s representation in the parliament. What social origins do women deputies come from? What kind of a picture do they have regarding their cultural capital and class position? In the course of time, what changes, if any, were seen in regards to their social characteristics? Concerning their constituencies, what kind of a picture do they present? In order to address these questions, the following data concerning women deputies since 1935 is examined: their earlier work experience prior to their election as a deputy, profession, level of education, the languages they speak other than Turkish, place of birth, electoral area, and political party. Although women gained the right to vote and stand for election at a relatively early time, they have still been represented in the parliament in lower rates. In the 2018 elections, the rate of women deputies went up to above 17 percent. As can be seen in Figure 8.1, since 1935, on the other hand, the average proportion of women deputies has been around 5 percent. Put in precise numbers, from 1935 until today, women occupied a total of 598 seats in the Turkish Grand National Assembly (TGNA). The political current-​based distribution of women deputies is provided below in Table 8.1. In this context, center-​right parties come first with a rate of over 45 percent.3 The largest share belongs to the JDP, which has been in power since 2002: it has 36 percent of all women deputies within its ranks. Center-​left and social democratic parties, on the other hand, rank second with 33 percent.4 Despite their rather short history in the TGNA, pro-​Kurdish parties have around 17 percent of women deputies, a relatively high rate.5 The political parties that had the least number of seats in their ranks are nationalist right-​wing parties,6 which had 4  percent of the women deputies. One can observe that women deputies’ attachment to their parties was rather

154  Mezher Yüksel Table 8.1 Women MPs by political current Political Current

Number

%

Center-​right Center-​left Pro-​Kurdish Turkish nationalist Other Total

270 195 100 22 11 598

45.2 32.6 16.7 3.7 1.8 100.0

Note: All the data about MPs in this table and the following tables and figures are obtained from TBMM (2010, 2012, 2018), Güneş (2001), and TUIK (2012).

Table 8.2 MPs and women MPs by political party, 1935–​2018 Political Party

Total number of deputies

Number of women deputies

Rate of women deputies

Pro-​Kurdish Center-​right Center-​left Nationalist Other Total

265 5.590 4.215 550 765 11.385

100 270 195 22 11 598

37.7 4.8 4.6 4.0 1.4 5.3

weak; particularly in the 1960–​2007 period, they were highly mobile between different parties. In other words, they frequently changed their parties. During this period, due to various reasons, such as the closure of parties by the Constitutional Court, unification of parties, or personal choices, approximately one-​fourth of women deputies left their party for another party.7 As important as the distribution of women deputies along political party lines is their rate in the total number of deputies of their parties. Viewed from this perspective, a striking picture emerges (see Table 8.2). The pro-​Kurdish parties rank the first with around 38  percent. The only political movement in Turkey which implements gender quota policy, the pro-​Kurdish political parties not only have high rates of women in their ranks but they also play a significant role in the shaping of the gender composition of the national parliament in favor of women. Although it has more than 45 percent of women members of parliament, the center-​right movement, on the other hand, does not perform equally well in terms of the number of women within the party itself. The proportion of women deputies within the ranks of center-​right parties is below 5  percent. A  similar picture emerges for center-​left parties. Although they have one-​third of all women deputies in the TGNA, within their total number of deputies, women’s rate is below 5 percent.

Engendering the parliament  155 Table 8.3 Women MPs by geographical region, 1935–​2018 Region

Marmara Central Anatolia Aegean South Eastern Anatolia Eastern Anatolia Mediterranean Black Sea Abroad Total

Constituency

Place of Birth

Number

%

Number

%

206 89 86 68 61 48 40 -​ 598

34.4 14.9 14.4 11.4 10.2 8.0 6.7 -​ 100.0

176 96 49 66 86 44 51 30 598

29.4 16.1 8.2 11.0 14.4 7.4 8.5 5.0 100.0

It is possible to examine the distribution of women deputies according to their geographical regions and the change seen in this regard in the course of time. This can be done through a view of such indicators as electoral area, place of birth, and the difference between metropolis and the countryside. Although there have been striking changes regarding distribution according to electoral areas, in total there seems a clear difference. As can be seen in Table 8.3, the Marmara Region ranks first with a percentage of more than 34. Central Anatolian and Aegean Regions come next with more than 14 percent. Located in these three regions and having more than half of the women deputies in their respective regions, the three metropolises of İstanbul, İzmir, and Ankara have an important share in this result. For example, İstanbul is the electoral area of about 70  percent of women deputies elected from the Marmara Region. The same rate is 55 percent and 58 percent for İzmir and Ankara, respectively. Southeastern Anatolian and Eastern Anatolian regions, on the other hand, come next with the striking rate of more than 10  percent. The least developed regions in Turkey in terms of their level of socio-​ economic development (Kalkınma Bakanlığı, 2013), these two regions have relatively high women representation mainly due to the fact that pro-​Kurdish parties have received high electoral support. Seventy percent of women deputies elected from the southeast and 50  percent of women deputies elected from the east belong to a pro-​Kurdish party. The coastal regions of the Mediterranean and Black Sea Regions are at the bottom of the list. Having the largest rate of women deputies during the single-​party period, the Black Sea Region had the smallest rate of women deputies after the transition to multi-​party political life in Turkey. Based on the indicator of birthplace, like the earlier indicator of electoral area, while the Marmara and Central Anatolian Regions are ranked first, Eastern and Southeastern Anatolian Regions occupy the third and fourth ranks. The Mediterranean, Black Sea, and Aegean Regions, on the other hand, are below 10 percent. Although it ranks third in regard to the largest number

156  Mezher Yüksel Table 8.4 Number and rate of women MPs by geographical region, 1935–​2018 Region

Total number of MPs

Women MPs

%

Marmara Southeastern Anatolia Aegean Eastern Anatolia Central Anatolia Mediterranean Black Sea Total

2.563 983 1.528 1.223 1.996 1.257 1.835 11.385

206 68 86 61 89 48 40 598

8.0 6.9 5.6 5.0 4.5 3.8 2.2 5.3

of women deputies, the Aegean Region ranks second from the bottom of the list on the basis of birthplace. The rate of deputies born abroad is 5 percent. Of these deputies born outside Turkey, around 40 percent were in the TGNA during the single-​party era and more than 50 percent entered parliament after 1999. The birthplaces of deputies born outside Turkey reflect the tendencies in population movements in the late Ottoman Empire and republican Turkey (Tekeli, 1990; Erdoğan & Kaya, 2019). While the majority of these members of parliament during the single-​party period was born in the Balkans, the large majority of those elected after 1999 were born in European countries. Table  8.4 shows that the distribution of women deputies based on the electoral areas and places of birth shows that there is a clear gap in regard to geographical regions. On the other hand, the distribution according to gender of the total quota that every geographical region has in the TGNA contributes other dimensions to the existing division. First of all, the rate of women deputies in all regions is below 10  percent. Like in the previous indicators examined earlier, the Marmara Region again ranks first. With its approximately 7  percent, the Southeastern Anatolian region comes second. Women have the least number and rate of representation in the Black Sea and Mediterranean regions. Although it has the second largest number of women deputies in terms of electoral area and birthplace, the Central Anatolian Region, on the other hand, in terms of the number and rate of women deputies within the region, is the third region from the bottom of the list. The categorization of women deputies along their professions is found in Table 8.5.8 Similar to their regional distribution, there have also been significant changes in their professional make-​up. I will go into a detailed discussion concerning these changes in the coming sections of the study. At this point, however, I would like to provide a broad picture. With about 21 percent, legal experts come first. With a striking increase, particularly after 1980, those trained in economic sciences follow with about 16 percent. The third rank is occupied by white-​collar professions such as doctors, architects, engineers, and educators. Their rate is about 12  percent, respectively. Those that have a profession in the fields of social science and humanities disciplines such

Engendering the parliament  157 Table 8.5 Women MPs by profession, 1935–​2018 Profession

Number

%

Law Economic sciences Doctor, pharmacist, dentist Education Engineer, architect Philosophy, sociology, psychology, history, archaeology, anthropology Journalism and public relations Literature, linguistics Businesswoman, merchant, industrialist International relations, diplomacy, political sciences Nurse Self-​employed Trade unionist Planner Artist, musician, painter Other Unknown Total

123 94 73 71 71 39

20.6 15.7 12.2 11.9 11.9 6.5

35 26 17 15 8 8 7 6 5 21 23 598

5.9 4.3 2.8 2.5 1.3 1.3 1.2 1.0 0.8 3.5 3.8

as philosophy, history, archaeology, anthropology, sociology, and psychology have a rate of representation of about 7 percent. While the rate of journalists and public relations is about 6 percent, those coming from the fields of linguistics and literature are more than 4 percent. While businesswomen, merchants, and industrialists are 3 percent, women deputies that run their own business are about 8 percent. 23 women deputies, on the other hand, have not disclosed their professions. A large number of these were elected after 1999. This study suggests that the history of parliament in Turkey can be examined in three periods and its historical development can be analyzed on the basis of these periods. These three periods are based on the change seen in the following criteria: rate of representation, social characteristics, socio-​ cultural capital, and constituencies. The first term covering the 1935–​1946 period is when women were represented in rather high rates and they were attached symbolic significance. The professional make-​up of women deputies in this period seems compatible with the anticipated gender roles for contemporary Turkish women by the founding Republican elites. Likewise, social class origins of the deputies have a class diversity, reflecting the understanding of society prevalent in the period under consideration, which was depicted as classless and unprivileged. From the point of view of constituencies, this period is characterized by the fact that women took their seats in the TGNA more often than not as representatives of the provincial cities. While the rate of those coming from the countryside was above 70 percent, those representing the three metropolises of İstanbul, Ankara, and İzmir was below 30 percent.

158  Mezher Yüksel Based on the criterion of regions, the Black Sea is the first with 22 percent. The Black Sea Region is followed by Central Anatolia and Aegean Regions, both with 19  percent. At the bottom of the list is Southeastern Anatolia, which has only one woman deputy. The second period covering half a century and following the transition to multi-​party politics, i.e. 1950–​1995, witnessed a dramatic fall in the proportion of women’s political representation. Moreover, there was a striking change in the social characteristics of deputies as well as their constituencies. Competitive multi-​party political systems brought about the exclusion of women from political representation in the male-​ dominated countryside. Sixty  percent of all the women deputies elected in this period were sent to the parliament as representatives of the three metropolises. Furthermore, there emerged a big gap in terms of geographical regions. While the Marmara and Aegean Regions had 70  percent of the entire women deputies, the Mediterranean, Eastern Anatolian, Southeastern Anatolian, and Black Sea regions went down below 20 percent. Finally, the predominant professional make-​up comprising teachers and civil servants was replaced with a more professionally diverse profile of deputies composed mostly of white-​collar professionals. The recent period following 1999 stands out with several characteristics. Most important of all, the rate of women’s representation in parliament regularly went up and reached a relatively high level. For the first time since 1935, their rate of representation went beyond 4 percent in the 1999 elections. This increasing trend continued in subsequent elections and reached beyond 17 percent. This is also a period when one can observe a diversification in professional composition and a change in class profile. There was an increasing presence of white-​collared professional women alongside an increase in the relative rise in the representation of lower-​class women. Furthermore, striking changes were seen in electoral regions. Since the transition to multi-​party political life, women took roles in the representation of metropolises and relatively developed regions. After 1999, on the other hand, they increasingly represented provincial towns and underdeveloped regions. Another characteristic is that women came to the public political arena as much with their gender identities as their religious and ethnic identities. While women were formerly seen as symbols of modernity and progress, during this period they turned to become representatives of peripheral identities. Two developments played an important role in the emergence of this picture:  first, the rise in women’s rights movement since the 1980s and 1990s. Second, the positive attitude adopted by the two significant peripheral political movements, namely the Islamism and Kurdish movements, towards women’s political participation and representation.

Political representation during the single-​party period: Women as a symbol, 1935–​1946 Women were granted the right to vote and stand for nation-​wide elections in 1934. In the national elections held the following year, a total of 17 women

Engendering the parliament  159 got to the parliament. In the by-​elections held in 1936, after one more woman deputy was also elected, the number of women deputies went up to 18 (TÜİK, 2012). This number meant an approximate 5 percent rate of representation, which was above the corresponding rates in many advanced countries of the era (Tekeli, 1981, p. 299). Although there was a relative decrease in the subsequent three elections, a total of 59 women deputies took their seats in the parliament until 1950, when the single-​party period came to an end. These deputies stand out with the following characteristics: a high level of education, socio-​cultural capital, experience in local politics prior to their election, and a high degree of civil society activities and different social class origins, representing different voting districts of the country. Nevertheless, both the relatively high rate of women’s representation and women deputies’ social characteristics to a great extent reflect the following points:  the dominant ideology of the single-​party period, the understanding of society of the period determined by the centralizing-​authoritarian political system that can be characterized as classless–​unprivileged–​corporatist, the symbolic meaning attributed to women in the story of contemporarization (çağdaşlaşma), and the new social roles envisioned for Turkish women. Rather than the preferences of the voters, it was the dominant political authority of the time that determined the proportion of women representatives in the parliament. To illustrate, Çakır notes that before the 1935 national elections, the number of women MPs to take their place in the parliament was decided following long-​lasting debates between Atatürk and several leading personalities of the single-​party period such as İsmet İnönü and Şükrü Kaya (Çakır, 2014, pp.  123–​4). Like they were planning an engineering project, the male members of the ruling party discussed the number of women to be elected for the parliament. The electoral system and centrist practice of the specification of candidates during the single-​party era reinforced this process. During this period, the elections were conducted according to the two-​ phased simple majority system.9 According to this system, candidates were designated by a small group in the center of the party and these candidates would be presented to the votes of secondary electors who had also been designated earlier by the center of the party. That seems to be why some candidates were nominated without even being informed about their candidacy, and sometimes candidates would even find out about their election after they were elected (Uyar, 1998; Koçak, 2005). It appears that it was not only women’s rate of representation that was planned by the male elites at the center of the party. The social class composition of women deputies also seems to bear the traces of the same centralizing social engineering mentality. In this context, the political elites of the time cared about the representation in the parliament of women coming from different social strata, and particularly from lower-​classes, albeit in symbolic proportions. As a result of the understanding of a society that, during the single-​party era, aimed to be classless, unprivileged, and corporatist,10 and the popular peasantist ideology reflected in the saying of “peasant is the master of the nation”,11 two women that were peasants and farmers12 entered

