UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Access to power : hegemonic party rule in Singapore and Taiwan Tan, Netina Clara 2010

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


24-ubc_2011_spring_tan_netina.pdf [ 3.83MB ]
JSON: 24-1.0071541.json
JSON-LD: 24-1.0071541-ld.json
RDF/XML (Pretty): 24-1.0071541-rdf.xml
RDF/JSON: 24-1.0071541-rdf.json
Turtle: 24-1.0071541-turtle.txt
N-Triples: 24-1.0071541-rdf-ntriples.txt
Original Record: 24-1.0071541-source.json
Full Text

Full Text

ACCESS TO POWER: HEGEMONIC PARTY RULE IN SINGAPORE AND TAIWAN by Netina Clara Tan M.A., The University of Regina, 2004 M.A., The National University of Singapore, 2000 B.A., The National University of Singapore, 1992 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in The Faculty Of Graduate Studies  (Political Science)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) December 2010  © Netina Clara Tan, 2010  ABSTRACT My dissertation investigates the sources of hegemonic party resilience. I ask why do some hegemonic party regimes persist, while others concede to multipartism? Building on party politics and electoral authoritarianism literature, I develop a mid-range theory based on the concepts of strategic coordination and institutionalization to explain why elites unite and oppositions fail to pose a credible threat. To demonstrate the utility of my explanation, I compare two similar hegemonic parties of different outcomes: the People’s Action Party (PAP) in Singapore and the Kuomintang Party (KMT) in Taiwan. I posit three factors to account for hegemonic party resilience. First, I contend that a hegemonic party that is adept in strategic coordination – by providing public goods and withdrawing political, civil liberties and media freedom – is more likely to win mass support and deter opposition coordination. Both the PAP and early KMT were high performing, strategic regimes that enjoyed growth and forestalled democratization. While the PAP remained the ruling party in Singapore, the KMT controlled the pace of liberalization during its long decade of transition, losing power after a party split. Second, I argue that the PAP is better than the KMT in keeping the ruling elites united because of its institutionalized leadership succession system. I develop a model to explain how a centralized, oligarchic and exclusionary leadership selection method fosters elite unity. My findings based on elite interviews, party publications and survey data support the counter-intuitive theory that the more intra-party democracy, the less party cohesion. Finally, in hegemonic party regimes, survival means increasing the certainty of winning. Through electoral engineering, the incumbent is able to institutionalize an uneven playing field that systematically disadvantages the opposition. By analyzing the mechanical and psychological effects of electoral reforms, I offer new empirical evidence to show how the PAP “manufactured” its legislative supermajority to rescue its declining popular votes. The contrasting study of the KMT highlights how a former hegemonic party transforms and adapts as a dominant party to survive the uncertainty of elections.  ii  PREFACE The UBC Behavioural Research Ethics Board (BREB) approved the elite interviews involving human subjects conducted for my dissertation. The Certificate of Approval, Ethics Certificate Number (H06-03753) to conduct research in Singapore and Taiwan is enclosed in Appendix Q: UBC BREB Certificate of Approval. Parts of Chapter 3 and 6 will appear as a book chapter entitled “Institutionalized Hegemonic Party Rule in Singapore” for Party and Party System Institutionalization in Asia, edited by Allen Hicken and Erik Kuhonta, supported by the Institute for the Study of International Development, the Social Sciences and Humanities Research Council, Singapore’s Institute for Southeast Asian Studies, the Department of Political Science at the University of Michigan and the Department of Political Science at McGill University. Parts of Chapter 7 will be revised as a book chapter entitled “Hegemonic Party Stability and Opposition Party Failure In Singapore” in Party Stability and Party Performance in Southeast Asia edited by Wolfgang Sachsenroeder, funded by Friedrich Naumann Foundation and Institute of Southeast Asian Studies, Singapore.  iii  TABLE OF CONTENTS  Abstract ..................................................................................................................................... ii! Preface...................................................................................................................................... iii! Table of Contents ..................................................................................................................... iv! List of Tables .......................................................................................................................... vii! List of Figures .......................................................................................................................... ix! List of Abbreviations .................................................................................................................x! Acknowledgements ................................................................................................................. xii! Dedication .............................................................................................................................. xiv! 1! Introduction ........................................................................................................................15! Implications......................................................................................................18! Case Selection ..................................................................................................21! Plan of the Dissertation ....................................................................................23! 2! Strategic Coordination, Institutionalization and Electoral Engineering ............................25! A Typology of Single-Party Regimes ..............................................................25! Theories of Authoritarian Durability ...............................................................29! Problems and Gaps ..........................................................................................32! Theoretical Framework ....................................................................................35! Part I: Origins, Organization and Coordination .......................................................................52! 3! The PAP: Organization Transformation and Strategic Coordination ................................53! Origins of the PAP ...........................................................................................55! Organizational Changes in 1958 ......................................................................63! Party Adaptation After 1963 ............................................................................65! Strategic Coordination: Dismantling Cleavages ..............................................67! 4! The KMT: Organizational Adaptation and Coordination Dilemmas ................................82! Origins of the KMT .........................................................................................83! Party Membership, Taiwanization and Co-optation ........................................97! Implications of Taiwanization .......................................................................102! Coordination Dilemmas .................................................................................105! Conclusion .....................................................................................................111  iv  ! Part II: Leadership selection and party cohesion ...................................................................113! 5! The PAP: Institutionalized Charisma and Party Cohesion ..............................................114! Westminster Parliamentary System and Party Cohesion ...............................115! Rules of the Game in Singapore ....................................................................117! Institutionalization, Leadership Selection and Party Cohesion .....................124! Key Features of the PAP’s Candidate Selection ............................................131! Implications and Potential Problems .............................................................138! Conclusion .....................................................................................................141! 6! The KMT: Un-Institutionalized Charisma and Factionalism ..........................................144! Selection of Chiang Ching-kuo as Party Chairman and President ................144! Intra-Party Democratization...........................................................................152! Rules of the Game in Taiwan.........................................................................159! Implications for Party Cohesion ....................................................................166! Re-Centralizing Leadership Selection ...........................................................177! Conclusion .....................................................................................................179! Part III: Elections, electoral engineering and party system ...................................................181! 7! Singapore: Electoral Engineering and Institutionalizing Certainty .................................182! Role of Elections in Hegemonic Party Regimes ............................................183! Why Elections in Singapore? .........................................................................184! Electoral Engineering in Singapore ...............................................................187! Mechanical and Psychological Effects ..........................................................198! Institutionalized Hegemonic Party System in Singapore...............................208! Conclusion .....................................................................................................212! 8! Taiwan: Elections, Electoral Reforms and Institutionalizing Uncertainty ......................215! Why Local Elections? ....................................................................................215! When Do Elections Matter? ...........................................................................219! Coordination Dilemmas and De-Institutionalization .....................................221! Origins of Electoral Reforms and Institutionalizing Uncertainty ..................225! Post-2005 Electoral Reforms .........................................................................228! Electoral Reforms and Unintended Consequences ........................................229! Conclusion .....................................................................................................244! 9! Conclusion .......................................................................................................................246! Implications....................................................................................................248! Gaps and Lessons Learnt ...............................................................................251! References ..............................................................................................................................255! Appendices.............................................................................................................................273! v  Appendix A: List of Interviewees and Archival Sites in Singapore And Taiwan ................ 273! Appendix B: BTI Democracy Status and Key Indicators of Singapore and Taiwan, 2010.. 275! Appendix C: Organization Structure of the PAP, 1958 and present..................................... 276! Appendix D: Significant Events During the PAP’s early rule, 1960s .................................. 277! Appendix E: Party Organization Structures of the KMT, 1952 and present ........................ 278! Appendix F: Ethnic Breakdown of the KMT Membership (1952-2005) ............................. 279! Appendix G: World Bank Governance Indicators for Singapore & Taiwan (1998-2008) ... 280! Appendix H: List of Registered Political Parties in Singapore............................................. 281! Appendix I: Significant Events Under Lee Teng-hui's early rule (1988-1989) .................... 282! Appendix J: TI Corruption Index for Taiwan and Singapore (1995-2009) .......................... 283! Appendix K: Old and New Electoral Systems in Taiwan..................................................... 284! Appendix L: Electoral Systems of Singapore and Taiwan ................................................... 285! Appendix M: List of Active Parties in Singapore ................................................................. 286! Appendix N: Voter Turnout in Singapore (1968-2001) ....................................................... 287! Appendix O: Taiwan's Constitutional Reforms in the 1990s ............................................... 288! Appendix P: Comparative Freedom House Scores of Singapore and Taiwan (1973-2009) 289!  vi  LIST OF TABLES Table 1-1: Dissertation Structure Based on Three-Level Analytical Framework .................. 21! Table 2-1: A Typology of Single-Party Regimes ................................................................... 29! Table 3-1: Singapore’s Population by Ethnicity, 1957 ........................................................... 59! Table 3-2: PAP Membership Size and Residential Population in Singapore (1958-2006) .... 60! Table 3-3: Trade Unions, Disputes and Strikes in Singapore, 1960s-70s .............................. 66! Table 3-4: Ethnic Composition and Education of the PAP Members, the 1960s ................... 67! Table 3-5: List of Para-Political Institutions under the PA in Singapore ............................... 70! Table 3-6: Number of Para-Political Organizations Under the PA (1996-6) ......................... 71! Table 4-1: Evolution of the KMT Regime (1945-2000) ......................................................... 85! Table 4-2: Ethnic Composition of Total Population in Taiwan (1952-64) ............................. 87! Table 4-3: Taiwanese-Mainlanders Composition in the Military, 1950-1987 ....................... 88! Table 4-4: Rural-Urban Geographical Distribution in Taiwan (1952-64) .............................. 89! Table 4-5: KMT Party Branches and Cells (1952 and 2010) ................................................. 94! Table 4-6: Growth of KMT Service Stations (1951-2) ........................................................... 95! Table 4-7: Demographic Trends of the KMT Recruitment (1952-1997) ............................... 99! Table 4-8: Ethnic Composition of Executive Yuan (1950-95) ............................................. 101! Table 4-9: Ethnic Composition of CSC and CC (1952-1994) .............................................. 101! Table 4-10: Timeline of Significant Events Under the KMT (1960s-80s) ........................... 107! Table 4-11: Indicators of Taiwan's Economic Growth (1953-87) ........................................ 108! Table 4-12: Public Protests in Taiwan (1983-1988) ............................................................. 108! Table 4-13: Press Censorship in Taiwan (1980-1986) ......................................................... 109! Table 4-14: Major Reforms Initiated During Last Years of Chiang Ching-kuo's rule ......... 110! Table 4-15: Breakdown of KMT Party Membership, Recruits and Staff (1952-present) .... 111! Table 5-1: Voting Behavior in Singapore Parliament (2007-9)............................................ 116! Table 5-2: Deposit Amount for Legislative Candidates in Singapore .................................. 120! Table 5-3: Occupational Background of MPs (1963-2006) ................................................. 121! Table 5-4: Comparative Salaries of MPs and Ministers, 2006 ............................................. 123! Table 5-5: List of PMs in Singapore ..................................................................................... 129! Table 5-6: Rate of Turnover in Parliament (1980-2006) ...................................................... 131! Table 5-7: Occupational Profiles of Cabinet Ministers (2001-9) ......................................... 134! Table 5-8: Female Representation in Singapore Parliament (1980-2006)............................ 137! Table 5-9: PAP Membership and Total Singapore Population ............................................. 140! Table 6-1: Violation of Party Discipline (1952-1995) .......................................................... 149! vii  Table 6-2: Selection Methods for the CC and CSC (1952-2005) ......................................... 154! Table 6-3: Social Composition of the CSC and CC (1952-1994) ........................................ 155! Table 6-4: Occupation Background of KMT’s CSC (1950-1992) ....................................... 156! Table 6-5: Female Representation in Legislative Yuan (1995-2008) ................................... 157! Table 6-6: Education Qualifications of the CSC .................................................................. 158! Table 6-7: Candidacy Requirements for Legislative Elections for Taiwan and Singapore .. 160! Table 6-8: Legislative Yuan Elections (1969-2001) ............................................................. 163! Table 6-9: Breakdown of Candidates and Elected Officials for LE (1989-2001) ................ 165! Table 6-10: Total Spending of Parties on Advertisements in LE and PE ............................. 170! Table 6-11: Comparative Electoral Campaigning Laws for Taiwan and Singapore ............ 172! Table 6-12: KMT's Presidential Nomination Methods (1996-2008) .................................... 174! Table 6-13: Direct Presidential Candidates and Results (1996-2008) .................................. 174! Table 6-14: Party Splits in the KMT..................................................................................... 175! Table 6-15: Selection Methods of Party Chairman (1949-present) ...................................... 178! Table 7-1: Key Electoral Changes in Singapore ................................................................... 189! Table 7-2: Effective Number of Parties Based on Vote and Seat Shares in Singapore ........ 200! Table 7-3: Electoral Disproportionality Based on Vote and Seat Shares (1968-2006) ........ 202! Table 7-4: Number of Contesting Candidates and Parties (1968-2006) ............................... 204! Table 7-5: Credibility of Political Parties in Singapore, 2006 .............................................. 211! Table 8-1: Breakdown of Vote and Seat Shares of Local Elections (1965-1993) ................ 221! Table 8-2: Vote and Seat Shares in National Assembly and Legislative Yuan .................... 231! Table 8-3: Effective Number of Parties Based on Vote and Seat shares .............................. 232! Table 8-4: Electoral Disproportionality of LEs (1986-2008) ............................................... 233! Table 8-5: Characteristics of Political Parties in Taiwan...................................................... 237! Table 8-6: Level of Political Trust in Taiwan and Singapore, 2006 ..................................... 241!  viii  LIST OF FIGURES Figure 2-1: Strategic Coordination of Hegemonic Party Regimes (HPRs) ............................ 37! Figure 2-2: Candidacy, Selectorate and Selection Method ..................................................... 44! Figure 2-3: Model of Leadership Succession and Institutionalization ................................... 45! Figure 4-1: Ethnic Breakdown of the KMT Membership .................................................... 100! Figure 5-1: The PAP’s Model of Leadership Succession and Party Cohesion..................... 126! Figure 6-1: The KMT’s Model of Leadership Succession and Party Cohesion ................... 146! Figure 6-2: The KMT's Seat and Vote Shares in the Legislative Yuan (1969-2008) ........... 166! Figure 7-2: Minority MPs in Parliament............................................................................... 194! Figure 7-3: Uncontested Seats in Singapore (1968-2006) .................................................... 206! Figure 7-4: Electoral Volatility of Singapore GEs (1968-2006) .......................................... 210! Figure 8-2: Electoral Volatility Based on LEs and PEs in Taiwan....................................... 236! Figure 8-3: Breakdown of Candidates and Independents in LEs (1989-2008) .................... 240!  ix  LIST OF ABBREVIATIONS AI APS BS BTI CCCs CCMCs CCs CD CDAC CDCs CDRA CDRA CEC CPI CSC CYC DP DPP EAB EIP EIU FH FPTP GE GNP GPWD GRC HDB IPS ISD KMT Lakas NUCD LDP MCP MM MMD MOA MP MRHA NA NCMPs NCs NMP NP NPS  Angkatan Islam Alliance Party Singapura Barisan Socialis Bertelsmann Transformation Index Citizens' Consultative Committees Community Club Management Committees Community Centre/Clubs Christian Democratic Party Chinese Development Assistance Councils Community Development Councils Chinese Democratic Reformers Alliance Chinese Democratic Reformers Alliance Central Executive Committee Consumer price index Central Standing Committee China Youth Corp Democratic Party Democratic Progressive Party East Asian Barometer Ethnic Integrated Housing Policy (EIP) Economist Intelligence Unit Freedom House First past the post General Election Gross National Product General Political Warfare Department Group Representative Constituency Housing Development Board Institute of Policy Studies, Singapore Internal Security Department Kuomintang Party (Nationalist Party) National Union of Christian Democrats Liberal Democratic Party Malayan Communist Party Minister Mentor Multi-member district Miscellaneous Offences Act Member of Parliament Maintenance of Religious Harmony Act National Assembly Non-Constituency Member of Parliament Neighbourhood Committees Nominated Member of Parliament New KMT Alliance/Chinese New Party National Party of Singapore x  NSP PA PAP PAYM PCF PE PEA PEMA PF PFP PKMS PLDP PM PP PPF PR PR PRI PRP PSC RCs RP RSAF SAF SCC SCMSSU SCP SCS SDA SDP SIDA SJP SM SMC SMD SMP SNF SNTV SPF SPP TSU UDP UNF UPF UPP USSR WP  National Solidarity Party People’s Association People’s Action Party People’s Association Youth Movement PAP Community Foundation Presidential Election Parliamentary Elections Act Public Entertainments Meeting Act The People’s Front Peoples First Party Pertubohan Kebangsaan Melayu Singapura People’s Liberal Democratic Party Prime Minister Progressive Party PAP Policy Forum Partai Rakyat Proportionate Representation Institutional Revolutionary Party People’s Republican Party Public Service Commission Residents' Committees Reform Party Republic of Singapore Air Force Singapore Armed Forces Secretariat of the Central Committee Chinese Middle School Students Union Singapore Chinese Party Singapore Civil Service Singapore Democratic Alliance Singapore Democratic Party Singapore Indian Development Agency Singapore Justice Party Senior Minister Single member constituency Single member district Single member plurality Singapore National Front Single Non-Transferable vote Singapore Police Force Singapore People’s Party Taiwan Solidarity Union Parti Kesatuan Ra’ayat/United Democratic Party United National Front United People’s Front United People’s Party Union of Soviet Socialist Republics The Workers’ Party xi  ACKNOWLEDGEMENTS  There are many people to thank for the completion of this dissertation. First and foremost, I want to express my gratitude to my supervisor, Max Cameron, who has stood by me from my first year of doctoral studies and offered me valuable advice and support throughout various “critical junctures”. Perhaps unbeknown to him, the advice that he gave has had enormous “path-dependent” effects. His positive feedback, encouragement, confidence and belief in my work have helped me stay on task, push myself and achieve beyond my expectations. I am also very fortunate and perhaps at the envy of my colleagues to have Yves Tiberghien and Ben Nyblade as my committee members. Yves is a mentor and a model teacher who has taught me the rules of the game and “nuts and bolts” of survival in the academic world. I am forever grateful to the encouraging emails that he has written to lift my spirits and many words of wisdom to navigate through difficult times. Ben has offered me the best and the most intellectually stimulating conversations on my research. His bountiful energy and insightful comments have always challenged me to think more comparatively, strategically and broadly. I cannot ask for a more supportive committee. Besides my committee, I have also benefitted immensely from the close mentoring of Paul Evans – who has been there from the beginning to offer support, encouragement and help to broaden my network of academic connections in Vancouver and beyond. I also want to thank Ken Foster, who “adopted” me after my comps and went through countless drafts of my prospectus. This dissertation has also drawn from many sources, conversations over meals and advice of scholars with similar research interests and expertise. Specifically, I am grateful to Angela O’Mahony, Diane Mauzy, Brian Job, Chris Kam, Mark Warren, Katia Coleman, Alison Bailey, Timothy Cheek and Michael Leaf who have taken time to talk to me about my work and ensure that I stay on course. At UBC, I have been blessed with many intelligent, fun and loyal friends who have helped me keep my head above water. I am especially thankful for the companionship and many dinner meals with Hyunji Lee, David Seekings, Erin Williams, Nathan Allen and Kate Hecht. Outside of UBC, I am grateful to have had the opportunity to work and engage with Joseph Wong, Allen Hicken, Erik Kuhonta, Cheng Tun-jen, Craig Townsend and Wolfgang Sachsenroeder - outstanding scholars on East and Southeast Asia - on book projects that I look forward to seeing them in fruition soon. My dissertation would not have been possible without financial support from many sources. Firstly, I want to thank the Faculty of Graduate Studies of UBC for four years of Ph.D Tuition Award and travel awards from both FOGS and Political Science Department that allowed me to present my work in conferences. Secondly, I am also very grateful to the Taiwan Foundation for Democracy for a Dissertation Fellowship that supported my 11 months of fieldwork in Taiwan, 2007. My interviews and data collection would not have been as successful without the help, warm hospitality and camaraderie of Kiel Downey, Bo Tedards, Russell, Mei Chen, Amy, Weichi, Grace, Angel and Lydia Tsai of the TFD. Additionally, I would like to express my deepest appreciation to Robert Chu Wen-Ching for his friendship and help in setting up interviews; xii  inviting me to KMT events and taking time to explain the internal KMT party politics to me over many lunches. I am glad to have connected with Kharis Templeman, who has taught me much about the mechanisms of electoral politics from the “dark green side”. In Taiwan, I am grateful to all the following interviewees and people who have taken time to speak to me on the KMT: Chang Yu-Tzung; Chen Peng-Jen; Cheng Tun-Jen; Cheng Su-Feng; Chu Yun-han; Michael Hsiao Hsin-Huang; Huang Min-Hua; Huang Su-Feng; Ho, Szu-Yin; David Hamilton; Ron MacIntosh; Shaui Hua-Min; Shelley Rigger; Shih Chengfeng, Wang Fu-Chang; Wu Nai-teh; Lin Jih-Wen; Shao Ming-Huang; Yu Ching-Hsing and Xu Xin-Sheng. Additionally, I would like to acknowledge the help of Marina Hu Hai-Ming, an exceptional librarian at the Kuomintang Archives Library for hunting down historical documents and recommending reading materials at crucial times. I am indebted to my adopted Taiwanese family, Jenny Hsia and Alan Tan who have looked after me during my stay and socialized me to live like a local in Taipei. We have spent endless hours debating thorny issues till late in the night. Without them, my time in Taiwan would not have been as colourful, meaningful and enjoyable. Third, I want to acknowledge the support of S. Rajaratnam School of International Studies, Nanyang Technological University for hosting me as a Visiting Research Associate during my 3 months of fieldwork in Singapore. I want to thank Amitav Archarya, Mely Anthony, Ralf Emmers, Harish, Luanne, Sue Chia and Sofiah for making me feel welcome and providing an excellent office support. For Singapore’s interviews, I am indebted to the following people who have enriched me with their views on the PAP government: Alex Au, Chee Siok Chin, Chee Soon Juan, Chua Beng Huat, Russell Heng, Lee Koon Choy, Catherine Lim, Sylvia Lim, Michael Palmer, Garry Rodan, Kenneth Paul Tan, Simon Tay, Perry Tong and Benjamin Wong. And for completing my dissertation, I want to acknowledge the financial support of the Taiwan Economic and Cultural Office for awarding me a Graduate Fellowship in Taiwan Studies, which freed me from teaching and allowed me to focus on writing. Last but not least, my family - Richard, Elizabeth, Victoria and Alexander - who have been through so much to support my seemingly endless years of grad school and put up with moving around the world. I would not have completed this dissertation without their patience, love and selflessness. My three wonderful children have learnt to survive with long periods of their mom’s absence and always brought home outstanding report cards that would make any parent proud. Their maturity, independence and many achievements have kept me going. I thank them for making me a stronger and better person. And to Rich, my pillar of strength, who has been my most faithful, dutiful and loyal supporter. I thank you for keeping me grounded, positive and happy.  xiii  DEDICATION  To Rich, Lisa, Vicky and Alex  xiv  1  Introduction Following decades of democratization since the Third Wave, the spread of democracy  has come to a halt (Freedom House 2009).1 Scholars observe that the decline of freedom in Asia and the rise of China and Russia are pointing to a growing “pushback” against democracy (Puddington 2007; Kagan 2007). With the recurrence of coups in Thailand and Fiji, suppression of protests in Burma, Pakistan and Malaysia, parts of Asia are showing signs of authoritarian reversion. In an era of liberal democracy, why are some authoritarian regimes stubbornly persistent while others breakdown? Current studies show that authoritarian regimes ruled by a single party are more durable than military or personalistic ones (Geddes 1999; Brownlee 2007).2 In particular, electoral authoritarian or hegemonic party regimes3 that combine democratic and authoritarian traits to govern are especially persistent.4 Indeed, since 1987, this single-party subtype has grown in strength and numbers and has become a modal authoritarian regime type (Roessler and Howard 2009, 103). My dissertation investigates the sources of hegemonic party resilience by comparing two exceptionally resilient hegemonic party regimes in East Asia, namely Singapore under the PAP (1959-present) and Taiwan under the KMT (1949-2000). The PAP has ruled Singapore since 1959 and remains in power today. On the other hand, the KMT, once the wealthiest party in the world, lost ruling power after 5 decades of rule. What explains their divergent political trajectories? What can be learnt from their successes and failures? Scholars have warned that apart from exogenous shocks that undermine the resilience of all authoritarian regimes, internal splits and leadership succession are the most likely  1  According to FH, 2008 was a setback for global freedom as the countries considered “Free” has stagnated for a decade. In 2008, the EIU found 51 autocracies and 36 hybrid regimes out of 167 countries. 