GOVERNMENT AND OPPOSITION AN INTERNATIONAL JOURNAL OF COMPARATIVE POLITICS Government and Opposition, Vol. 45, No. 4, pp. 505-530, 2010 doi:10.1111/j.l477-7053.2010.01325.x Thomas Sedelius and Joakim Ekman Intra-executive Conflict and Cabinet Instability: Effects of Semi-presidentialism in Central and Eastern Europe THE CONSTITUTIONAL FRAMEWORK IN TRANSITIONAL COUNTRIES IS A terrain on which political incumbents struggle to expand and define their influence. Under semi-presidentialism - with two separately chosen chief executives - this struggle is particularly manifested in conflicts between presidents and prime ministers.1 Experiences of semi-presidentialism in Central and Eastern Europe have resulted in numerous disputes between presidents and prime ministers, where some of the most salient examples are the stalemating conflicts between President Lech Walesa and several prime ministers in Poland 1991-95, between President Leonid Kuchma and several prime ministers in Ukraine 1994-2004, as well as the critical clashes between President Viktor Yuschenko and Prime Minister Viktor Yanukovich in Ukraine in 2006-07. In this article we ask to what extent and in what ways intra-executive conflicts increase the risk of pre-term termination of governments. Comparing eight semi-presidential systems in Central and Eastern Europe - Bulgaria, Croatia, Lithuania, Moldova, Poland, Romania, Ukraine and Russia, comprising a total of 65 instances of intra-executive (president-cabinet) coexistence between 1991 and 2007 - we examine the link between intra-executive conflict and cabinet instability. This sample represents a deliberate attempt to cover transitional countries with different forms of semi-presidential regimes (premier-presidential and president-parliamentary). Also, in the period under 1 Cf. T. Baylis, 'Presidents Versus Prime Ministers: Shaping Executive Authority in Eastern Europe', World Politics, 48: 3 (1996), pp. 297-323; R Elgie, 'The Perils of Semi-Presidentialism: Are They Exaggerated?', Democratization, 15: 1 (2008), pp. 49-66. © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd Published by Blackwell Publishing, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Maiden, MA 02148, USA. 506 GOVERNMENT AND OPPOSITION review, all these countries have experienced an uncertain transitional phase, making them especially appropriate for the kind of analysis we have in mind. In transitional countries, effects conventionally associated with cabinet instability - disruptive policy making, political unpredictability and lack of political accountability - are generally considered undesirable, as potential threats to the fragile democratization process. The article is thus related to the rich body of literature on institutional design and democracy performance.2 More specifically, it contributes to the more limited albeit growing research on the pros and cons of semi-presidentialism.3 Proponents of semi-presidentialism 2 Perhaps most salient in this literature is the longstanding presidentialism vs parliamentarism debate, e.g. J. M. Colomer and G. L. Negretto, 'Can Presidentialism Work Like Parliamentarism?', Government and Opposition 40: 1 (2005), pp. 60-89; J. A. Cheibub, Presidentialism, Parliamentarism and Democracy, Cambridge, Cambridge University Press, 2007; K. von Mettenheim (ed.), Presidential Institutions and Democratic Politics: Comparing Regional and National Contexts, Baltimore, Johns Hopkins University Press, 1997; S. Mainwaring and M. S. Shugart (eds), Presidentialism and Democracy in Latin America. Cambridge: Cambridge University Press, 1997; T. Power and M. Gasiorowski, 'Institutional Design and Democratic Consolidation in the Third World', Comparative Political Studies, 30: 2, (1997), pp. 123-55; J. J. Linz, 'The Perils of Presidentialism', Journal of Democracy, 1: 1 (1990), pp. 51-69; J. J. Linz, 'Presidential or Parliamentary Democracy: Does it Make a Difference?', in J. J. Linz and A. Valenzuela (eds), The Failure of Presidential Democracy: The Case of Latin America, Volume 2, Baltimore, Johns Hopkins University Press, 1994; A. Stepan and C. Skach, 'Constitutional Framework and Democratic Consolidation: Parliamentarianism Versus Presidentialism', World Politics, 46: 1 (1993), pp. 1-22. 