160  Mezher Yüksel the parliament. On the other extreme of the diversity reflecting women’s social class status, three women who had their own businesses were elected as deputies. The relationship between the constituencies and birthplaces of the deputies likewise reflects the conception of society as an integrated and unprivileged mass. Only 25 percent of the deputies represent their birthplaces. It must be pointed out that this practice is one of the distinctive traits of the single-​ party period and that, especially in the 1927–​1939 period, in some regions this rate went up to around 60 percent (Frey, 1965; Demirel, 2013). In his work entitled Turkish Political Elites, Frey (1965) views this picture as a reflection of the fact that politics was no longer dependent on local values and dynamics but rather it became tied to an axis revolving around universal principles and values. Frey thus views this process as a positive development experienced by Turkey’s political modernization. As a conscious strategy, the main goal of this mechanism was as follows: to consolidate the power and control of the center in the countryside and to take under control the activities of local notables in politics, particularly in the regions where feudal social structure was powerful.13 The distribution of deputies according to their birthplaces also presents a stark difference. Those who were born in the Marmara region come first with more than 40  percent. Those that were born outside the borders of the Turkish Republic came second with around 20  percent, which was almost twice the rate of those born in the Mediterranean, Central, Eastern, and Southeastern Anatolian regions. There were no deputies born in the Eastern and Southeastern Anatolian regions. The rate of those born in the Mediterranean and Central Anatolian regions was around 5 percent each. Compared to the distribution according to their birthplaces, there seems a relatively more balanced picture in terms of their constituencies (see Figure 8.2). With the exception of the Southeastern Anatolian region, which had only one MP, all regions had a representation rate of above 10 percent. The Black Sea region ranked first with 22  percent. It was followed by the Aegean and Central Anatolian regions, 19  percent each. According to the criterion of birthplaces, the Marmara region had more than 40  percent of the MPs. As to the principle of constituency, the Marmara region’s share was 15 percent. Coming after the Marmara region, the Mediterranean and Eastern Anatolian regions each had 12 percent of women deputies. Nonetheless, in terms of the total number of deputies in all regions, the proportion of women MPs is below 5 percent. Although there seems a balanced allocation on the basis of geographical regions, there is a stark gap on the basis of the distribution to metropolises and provincial cities. During the single-​party period, women mostly entered the parliament as the representatives of provincial cities. Approximately 75  percent of the MPs were the representatives of provincial towns rather than the metropolises of İstanbul, Ankara, and İzmir. In short, other than the largest three cities, 21 Anatolian provinces were represented by women during

Engendering the parliament  161 45 40 35 30 25 20 15 10 5 0

Constituency Birthplace

Figure 8.2 Women MPs by constituency and birthplace, 1935–​1946 (percent)

the period under consideration. Similar to the rate of women’s representation and their professional composition, the geographical distribution of women MPs also reflects the choices of the period’s centrist-​authoritarian single-​ party regime. When one looks at the women’s professional backgrounds and the jobs they did before elections (see Table 8.6) in addition to the above arguments, a few points come to the fore. First, compared to others, the prevalence of those who were teachers is clear. This is compatible with the new gender roles attributed to women during the period. While 66  percent of women deputies indicated that they were teachers by profession, 75  percent point out that they had taken on the position of teaching at different levels. This is hence a reflection of the new social role attached to modern Turkish women. The new primary social role attributed to women was enlightened motherhood and childrearing (Tekeli, 1982), and an extension of such a role would be teachership (Saktanber, 2004, p.  327). As enlightened mothers, Turkish women were primarily required to bring up their own children. And now they would fulfill the role of educating society as the mothers of the nation under the leadership of the headteacher, Atatürk. Furthermore, they were not only seen to be responsible for training future generations as good citizens in both family and school, but they were also expected to be role models for the “other”, “traditional” women, i.e. those women who lagged behind modernization (Saktanber, 2004, p.  327). However, women were targeted to be involved in social life particularly by way of certain traditional roles attributed to women. Women deputies adopted this goal and they reflected it in their activities in the Parliament. As Arat points out, a considerable part of the parliamentary activities of this period’s women deputies are about such matters as family, education, health, and social security, which are about or

162  Mezher Yüksel Table 8.6 Women MPs by profession, 1935–​1946 (percent) Profession

Number

%

Education Literature Doctor, dentist Sociology, philosophy Sciences Law Farmer Musician Couture Unknown Total

39 9 7 6 3 2 2 1 1 1 59

66.1 15.3 11.9 10.2 5.1 3.4 3.4 1.7 1.7 1.7

are related to the roles traditionally assigned to women (Arat, 1998, p. 256). Perhaps the most important characteristic of this period’s women deputies is that they speak as the members that represent women in the TGNA (Arat, 1998, p.  254). In their professional categorization, Literature comes second following Education. Then, in a similar proportion, come the white-​collar professions of Medicine, Dentistry, and Law. Those who indicated that their profession was Sociology and Philosophy are around 10 percent. In brief, the professional backgrounds of women deputies reflect strong parallels with the gender roles assigned to women during the single-​party era. Second, women deputies showed a great interest in civil society activities and local politics. More than 50 percent of the deputies had been members or directors in civil society organizations before they were elected, and they took an active part in their activities. Among these organizations were such official aid agencies as Kızılay (The Red Crescent) and Çocuk Esirgeme Kurumu (Society for the Protection of Children) in addition to various associations protecting and supporting the poor. Membership to women’s associations, on the other hand, was relatively low. Indeed, the fact that women were directed to social aid agencies rather than the market, trade, and political parties is a significant trait of societies in an early stage of industrialization, and this was the most important point of opposition raised by the first-​wave feminist movement (Sancar, 2011, p. 63). Likewise, around 35 percent of women deputies took a position—​most often as members of town council (belediye meclis üyesi)—​in local politics prior to their access to the Parliament. Third, Table  8.7 shows that women deputies in this period stand out in regards to both their level of education and cultural capital. Of those surveyed, 56 percent had a university degree, while 32 percent had high school education. Only two deputies did not have a diploma. In terms of their educational level, therefore, they were far beyond their society in general and the average female population in particular. To illustrate, the literacy rate of the

Engendering the parliament  163 Table 8.7 Women MPs by level of education and non-​Turkish languages they speak, 1935–​1946 Level of Education

Number

%

Number of Number languages spoken other than Turkish

University 33 High 19 School Middle 5 School No 2 diploma

55.9 One 32.2 Two

Total

100.0

59

30 14

8.5 Three

3

3.4 None

12

59

%

Languages Number spoken other than Turkish

50.8 French 23.7 English

%

43 14

72.9 23.7

6

10.2

20.3 Greek

3

5.1

Arabic

1

1.7

5.1 German

100.0

adult population in Turkey in 1935 was 19 percent, while it was 8 percent for women (TÜİK, 2005, p. 15). Yet compared to the average of the Parliament at the time, the rate of university graduates was low. In the 1935–​1950 terms of the Parliament, while the rate of university graduates was above 70 percent (Frey, 1965, p. 176), this was below 60 percent for women deputies. Another significant indicator of cultural capital is languages spoken other than Turkish. Although in terms of university degrees, women deputies have a rate below the average of the TGNA, they are above the average of the Parliament in the knowledge of languages other than Turkish. Eighty ­percent of women deputies indicated that they speak at least one foreign language. Almost all of these non-​Turkish languages are western European languages: while 73 percent speak French, 24 percent speak English. It is notable that, though spoken by a significant part of the population in Turkey, neither Kurdish14 nor such languages as Arabic and Persian, which had a significant place in the history and culture of Turkey, were spoken by almost any woman deputy. In this context, with comparison to the average of the parliament, women deputies provide a different picture. In the 1935–​1946 parliamentary terms, the rate of non-​European languages spoken by MPs was close to 15 percent (Frey, 1965, p. 178). In short, women deputies of the single-​party period stand out with their western cultural capital.

More political parties, less representation: 1950–​1995 In 1945, there was a transition to a multi-​party political system in Turkey, and in five years’ time, the single-​party period effectively came to an end. From this period onward, mainly due to the competition between political parties, political life became more closely tied to local dynamics. This

164  Mezher Yüksel picture was more pronounced in the countryside, where especially patriarchal notables were influential. One of the most important indicators of the localization of political life is the increase in the proportion of the deputies’ being natives from their own electoral areas, i.e. that were born in their own constituencies. While the rate of those born in their electoral districts was on average 40 percent during the single-​party period, following the transition to the multi-​party political system, this rate more than doubled, rising above 80 percent. Another characteristic of this period is the instability of political life. Military interventions and violence had a great impact on the political instability prevailing during the 1960–​1980 period. Transition to multi-​party politics had striking consequences concerning the participation of women in politics, their representation in the parliament, and the social characteristics of women deputies. The first and perhaps the most important consequence was as follows:  women lost the privilege they enjoyed in their designation during the single-​party era (Tekeli, 1981, p. 300). Likewise, the increasing competition between political parties had various negative results for women during the process of nominating candidates from a pollable ranking. In other words, as women were placed in lower ranks of lists of nominees, they would not likely be elected. Therefore, this was a period during which women had to struggle harder to be both nominated and elected. As a result, for more than half a century, women were almost nonexistent in the parliament. In a total of 12 general elections until 1999, only 80 out of five thousand deputies were women. Taking together the election results, the allocation of seats in the parliament, and the rate of women’s representation, which was around 1 percent, it seems that the choices of the political parties were influential in the shaping of the gender composition of the parliament in the 1950–​1995 period. In other words, the most important reason behind women’s underrepresentation seems to have been woman policies of the center-​right political parties that did quite well in the elections throughout this period. Some scholars interpret the decrease in the rate of women’s political representation as the loss of women’s symbolic characteristics (Tekeli, 1982; Güneş-​Ayata, 1998). Relatively more liberal and open development policies accompanied the change in the political system. The symbols of westernization and modernization changed; economic criteria such as development, growth, and industrialization became the new indicators of modernity and development during this period. Put differently, while the dominant sign of contemporariness during the single-​party era was the modern woman participating in social life on an equal basis with men, it was replaced with such indicators as industrialization, economic growth, and urbanization. Nevertheless, some researchers claim that the symbolic use of women was seen, albeit discontinuously (Arat, 1998; Tekeli, 1982). These researchers interpret the case of the relatively high rate of women deputies in the first parliament following the post-​1980 military intervention as the putchists’ wish to give a democratic and pluralist image to the system (Güneş-​Ayata,  1998).

Engendering the parliament  165 Table 8.8 Women deputies by profession, 1950–​1995 Profession

Number

%

Law Economic sciences Education Philosophy, sociology, psychology, history Engineering, architecture Businesswoman Medicine, pharmacy Linguistics, literature Trade unionism Planning Journalism, painting Diplomacy Unknown Total

15 14 13 9 8 6 5 5 4 3 2 1 6 80

18.8 17.5 16.3 11.3 10.0 7.5 6.3 6.3 5.0 3.8 2.5 1.3 7.5

It seems that women’s indifference to politics also had its significant share in the decrease of the rate of women’s political representation. It appears that the increasing connection of political representation to local dynamics alongside political instability negatively affected women’s interest in politics. For instance, during the 1961–​1977 period, among all presumptive nominees of deputies, the rate of women is less than 3 percent (Tekeli, 1981, p. 302). One can also see women’s disinterest in active politics during the period under consideration in the rates of the deputies’ active involvement in local political life prior to their election. While almost 35  percent of the deputies during the single-​party era had some experience in local political life prior to their election, in the period under consideration, this proportion fell as low as around 16  percent. A  similar trend to fall can also be observed in memberships of NGOs. During the single-​party period, more than 50 percent of the deputies were active members of NGOs prior to their election. From 1945 to 1999, on the other hand, one can observe that this rate fell to the level of around 20 percent. Following the transition to multi-​party politics, there were also significant changes in women’s social characteristics (see Table 8.8). First, one can note the change regarding professional background. During the single-​party period, women were ascribed new gender roles that were limited to specific professions and, hence, there was a strong parallel between professions of the deputies and these roles. In the following period, this picture underwent a striking shift. The professional make-​up of women deputies has been varied, and it got close to the average of the parliament. Women deputies of this period were professionals in the fields of law, architecture, engineering, medicine, pharmaceutics, journalism, public administration, social sciences,

166  Mezher Yüksel and economic sciences. The most striking change has been in the context of deputies who were coming from the field of education. While the deputies that were professionals in the field of education during the single-​party era comprised the largest group of deputies with a rate of 65 percent, this rate decreased to 16  percent. Nevertheless, in regards to their jobs before their election, the proportion of those who took positions in different levels of education was around 40 percent and among them, academics comprise the largest group. While the rate of educator deputies fell, the rate of white-​collar professionals went up. Among the latter, the rate of legal practitioners was the highest. The increase in the rate of legal practitioners, who comprised the largest group among all professions, was not limited to women deputies; this was the case for the entire parliament. As some researchers suggest, the rise in the rate of professionals of law can be taken as a sign of change in political system (Tachau, 1980). Yet it must be pointed out that jurisprudents (hukukçu) were predominantly lawyers, who comprised the biggest group in political representation, thanks to various professional advantages that they had. The strong relationship between the main function of the parliament, i.e. legislation and the professional knowledge and accumulation of legal practitioners is their leading professional advantage. Moreover, with the roles they play in the resolution of juridical and legal questions, they acquire respectability and reputation in society; therefore, the candidates within this professional category become more advantageous to be elected in the eyes of both political parties and the electorate. In addition, if they do not get elected or once they end their active political life, the fact that it is relatively easy for them to continue their professional practice can be counted as another professional advantage. In parallel with the general trends in the parliament, the number of women deputies coming from jurisprudent origins saw a sharp decrease in the 1980– 1995 period. Only four of the fifteen women deputies serving in the parliament in the 1950–​1995 period were elected in the post-​1980 term. A great rise in the number of deputies educated in economic sciences accompanied the sharp decrease in the number of jurisprudent deputies. In line with the general tendencies of the period and comprising the second largest professional category with a rate of 18 percent, the rate of economists and managers in the TGNA underwent a striking increase, particularly after 1980. This picture reflects the great shift that was viewed in the rise of neo-​liberalism in Turkey in the post-​1980s, when the state was perceived as a business enterprise; politics was seen as an art of managing this business enterprise with great efficiency and profitability. In this context, the professional profile of deputies also took shape in line with this understanding; therefore, the candidates coming from economics and management backgrounds were seen to have the necessary skills for this kind of understanding of state and politics, and these candidates thus were represented more. Although it was not as striking as the increase in the rate of jurisprudents, there has also been a considerable rise in deputies with a background in