2 Geddes found single-party regimes to have an average life span of 34 years, 16 years longer than personalist and 24 years more military regimes (2003, 79). 3 Following Sartori, a hegemonic party regime is defined as a polity where a party dominates policy, controls access to political office, even though other parties may exist and compete for power. It refers to a semi-competitive party system where a hegemonic party exercises tight control over the players; rules of the game in the electoral arena and leaves little room for opposition and contestation (1976, 230). 4 Schedler classifies electoral authoritarianism as regimes that hold multi-party elections to select the chief executive as well as a legislative assembly and earn average FH ratings between 4 and 6. In 2001, he identified 58 electoral autocracies; 32 electoral democracies and 25 closed autocracies (2002, 47).  15  causes of authoritarian breakdown.5 To explain why hegemonic party regimes, such as Singapore’s PAP, are successful in staying cohesive and winning elections, I ask the following: what builds party cohesion? And if hegemonic party regimes hold elections, when do these elections matter and bring about democratizing outcomes? To answer these questions, I move away from large-scale theories of regime change and focus on investigating the sources of hegemonic party resilience. Drawing from electoral authoritarianism6 and party politics literature, I develop a mid-range theory to explain what behavioural incentives keep elites united, maintain mass support and deter opposition challenge, and how they do so. Using the concepts of strategic coordination and institutionalization, I posit three explanatory factors. First, I argue that a hegemonic party regime with an institutionalized incentive distribution system that ensures the long-term security of its ruling elites is more likely to stay cohesive. My dissertation offers a leadership succession theory to explain how a centralized, oligarchic and exclusionary leadership selection model fosters elite unity; a decentralized, democratic and inclusionary model leads to disunity. Second, I contend that a successful hegemonic party that is adept in strategic coordination – by providing public goods and withdrawing political and civil liberties and media freedom – is more likely to win mass support and deter opposition coordination. Finally, when faced with declining vote shares, I show how a hegemonic party can turn to electoral engineering to institutionalize a semi-competitive party system to “manufacture” its legislative supermajority and systematically repress the opposition. I demonstrate the utility of my explanation through a “structured and focused”7 comparative study of Singapore’s PAP and Taiwan’s KMT over the last 4 decades (George and Bennett 2004, 67). Through careful process tracing8, I examine how a selected aspect of  5  For a sample of the scholarship on authoritarian breakdown, see O’Donnell and Schmitter (1986), Huntington (1991) and Haggard and Kaufman (1995). 6 This dissertation builds on the emerging scholarship on electoral authoritarianism to examine the hybrid nature of single-party regimes (Schedler 2003; Levitsky and Way Lindberg 2009; Carothers 2002). 7 The method is “structured” as I pose the same questions for each case and these questions guide my data collection. It is “focused” as I only address a particular aspect of the two cases. 8 Process-tracing refers to the method of within-case analysis to evaluate causal processes. It investigates the decision-making processes by which various initial conditions are translated into outcomes (George and McKeown 1985; Falleti 2006).  16  institutional structure in the two cases, drives change (or not) and investigate how the choices made at critical junctures affect the final regime outcome. Based on qualitative analysis, in-depth elite interviews, and electoral and public opinion data, I study the strategies hegemonic parties use to stay in power and the conditions under which they give it up. My evidence is drawn from 60 elite interviews with party members, politicians, subject experts and archival sources collated over 12 months of fieldwork in Singapore and Taiwan during 2007-8. See Appendix A for the list of interviewees and archival sites in Singapore and Taiwan. Key Arguments In an age of democracy, the resilience of hegemonic party regimes depends more on institutions than coercion, charisma or ideological commitment. Unlike a military or personalistic authoritarian regime, a successful hegemonic party regime is characterized by its selective use of coercion9 and relies on the party as the key organizational tool to co-opt dissidents and mobilize mass support through semi-competitive elections. In this dissertation, I argue that a modern, successful hegemonic party is one that is effective in its strategic coordination – responsive in public good provision to earn mass support and selective in the withdrawal of coordination mechanisms to prevent the opposition from becoming a credible threat. When basic public goods such as healthcare, primary education, housing, national security and basic infrastructure are provided to a large proportion of its population, the hegemonic party is likely to be viewed as competent and legitimate.10 However, the withdrawal of coordination mechanisms such as freedom of speech, assembly and media freedom prevent the opposition parties from coordination and mobilization. Leadership succession is one of the gravest threats to hegemonic party stability. I argue that a key strategy for the hegemonic party to counter this threat is to “institutionalize charisma”, that is, to ensure that the charismatic leadership that pulls the party together is replaced by an institutionalized incentive structure that is centralized, autonomous and  9  In hegemonic party regimes, force is usually not manifested, as it is rarely necessary to maintain the capitalist organization of society (Przeworski 1986, 137). Calibrated coercion works better than brute force as the latter damages international image. 10 Legitimacy is defined as “the belief in the rightness of a state…so that commands are obeyed not simply out of fear or self-interest [but] because subjects believe that they ought to be obey” (Barker 1990, 11).  17  routinized. Based on candidate selection literature, I develop a leadership selection model to explain how the PAP’s elitist and oligarchic leadership selection system asserts centripetal forces to encourage the party elites to rally around party goals, display loyalty and step down without protest. In comparison, I show how the KMT’s leadership selection model of “uninstitutionalized charisma” during Chiang Ching-kuo’s rule (1975-1988) was unstable as the lack of routine, procedures and criteria opened up space for speculations, factionalism and party splits after the strongman’s passing. Unlike single-party states, hegemonic party regimes permit multi-party elections and claim political legitimacy through elections. In electoral democracies, democratization involves institutionalizing uncertainty – subjecting all interests to uncertainty (Przeworski 1986, 58). But in hegemonic party regimes, regime consolidation involves institutionalizing certainty – to increase the incumbent’s certainty of winning. When faced with declining electoral support, a strategic hegemonic party will tend to fall back on electoral engineering as its last line of defence to ensure that it ultimately prevails.11 Implications My dissertation explains a number of theoretical and empirical puzzles in post-ThirdWave studies. These range from institutional sources of hegemonic party resilience to the importance of leadership selection and the failure of institutionalization to ensure democratization First, my strategic coordination theory explains why high performing hegemonic party regimes are more resilient. Based on the comparative study of the PAP and the KMT, I show how hegemonic parties that invest in institutions and are strategic in their coordination of incentive distribution system that secures benefits to its ruling elites, supporters and voters are more stable and persistent. Both the PAP and KMT built a strong bureaucracy, powerful military, and a robust export-oriented economy. The mass legitimacy earned from providing public goods and selective restriction of civil and political liberties helped to insulate them  11  Confronted with new electoral realities, former hegemonic party needs to adapt, accept “institutionalized uncertainty” and “learn to lose” (Cheng and Lin 2008; Wong and Friedman 2008). While for others, survival means turning to electoral engineering and institutional manipulation as its “last line of defence” to institutionalize the certainty of winning (Schedler 2010).  18  from the social pressures for democratization that modernization brings. While Singapore remained resistant to liberal democracy, Taiwan took more than a decade to fully democratize. Both cases show that high-performing authoritarian regimes are better in controlling the pace of political liberalization and shaping the institutional design of electoral arrangements to their benefits than low performing ones. Second, my study on the origins of the PAP and the KMT lend support to the hypothesis that early formative years have an indelible impact on the party organization and chances of survival (Smith 2005; Huntington 1970). Likewise for the PAP and the KMT, their early electoral participation gave them a distinctive advantage. Both parties built massbase organizations to grow a wide network of para-political institutions to penetrate into the grassroots levels to police public discourse. Both capitalized on their incumbency and drew on state resources to develop their organizational and mobilizational capacities. My study also finds that early socio-economic structural pre-conditions are not deterministic. Contrary to Lipset and Rokkan’s thesis, the socio-ethnic cleavage structures in Singapore are not as immutable and did not “freeze” the party system (1967). Unlike the KMT, the PAP put in place a series of measures to dismantle and depoliticize socio-ethnic cleavages after coming to power. Through strategic coordination, the PAP prevented partisan alignment or mobilization based on ethnic or religious consciousness. On the other hand, the KMT exacerbated the sub-ethnic cleavage between the majority Taiwanese and minority Chinese from Mainland China. The KMT’s discriminatory policy that prevented Taiwanese from assuming key party and political positions reified the ethnic cleavage and provided the basis for its party system formation. The diametrically different approaches, which the PAP and the KMT took to address the socio-ethnic cleavages in their countries, are what I found to be one of the key factors that explained their divergent political trajectories. Third, my theory of leadership selection explains how certain hegemonic party regimes avoid the perils of leadership succession. How the PAP stays united and renews itself through an institutionalized leadership selection system is significant as it could become a model for other authoritarian regimes such as China, Vietnam or Burma to emulate. Unlike the KMT, which suffered factionalism and party splits after the passing of its strongman, the PAP defied all odds by engineering two smooth national leadership 19  transitions. In Chapters 3 and 4, I show how the oligopolistic PAP with centralized and exclusive selection methods is more cohesive than the KMT that experimented with inclusive and democratic selection methods in early 1990s. The PAP leaders put in place many institutional incentives and constraints at the systemic and party organization levels to ensure elite loyalty and cohesion. Through a complex elite recruitment and candidate selection process based notionally on meritocracy, the PAP institutionalized a process by which the Party as an organization incorporated the founder’s values. If the KMT had institutionalized similar leadership selection process in the 1980s, then its political fate may be quite different today. Fourth, current democratization and party politics literature often assumes that party system institutionalization leads to democratic consolidation.12 But in this dissertation I argue that the concept of institutionalization ought to be distinct from democratic consolidation. My study of the PAP shows that party and party system institutionalization can occur in electoral authoritarian regimes. In Chapter 7, I argue that institutional manipulation and electoral engineering helped to institutionalise a semi-competitive party system that systematically disadvantaged the opposition in Singapore.13 Faced with declining vote shares, the PAP “manufactured” its legislative supermajority through a series of electoral reforms that fundamentally altered inter-party competition and voting behavior. The notion that a hegemonic party regime may be institutionalized is significant as it highlights the teleological bias in transitology paradigm and demonstrates how hegemonic party regime may be more persistent despite regular elections. Fifth, my study contributes to the emerging studies on electoral authoritarianism to show when elections might bring about democratization (Lindberg 2009). My theory of strategic coordination shows that when the coordination mechanisms (freedom of press and assembly) increase to level the playing field, more competitive elections are likely to bring  12  For example, Mainwaring and Scully have argued that an institutionalized party system is necessary for democratic consolidation (1995, 1). Likewise, Dix also assesses the prospects of democratic consolidation in Latin America based on the institutionalization of its party system (1992). 13 Besides Singapore, Brazil (1967-1969) and Chile (1980) also show that military regimes may be institutionalized as new constitutions were adopted and formal rules were established to regulate the power structure within the regime and government functions were assigned to the armed forces (Aguero 2000).  20  about democratizing outcomes, as in Taiwan during the late 1990s. If high-performing regimes such as Singapore are able to maintain mass support and selectively withdraw coordination mechanisms to prevent opposition mobilization then the opposition will remain small, inferior, and incapable of posing a credible threat. Under these conditions elections are unexpected to bring about party alternation. Finally, my dissertation offers a three-level analytical framework to integrate agency and structure by showing how incentive structures shape behaviour at the party organization, party system and regime level. Building on existing theories, I identify the key processes and mechanisms that encourage elite unity, opposition party failures and mass support of the hegemonic party – conditions necessary for the survival of the hegemonic party. Unlike most studies that theorize on the individual level or too broadly at the regime level, my analytical framework highlights the incentive structures within party organization and the party system to show how they affect the calculations of key actors. My dissertation shows how disaggregating the broad processes may help us understand how the incentive structures shape the preferences and behaviour of actors. See Table 1-1 for a summary. Table 1-1: Dissertation Structure Based on Three-Level Analytical Framework Party Organization Party System Regime Focal point Interactions between Interactions between parties Interactions between party elites; elites and parties & voters members Unit of Individuals and factions Parties as unitary actors Parties and voters analysis Structure Socio-ethnic, economic Electoral system Political system context Processes Party institutionalization Electoral Engineering Strategic coordination Mechanisms  Elite recruitment Candidate selection  Selective tweaking of electoral rules  Outcome  Party cohesion  Semi-competitive party system  Public goods provisions and withdrawal of coordination mechanisms for opposition Electoral hegemony  Case Selection My case selection has strong methodological and theoretical rationale. Based on JS Mill’s method of difference, I have chosen to compare Singapore with Taiwan as they share many similarities but have different regime outcomes. For example, both regimes share 21  similar Sinitic roots, Confucian ethos and formidable militaries. Ethnically, they have a large population of migrant Chinese from China such as Fujian, Guangdong and Jiangxi. Economically, the two capitalist, export-oriented economies were also leading Tigers in the region that grew under authoritarian rule. These similarities allow for some qualitative control over cultural, socio-economic and ethnic factors. However, Taiwan abandoned the yoke of authoritarianism in the 1980s; Singapore remains a hegemonic party regime. My focus on the varied incentive structures of the two regimes helps to explain why this is so. Singapore and Taiwan represent the failure and success of the Third wave. Singapore is the world’s wealthiest non-oil producing country that is not a democracy. Its refusal to embrace competitive multipartyism defies democratization theories. Indeed, the story of how the PAP maintains elite cohesion and wins elections again and again after 5 decades is an exceptional one that deserves explanation. On the other hand the KMT was one of the wealthiest and most durable hegemonic parties in the world. After 51 years a combination of leadership struggles, rising middle-class consciousness and international pressures led to the regime’s electoral opening and lost of government power in 2000. The KMT’s adaptation from a hegemonic party to a dominant party provides an excellent counter-case to test my institutionalist theory of hegemonic party resilience. Studies indicate that the trend of democratization is weak in Asia (Croissant 2002, 1; FH 2008). Yet, no studies have satisfactorily explained why this is so. One reason could be because Euro-centric party politics theories do not neatly apply to the developments in Asia (Blondel 2006).14 Based on the comparative study of Singapore and Taiwan, my dissertation hopes to offer more refined hypotheses and generalizable mid-range theories that can be further tested in a larger sample and explain the persistence of hegemonic parties in Asia and other regions in Latin America and Africa (see Lijphart 1975, 685). Presently, no work has been done to compare the party organization and electoral strategies of the PAP and the KMT in a systematic way.15 My comparative study addresses this gap and contributes to the democratization literature by moving away from the single-  14  And others would argue in Latin America too (Mainwaring and Scully 1995). Mathew Towner’s dissertation (1997) compares Taiwan and Singapore. But his work focuses on Chiang Ching-kuo and Lee Kuan Yew and makes little reference to party organization and electoral strategies. 15  22  case study method, more commonly used in the study of democratization in Asia. The divergent political trajectories of Singapore and Taiwan hope to address the teleological bias in the transitology scholarship and help expose the selectivity problem in the “Asian values” cultural thesis. Plan of the Dissertation My dissertation contains 9 chapters and the research chapters are divided into 3 parts. I begin with an Introduction and Theoretical Chapter that lay out the key questions, gaps in the literature, rationale behind the case selection and offers a three-level theoretical framework on strategic coordination to explain hegemonic party resilience. In Part I, I trace the sources of institutionalization by examining how party origins and organizational models of the PAP and the KMT affect strategic coordination of the party elites. In Chapters 3 and 4, I show how socio-ethnic cleavages affect party system formation in Taiwan, but not in Singapore. The empirical evidence is drawn from party publications, archival data and elite-interviews. In Part II, I consider the effects of intra-party mechanisms and processes and compare how different leadership selection models in the PAP and the KMT affect party cohesion. In Chapters 5 and 6, I examine how elite recruitment and candidate selection methods for top party and political positions become tools for distributing patronage, settling disputes and reinforcing elite unity; and the conditions under which these processes fail to bind behaviour, leading to elite defections and party splits. Empirical data on factionalism, party defections and party discipline will be used to support my analysis. In Part III, I examine the role of elections and electoral institutions on party system formation in Singapore and Taiwan. I study the conditions under which the electoral institutions are manipulated or de-institutionalized, bringing about unintended regime outcomes. In Chapter 7, I demonstrate how the PAP institutionalized an uneven playing field and “manufactured” its legislative supermajority despite its declining vote shares. In Chapter 8, I examine the role of local elections and study how recent electoral reforms concentrate the party system and reward the larger parties in Taiwan. Measures of electoral proportionality, volatility, fragmentation and electoral legitimacy are used to assess the mechanical and 23  psychological effects on parties and voters. I end by comparing the degree of party system institutionalization in Singapore and Taiwan and prospects for regime consolidation. In the concluding chapter I summarise the key findings of the dissertation and review the lessons learnt from Taiwan’s evolutionary party system change and implications for Singapore’s prospects for more competitive, multi-party system.  24  2  Strategic Coordination, Institutionalization and Electoral Engineering Apart from exogenous shocks that destabilize all authoritarian regimes, internal splits  and leadership succession are the two biggest threats to hegemonic party stability (Haggard and Kaufman 1995; Huntington 1970; O’Donnell and Schmitter 1986). To explain how some hegemonic parties are better equipped to avoid these threats, this chapter introduces a threelevel theoretical framework to show how strategic coordination and institutionalization of leadership selection maintain elite unity and mass support. As hegemonic parties claim their legitimacy to rule through elections, my explanatory framework considers the conditions under which institutional manipulation might disadvantage the opposition and increase the incumbent’s chances of winning. This chapter begins by introducing a typology of single-party regimes that allows us to study how hegemonic parties evolve from one sub-type to another. Then, I review existing theories and problems of authoritarian durability before presenting my theoretical framework that will be used to explain the subsequent chapters on Singapore and Taiwan. A Typology of Single-Party Regimes Single-party regimes vary in party competitiveness and electoral certainty. Here, I propose a typology that includes 3 sub-types of varied competitiveness and pluralism: 1) dominant party, 2) hegemonic party and 3) one-party state. The point of this tripartite typology is to merge competitive and non-competitive regimes into one framework. The focus is on the regime1 rather than the party system because the ‘one-party system’ is a misnomer (Sartori 1976, 44). It is important to distinguish the different single-party sub-types as varied electoral competitiveness could lead to different liberalizing outcomes (Howard and Roessler 2006; Norris 2009). My typology, designed based on the following 4 dimensions: 1) political participation; 2) contestation; 3) legislature dominance and 4) party alternation would allow us to compare the various single-party regime sub-types in a meaningful way.  1  My analysis is on the regime-level so to highlight a system of governance that determines the methods of access to the political offices; the characteristics of the actors admitted to or excluded from such access; the rules of the game for access (Schmitter and Karl 1996, 50; Munck 1996, 6).  25  Regimes are not static. They adapt and evolve. My typology allows us to assess the transformation or adaptation from one sub-type to another.2 Political participation refers to the right to vote. This dimension refers to: fairness of the electoral process, access of parties to public funding and the extent of suffrage - the proportion of the population and entitlement to participate in the contestation for government office (Dahl 1971; Munck 2002). The way the regime impinges on the procedures affect the opposition’s access to public offices. In a dominant party regime, voters enjoy universal suffrage and the opposition candidates can participate freely in elections. In a hegemonic party regime, voters enjoy universal suffrage but the opposition candidates face formal, legal barriers to entry such as the lack of independent electoral commission or legal restrictions. On the other hand, in a one party state, opposition parties are banned or excluded. The second dimension, contestation, refers to the degree to which the regime permits opposition. This includes the right to compete, the freedom to form parties, freedom of the press and access to alternative information (Dahl 1971; Munck 2002). Electoral rules matter and may be categorized into three types: autocratic, cartel and egalitarian (Norris 2009, 1523). While autocratic regulations are those that are explicitly skewed towards the ruling party and restrict the opposition to prop up the one-party state, cartel regulations limit party competition through a variety of practices designed to benefit established parties such as ballot access; allocation of public funding, access to campaign subsidies or high minimal vote thresholds to achieve elected office. Egalitarian rules are more permeable and open, facilitate pluralism, and competition among contenders enabled their access to resources. In a dominant party regime, opposition parties are free to contest in an open and fair environment. Parties and candidates can nominate and campaign freely with minimal restriction on civil rights and freedom. As for hegemonic party regimes, elections are free but unfair. Free elections induce opposition actors to form parties and compete for votes. However, elections are unfair as there are biases in partisan competition that tilt the playing field which means the opposition parties are unlikely to win. Opposition parties are disadvantaged and antagonistic contestation is not permitted through legal means or else  2  See Tien (1989, 7-12) for an excellent discussion of the various types of single-party rule.  26  deprived of rights of expression and organization (Sartori 1976). Skewed playing fields enable the incumbents to retain power without resorting to blatant electoral fraud, which would undermine their international reputation (Levitsky and Way 2010). A hegemonic party sustains power through asymmetric access to state resources and selective restrictions in civil and political freedoms which impair the opposition’s ability to organize and compete. In oneparty state, contestation is minimal or non-existent as opposition parties are disallowed and independents have slim chance of winning seats. The third dimension measures the legislative dominance of the party. If a major party receives a plurality of votes and manages to win a majority or more than 50% of seats (not necessarily of votes) in Parliament, then it qualifies as a dominant party. If a party receives a majority of votes and occupies more than 70% of seats (or two-thirds majority) in Parliament and enjoys a dominant bargaining position, then it ought to be viewed as a hegemonic party.3 Unlike the dominant party, a hegemonic party strives for a supermajority (more than 70%) in legislature so as to control institutional changes to their advantage. As opposition parties are formally banned in a one party state, the incumbent is expected to gain more than 90% of the votes and nearly all the seats in the legislature. The ability of citizens to change their government is the hallmark of democracy. The prospects of party alternation may be used to distinguish the different single-party sub-types. While turnover is expected in a dominant party regime, in hegemonic party regimes alternation can occur but does not (Sartori 1976, 230). Constraints in civil-political liberties and media freedom prevent the possibility. While there is an avenue for party alternation, it seldom occurs as a semi-competitive party system suppresses the rise of a credible replacement. In a one-party state, alternation through elections cannot occur. Based on these 4 dimensions, a one-party state refers to a closed, non-competitive autocracy where national elections are not permitted and outcomes are certain. Examples include former Soviet Union, China, Vietnam and Laos. On the other hand, a dominant party regime4 refers to a competitive, electoral party system where a party regularly wins multiple  3  Any threshold of quantitative support is inevitably arbitrary. I use the percentages of seat shares that are widely used in the literature. See Sartori (1976, 195); Pempel (1990, 3) and Magaloni (2006, 15). 4 For simplicity, I retain the use of “dominant” rather than “predominant” as Sartori suggests.  