3 E.g. R. Elgie, 'Semi-Presidentialism, Cohabitation, and the Collapse of Electoral Democracies 1990-2008', Government and Opposition, 45: 1 (2010), pp. 29-49; P. Schle-iter and E. Morganjones, 'Citizens, Presidents and Assemblies: The Study of Semi-Presidentialism beyond Duverger and Linz', British Journal of Political Science, 39: 4 (2009), pp. 871-92; R. Elgie and S. Moestrup (eds), Semi-Presidentialism in Central and Eastern Europe, Manchester, Manchester University Press, 2008; R. Elgie and S. Moestrup (eds), Semi-Presidentialism Outside Europe, London, Routledge, 2007; O. A. Neto and M. C. Lobo, 'Portugal's Semi-Presidentialism (Re) Considered: An Assessment of the President's Role in the Policy Process, 1976-2006', European Journal of Political Research, 48 (2009), pp. 234-55; M. Tavits, Presidents with Prime Ministers: Do Direct Elections Matter?, Oxford, Oxford University Press, 2008; T. Sedelius, The Tug-of-War Between Presidents and Prime Ministers: Semi-Presidentialism in Central and Eastern Europe, Saarbrücken, VDM Verlag, 2008; M. S. Shugart, 'Semi-Presidential Systems: Dual Executive and Mixed Authority Patterns', French Politics, 3: 3 (2005), pp. 323-51; C. Skach, Borrowing Constitutional Design: Constitutional Law in Weimar Germany and the French Fifth Republic, Princeton, Princeton University Press, 2005. There are a number © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd INTRA-EXECUTIVE CONFLICT AND CABINET INSTABILITY 507 have declared it to be an advantageous constitutional arrangement for new democracies4 while sceptics have associated it with institutional conflict and political risks.5 The few existing previous studies of intra-executive conflicts in semi-presidential regimes have mainly been occupied with explaining why such conflicts occur in the first place. In particular, Oleh Protsyk has analysed the coexistence of popularly elected presidents and prime ministers in semi-presidential regimes, focusing on what specific institutional features excerbate intra-executive conflicts.6 Still, there are no comparative and systematic analyses of how intra-executive conflicts are actually related to political outcomes in semi-presidential regimes. By empirically exploring the link between intra-executive conflict and cabinet duration, this article contributes to our understanding of semi-presidentialism. The analysis is partly based on data from an expert survey, conducted in 2002-5 among scholars and officials with expertise on executive-legislative issues in the countries under examination. The article also utilizes a number of indicators of constitutional practices (1991-2007) to assess the conflict level in the eight countries under review. of older contributions that have highly influenced more recent analyses of semi-presidentialism, e.g. G. Sartori, Comparative Constitutional Engineering: An Inquiry into Structures, Incentives and Outcomes, 2nd edn, London, Macmillan Press, 1997; M. S. Shugart and J. M. Carey, Presidents and Assemblies: Constitutional Design and Electoral Dynamics, Cambridge, Cambridge University Press, 1992; M. Duverger, 'A New Political System Model: Semi-Presidential Government', European Journal of Political Research, 8 (1980), pp. 165-87. 4 Cf. M. Duverger, 'Reflections: The Political System of the European Union', European Journal of Political Research, 31 (1997), pp. 137-46; Sartori, Comparative Constitutional Engineering. 5 Cf. S. Fabbrini, 'Presidents, Parliaments, and Good Government', Journal of Democracy, 6: 3 (1995), pp. 128-38; A. Lijphart, 'Constitutional Design for Divided Societies'', Journal of'Democracy, 15: 2 (2004), pp. 96-109; J. J. Linz, 'Introduction: Some Thoughts on Presidentialism in Post-Communist Europe', in R. Taras (ed.), Postcom-munist Presidents, Cambridge, Cambridge University Press, 1997. 6 O. Protsyk, 'Politics of Intra-executive Conflict in Semi-presidential Regimes in Eastern Europe', East European Politics and Society, 18: 2 (2005), pp. 1-20; O. Protsyk, 'Intra-executive Competition Between President and Prime Minister: Patterns of Institutional Conflict and Cooperation in Semi-Presidential Regimes' Political Studies, 56: 2 (2006), pp. 219-41. © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd 508 GOVERNMENT AND OPPOSITION DEFINITIONS OF SEMI-PRESIDENTIALISM AND THE FOCUS ON CENTRAL AND EASTERN EUROPE To define semi-presidentialism and account for crucial differences between semi-presidential regimes, we adhere to the commonly used distinction between premier-presidential and president-parliamentary systems originally suggested by Matthew Shugart and John Carey.7 The criteria are as follows: under premier-presidentialism (1) the president is elected by a popular vote for a fixed term in office; (2) the president has considerable constitutional authority; (3) there also exist a prime minister and cabinet, subject to the confidence of the assembly majority. In president-parliamentary systems (1) the president is elected by a popular vote for a fixed term in office; (2) the president appoints and dismisses the prime minister and other cabinet ministers; (3) the prime minister and cabinet ministers are subjected to both parliamentary and presidential confidence; and (4) the president typically has some legislative powers and the power to dissolve the parliament.8 In this study, the premier-presidential cases are Bulgaria 1991-2007, Croatia 2000-7,9 Lithuania 1991-2007, Moldova 1991-2000,10 Poland 1991-2007, Romania 1991-2007 and Ukraine 2006-7.11 The president-parliamentary cases are Croatia 1992-2000, Russia 1991-2007 and Ukraine 1991-2006. 7 Shugart and Carey, Presidents and Assemblies. 