Engendering the parliament  167 engineering and architecture. This group is ranked fifth with a rate of 10 percent. The rise in the number of women architects and engineers is a sign of the fact that the relationship conceived between gender roles and certain professions began to be undermined. This was also a result of women’s participation in social life with an increased professional diversity. On the other hand, similar to the general tendencies of the period, this picture reflects the change in the understanding of modernization during the period. The early Republican view of modernization shaped by the belief in transforming social life mostly by means of formal regulations was replaced by the view of development through industrialization. Engineers were welcomed, and they gained importance as the pioneering staff in the way to contemporary civilization. Consequently, the following tendency in the post-​1960 period can be highlighted: engineers came to the fore in political life and decision-​making processes. This current was also compatible with the dominant ideology of the era: an understanding of modernization oriented toward industrialization (Göle, 1998). The second important change in women deputies’ social characteristics is about their social class. The social class profile of women deputies shifted to the disadvantage of lower classes. Symbolizing the corporatist view of society during the single-​party era, the representation of farmers and peasants came to an end in this period. An exception to this rule was Hatice Mualla Akarca, who was elected from the province of Muğla from the ranks of the RPP during the 14th term of the Assembly. Although she indicated farming as one of the jobs she did prior to her election, Akarca received her master’s degree in the USA in the field of agricultural engineering. She also had a diploma in German Language and Literature. With these qualifications, it really seems difficult to find a shared social class background between Akarca and Hatı Çırpan, who was one of the female deputies symbolizing the single-​party period. In terms of the representation of farmers and peasants, women deputies differ remarkably from their male counterparts. Although it shows a trend that would decrease in time, the rate of all MPs who pointed to farming as their profession or as one of the jobs they did prior to their election, was above 10  percent in the 1950–​1995 period. Another notable indicator concerning their class characteristics is as follows: more than 10 percent of these women deputies were employers, the boss of their own businesses, or were managers in the private sector. After the transition to multi-​party politics, there were also important changes in the deputies’ cultural capital. The proportion of those who received tertiary education is about 74  percent; one-​third of these women deputies had a master’s degree and six of them also had two separate undergraduate diplomas. This way, the rate of university-​educated women deputies caught up with the mean average of the parliament. Another indicator in cultural capital is knowledge of languages other than Turkish. In parallel with the rise in educational levels, there has also been a large increase in both the rate and diversity of the languages spoken other than Turkish. Table 8.9 shows that

newgenrtpdf

Level of education

Number

University

59

73.8

One

35

High school

21

26.2

Two

Total

80

%

100.0

Number of languages Number spoken other than Turkish

Languages spoken other than Turkish

Number

%

43.8

English

52

65.0

27

33.8

French

37

46.3

Three

3

3.8

German

19

23.8

Four

7

8.8

Greek

8

10.0

None

8

10.0

Latin

3

3.8

Italian Spanish Kurdish Persian Arabic

3 1 1 1 1

3.8 1.3 1.3 1.3 1.3

80

%

100.0

168  Mezher Yüksel

Table 8.9 Women MPs by level of education and knowledge of non-​Turkish languages, 1950–​1995

Engendering the parliament  169 during the 1950–​1995 period, the rate of women deputies who spoke at least one foreign language was 90 percent, which was way higher than the mean average of the parliament. Around 50 percent of them, on the other hand, spoke two or more foreign languages. Needless to say, knowledge of a foreign language is a significant asset in order to follow the news all around the world, as well as to do politics. Nevertheless, one can see that an important feature of the preceding period continues in this period, too. Almost all the foreign languages were western European languages. On the other hand, the number of women deputies who spoke Arabic, Persian, and Kurdish was one each. Transition to a competitive multi-​party political system also brought about dramatic changes in the regional distribution of deputies. While there has been a large increase in the number of those elected from the Marmara and Aegean regions, there was a decrease in the number of deputies of almost all of the remaining regions. In the 1950–​1995 period, 70 percent of women deputies were elected from the Marmara and Aegean Regions, which is double that of the previous term. In other regions, by contrast, there was a decrease in the proportion of women deputies. The most dramatic decrease was seen in the Black Sea region. While it had the largest number of women deputies with a rate of 22 percent during the single-​party period, the Black Sea region had a mere 5 percent of the women deputies in the TGNA. The respective rates in the Mediterranean and Eastern Anatolian regions went down almost by half. Like the former period, the Southeastern Anatolian region was again at the bottom of the list with its two women deputies. There are big differences regarding the regional distribution of women deputies by their political party affiliation. For example, while almost all women deputies elected from the Eastern and Southeastern Anatolian regions were affiliated with center-​left (Republican People’s Party (the RPP), Populist Party (the PP), Social Democratic Populist Party (the SDPP)) and leftist parties (the Workers’ Party); the women MPs of the Mediterranean, Aegean, and Marmara Regions belong to the center-​right parties (the Democrat Party—​ the DP, the Justice Party—​the JP, the Motherland Party—​the MP, the True Path Party—​the TPP). Taken together, the election results, the allocation of the seats in the parliament, and the rate of women’s representation which was around 1 percent, demonstrating that the choices of the political parties have been influential in the gender composition of the TGNA in this period. In other words, the most important reason underlying the underrepresentation of women during this period was the woman policies of center-​right parties that were highly successful in elections. As in the case of regional distribution, a similar picture emerges in the context of place of birth. Figure 8.3 demonstrates that the Marmara region ranks first as the region where more than 50 percent of women deputies were born. This rate was about 14 and 13 percent for the Central Anatolian and Aegean regions, respectively. The rate of women deputies born in the Mediterranean, Black Sea, Eastern and Southeastern Anatolian regions was 16  percent. In other words, the total share of the latter two regions is below four percent. In

170  Mezher Yüksel 60.0 50.0 40.0 30.0 20.0 10.0 0.0

Constituency Birthplace

Figure 8.3 Women MPs by constituency and birthplace, 1950–​1995 (percent)

short, by the standards of both constituency and birthplace, women deputies were concentrated in the Marmara and Aegean regions. However, the rate of women MPs in the total number of deputies in these two regions is below three percent. In parallel with the change in regional distribution, there has also been a striking shift on the basis of countryside–​metropolis. While women were most often representatives of the countryside, this changed after the transition to multi-​party political life and they entered the TGNA mainly as the representatives of the three metropolises. While the rate of women deputies elected from the three metropolises was below 30 percent during the single-​ party period, it more than doubled, going up to 60 percent. The main reason seems to be the fact that competitive multi-​party politics came to be much more tied to local dynamics and especially the increasing power of the provincial male-​dominated notables in politics. Moreover, the choices of the center-​ right parties that had 50 out of 80 deputies during this period seem to have been influential in this result. Of these 50 deputies, 30 were elected from the three metropolises. In brief, the increasing presence of local dynamics in politics paved the way for the growing exclusion of provincial women from political representation and it limited women’s representation predominantly to metropolises.

More women, more representation: 1999–​2018 From the 1990s onward, a new stage began in women’s representation in the Turkish parliament. Crystallizing particularly with the 1999 elections, this new period comes to the fore with two outstanding characteristics. First, women’s representation in the parliament regularly went up. The rate of women deputies rose from 4 percent in 1999 to above 17 percent in the 2018 nation-​wide

Engendering the parliament  171 elections. Second, women came to the public political arena as much with their gender identities as their religious and ethnic identities. This period’s women deputies took their seats in the parliament as actors symbolizing politicized ethnic and Islamic identities. While they were previously viewed to symbolize progress and contemporariness, they now came to symbolize the repressed and peripheral identities. The greatest share in the increasing representation of women in parliament without a doubt is due to the struggles of the women’s movement, which has been on the rise in its varied manifestations since the 1990s. Despite certain divisions among different women’s groups,15 the women’s movement in general voiced demands for more participation of women in social and political life on an equal footing. Perhaps for the first time in history, women fought for their rights in a massive and organized way and achieved this result. The change introduced in the male-​dominated and sexist appearance of the parliament had its notable share in the two political movements that were on the rise since the 1990s and were represented more often in parliament:  Islamism and Kurdish nationalism. Both movements had strong women’s organizations.16 One must particularly point to the Justice and Development Party (the JDP), which has Islamist origins and has been in power since 2002. The pro-​Kurdish parties, on the other hand, were the Peace and Democracy Party (the PaDP) and the Peoples’ Democratic Party (the PDP). To give an example, in the 2018 nation-​wide elections, out of a total of 104 women deputies, 79 were JDP and PDP members. In its list of candidates of the PDP, women were not only reserved more than 40 percent of the seats,17 but the PDP, with its rate of 40 percent of women MPs, also ranks first among all the parties in the TGNA (TBMM, 2018). These two movements also occupy a special place in highlighting women as representatives of religious and ethnic identities. At this point, two personalities come to the fore as symbols of these two stances.18 The first is Leyla Zana, who was elected to parliament in 1991 and who, at the end of her parliamentary oath, stated the following in Kurdish: “I take this oath for Turkish and Kurdish peoples”. This caused huge reactions, subsequently leading to ending Zana’s deputyship and the ban on her political activities on account of the accusations of separatism by the Constitutional Court. After many years of imprisonment and a ban on her political participation, Zana was elected to the parliament in 2011 from Diyarbakır. Following the November 1, 2015 elections, she entered the parliament from the province of Ağrı. In the ceremony of parliamentary oath, since she read the “Turkish nation” as “Nation of Turkey” her oath was considered invalid and subsequently, due to her inattendance in parliamentary activities, i.e. absence in the parliament, she was stripped of her deputyship with a parliamentary decision. The second personality was Merve Kavakçı, elected to the parliament as a member of the Virtue Party (the VP), and she was not allowed to take her parliamentary oath because she was wearing a headscarf. Unlike Zana, Kavakçı was not sent to jail but was dismissed from the parliament and

172  Mezher Yüksel 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0

Constituency Birthplace

Figure 8.4 Women MPs by constituency and birthplace, 1999–​2018 (percent)

then subsequently denaturalized.19 This was not the only feature that made Kavakçı different. She rejected the identity anticipated for modern Turkish women by the founding elites of the Turkish Republic. The prime minister at the time, Bülent Ecevit, viewed Kavakçı’s appearance in the parliament while wearing a headscarf as a challenge for the state and asked her to be put in her place.20 The case of Merve Kavakçı points to a rupture and the beginning of a new stage in the representation of women in the TGNA and its symbolic meaning. Today 25 percent of the women MPs wear a headscarf.21 The symbolic meaning embodied in the cases of Zana and Kavakçı includes a challenge targeted at both the founding values of the Republic and the ethno-​ secular boundaries22 of the collective identity built upon these values.23 As much as their gender identities, these women take their place in the TGNA by their ethnic and religious identities, and they represent the two peripheral identities repressed throughout the Republican history. Both the JDP and pro-​Kurdish parties have a great share not only in the rise of women’s representation but also in more representation of the countryside by women MPs in general, and more representation of the regions with low women deputies in particular. Thanks to the JDP, the PaDP, and the PDP the regional distribution of women deputies changed, and women got to be represented in increased rates in the regions where they had not been represented throughout the multi-​party political life. As can be seen in Figure 8.4, the most dramatic shift in this context was seen in the Southeastern Anatolian region. While it had only one female deputy during the single-​party period and two in the following period, the region in the post-​1999 period had a total of 65 women MPs, corresponding to a rate of more than 14 percent, which makes this region rank third in terms of women’s representation among

Engendering the parliament  173 all regions in Turkey. Of these women deputies in the Southeastern Anatolian region, 72  percent are affiliated with pro-​Kurdish parties, and 28  percent with the JDP. A similar case emerges in the context of the Eastern Anatolian region. While in the preceding period it had only 6 percent of all women deputies, it had now risen to above 11 percent, all of them again affiliated with pro-​Kurdish parties and the JDP. Their rates are 63 percent and 37 percent respectively. Coming to other regions, the contribution of the JDP appears even more striking. More than 80  percent of the women deputies elected from the Black Sea region are members of the JDP. This rate is 58, 54, and 51 percent for the Central Anatolian, Mediterranean, and Marmara regions, respectively. Fifteen provinces found mostly in the Eastern and Southeastern Anatolian regions for the first time in Republican history were starting to be represented by the JDP and pro-​Kurdish parties.24 Some provinces, on the other hand, were represented by the women deputies of these parties for the first time since the transition to multi-​party politics. Consequently, women’s rate in the representation of the countryside went up, making women more important in the political life of the countryside. The increasing role of women in provincial politics and the rising significance of local connections in political representation were seen as a shift in the marital status and socio-​cultural traits of women deputies. As pointed out by various researchers (Tekeli, 1982; Çakır, 2014), in Turkey, unlike for men, for women to be married has had a negative impact on their political involvement and their assuming active roles in political life. As a result, the rate of married women politicians has been much lower than the rate of married male politicians. For example, until the 2000s, while the proportion of married women deputies was 65 percent, it was above 90 percent for male deputies. This situation changed, however, starting with the 2000s, and this rate went up 70 percent for women deputies. The increasing rate of women’s political representation in the countryside has played a role in this change since the rate of the married among women deputies representing the provinces other than the metropolises is above 75 percent, while it is about 65 percent among the representatives of the metropolises. The most striking change in socio-​ cultural characteristics, however, was seen in the rate of the non-​Turkish languages spoken by deputies (see Table  8.10). First, the rate of those who cannot speak a language other than Turkish went up to close to 15 percent. In the preceding period, it was 10 percent. Likewise, the rate of those who could speak at least two or more languages was halved, going down to about 23 percent. Moreover, while there was a dramatic decrease in the rate of the speakers of western European languages, especially of French and German, the rate of Kurdish and Arabic speakers went up. There was an especially impressive increase in the number of Kurdish speakers. With a rate of more than 11 percent, Kurdish became the third most spoken language. Needless to say, the role of pro-​Kurdish parties has been influential in the rise in the proportion of Kurdish speakers.

newgenrtpdf

Level of education

Number

%

University High school Middle school Primary school

416 34 4 5

90.6 7.4 0.9 1.1

Total

459

100.0

Number of non-​Turkish languages

Number

%

One Two Three Four Six None

285 77 19 8 3 67

62.1 16.8 4.1 1.7 0.7 14.6

459

100.0

Non-​Turkish languages

Number

%

English German Kurdish French Arabic Italian Spanish Latin Persian Armenian Other Total

341 53 53 43 24 7 5 5 4 2 9

74.3 11.5 11.5 9.4 5.2 1.5 1.1 1.1 0.9 0.4 2.0

174  Mezher Yüksel

Table 8.10 Women MPs by level of education and knowledge of non-​Turkish languages, 1999–​2018

Engendering the parliament  175 Fifty-​one of the 53 deputies who declared their knowledge of Kurdish are members of pro-​Kurdish parties. Pro-​Kurdish parties also have 50 percent of the deputies who cannot speak a language other than Turkish. Another change viewed in the socio-​cultural profile of women deputies is as follows: on the one hand, the rate of those who received a university degree went up above 90 percent; on the other hand, those who received only primary and secondary education once again began to have a seat in the TGNA. Although their rate is negligible, it is important to note that after a half century the graduates of primary and secondary schools once again assumed a position in the parliament. This was due mainly to the pro-​Kurdish parties that deputies with only primary and secondary education as well as those without university education got a seat in the parliament. Forty-​one of the 43 deputies without a university education were members of pro-​Kurdish parties. Compared to earlier periods, while they demonstrate differences regarding such issues as marital status, in the post-​1999 period, women deputies have strong parallels with the single-​party period in the context of their electoral regions and participation in civil society activities. Like the single-​party period, in this period, too, the rate of participation in civil society activities was above 50 percent. Nevertheless, while the civil society experience of the MPs during the single party era was mostly concentrated on charity activities, the women MPs of the period under consideration come to the fore mostly with their active participation in the activities of women’s organizations. Moreover, in parallel with their political views, the experiences of this period’s deputies in this regard show a rich range. Among the associations to which they were members are as follows, among others: human rights, environmental associations, Association for Atatürkist Thought, Association for Support of Contemporary Life, and associations for aid for the families of martyrs. Unlike experience in civil society organizations, experience in local politics seems to have lost its significance. During the single-​party period, one-​third of women deputies had political experience at a local level before they were elected. In this sense, local political experience played an important role as an intermediate stage before national political representation. Yet this changed after the transition to multi-​party politics and experience in local politics started to lose its importance. The process of change increasingly continued during the period under consideration and the rate of those with experience in local politics prior to their election went down to around 10 percent. As I  pointed out earlier, after the transition to multi-​party politics, the established relationship between gender roles and certain professions was loosened, and women started to participate in social life with an increasing professional diversity. This was also reflected on the distribution of women deputies’ professions and of the jobs they had before they were elected, and their professional composition changed significantly. The process of change continued during the period under consideration. In this context, while professional diversity increased, the rate of the representation of professions changed considerably. The increase in the proportion of