27  or successive elections which are relatively free and fair and is supported by a winning majority (absolute majority of seats). Minor opposition parties are legal, legitimate and independent antagonists of the dominant party, as in TJ Pempel’s Uncommon democracies.5 Sometimes, the dominant party has to form coalitions to maintain legislative majority. Examples include Japan under the Liberal Democratic Party (LDP) rule and Italy under the Christian Democratic Party (CD). Finally, the term hegemonic party regime refers to a polity where a party exercises tight control over the electoral arena, rules and players, leaving little room for contestation. Opposition parties serve as second-class, licensed parties that are not permitted to compete with the hegemonic party in antagonistic or equal terms. The ruling elite puts up the appearance of competitive elections, but in reality do not tolerate open contestation. Harmless opposition candidates are allowed to run but party alternation is not viable because of an asymmetry of power between the hegemonic and satellite opposition parties. The hegemonic party system functions as two level system in which one party tolerates and allocates a fraction of its power to subordinate political groups (Sartori 1976). In short, a hegemonic party regime is a “diminished” form of authoritarianism rather than a “diminished” democracy (Carothers 2002). Examples include Taiwan under the KMT (1949-1991), Mexico under the Institutional Revolutionary Party6 (PRI, 1929-1997) and Singapore under the PAP (1968-present). See Table 2-1 for a summary.  5  The quality of democracy in dominant party regimes is controversial. See Sartori (1976, 48-51), Scheiner (2006, 7-30) and Ware (1996, 245-254). 6 The PRI won every presidential election from 1929 to 2000, held the majority in Congress until 1997, won every governorship until 1989 and controlled the vast majority of municipalities. For excellent studies on the PRI rule, see Magaloni (2006) and Greene (2007).  28  Table 2-1: A Typology of Single-Party Regimes Degrees of Inter-Party Competitiveness Non-competitive Competitive Dimensions One-Party Hegemonic Party Dominant Party Participation No suffrage Universal suffrage Universal suffrage Contestation Autocratic or no Relatively free but unfair Free and fair elections under elections elections under cartel egalitarian regulations regulations Legislative Absolute control Supermajority Absolute majority dominance (90 -100% of seats) (more than 70% of seats) (50% or more seats) Party Cannot occur Can but does not occur Can and may occur alternation Examples China (CCP) Singapore (PAP, 1968-) Japan (LDP, 1955-1993) Vietnam (VCP) Mexico (PRI, 1929-1997) Taiwan (KMT, 1990-present) Taiwan (KMT, 1978-1989)  Theories of Authoritarian Durability Explanations of authoritarian durability and breakdown are two sides of the same coin. Within the democratization literature, theories of regime breakdown often fall within three camps: 1) structuralist, 2) institutionalist and 3) culturalist. To begin with, structuralists usually build on Marxist or Weberian tradition to focus on large-scale phenomena such as: socio-economic conditions (Lipset 1959; Dahl 1971); rise of a robust middle class (Moore 1966); class relations (Rueschemeyer 1980; Stephens 1989); revolutions (Skocpol 1979) or stateness (Linz and Stepan 1996) to explain regime change. Amongst these, Lipset’s (1959) modernization thesis is most widely cited to link economic growth to single-party breakdown in post-Communist states in East Europe (Przeworski 1991). Like regime studies, party politics literature often turns to structuralist approaches to explain how social cleavage structure affects party system stability and change (Lipset and Rokkan 1967). Based on Western European party formation experiences class structure, religious or ethnic differences are considered as causal factors for ideological, strategic or electoral shifts in the party system (Mair 1997). Institutional factors such as type of electoral system, size of district magnitude and type of electoral engineering strategy are used to explain party system change and stability (Duverger 1954; Cox 1997). Dissatisfied with the over-determinism in structural analysis, scholars have merged elite-based explanations with structural analysis to show how regime transitions may be 29  “crafted” or negotiated between hard and soft-liners in the form of “pacts” (O’Donnell and Schmitter 1986). From this perspective, authoritarian stability and transition depends on the bargaining power or short-term calculations of negotiating elites. For example, building on O’Donnell’s (1988) seminal work on bureaucratic authoritarianism, later structuralinstitutionalists such as Haggard and Kaufman (1995), Remmer (1990), Liberthal and Oksenberg (1988) and Wallerstein (1980) highlight the institutional conditions of authoritarian regimes that shape elites’ preferences and capacity to deal with economic and legitimacy crises in single-party regimes. Besides structuralist analyses, institutionalist7 explanations have also dominated the literature. Generally, institutions refer to the formal or informal procedures, routines, norms and conventions embedded in the organizational structure of the polity (Hall and Taylor 1996). Broadly, institutionalists focus on the organizational and institutional features such as the decision-making authority in the executive (Geddes 1999); role of party organization (Panebianco 1988); degree of party system institutionalization (Huntington 1968; Levitsky 1998; Mainwaring and Scully 1995); inter-institutional relations in presidential or parliamentary systems (Linz 1994); separation of coercive and political resources (Haggard and Kaufman 1995); use of legal barriers to prevent opposition challenge (Molinar 1991) or technocratic or bureaucratic agencies that equipped elites to co-opt social groups during crisis to explain the resilience. Within this tradition, Perlmutter’s work is perhaps most representative to explain how the institutional structures, instruments of control secure mass support and legitimacy in authoritarian regimes (1981). Apart from structural and institutional analyses, coercion and tyrannical leadership are also used to explain authoritarian resilience (Alagappa 2001; Bellin 2005; Wintrobe 1998, 2007). In this school, scholars focus on systematic state repression, violations of human rights, torture, arbitrary imprisonment, restrictions of civil liberties and freedom of speech as explanatory factors for authoritarian persistence. But not all authoritarian regimes are equally repressive.8 For example, Davenport’s (2007) study of 137 regimes shows that one-party regimes are less repressive than other authoritarian regimes. Besides, Smith’s (2005, 427)  7 8  See Thelen and Steinmo (1992), Skocpol and Somers (1980) and Pierson and Theda (2002). For an excellent quantitative study on repression and human rights violation, see Davenport (2007).  30  study on repression based on Polity III scores found no systematic relation between the level of political violence and regime viability.9 In fact, hegemonic-party autocracies are a comparatively more benign form of authoritarianism (Magaloni 2006, 10). The decision to use coercion is often a question of agency and strategic interactions between power brokers such as the military and the opposition elites. The focus on coercion alone is thus insufficient, as it needs to explain the timing and selective use of coercion to govern. Drawing from anthropology, culturalists posit that a common set of shared meanings, values and belief systems, embodied in a society’s institutions and organizations, shape the way of life and worldview in a regime (Almond and Verba 1989; Diamond 1993). Culturalists contend that a collection of attitudes or values, such as Confucianist-inspired Asian values or democratic culture, are important tools for legitimacy or mass mobilization. For example, Ottaway’s (2003) work on semi-authoritarianism shows how racial, cultural and religious differences explain authoritarian persistence in the era of neo-liberal democratization. In East Asia, scholars and political leaders have argued that East Asian countries with similar group consciousness and political culture are unlikely bedfellows of liberal democracy (Harrison and Huntington 2000; Kausikan 1993).10 For example, Pye claimed that East Asian countries are more likely to exhibit certain group consciousness in their political culture that is distinct from the kind of individualism found in Western culture, the source of liberal democracy. Advocates of Asian values11 stressed that the existence of a shared identity and values among East Asians are different from the West. The inference is that democratization is antithetical to East Asians as they have different view of human rights.12  9  Polity III score is a composite measure of different kinds of repressive or coercive government action. Smith’s study on the impact of fiscal and political constraints on regime durability found no relationship between economic performance and regime survival (2005). 10 Pye asserted that: “There do seem to be some features of Asian civilizations that have set them apart from western civilization. Probably the most significant of these is the Asian tendency to place more value on the collectivity and to be less sensitive than the West to the values of individualism” (1985, 26). 11 Asian values are based on Confucianism that include: an emphasis on the community; a strong state; deference toward authority; diligence; strong work ethic; focus on harmony and consensus. 12 A more extreme strand of culturalist approach suggests that authoritarianism is necessary for the growth and stability as liberal democracy is disorderly, confrontational and inefficient. The IMF has cited negative Asian values such as nepotism, cronyism and patrimonialism as causes of the Asian Economic Crisis in 1997. See Pye (2000, 244-255) and IMF Finance & Development, June 1998.  31  The controversy over the compatibility between Confucian societies and liberal democracy, otherwise known as the Asian Values debate, has pitted proponents justifying the exigency of authoritarian governance based on cultural relativism against critics arguing for the universality of democracy (Kausikan 1993; Mahbubani 1995; Sen 1997).13 There are flaws with the cultural argument. Apart from the selective reading of Confucianism and oversimplification of the cultures deriving from Islam, Buddhism and Christianity, there are no quintessential values that apply to a large and heterogeneous Asia (Diamond 1988, Tay 1997).14 The fact that South Korea and Taiwan have embraced liberal democracy shows that it can thrive in East Asian Confucian societies.15 Problems and Gaps In an age of democracy, the resilience of Singapore’s hegemonic party rule is a puzzle (Haas 1999). A small island with a population of less than 4.6 million, Singapore is the wealthiest non-oil producing country in the world that is not a democracy.16 Despite its affluence and ideal socio-economic preconditions for democracy, the PAP has ruled the country uninterrupted since 1959. No substantial opposition party has existed since and electoral competition is weak. Contrary to expectations, the large middle-class in Singapore does not agitate for change but remains “passive, deferential, acquiescent, and lacking political mobilization” (Sinnott 2006, 45).17 Singapore’s refusal to embrace competitive party politics confounds democratization theorists, leaving some to exclude it as a deviant case that permits no meaningful comparison (Neher 2002, 174). As Huntington said, “the anomaly remains Singapore” (1993, 38). The case of Singapore shows that social pressures arising from development are necessary but insufficient for democratization. As Bueno De Mesquita and Brown caution, modernization theorists underestimate the ability of oppressive governments to thwart  13  To critics, cultural argument based on Asian values is an “illusion concealing the iron grip of petty despots”, a defence for authoritarianism (Lingle 1995, 193; Kessler 1999, 7). 14 There is no authoritative study to demonstrate that Asian values are hostile to liberal democracy. The literature is replete with contradictory public opinion data. See Dalton and Ong (2003) and Blondel (2006). 15 The rejection of the Asian values discourse by former leaders of South Korea and Taiwan, Kim Dae Jung and Lee Teng-hui also points to the intra-regional differences within Asia (Kim 1997). 16 Per capita income surpassed US$10K in 1990s and US$20K in 2000. See Statistics Singapore (2008). 17 See World Values Survey 2008 that found Singaporeans to have no strong demand for more democracy.  32  demands for social and political change (2005). The idea that autocrats could embrace economic growth while postponing democracy through strategic coordination and suppressing coordination mechanisms such as political rights, human rights and media freedom is important and will be further developed in the next section. Current literature has overlooked how effective authoritarian regimes turn democratic institutions to their advantage and manipulate the rules of the game through constitutional and electoral engineering to suit their interests. With a supermajority of seats in the legislature, autocrats enjoy a marked advantage over the average citizen and opposition in their ability to shape institutions and politics. Without turning to electoral fraud and ballot rigging, autocracies can institutionalize a semi-competitive electoral arena and deny the opposition the ability to coordinate. Consequently, voters are constrained by a series of dilemmas that compel them to vote for autocracy, as in Mexico from 1940 to 198018 (Magaloni 2006, 19). The ability of hegemonic-party regimes to survive exogenous shocks such as economic crisis also depends on the capacity of rulers to retain the loyalty of elites who have control over the coercive, administrative and political resources (Haggard and Kaufman 1995, 64). Indeed, in response to the global credit crisis in 2008, Singapore’s PAP was quick to dip into its national reserves, an estimated amount of US $170 billion to fund a US$13.74 billion “Resilience Package” to stabilize its economy.19 The decision to use the national savings was made for the first time in the country’s history and the initiative was passed without much debate. As Nobel Prize winner Michael Spence, says, “The use of reserves to stabilise the net capital flows is the most important domestically controlled circuit breaker” (Straits Times 31 Oct 2009). With the PAP government’s direct control over the national resources, it is clearly in a better position to ward off economic shocks.20  18  Also see Greene (2007) who contends that opposition coordination failure contributed to PRI’s dominance in its final decade and PRI’s dominance was unsettled by economic conditions beginning in 1982. 19 By April 2009, estimates show Singapore’s economy have shrunk by 11.5% compared to 2008. The “resilience package” was designed to save jobs and keep local companies afloat. “Singapore Taps Reserves for S$20.5b Economic Stimulus Plan,” Channel NewsAsia, 20 Jan 2009. 20 Some argue that mass support in times of crisis reflect Singaporeans’ trust the government’s ability to lead them out of recession (Straits Times 31 Mar 2009). While others contend that the 1998 economic crisis was not as bad as others such as the one that Mexico experienced in the 1982 and 1997 fiscal crises under the PRI’s rule. Bratton and Van de Walle demonstrated that mass support for some authoritarian African regimes doubled  33  Institutionalist explanations have been used to highlight how institutions and strategic control over economic, political and coercive institutions explain hegemonic party resilience. For example, theorizing at the individual-level, rationalist-institutionalists focus on the behaviour of party elites and show how incentive structures, incumbency advantage and opposition coordination dilemmas explain the survival of hegemonic parties (Geddes 1999; Greene 2007, Magaloni 2006). In this approach, structural factors such as ethnic cleavage, class relations or historical factors become secondary contextual conditions that affect the cost-benefit calculations of actors. For example, Geddes’ study shows how single party regimes are equipped with the incentive structures to prevent elite defection. Based on gametheoretical logic, intra-party struggles are modelled as a stag-hunt game. Party elites are expected to co-operate as they have little incentives to defect and have no barracks to return to, unlike military leaders (Geddes 2003). My dissertation builds on the institutionalist approach to highlight the importance of behavioral incentives. However, it goes beyond Geddes’ game-theoretical approach that takes the origins of party and electoral institutions as given. As Smith notes, “treating party institutions as prior variables makes it nearly impossible to figure out the incentives within such regimes might come or not come, to look very much like the Stag Hunt game” (2005, 427). An account linking party institutions to regime outcome is missing in rationalistinstitutionalist story.21 Existing institutionalist explanations tend to overlook the origins of party organization foster elite unity and show how electoral institutions could dampen interparty competitiveness – preconditions for hegemonic party durability. I build on institutionalist explanations by showing how strategic coordination and institutionalization22 of a self-reinforcing incentive distribution system that ensure the long-  during crises as the public looks to the government to restart growth (1997, 36). On average, single-party regimes have been remarkably resilient in the face of long, severe economic crises. However, economic shocks undermine all authoritarian regimes and may impede the distribution of benefits to supporters and allies, and destroy coercive capacity (Geddes 1999, 24). 21 In Geddes’ explanation, leadership struggles do not affect the party cadres’ desire to defect. As long as benefits of cooperation are sufficient to ensure co-operation, single-party regimes will endure. See Elster’s (1989) critique of rational choice theory and assumption of methodological individualism that fail to explain motivation and behavior. He argues that behaviours such as wishful thinking, sour grapes and lack of foresight may defy rationality and result in non-cooperative behavior. 22 Institutionalization is a process by which rules or patterns become routinized, stabilized or entrenched (Levitsky 1998; North 1990).  34  term benefits of ruling elites, supporters and voters build regime resilience. Institutions23 matter because they serve as the rules of the game that shape interaction and “reduce uncertainty by providing a structure to everyday life” (North 1990, 3). They provide stability and predictability in hegemonic party regimes. The mechanisms and processes that bring about elite unity are different from those that enable a party to win elections. A complete institutionalist argument needs to identify the organizational structure in which actors operate, the rewards and punishments that motivate elite co-operation or defection, and the rules and regulations that shape the parties’ chances in the electoral market. In any study of single-party rule, scholars have to decide on the level of analysis, to theorize on the party organization, party system or regime level. But to focus on one and not another risks oversimplification and misses the link between institution and regime outcome. To address this, my dissertation offers a three-level theoretical framework that integrates analysis at the party organizational, party system and regime level (Barnea and Rahat 2007). This framework focuses on how behavioural incentives shape the preferences of the party elites, parties and voters. It highlights how the interactions of the political processes and control mechanisms24 at three levels that bring about the final outcome – regime resilience.25 This framework is useful as it allows one to 1) identify and study the underlying political processes and mechanisms at each level; 2) examine the utility of existing explanations and propose hypotheses that contribute to broader, complex theories; 3) compare the effects of specific process or mechanism in the two selected cases in a structured and focused way. Theoretical Framework Why are some hegemonic parties better in maintaining cohesion and winning elections? I argue that an institutionalized hegemonic party regime with an effective strategic coordination of incentive distribution system is more likely to keep elites united, maintain  23  Institutions are formal and informal procedures, routines, norms and conventions embedded in the organization structure of the polity or political economy (Hall and Taylor 1996, 398). 24 I treat mechanisms as “intervening variables” that link the independent and dependent variables together in a causal inference. They are intermediate between laws and descriptions, which allow us to explain but not to predict (Elster 1989, 45). 25 Resilience refers to the ability of the regime to recover quickly to its original institutional arrangement when challenged; durability refers to the ability of the hegemonic party to withstand challenges and remain in government, measured by its longevity in government. I use the two terms interchangeably.  35  mass support and deter opposition challenge. Strategic coordination refers to a set of activities that the ruling elite actively engages in to maintain political power (Bueno De Mesquita and Brown 2005). Hegemony is never complete but rests on the ruling party’s ability to improve the material life of the governed (Gramsci 1971; Weber 1958). Strategic coordination is necessary to ensure that material incentives, socio-legal and electoral constraints are in place to appease the ruling elites, supporters and voters. For its long-term survival, the hegemonic party needs to selectively suppress opposition activity without undermining economic growth or the provision of public goods. If it is able to maintain performance legitimacy and limit the opposition mobilization by restricting the availability of coordination mechanisms26, then, it can reduce its chances of being replaced. To assess the regime’s provision of public goods, indicators from the World Bank Governance data on “Government effectiveness” may be used. For example, a regime with 60% and above rating may be considered effective in public good provision; rating below 60% may be classified as ineffective that lowers its legitimacy and mass support. Coordination mechanisms refer to a system of rules, laws and processes that govern civil, political liberties and media freedom. Coordinating mechanisms benefit the opposition parties as greater civil and political liberties increase their ability to recruit, communicate with and mobilize their supporters. However, restrictions over campaign rules, state control media and biased reporting impair the opposition’s publicity, fund raising and recruitment capabilities. As a measure of the availability of coordination mechanisms in the regime, FH’s scoring of political rights and civil liberty may be used.27 While the scores of 3 and below mean a freer electoral environment with fairer inter-party competition, regimes with scores of 4 and above imply a more repressive environment that stifles opposition. Hegemonic party regimes differ in their provision of public goods and restriction of coordination mechanisms. The different capacities to provide and repress, and the consequent effects on regime outcomes are presented in a two by two matrix in Figure 2-1. This strategic  26  This is similar to the Dahlian calculus of the cost of repression as compared to the cost of toleration (1971). The withdrawal of coordination mechanisms are less costly than brute force as it is difficult to isolate its effects and do not directly threaten economic activities. 27 Also see Howard and Rossler (2009) and Hadenius and Teorell (2007) who use both FH and Polity data to measure the degree of civil and political liberties in electoral authoritarian regimes.  36  coordination matrix is designed to explain the conditions under which: 1) hegemonic party regimes are most stable and likely to persist; and 2) opposition parties are likely to coordinate and increase the competitiveness of elections to bring about democratizing outcomes. For example, the upper left column in Figure 2-1 shows that hegemonic party regimes are stable when there is an adequate supply of public goods to earn mass support and where restriction of political and civil liberties prevent the opposition mobilization such that they remain small, weak in organization and ephemeral (Rakner and Van de Walle 2009). A regime is stable when there is an absence of preferable alternatives (Przeworski 1986, 51-2). Under this condition, the opposition that is systematically disadvantaged is prevented from becoming a credible replacement. Singapore under the PAP (1968-present) and Taiwan under the KMT (1975-1988) are prime examples of this regime type. Figure 2-1: Strategic Coordination of Hegemonic Party Regimes (HPRs) Availability of Coordination Mechanisms (CM) for Opposition Stable Competitive  Public Goods Provision (PG)  Hi PG, Lo CM  Hi PG, Hi CM  HPRs enjoy mass support; selective repression results in opposition failure  HPRs enjoy mass support; more competitive elections may lead to liberalizing outcome  Singapore PAP (1968-present) Taiwan KMT (1972-1989) Malaysia UMNO (1959-present)  Taiwan KMT (1990s) S Korea (1961-1987)  Unstable and Uncompetitive  Unstable and Competitive  Low PG, Low CM  Low PG, Hi CM  Unresponsive and repressive HPR  Unresponsive HPR and effective opposition, most likely to lead to regime change  Burma (2010) Cambodia (1990s) Congo, Angola, Zimbabwe, Sudan  Indonesia’s Golkar Party (1990s) The Philippines’ KBL (1980s) Mexico PRI (1990s)  If high performing hegemonic party regimes are able to provide public goods but coordinating mechanisms are also available the opposition are in better position to mobilize support and compete effectively in a more level playing field. See upper right column in 37  Figure 2-1 for effects of elections in competitive hegemonic party regimes. With increased space and mechanisms for opposition coordination, competitive elections in hegemonic party regimes are likely to bring about non-violent, liberalizing outcomes as in the case of Taiwan in the 1990s after electoral opening. As transitology studies show, single-party and hegemonic party autocracies are more likely to negotiate its withdrawal from authoritarian rule (Geddes 1999). Under this condition, the opening of the electoral market may result in the incumbent’s gradual loss of power through electoral means.28 Regime transitions may occur in the form of elite “pact negotiations” to abolish old institutions and to erect new ones. Given that voters are utility maximizers, they cannot simply “throw the rascals out” as their choices are constrained by a series of dilemmas that compel them to support the hegemonic party (Magaloni 2006, 20). For high performing hegemonic party regimes, voters may continue to vote for the incumbent because of fears of being made worse off with an untested contender and preference for predictability.29 Having been ruled by the same party over a long time, voters may not have any prior knowledge or experience to mitigate their fear of the unknown. They may avoid voting for opposition out of fear of instability, reprisals or negative impact on the economy. Even as they acknowledge compromise in their civil liberties, voters vote for the hegemonic party as it is the “known evil” that has a record of delivering goods.30 Without fair media coverage, voters may not have all the necessary information to assess the credibility of the opposition. The availability of coordination mechanisms is thus one key factor that tilts the level playing field between the incumbent and the opposition. In the lower left column, hegemonic party regimes are unstable and uncompetitive as the incumbent party fails to deliver public goods and resorts to more repression to prevent opposition mobilization from challenging the ruling party. These low performing and repressive regimes are usually stuck in the “grey zone” and conduct façade elections to pay lip service to democratic norms. Examples of this regime type are found in post-communist regimes and electoral authoritarian regimes in Sub-Sahara Africa such as Burkina Faso,  28  For studies on the liberalizing effects of regular elections in authoritarian regimes, see Lindberg (2009). In Singapore, there is a popular, colloquial Hokkien phrase - “kiasu” which means “afraid to lose” that captures the essence of the voter’s dilemma in a hegemonic party autocracy. 30 In economic crisis, voters usually rally behind the “known evil” with the past record of delivering (Geddes 2003; Haggard and Kaufman 1995; Magaloni 2006). 29  38  Angola, Zimbabwe and Congo (Lindberg 2009; Schedler 2009). The incumbent party remains in power despite its low legitimacy. Voters are constrained, as there is a lack of viable alternatives and repression causes opposition coordination failure. Under this condition, the hegemonic parties are more likely to breakdown as a result of internal splits and struggles over spoils; rather than through electoral means. Finally, low performing hegemonic parties that are unable to deter the opposition challenge are likely to be replaced. As lower right column of Figure 2-1 shows, in unstable and more competitive hegemonic party regimes, the opposition are likely to cooperate as their prospects of winning are increased. The availability of coordination mechanisms enables the opposition to mobilize and boosts their public image as a credible alternative. Presented with a viable replacement, voters may vote for the opposition if the hegemonic party fails to deliver. Under this condition, party alternation is likely to result if the incumbent chooses not to repress, as in the case of Mexico under PRI – which gave up power through elections. And if the hegemonic party represses and fails, then, the low performing hegemonic party may be ousted through mass uprising as in the case of the Golkar Party in Indonesia during Surharto’s rule or the Kilusang Bagong Lipunan (KBL) in the Philippines during Marco’s rule. The next section will now consider the incentive structures within the party organization that foster elite unity and party cohesion. Party and Party System Institutionalization Institutions matter. I argue that the underlying strength of hegemonic party regime depends largely on the degree to which the ruling party is institutionalized or moved through the phases of transformation, adaptation and consolidation. Party institutionalization31 refers to the process by which a party acquires stability and values, and becomes established in organizational terms as well as in patterns of behavior and attitudes. An institutionalized hegemonic party regime is one with an entrenched two-tier party system whereby the  31  Panebianco defines party institutionalization as a process where a party ceases to be a means to certain ends; the preservation and survival of the party becomes a goal and is “valuable in and of itself” (1988, 53).  