8 Duverger, European Journal of Political Research, p. 166; Sartori, Comparative Constitutional Engineering, p. 131; Shugart and Carey, Presidents and Assemblies, pp. 23-4; Shugart, 'Semi-presidential Systems: Dual Executive and Mixed Authority Patterns', pp.324-35. 9 In 2000-1, the Croatian constitution was revised from a president-parliamentary to a premier-presidential type of system. Since then, the government is responsible to the parliament only, and not, as the case was under the 1990 constitution, to both the president and the parliament. 10 The Moldovan parliament amended the 1994 constitution in 2000, envisioning a shift away from premier-presidentialism to parliamentarism by changing from direct to indirect presidential elections. Thus, since then, the president is elected by the parliament. 11 In the aftermath of the Orange Revolution in 2004, agreements were reached to reduce the president's power and to subordinate the government to the parliament, which de facto entailed a change from a president-parliamentary to a premier-presidential system. © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd INTRA-EXECUTIVE CONFLICT AND CABINET INSTABILITY 509 There are good reasons for focusing exclusively on post-communist countries in this study. Semi-presidentialism has become an extremely popular form of government in the third wave of democratization and has emerged as the most common regime type in Central and Eastern Europe. Moreover, analysing different forms of semi-presidentialism in transitional societies adds to our understanding of the potential risks and advantages associated with this particular regime type. A possible objection would be that our sample entails an examination of countries that are both democratic and less than fully democratic and, at the same time, covers different types of semi-presidentialism; and that the empirical results of our investigation may partly be attributed to the fact that some of our countries today are at best semi-democracies or hybrid regimes (e.g. Russia).12 Here, however, the issue of different levels of democratization is considered to be only of secondary importance. Our sample is exclusively made up of transitional countries, and the important thing is to investigate the pros and cons of both types of semi-presidentialism under such conditions. SEMI-PRESIDENTIALISM, INTRA-EXECUTIVE CONFLICT AND CABINET INSTABILITY How are the two types of semi-presidential arrangements related to intra-executive conflict? In premier-presidential systems, underlying executive-legislative antagonism between president and parliament often appears in the form of intra-executive conflict between the president and the cabinet. Since premier-presidentialism provides the legislature with the exclusive power of prime minister dismissal, the cabinet is primarily dependent on parliamentary support for claiming authority to control the executive branch, and its political orientation is likely to be in the parliament's favour rather than in the president's. This is particularly the case if there is a stable and coherent majority in parliament. Even if the cabinet initially emanates from a compromise outcome of strategic interactions between the president and the parliament, it is likely to drift during its 12 Cf. J. Ekman, 'Political Participation and Regime Stability: A Framework for Analyzing Hybrid Regimes', International Political Science Review, 30: 1 (2009), pp. 7-31. © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd 510 GOVERNMENT AND OPPOSITION tenure in office towards the ideal point of the parliament. Intra-executive conflicts are thus to be expected.13 In president-parliamentary systems the institutional lines of conflict are more uncertain. Here, the constitutional framework provides both the president and the parliament unilateral power to dismiss the cabinet, which makes the distribution of dismissal powers a somewhat less effective predictor of the likely alliances and possible lines of conflict. Generally though, in president-parliamentary systems the president has an overall stronger position in the cabinet formation process than is the case in premier-presidential systems. Typically, the president has the power to nominate a prime minister candidate without consulting or having recommendations from the parliament. Thus, the likelihood of a 'united' executive without intra-executive conflicts is generally higher in a president-parliamentary regime than in a premier-presidential one.14 Still, intra-executive conflict has been a quite frequently occurring phenomenon in some of the contemporary president-parliamentary systems as well. Our main theoretical expectation is that intra-executive conflicts have a negative impact on the durability of cabinets, especially in a context of transition. As already noted, the effects of intra-executive conflict have so far been largely neglected in the research on semi-presidentialism. The literature on