176  Mezher Yüksel Table 8.11 Women MPs by profession, 1999–​2018 Profession

Number

%

Law Economic sciences Engineering, architecture Medicine, dentistry, pharmacy Social sciences, sociology, psychology, history, archaeology, anthropology Journalism Education International relations, diplomacy, political sciences Philology, literature Communication and public relations Trade, industry, tourism, businesswoman Self-​employed Nurse Painter, artist Planning Trade unionism Other Unknown Total

106 80 63 61 24

23.1 17.4 13.7 13.3 5.2

23 19 14 12 11 11 8 8 3 3 3 15 16 459

5.0 4.1 3.1 2.6 2.4 2.4 1.8 1.8 0.7 0.7 0.7 3.3 3.5

white-​collar professions is particularly striking. There was a rise in the rate of jurisprudents, which had relatively gone down in the post-​1980s, and it ranked first among all professions. The largest rise was seen in the rate of doctors, dentists, and pharmacists, whose rate went up above 13 percent, which places them in a ranking of fourth among other professions. Despite their relatively little increase, engineers and architects went up to the third ranking. In short, approximately half of all women in the parliament in this period were comprised of white-​collar professionals. While the rate of white-​collared professionals went up, the rate of those in professions coming from the fields of education, humanities, language, literature, and science went down. The most dramatic decrease was seen in the rate of educators. While, with a rate of 65 percent, they constituted the largest group, there was a sharp decrease down to around 12 percent after the transition to multi-​party politics. In the period under consideration, it further went down to about 3 percent. Nevertheless, when one views the jobs they had, one can see that a significant number of women deputies had a background in education. Indeed about 25 percent of women deputies had been academics. Finally, regarding the question of social class representation, this period presents a relatively more democratic and pluralist picture. There has been a big increase in the number of women deputies who were the boss in their own work and those who were managers in the private sector. The sharp rise

Engendering the parliament  177 in proportion of those who were managers in the private sector is particularly notable: more than 17 percent of them were managers in the private sector prior to their election. In other words, approximately one-​fourth of women deputies were either members of the capitalist class or had strong ties to this class. Likewise, women coming from lower-​middle class origins also acquired more chances to be represented. In this context, the rate of those declaring themselves to be a civil servant, technician, nurse, public relations expert, or self-​employed together with those who did not state their profession is about 15 percent. Pro-​Kurdish parties played a great role in the increasing representation of lower classes:  those who did not reveal their former profession—​ these are also those who did not receive university education—​and those who declared to have been self-​employed are all members of the pro-​Kurdish parties.

Conclusion To conclude, a few points must be highlighted. First, although there have been significant fluctuations in the regional distribution of women deputies in Turkey, there is still a clear gap according to regions. The Marmara, Central Anatolian, and Aegean regions have the highest representation of women. The Marmara region has one-​third of all women deputies. The share of the coastal regions of the Mediterranean and Black Sea regions is below 10 percent each. On the other hand, with their lowest level of socio-​economic development, the Eastern and Southeastern Anatolian regions come to the fore with their relatively high rate of women deputies. Nevertheless, the proportion of women in every region remains below 10  percent within their respective regions. Another aspect of regional distribution of women deputies is the division along the metropolis-​countryside. Women were predominantly the representatives of the three metropolises. Although women played more roles in the representation of the countryside in the post-​1999 period, they have been more advantaged in the representation of the metropolises. Second, women deputies have been present in the political arena as both the symbols and actors of their gender identity, yet they have also been the symbols and agents of ideological, religious, and ethnic identities. While they were the symbols of westernization and modernization in the single-​party period, they were the symbols of democratization during the periods of military intervention. From the 1990s onward, on the other hand, they have been the symbolic agents of the repressed religious and ethnic identities. Third, while factors such as the choices of the electorate, the electoral system, and women’s demands have been influential, the most decisive factor in the political representation of women in Turkey and the gender profile of the TGNA seems to be the choices and attitudes of political parties. The fact that the RPP, the JDP, and pro-​Kurdish parties reserved relatively more seats for women during the single-​party and the post-​1999 periods, respectively, resulted in women’s growing representation in parliament. Among these

178  Mezher Yüksel political parties, one must especially point to the pro-​Kurdish parties as they are the only political movement that implemented the policy of equal gender quota in all the elections they participated in. Fourth, there emerged striking changes in the professional profile of women deputies throughout the Republican period. In line with the gender roles attached to women during the single-​party period, they were predominantly educators in this period. Following the transition to multi-​party politics, not only professional diversity went up but also the rate of white-​collared professionals increased. This shift indeed reflects a change in both the gender roles attributed to women and the level of women’s participation in social life in Turkey. Fifth, a notable point concerning women’s social characteristics is that the rate of married women deputies is relatively low. While the rate of married male deputies is above 95 percent, it reached 70 percent for women deputies only after the relative increase seen in the post-​1999 period. These figures suggest that there is a significant relationship between marital status and active political participation in Turkey and that it changes according to one’s gender. While being in a marriage is an advantage for male deputies, the same serves as an obstacle for women representatives. Indeed, field research undertaken in different periods also reveals that being married makes it difficult for women to actively participate in politics. Another point that differentiates female deputies from their male counterparts is the weak ties between their birthplace and constituency. While 80  percent of male deputies are representatives of their places of birth, it is merely around 40 percent for women deputies. A final remark needs to be made concerning the fundamental changes introduced in the Turkish political system after the constitutional changes accepted by the 2017 referendum. With the nationwide elections held in 2018, the new system called the presidential governmental system (Cumhurbaşkanlığı Hükümet Sistemi) has been put into practice. Following the first elections held in the new system, 104 women were elected to parliament. This is the highest number ever reached throughout Republican history. Hence, it seems that it can be argued that the change in the system will not influence women’s political representation in a negative way. Nevertheless, there is still a need for legal regulations as well as for changes in the policies of political parties in order to catch up with the goals set by the UN.

Notes 1 This article is based on a research project (#2016/027) funded by Kırıkkale University’s Scientific Research Projects Coordination Unit. 2 The association was founded by the name of the Turkish Women’s Union, which continued its activities until 1935, when women first entered the Parliament (Bora & Günal, 2011; Toprak, 1988). 3 They are: the Democrat Party (the DP), the Justice Party (the JP), the True Path Party (the TPP), the Motherland Party (the MP), and the Justice and Development Party (the JDP).

Engendering the parliament  179 4 These parties are: the Republican People’s Party (the RPP), the Social Democratic Populist Party (the SDPP), the Democratic Left Party (the DLP), and the Populist Party (the PP). The fact that all of the 59 women deputies were RPP deputies during the single-​party period and that the RPP has been the longest-​living political party in the TGNA played a role in this result. 5 The pro-​ Kurdish parties are:  the Democracy Party (the DEP), the People’s Democracy Party (the PDemP), the Peace and Democracy Party (the PaDP), and the Peoples’ Democracy Party (the PDP). 6 They are:  the Republican Peasant Nation Party (the RPNP), the Nationalist Action Party (the NAP), and the Good Party (the GP). 7 It must be noted that shifting between different parties was not limited to women deputies, it was a general tendency of the period (Turan et al., 2005). 8 Two points regarding this table need to be highlighted. First, the table about professions is based on the deputies’ own declarations. Nevertheless, some MPs stated their jobs (like mayor or academic) to be their profession. In such cases, rather than their statements, their education is taken as the basis for their professional classification. Secondly, some deputies pointed to more than one profession as their profession. If one takes into account both professions, then the number of professions exceeds that of deputies. For instance, since 44 MPs have two different professions, the total number of professions appears to be 642. Reflected on the column of proportions, this situation results in a rate that goes beyond 100 percent. This explanation is also valid for Tables 8.6, 8.8, and 8.11. 9 For a detailed analysis of the two-​phased simple majority system, see Uyar (1998) and Koçak (2005). 10 For the conception of society during the single-​party period as classless, unprivileged, and corporatist, see Parla (2006). 11 For a comprehensive assessment of the period’s peasantist ideology, see Karaömerlioğlu (2006). 12 They are as follows: deputy of Ankara, namely Hatı Çırpan (her original name Satı was changed by Atatürk to Hatı) and Şekibe İnsel, deputy of Bursa (Güneş, 2001). 13 The most consistent example of this strategy was seen in the Eastern and Southeastern Anatolian regions, characterized by their predominant feudal characteristics as well as long-​standing revolts against the center. For more on this point, see Mezher Yüksel (2007) and Demirel (2013). 14 According to the data provided by the 1935 censuses, the rate of Kurdish speakers is 9.2 percent (TC Başbakanlık İstatistik Genel Direktörlüğü, 1937, p.145). 15 For a variety of women’s movements called Kemalist, Islamist, Socialist, Marxist, and Kurdish, see Bora and Günal (2011), Metin Yüksel (2006), and Çağlayan (2013). 16 For a work on the Kurdish women’s movement, see: Çağlayan (2013). 17 To illustrate, approximately 45 percent of the candidates of the Peoples’ Democratic Party in the November 1, 2015 national elections were women. See: www.aljazeera. com.tr/​al-​jazeera-​ozel/​kadin-​vekil-​sayisi-​azaldi (Date accessed: May 25, 2019). 18 Another important case of women’s symbolic use in Turkish politics was Tansu Çiller, who would become the first female Prime Minister in Turkey after she became the head of the center-​right party called the True Path Party (Güneş-​ Ayata, 1998, p. 244). 19 Kavakçı recently was granted her citizenship and was appointed as the Turkish ambassador to Malaysia:  www.diken.com.tr/​buyukelci-​atamalari-​yapildi-​merve-​ kavakci-​da-​listede/​ (Date accessed: May 25, 2019).

180  Mezher Yüksel 20 www.youtube.com/​watch?v=zhQGlDyBIhU (Date accessed: May 25, 2019). 21 One of the women deputies wearing a headscarf is a member of the PDP, while the rest—​including Kavakçı’s sister Ravza Kavakçı Kan—​are members of the JDP. 22 For a study on the ethno-​secular boundaries of the modern Turkish identity, see Yıldız (2001). 23 The challenge at times also found its expression in the total dismissal of modern Turkey. The JDP deputy from the province of Balıkesir, namely Tülay Babuşçu, shared the following on her Twitter account: “90 year-​long ad break of the 600  year-​long Ottoman [imperial] era has come to an end”. This way, she provided a glimpse into the political-​ideological picture of the new period’s deputies. See:  www.cumhuriyet.com.tr/​haber/​siyaset/​185737/​AKP_​li_​vekil_​_​ Osmanli_​nin_​90_​yillik_​reklam_​arasi_​sona_​erdi.html# (Date accessed:  May 25, 2019). 24 For the first time throughout Republican history, the following provinces were represented by women deputies:  Ağrı, Aksaray, Amasya, Batman, Bilecik, Elazığ, Gaziantep, Iğdır, Kahramanmaraş, Mardin, Sakarya, Siirt, Şırnak, Uşak, and Van.

Bibliography Arat, Y. (1998). Türkiye’de Kadın Milletvekillerinin Değişen Siyasal Rolleri, 1934–​ 1980. In Berktay-​Hacımirzaoğlu, A (Ed.), 75 Yılda Kadınlar ve Erkekler, (pp. 249–​ 66). İstanbul: Tarih Vakfı. Bora, A., & Günal, A. (2011). 90’larda Türkiye’de feminizm. İstanbul: İletişim. Çağlayan, H. (2013). Analar, yoldaşlar, tanrıçalar: Kürt kadın hareketinde kadınlar ve kadın kimliğinin oluşumu. İstanbul: İletişim. Çakır, S. (2014). Erkek kulübünde siyaset. İstanbul: Versus. Demirel, A. (2013). Tek partinin iktidarı: Türkiye’de seçimler ve siyaset (1923–​1946). İstanbul: İletişim. Erdoğan, M.M., & Kaya, A. (2019). Türkiye’nin göç tarihi. İstanbul: Bilgi Üniversitesi Yayınları. Frey, F. (1965). The Turkish political elite. Cambridge, MA: MIT Press. Göle, N. (1998). Mühendisler ve ideoloji. İstanbul: Metis. Güneş-​Ayata, A. (1998). Laiklik, güç ve katılım üçgeninde Türkiye’de kadın ve siyaset. In Berktay-​Hacımirzaoğlu, A (Ed.), 75 Yılda kadınlar ve erkekler, (pp. 237–​48). İstanbul: Tarih Vakfı. Güneş, İ. (2001). Türk parlamento tarihi, TBMM V. Dönem II. Cilt. Ankara: TBMM Vakfı Yayınları. Kadıoğlu, A. (1999). Cumhuriyet iradesi demokrasi muhakemesi. İstanbul: Metis. Kalkınma Bakanlığı. (2013). İllerin ve bölgelerin sosyo-​ekonomik gelişmişlik sıralaması araştırması (SEGE-​2011). Ankara: Kalkınma Bakanlığı. Karaömerlioğlu, A. (2006). Orada bir köy var uzakta:  Erken cumhuriyet döneminde köycü Söylem. İstanbul: İletişim. Koçak, C. (2005). Parliament membership during the single party system in Turkey (1925-​ 1945), Journal of European Turkish Studies, 3. Parla, T. (2006). Ziya Gökalp, Kemalizm ve Türkiye’de korporatizm. İstanbul: İletişim. Saktanber, A. (2004). Kemalist kadın hakları söylemi. In İnsel, A. (Ed.), Modern Türkiye’de Siyasi Düşünce: Cilt: 2 Kemalizm, (pp. 323–​33). İstanbul: İletişim.