39  opposition parties are systematically disadvantaged and unable to compete with the incumbent on antagonistic or equal terms.32 Scholars use the terms institution and institutionalization differently.33 While Huntington defines institutionalization as the “process by which organizations and procedures acquire value and stability” and measures institutionalization by its “adaptability, complexity, autonomy and coherence of its organizations and procedures"(1965, 394); Mainwaring and Scully identify stability; stable roots in society; legitimacy and party organization as 4 criteria for the institutionalization of democratic party systems (1995). To complicate matters, the same dimension is sometimes analyzed under different labels. For example, Huntington’s “complexity” is similar to Mainwaring and Scully’s “party organization”. Apart from definitional ambiguity, the concept has been criticized for tautology and failing to separate the explanation from its outcome (Levitsky 1998). Despite the debates, there is a general consensus that the concept retains its utility, as long as its definition, unit of analysis and dimensions are laid out. The challenge is to “unpack” the concept and demonstrates that the outcome to be explained is not treated as an aspect of institutionalization (Levitsky 1998). Instead of treating institutions as “sticky”34 and electoral institutions as “very stable and resistant to change” (Pierson 2004; Lijphart 1994), I view institutions as stable because of the power of the current supporters and not because of the binding constraints of the original agreement (Moe 2005). An institutionalized party must have an incentive distribution system that builds organizational interests, loyalty and cohesion, without which it will be prone to power struggles and internal splits. To build cohesion, the party must distribute selective incentives (prestigious positions, career opportunities) to its ambitious members and collective incentives (sense of belonging) to its activists and supporters (Panebianco 1988). Without the consolidation of this incentive structure party institutionalization cannot take place and  32  As Panebianco said: “All parties must institutionalize to a certain extent in order to survive” (1988, 54). Some view institutions as “rules of the game” while others see them as formal organizations; patterned behaviour or as “myths” and ideational structures. See Hall and Taylor (1996). 34 Pierson has argued that institutions “are typically not plastic. They do not adapt swiftly and effortlessly. They are subject to change, but the multiple sources of resilience suggest that in many circumstances they will exhibit very substantial inertia. It is this inertial quality that makes them important contributors to an understanding of long-term process of institutional development” (2004, 157). 33  40  organization survival is at stake. In this study, party cohesion refers to “the extent to which, in a given situation, group members can be observed to work together for the group’s goals in one and the same way” (Ozbudun 1970, 305). It embraces the degree to which elites cooperate (unity), and the compliance of members with party goals or leader’s preferences (loyalty). Cohesion is different from party discipline as the latter refers to a special cohesion achieved through enforcement or sanctions by which enforced cohesion is attained. According to Mainwaring and Scully, an institutionalized party system must have: 1) regular and stable inter-party competition, 2) rootedness 3) party organizations that matter and 4) electoral legitimacy (1995). It is through time that a party as an organization becomes institutionalized (Randall and Svasand 2002, 14). The longer the party is in existence, the more likely it will be rooted and institutionalized. The age of the party may be measured: 1) chronologically, 2) generationally - replacement of party founding leaders and 3) functionally – its participation in national elections. A party is likely to be institutionalized if it has contested at the first national competitive elections and fought in more than three national elections (Rose and Mackie 1988; Huntington 1968). The party origins, age and rootedness of the PAP and the KMT are important sources of its organizational resilience and will be examined in Chapters 3 and 4. I will compare the degree of party system institutionalization in Singapore and Taiwan based on the 4 dimensions in Chapters 7 and 8. Institutionalization does not progress in a neat, linear fashion (Randall and Svasand 2002, 15). While institutionalization increases a regime’s stability and resilience, there is no guarantee against regression, manipulation or de-institutionalization.35 As Mainwaring and Scully remind us “evaluations of party system institutionalization are not “static”, “unilinear nor irreversible” (1995, 21). While institutions have the capacity to shape behaviour, they are also partially endogenous36 and subject to change as a result of strategic calculations by political actors (Magaloni 2006, 12). When self-interested elites set out to redesign electoral institutions or constitutional framework, we need to recognize the embedded cultural, sociopolitical constraints that led to the final choices (Grofman et al. 1999, xi). A key contribution  35  De-institutionalization refers to the loss of autonomy and systemness in response to sudden, exogenous change in the external environment such as economic shock or electoral loss (Panebianco 1982). Mainwaring and Scully argue that for hegemonic party regimes (such as Mexico and Paraguay) to democratize; deinstitutionalization has to first occur (1995, 20). 36 For more on the endogeniety problem, see Greif and Laitin (2004).  41  of my dissertation is to highlight the sources of institutionalization (Chapters 3 and 4) and the conditions under which these institutions are manipulated or de-institutionalized bringing about un-intended consequences (Chapter 7 and 8) (Pierson 2004). Leadership Selection and Party Cohesion The rules governing leadership selection have a direct impact on party cohesion (Katz and Mair 2002; Lovenduski and Norris 1993; Panebianco 1988; Przeworski371986). Leadership selection is important as it highlights the interests of power brokers. As Schattschneider reminds us, “The nature of nominating process determines the nature of the party; he who can make the nominations is the owner of the party. This is therefore one of the best points at which to observe the distribution of power within the party” (1942, 64). Parties have different rules and selection criteria on who can or cannot join the party, gain access to top party leadership positions. Leadership selection consists of elite recruitment and candidate selection.38 Elite recruitment refers to the process by which individuals are inducted into active, high profile political roles within the party or at the national level; candidate selection is part of elite recruitment where parties nominate and select their candidates before the general elections (Hazan and Rahat 2006, 109). On the party system level, the rules and procedures guiding candidacy directly affect the composition and representativeness of the national legislature. I posit that a hegemonic party with an institutionalized leadership selection system is less vulnerable to power struggles and defections. In highly institutionalized parties, elite recruitment has a centripetal movement or a strong “centre” that monopolizes incentive distribution. To make one’s career in the party is to allow oneself to be co-opted by the centre. Centralized incentive structures governed by a clear set of rules are more predictable and stable. In weakly institutionalized parties, elite recruitment has a centrifugal movement. To succeed, the candidate has to define himself against others, which fuels intraparty competition and factionalism while undermining cohesion.  37  As Przeworksi also notes, when the problem of succession appears and mechanisms of succession are not institutionalized, a conflict is imminent (1986, 55). 38 Candidate selection is “the process by which a political party decides which of the person legally eligible to hold an elective office will be designated on the ballot and in election communications as its recommended and supported candidate or list of candidates” (Ranney 1981, 75).  42  I propose to measure the degree of hegemonic party institutionalization based on its leadership succession system. If the party selectorate39 enjoys high decisional autonomy and the process displays high systemness40 then, its leadership selection system is institutionalized. Autonomy is high when the selectorate can set its own requirements for candidacy, exclude external intruders who do not meet its admission requirements and nominate or appoint its candidates without external interference from outside organizations such as the unions or the church (Lovenduski and Norris 1993, 321). Exclusiveness means that the party selectorate has the rights and jurisdiction over the recruitment eligibility and criteria. Parties with candidates selected by the party leader or a small group of elites through methods such as closed nominations and party conventions, are more likely to produce loyalists and foster cohesion (Siavelis and Morgenstern 2008, 15). In institutionalized hegemonic parties, leadership selection has a centripetal movement – towards whoever has the monopoly to nominate, select candidates and distribute incentives. To climb the party ladder, careerists will need to show loyalty, unite and rally around party goals. Conversely, if the selectorate is too inclusive41 with permeable borders, or has little control over the quality of candidates, then its autonomy will be low. For example, if the candidates are selected through an election that involves a large selectorate, the candidates are likely to respond to various interests and less inclined to comply with the party directives (Rahat 2007, 159). For example, in open primaries, party leaders have little control over the candidates and candidates have little incentives to be loyal or unite. See Figure 2.2 for a summary of the inclusiveness of the candidate selection method.  39  Party selectorate refers to the body that selects the candidates. It composes of one person or many people (including the whole nation) (Harzan and Rahat, 2006, 110). 40 Systemness refers to the “increasingly scope, density, and regularity of the interactions that constitute the party as a structure” (Randall and Svasand 2002, 13). 41 Canadian parties are an exception as their candidate selection methods are highly decentralized and inclusive (less than the US parties), but Canadian MPs show high levels of discipline (Carty 2004, 5-24).  43  Figure 2-2: Candidacy, Selectorate and Selection Method Inclusiveness 1 Party Selectorate  General Electorate  Party membership  2 Candidacy  All citizens  Party members  3 Electoral Method  Voting  Selected party agency  Exclusiveness Nonselected party agency  Single leader Party members & additional requirements  Appointment  Source: Harzan and Rahat (2006)  Systemness is another dimension of party leadership succession. Systemness refers to the routinization of charisma or regularization of patterns of social interaction, or the entrenchment of the formal and informal rules of the game (Levitsky 1998, 88; Panebianco 1988, 53). If there are formal party charters or informal rules (patronage network), established guidelines (education qualification) governing candidate selection for top party and national positions and these rules are accepted without contest, then, leadership selection displays high systemness. Conversely, if these rules are circumvented or manipulated to suit the short term needs of one person or a group; or is challenged by a majority of members, then, systemness is low. A party may be unevenly institutionalized, displaying high autonomy and low systemness, or vice-versa. Uneven autonomy and systemness signifies uneven institutionalization of leadership selection that could lead to leadership struggles. See Figure 2-3 for the combinations in a four-cell matrix.  44  Figure 2-3: Model of Leadership Succession and Institutionalization Autonomy (A) Institutionalized Charisma  Institutionalized Conflict  High S & High A  High S & Low A  Regular successions based on clear rules of the game, and facilitated by a small, exclusive selectorate  Regular successions based on clear rules of the game and facilitated by an inclusive electorate  Singapore’s PAP (1990s- present)  Un-institutionalized Charisma  Taiwan’s KMT (1989-1990s) American parties British Labour party Most Nordic parties Un-institutionalized Conflict  Low S & High A  Low S & Low A  Irregular successions based on a small incumbent selectorate  Irregular successions based on patronage or informal clientelistic network with no clear rules; ad-hoc selectorate  Systemness (S)  Taiwan’s KMT (1945-1988) Mexico’s PRI (1988-1994) Indonesia’s Golkar (1966-1998) Thailand’s Thai Rak Thai (2001-8) Centripetal  Thailand’s Chart Thai (1974-2008) Philippines’ Lakas-NUCD (1998-2004) Cambodia’s Funcinpec Party (1980s) Centrifugal  The upper left quadrant refers to parties with leadership selection that display high systemness and high autonomy. In these parties, successions are institutionalized and routinized and a small selectorate retains the exclusive right to decide the admission criteria and quality of candidates into the top party or national political positions, without external interferences. A clear set of rules regulates the expectations and the party’s opportunity structure. Careerists who want to be elected will rally around the selectorate or party goals, demonstrate loyalty, cohesion and discipline. Intra-party competition takes a centripetal pattern and manifests itself in the form of “procedural battles”.42 Charisma is institutionalized as it is objectified and access to power is not tied to the idiosyncrasies of any individual. Rather, the party organization becomes a bureaucracy for facilitating leadership renewal. Parties with “institutionalized charisma” model are typically unified and cohesive.  42  Rules become instruments of control and represent a source of guarantee for other careerists who can appeal to the rule to defend themselves from the whims and fancy of the ruler (Panebianco 1988).  45  Leadership successions are less prone to power struggles as the outcome is predictable. Singapore’s PAP is a prime example. On the upper right quadrant, parties display high systemness as leadership selection is routinized and governed by clear selection rules and admission criteria. However, the parties suffer from low autonomy as the selectorate can include all the party members or the voting electorate. Inclusive methods are likely to encourage politics of personality and cooperative behavior. For example, in the U.S. open primaries, voters may select congressional nominees without deferring to the preferences of the party organizations. This increases the value of cultivating constituent loyalties and fosters personality-based politics (Siavelis and Morgenstern 2008). Parties become platforms for the candidates, who are responsive to their electorates rather than to their parties. Candidate centered politics are prone to dissension and infighting. For example, the U.S. Democratic or Republican parties that facilitate open primary elections typically lack cohesion. Likewise, the British Labour Party also falls within this quadrant as it has weak autonomy and the organizational stability depends on its links to the trade unions. The unions can assert influence over the party selectorate as they have veto powers over unwanted parliamentary candidates through proposing their own nomination lists via the National Executive Committee (where the unions are a majority). Consequently, the Labour party is factionalized by various power centres and competing interests (Panebianco 1988, 93-4). Leadership selection based on “institutionalized conflict” leads to low party cohesion and is open to contention with uncertain outcomes.43 On the bottom left quadrant, parties are characterized by low systemness and high autonomy, as successions are dependent on a single-ruler or a small, exclusive group of selectorate. While most parties have charismatic leaders, this model is concerned with parties with leaders who impose themselves as “the undisputed founder, conceiver and interpreter of a set of political symbols” (Panebiano 1988, 52). This characterization typifies most charismatic or leaderist parties where succession is based on personal ties to the incumbent leader, with an absence of rules or internal career patterns. Unlike the “institutionalized charisma” model, this opportunity structure relies on informal networks. To climb the party  43  This label “institutionalized contention” draws from Przeworski’s view that democracy is a process of institutionalizing uncertainty - where all groups subject their interests to uncertainty. In democracy, no group is able to intervene when outcomes of conflicts violate their self-perceived interests (1986, 58).  46  ranks, party careerists will need to mobilize around the leader and compete continuously for his attention or favour. In this model, centripetal competition between “tendencies” 44 may get intense; it will not involve the leader and takes place at the level below him. Careerists compete amongst themselves to get into the inner power hierarchy. Publicly, they display high levels of unity and loyalty as no one expects to openly challenge the leader and win. “Uninstitutionalized charisma” is inherently unstable. In the event of the sudden death of the leader, the “tendencies” could turn into factionalism and result in intense leadership infightings. As Huntington says, “The institutional strength of a party is measured in the first instance by its ability to survive its founder or the charismatic leader who first brings it to power” (1968, 409). Un-institutionalized charismatic parties are vulnerable to the death of the leader and violent overthrow. Leadership successions could lead ugly contentious politics with uncertain outcomes. Examples of this model include the KMT during Chiang Chingkuo’s rule in Taiwan and the Golkar Party under Suharto’s era in Indonesia. Finally, on the bottom right quadrant, parties lack both autonomy and systemness in its leadership succession system. In this model, successions are irregular and the opportunity structure is based on informal networks, patronage or clientelism. Rules governing candidacy are arbitrary or inconsistently applied based on network ties. Both the party selectorate and admission criteria vary depending on the relations with the party elite, resources or candidate popularity. Examples of this model include Thailand’s Chart Thai or Thai Nation Party (“Generals’ Party”). Thai Nation Party was described as a “family” rather than an institution. It has no real candidate election by the party members as it believed that conflict would ensue, if, candidates were determined by the branches (Thornton 2003). Other examples include the Funcinpec Party, led by Prince Sihanouk and later Prince Ranariddh. The Party lacked autonomy as it had depended on foreign and American aid, funnelled through ASEAN. Without strong leadership and clear rules governing successions, these parties are unstable, un-cohesive and organizationally weak. In these parties, irregular successions are prone to internal conflicts with uncertain outcomes. In the next section, I will explain how varied  44  Tendencies refer to loosely organized groups or aggregations at the top without organized rank and file. Factions on the other hand are organized groups that may be cut either vertically (from the top to rank and file) or geographically, organized at periphery (sub-coalitions) (Panebianco 1988).  47  electoral systems assert mechanical and psychological effects to shape the degree of competitiveness in hegemonic party systems. Electoral Engineering and Institutionalizing Certainty Earlier, party institutionalization was assessed based on party origins, autonomy and systemness of leadership succession. On the party system level, the unit of analysis is on the electoral system, the key incentive structure that regulates inter-party competition.45 As Sartori reminds us, the electoral system is “the most specific manipulative instrument of politics” (1968, 273). The electoral system serves as “redistributive” institution while electoral rules are distributive institutions to improve the share of one group at the direct expense of another (Tsebelis 1990; Benoit 2004). In hegemonic party regime, the incumbent will seek to increase electoral certainty through electoral engineering as its “last defence of authoritarianism” (Schedler 2010, 69). Electoral engineering is marked by a flexible electoral system that is strategically altered and tweaked in small parts, but not overhauled by the incumbent party. The aim is to institutionalize the semi-competitive party system such that the opposition parties are systematically disadvantaged. As North suggests, “institutions are not necessarily or even usually created to be socially efficient; rather, they are created to serve the interests of those with the bargaining power to create new rules” (1990, 360-1). Hence, electoral rules tend to be written by strong parties or “winners”.46Changes in the electoral rules tend to reflect the self-interest of dominant parties in the face of rising electoral uncertainty (Remmer 2008, 9). As long as the electoral system serves the incumbent well, there is no incentive to change. If the partisan competition heats up, the stronger party will modify its electoral system to maintain its  45  I view parties as self-interested actors with the primary aim of maximizing their overall vote and seat shares in the legislature (Remmer 2008). Voters are strategic actors who will avoid voting for candidates who will do badly in the election, even if it means supporting the second-ranked candidate in their preference orderings. Parties will avoid wasting votes or resources on hopeless candidates (Boix 1999). 46 Electoral rules are formal institutions that encourage the strategic behavior of both elites and voters and hence force their coordination around a set of viable candidates (Boix 1999, 609).  48  advantage.47 Electoral engineering is likely to occur when the hegemonic party supports an arrangement that brings it more seats than the status quo electoral system (Benoit 2004, 374). Electoral institutions can shape party systems by asserting mechanical and psychological pressures on parties and voters (Duverger 1951).48 Mechanical effects concern how electoral rules constrain the seats from the distribution of votes; psychological effect deals with the shaping of party and voter strategies in anticipation of the electoral function’s mechanical constraints (Benoit 2004, 364). Broadly, the electoral system may be divided into 3 parts: 1) the ballot structure that may or may not allow voters to split their votes between parties; 2) district magnitude (average number of legislators elected per district) that constrains factor in the translation of votes to seats, and 3) electoral formulae (plurality or PR systems) for translating votes to seats (Rae 1967; Lijphart 1990). Electoral engineering occurs when the incumbent party seeks to manufacture its electoral dominance by altering specific aspects of the constitution or electoral system to institutionalize an un-level playing field. For example, Mexico held 5 congressional elections between 1985 and 1997, each under different electoral formulae to prop up the PRI. In Singapore, its single-member plurality (SMP) system was changed to include an additional block voting in Group Representative Constituencies (GRC) in 1988, which had disastrous effects on the opposition. Taiwan, on the other hand, has had 7 constitutional revisions since 1991, making its electoral system one of the most complicated in the world. Electoral systems are subject to partisan manipulation because they have non-neutral effects: larger parties tend to do better with SMP electoral formulas, while smaller parties do better under the PR system. Electoral engineering shapes the incentives facing parties and actors and alters inter-party competitiveness. Parties hold or derive preferences for alternative institutions based on expectations of the payoffs of the electoral institutions. Hence, parties fight over the rules and practices that govern the electoral process in a complex “nested  47  Boix argues that the impetus to change from plurality to PR system is contingent on 2 conditions: 1) strength of the new challenging parties and 2) coordinating capacity of the ruling parties (1999, 609). There will be no change to PR if new parties are weak. 48 Mechanical effects show the electoral system’s systematic underrepresentation of “third parties”, while psychological effects shows the tendency of voters to rally what they consider the least unacceptable of the two major parties as they realize the votes for minor parties are not translated to votes. See Blais and Carty (1991) for attempts to measure the psychological effects of electoral laws.  49  games” (Tsebelis 1990). By altering the electoral system, the incumbent parties seek to maximize their electoral “bang for the buck” to win more seats with the same number of votes (McElwain 2008, 32). In hegemonic party autocracies, survival hinges on the incumbent’s ability to win supermajority so to control institutional change to their advantage and project an image of invincibility to deter opposition (Magaloni 2006). With legislative supermajority, the constitutional order and judicial independence tend to be weak in hegemonic party regimes. The opposition movement is systematically disadvantaged with little protection from the rule of law. The hegemonic party with the monopoly over rulemaking will thus shape the rules of the game to its advantage. Given access to state resources, infrastructure, personnel and information, the hegemonic party can calibrate its electoral strategies to cripple but not eliminate the opposition, so that it does not become a credible replacement. The position of the hegemonic party is maintained as the challengers are better off playing by the rules set by the incumbent rather than defect and be completely excluded. The opposition thus face a dilemma in an electoral autocracy: to be an inferior party and play by the rules or boycott the election altogether (Schedler 2009, 200-1). To boycott the election comes with a high cost. The withdrawal of opposition or failure to co-ordinate will often strengthen the status quo - a desired outcome for the incumbent. The view that electoral institutions may be manipulated is important as it: 1) highlights the structure of unequal power relations between the incumbent and opposition parties that constrained their ability to compete (Moe 2005); 2) allows us to move beyond static, structural explanations that focus on historical “founding moments” (Lipset and Rokkan 1967); 3) avoids tautological propositions that link ex ante shifts in electoral outcomes to reforms (Remmer 2008), and 4) shows the adaptability of hegemonic parties in response to rising opposition threats (Benoit 2004). Repression is costly and electoral fraud requires substantial resources that may result in unintended outcome (Lehoucq 2003). Electoral engineering is less costly and safer that allows the incumbents to retain power without undermining their international reputation (Levitsky and Way 2010). The hegemonic party does not have full control over the electoral outcome as electoral strategies is only one of the institutional constraints that shape the behaviour of party leaders, candidates and their followers. Regular elections offer feedbacks to the 50  incumbents on the effectiveness of its strategies. Despite the extensive literature and studies on the expected effects of electoral rules on party system, “electoral design remains more an art than science” (Norris 2004, 23). Indeed, elections and electoral engineering may bring about unintended consequences49 and end up enlarging the spaces of interparty competition by “default” (Way 2006; Case 2005; Hicken 2008).50 In Chapter 7 and 8, I will compare the effects of electoral engineering in Taiwan and Singapore. Measures such as party system fragmentation, party polarization and electoral volatility will be used as indicators of mechanical and psychological effects on party system.  49  Schedler argues that electoral engineering may be ineffective and fail to persuade voters as 1) the incumbent party may lack personnel or resources to put their strategies into effective practice, and 2) opposition parties may employ more sophisticated means to demand for more accountability (2009). 50 Case argues that the persistence of electoral autocracies is a function of artful manipulation as opposed to “unskilful” manipulation that result in regime replacement (2005).  51  PART I: ORIGINS, ORGANIZATION AND COORDINATION  52  3  The PAP: Organization Transformation and Strategic Coordination Origins of strong parties are often to be found in the struggles that brought them to  power (Lapalombara and Weiner 1966). This is especially true for one-party regimes, which are “the product of the efforts of a political elite to organize and to legitimate rule by one social force over another in another bifurcated group” (Huntington, 1970, 11). Early political experiences and societal cleavages have important implications for a party’s organizational model and long-term survival (Lipset and Rokkan 1967; Smith 2005). In Asia, the period between World War II and the onset of authoritarian regime was a critical juncture for the formation of states, regimes and parties (Slater 2005). To understand how socio-economic cleavages and historical legacies affect the hegemonic party systems in Singapore and Taiwan, Chapter 3 and 4 will highlight the conditions under which the PAP and the KMT arose. This is not a historical study of the two parties as it has been done elsewhere in more depth.1 Rather, my aim is to build on established insights on party system formation and trace how historical experiences and early development of the two parties laid critical foundations for its institutionalization and resilience (Rose and Mackie 1988). Within the study on Western European party systems, Lipset and Rokkan’s “freezing hypothesis” is most widely cited to explain how socio-economic cleavages, industrialization and national suffrage affect the formation and stability of party system (1967).2 Socio-ethnic cleavages such as class, culture, religion and ethnicity are usually viewed as contributing factors that lead to sharply differentiated groups within society and also determinants of the social bases of political conflict, institutionalizing a pattern of politics that was more or less immutable (Lijphart 1999; Rose and Urwin 1969).  