cabinet durability and termination, however, is quite considerable15 although comparative researchers have mainly focused on parliamentarism and presiden-tialism, rarely on semi-presidential systems16 (or on the institutional variation within semi-presidentialism). This may also explain why previous studies of cabinet survival have not systematically included intra-executive conflict as an explanatory variable. Here the idea is to add to our knowledge by providing actual empirical substantiation of such instances of cabinet instability. What are the theoretical arguments underpinning our hypothesized relationship between the level of intra-executive conflict and cabinet durability? In president-parliamentary systems the logic 13 Cf. Shugart and Carey, Presidents and Assemblies. 14 Cf. ibid. 15 For quite extensive reviews of this literature see B. Grofman and P. van Roozendaal, 'Review Article: Modelling Cabinet Durability and Termination', British Journal of Political Science, 27 (1997), pp. 419-51; M. Taver, 'Government Termination', Annual Review of Political Science, 6 (2003), pp. 23-40. 16 Shugart, French Politics, p. 345. © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd INTRA-EXECUTIVE CONFLICT AND CABINET INSTABILITY 511 behind this is quite straightforward. Since the president in these systems has unilateral power to dismiss the prime minister, cabinet resignation is to be expected under intra-executive conflict. When challenged by a rebellious prime minister, the president may quite easily use the threat of dismissal in order to sub-ordinate the prime minister under his firm control. If the prime minister still does not comply, the president has the choice to go all the way and replace the prime minister. In premier-presidential systems, by contrast, authority to dismiss the cabinet rests exclusively with the parliamentary majority. Consequently, the prime minister can be expected to stand up to the president and survive any intra-executive conflict as long as the parliament is supportive. Actually, this factor has been put forth as an inherent advantageous mechanism of premier-presidentialism, i.e. that the system can still function through a prime minister supported by a parliamentary majority.17 Thus, given the formal distribution of power, we expect cabinet turnover to be relatively more frequent under intra-executive conflict in president-parliamentary systems than in premier-presidential systems. Even so, we can identify informal norms and practices related to the dual executive structure of premier-presidentialism that may enhance cabinet instability under intra-executive conflict. Both the president and the prime minister can claim their legitimacy on popular elections, but the former leans on a direct electoral mandate while the latter is 'only' indirectly elected through parliamentary elections. Typically in premier-presidential systems, the relatively limited constitutional powers provided to the presidency do not correspond to the prestigious and popular legitimacy upheld by its incumbent. But the president can exert strong influence on the cabinet by leaning on his popular mandate and - as is particularly salient in a transitional context - his greater popularity as compared to the other branches. Under intra-executive conflict a rewarding strategy to the president is thus to 'go public' and criticize the government with negative effects on the cabinet's credibility.18 This may ultimately force the prime minister out of office despite the fact that the president lacks formal dismissal powers. If, in addition, the 17 Sartori, Comparative Constitutional Engineering, pp. 153-4. 18 See R. Elgie, 'Cohabitation: Divided Government French-Style', in R. Elgie (ed.), Divided Government in Comparative Perspective, Oxford, Oxford University Press. © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd 512 GOVERNMENT AND OPPOSITION president also possesses informal partisan power this affords him extra-constitutional powers to censor the cabinet.19 In a transitional context where the institutionalization process is still incomplete, the importance of personal authority and informal powers is even more profound and the president's opportunities to undermine the survival of cabinets thus more favourable. To better illustrate presidential influence over prime ministers in premier-presidential systems, we will explore some case study examples in the final sections of the article. First, however, we need to provide empirical support for the hypothesized link between intra-executive conflict and cabinet instability in semi-presidential regimes. DATA AND MEASUREMENT ISSUES In order to collect information on how post-communist semi-presidential systems have worked in practice, a total number of 50 country experts on constitutional and political issues were included in an expert survey conducted in 2002-5.20 In addition to the expert 19 Cf. D.J. Samuels and M. S. Shugart, Presidents, Parties, and Prime Ministers: How the Separation of Powers Affects Party Organization and Behavior, New York, Cambridge University Press, 2010. 