Engendering the parliament  181 Sancar, S. (2011). Türkiye’de kadın hareketinin politiği:  Tarihsel bağlam, politik gündem ve özgünlükler. In Sancar, S. (Ed.), Birkaç Arpa Boyu, (pp. 61–​117). İstanbul: Koç Üniversitesi Yayınları. Tachau, F. (1980). Parliamentary elites: Turkey. In Landau, J.M., Özbudun, E., & ve Tachau, F. (Ed.), Electoral Politics in the Middle East, (pp. 205–​43). London: Hoover Press Publication. TBMM. (2010). TBMM albümü 1920–​2010, 4 Cilt. Ankara: TBMM. —​—​—​. (2012). 24. Dönem TBMM albümü. Ankara: TBMM. —​—​—​. (2018). Türkiye Büyük Millet Meclisi milletvekilleri dağılımı. < www.tbmm. gov.tr/​develop/​owa/​milletvekillerimiz_​sd.dagilim> TC Başbakanlık İstatistik Genel Direktörlüğü. (1937). Genel nüfus sayımı. Ankara: Mehmet İhsan Basımevi. Tekeli, İ. (1990). Osmanlı İmparatorluğu’ndan günümüze nüfusun zorunlu yer değiştirmesi ve iskan sorunu, Toplum ve Bilim, 50,  49–​71. Tekeli, Ş. (1981). Women in Turkish politics. In Abadan-​Unat, N. (Ed.), Women in Turkish Society, (pp. 293–​310). Leiden: E. J. Brill. —​—​—​. (1982). Kadınlar ve siyasal toplumsal hayat. İstanbul: Birikim. Toprak, Z. (1988). Halk Fırkası’ndan önce kurulan parti:  Kadınlar Halk Fırkası. Tarih ve Toplum, 9(51),  30–​32. Turan, İ., İba, Ş., & ve Zarakol, A. (2005). Inter-​party mobility in the Turkish Grand National Assembly:  curse or blessing? European Journal of Turkish Studies, Thematic Issue N°3, Being a MP in contemporary Turkey. < www.ejts.org/​document400.html> TÜİK. (2005). İstatistik göstergeler 1923–​2004. Ankara: TÜİK. —​—​—​. (2012). Milletvekili genel seçimleri 1923–​2011. Ankara: TÜİK. Uyar, H. (1998). Tek parti dönemi ve Cumhuriyet Halk Partisi. İstanbul:  Boyut Yayınları. Yıldız, A. (2001). Ne mutlu Türküm diyebilene:  Türk ulusal kimliğinin etno-​seküler sınırları. İstanbul: İletişim. Yüksel, M. [Metin]. (2006). The encounter of Kurdish women with nationalism in Turkey. Middle Eastern Studies, 42(5), 777–​802. Yüksel, M. [Mezher]. (2007). An unintended consequence of modernization in Turkey:  Nationalist reactions from its periphery. Unpublished PhD Dissertation. Ankara: Institute of Social Sciences, METU. Zihnioğlu, Y. (2003). Kadınsız İnkılap. İstanbul: Metis.

Websites Al-​jazeera Turk. (2015). Kadın vekil sayısı azaldı. www.aljazeera.com.tr/​al-​jazeera-​ ozel/​kadin-​vekil-​sayisi-​azaldi Diken. (2017). Büyükelçi atamaları yapıldı: Merve Kavakçı da listede. www.diken.com. tr/​buyukelci-​atamalari-​yapildi-​merve-​kavakci-​da-​listede/​ YouTube. (2008). Bülent Ecevit:  Bu hanıma haddini bildiriniz. www.youtube.com/​ watch?v=zhQGlDyBIhU Cumhuriyet. (2015). AKP’li vekil: Osmanlı’nın 90 yıllık reklam arası sona erdi. www. cumhuriyet.com.tr/​haber/​siyaset/​185737/​AKP_​li_​vekil_​_​Osmanli_​nin_​90_​yillik_​ reklam_​arasi_​sona_​erdi.html#

9  An analysis of Turkish overseas electoral districts Opportunities and challenges1 Necati Anaz and Mehmet Köse

Introduction Over the last few decades, there has been a rise in the number of countries that have implemented external voting. Some of the nations who allow their diaspora2 to vote from abroad also elect representatives for the diaspora through an extraterritorial electoral district system. This new trend is considered an expansion of the democratic rights of transnational citizens (Laguerre, 2013; Collyer, 2014), an increase in diaspora policies, and an implementation of engagement practices for citizens living abroad. Currently there are 150 countries that regulate external voting rights (Fierro, 2016), and out of these nations 17 countries have overseas electoral districts, including France, Italy, Croatia, Macedonia, Romania, and Portugal in Europe; Algeria, Tunisia, Senegal, Cape Verde, Guinea Bissau, Mali, Mauritania, and Mozambique in Africa; and Colombia, Ecuador, and Dominican Republic in the Americas. Turkey is one of the nations that allows emigrants to vote from abroad. It has taken almost half a century to fully implement this practice. Currently, the debate is in its next phase of whether to allow Turkish emigrants to be represented directly in the national parliament. This paper, for that reason, explores the systemic consequences of adopting an overseas electoral district system and what possible opportunities and challenges lay ahead if Turkey implements such a system. To set the stage, in this paper we will examine the discussions about Turkey’s overseas voting practices beginning in 1965, as well as emigrant participation rates in those elections. Subsequently, we will discuss ways in which voting from abroad can be turned into an overseas electoral district system that may cultivate significant privileges for Turkey and the Turkish diaspora.

Voting from abroad Although studies about elections and electoral behaviors have a significant place in electoral geography literature, research on diaspora representation systems is relatively new (Honohan, 2011; Lafleur, 2013; Shain & Barth, 2003; Tager, 2006). When we examine electoral research studies about overseas

Turkish overseas electoral districts  183 voting and diaspora parliamentary representation, we see that studies on these subjects are limited in number and are confined to a narrow area. In general, the literature on external voting focuses on four study areas: normative political theory (Gamlen, 2015; Lafleur, 2011), comparative studies (Rojas, 2007; Laguerre, 2013; Farmani & Jafari, 2016), voting practices (Brand, 2010), and electoral behaviors (Jaulin, 2015). Normative political studies include discussions about permitting or denying extraterritorial citizens the right to vote from abroad (López-​Guerra, 2005). In this debate, some argue that voting abroad has its own challenges, including the risk of fraud, the cost of the electoral procedure, and the “threat to sovereignty when the diaspora outnumbers the domestic population” (Jaulin, 2015, p. 2). Additionally, the topic of overseas voting generates discussions on citizenship and loyalty to a sovereign state (Laguerre, 2005). Ideally, a citizen can vote in an election if they hold citizenship in a state or at least wish to remain a permanent resident of the country. In this matter, Rainer Baubock draws the line to determine if one’s interest in their own state is threatened. Baubock argues that those who live abroad temporarily or become a refugee should have rights to vote in the elections at home because “they still have a significant stake in the country, through family, regular visits, property, and often the hope of returning one day” (Collyer & Vathi, 2007, p. 8). Baubock conceptualizes this as stakeholder citizenship, meaning that those who do not have stakes in their country of origin should not be permitted to vote from abroad (Bauböck, 2007). For this reason, some states exclude citizens who spend certain amount of time abroad from voting. For instance, after 15 years abroad, citizens of the United Kingdom cannot vote from abroad in homeland elections. Comparative studies on overseas voting probe into questions, such as, why is external voting permitted and how it is executed? Jean-​Michel Lafleur (2013) gives three reasons for states to grant their citizens the right to vote from abroad. The first one is emigrants’ financial contribution to their state through remittances. For some states, remittances are a major part of the GDP. For example, Colombia’s emigration engagement policies include voting from abroad, which is to some extent tied to the continual flow of remittances. Secondly, some states enact domestic policies that necessitate the inclusion of their citizens who live abroad. In other words, states work to attract diaspora’s contributions at times of political reform and democratic transition. During the Arab Spring, some states (Egypt, Tunisia, and Libya) in the Middle East and North Africa integrated diaspora policies that included extraterritorial voting (Baradei et al., 2012). Related to the first two reasons, when citizens living abroad lobby and publicize their ethnic and cultural connection to their mother state, the state is more likely to extend voting rights. Research shows that emigrant organizations and representatives that are active in pursuing extraterritorial voting and political representation policies oblige their mother states to enact emigration policies. Laurie A. Brand argues that when the diaspora population reaches a critical number that the state cannot ignore, the state is more likely to try to develop emigration policies (Brand, 2014). Brand

184  Necati Anaz and Mehmet Köse also emphasizes that if a political party believes that emigrants will support their party more often than their competitors, they are also more likely to lobby for developing emigration policies (Brand, 2014). The literature also includes “the heterogeneity of the rules and the mechanism” of external voting (Jaulin, 2015 p.  3); in particular, registration, casting ballots, and counting the votes from abroad (Erben et al., 2012). This encompasses the technicality of voting from abroad and includes topics such as how voting procedures and the proximity of the constituency to the voting locations affect turnouts. Finally, the literature examines questions about how the overseas constituency behaves when they vote from abroad. In this sense, scholars explore external voting using electoral sociological methods to explain how immigrants’ electoral behavior changes depending on pre-​migration variables, such as their socio-​demographic criteria, region of origin, and post-​migration variables such as legal status in the host country, length of time abroad, employment, integration, and discrimination (Jaulin, 2015, p.  3; Mencutek, 2015). Additionally, the literature on voting from abroad also includes discussions related to citizenship studies. Specifically, the way that states are keen to build long-​term relationships and to create and maintain links with their emigrant population around the world to foster a common identity (Collyer, 2014). Collyer’s territorial conceptualization model of citizenship can be grouped into four perspectives:  (1) bounded citizenship in which the understanding of citizen–​state is a territorial relationship and is restricted to a single state; (2) post-​national citizenship in which the citizen–​state territorial relationship is irrelevant as citizens and non-​citizens enjoy similar rights; (3) transnational citizenship in which the citizen–​state relationship is irrelevant as state’s territory follows citizens; and (4) extra-​territorial citizenship in which the citizen–​ state relationship is understood to be central and fluid (Collyer, 2014, p. 61). Our study falls in Collyer’s fourth model and emphasizes that citizenship is not completely bound because certain rights, including voting from abroad, are maintained beyond the territory. To see the worldwide trends in electoral applications, we must look at Collyer’s classification of global voting systems. This classification can be separated into five points:  (1) countries where no elections are held, (2) elections are held but expatriates are prevented from voting, (3) diaspora must return to their country of origin in order to vote. In this case, expatriates need to return to vote in the district where they were most recently resided. (4) Citizens can vote from abroad in polling stations located in the embassy or other diplomatic missions, but their votes will be counted in the district where they most recently resided or their vote will be counted and distributed among political parties’ nationwide scores. (5)  Expatriates elect their own representatives in an overseas electoral district (Collyer, 2014, p.  64). This paper explores implementing the fifth point in Turkey by analyzing voting practices and voter participation rates in other countries.

Turkish overseas electoral districts  185

Voting from abroad in the Turkish case Although Turkey has been discussing external voting for almost half a century, substantive academic literature only began to appear after 2014 with the first voting implementation abroad. Thus, the few numbers of studies that appeared focused on Turkey’s first experience of overseas voting (Sahin-​ Mencutek & Erdoğan, 2015; Yıldırım, 2016). For example, Şahin-​Mencütek and Erdoğan test three hypotheses that apply to Turkey’s implementation of voting from abroad:  (1) “migrants from Turkey have long been active in making demands and lobbying for VFA (voting from abroad)”, (2)  “the implementation of VFA reflects the aim of home states to maintain ties with and cultivate loyalty among citizens and co-​ethnics abroad”, (3)  the ruling Adalet ve Kalkınma Partisi (Justice and Development Party, the JDP) “might have utilized VFA in an attempt to secure a majority vote for its presidential candidate” (Sahin-​Mencutek & Erdoğan, 2015, p. 176). In their conclusion, the authors discuss their hypotheses and how they support their testing objective. However, referencing the same literature, the authors downplay the last decade’s worth of comprehensive diaspora policy development. We argue that diaspora policies implemented over the last decade are an obvious result of domestically obtaining political stability and economic development. Şahin-​Mencütek and Erdoğan fail to examine how Turkey’s diaspora engagement policies parallel with their policy development in the areas of social, cultural, economic, and foreign policy, that began with the governing Justice and Development Party (Anaz, 2018). Related literature also ignores the eagerness of the Turkish diaspora to be able to vote in homeland elections throughout emigration. After the introduction of voting at the customs in 1987, Turkey witnessed gurbetçi3 convoys headed to Turkey’s borders to cast their votes. Şahin-​Mencütek and Erdoğan’s arguments that the governing JDP prioritized voting from abroad for possible electoral advantages is misleading. This argument either ignores or is unaware of initiatives taken by members of parliament since 1965 to legislate external voting, which will be detailed below. Besides, external voting is not limited to presidential elections; parliamentary elections and referendums are also included in the practice. More importantly, the claim that it was done for an advantageous election result is groundless, as it was in the case of Italy. As we saw in Italy in 2006, emigrant votes shifted the outcome of homeland elections when incumbent Prime Minister and leader of the center-​right coalition, Silvio Berlusconi, was voted out of office. Emigrant votes do not always follow right-​wing politics or always lean toward the governing party (Anaz, 2018). Similarly, Deniz Yıldırım investigates external voting behaviors in the June 7 and November 1, 2015 general elections. Yıldırım elaborates on the differences and similarities in external and domestic voting patterns and how party performances, in the case of external voting, vary according to the

186  Necati Anaz and Mehmet Köse country. Yıldırım concludes that when emigrants go to the polls, their initial preference leans toward Turkey’s governing JDP and the third largest party, Halkların Demokratik Partisi (the Peoples’ Democratic Party, the PDP). Yıldırım states in most countries that the JDP wins the most votes with a large margin but in a few countries, the PDP takes the most votes with a small margin. As we examine the related literature on Turkey’s voting from abroad, we notice that the few studies that are available on overseas voting are very recent and do not analyze external voting matters in a comprehensive manner. This shows the depth of diaspora studies in Turkey. Although Turkey’s emigration policies go back half a century, academic research on the notion of Turkish diaspora and diaspora policies is very recent. Recently, Turkey’s transition to external voting practices has attracted some researchers to this area. Alongside the discussion on external voting, the issue of a diaspora electoral district is raised from time to time. Most recently was during the 2012 parliamentary discussion (TBMM, 2012). Some parties included it in their election program in 2015 as well. As of yet, academia has not paid attention, so the research on this is almost nonexistent. Research on the implication of Turkey’s shift to an overseas electoral district system does not exist. For this reason, this paper is an invitation to explore the advantages and disadvantages of establishing an overseas electoral district in Turkey. As of today, Turkey does not recognize overseas as a separate electoral district. This paper emphasizes the necessity to create and analyze scholarly discussion about the opportunities and challenges that emerge from creating an overseas electoral district, considering the record-​breaking overseas voter turnout in the past few elections.