1  For the PAP’s early history, see Pang (1971); Chan (1976); Fong (1979); and Drysdale (1984). For the history of the KMT after 1949, see Dickson (1997); Taylor (2000) and Tien (1989). 2 As Lipset and Rokkan say, “The decisive sequence of party formation took place at the early stage of competitive politics, in some cases well before the extension of the franchise, in other cases on the eve of the very rush to mobilization of the finally enfranchised masses” (1967, 34). Cleavages are issues, policy differences or identifications related to conflicts in a society; associated with demographic attributes such as race, religion and occupation. See Bartolini and Mair (1990); Mair (1990); and Kitschelt (1995).  53  In contrary to Lipset and Rokkan’s thesis, I argue that socio-ethnic cleavage structures are not deterministic or immutable.3 Based on my study of the PAP and KMT, I argue that socio-ethnic cleavages have varying impacts on party system formation. In Singapore, the PAP leaders were able to put in place a series of measures to dismantle and depoliticise socio-ethnic cleavages. The PAP’s strategic coordination prevented the rise of partisan alignment based on ethnic or religious consciousness. But in Taiwan, the KMT exacerbated the sub-ethnic cleavage between the majority local Taiwanese and minority Chinese from Mainland China by adopting discriminatory policy to prevent Taiwanese to assume key party and political positions since 1945. In Chapter 4, I show that while the KMT attempted to address this sub-ethnic cleavage through Taiwanization policy - recruiting more local Taiwanese to key party and official positions in the 1970s and 80s, the policy reversal came too little, too late and had dire consequences on party cohesion. Emboldened by Taiwanization, the “Dangwai” or opposition movement formed the Democratic Progressive Party (DPP) in 1986 against the legal ban. Capitalizing on the sub-ethnic cleavage, the DPP was able to mobilize the disenfranchised Taiwanese based on ethnic justice. Unlike the KMT model that allowed the sub-ethnic cleavage to translate into partisan alignment, the PAP repressed socio-ethnic tensions through strategic coordination – the provision of public goods and selective withdrawal of civil-political liberties. After the PAP government came to power, it systematically de-politicized ethnic cleavages, fostered a national identity and built a “catch-all” party platform to appeal to a wide-sector of Singaporeans. The PAP was intrusive and deliberate in its public policies and use of state resources to build a network of para-political institutions to maintain ideological and racial harmony. As Chua argued, the “depoliticization” effect was an ideological achievement of the PAP (1997, 127). In Chapter 5, I contend that the PAP’s ability to set the political agenda and dictate the parameters for public discourse gave it a distinct advantage over its challengers. Pre-democratic parties formed during Lipset and Rokkan’s “early period” are more likely to persist (1967). In Singapore and Taiwan, the early establishment of the PAP and  3  This is based on the view that ethnic identities are pragmatic or opportunistic constructions by political entrepreneurs and need to be understood within its social, cultural contexts (Laitin 1998).  54  KMT gave them a distinctive advantage. In both cases, the parties have had a critical role in fighting for the country’s independence and international recognition. In the 1950s, both parties experienced working with, and later fought against, the communist parties. These early experiences left an indelible mark as seen in the KMT’s mass-based, quasi-Leniniststyle organization and democratic centralism as its decision-making model. Likewise for the PAP, its left-wing experiences led to its transformation from a mass party to a cadre one - a critical juncture in its party development that removed intra-party democracy from its leadership selection. As Chapters 3 and 4 posit, the PAP and KMT’s quasi-Leninist party organization based on cadre system and penetration into the grassroots levels through parapolitical organizations were common traits that foster party discipline and elite cohesion – institutional sources that explain their resilience in the 1960s and 70s. The arguments in this chapter are organized as follows. The first section considers the origins of the PAP and demonstrates how Singapore early socio-economic cleavages affect the PAP’s organization. It focuses on the PAP’s early party formation experiences and impact of the party splits on its cadre-party organizational model on cohesion. Based on the idea of strategic coordination, I argue that the provision of public goods such as housing and social services via the para-political institutions helped to build ideological consensus and performance legitimacy of the PAP. By constraining civil liberties through a myriad of legal rules, the PAP was able to turn the negative ground swell around in the 1960s and lay the institutional foundation for its hegemonic rule. Origins of the PAP Singapore was returned to its British colonial master after 3 years and 8 months of Japanese occupation (Feb 1942- Sep 1945). After the end of Second World War, the British resumed control of Singapore, first as a British Military Administration (Sep 1945-Apr 1946) to restore law and order and later, as a separate crown colony. However, the inability of the British to defend Singapore had severely damaged its credibility as the ruler of Singapore. Post-war Singapore saw an increasing political awakening amongst its local population and rising anti-colonialist and nationalist sentiments (Fong 1979).  55  Meanwhile in London, a group of English-speaking, middle-class men4 from Singapore and Malaya came together to form the Malayan Forum to discuss the ideas of forming an independent Malaya. Upon their return from Britain after completing their studies the young men, imbued with Fabian Socialist idealism, decided to form an anti-colonial leftwing party to fight for equal treatment, social justice and local representation. Lee Kuan Yew, the PAP’s founding leader, was a trained lawyer with double-first class honour from Cambridge University5. As automatic registration of voters under the Rendel constitution6 enabled more Chinese-speaking voters into the Singapore rolls, Lee knew his middle-class background and limited Chinese language knowledge would hinder his reach to the Chinese supporters (Lee 1998). To capture the Chinese and dialect-speaking voters for the Legislative Elections in 19557, he decided to “ride the tiger” and invite leftwing Chinese-speaking, communist leaders from the trade unions to join the PAP. In a ceremony attended by 1,500 members, supporters and observers, the PAP was inaugurated on 21 November 1954 (Lee 1998). This ceremony marked the beginning of what Duverger termed an “externally-created party”8 that rejected colonialism, imperialism, the Rendel Constitution, the Emergency Regulations and Malayanization of the civil service. The PAP had sought recognition of Chinese, Tamil, Malay and English as official languages and equal citizenship regardless of race, religion or language to all who were prepared to pledge loyalty to Malaya9 (Pang 1971). Ideologically, the PAP projected itself as a radical, left-winged party with the support of the trade unions and student movement. While the founding leaders were from a middleclass background, the PAP distinguished itself from the other more established, right-winged  4  The PAP’s founding leaders were Lee Kuan Yew, Toh Chin Chye, Goh Keng Swee; members of the Malayan Forum - a group formed in London. Goh later recruited two more “lieutenants”, Joe Pillay and Chua Sian Chin (Lee 1998, 231). These pioneers held key government positions in Singapore’s early government. 5 A legal advisor to trade unions, Lee was self-described as a “golf-playing, beer-swilling bourgeois”, brought up in a family who thought “everything English was the acme of perfection” (Lee 1998, 138). 6 The Rendel Constitution was introduced in 1953 to allow Singapore more self-governance. 7 There were two previous elections to the Legislative Council (1948 with 6 elected seats and 1951 with 9 elections) with small electorates (Fong 1979, 27). 8 “Externally created parties” refer to those that emerge outside the legislature and involve some challenge to the ruling group and a demand for representation. They are expected to be more centralized than “internally created” parties and more ideologically coherent and disciplined, less willing to ascribe importance to be deferential toward the Parliament (Duverger 1954, xxiii). 9 The PAP considered Singapore to be part of Malaya from its inception (Pang 1971, 3).  56  parties such as the Progressive Party (PP) - Singapore’s first political party led by English educated professionals and Englishmen - and the wealthy Democratic Party (DP) which consisted of Mandarin-speaking bourgeois from the Chinese Chamber of Commerce. To legitimize itself as a multi-racial, multi-lingual mass party, the first PAP Central Executive Committee (CEC) - the highest executive body in the Party - had 11 members of different ethnicities and backgrounds who were a “consciously radical and anti-colonial coalition of moderate socialists and left-wing communist forces” (Chan 1989, 71). Then, the CEC composed of 7 Chinese, 2 Malays and 2 Indians; young men in their 20s and 30s of varied education levels (3 with university degrees, 5 secondary and 3 primary school education) (Pang 1971, 38). From the onset, it was clear to outsiders that the Party had two factions: the moderates and the communists. 3 out of the 11 CEC members were known communists or communist sympathizers.10 While the moderates led by Lee needed the communists for mass support; the communists led by Lim Chin Siong needed the PAP to pursue the cause of their illegal Malayan Communist Party (MCP).11 Lim was a prominent leader of the militant Middle Road Group of Unions12 and Chinese Middle School Students Union – training grounds and channels of recruitment for the underground Malayan Communist Party (Clutterbuck 1985). Despite the ideological and language differences between the factions, the Party was united in the common goal to represent the working-class and end colonial rule. Singapore’s early electoral politics was vibrant, competitive and rambunctious. As a result of the Rendel Constitution, the island was divided into 25 constituencies and automatic registration of voters in 1955 increased the limited electorate from 48,000 in 1951 to 300,000 voters. Then, 60% of the voters were Chinese-speaking.13 The 1955 Legislative Assembly  10  Lee Kuan Yew, Toh Chin Chye, Goh Keng Swee were considered the moderates, while Lim Chin Siong, Fong Swee Suan and Devan Nair were said to be leaders of the communist faction (Pang 1971, 3). 11 Lim has never publicly admitted to being a communist. See Clutterbuck’s interview with Douglas Hyde, Lim’s prison cellmate from 1964-8 (1985, 99). 12 The “Middle-Road” Group was a group of unions that included Singapore Factory and Shop Worker’s Union (SFSWU) under Lim Chin Siong and Singapore Bus Workers’ Union (SBWU) under Fong Swee Suan. 13 See Lee (1998, 184) and Yeo (1973, 252-4) for Singapore’s early electoral framework.  57  election was the PAP’s first election and marked the Party’s foray into Singapore politics as an opposition force.14 Fight for Independence and Cleavages, 1950s From 1955 to 1959, Singapore was a self-governing entity and in a state of flux. The Labour Front government, under its first Chief Minister David Marshall and then later under Lim Yew Hock, pushed for independence. However, their efforts were met with resistance as Singapore was deemed a vital British commercial interest and strategic naval base in the region. Besides, the outbreak of the Cold War in 1947 and the defeat of the Chiang KaiShek’s KMT government in China in 1949 also raised fears that Chinese chauvinism and wave of nationalist sentiments could turn Singapore into a hotbed for communism in Southeast Asia. Pre-independence Singapore had all the elements of value dissension (Chiew 1990, 46). The small island’s struggle to independence was plagued by anti-colonial agitation, communist subversion and racial unrests. The Rendel Constitution that extended the franchise of Singaporeans changed the electoral techniques, composition and organization of parties such as the PAP (Yeo 1973, 253). Elite-based politics dominated by English-speaking locals and British elites were quickly replaced by mass politics that altered the pattern of electoral contests. In the 1950s, Singapore was filled with immigrants from impoverished rural South China, South India and Malaysia as a result of an unrestricted immigration policy.15 Problems of overpopulation, poverty, unemployment and inadequate housing fuelled discontent amongst the population of 1.4 million. Low literacy rate, poverty and lack of job opportunities pushed the immigrants to lowly paid menial work such as hawkers, rickshaw pullers and domestic servants (Huff 1994, 277).16 In the 1950s and 60s, the majority of Chinese population (75.4%) was comparatively, the least literate (46.2%). See Table 3-1.  14  In the 1955’s hotly contested Legislative Elections, a total of 79 candidates competed: 69 nominated by parties, while 10 stood as independents. The PAP fielded 4 candidates and won 3 out of the 25 seats. 15 After 1965, Singapore's government imposed strict controls on immigration, granting temporary residence permits only to those whose labor or skills were considered essential to the economy. 16 In 1957, unemployment stood at 4.9%. 38.5% of the total employed population were in the production, transport and manual work occupation. Only 5.2% were in the professional sector (Huff 1994, 91 and 291).  58  Table 3-1: Singapore’s Population by Ethnicity, 1957 Chinese Indians 1957 75.4 % 9.0 % 1957-70 (Growth rates) 2.9% 3.6% 1957 (Literacy rates) 46.2% 75.2%  Malays 13.6 % 0.9% 62.2%  Others 2.0 % 94.0%  Source: Huff (1994) and Kuo (1990).  Rising Chinese nationalism polarized the multi-racial Singapore society. Dissatisfied with the income disparity between the minority English speaking colonial subordinates and the rest of the population, many working-class Chinese joined unions, parties and student movements to agitate for change. The earliest record shows that in the mid-1960s, Singapore’s personal income distribution or Gini coefficient17 was 0.50, improving to 0.44 only in 1975 (Huff 1994, 351). however, the orientation and loyalty of the Chinese students and unionists were to China and not Singapore (Chiew 1990, 55). The Chinese textbooks used in the Singapore Chinese schools were mostly from China and Taipei.18 As a result, unions and student groups were prime recruit grounds and infiltrated by communists to further their agenda (Clutterbuck 1985, 75-98). The Chinese working-class was dispossessed as they were excluded in the official life of the colony, which employed only Englisheducated Caucasians, Indians and others as subordinates (Lee 1998, 167). As anti-colonial sentiments grew, the gap between the English middle-class and the poor, Chinese-speaking working class widened. Low wages and unfair welfare treatment of labourers also led to a series of labour union strikes, civil unrests, student “sit-ins” and anti-colonial demonstrations, many instigated and led by the unions and the PAP’s left faction. The riots and grievances of the Chinese movement highlighted the problem of nation building in a plural society (Chiew 1985, 55). Some of the strikes and protests that turned bloody include Maria Hertogh19 riot in 1950; National Conscription riot on 13 May 1954; Hock Lee Bus Riot on 12 May 1955 and Chinese middle-school riots in 1956.20 It was estimated that 275 strikes were called and 162  17  There was no income distribution index or statistics available for 1950s. After China turned communist in 1949, Chinese nationalism in Singapore was split into the left and right. The left was led by the Communist Party of Malaya, and right, by the Kuomintang Party in Taiwan. Both competed for the allegiance of the Chinese students in Singapore (Chiew 1990, 55). 19 Maria Hertogh ritos was sparked by a custody battle over a Dutch Catholic girl allegedly informally adopted by a Muslim Malay woman which resulted in 18 deaths and 173 injured (IBA Report 2008, 62). 20 See Clutterbuck (1985) for causes of the riots in the 1950s and Vasil (1989, Chapter 7) on the role and development of trade union movement in Singapore. 18  59  of them in the 5 months from Apr to Sep 1955 were attributed to the “Middle-Road” Group under Lim Chin Siong – the PAP’s communist faction leader (Clutterbuck 1985, 100). The PAP Membership and Organization As Duverger posits, the mass party model usually emerges in a highly polarized society with key objectives of large-scale recruitment and socialization of individual members (1964, 60-71). The PAP was initially organized as a mass party with a broad and inclusive recruitment policy that emphasized quantity rather than quality. As early sources show, the party membership was only a few hundred and grew to 4,860 after 1959 (Pang 1971, 6). See Table 3-2. The PAP membership in the early 1960s consisted mostly of lowly educated, working-class Chinese, between 20 to 29 years old. While there were other ethnicminority representatives in the PAP, the percentages were low, compared to the ethnic minority groups of the total Singapore population in 1961. Table 3-2: PAP Membership Size and Residential Population in Singapore (1958-2006) Year 1958 1959 1963 1965 1971 1997 2009 Total membership 2000 4,860 2,183 4,000 14,830 10,000 15,000 Total no of residents / no. of citizens (‘000) Estimated % of total residents/citizens  1,445 (1957) 0.1%  1,646  1,795  1,886  0.3%  0.1%  0.2%  1,874 (1970) 0.8%  2,623* (1990) 0.4%  PAP Membership and Singapore's Population by ethnicity (%), 1961 Chinese Indians Malays Others PAP’s membership (before 1961) 92.9 2.8 4.1 0.2 Singapore’s total population (1961) 76.0 7.6 14.2 2.0 PAP Membership by Age and Education, 1961 Age Before 1961 Under 20 19.3 20-29 36.5 30-39 20.8 40-49 14.5 50-59 7.3 60+ 1.4  Education level Primary Secondary College or University Vocational and Classics Unknown  3,164 (2008) 0.5%  Unknown 0.1 0 1961 65.4 33 0.8 0.8 0.0  Sources: Calculated based on data from Pang (1971), Petir and Singapore Statistics, various years.  PAP’s recruitment has declined since its heydays in 1960s and not kept with the country’s population growth. In fact, the proportion of PAP members to the total population has dropped from 0.8% to 0.5% as the population grew from 1.9 mil in 1971 to 3.1 mil in 2009. See Table 3-2. Presently, the party membership is 15,000, approximately 0.1% of total 60  number of Singapore citizens21 (3.2 mil, 2009). There are insufficient new recruits to replace the aging party members. In 2006, more than 50% of its party members were above 40 years old. As one PAP member observed, “By 2030, 15 years from now, 25 per cent of the population will be more than 65 years old. The situation will be the same in our party membership and in each branch. Without a focused, sustained influx of young activists in the next five years, we will have a weak party” (Petir, Sep/Oct 2006).22 With an increase in the total population of Singaporeans, the PAP has attempted to recruit more aggressively to keep up with the changes in demography (Petir, Sep/Oct 2006). In 2009, the PAP Youth Wing reported a surge of over 1,000 new, young members - an increase from previous years of 600 to 700 recruits annually (Straits Times, 4 Feb 2010). The PAP was understaffed and underfunded in its early years.23 As Lee Kuan Yew said in his autobiography: “The PAP organization was weak, almost nonexistent: no paid staff, branches or grassroots leaders. For canvassing and help at election rallies, we would call upon the unions and Chinese middle school students”. To expand membership, the party welcomed all members and volunteers that included “young people, mostly workers, trade union officials and students” (Lee 1998, 182). The PAP’s open and inclusive recruitment policy meant that the party was vulnerable to communist infiltration (Fong 1979, 32). In 1961, it was estimated that 677 (29.1%) of total PAP members were trade unionists. After 1961, the total number of unionists began to decline to 570 (23.5%) (Pang 1971, 67). Between 1954 and 1957, the PAP had a simple party structure with direct communication channels between the Central Executive Committee (CEC) and cadre members (Shee 1971, 85). In 1955, the PAP organization consisted only of a CEC and 6 branches (Fong 1979, 33). A headquarter (HQ) Committee was only set up to re-organize and streamline the various branches after 1958. See Appendix C for the organizational structure of the PAP in 1958. In the PAP, the CEC is the main decision-making body and the branches are instrumental in implementing the policy.  21  According to Article III of the PAP’s Party constitution, only citizen of Singapore above 17 years old is eligible to be a member. Refer to the PAP’s Constitution at http://www.pap.org.sg/ourconstitution.php. 22 See “Speech by Dr Vivian Balakrishnan, Chairman,” YP 20th Anniversary YP, 15 Apr 2006. 23 For party finance, see Fong (1971, 288) and Pang (1971, 29).  61  The party organization has grown in complexity since the 1960s. With the CEC holding the key executive power, there is now a highly organized Executive Committee at the Headquarters that is tasked with 12 core functions, namely: Constituency Relations, Information and Feedback, Malay Affairs, Membership Recruitment and Cadre Selection; New Media, PAP Awards, Political Education, Publicity and Publications, Social and Recreation, Women’s Wing (set up in 1989), Young PAP (Youth wing, set up in 1986) and External Relations. Aside from running these 12 functions, the HQ Executive Committee also oversees the organization of the PAP policy Forum and five Districts. Unlike wealthy parties such as the KMT, the PAP operates out of a modest headquarters in Upper Changi, on the outskirts of the city centre. And from just 10 party branches in 1950s, there are now a total of 84 party branches, strategically located in various parts of Singapore. The expansion of party branches indicates a high degree of party rootedness and institutionalization. Intra-party Leadership Struggles and Implications Before 1957, the PAP constitution provided that the CEC was to be elected annually in the Party Conference by all party cadres and members. While the communist faction leaders stood aside24 in the Party’s first CEC election in June 1955, the tension between the moderates and communist factions grew. By the 3rd Party Conference in July 1956, the Communist faction looked set to wrestle control from the moderates, as it won 5 out of 12 CEC seats during the CEC election (Pang 1971, 4).25 As the two factions headed for a showdown for the 4th Party Conference in Aug 1957 to elect the next CEC, the rift between the factions became irreconcilable26. Inclusive party membership and open, electoral policy meant that the communists could stack the assembly hall with their supporters, many whom were registered as party members at the last minute. As a result, 6 left-leaning members were elected into the 12 men CEC with 4,858 votes. The moderate faction led by Lee only received 5,380 votes. As the moderates realized that they have lost majority control of the CEC, they refused to take leadership. As Lee said, it was a  24  There were conflicting reports to why Lim and his comrades decided not to contest (Pang 1971, 4). In response, Lee recruited Ong Pang Boon, an educated bi-lingual man, who could converse in English, Mandarin and dialects to help Lee connect with the Chinese speaking mass-base (Lee 1998, 242). 26 The control of the CEC was critical as the communist faction was against the moderate’s call for independence through merger with the Federation of Malaya and the Internal Security Council (Pang 1971, 4). 25  62  tactic to leave the pro-communist in charge, so that they will be left without a front (1998, 270). But the victory of the pro-communist faction was short-lived. Ten days after the 4th Party election, 5 PAP CEC members were among the 35 people arrested under the Preservation of Public Security Ordinance – which gave the ruling Labour Front government powers to detain without trial.27 The purge was a “welcome respite to the moderates who resumed control of the Party” (Pang 1971, 4). The near capture of the CEC by the communist faction was a turning point in the history of the PAP as the moderates learnt that inclusiveness and intra-party democracy could severely weaken internal cohesion and control. As Lee said, “the folly of adopting a democratic constitution that had left it open to capture through the penetration of its own party branches. We discussed several possible changes to ensure that it could never happen again” (Lee 1998, 271). Indeed, the PAP learned a hard lesson from this incident and has not opened up the election of the CEC to the party members since. As shall be discussed in Chapter 5, the near take-over by the pro-communist faction was a critical juncture that led to the permanent removal of a democratic and inclusive leadership selection method – a significant organizational change that fostered party cohesion. Organizational Changes in 1958 After the failed communist takeover attempt in 1957, Lee instituted a cadre system and introduced a bloc-voting system for the CEC selection and maintained regular reregistering of party membership to prevent outsiders from takeover (Pang 1971, 35). Reminded of the resilience of the Catholic Church after a visit to Rome in 1958, Lee overhauled the PAP’s CEC election system and applied the Pope selection method to the selection of the party’s CEC.28 In addition to this, an Ordinary Party Conference was introduced where cadre members meet every two years to elect the CEC. Significantly, the  27  The 35 people were detained because of their communist subversive activities and instigation of strikes. Lee Kuan Yew claims that the arrests were an attempt by the Lim’s government to prevent the communist in capturing the Singapore Trade Union Council (Lim’s support base) (1998, 271). 28 As Lee recounts in his autobiography: “The Church must have got many things right to have survived for nearly two thousand years... Soon after I returned from Rome, I proposed that PAP elections to the central executive committee be modeled on the system for electing the Pope…Only cadres who had been chosen by the CEC could in turn vote for candidates to the CEC, just as only cardinals nominated by a Pope could elect another Pope. This closed the circuit, and since the CEC controlled the core of the party, the party could not now be captured” (1998, 287).  63  PAP leaders altered the party constitution to remove intra-party democracy. The new party constitution: 1) allowed only full cadres to vote in the election of the CEC; 2) became more stringent in its membership recruitment policy; and 3) introduced four categories of membership to distinguish between the cadre and ordinary members (probationary, ordinary, probationary cadre and cadre) (Chan 1989, 73). 1958 marked the end of 4 years of intra-party democracy where party members came together annually to elect the CEC. The implications of the PAP’s leadership selection will be discussed in Chapter 5. Party Splits in 1960 and 1961 Unlike the present day leadership, the PAP in the 1960s was factionalized. In fact, leadership and factional struggles led to two party splits. While the first split was the result of a personality clash between Ong Eng Guan29 and Lee Kuan Yew, the second was driven by ideological differences between the moderate and pro-communist factions. Perceiving attempts by Lee to curtail his power, and out of a general dissatisfaction with the leadership, then Minister for National Development, Ong Eng Guan tabled “Sixteen Resolutions” 30 in the Legislature and challenged Lee’s leadership. Ong was later suspended as a Minister and expelled for “attempts to disrupt party unity and destroy collective party leadership” (Pang 1971, 8). Ong’s expulsion triggered the PAP’s first split and the formation of the United People’s Party (UPP). Consequently, a by-election was held in Hong Lim constituency and a PAP candidate, Jek Yuen Thong, lost to Ong by a large majority. Ong’s expulsion and the loss of the Hong Lim Branch marked the beginning of the PAP’s unravelling as the party leaders were losing control over the CEC and coordination among the branches (Shee 1971, 87). In 1961, disagreements over the terms towards independence through merger with Malaya and the abolishment of the Internal Security Council between the pro-communists and moderates led to the second historic party split. On July 1961, 13 PAP Assemblymen left the PAP and formed the Barisan Socialis (BS) party, fronted by Dr. Lim Siew Choh.31 This  29  Ong was the Party Treasurer and Mayor of the City Council who stood in the Hong Lim constituency in 1959 elections. Ong’s fiery and anti-imperialist speeches in Hokkien were a crowd puller (Pang 1971). 30 The “Sixteen Resolutions” accused the Party of moving to the right and lack of intra-party democracy. Ong gathered 10,000 supporters in a stadium to support the Resolution on 12 Jul 1960 (Pang 1971). 31 See Pang (1998, 12-4); Vasil (2000, 25-7); Fong (1970, 98-9) and Mutalib (2004, 78-84) for the events leading to the split and formation of the BS.  64  defection resulted in the mass exodus of PAP members, including one of Lee’s parliamentary secretary Chan Sun Wing, who was in charge of the People’s Association (PA). The PA was created as a national grassroots organization to promote racial harmony, social cohesion and nation building32 through community recreation and activities. After the split, the PA and the Work Brigade33 went on strike (Bloodworth 1986, 243). The second split nearly decimated the Party. As Chan observed, the split was a “major watershed in party organizational and ideological development, leading to a creative phase in political leadership. The moderates, who overnight had lost the existing mass base to the new party, were forced to develop new organizational resources” (Chan 1989, 72). Reportedly, 20 out of 25 party branch organizing secretaries and their committees defected and as much as 35 out of 51 party branches went to BS. The damage to the Party extended beyond membership numbers as branch property such as typewriters, sewing machines and furniture were looted and many properties were vandalised and sabotaged (Fong 1979, 105). Party Adaptation After 1963 As Huntington reminds us, “the strength of one-party systems usually depends upon the duration and intensity of the struggle to acquire power or to consolidate power after taking over the government” (1970, 14). For the PAP, the intra-party leadership fights and struggle towards independence through merger with the Malaysia Federation in early 1960s were intense and long drawn. After it assumed government, pressures mounted from within and without. Parties from the left, such as the BS, campaigned against the PAP’s proposal for merger. Union strikes and student protests were staged to undermine the government. As Table 3-3 shows, in 1961 and 62, the number of trade disputes reached its highest levels of 1, 225 and 1064 respectively. Pressures were mounting. Besides the party splits, the PAP also lost its second by-election in Anson constituency to David Marshall.  