20 The expert survey consisted of two main parts, conducted in two steps: (i) a questionnaire including 29 questions covering a broad set of issues concerning executive-legislative relations, presidential powers and party system factors; and (ii) a number of interviews and/or e-mail questions in which the specific issue of intra-executive and executive-legislative conflicts were the main topic. While the questionnaire was designed rather broadly and inductively, the follow-up interviews were more focused and delimited in scope. The questionnaire was sent to 8-15 experts in each of the countries and contained questions with both pre-defined and open-ended answers. The interviews were undertaken in a semi-structured manner under which questions relating to intra-executive and executive-legislative relations were asked. Tracking the appropriate kind of 'experts' to be included in the survey was not an entirely straightforward task. We started from a list of social scientists with expertise in political and constitutional issues in the countries under examination, and then added a number of potential respondents by screening the literature and by employing a snowball strategy, asking the first round of respondents to suggest other potential experts. An overall response rate of about 50 per cent was reached. All in all, 50 different experts finally took part in the survey (36 academic scholars and 12 high officials, e.g. judges and constitutional court officials). The number of experts for each country was: Bulgaria 8, Croatia 3, Lithuania 12, Moldova 3, Poland 16, © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd INTRA-EXECUTIVE CONFLICT AND CABINET INSTABILITY 513 survey data, we also draw on conventional country reports and up-dates as well as previous research, in order to determine the level, character and outcome of institutional conflicts. Country reports from the East European Constitutional Review (EECR),21 Radio Free Europe/Radio Liberty (RFE/RL)22 and Freedom House's Nations in Transit (NiT)23 are among the sources used. Furthermore, a number of case-study volumes on the post-communist countries have been particularly useful.24 In this article, the term intra-executive refers to the relation between the president and the cabinet. Intra-executive conflict is generally understood here as struggles between the president and the prime minister/cabinet over the control of the executive branch. More specifically, and in order to reach an operational definition, the relationship between the president and the cabinet has been considered as conflict ridden when there has been an observable clash between the president and the prime minister and/or between the president and other government ministers, manifested through obstructive or antagonistic behaviour from either side, directed towards the other. Conflicts can thus be observed in concrete disputes, such as in public statements where criticism is levelled against the other side, in disagreements over key appointments or dismissals, in different interpretations of constitutional prerogatives, in interference in each other's political domains, or finally, in personal disputes or in strong disagreement over policy directions. By this definition, we encompass a relatively broad set of conflicts. Admittedly, some conflicts are more Romania 2, Russia 4, and Ukraine 2. Among the researchers, the majority are senior scholars of political science, constitutional law or sociology. Most of them are living and working in the country of which they have been contacted as experts. 21 East European Constitutional Review, 'Constitution Watch', 6: 2; 6: 3; 6: 4; 7: 4; 8: 3; 9:1/2; 11: 3; 12: 2 (1997-2003), www.law.nyu.edu/eecr. 22 Radio Free Europe/Radio Liberty Newsline, country reports, (2000-8), www.rfer-l.org. 23 Nations in Transit, Freedom House (2002-07), www.freedomhouse.org. 24 E.g. S. Berglund, J. Ekman and F. H. Aarebrot (eds), The Handbook of Political Change in Eastern Europe, 2nd edn, Cheltenham, Edward Elgar, 2004; J. Blondel and F. Muller-Rommel (eds), Cabinets in Eastern Europe, London, Palgrave, 2001; R. Elgie (ed.), Semi-Presidentialism in Europe, Oxford, Oxford University Press, 1999; Elgie and Moestrup, Semi-Presidentialism in Central and Eastern Europe, R. Taras (ed.), Postcommu-nist Presidents, Cambridge, Cambridge University Press, 1997; J. Zielonka (ed.), Democratic Consolidation in Eastern Europe: Volume 1, Institutional Engineering, Oxford, Oxford University Press, 2001. © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd 514 GOVERNMENT AND OPPOSITION critical than others for the overall stability of the political system, but the manifestations of these critical conflicts can still be different from case to case, and we have accordingly chosen to capture as many kinds of conflict as