Turkey’s diaspora policies and overseas electoral system A major turning point in the history of Turkey’s electoral system was when Turkish emigrants obtained the right to vote from abroad in 2012. However, the diaspora’s participation in Turkish elections was possible only in their residential district in Turkey until 1987 when voting was introduced at customs gates. For the first time in 2014 at 103 polling stations in 54 countries, Turkish citizens living abroad were able to vote in the presidential election from where they live. After the August 2014 presidential election, the first parliamentary election that allowed voting from abroad was in June 2015, followed by the November 2015 parliamentary reelection. In 2017, almost 1.5  million emigrants became part of a historic referendum that changed Turkey’s governmental system from parliamentary to presidential. The latest election was in June 2018 to elect the president and members of parliament in the new system, and the diaspora participation rate was more than 50 percent. This makes a total of five elections in five years. During this series of elections, there was some criticism in handling elections referred to the

Turkish overseas electoral districts  187 Supreme Election Council (YSK), but overall the voting process in more than 50 countries and custom gates was successfully executed. Although mass emigration began in the 1960s, there was a 50-​year delay in implementing external voting. To explain this, we need to look at both home and host country perspectives. First, in the beginning, most labor migration agreements between Turkey and western European countries were not planned for long-​term migration. Two-​year rotated migration was planned and guest workers traveled alone without family. Second, the political situation in Turkey was not stable and the democratic process was halted by a coup d’état and military interventions. Third, Turkey’s economic capacity and logistical issues did not allow planning expensive and complex external voting operations. Fourth, Germany was reluctant to allow 1.5  million Turkish voters to pour into ballot boxes on its soil. Official labor migration was stopped in 1973 with the oil crisis, but migration continued through family unification, students, and asylum seekers. The first appeal to allow emigrants to vote in homeland elections was in the late 1970s. However, at the time, violence within the country and political instability prevailed during and after the 1980 military coup, and many fled to Europe. In this era, Turkey’s emigration policy was obscured by security concerns and the desire to tap economic resources. Host countries’ attitude toward foreigners in general also changed in the 1980s. During this time, some host countries spent considerable effort to encourage immigrants to make a definite return to their homeland, but these incentives did not receive enough attention. Host countries like France, Germany, and Austria were forced to acknowledge that immigrants were there permanently. In light of this new situation, both the host and migratory countries began to seek new policies. The process of integration in the host countries was formalized in policies that encouraged voluntary assimilation. As homeland country, Turkey began to employ cultural, educational, and social programs for the Turkish migrant community, and over the years Turkey’s gurbetçi policies evolved to another level and institutionalized both at home and abroad. Political participation and the utilization of the diaspora’s democratic rights entered into a new phase with the 1987 policy changes. In these years, Turkey witnessed a transformation in the political arena from the post-​ military era to extended political freedoms and economic transformation to more liberalized policies. In 1987, for the first time, citizens had the opportunity to vote at customs gates in the general election. In this election, about 40 thousand Turkish emigrants exercised their right to vote. In the following years the turnout increased to several hundred thousand. In the 2007 election, the turnout reached close to 228  thousand. However, considering the total Turkish diaspora is about six million, these figures remain insignificant. A constitutional amendment was required to extend external voting beyond borders. In this sense, in 1995, Article 67 was changed to allow

188  Necati Anaz and Mehmet Köse voting from abroad. These changes came with other democratically desired amendments:  the voting age was reduced from 21 to 18  years old, and the restrictions placed on the activities and programs of associations, foundations, and unions were abolished. The diaspora’s demands and democratic openings in many areas came into popular awareness and were the preconditions that lead to external voting in Turkey. In this paper, we also need to highlight that external voting in Turkish national elections initially was pushed through from two directions: one came from Turkish workers (gurbetçi) in Europe and the other came from Turkish legislators. Emigrants demanded that they needed to be included in the democratic process in Turkey. For this purpose, emigrants collected signatures to be sent to Ankara and asked for their external voting rights to be granted. A petition sent to the Turkish National Grand Assembly included a Turkish worker’s demand for the right to vote from Germany. The letter sent in 1975 asserts that extraterritorial voting is a constitutional right and the natural component of universal human rights (Özhan, 1975). Beyond individual submissions, the gurbetçi organizations in Germany pioneered by requesting voting privilege from Turkish politicians and legislators. The reason behind asking for such privilege was interpreted to be that the right-​to-​vote may be the shortest way to problem-​solving the issues that emigrants encountered, especially during the consulate affairs in Turkish diplomatic spots and when they feel the need for fewer bureaucratic obstacles during visits to Turkey. The second direction for the rights for external voting came from individual exertion from a few legislators, especially senator and later parliamentarian Reşid Ülker of Cumhuriyet Halk Partisi (Republican People’s Party, the RPP). Senator Ülker repeatedly brought the emigrants’ voting rights issue to the attention of the parliament beginning as early as 1965 (Ülker, 1965). Ülker reasoned that the right to vote from overseas is part of emigrants’ constitutional rights and Turkey’s only way to connect with its diaspora. Ülker never saw the Turkish diaspora in Europe as a lost population or as permanent settlers in Europe. He thought that they always had a stake in their country of origin and that Turkey should give them rights to be part of domestic politics for the assertion that emigrants will return to their soil in the end. Considering Turkey’s test with military interventions almost every ten years in its democratic life, giving the right-​to-​vote to emigrants always became a matter of security. Especially after the number of political asylum seekers reached a critical point, security bureaucrats in Turkey lobbied legislators not to open the door for such a risky current in the Turkish electoral system. Unspoken rationality of such resentment was that the number of the Turkish minority’s electoral choice could have jeopardized the Turkish political system and controlled democracy. However, a high number of politicians and emigrant organizations disputed such claims that extraterritorial voting should not be a matter of security. Furthermore, they insisted that asking emigrants to invest in Turkey without compensating their right-​to-​vote seemed duplicitous. For example, in his speech during one of the constitutional committee

Turkish overseas electoral districts  189 meetings in 1979, Senator Ahmet Tahtakılıç emphasized that external voting remains to be a matter of constitutional rights and its implementation can only strengthen Turkish democracy. He further critiqued those who opposed such regulation for being double-​tonged, saying they could approve external voting when they visit workers in Europe but backtrack when they return Turkey. Tahtakılıç also comments that descending upon workers savings like an eagle but not granting vote-​to-​right remains unacceptable (Tahtakılıç, 1979). As it is understood, political debate on external voting began in the early 1960s but putting it into practice took decades.

Polling stations beyond the border Although external voting was constitutionally allowed since 1995, the legislative changes to implement it only came in 2008. Political turmoil and an economic crisis in the second half of the 1990s and early 2000s were reasons these regulations were delayed. With the changes in the electoral code in 2008, an overseas electoral board was established under the Ankara Provincial Electoral Board. Voting procedures were defined and three voting methods were introduced; electronic, postal, and ballot box. However, the constitutional court later cancelled postal voting because of security and voting privacy concerns. After these amendments to the electoral code, the Turkish diaspora had to wait for another amendment to be part of their homeland’s democratic process. In the 2010 referendum and the 2011 parliamentary elections, the YSK ruled out external voting due to legislative shortcomings and the position of the Ministry of Foreign Affairs. Execution of elections did not go as planned because the electoral code gave power to the YSK, with the counsel from the Ministry of Foreign Affairs (MFA), to decide if external voting could take place, where ballot boxes could be located, and the length of time they could be installed for. To ensure the participation of citizens living abroad and to strengthen democracy in Turkey, elections and polling issues continued to be a crucial topic in the political arena and the political rights of citizens living abroad were indeed gaining traction. The Presidency for Turks Abroad and Related Communities (YTB) was established in 2010 and became the first to take ground-​breaking and concrete steps toward the implementation of a comprehensive Turkish diaspora policy. The YTB played a vital role in pushing for the implementation of external voting as a diaspora agency. However, the Supreme Election Council rejected voting abroad in 2011 for parliamentary elections because of the inadequate legal and administrative infrastructure. After the June 2011 elections, in coordination with the YTB and the government diaspora agency, comprehensive amendments were made to the electoral code, which paved the way for external voting in the 2014 presidential election. The presidential election in August 2014 was the first time external voting was accomplished, and the president was elected by popular vote for the first

190  Necati Anaz and Mehmet Köse time in Turkish democratic history. The turnout in this election was not as high as expected. Over half a million diaspora members casted their votes abroad and at customs gates. The low turnout was blamed on the appointment system, the host countries’ attitude, and the summer holiday. In Germany, for example, half of the 1.4 million registered voters were allowed to vote at only seven polling stations. Voters had to travel hundreds of kilometers to reach the polling stations, which in turn, negatively affected participation. The inadequate number of polling stations for the diaspora also adversely affected voter turnout. According to electoral code, it was optional to require voters to get an appointment to vote if there was a risk of overcrowding, but the YSK applied the appointment system to all overseas voters in 53 countries and required them to register for an appointment. Only a few exceptions were granted. It was also advised for all emigrants to make their appointments through the internet, which made the procedure unnecessarily complicated. Another reason for the low turnout was the summer holiday; almost a million people from the diaspora intended to spend their holiday in Turkey. This was obvious from the turnout figures at the customs gates. More than half of the diaspora casted their votes at the customs offices. Despite all this, voter turnout was still two times higher than the previous highest participation. In the June 2015 parliamentary election, over a million external votes were cast. This was twice the previous years’ election participation. The abolishment of the appointment system was remarkably visible in the results. Political parties were also more active because the outcome of the election would have an impact on the number of seats, unlike in the 2014 presidential election. Election participation continued to increase in the November 2015 elections; from 1.05 million votes cast from abroad in June to 1.3 million in November in 2015. The diaspora watched closely from abroad and casted their votes in an effort to help resolve the political deadlock in Turkey (See Table 9.2). A record 1.42 million voters showed up at the polls in April 2017 for the referendum (See Table 9.3). After the attempted coup d’état and the referendum decision to change the governmental system, the diaspora once again turned up at polling stations. In the 2018 presidential and general elections, Turkey witnessed another remarkable voter turnout (See Table  9.1 and Table  9.4).

Table 9.1 2014 and 2018 presidential elections 2014 Custom Gates Number of Registered 2,780,739 Number of Voters 297,340 Voter Turnout 9.3%

2018 Overseas

Custom Gates Overseas

2,780,739 232,795 8.4%

3,044,837 168,031 5.5%

3,044,837 1,358,584 44.62%

Turkish overseas electoral districts  191 Table 9.2 2015 general elections June 7

Number of Registered Number of Voters Voter Turnout

November 1

Custom Gates

Overseas

Custom Gates

Overseas

2,866,979 124,432 4.3%

2,866,979 931,646 32.5%

2,899,069 138,454 4.8%

2,899,069 1,159,871 40.0%

Table 9.3 2017 referendum results

Number of Registered Number of Voters YES NO Voter Turnout

Custom Gates

Overseas

2,972,676 98,597 54.2% 45.8% 3.3%

2,972,676 1,325,682 59.5% 40.5% 44.6%

Table 9.4 2018 general election results

Number of Registered Number of Voters Voter Turnout

Custom Gates

Overseas

3,044,837 168,075 5.5%

3,044,837 1,357,204 44.6%

Participation came from 60 countries in 123 diplomatic missions with 3,386 ballot boxes and 34 border posts with 4,098 ballot boxes (YSK, 2018). This is a rare occurrence in today’s elections around the world. Compared to other diasporas’ turnout in homeland elections, the Turkish diaspora is far more active than any other nation. For example, in Mexico, because of the complicated voter registration requirements, only 667,000 of the 12 million Mexican diaspora registered for the July 2018 election. Only 181,000 of them collected their registration card and nearly 100,000 cast their votes. In Brazil, the number of emigrants registered to vote from abroad is very low, but their participation rate is high. Only 135,000 of Brazil’s 3 million diaspora are registered to vote. In the 2014 Brazilian presidential elections, 95  percent of the registered diaspora voters showed up to vote. In Italy’s March 2018 elections 1.25 million diaspora voted. This was only a 30 percent participation rate, even though Italy reserves 12 seats in the parliament and six seats in the senate for diaspora representation. Participation rates are crucial to our analysis because in the next section we will build our arguments based on overseas participation rates and their

192  Necati Anaz and Mehmet Köse 16,00,000 14,00,000 12,00,000 10,00,000 8,00,000 6,00,000 4,00,000 2,00,000 0

7 June 2015

1 November 2015 16 April 2017 Registered Voter

24 June 2018

Turnout

Figure 9.1 Participation in Germany

ability to impede the electoral process if votes are distributed among the overall results. If overseas participation is considered a separate electoral district in Turkey, it would be the fourth largest district after İstanbul, Ankara, and İzmir. This is partially why, when we analyze an electoral district system for Turkey, we put more emphasis on the overseas participation rate. When delving into the country-​base data set (Figure  9.1), it can be seen that citizens who live in large communities showed the greatest interest in participation. For example, in Germany, more than half a million voters have participated in elections since 2015. When the data is detailed, consulate-​base turnout numbers show that the June 24, 2018 presidential election had the highest participation rate for an election since external voting was implemented in 2014. Among the consulates, Essen region received the highest turnout in the presidential election (Figure  9.2). In the Essen Consulate, President Erdoğan received 76.2 percent of the votes. This turnout for the president is even higher than the total number that his party (the JDP) received (67.49 percent ) and the number of votes in the April 16, 2017 referendum in which the JDP and the NAP coalition received 75.89 percent of the vote in Essen. Except a slight decline in Nuremberg city, among all other districts, participation increased in Germany. These numbers tell us that turnouts for external voting were impacted by institutionalization of the diaspora and the density of Turkish emigrants in the consulate region. This is clearly seen in the case of United States of America (USA). The turnout in the USA remained just below 30 percent as opposed to above 50 percent in Germany.