32  See Fong (1979, 108-9) on the formation of the PA on July 1960 and role of party branches. Lee later appointed Ahmed Ibrahim, a hardliner, to deregister the Trade Unions Congress and put down a mutiny by the pro-communist operators in the Works Brigade (1998, 388). 33  65  Table 3-3: Trade Unions, Disputes and Strikes in Singapore, 1960s-70s Year Trade Unions Union Members Industrial stoppage/strikes 1953-4 13 1955 275 1956 29 1957 27 1960 130 144, 770 45 1961 124 164, 462 116 1962 122 189, 032 88 1963 112 142, 936 47 1964 106 157, 050 39 1965 108 154, 052 30 1966 108 141, 925 14 1967 106 130, 053 10 1968 110 125, 518 4 1969 110 120, 053 1970 102 112, 488 5  Trade disputes 804 1, 225 1, 064 967 792 801 702 790 649 449 486  Source: Data for 1953-55 from Clutterbuck (1985, 100); 1955-7 from Fong (1979, 143) and Singapore Statistics (1983, 40-1).  According to Pang’s estimates, the 1961 party split caused the PAP to lose as much as 80.4% of its members. In 1961, a total of 565 had either resigned or were expelled from the party. Many left for ideological reasons; dissatisfaction with the PAP’s leadership or a loss of faith in a non-communist PAP (Pang 1971, 388-9). The PAP had to face the formidable task of rebuilding the embattled Party that was left with only 26 members in a 51 member Legislative Assembly (Lee 1998, 385-6; Fong 1970, 105). See Appendix D for the significant events during the PAP’s early rule. The 1961 split altered the size and composition of the party membership. Prior to the split, the PAP consisted mainly of young, Chinese working class members. By mid 1966, the PAP became more heterogeneous with more Malays and Indians. There was also a noticeable downgrading of mass party expansion, as the Party moved towards a more balanced social profile that reflected the socio-ethnic demography of Singapore’s population. By the mid 1960s, party recruitments show that the PAP was attracting older and better educated, English speaking members from other minority groups. See Table 3-4.  66  Table 3-4: Ethnic Composition and Education of the PAP Members, the 1960s Ethnic Composition of PAP members (%) Before 1961 Joined after 1961 Mid 1966 Chinese 92.9 63.1 67.9 Malays 4.1 16.0 14 Indians 2.8 19.5 16.9 Others 0.2 1.2 1.1 Unknown 0.1 0.1 0.1 Medium of Education of the PAP members (%) Before 1961 Joined after 1961 Mid 1966 Chinese 69.6 33.9 38.7 English 7.3 23.5 21.1 Malay 2.9 9.4 8.3 Tamil 1.0 6.0 5.3 Nil 13.0 17.0 14.0 Others/Unknown 6.2 16.5 12.5 Changes In The Education Level Of The PAP Members (%) Before 1961 Mid 1966 Mid 1970 Elementary or 98.4 97.1 77.6 Secondary College or University 0.8 2.2 2.4 Vocational and Classics 0.8 0.7 0.4 Unknown 0.0 0.0 19.6 Source: Shee (1971, 166) and Pang (1971, 61-2).  Strategic Coordination: Dismantling Cleavages How did the PAP recover from the bloodbath from the 1960s splits? How did the frail party manage the volatile social-ethnic relations in a polarized society? I argue that through strategic coordination - provision of public goods and selective withdrawal of coordination mechanisms - the PAP was able to rebuild the Party and regain the mass support of Singaporeans. Here, party building refers to the extent to which the Party was able to penetrate from the centre to the periphery (territorially and organizationally) (Randall and Svasand 2002, 17). Ideologically, the PAP moved towards the centre and projected itself as a centrist “catch-all” party to appeal to wider spectrum of Singaporeans.34 Organizationally, the PAP’s reforms after 1963 to change from mass party to cadreparty model helped to tighten party cohesion after the mass exodus of members.35 As Chan  34  According to Kirchheimer, “catchallism” meant a reduction of the party’s ideological baggage; b) strengthening of the top leadership groups, whose actions are judged based on their contribution to the entire social system; c) downgrading of the role of the individual member; d) in favour of recruiting voters among the population and e) securing access to interest groups for financial and electoral reasons (1966, 190). 35 See Lee’s account of how the moderates regained grounds after the communists left (1998,174).  67  notes, the PAP is “a cadre party in a mass party guise” (1985, 159). Party building includes the use of para-political institutions for integrative functions. As Panebianco says, territorial penetration occurs when the “centre” controls, stimulates or directs the development of the “periphery” (1988, 50). The greater the degree of penetration to the ground, the more institutionalized a party is likely to become. The following sections will examine how the PAP’s strategic coordination laid the foundation for its institutionalization. Propaganda Campaigns. To begin with, a Special Committee headed by Dr. Goh Keng Swee was formed in July 1961 to find ways to improve party morale. The Party’s special committee was asked to weed out career-seekers and disloyal elements by holding more meetings between party leaders and branch officials and increased political indoctrination (Shee 1971, 92). Besides, Lee and his colleagues also went on massive constituency tours in 1962 to 1963 to visit all 51 constituencies to rally grassroots support and boost morale (Fong 1979, 107). As a result of the early partnership with the procommunists, Lee learnt that the communist styled mass psychology and adroit manipulation of stage management were critical in winning the hearts and minds of the people.36 Propaganda campaigns were one of the mobilizing strategies that the PAP used to build civic consciousness. As Lee said: “We mounted a series of well-publicized campaigns to clean the streets of the city, clear the beaches of debris and cut the weeds of unkempt vacant land. It was a copycat exercise borrowed from the communists” (1998, 322). Lee gave 12 broadcasts over the radio in English, Malay and Mandarin to explain the reasons behind the infightings, the events leading to split between the two factions and rationale behind the PAP’s proposal for the merger (Lee 1998, 398-9). In an era without TV, Lee’s broadcasts had a far-reaching public relations impact and changed the mass perception of the PAP. Besides, the party’s bilingual organizing secretary, Ong Pang Boon also started a recruitment drive. Building Para-Political Institutions. To rebuild the party branches, the party leaders appointed the loyal 26 Assemblymen as Party branch chairmen in their respective constituencies (Pang 1971, 15). The tradition of appointing a PAP MP as a chairman of the party branches remained till date. Under their guidance, the branches regrouped the cadres,  36  As Lee said, “It took me two years from 1954 to 1956 to fathom their methods, to get glimpses of their intrigues and deviousness and to understand the dynamics of the communist united front (CUF) (1998, 174).  68  supporters and re-established networks (Fong 1979, 106). Interestingly, the PAP did not expand or invest more resources at the branch level. Instead, the PAP decentralized and left the branches to be financially self-sufficient, autonomous in recruitment and solving local constituency problems (Mauzy and Milne 2002, 43). The PAP leaders learnt that they could not rely on mass-party organizational techniques and the concentration of activities in one single organization would leave the PAP vulnerable to one blow (Chan 1985, 161). With the impending referendum in 1962 and Assembly Elections in 1963, the PAP used the People’s Association (PA) to re-establish links with the clan associations, civil and cultural groups, the Community Centres and the Works Brigade (Bellows 1968). Established as a statutory board37, the PA then draw on state resources to build a sprawling network of Community Centres (CCs); Citizens' Consultative Committees (CCCs); Community Club Management Committees (CCMCs), Residents' Committees (RCs) in public housing estates and Neighbourhood Committees (NCs) in private estates for political communication and political participation purposes38 (Bellows 1968, 36-7; Seah 1985, 174-6). See Table 3.5. These para-political institutions are critical to PAP’s consolidation of power and building of mass support.39 Indeed, the CCs provided a wide range of recreational, social and sports courses, activities, programmes and facilities in the local constituencies, while the CCMS ran the CCs. On the other hand, the CCCs were created as institutionalized feedback mechanisms for the masses to relay messages from the ground to the government.  37  The PA is anchored in the government structure. The Chairman of its board of management is the Prime Minister and the Deputy Chairman is a Cabinet minister. 38 See Tan (2003) for a critique of the CDCs as tools to mute the antagonistic tendencies of civil society. 39 As Seah notes: “the government realized that it could afford not wait for the party to regain its former organizational and mobilizational capabilities. The parapolitical organization was to perform this vital task of consolidating the ruling government…institutionalized channels where the norms of the new political community envisaged by the PAP leaders would be fostered” (1985, 177).  69  Table 3-5: List of Para-Political Institutions under the PA in Singapore Para-Political Institutions Aims 1960 People’s Association (PA) Established as a statutory board, chaired by the Prime Minister to promote closer ethnic ties 1962 Community Centre/Clubs (CCs) To provide recreational, social and sports courses, activities, programmes, facilities and kindergartens. 1964 Community Centre Management To manage the CCs. Local leaders appointed to Committees (CCMCs) newly established CCMCs. 1965 Citizens' Consultative Committees Committees led by the MPs. CCCs members are (CCCs) selected by the MPs to provide feedback 1972 PA Youth Movement (PAYM) To encourage youths to participate in local, communal activities 1978 Residents' Committees (RCs) Created to improve the living environment, safety and security of their estates; run by volunteers 1987 PA Indian Cultural Groups Coordination Council 1988 PA Malay Activity Groups Coordination Council 1990 CCs upgraded and renamed Community clubs 1993 CCCs and RCs Officially joined the PA network 1997 Community Development Councils 9 CDCs were first set up, later regrouped into 5 (CDCs) districts that administer programs to promote community bonding and cohesion. 5 appointed mayors lead the CDCs. 1998 Neighbourhood Committees (NCs) Part of CDCs, set up to improve the social and physical environment in private estates. 2002 Full time mayors appointed to lead 5 CDCs. Source: “PA’s Milestones.” Straits Times, 29 Jan 2010.  To counter the opposition BS’s provision of pre-school education in the 1960s, the PAP also operated a chain of low-cost kindergartens, ran directly by the branches until 1986, where a charitable arm of the PAP (PAP Community Foundation) took over. Organized similarly to the PA, the PCF is a large organization with 84 branches, over 300 centres and 3,000 employees. Together, the para-political institutions played a “vital role in political communication and brokerage between the masses and the government” and helped foster a sense of community and well-being (Seah 1985, 174). For example, in 1969, the CCs were mobilized to make house visits to calm fears and soothe tempers to avert violence resulting from riots between Malays and Chinese in Kuala Lumpur, Malaysia (Clutterbuck 1985, 324). CCs were set up in all 51 constituencies by 1961. Starting with only 28 CCs, the PA has expanded to 105 CCs, 550 RCs and 100 NC centres. As of this writing the PA has over 1,800 grassroots organisations with more than 25,000 volunteers or grassroots leaders. 70  The establishment of the PA was the most effective strategy of controlled mobilization that served the critical functions of co-opting the local talents and directing them to the government’s socio-political development programs (Seah 1985, 174). Besides, the CCCs act as brokers between the MPs, the PAP government and the people and replaced some of the traditional party functions by connecting the Party with the masses.40 They provide outlets for the people to express their interests and avenues to be socialized and abandon their sub-group nationalism and anti-system beliefs. In sum, the CCCs and CCMCs are channels where people can participate in local politics and provide feedback to the MPs. They are tools used to rationalize government policies so that they are more palatable and acceptable to the people. For example, in the 1960s, the CCs facilitated the acceptance of national conscription among the Singaporeans. It was also through the CCs that the PAP launched campaigns such as “Keep Singapore Clean and Pollution Free” and exhibitions to rationalize the significance of a citizen soldier for the defence of Singapore (Chan 1991, 165). For the growth of para-political grassroots organizations under the PA, see Table 3-6. Table 3-6: Number of Para-Political Organizations Under the PA (1996-6) Para-political Institutions 1996 2001 2002 2003 2004 Citizens’ Consultative 81 83 84 84 84 Committees (CCCs) Community Centre/Club 115 106 110 108 106 Management Committees (CCs/CMCs) Residents’ Committee (RCs) 454 529 534 536 544 Total No. GR Organizations 728 728 734 Citizens’ Consultative 3,034 3,054 3,165 Committees (CCCs) Community Centre/Club 2,418 2,786 2,858 2,879 2,861 Management Committees (CCs/CMCs) Residents’ Committee(RCs) 10,869 11,439 11, 281 11,343 11,453 Total Membership 17, 173 17,276 17,479  2005 84  2006 84  104  105  545 733 3,223  549 738 3,335  2,964  3,092  11,643 17,830  11,709 18,136  Source: Singapore Social Statistics in Brief, 2004, 2005 and 2007; Singapore Yearbook of Statistics 2007 and National Volunteerism Survey 2004.  Instead of building party branches, the PAP scaled down branch activities. Organizationally, the PAP has only 8-9 paid full-time staff to run a small party headquarters located in an isolated location in Changi. With the PA created as a statutory board, the PAP  40  As Chan notes: “The Government had nurtured a viable local leadership whose contribution in the implementation and feedback process enhanced the capacity of the political leaders to govern” (1991, 165).  71  was able to draw government resources and free its party branches of the burden of organizing educational, sports and welfare activities at the grassroots level. Doing so meant that the line between the party and state becomes unclear. While the para-political organizations under the PA are quasi-governmental organizations, the opposition is excluded and prevented from using the bases to mobilize support (Mauzy and Milne 2002, 95; Seah 1985, 177 and 191).41 Opposition WP leader, Sylvia Lim complains that: Under the Town Councils Act, the incumbent MP of a constituency will be in charge of the Town Council which controls the use of common space. As for the Community Clubs, these are in the hands of the People’s Association. It is next to impossible for an opposing candidate to be allowed to use a space to organize activities or dialogues. We have applied for permission to use spaces in PAP wards, and received expected rejections. (Constitutional Amendment Bill, 27 Apr 2010).  Some scholars suggest that the PAP’s strategy of building para-political institutions has undermined the PAP’s organizational capability (Chan 1985, 160; Lam 1999, 266; Mauzy and Milne 2002, 49). However, I argue that these institutions are extensions of the PAP as they propagate its party policies. The para-political institutions also have indirect partisan linkages as they are headed by PAP MPs. Under the PA’s Act, the PM is the Chairman. A senior minister would typically be appointed as Deputy Chairman. Having celebrated its 50th anniversary recently, the PA’s network of organizations is like the PAP’s penetrative tentacles that connect the state, party and society. It is unclear whether the PAP will give up control over the PA should party alternation occur. As E. Tan concurs, voters have grown to perceive the CDCs, Town Councils and the GRCs as PAP related entities. The PA’s successes in delivering social and recreational services are claimed as the Party’s achievements (2005, 419). To appreciate the extent of the PAP’s penetration to the grassroots level, I suggest looking beyond the party’s formal, lean party structure to consider how these quasi-political institutions touch the people’s daily lives in concrete ways. As extensions of the Party’s formal organization, the PA has blurred the line between the government and party reinforcing the PAP as a party-state. Like Seah, I contend that “unless the opposition is in a position to duplicate a similar set of parapolitical institutions, it will be extremely difficult for  41  Milne and Mauzy contend that while the BS gained control of the secondary associations, they lacked the PAP’s astute leadership (1990, 58). Also see Mutalib (2004, 96-111) for an explanation of the BS’s decline.  72  these parties to make any significant headway in the political process, especially if continued economic and social development is assured by the ruling party” (1985, 179). “Broker Institutions”. Once in power, the PAP took active steps in “de-pluralization” – breaking down ethnic boundaries and exclusiveness of social and political organizations to foster a common Singaporean national identity. De-pluralization refers to the twin process of decreasing the significance of “parallel institutions” (an institution not shared among the ethnic groups) and increasing the significance of “broker institutions” (an institution shared by members of different ethnic groups) that bridge two or more ethnic groups such as family, education, government and parties. The PAP inaugurated 7 broker institutions to serve integrative functions and broke down socio-economic cleavages to foster national identity. For example: 1) integrated schools of different ethnicities; 2) inter-ethnic participation was actively promoted in schools; 3) school textbooks were Singaporeanized; 4) state symbols such as Singapore Flag, National Day Parade and National Anthem were designed to foster national identity; 5) bilingualism schools were made compulsory in 1966; 6) Housing Development Board was set up in 1960 and integrated public housing estates that consisted of all ethnicities; 7) technical and secular subjects were introduced to build a new Singaporean culture that is shared by all members of ethnic groups (Chiew 1990, 55). All these nation-building efforts helped to develop cross-cutting ties, break down negative stereotypes and neutralize ethnic strife. With the help of organizations under the PA, the PAP was able to dismantle ethnic barriers and encourage communal, integrative activities for peoples of different ethnicities. As Brown argued, since the 1980s, the PAP has employed inclusionary corporatism that encouraged co-operative relations between state and private enterprises42 (1994, 76). Besides, the PAP government has also imposed corporatist controls on its workers and placated the workers through the redistribution of the surpluses earned (Case 2002, 87). Corporatist techniques were extended to the realm of ethnic identities which included reinforcing ethnic identities and engaging the Chinese community through the creation of Chinese Development Assistance Councils (CDAC); the dispensing of benefits to  42  According to Brown, the PAP’s corporatist strategy involves the state’s attempt to structure political around three facets: loyalty, value and interest. He contends that the state’s corporatist tendency has developed to such an extent that the ruling elites seeks to depict and organize every party of Singapore society along ethnic lines; even economic, political and social areas unrelated to ethnicity (1994, 76-7).  73  the Malays through the Mendaki program and the Indian community through the Singapore Indian Development Agency (SIDA). But critics such as NMP Viswa Sadasivan argue that these self-help groups reinforce race-consciousness and racial segregation.43 Legal Constraints. To quell union strikes, social and ethnic unrests, the PAP put in place a range of legal and social control mechanisms to maintain social order. The nationalist movement, backed by the trade unions and Chinese students that first propelled the PAP into the political scene, was later systematically dismantled by the PAP. As most scholars observed, Singapore labour-capital relations were channelled, institutionalized and regulated by a tight framework of labour laws and regulations (Luther 1978, 221). To begin with the PAP introduced the Trade Unions (Amendment) Bill and the Trade Union Bill in 1960 to empower the Registrar of Trade Unions to refuse registration or deregister unions in 1959. Between 1959 and 1960, a total of 106 unions lost their registration because of demands for high wages or were considered “infiltrated by communists” (Luther 1978, 225). By 1965, union movement and strikes were brought under control (Vasil 1989, 149). In 1970, the number of strikes and union disputes were contained at 5 and 486 respectively, a far cry from a high of 116 strikes and 1, 225 disputes in 1961. See Table 3-3. The transformation of the Singapore labour movement, its reorganization by the state was arguably one of the “most important single factor(s) in the shaping of contemporary Singapore” (Luther 1978, 229). Presently, Singapore’s industrial labour relations are characterized by a tripartite structure with joint decisions by representatives of labour, employers and government; issues such as wages are resolved through the tripartite National Wages Council.44 The scope of collective bargaining is restricted by legislation, and bargaining over transfers, promotions, layoffs and job assignments is disallowed. As Surin observes: “The Singapore labour system is a greatly oppressive one in which highly regulated trade unions appear to be engaged in collective bargaining. However it is a bargaining process that is strictly controlled by the state. In addition, union activity such as striking is closely regulated, union leadership is scrutinized by the ruling party, and wage increases are capped by government policy” (1996, 145).  43  See “NMP as the harbinger of controversial and divisive issues,” New Asia Republic, 1 Jan 2010. For a critique of the top-down corporatist model of the Singapore’s industrial relations, see Barr (2000) and Leggett (1993). For a legal interpretation, see Surin (1996). 44  74  Singapore’s last racial riot was in 13 May 1969 when the election results in Malaysia led to clashes between the Chinese and Malay communities in Singapore.45 To prevent a replay of the ethnic or religious riots, the PAP government introduced a range of legislations that include the Public Entertainments Meeting Act (PEMA)46, the Miscellaneous Offences Act (MOA), the Newspaper and Printing Presses (Amendment) Bill47, the Singapore Societies Act, the Maintenance of Religious Harmony Act (MRHA, 1992), and the Seditious Act (1985) to ensure ethno-religious harmony and control over public gatherings.48 In 1970, the Presidential Council (later renamed Presidential Council for Minority Rights) was provided for in the Constitution to scrutinize the Bills passed by the House to ensure that the proposed law does not discriminate against any race, religion or community.49 The Sedition Act was also introduced in 1985 to curtail any attempt that might “promote ill-will or hostility between the different races or classes” (Subsection 3 of Sedition Act, 1985).50 See the list of legislations that prevent the politicization of ethno-religious cleavages in Singapore below. Significant Legislations on Public Assembly and Organization Year Legislations 1960 Trade Union Bill 1979 Public Entertainments Meeting Act (PEMA) 1985 Seditious Act Singapore Societies Act 1986 Newspaper and Printing Presses (Amendment) Bill 1992 Maintenance of Religious Harmony Act 2009 Public Order Act Source: compiled by author.  45  See Pang, Tan and Cheng (1989, 128-143) and Vasil (1989, 144-170) on how the PAP tamed the labour movement and unions in Singapore. 46 Outdoor protests and marches are required to obtain a permit under the PEMA. The only outdoor place that does not require a permit to demonstrate is at the Speaker’s Corner, venue for outdoor political speeches. See Public Entertainments and Meetings Act, Chapter 257 of Singapore Statutes Online at: http://statutes.agc.gov.sg. and Public Entertainment (Amendment) Bill 2000 at http://www.mha.gov.sg/basic_content.aspx?pageid=65. 47 This bill, passed on 1 Aug 1986 prevents foreign interference in Singapore politics. It forbade newspapers from receiving foreign funds and allows the PAP government to restrict the sale of publications deemed to have interfered in domestic affairs. Magazines such as Time, Asiaweek and Far Eastern Economic Review have had their circulations restricted in 1986 and 7. See Chronicle of Singapore (2009, 128 and 210). 48 See IBA Human Rights Institute Report, July 2008. 49 See Tan (1989, 24) and Chapter 167A in the Singapore Constitution. 50 See all the legislations on Singapore Statutes Online at: http://statutes.agc.gov.sg  75  Amongst all the legislations, the MRHA is the most unique and powerful invention that gives the government an exceptional jurisdiction to restrain leaders and members of religious groups from carrying out activities, "exciting disaffection against" the government, creating "ill will" between religious groups, or carrying out subversive activities. In 2009, prior to the Asia Pacific Economic Co-operation meeting in Singapore, the Government passed a Public Order Act51 to bar all public assembly without prior police permit. Most of the legislations have been designed to empower the government to police race and religion in the public arena.52 With the advent of information technology, the PAP has introduced a range of legislations to regulate political discussions on chatrooms, blogging and social network websites on the Internet. Unlike the Chinese government, the PAP government engages in minimal Internet filtering, blocking only a small set of pornographic web sites. Instead, Singapore employs a combination of licensing controls and legal pressures to regulate internet access and limit the presence of objectionable content and conduct online (OpenNet Initiative Report Jun 2009; Hachigian 2002). In Sep 2005, the Sedition Act was first used when 3 men were charged for making inflammatory racist comments on the Internet. In 2006, a young man received a police warning under the same Act for putting up offensive cartoons of Jesus Christ on his blog while 3 teenagers were arrested over racially insensitive remarks on Facebook. 53 More recently, an evangelical Pastor was called up by Internal Security Department for disparaging Buddhism and Taoism during a sermon that was posted on YouTube.54 See list of charges made under Sedition Act.  51  See “Singapore enacts new law to limit freedom of assembly,” Reuters, 15 Apr 2009 and Public Order Act 2009 at Singapore Statutes Online. 52 The MRHA was the brainchild of former Law Minister Jayakumar. See “Foolhardy to take harmony for granted,” Straits Times, 25 July 2009; U.S. Dept. of State. Human Rights Report 2008: Singapore. 25 Feb 2009 and U.S. Dept. of State. International Religious Freedom Report 2009: Singapore. 26 Oct 2009. 53 The PAP has been successful in curbing dissenting voices amongst traditional print and broadcast media, it is less so with regulating personal blogs, vodcasts and podcasts. See Channel NewsAsia, 5 May 2006. 54 “Online joke turn nasty,” and “Pastor called up by ISD,” Straits Times 5 Feb and 8 Feb 2010.  76  Charges made under Sedition Act in Singapore Date Events Sep 2005 Sedition act first used on 2 men for making seditious and inflammatory racist comment on internet 2006 Young man received police warning for putting offensive cartoons of Jesus Christ on his blog Apr 2008 A middle-aged Christian couple charged for distributing seditious publications to two Muslim women in 2007 Jan 2009 3 teenagers charged over racially sensitive remarks on Facebook Feb 2010 A pastor called up by ISD for disparaging remarks about Buddhism and Taoism Aug 2010 A man jailed for casting aspersions on Prophet Mohammad55 Source: Compiled based on various media and newspaper sources.  Presently, ethnic and religious politics are banned in Singapore and ethnic parties are nearly non-existent. Subsumed under a nation-building rationale, strict rules regulating public and political discourse have deprived the potential opposition challengers of any space for social or political mobilization. Opposition parties with representation in Parliament such as the Workers’ Party56 (WP) and the Singapore People’s Party (SPP) are led by Chinese leaders. The Singapore Malay National Organisation, or Pertubuhan Kebangsaan Melayu Singapura57 (PKMS) is the only Malay-based political party representative of Malay interests that is plagued by power struggles and a spent force. The newly formed Reform party (RP), led by JB Jeyaratam’s son, Kenneth Jeyaretam, an ethnic-minority leader of mixed Indian and English descent is inclusive and open to members of all ethnicity and religions. Public Goods. The PAP relies heavily on bureaucratic expertise58 for policy formulation and execution to improve its performance legitimacy (Chan 1989, 75-6). To boost economic growth, Lee called on former civil servant Dr. Goh Keng Swee to restructure the economy and tourism sector in the 1960s. To provide affordable housing, Lim Kim San was tasked to set up Housing Development Board (HDB) – a crucial project that launched two 5-year plans for urban renewal and provided subsidized housing to replace squatters.  55  This is the first case in Singapore’s history where a man was jailed under the Sedition Act. See “Jailed for religious insults.” Straits Times, 6 Aug 2010. 56 The WP is the longest surviving party and was first led and David Marshall and JB Jeyaretnam, an Indian lawyer. JBJ was an MP from 1981-6 and again as a Non-Constituency MP from 1997-2001. 57 PKMS started as a branch of Malaysia's UMNO party in Singapore in 1951. Then it made efforts to help provide affordable housing for Malays. Yet, none of the PKMS candidates who contested in past general elections has ever won a seat. Since 2001, the PKMS has contested as part of the SDA coalition. See “Party once had a part in helping Malays” and “Is this $10m site the reason behind the PKMS fracas? Straits Times 13 Sep 2009 and “PKMS factions to meet,” Straits Times 7 Aug 2010. 58 A political study centre was set up to train and socialize civil servants about the social and economic threats to Singapore. It was later known as Civil Service College.  77  50,000 units in high-rise blocks were built by 1965. Within 5 years, another 60, 000 were completed. By 1972, 42% of the population were living in HDB housing and by 1981, 70%. Since 1988, 16 town councils59 were formed to manage the 900,000 HDB flats – 14 town councils by PAP members of parliament (MPs), while the other 2 are by opposition MPs. The New Towns housed around 600 to 1000 (25000 to 50000 inhabitants) units with its CCs, swimming pools, social and recreational amenities. If the PAP had not been able to deliver results and build houses in prominent and strategic locations, Lee Kuan Yew may not have been re-elected in 1963 GE (Lee 1998, 344). The provision of affordable public housing was a visible, public good that earned the PAP extensive legitimacy and ground support (Quah 1985, 247; Chua 1997, 129). Singapore’s success in public housing has won the country much international acclaim.60 It is a public good that the PAP is wont to remind the electorate of in order to make political capital out of it. A lot has been written on the national housing policy and the PAP’s electoral successes and will not be repeated here.61 It suffices to say that the PAP’s provision of public good such as housing reinforced the Party’s governing capability and legitimacy. Besides generating legitimacy, the HDB public housing programme was also a critical nation-building tool. As a result of early British colonial town planning, various immigrant ethnic groups were separated spatially in Singapore (L.L. Sim et al. 2003). To prevent the growth of ethnic enclaves, the PAP introduced an Ethnic Integration Policy (EIP) in the 1960s to integrate Singapore’s multi-racial population.62 Under the policy, maximum proportions are set for all races - Chinese, Malays, Indians in each HDB block and neighbourhood. Once the limit for an ethnic group is reached, no further sale of HDB flats to  59  In 1988, A Town Council Bill was passed such that the town council will take over the management of HDB estate from the HDB. There were considerable debate regarding the role of the Town Council – whether it should serve as an estate management corporation or meant as a form of local authority with wider residential participation. See “Missing Link Between Town Councils And Residents,” Straits Times, 27 Jun 2009. 60 In 2009, Singapore was ranked first, amongst 215 cities, to having the best city infrastructure by Mercer’s 2009 Quality of Living Survey. 61 See Chua (1995, 124-146) and Quah (1985, 233-258). 62 The EIP is applicable to the purchase of new flats, resale flats, replacement flats, as well as the allocation of rental flats in all HDB estates. “PRs may be subjected to ethnic integration policy in buying flats,” Channel NewsAsia, 29 Jan 2010.  78  that ethnic group is allowed.63 For example, when Chinese quotas are binding, non-Chinese sellers cannot sell to Chinese buyers because the transaction increases the Chinese proportion above the quota. Research shows that the EIP is effective in reducing the intensity of the ethnic enclaves while increasing social integration (L.L. Sim et al. 2003). With more than 80% of Singaporeans in public housing, issues concerning the prices and governance of the housing estates have the potential of being politicized.64 The PAP government has introduced legal restrictions to prevent the opposition claiming credit for housing or from politicizing it to gain leverage with the electorate. Chapter 7 will consider the effects of these strategies on the opposition from setting political agenda. Racism and Implications. While the PAP government has gone to significant lengths to foster racial harmony and to discourage intolerance, this does not mean that racism or discriminatory policies do not exist. For example, the Malays, the largest ethnic minority community constituting around 14% of Singapore’s total population, are known to be excluded from top, sensitive military positions. From 1969 to 1973, no Malays were conscripted into the Singapore Armed Forces (SAF) and regular Malay soldiers were removed from combat posts. Early retirement was encouraged and promotional prospects curtailed. Malay loyalty is of question as they are expected to fight against Muslim brothers in Malaysia and Indonesia.65 Limited conscription of Malays into the SAF only recommenced in 1973 but restrictions remained on their placement in critical positions (Huxley 2000, 1024).66 Critics argue that the some of the PAP’s policies are counter-productive and have an effect of reifying racial stereotypes and categories of the ethnic minority groups. Official procedures such as form-filling, government data collection and Singapore identity cards required the filling of one’s “Race” are inconsistent with the “regardless of race” tenet  63  As a result of the increased influx of immigrants, the government has also imposed similar EIP scheme on immigrants to prevent the congregation of permanent residents. See Channel NewsAsia, 29 Jan 2010. 64 In 1981 Anson by-election, the PAP lost a seat to J.B. Jeyaratnam of the WP as a result of voters unhappiness with increases in the HDB flat prices and housing plans (Quah 1985, 248). 65 For the PAP government’s position, see “Responding to the Observations of the MPs,” Petir, Apr 1987 and “This is a Singapore problem. We will solve it ourselves,” Straits Times, 18 Mar 1987. 66 All eligible Malays were only called for to serve National service in 1985. More Malays are allowed to participate in the SAF. Bur very few Malays are admitted to the Air Force. For a critical study of the restriction of Malay recruitment in the SAF, see Walsh (2007).  79  enshrined in the national pledge. Recently, a fact finding mission conducted by the UN in April 2010 found that ethnic marginalisation, entrenchment of minority status through ethnic categorisations, the minority political representation dimension to the GRC system, the academic under-performance of Malay students and under-representation of minorities in the SAF, Police, intelligence services and the Judiciary are pervasive, even today.67 The UN Report is important as it highlights how strict controls over political debates on ethnic, racial and religious issues have had the effect of stifling the understanding of the social issues facing Singapore. However, the PAP was swift to dismiss calls for greater openness in the discussion of ethnic, religious issues and defend its right to police ethnicity and religion.68 Defending its right and “responsibility” to police ethnicity and religion, the statement from the Ministry of Foreign Affairs to the UN report says: “Race, language and religion will always be sensitive issues in Singapore. This does not mean that they cannot be discussed, but a balance must always be struck between free expression and preservation of racial and religious harmony. This balance is only for the Singapore Government to determine because only the Singapore Government bears the responsibility should things go wrong. The UN bears no such responsibility and we see no reason to take risks for the sake of an abstract principle. We believe most Singaporeans agree with the Government's approach” (Straits Times, Apr 29, 2010).69  Conclusion This chapter has attempted to show how early political experiences and societal cleavages have important implications for a party’s organizational model and long-term survival. In Singapore, early struggles toward independence and painful party formation experiences were critical junctures that led to the PAP’s adoption of the exclusionary, cadre system that tightens party discipline and rewards loyalty. To voters and supporters, the PAP has projected itself as an inclusive, pragmatic “catch-all” party, representative of peoples of all ethnicities, religions and social class. I have argued that one of the PAP’s strengths lies in its wide network of para-political institutions and the introduction of a myriad of legal rules and legislations to maintain social order and racial harmony. The para-political institutions  67  See “Certain ethnic groups marginalised by government policies,” Online Citizen, 28 Apr 2010 and “SFD written submission to UN Special Rapporteur,” Singaporeans for Democracy, 22 Apr 2010. 68 For the Singapore’s government full response to the UN report, see “UN expert's comments draw swift Government reply,” and “Foreign Ministry responds to UN expert's comments,” Straits Times, Apr 29, 2010. 69 For the Singapore’s government full response to the UN report, see “UN expert's comments draw swift Government reply” and “Foreign Ministry responds to UN expert's comments,” Straits Times, Apr 29, 2010.  80  are now part of the regime’s strategy that polices public discourse and prevents partisan mobilization based on socio-ethnic tensions or consciousness. In Singapore, Lipset and Rokkan’s “freezing thesis” is limited in explaining how socio-ethnic cleavages affect party system formation. Based on the ethnic fractionalization index score by Fearon, Singapore has higher respective ethnic and cultural fractionalization scores (0.388 and 0.388) than Taiwan (0.274 and 0.169) (2003).70 One would consequently expect a higher party system fragmentation and polarization in Singapore than in Taiwan. This was the case in pre-independent Singapore where ethno-linguistic and class cleavages mobilized citizens into various parties and voter alignment that fostered a competitive party system. However, after the PAP came into power, the predicted pattern of socio-ethnic cleavages shows severe limitations when it tries to explain the rise and persistence of the PAP. Presently, ethnic politics are disallowed and ethnic parties are nearly non-existent in Singapore. The early ethnic-linguistic cleavage structures did not “freeze” but were “thawed” and deliberately reconfigured to foster a common Singapore national identity. As Huntington says, “Party systems originate in the patterns of cleavage and alignment among social forces. Different relationships among social forces and different sequences in the development of cleavages among them give rise to different types of party systems. Once the system takes root, however, they develop a life of their own” (1970, 10). The PAP’s successful turn-around from near decimation to hegemony hinged on its strategic coordination - supply of public goods such as para-political grassroots organizations and public housing to fulfil the needs of the electorate. The restriction of coordination mechanisms through rules and legislations prevented the opposition from politicization based on ethnic or religious grounds. Institutional building was key in constraining dissent. In the next chapter, I compare the experiences of the KMT and the PAP and examine how its early civil war experiences affect its party organization and early exclusionary policy widened Taiwanese society’s sub-ethnic cleavage. Specifically, I examine the KMT’s policy reversal to adopt inclusionary “Taiwanization” policy to address the sub-ethnic tensions heightened intra-party tensions and triggered regime “de-institutionalization” in the 1980s.  70  Singapore is ranked 10th while Taiwan is ranked 15th in the region for cultural diversity (Fearon 2003).  81  4  The KMT: Organizational Adaptation and Coordination Dilemmas The last chapter demonstrated how early historical experiences and organizational  development lay critical foundation for the PAP’s institutionalization in Singapore. In this chapter, I examine how the KMT’s organizational reforms in the early 1950s transformed the revolutionary and exclusionary one-party state to a pragmatic and inclusionary hegemonic party regime. It is argued that historical and socio-ethnic cleavages are not deterministic but contextual factors that shape the strategic calculations of the ruling elite. In response to the changing socio-economic developments, Chiang Ching-kuo initiated inclusionary policy and triggered the regime’s de-institutionalization – a dismantling of authoritarian institutions that facilitated the economic and political liberalization of the 1980s. De-institutionalization was de-stabilizing. The implementation of Taiwanization that reversed the discriminatory policy against Taiwanese posed new challenges to the KMT. As Chapter 6 will show, after Chiang’s passing, the KMT under Lee Teng-hui’s leadership experienced the worst factionalism. Torn apart by irreconcilable visions of national identity and sub-ethnic cleavage, intense leadership struggles led to the KMT’s first split in 1993. Huntington has hypothesized that as a result of socio-economic changes, the revolutionary and exclusionary party may transform itself into an established and inclusionary party. He divided the evolution of a revolutionary one-party system into three periods: 1) transformation, characterized by strong ideological commitments and an autocratic leader; 2) consolidation, characterization by pragmatism and institutionalization; and 3) adaptation, characterized by reliance on technocrats, with more room given to interest groups, intelligentsia and popular participation.1 The evolution of the KMT appears to follow Huntington’s model of party change (Jian and Wu 1992; Meaney 1992). For example, the KMT under Chiang Kai-shek (19491975) signified a time of organizational transformation and consolidation as a one-party, quasi-Leninist monolithic state. But under Chiang Ching-kuo’s rule (1975-1988), the KMT’s adapted as a pragmatic hegemonic party and initiated liberalization. Significantly, the lifting  1  Huntington’s classification is similar to Sartori’s, who divided single-party systems into three types: 1) totalitarian, 2) authoritarian and 3) pragmatic (1970, 32-40).  82  of martial law and the ban on formation of political organizations under Junior Chiang’s rule signified a gradual transition from “hard to soft authoritarianism” (Winckler 1984). As a result, the KMT regime was able to fend off international and domestic pressures to fully democratize and prolong its rule through capitalizing on its incumbency advantage and relative strength to the opposition (Cheng 1989; Cheng and Lin 1999; Tan 2002). The analysis of the KMT’s organizational adaptation and strategic coordination proceeds as follows. The first section highlights the KMT’s origins and party building experiences after its move from Mainland China to Taiwan in 1949. Specifically, I focus on the effects of expanding party membership and exclusionary discriminatory policy that exacerbated Taiwan’s sub-ethnic cleavage and partisan alignment. The second section outlines the series of external and internal pressures that presented a coordination dilemma to the regime. Chiang’s decision to liberalize rather than repress at the critical juncture triggered the process of “de-institutionalization”, signified by the reversal of ethnic policy or Taiwanization, constitutional reforms and negotiations with the opposition. Instead of repression to starve off opposition, Taiwanization and negotiation with the opposition also meant that the KMT was able to co-opt and control the pace of transition. The unintended effects of “de-institutionalization” are presented in the final section. Origins of the KMT Like the PAP, the KMT party was an “externally created party” with origins dating back to the last century (Duverger 1954). Founded by Sun-Yat Sen and a group of Chinese in-exile in opposition to the Manchu monarchy during the Qing dynasty, the KMT began as a dissident force against the Yuan Shi-kai’s government in 1912. Embattled by civil war against warlords, the battle fatigued KMT was offered a lifeline after the Communist government of the USSR extended military and economic aid and training2 to the KMT in the early 1920s. Then, the Soviets provided 300,000 roubles, 40 Soviet officers as instructors, political indoctrination as well as military training (Long 1991, 41-5). Soviet aid permitted Sun to establish a KMT army and a military academy at Whampoa in 1924 and Chiang Kai-  2  Chiang kai-shek was one of the generals sent to Moscow to study Soviet military and political organization. Chiang was the director of the academy and trained 2,000 cadets in 3 years. These officers were known as the Whampoa Clique, the core of the KMT army and Chiang's political base (Dickson 1997, 59).  83  shek was appointed as its first commandant.3 The massive aid that the KMT received facilitated the party re-organization and transformation into a quasi-Leninist-style party. The KMT was formed as a revolutionary party that struggled against the Manchu dynasty and later against the Chinese Communist Party. As Tai notes, “the fact that the KMT was conceived as a revolutionary organization dominated by the military created an inevitable trend toward the personalization of power” (1970, 408). As a result of its revolutionary aims, power was concentrated in the hands of the man who held the highest command and security apparatus was used to ensure compliance while the Party usurped state functions. The strong personalistic rule of Chiang Kai-shek, and later his son Chiang Ching-kuo, led many to describe the KMT rule as “leaderist”4 or even a “one-party dictatorship” (Winckler 1988; Moody 1992; Meaney 1992). On a regime level, the KMT’s rule during the Chiangs’ era was closest to being personalistic5, as major decisions concerning access to office and fruits of office depended largely on the discretion of an individual leader. As Huang notes, “Neither Chiang Kai-shek nor Chiang Ching-kuo permitted challenges to their authority within the KMT, and although both were known to consult with close advisers, both made almost all important decisions within the party alone” (1996, 122). Tien shares this view as he says: “to ensure the regime’s political stability, the ruling political elite, particularly the Mainlanders, have relied heavily on the charisma of Chiang Kai-shek and Chiang Ching-Kuo to provide legitimacy” (1989, 75). While Chiang Kai-shek ran the KMT in a dictatorial way, he paid lip service to the 1946 Constitution and upheld fa-tong or constitutional legitimacy. My study characterizes Chiang’s leadership as “situational charismatic”6, or one in which there is a total symbiosis between the leader and organizational identity (Panebianco 1988). Organizationally, the early KMT under Chiang Kai-shek’s rule behaved as one-party  3  See Tai (1970, 406-411) for the party structure and functions of the KMT back in the mainland. Meaney argues that leaderist quality of the KMT makes it more comparable to a Leninist party than to PRI in Mexico that does not permit or allow dominant figure (1992, 98). 5 A personalist regime may be defined as one, which the leader, himself maintains near monopoly over policy and personnel decisions despite the existence of a support party (Geddes 2003, 53). 6 The significance of Chiang’s “situational charisma” on leadership succession and elite cohesion will be discussed in Chapter 6. 4  84  state as opposition parties7 are not tolerated and viewed as traitors to its revolutionary or nationalist cause (Tien 1989, 10). Since national elections were banned till 19698, party alternation could not occur. However, local elections were permitted as a means of maintaining legitimacy and keeping tabs on the local population (Huang 1995). And by the late 1960s, the one-party state was transiting to a hegemonic party9 regime; subsequently, the Party allowed elections of the Legislative Yuan. Refer to Table 4-1 for a summary of the evolution of the KMT regime based on the typology introduced in the Theoretical chapter. The following sections will now focus on the KMT’s early organizational changes and implications on sub-ethnic cleavage in Taiwan. Table 4-1: Evolution of the KMT Regime (1945-2000) Dimensions One-Party Hegemonic Party 1945-1969 1969-1986 Participation No suffrage Universal suffrage Contestation No national election Relatively free but unfair till 1969 elections under cartel regulations Legislative Absolute control Supermajority dominance (90 -100% of seats) (more than 70% of seats) Party Cannot occur Can but does not occur alternation  Dominant Party 1986-2000 Universal suffrage Free and fair elections under egalitarian regulations Absolute majority (50% or more seats) Can and may occur  The KMT’s Early Rule in Taiwan (1945-1949) Taiwan was under Japanese colonial rule for fifty years (1895-1945). After the end of World War II, Taiwan was transferred to China and ruled under the authority of a military governor, General Chen Yi, a corrupt and incompetent administrator. While Taiwan was decolonized, the political order in the postwar period resembled a colonial one under the KMT (Cheng and Haggard 1992, 3). As an outside power, the KMT imposed political control over domestic politics where the local Taiwanese population was largely excluded. The KMT denied local Taiwanese from holding key party, military and political positions  7  Two other parties, Young China Party and the Chinese Democratic Socialist Party were allowed to exist as they posed no threat to the regime and were internally split (Huang 1995, 954). 8 While local elections were allowed since 1954, limited national elections were held only in 1969 for the limited number of seats in the Legislative Yuan. 9 LaPalombara and Weiner describe the KMT as a “one-party pluralistic system” - quasi-authoritarian regime dominated by a single party is pluralist in organization, pragmatic rather than rigidly ideological in outlook, absorptive, rather, than ruthlessly destructive in its relation with other groups (1966, 38).  85  because of an on-going civil war with Communist China. The political atmosphere was tense and highly-charged as the KMT was suspicious of anyone with loyalty problems or socialist inclinations. The Party’s discriminatory practice and brutal treatment of the local Taiwanese soon gave rise to the “228 Incident”10 - a series of bloody riots that left thousands dead on 28 Feb 1947. This incident marked the beginning of the “White Terror” era where thousands of Taiwanese and Chinese, suspected of being dissidents or communist sympathizers, were jailed, tortured and killed. To date, the “228 incident” remains controversial and symbolizes the cruelty and injustice of the KMT towards the local Taiwanese population. During the first four years of rule in Taiwan, the KMT had little party activity. After its civil war experiences against the warlords and its humiliating loss of the mainland to the Chinese Communists, the KMT government under Chiang Kai-shek fled to Taiwan in 1949. After Chiang’s arrival, the KMT was committed to liberating the Chinese mainland from Communist control.11 In 1948, the KMT government passed the “Temporary Provisions Effective During the Period of Communist Rebellion” and imposed martial law in May 1949. These laws put the country in the state of emergency and exempted the President and Vice President from the two-term limit as provided in the Article 47 of the Constitution.12 Chiang also eliminated national and intra-party factionalism, which he thought led to the defeat of the KMT. Factional strife petered out after 1965, not to resurface again until mid-1980s (Cheng and Chou 2000, 42-66). To maintain civilian control over the military, the warlord or province-based military units were disbanded; rotation and fixed tenure for command was established (Dickson 1997, 60). Sub-Ethnic Cleavage in Taiwan From 1948 to 1950, an estimated half a million Chinese Mainlanders followed Chiang Kai-shek to Taiwan after the civil war. According to 1948 census, there were 316, 642 aborigines, 5,523,912 permanent Chinese settlers, 46, 190 Mainland Chinese in Taiwan (Hu 1989, 2). By 1952, the Mainlanders constituted less than 10% of total population while  10  For more on the “228 incident”, see Formosa Betrayed (Kerr 1965). Also see Roy (2002) for more on the political history of the KMT. 12 The extra-constitutional arrangements were expanded over 1950s and 1960s which suspended the reelection of the three national representative bodies – the National Assembly, the Legislative Yuan and the Control Yuan – extended the tenure of the incumbent members for life and deterred the election of the provincial and municipal heads indefinitely (Chu and Lin 2001, 114). 11  86  the Taiwanese consisted a majority of 90% of the total population in Taiwan. See Table 4.2. As a minority group, the Mainlanders dominated the key political, military and central government positions and enjoyed special privileges. Since 1949, the KMT government laid a wide, social network as well as numerous rules to discriminate against Taiwanese and ensure better social status for Mainlanders. Financial subsidies and unfair screening rules in schools as well as in government reified the sub-ethnic cleavage between the two groups.13 Table 4-2: Ethnic Composition of Total Population in Taiwan (1952-64) (‘000) 1952 1958 1963-4 Taiwanese (%) 7,478 (92) 8, 943 (89.1) 10, 349 (87.1) Mainlanders (%) 650 (8.0) 1, 096 (10.9) 1, 535 (12.9) Total Population 8,128, 374 10,039,435 11,883,523 Source: Tai (1970, 416) and Kung (1995, 53).  The label “Mainlanders”14 refer to Chinese from China and those Taiwan born, native Taiwanese who speak the Fukien dialect and whose parents came from China after the civil war (Tien 1989, 36). On the other hand, Taiwanese refer to the permanent settlers in Taiwan that include aborigines and Chinese who speak Hoklo/Hakka. These two categories are affected but not determined by their political or economic status. Broadly, Mainlanders harbour anti-Japanese sentiments as a result of experiences through the Second JapaneseSino War and tend to view Taiwanese, educated and brought up under the Japanese as traitors and with suspicion. On the other hand, Taiwanese tend to view the Mainlanders as poor, backward and corrupt.15 Despite the stereotypes and ambiguities in the sub-ethnic classification of Mainlanders and Taiwanese, the two categories persist. Like the PAP that initiated a series of economic development projects in Singapore, the KMT launched a series of social welfare programs in the 1950s. For example, the KMT government constructed public financed national housing units, mobilized the Armed Forces to develop social communities, created a cabinet commission to provide job retraining and employment, and extended free education from six to nine years (Tai 1970, 428). These social welfare programs were well-received and generated performance legitimacy for the mainlander-dominated KMT regime. However, during the course of economic development,  13  For more on ethnic politics in Taiwan, see Wang (2004) and Wu (1995). Mainlanders also include those with Taiwanese wives and children (Hu 1989, 255). 15 I found similar negative stereotypes and mutual suspicions to persist between the DPP and the KMT supporters during my fieldwork in Taiwan in 2007. 14  87  Mainlanders with official or party connections used government prerogatives for personal gains, which the local Taiwanese detested (Huntington and Moore 1970, 29). Comparing the KMT and the PAP’s Ethnic policy Comparatively, the KMT government’s early discriminatory policy was more extensive and systematically applied than the PAP’s. In Singapore, the most evident and clear discriminatory policy against ethnic minority is the exclusion of Malays from sensitive and high ranking positions in the Singapore Armed Forces. For fear of split loyalty in times of war against neighbouring Muslim neighbours like Indonesia or Malaysia, Malays were excluded from national conscription in early years (Huxley 2000, 102).16 However, since this discriminatory policy began to be phased out in the 1970s, the proportion of Malays doubled over the course of 7 years. In 1985, all eligible Malays were called up for national service and more assumed higher-ranking military positions, with two Malay fighter pilots in the Air Force (Interview with RSAF Officer, 15 Jan 2010). In contrast, the KMT kept local Taiwanese away from sensitive postings in the military, the central government and also the party bureaucracies from 1950s to 80s. To prevent coups, the Taiwanese military only enlisted Taiwanese as soldiers and Mainlanders as officer corps.17 See Table 4-3 that shows Mainlanders occupying more than 90% of leadership positions, while more than 50% Taiwanese were foot soldiers from 1950-1965.18 Table 4-3: Taiwanese-Mainlanders Composition in the Military, 1950-1987 Generals Colonels Lieutenants M T M T M T 1950-65 97.7 1.3 90.4 9.6 86.2 13.8 1965-78 92.6 7.4 81.2 65.3 34.7 34.7 1978-87 84.2 15.8 67.4 32.6 51.7 48.3 Source: The Journalist, 7 Dec 1987, 9. *M = Mainlanders; T = Taiwanese  Soldiers M 47.2 31.6 21.3  T 52.8 68.4 78.7  As mentioned, in Singapore, the PAP has suppressed ethnic tensions in Singapore through the use of “parallel” and “broker institutions” and a range of socio-legal constraints  16  See Walsh (2007, 265-285) who argues that the SAF’s exclusion of Malay minority and women from the military has reduced its effectiveness and professionalism. 17 Critics of the PAP argued that the SAF is likewise discriminatory as ethnic Malays are excluded from high-level positions or sensitive postings such as fighter pilots. For more see Walsh (2007) and Huxley (2000). 18 A 1967 survey showed that Mainlanders held 82% of the positions in the military, police and national security agencies while 34% in public administration and the professions (Long 1991, 63).  