Discussions on an overseas election district system for Turkey The introduction of external voting and diaspora electoral district practices have gained more ground over the last half century, which coincides with the

Turkish overseas electoral districts  193 60,000 50,000 40,000 30,000 20,000 10,000 0

7 June 2015

1 November 2015

24 June 2018

Figure 9.2 Consulate base turnout in Germany

dramatic expansion of the diaspora phenomenon and the embarkment of diaspora policies by states. The current literature on diaspora policies based on extraterritorial citizenship presents us very little opportunity to develop a scholarly concept of external electoral geography (Warf & Leib, 2011). Furthermore, comprehensive research on diaspora representation systems related to extraterritorial citizenship remains limited (Laguerre, 2013). This paper does not specifically focus on the theoretical aspect of diaspora policies but instead aims to contribute to the discussion on diaspora policies by studying external voting and diaspora representation debate and experiences in the Turkish case. As Michel Laguerre (2013) puts it, moving the boundaries of democratization beyond national borders is now a generally accepted trend. In political decision-​making, it is essential for countries to include the participation of citizens in the homeland as well as those living abroad, regardless of their reasons for emigration. This political participation is generally granted to citizens living abroad by recognizing their right to vote. At the end of the 20th century and early 21st century, we began to witness various systems that allowed citizens living abroad to take their participation one step further in the political decision-​making process (IDEA, 2007). In the new system, 17 countries allow citizens living abroad to choose their own representatives directly in the homeland parliament. This new practice plays a vital role in increasing extraterritorial citizens’ participation in homeland politics and in the democratic process, but more importantly, it establishes a new transnational space alongside cultural, social, and economic spaces between homeland and diaspora. According to Laguerre, political participation is expanding in magnitude by shifting citizenship across borders and also increasing countries’ commitment to their citizens (Laguerre, 2013). As countries’ cross-​border political activities increase, the rhetoric based on the notion of temporariness

194  Necati Anaz and Mehmet Köse decreases and emigrant/​guest/​expatriate is being replaced by diaspora, which emphasizes permanency of living beyond homeland territory. This is also true in the Turkish experience: the development of diaspora policies and the increasing of the myriad of transnational spaces has resulted in a decrease in the discourse of “gurbetçi/​emigrant” and an increase in the usage of “Turkish diaspora”. This can be seen in the election programs of the major political parties as well. However, a representation system that transcends borders must not only be seen as an adaptation that concerns the homeland. In this sense, we argue that Turkey’s decade-​long diaspora politics must meet with spatial organization in the form of a representation system. Although emigration increased in numbers in the second half of the 20th century, neither side of the process thought that this would be for long, including the emigrant/​gurbetçi him/​herself. Gradual change in this thinking over the next decades has been reflected in the states’ policies as well as in individual’s behavior. The Turkish state authorities’ engagement with its citizens living abroad has been the policy of many governments in one way or another. Economic gains were a major motive most of the time. Cultural and social engagement and security perspectives were also prevalent in some periods. However, a comprehensive approach to the diaspora as a state policy is a recent phenomenon. Beginning with Turkey’s political stabilization, economic development, and openings in foreign policy during the first half of JDP governance, diaspora-​related issues captured attention of state elites as well as emigrant associations and diaspora leadership. The necessity of establishing a state institution with the task of drafting policies and coordinating the implementation of these policies directly under the prime minister’s office was one of the critical recommendations made by the parliamentary committee, which was mandated to prepare a report on the situation of Turkish citizens living abroad in 2003 (TBMM, 2003). This political pressure on the government from both sides paved the way for institutional formation and allowed the creation of a presidency for the diaspora on April 6, 2010. This was an effort to strengthen Turkey’s ties and establish closer economic, social, and cultural relations with citizens living abroad and related communities (YTB, 2018). The institutionalization of the Turkish government’s diaspora policy through the YTB resulted in solving issues and meeting the long-​awaited demands of the diaspora. The implementation of external voting was one of them. The issue of taking ballot boxes to the diaspora was another important task the YTB focused on for a couple of reasons: first, it was the long-​time desire of the diaspora to have such a system, and second, to make diaspora issues in elections matter so that the political parties who form the government would have to take diaspora-​related issues into government programs (Kose, 2018). Beginning with the June 2015 elections, all major political parties made considerable election promises for diaspora programs and this was also reflected in government programs. Education, Turkish language, cultural

Turkish overseas electoral districts  195 identity, social security benefits, and overseas representation were the significant issues for the diaspora that appeared in party programs. In the June and November 2015 elections, two major rival parties promised to reserve seats in the parliament for diaspora. The RPP promised to allocate one seat for every 300,000 electorate, which makes ten seats altogether. The JDP did not identify any figures but after the June 2015 elections, JDP İstanbul MP, Mustafa Yeneroglu (who is himself from the diaspora) and his colleagues proposed 15 MPs to be allocated for the diaspora. However, in the last June 2018 election, the JDP omitted an electoral district from its election platform, unlike the RPP. Even though the number of seats in parliament went from 550 to 600, none of them were reserved for the diaspora. When we examined the topics brought to the attention of the Grand Assembly, we found none of the major parties brought up the issue of a separate electoral district. We found that the parties purposely froze the case indefinitely. Especially disappointing was the governing party’s (JDP) indifference to the issue. Parties still do have a diaspora representation system attached to their party programs, but their willingness to bring the case to law makers’ consideration seems dormant. However, we continue to believe that the major political parties will recognize the importance of a diaspora representation district system and it will likely be the topic of their party agendas once again. Albeit, parliamentary initiatives begun in 1965, such as legal steps and practical implementation of external voting with conventional methods were only successfully implemented in 2014—​to come to this level took half a century. The introduction of voting at customs gates in 1987 was not a big step although it was considered a partial solution. Before that, anyone who registered on the electoral list could only have had a chance to vote in their residential district in Turkey. A person eligible to vote from the diaspora still had to travel all the way to Turkey to vote, to their hometown only on the election day, or to any border within 75 days. Constitutional amendments to clear some hurdles were made in 1995, but a change to the electoral act by the YSK was not made until 2008, which wasn’t adequate to implement voting abroad. 2012 amendments to the electoral act made it possible to practice external voting. The question of allocation of the votes casted by the diaspora has been debated all the way. Some members of the parliament as well as the diaspora itself raised the issue of a special electoral district for the diaspora. As studies on extraterritorial citizenship suggest, Turkey needs to implement political participation of citizens living abroad in a holistic way. The right to vote shall be complemented with the right to elect their own representative directly. This will not only enhance the belongingness of the Turkish diaspora but also mobilize political parties to engage more with the diaspora and its agenda. In other words, establishing an overseas electoral district system will empower the diaspora and make their case stronger in Turkish politics. The introduction of a diaspora representation and establishing an external electoral district is the next phase of a process that has already begun and that opens space for the diaspora in the Turkish Political system. As briefly

196  Necati Anaz and Mehmet Köse mentioned above, the participation of the Turkish diaspora is one of the highest in terms of turnout number and percentage among its peers. Since 2014, the number of diaspora votes and the participation percentage increased dramatically and passed 1.5 million and 50 percent respectively. To respond to the enthusiasm shown by the Turkish diaspora in the elections entails the introduction of an external electoral district that major parties have promised to add to their election programs. The current distribution system of votes casted abroad fails to fairly reflect voter choice. By implementing diaspora representation, this faulty side of the electoral system will be fixed. In the last three parliamentary elections, 17 members of parliament were shifted among parties after the distribution of diaspora votes. External votes are added on top of the party votes in each of the 85 domestic electoral districts according to specific calculation methods. Three seats in June 2015, three seats in November 2015, and 11 seats in June 2018 shifted between parties after overseas votes were distributed. For example, in the last election Çanakkale, İstanbul, Hatay, Hakkari, Osmaniye, and Aydın were some the electoral districts whose residents’ preference was shifted after the distribution of external votes. Establishing a special electoral district for external votes would have kept the preferences of these districts’ voters, which is one of the principles of democratic representation. Not having a special district for diaspora votes first and foremost shortens the diasporas democratic participation by not knowing for whom they vote other than political parties. On the other hand, it has negative impact on other domestic voters’ preferences, which they would have never imagined and thought. Some political parties nominate candidates among diaspora members in domestic districts to attract more diaspora support. If elected, that person officially represents İstanbul or Ankara in parliament, but practically an MP is asked to represent the diaspora. This way the system puts more responsibility and burden on the MP who is supposed to represent the diaspora but at same time is responsible for his electoral district, which often has different needs. This job needs the representation and responsibility of two distinct people, so their agendas and expectations are completely separate. Another question raised here is the effectiveness of nominating someone who is unfamiliar with a domestic district and unknown to residents. This adds another level of violation of political choice in that electoral district. The introduction of an external district would eliminate another electoral practice defect, which is the inability to vote for an independent candidate. The diaspora can only vote for political parties for practical reasons. The current system distributes diaspora votes to 85 domestic districts, which makes it impossible to present independent candidates to external voters. Therefore, with the establishment of a special electoral district, people can run as an independent from the diaspora and the diaspora can vote for independent candidates as well. Above all, the Turkish diaspora eagerly expected to vote for over half a century, to make their voices heard, and to have their problems, issues, and

Turkish overseas electoral districts  197 ideas addressed. Since 2014, they have heartily participated in every election, and participation has increased each time. Political right should include both electing and being elected. The current picture does not fill the void. There are a few MPs from the diaspora elected in local districts, but this fails to address the diversity of the diaspora and the dispersal of issues and does not fairly represent 1.5 million diaspora votes. Alongside aforementioned arguments for an external electoral district, there are also several challenges associated with the implementation of this kind of system. One of the most debated challenges is associated with the practicality of voting from abroad. The cost of organizing an election abroad is one of the most visible challenges. Critics argue that votes cast from abroad cost more than those cast in the country of origin (Laguerre, 2013). In other words, voting from abroad is a luxury democratic practice for Turkey (Anaz, 2018). This might seem to be a reasonable argument, but we argue that there is not a price that can be placed on a more representative democracy. If the constitution mandates the right to vote for every citizen, then the geography of their presence should not be up for debate. Another practical challenge that voting from abroad faces is security, because voting takes place in a vicinity that is not controlled by the home country. According to the law enforced by the YSK, no party is allowed to campaign on Election Day. However, in some countries, campaign tables were set up outside of election centers on the day of election, openly violating the law. In this scenario, there is little that the Turkish authorities can do because outside of the ballot station is not controlled by Turkish authorities. When the voting site is outside of Turkish diplomatic missions, Turkey has no sovereignty to intervene and prevent possible breaching. Related to the security issue above, candidates’ election campaigns abroad can be perceived as violating the sovereignty of another country. In the 2015 elections, some European states tested already shaky bilateral relations by blocking official visits from Turkey and turning these visits into sovereignty and security issues. Although it is clear that Turkish election law strictly forbids any campaigning outside of its borders, some European states interpreted even the regular official visits from Turkey as part of the election campaign. This caused further tension between the states. Now that more states are implementing voting from abroad, this matter can be managed through bilateral responsibilities and diplomatic understandings. Other nations locate ballot boxes outside of their diplomatic missions in Turkey, so any matter regarding sovereignty and security can be resolved with mutual respect.

Conclusion In this chapter, we argue that the voting from abroad system needs to be taken a step further to establish a Turkish diaspora electoral district. Turkish emigrants’ participation and longtime demand for such a system deserve to be heard and put into action accordingly. When we compare the overseas

198  Necati Anaz and Mehmet Köse participation rates of the 17 countries that already have a diaspora electoral district, we see that the Turkish emigrants’ participation compellingly exceeds all of them. However, it should be noted that the gradual increase in the participation rate may not continue to increase unless election promises about a special election district for emigrants is fulfilled. The overseas voter turnout and participation rate, for example, in Croatia, France, Portugal, and Italy is much less than the overseas voter turnouts in Turkey, even though they have a diaspora representation system. Turkey has the capacity to implement such a system and to have an active diaspora to participate in the effort. Further analysis of the countries who implement external representation systems negate the perception of diaspora policy, implementing states as developing, democratically underdeveloped, and nationalist. The 17 countries listed above that have diaspora seats in their parliament are in Europe, Africa, and Latin America. France, Portugal, Italy, Croatia, and Romania are members of the European Union, and most are considered developed countries. Croatia and France initiated diaspora voting and representation system as a policy of appreciation for what the diaspora sacrificed during the liberation and wars in their home countries. Romania, Portugal, Macedonia, and Tunisia embarked on this system after the demise of authoritarian regimes and with the democratization process. Italy, Colombia, Ecuador, and the Dominican Republic can be used as an example of large diaspora mobilization. The remaining countries, Francophone and Lusophony nations, may import old colonial countries’ state experiences in addition to diaspora engagement policies. We emphasize throughout the paper that Turkey needs to take a diaspora electoral district system seriously. We believe that diaspora issues will only be taken into serious consideration when some kind of representation system is in place. Politicians give better voice to political inclusion when their policies involve emigrants. In our paper, we also highlight that debates on governmentality, transnationalism, and citizenship are taking a new turn. A state-​led citizenship is now outdated. New opportunities, responsibilities, and possibilities for transnational citizens emerge when extraterritorial political activities expand. This, in turn, not only produces new challenges for policymakers but also presents new opportunities as well. Creating an overseas electoral district system will resolve some of the difficulties that come with voting from abroad and help to enhance Turkish democracy in the future.

Notes 1 Tables and Figures in this chapter were created with data extracted from Turkey’s official statistical institute (tuik.gov.tr) by the authors. 2 The authors acknowledge various ways of defining diaspora, but in this paper they use the term in its most inclusive way. 3 Gurbetçi means “emigrant” in Turkish usage.

Turkish overseas electoral districts  199

Bibliography Anaz, N. (2018). Diaspora temsil sistemi. Kriter Yayinevi. Baradei, L.E., Wafa, D., & Ghoneim, N. (2012). Assessing the voting experience of Egyptians abroad:  Post the January 25 revolution. Journal of US-​China Public Administration, 9(11), 1223–​43. Bauböck, R. (2007). Stakeholder citizenship and transnational political participation: A normative evaluation of external voting. Fordham Law Review, 75(5), 2393–​447. Brand, L.A. (2010) Authoritarian states and voting from abroad:  North African experiences. Comparative Politics, 43(1),  81–​99. —​—​—​. (2014). Arab uprisings and the changing frontiers of transnational citizenship:  Voting from abroad in political transitions. Political Geography, 41(Special Issue),  54–​63. Collyer, M. (2014). A geography of extra-​ territorial citizenship:  Explanations of external voting. Migration Studies, 2(1),  55–​72. Collyer, M. & Vathi, Z. (2007). Patterns of extra-​territorial voting. Sussex Centre for Migration Research, October 2007 (T22), 4–​10. Erben, P., Goldsmith, B., & Shujaat, A. (2012). Out-​ of-​ country voting:  A brief overview. The International Foundation for Electoral Systems, IFES White Paper(April),  1–​11. Farmani, M., & Jafari, A. (2016). A comparative approach to study the electoral Systems of Selected Countries. International Journal of Humanities and Cultural Studies, 2(March), 1913–​24. Fierro, C.N. (2016). Electoral studies in compared international perspective, Mexico: National Electoral Institute. Gamlen, A. (2015). The impacts of extra-​territorial voting: Swings, interregnums and feedback effects in New Zealand elections from 1914 to 2011. Political Geography, 44(October),  1–​8. Honohan, I. (2011). Should Irish emigrants have votes? External voting in Ireland. Irish Political Studies, 26(4), 545–​61. IDEA. (2007). Voting From Abroad, s.l.: International IDEA Handbook. Jaulin, T. (2015). The geography of external voting: The 2011 Tunisian election abroad. The Transnational Studies Initiative, January 2015(1), pp. 1–​22. Kose, M. (2018). Seçim beyannamelerinde Türk diasporası. Star. www.star.com.tr/​ acik-​gorus/​secim-​beyannamelerinde-​turk-​diasporasi-​haber-​1354449/​ Lafleur, J.-​M. (2011). Why do states enfranchise citizens abroad? Comparative insights from Mexico, Italy and Belgium. Global Networks, 11(4), 481–​501. —​—​—​. (2013). Transnational politics and the state:  The external voting rights of diasporas. New York: Routledge. Laguerre, M.S. (2005). Homeland political crisis, the virtual diasporic public sphere, and diasporic politics. Journal of Latin American Anthropology, 10(1), 206–​25. —​—​—​. (2013). Parliament and diaspora in Europe. New York: Palgrave Macmillan. López-​Guerra, C. (2005). Should expatriates vote? The Journal of Political Philosophy, 13(2), 216–​34. Mencutek, Z.S. (2015). Mapping motivations of emigrants and concerns of host countries about external voting. Insight Turkey, 17(4),  1–​25. Özhan, S. (1975). T.B.M.M. Başkanlığına (To the Turkish National Grand Assembly), Hamburg: Proceedings of Turkish National Grand Assembly.