88  to breakdown ethnic barriers. But the KMT did not employ any of these measures. Within the state bureaucracy, discriminatory policy meant that Taiwanese only accounted for 56.5 % of the overall civil service and 37.3 % of the central government in 1959. The proportion of Taiwanese in the civil service and central government only increased to 71 and 66.2% respectively by the end 1991 (Chu and Lin 2001, 119). One of the key differences between Singapore and Taiwan in terms of ethnic policy lies in the distribution of public goods such as housing. As discussed in Chapter 3, the PAP instituted Ethnic Integrated Housing Policy (EIP) to enforce inter-ethnic interaction in the housing estates to prevent the rise of ethnic enclaves or local partisan alignment. The enforced assimilation of different ethnicities in the same housing estate increased social integration and understanding between the different ethnic groups. In Taiwan, Mainlanders with occupations in the central government resided largely in the urban areas in eastern and parts of northern Taiwan, while Taiwanese are predominantly rural folks with positions at the provincial, county and municipal levels were located in the southern parts of Taiwan. See Table 4-4 for the rural-urban geographical distribution in Taiwan. As Chapter 8 will show, the two sub-ethnic groups that are geographical differentiated are to contribute to regional variation of partisan alignment and voting behaviour in the 1980s (Lay, Yap and Chen 2008). Table 4-4: Rural-Urban Geographical Distribution in Taiwan (1952-64) (in thousands) 1952 1958 1963 and 4 Urban (%) 2, 800 (31) 4, 300 (35) Rural (%) 6, 200 (68) 8, 000 (65) Literacy (%) 3, 694, 000 (57.9%) 5, 404, 000 (69.1) 7, 653, 000 (77.6) Source: Tai (1970, 416)  In terms of economic achievement and legitimacy, the KMT regime has done remarkably well and provides a good case of exceptional growth under authoritarian regime. As Tien noted, 3 major policies namely, land reform, price stabilization and import substitution policy pursued in early 1950s were critical in propelling Taiwan’s economy (1989, 19). To pursue economic development, the KMT launched a major land reform  89  program19 (1949-1953) that benefitted tenant farmers. Essentially, the land reforms boosted the KMT’s legitimacy as it brought “a great of rural good will and developed among the farmers a vested interest in the government’s agricultural program” (Tai 1970, 428). However, reform programs have also stirred up strong resentments amongst the elite, largely Taiwanese landlords against the KMT as the reforms redistributed income; increased the status of small-owner families and undermined the status of landlords. On the whole, the land reforms laid a solid foundation for Taiwan’s agricultural growth throughout the 1950s, improved the lifestyles of small farmers, resolved land tenure problem and forestalled potential rural unrests (Tien 1989, 22-4). As White observes: the land reforms “allowed the KMT to give traditionally prestigious Taiwanese political families political leadership in nonstate enterprises, delaying their political leadership in government” (2009, 69). He argues that the equalization effects of land reforms and the redirection of indigenous Taiwanese from landowners to small and medium size firm entrepreneurs contributed to the Taiwanese economic boom in the 1980s. To fend off competition from local factions and conglomerates to maintain their oligopolistic status, the KMT had also built alliances with conglomerates and left Taiwanese to run small and medium enterprises. For many decades, the political economy of Taiwan also reflected the sub-ethnic division. While Mainlanders were largely in charge of national politics and large-scale corporations, Taiwanese remained as foreign exchange earners, entrepreneurs and businessmen from industrial production and exporting industries. The political and socio-economic elite thus overlapped the sub-ethnic cleavage between mainlander and Taiwanese. Taiwan’s ethnic division was not a zero-sum as economic opportunities were still available to the Taiwanese (Cheng and Haggard 1992, 9-10). The KMT’s Wealth and Party Assets As a quasi Leninist party state, there was overlapping character of party, state and military political authority in Taiwan under the KMT’s hegemonic rule (1949-1988). As a result of long-term incumbency and assuming national assets from the Japanese colonial  19  There were three aspects to the land reform: 1) rent reduction; 2) sale of public farmland to tenant farmers; 3) land to the tiller program which requires each landlord family to sell to the government any land that it owned in excess of 2.9 hectares of paddy field (Tien 1989, 23).  90  government after 1945, the KMT is now one of the world’s richest parties. The KMT is said to have illegally seized land and property worth millions of pounds when the Japanese left Taiwan, which they ruled as a colony from 1895-1945.20 However, “the extent of the KMT empire is Taiwan’s best kept secret” (Economist 27 Aug 1993). News reports in 2001 estimated the party assets to be between US$7 billion and US$16 billion (Schafferer 2004, 47). During Chiang’s rule, the amount of party assets was large enough to finance a huge party apparatus to prop up a strong central leadership (Kuo 2000, 94). Recent report by the Ministry of the Interior shows that in 2006, the KMT has 100 times more asset value, or roughly NT$25.5 bil (Taipei Times 18 July 2007). To support the huge party machinery for political control, the KMT relied on government resources and privileges to build a colossal business empire.21 Essentially, the KMT is funded by revenue generated by party enterprises rather than membership dues. In 1992, the income from membership was NT$70 million, which could only cover 1.4% of the party’s expenditure (Xu 1997, 410). As Chu describes it, the KMT is an “oversized, richly endowed, and autocratically run political machine” with a business empire estimated at about US$9 billion with yearly dividends exceeding US $140 mil (1996, 76). A large part of the KMT businesses concentrated in highly regulated sectors such as insurance, leasing, banking, brokerage, investment trust, mass media, public utilities and real estate (Chu 1996, 76). In the 1980s, the KMT government had owned or controlled most of the island’s media that include three television stations, four national daily newspapers (the government owned 2, and the military 5). Besides, privately owned newspapers also enjoyed close corporate ties with the KMT. For example, the owners of the two newspapers with the highest circulation, China Times and United Daily News were members of the KMT’s CSC (Rawnsley 2006, 137).  20  The KMT was reported to have taken TW$11 billion worth of private property and 860 Japanese owned or joint ventures from Mar 1947 to Dec 1950 from the Japanese (Xu 1997, 401). For more critical reports of the KMT’s party enterprises, see “Wealth probe for 'world's richest' party,” BBC, 26 Oct 2001 and “Kiss your assets goodbye,” Time, 7 Oct 2002 and “KMT assets’ magical vanishing act,” Taipei Times, 5 May 2010. 21 In Feb 1993, it was reported that the KMT party enterprise management committee has 7 major holding companies with total assets of NT $20 billion (Xu 1997, 404).  91  The KMT’s Organizational Changes, 1951-2 As Dickson notes, it “was the lessons learnt in its defeat in the civil war and the steps taken from in the early 1950s to overcome its failure that allowed the KMT to endure” (1997, 63). Obsessed with unifying China under the KMT’s rule, the KMT underwent a critical party re-organization from 1950 to 1952.22 To address the party’s failings and loss of China, Chiang launched a “Reconstruction Campaign” to re-organize the KMT (gai-zhao-yun-dong). A Central Party Reconstruction Committee was established on 5 August 1950 to transform the organizational structure and the political socialization of party members.23 The aim was to revive the esprit de corps of the cadres and rebuild the KMT’s revolutionary spirit. Like the PAP, the KMT adopted communist organizational and mobilization methods. As Chiang Kai-shek said: “If we want to beat the communist bandits, we must understand their methods and use their techniques to overcome them” (cited in Tsang 1999, 2). During these two years, the KMT strengthened its party organization, eliminated all divisive factions, created a network of party organizations in the military and government, a cadre system, a cadre school and made the party cell the basic working and training unit of the KMT.24 The KMT’s re-organization reflected Leninist features as: 1) party membership was only open to those who pledged that they would carry out party decisions; 2) a pyramidal party structure that parallels government hierarchy with the National Congress as the supreme organ and small cell groups act as base; 3) the formation of party groups in social organizations that asserted party policies, and finally, 4) the establishment of Control Committees in all the rank and file of the party organs to enforce discipline (Tai 1970, 409). These Leninist organizational25 features were key in fostering party cohesion during the KMT’s early rule in Taiwan. The KMT has a pyramidal party structure, parallel to the government hierarchy. At the top of the party, an elected National Congress serves as the supreme organ, which  22  For an excellent study of the organizational transformation of the KMT, see Dickson (1993) and Lin (1998, Chapter 3). For an in-depth analysis in Chinese, see Kung (1995). 23 For more on the aims and course of actions to be taken by the Central Committee for Party Reform Commission, see 7th Plenum of the KMT Central Executive Committee Party Report, 1952. 24 Also see Cheng and Haggard (1992, 4-8) on the implications of party reforms. 25 The KMT’s party organization exhibited many Leninist features because of early Soviet aid and training in the 1920s (Tai 1970, 409).  92  discharges its functions between sessions through the Central Standing Committee (CSC) the most powerful decision making body of the party. The National Congress meets once every 2 years and the delegates26 are selected to serve 4-year terms. Essentially, the Congress amends the party charter, determines the party platform and policies, elects the party chairman and the Central Committee members, and approves candidates nominated by the chairman to serve as vice chairmen and members of the Central Advisory Council. In 1952, the 7th National Congress was held (first time on the island) and renewed Chiang Kai-shek’s term as party leader. Besides, the Congress also elected a Central Committee (32 members) and a Central Advisory Committee (48 members). See Appendix E for the party organization structures of the KMT, 1952 and present. The Central Standing Committee (CSC) is the most influential organization in the KMT. It represents the CC when that body is not in session and meets every week to discuss and approve important policies and nominate candidates. Currently, the CSC consists of 39 members with a one-year term. Among these, 32 members are directly elected by the party delegates while 7 members are designated by the KMT chairman (including one for the Youth League and the other for Youth Affairs). As a “revolutionary democratic party”, the KMT adopted democratic centralism as a party orthodoxy where the party leaders held the lion’s share of decision-making power (Hood 1997, 26-8). Democratic centralism meant that tactical and ideological decisions were made with internal democracy and centralization. Once the top party leaders make a decision, all members must support and actively promote it. Decisions flowed from one central leader from the Central Committee downwards, all the way down to the party cell (Dickson 1996, 45). Along with these organizational changes, Chiang also upheld Sun Yat Sen’s Three Principles of the People27 (San-Min-Zhu-Yi) as the party’s official ideology that promoted a set of national values to build party loyalty. In the early years, power was highly centralized and anchored on the paramount leader. The 1950 re-organization created a structure based on a stable, homogenous and non-  26  The delegates were elected from all levels of the Party. See Article 23 of the KMT’s Party Constitution. This consists of 3 lectures which deals: 1) Principle of Nationalism; 2) Principle of Democracy; and 3) Principle of People’s Livelihood - the most important as it proposed a planned, mixed economy with the public controlling the major share, resembling British Fabianism (Tai 1970, 410). 27  93  competitive elite recruitment process that institutionalized the hegemonic rule of the KMT in society (Chu and Lin 2001, 114). Structurally, a hierarchy of party apparatus are dispersed throughout the state structures and grass roots organizations based on network of cells. The party cell28 functions as the basic, training unit of the party and penetrates into all state institutions and organized social groups such as schools, labour unions29, youth groups, religious groups, professional associations, business associations, famer’s associations, schools and mass media. Similar to the PAP’s “corporatist” strategy, the KMT built mass support by creating social organizations, subsidizing and inserting party cadres to represent their interests (Wang 2006, 110). As Cheng and Haggard note, the “pre-emptive corporatism” reduced the possibility for working-class politics to emerge (1992, 5). With the growth of sub-leaders, cadres and small party cells, the KMT became more organizationally complex and hierarchical. By the end of the 1952, there were a total of 34, 476 party cells, 6, 542 party sub-branches, 1,405 party branches and 56 party bureaus throughout the 23 counties and cities of Taiwan (Lin 1998, 75). Presently, the party cell remains the workhorse of the party. My interview with a KMT party staff reveals that there are an average of 50,000 party cells throughout Taiwan now (Interview with R. Chu, 11 Apr 2010). See Table 4-5. Table 4-5: KMT Party Branches and Cells (1952 and 2010) 1952 1984 2008 Party sub-bureaus 56 Party branches 1,405 Party sub-branches 6,542 34,476 14,498 48, 877 Party cells 小组 Party members 281,947 General population 8,128,000 Ratio of members to population 3.5%  2005  1,039,752  2010  50,000 1,040,000 (Est.) 23,127, 845 4.5%  Source: Data for 1952 from Lin (1998, 76), Taiwan National Statistics and the KMT Party Congress Report, various years.  Building Para-Political Institutions The creation of para-political organizations helped the party to boost the morale of a defeated party after the civil war. While the KMT was organized as a mass party, it became  28  Each party cell consists of about 3-11 people. All party members were required to attend cell meetings every other week and pay their membership fees regularly (Lin 1998, 75). 29 For the myriad of political and legal strategies to control labour movement, see Wang (1998, 250-74).  94  “a cadre party design to facilitate the exercise of power” (Tai 1970, 409). The creation of para-political organizations was what Winckler calls “institutional leadership” where leaders build institutions, assigns tasks and cultivate their own policies and personnel through them – an authoritarian form of social control to expand the KMT power base (1993, 113). Like the PAP, the KMT treated the mass organizations as “transmission belt of the party-state and remains an “overarching institution of interest aggregation” (Tien 1989, 14-5). In an effort to increase party membership, the KMT established a network of parapolitical organizations. In 1951, the KMT built “people’s service stations” (Min zhong fu wu zhan) throughout the island. Within a year, 234 stations were set up in every city and township under the provincial party office to provide social services to the public which include mandarin classes, reading classes and travelling advice (Wu 1987, 54). Organized under the Special Sector Party Office, party organs mushroomed all over the military, stateowned and public enterprises such as the aviation and shipping industry, railway and highway industries. See Table 4-6 for the degree of the KMT penetration into the public and grassroots organizations in 1951-2. By 1952, there were a total of 1,839 service station party cells installed in the various public enterprises, unified in 1954 (Wu 1987, 56). Table 4-6: Growth of KMT Service Stations (1951-2)30 1951 Party service stations at city and county level 31 Sub-party service stations at city and county level 304 Service station party cells 1,638 Total no. of party members 12,097 Source: KMT 7th Party Congress Report, 1952.  1952 35 325 1,839 15,007  Like the Soviet Communist Party, the KMT formed two “organizational weapons”31: the China Youth Corp32 (CYC) and the General Political Warfare Department (GPWD) to concentrate power in the hands of a small ruling elite and cultivate party loyalty. Then, Chiang Kai-shek tasked his son, Chiang Ching-kuo, to head the CYC and the GPWD to  30  These figures are sum total of service stations from the aviation, shipping; railway and highway industries from 1951-2. 31 Selznick (1960) calls “organizational weapon” - institutions for national integration and political socialization. Some subversive tactics include divide and rule, propaganda and agitation (Bullard 1997, 4-11). 32 The CYC was also known as China Youth Anti-Communist National Salvation Corps. The CYC began with 15 detachment units in the various universities32 and youth organizations in the early 1950s (Li 2005, 133). Structurally, the CYC became independent of the GPWD after 1969.  95  recruit and transform the mentality of youths and soldiers – 2 most important social groups in Taiwan.33 Both the CYC and the GPWD were auxiliary organizations to expand party membership, improve political socialization, training and mobilization of youths. Under Chiang’s supervision, the CYC and GPWD combined the Leninist revolutionary method and neo-Confucian traditional ethics to rear a group of followers who pledge their loyalty to the Party. And 2 of the pioneer CYC and GPWD members such as Li Huan and Wang Sheng34 were Chiang’s close aides and became influential leaders in the KMT in the 1970-80s. Established in 1952, the CYC provided military training to youths before they were drafted into the Nationalist Armed Forces. Taiwan has an average total of 150,000 youths conscripted each year. It was a captive audience to undergo political socialization (Bullard 1997, 81). The CYC offers a range of recreational and social activities such as summer field trips that became important channels for the KMT to recruit and co-opt talented and bright students to join the Party. Around 300 students were selected from the summer programs of the CYC to receive one year of military-based training annually. Under Chiang’s leadership, he introduced the harsh discipline that he’d experienced in Russia (Hsu 1993, 8).35 The stringent selection and the prestige in joining the CYC and military training helped to instil party loyalty (Interview with Chang, 26 Sep 2007).36 The GPWD, in coordination with the Ministry of National Defence and Education was responsible of training programs, the selection and training of political instructors for the military, civilian schools and colleges (Bullard 1997).37 The GPWD’s role was to instil in the military personnel an appropriate ideology and loyalty in times of conflict with China (Chang 1984, 434). Through the GPWD38, Chiang monitored the reliability of all military officers, gathered covert information of corrupt or unethical practices and maintained control over the  33  Taiwan has an estimated average total of 150,000 youths conscripted each year. It was clearly a captive audience to undergo political socialization (Bullard 1997, 81). 34 Wang Sheng was one of Chiang Ching kuo’s possible successors. See Chang (1984, 434). 35 Chiang went through considerable hardship during his twelve years in Russia. His training as a former graduate of Soviet Central Tolmatchev Military and Political Institute and life in Soviet Union had an impact on his populist leadership style and his organization of the CYC and the GPWD. See DVDs on Finding Chiang Ching-kuo, 2006; President Chiang Ching-kuo 2006; Chiang Ching-kuo and Chiang Kai-shek in Taiwan, 2007. 36 For more on the China Youth Corp, see Bullard (1997, 143). 37 The responsibility for training was transferred to the Ministry of Education in 1 July 1960. Structurally, the CYC became independent of the GPWD after 1969. 38 Officially, Chiang had no operational authority over the military commanders. But through the GPWD, Chiang was able to supervise the intelligence and paramilitary activities against the mainland.  96  military. As GPWD was not an operational combat unit, its intrusion into the different military services brought Chiang in direct conflict with career commanders (Interview with Shuai, 22 Nov 2007). Since Chiang’s passing in 1988, the youths’ interest in the CYC has declined drastically. As the country democratized, the government’s spending is more closely scrutinized and funding for the CYC is harder to come by. While the CYC remains active today, youths are showing less interest in the CYC as compared to the 1960s. Reports indicate that an estimated number of 130,000 young people participated in the CYC programs in 2006, a dismal performance compared to an average of one million registrations per year two decades ago (Chou, 2007).39 Since Ma Ying-Jeou became chairman in 2005, he had tried to revive the KMT youth corps and created “Young KMT” – an umbrella of nonpartisan youth organizations in high schools and universities to provide recreational and social activities to re-ignite the youths’ interest in the Party.40 However, party reports and interviews with party officials indicate that the interests in “Young KMT” remained lacklustre (Interview with R. Chu, 13 Apr 2009). Party Membership, Taiwanization and Co-optation There are no reliable figures on KMT’s membership in 1949. Early estimates show that the KMT had around 34,382 members when it first moved to Taiwan in 1949 (Wu 1987, 367). To address the problems of factionalism and leftist sympathizers, the KMT Reorganization Commission imposed stringent recruitment guidelines for cadres and party members. Like the PAP, the creation of the KMT’s cadre system in 1950 was a major achievement in the KMT’s reorganization. Selected from party activists, cadres are those who demonstrate leadership abilities and loyalty in non-party organizations. All cadres, regardless of their background had to be inspected.41  39  See “China Youth Corps adapts to post-Chiang times,” Taiwan Journal, 20 Jul 2007. For a critique of Ma’s efforts, see “Ma Youth corps set to produce dictators,” Taipei Times, 4 Feb 2006. 41 Inspections cover moral character, academic history, administrative accomplishments and family background. Some have to undergo examinations and appointed cadres have to demonstrate competence in skills appropriate to the posts (Dickson 1997, 56). 40  97  In 1951, 5,491 political cadres were inspected (Dickson 1997, 56). Regular inspection42 fosters cohesion as it filters out undesirable members of split loyalties. Party membership is open only to those who pledge to carry out party decisions, participate in party activities and pay regular dues. To qualify, one has to be endorsed by two party members and undergo two-months of training. In the early years, the KMT recruited almost exclusively from urban elites and overseas Chinese. Party members were mostly blue-collar workers and young males below forty years of age from the military, government sector or education backgrounds (Jiang and Wu 1992, 81). By early 1970s, the KMT had begun to put more emphasis on professionalism and expertise, as opposed to mobilization skills formerly valued by revolutionary movement. This was reflected in the recruitment policies. By 1969 to 1976, over 57% of all new recruits were students. Another 6% were from among public servants and teachers. This means that nearly two-thirds of new recruits were intellectuals. The largest occupational groups are government, party officials, soldiers and workers (Dickson 1996, 56-7). See Table 4-7. By 1970s, the efforts to recruit youths and intellectuals had resulted in a marked improvement in the educational levels of the party members. The number of recruits with high school and tertiary education has increased in the 1980s. Party reports show that by 1997, more than 85% of recruits were below the age of forty.  42  Article 10 of the Party Constitution stipulates that party membership rectification has to be conducted every 2 years and a membership inspection, every 4 years.  98  Table 4-7: Demographic Trends of the KMT Recruitment (1952-1997) Gender Male Female Total Recruits  1952 No.  %  1969-1976 No. %  1986 No.  %  1993 %  1994 %  1995 %  1997 %  40,097 2,514 42, 611  94.1 5.9 -  707,020 277,242 984,262  71.8 38.2 -  104,614 42,528 147,142  71.10 28.90 -  76.6 23.4 95,710  75,973  80.26 18.74 53,693  75.26 24.74 615,539  36,614 5,997  85.9 14.1  915,977 68,285  93.1 6.9  142,377 4,765  96.76 3.24  82.93 17.07  90.02 9.98  89.7 10.3  85.3 14.7  Age  Under 40 40 and over Occupation  6,533 15.3 42,345 4.3 2, 742 1.86 18.22 5.60 5.96 10.89 Farmer/Fishermen 9,980 23.4 4,407 6.5 6, 742 4.58 Workers 2,427 5.7 30,529 3.1 4, 558 3.10 3.56 4.78 Industry & Commerce 3,164 7.4 Politicians/C.serva nts 3,061 7.2 562,850 57.2 85,020 57.78 22.17 25.47 29.69 26.78 Students 6.8 58,850 6.0 8,146 5.54 47.04* 56.34 50.8 46.13 Teachers/Educators 2,900 5,824 13.7 181,622 18.5 34,561 23.49 Soldiers 0 4,690 0.5 71 0.05 Retired Soldiers 4895 11.5 7,163 0.7 1,697 1.15 Self-Employed 12,545 1.3 1,865 1.27 Housewife 3806 8.9 19,441 2.0 1,740 1.18 12.15 9.03 13.55 11.48 Others Source: Data for 1952 from 7th Party Congress Report, 1952; 1969-1976 from Dickson (1997, 128); 1986-1997 from Party Congress Reports, various years.  In the 1950s, the KMT was viewed as an “outsiders” party or émigré regime as more than 70% of its members were Mainlanders. From 1950-2, the total membership in KMT increased eight fold to 282, 959 members. However, out of this total population, only 26% (73, 852) were Taiwanese. See Appendix F for the ethnic breakdown of the KMT party membership (1952-2005). To address the internal inequalities, the KMT began Taiwanization - the deliberate strategy of placing more Taiwanese elites and local factions into the local party officers and central decision making bodies of the party. From 1952 to 1963, the proportion of Taiwanese was raised to 30%. After 1972, there was a significant jump in the number of Taiwanese, hitting 553, 215, nearly half of the total membership of 1,448,106. See Figure 4.1 for the surge of Taiwanese after 1963. By 1975, Taiwanese already constituted 53% of the total KMT membership. The total number of Taiwanese increased another 10% and exceeded 65% of total membership by 1986. Also see Appendix F. Taiwanization was Chiang’s strategy to reshape Taiwan’s political culture and control dissent by co-opting43 Taiwanese elites and local factions into the party (Dickson 1996). As Geddes says: “Factions  43  Co-optation refers to the “process of absorbing new elements into the leadership or policy-determining structure of an organization as a means of averting threats to its stability of existence” (Gamson 1968, 135).  99  form in single-party regimes around policy differences and competition for leadership positions, as they do in other kinds of regimes, but everyone is better off if all factions remain united and in office. This is why co-optation rather than exclusion characterizes established single-party regimes” (2003, 59). Figure 4-1: Ethnic Breakdown of the KMT Membership  At the local level, the KMT created a new position of vice-chairman in every city and county party committee and required that it to be filled by a “good local” party member, that is, a Taiwanese. Even though the vice-chairman was without significant power, the move showed that the KMT was trying to mollify the Taiwanese’ bad feelings by recruiting more Taiwanese into the party’s leadership stratum (Wu 1988, 67). At the highest level, Chiang surprised many conservative members when he nominated Lee Teng-hui, a low profile Taiwanese as Vice-President on 15 Feb 1984. The nomination of Lee as Chiang’s VP running mate was a critical juncture that gave Lee his political opportunity to reshape the KMT (Jacobs and Liu 2007, 380-2). Lee became President when Chiang passed away in 1988. After Lee assumed presidency, Taiwanization escalated, raising the total proportion to 69% by 1992. In the Executive Yuan44 (cabinet), under Lee’s presidency, the proportion of Taiwanese increased to 44%. See Table 4.8.  44  The Executive Yuan (Cabinet) is headed by a premier (appointed by the President) and consists of a vice premier; a number of ministers and chairpersons of commissions; and five to seven ministers.  100  Table 4-8: Ethnic Composition of Executive Yuan (1950-95) President Premiers 1950-1954 Chiang Kai-shek Chen Cheng 1954-8 Yu Hung-chun 1958-63 Chen Cheng 1963-72 Yen Chia-kan 1972-8 Yen Chia-kan Chiang Ching-kuo 1978-1984 Chiang Ching-kuo Sun Yun-hsuan 1984-9 Yu Kuo-hua 1989-1990 Lee Teng-hui Lee Huan 1990-3 Hau-Pei-ysun 1993-7 Lien Chan 1997-2000 Xiao Wan-zhang Source: Extracted from Huang (1995).  Taiwanese (%) 1 (50) 1 (5.3) 2 (7.7) 4 (11.1) 8 (30.8) 8 (27.6) 12 (34.3) 9 (45) 10 (47.6) 11 (44)  Mainlanders (%) 19 (95) 18 (94.7) 24 (92.3) 32 (88.9) 18 (69.2) 21 (72.4) 23 (65.7) 11 (55) 11 (52.4) 14 (56)  Soong argues that the KMT demonstrated its willingness to bring about changes to accommodate diverse views (1992, 65). But critics contend that the composition of the staff of the Party HQ and the CCs still reflected the discrimination and Taiwanization only occurred in the lower levels (Huang 1995, 102; Wu 1987, 68). But the KMT did recruit more Taiwanese into the higher party organizational levels. While there were no Taiwanese in the CSC from 1950 to 1952, from 1972 to 1981, the number of Taiwanese in CSC doubled from 14.3% to 33%. See Table 4.9. By July 1988, Lee Teng-hui had named 16 Taiwanese out of 31 members CSC. Under Lee, Taiwanization accelerated and more Taiwanese were appointed to CSC. For the CC, the proportion of Taiwanese was raised to 25% for the same period. See Table 4.9. Table 4-9: Ethnic Composition of CSC and CC (1952-1994)45 Chiangs’ Era 1952 1957 1960 1967 1972 1976 1981 1984 CSC Total No. 10 16 16 19 21 22 27 31 Taiwanese 0 1 2 2 3 5 9 12 % 0 8.3 12.5 10.5 14.3 22.7 33 39 CC Total No. 32 50 74 74 99 130 150 Taiwanese 1 3 4 4 6 19 25 % 3.1 6.0 5.4 5.4 6.1 14.6 19.3 Source: Extracted from Huang (1995). Years  45  Lee’s Era 1988 1993  1994  31 16 51.6  35 20 57.1  35 21 60  180 62 34.4  210 112 53.3  The KMT stopped the practice of keeping official records of the members’ background after 1992.  101  Implications of Taiwanization As the KMT 15th party congress reports: “da y da de nan chu”, which roughly translates to “big parties come with their own big problems”. While the 1952 re-organization increased the party’s organizational complexity, hierarchical structure, the membership expansion had strained the relationship between the old and new members. As Panebianco reminds us, a principal cause of intra-party conflicts lies in the party’s internal system of inequalities (1988, 4 and 43). It is argued that the KMT’s early discriminatory policy has led to: 1) the antagonism between the two sub-ethnic groups that translated into polarizing partisan alignment, 2) contradictory and destabilizing internal tensions which undermined the KMT’s