200  Necati Anaz and Mehmet Köse Rojas, H.S. (2007). A comparative study of the overseas voting laws and systems of selected countries. Kyiv: USAID/​Ukraine’s Strengthening Electoral Administration. Sahin-​ Mencutek, Z., & Erdoğan, M.M. (2015). The implementation of voting from abroad: Evidence from the 2014 Turkish presidential election. International Migration, 54(3), 173–​86. Shain, Y., & Barth, A. (2003). Diasporas and international relations theory. International Organization, 57(3), 449–​73. Tager, M. (2006). Expatriates and elections. Diaspora:  A Journal of Transnational Studies, 15(1),  35–​60. Tahtakılıç, A. (1979). Proceedings of Parliament. Ankara:  Turkish National Grand Assembly. TBMM. (2003). Yurt dışında yaşayan vatandaşlarımızın sorunlarını araştırma komisyonu raporu. Ankara: TBMM. —​—​—​. (2012). Proceedings of Parliament. Ankara: Turkish National Grand Assembly. Ülker, R. (1965). Proceedings of Parliament. Ankara:  Turkish National Grand Assembly. Warf, B., & Leib, J. (2011). Introduction. In:  B. Warf & J. Leib (Eds.) Revitalizing Electoral Geography (pp. 3–​8). Ashgate Publishing House. Yıldırım, D. (2016). Votes of expatriates: External voting behavior from 7th June to 1st November. Alternatif Politika, 8(2), 376–​403. YSK. (2018). Cumhurbaşkanı Seçimi ve 27. Dönem Milletvekili Genel Seçimi Sonuçları. Ankara: Resmi Gazeteye Gönderilen Tutanaklar. YTB. (2018). Presidency for Turks Abroad and Related Communities Related Legislation. Ankara: Official Gazette.

Index

Ankara’s Votes (AV) 121–​31, 133–​4, 136–​7, 138n5, 139n14, 140n27 Association for Atatürkist Thought 81, 175 Association for Human Rights and Solidarity for the Oppressed 90 Association for Support of Contemporary Life 175 Atatürk, Mustafa Kemal 159, 161 authoritarian 96, 159, 161, 198 Ayna, Emine 92 Bahçeli, Devlet 58, 62n25 Bilgen, Ayhan 90 Bölükbaşı, Osman 11–​12, 53–​4 Chavez, Hugo 136 clientelism 17, 41, 46, 49–​51, 60–​1, 105–​7 Confederation of Public Employees’ Trade Unions (KESK) 89, 110 Congress Party of India 47 conservatism 1, 5, 7, 16, 20, 37, 41, 67, 69, 70, 81–​2, 97, 103, 121, 127, 136–​7, 139n23, 140n34 constitutions: 1924 Constitution 10; 1961 Constitution 10–​11, 55; 1982 Constitution 9, 13, 55 coup d’état: 1960 coup 10, 39, 46; 1960 military intervention 55; 1980 military intervention 8, 33, 68, 164; 1980 (September 12) coup 7, 9, 13, 32–​3, 39, 55–​6, 66, 72, 158, 187; 2017 coup attempt 190; post-​modern coup, February 28, 1997 71 Çiller, Tansu 58, 179n17 Demirel, Süleyman 14, 56–​8 Demirtaş, Selahattin 91, 93, 95–​7, 99n22 Democracy Party (DeP) 21, 179n4

Democrat Party (DP) 8–​12, 20n4, 29, 33, 35, 37, 39, 46–​55, 61n7, 62n12, 169, 178n1 Democratic Left Party (DLP) 30, 36–​7, 57, 71–​3, 75–​6, 78, 80, 83n9, 179n3 Democratic Party (DkP) 35 Democratic People’s Party (DPP) 42n9, 73, 75, 87–​8, 90 Democratic Regions Party (DRP) 87, 92 Democratic Society Party (DSP) 41n1, 42n9, 47, 87, 90 D’Hondt method 39, 55 dynamic nationalization 28 Ecevit, Bülent 11, 57, 172 effective number of parties 31–​2, 34 Elçi, Şerafettin 92 election monitoring 19, 121–​3, 125, 129, 131–​2, 136–​8 elections, general: 1935 elections 159; 1946 elections 50, 122; 1950 elections 4, 8–​10, 26, 29, 31, 35, 38, 46, 52; 1954 elections 8, 10, 50–​1, 62n21; 1957 elections 10, 35, 52, 54; 1961 elections 10, 26, 33–​5, 39–​40, 51, 61; 1965 elections 8, 11–​2, 182, 185, 188; 1969 elections 12, 52; 1973 elections 29; 1977 elections 4, 9, 12, 33, 35, 38, 40; 1983 elections 9, 13–​14, 33–​4, 36, 39, 66, 70–​1, 76; 1987 elections 14, 36, 57, 71, 76; 1991 elections 14, 38, 40, 41n6, 71–​2, 76, 78; 1995 elections 15, 21n9, 57–​8, 71–​2, 76; 1996 mid-​elections 58; 1999 elections 18, 29–​30, 72, 76, 88, 158, 170; 2002 elections 15, 36, 40, 66, 82, 87–​9, 138n1; 2007 elections 24, 42n9, 72–​3, 77, 80, 87, 90, 97; 2011 elections 24, 32, 42n9, 66, 73, 83n9, 87, 92, 189; 2014 presidental elections

202 Index 19; 2015 elections (June–​November) 32, 34, 41n7, 59, 73, 75, 82, 87, 92, 94–​5, 97, 123, 132, 170, 190, 195, 197; 2018 elections 1, 28, 31–​2, 34, 38, 41n6, 69, 73, 87, 131–​2, 153, 170, 178, 185, 190 elections, local: 1930 elections 49; 1955 elections 53; 1963 elections 54; 1979 elections 8; 1984 elections 56; 1989 elections 56–​7, 62n19; 1994 elections 57–​8; 1999 elections 57–​8, 71; 2004 elections 90; 2014 elections 15, 58, 87, 92, 123; 2019 elections 16, 60, 82, 131 electoral integrity 122, 131, 136, 150 electoral monitoring 138 Erdoğan, Recep Tayyip 24, 41n7, 81–​2, 85, 93, 95–​7, 115, 193 ethnographic specificity 103 Evren, Kenan 13 Felicity Party (FP) 38 fragmentation 1, 11, 14, 27–​8, 34, 36, 41, 55, 103, 122 Free Republican Party (FRP) 49–​50 Freedom and Solidarity Party (FSP) 90 Freedom House 122 gender 3, 4, 16, 18, 20, 69, 85, 101–​5, 110, 117, 138n7, 154, 156–​8, 161–​2, 164–​5, 167, 169, 171, 175, 177–​8 Gerry, Elbridge 46 gerrymandering 4, 16–​17, 46–​9, 51, 53–​61, 63n31 Gezi Park protests 123, 136, 138n5 Good Party (GP) 16, 136–​7, 179n5 İnönü, İsmet 51, 159 Islamism 158, 171 Justice and Development Party (JDP) 1, 15–​16, 18, 24, 29, 30–​1, 36–​7, 40–​1, 58–​60, 62n30, 66–​8, 72–​8, 80–​2, 92–​6, 111–​12, 128, 137, 139n15, 153, 171–​3, 177, 178n2, 180, 185–​6, 192, 194–​5 Justice Party (JP) 10–​12, 29, 33, 35–​7, 40, 51–​2, 54, 169, 178n2 Kalkan, Duran 93, 99n17 Kavakçı, Merve 171–​2, 179n18, 181 Kurdistan Workers’ Party (PKK) 85, 88, 90–​5, 97, 98n1, 99n17 Kulturkampf 81

Labor Party (LP) 71, 88, 90 Labor, Freedom, and Democracy Bloc 91 LGBT community 92, 97 Liberal Democrat Party (LDP) 132, 140n29 Liberty Party 53 media use 18, 124–​6, 133, 135, 143 Ministry of Foreign Affairs (MFA) 189 modernization 19, 80, 160–​1, 164, 167, 177 Motherland Party (MP) 13–​14, 36–​7, 39–​40, 55–​8, 60, 62n17, 71–​4, 76, 78, 80, 111, 160, 169, 178n2 Nation Party (NP) 37 National Development Party 61n5 National Outlook Movement (NOM) 38 National Salvation Party (NSP) 8–​9, 29, 37–​8 nationalism 13, 18, 20, 47, 96–​7, 103, 171 Nationalist Action Party (NAP) 7, 15, 18, 29, 30–​1, 37–​8, 40–​1, 58–​60, 62n30, 67, 71–​3, 75–​8, 80, 136–​7, 139n15, 140n34, 179n5 Nationalist Democracy Party (NDP) 13, 36, 41n1, 71–​2 Nationalist Working Party (NWP) 38 nationalization of political parties 25, 31, 40 New Turkey Party (NTP) 10, 33, 35, 51–​2, 61n9 Önder, Sırrı Süreyya 92 Özal, Turgut 13, 56–​8 participation 18, 102–​3, 110, 123, 130, 132–​3, 137, 164, 171, 175, 178, 182, 184, 186–​7, 189–​93, 195–​8 party nationalization: Party Nationalization Score (PNS) 25–​6, 32; Party System Nationalization Score (PSNS) 25–​6, 32–​4, 39, 41; Standardized Party Nationalization Score (sPNS) 25–​6, 32–​40, 42n9 Peace and Democracy Party (PaDP) 42n9, 86–​7, 91–​2, 171–​2, 179n4 Peasant Party 53, 62n12 People’s Democracy Party (PDemP) 15, 21n9, 42n9, 60, 72, 76, 88, 179n4 People’s Democratic Party (PDP) 1, 15–​16, 18, 31, 38, 41n1, 60, 73, 76, 85,

Index  203 87, 92–​6, 98, 99n17, 136–​7, 139n15, 171–​2, 179n4, 180, 186 People’s Party of Women 152 polarization 1, 11, 13–​14, 16, 29, 41, 48, 69, 80–​1, 113 political elites 159 political entrepreneurship 13 Political Parties Law 26, 41 populism 55, 137 Populist Party (PP) 14, 36, 40, 54, 71, 169, 179n3 Presidency for Turks Abroad and Related Communities (YTB) 189, 194 proportional representation 10, 28, 33, 46, 55, 60 referendums: 1961 constitutional referendum 10; 2010 constitutional referendum 189; 2017 constitutional referendum 25, 78, 122–​3, 178, 186, 190–​2 regions in Turkey: Aegean Region 7, 10, 29, 30, 66, 68–​9, 155–​6, 158, 160, 169, 170, 177; Black Sea Region 8, 10, 19, 20, 155–​6, 158, 160, 169, 173, 177; Central Anatolia Region 7–​9, 11–​12, 14–​15, 20, 29, 30, 67, 112, 133, 155–​6, 158, 160, 169, 173, 177; Eastern Anatolia Region 7, 9, 10, 12, 15, 29, 30, 67, 82, 85, 88, 97, 98n1, 106–​7, 137, 150, 155–​6, 158, 160, 169, 173, 177; Marmara Region 10, 12, 14, 19, 29, 30, 66, 69, 155–​6, 158, 160, 169, 170, 173, 177; Mediterranean Region 9–​11, 17, 19, 29, 66, 68–​9, 155–​6, 158, 160, 169, 173, 177; Southeast Anatolia Region 1, 10, 12–​15, 18, 24, 29, 30, 40, 66–​8, 87–​9, 91, 95, 99n22, 106–​7, 150, 109, 121, 137, 155–​6, 158, 160, 169, 172–​3, 177, 179n12; Thrace Region 69 Reliance Party (RP) 35 Republican Nation Party (RNP) 37, 51, 53–​4, 61n8 Republican Peasant Nation Party (RPNP) 8, 37, 51, 54, 179n5 Republican People’s Party (RPP) 1, 9–​12, 15–​18, 24, 29, 31, 35–​7, 40–​1, 46, 49–​54, 57–​60, 61n8, 62n30, 66–​8, 71–​8, 80, 82, 83n7, 92, 94, 111–​12, 127–​8, 131, 136–​7, 139n15, 167, 169, 177, 179n3, 188, 195 Republican Reliance Party (RRP) 35

resolution process 92–​3 Revolutionary Path 8 secularism 81, 130, 139n22 senate 10 Sezer, Ahmet Necdet 62n28 simple majority 9, 10, 47, 51, 60, 159 single-​party era 9, 10, 50, 152, 155–​6, 159, 160, 162–​7, 169, 172, 175, 177–​8, 179n3: republican era 19, 96; republican period 49, 92, 152, 178; republican Turkey 156 Social Democracy Party (SDP) 36, 40 social democrat 67, 85, 153 Social Democratic Populist Party (SDPP) 14, 56–​8, 62n18, 71, 169, 179n3 Socialist Democracy Party 88 Sönmez, Fikri 8, 20n1 stakeholder citizenship 183 state of emergency 124–​5, 140n30 Support of Women’s Work (KEDEV) 110 Supreme Election Council of Turkey (YSK) 122, 150, 187, 189, 190, 195, 197 Tan, Altan 92 Thousand Hope Candidates 90 Toptan, Köksal 57 transitional election 11 transnational citizenship 182, 184, 198 transnational space 193–​4 True Path Party (TPP) 14, 36–​7, 40, 56–​8, 60, 62n18, 71–​4, 76, 78, 169, 178n2, 179n17 Tuğluk, Aysel 91 Turkish Constitutional Court 42n9, 61n7, 62n17, 71, 86, 90–​2, 154, 171 Turkish diaspora 182, 185–​9, 191, 194–​7 Turkish Grand National Assembly (TGNA) 13, 14, 41n7, 50, 132, 152–​4, 156–​7, 162–​3, 166, 169–​72, 175, 177, 179; Turkish Parliament 10, 11, 15, 19, 20n3, 37, 39, 51, 53, 82, 85, 88, 91, 93–​5, 102, 105, 153–​4, 156–​67, 169–​71, 175–​7, 178n1, 182, 185–​6, 188, 191, 193, 195–​6, 198 Turkish Women’s Union 178n1 Türk, Ahmet 90–​1, 98n7 Türkeş, Alparslan 37

204 Index Union of Chambers of Turkish Engineers and Architects (TMMOB) 110 United Nations 110; United Nations Development Programme (UNDP) 110 Virtue Party (VP) 38, 70–​1, 74–​5, 171 volatility 1, 4, 7, 14–​15, 18, 29, 30, 41, 48, 60–​1, 70, 72, 122 Vote and Beyond (VB) 121–​30, 133–​4, 136n3, 137, 138n5, 140n30

Votes Abroad 123 Votes of Turkey 123 Welfare Party (WP) 7, 8, 15, 19, 37, 58, 62n18, 70–​2, 74–​7, 80, 82 westernization 164, 177 Workers’ Party 169 Yılmaz, Mesut 57 Young Party (YP) 72 Yüksekdağ, Figen 94–​5 Zana, Leyla 90, 92, 171–​2