possible at the outset. Providing accurate judgements about the level of conflict between institutions across a wide range of countries is of course a tricky task. What one observer might consider an open fight between two branches of government may be just a small bickering in another viewer's eyes. Furthermore, an institutional conflict might be latent or covert and may therefore not reach the wider society. Such conflicts are not easily accessible and cannot be uncovered by, for example, text analysis or by examining public statements. Our strategy has been to capture only manifest and observable conflicts. Based on the results from the expert survey, from reports as well as from previous research, the level of conflict is compressed into ordinal estimations. In some recent works, Protsyk uses a simple dichotomy (low and high) to determine the level of conflict in a number of semi-presidential systems.25 Here, we have used a similar strategy, but initially coded on an ordinal scale: low, medium and high conflict. When no significant conflict between the president and the cabinet has been reported, the relationship has been considered as non-conflictive and hence the conflict-level estimated as low. Under periods with episodic but manifest and observable conflict, the level has been estimated as medium. Finally, instances of durable and severe tension between the president and the cabinet have been estimated as high levels of conflict. In some cases we have also consulted experts (from the expert survey) for their opinion on the final coding. Given that the number of country experts for some of the countries was insufficiently small, all estimations are backed up by secondary data sources. As a rule we have recorded an observation of conflict only if it could be substantiated by at least two independent sources. Whenever it has been too difficult to obtain a clear picture of the institutional relations, we have abstained from making a definitive assessment. This has particularly been the case for very short-lived 25 Protsyk, 'Politics of Intra-Executive Conflict in Semi-Presidential Regimes in Eastern Europe'; Protsyk, 'Intra-Executive Competition between President and Prime Minister'. © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd INTRA-EXECUTIVE CONFLICT AND CABINET INSTABILITY 515 cabinets with an interim character, i.e. caretaker governments that have held office in a limited period after the dismissal of the old cabinet up until new parliamentary elections have been completed. For the purpose of reliability a few case examples could demonstrate how the coding decisions were made. The relation between President Walesa and Prime Minister Pawlak in Poland in 1994—95 is an illustrative example of a quite straightforward high-conflict case. Here, severe tension was manifested by several different attributes, such as negative public addresses and a long-lasting institutional tug-of-war (e.g. frequent remittances of laws to the constitutional court, presidential vetoes, disagreements over powers and prerogatives) .26 The intensity of this conflict has been well documented by both expert accounts and secondary literature,27 making us quite confident in coding the intra-executive coexistence as 'highly' con-flictive. In some other instances though, it was less obvious how to make a distinction, especially when the level of conflict shifted considerably over the cabinet period. In such cases we decided to code within a simple range, i.e. as low-medium or medium-high. The coexistence between President Snegur and Prime Minister Sangheli in Moldova in 1994-96 is a case in point. Various sources reported a rather compliant and cooperative intra-executive relationship in the initial year, but a gradual deterioration in the following year, resulting in an open conflict in 1996 under which the president publicly accused the cabinet of incompetence and called on a parliamentary vote of no-confidence.28 Accordingly, this case was coded as a 'medium-high' conflict. It should be noted that in our subsequent analysis the ordinal scale described above is collapsed into a high-low dichotomy where medium, medium-high or high levels of intra-executive conflict For a more extensive description of this conflict, see Sedelius, The Tug-of-War between Presidents and Prime Ministers, pp. 136-8. 27 E.g. K. Jasiewicz, 'Poland: Walesa's Legacy to the Presidency', in R. Taras (ed.), Postcommunist Presidents, Cambridge, Cambridge University Press, 1997, pp. 152-4; A. Krok-Paszkowska, 'Poland', in Elgie, Semi-Presidentialism in Europe, pp. 182-5. 28 Cf. RFE/RL Newsline, various reports on Moldova, 1994-96. For reports on the escalating conflict, see RFE/RL Newsline, March^July 1996, www.rferl.org; cf. W. Crowther and Y.Josanu 'Moldova', in Berglund et al. The Handbook of Political Change in Eastern Europe, pp. 557-63. © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd 516 GOVERNMENT AND OPPOSITION qualify as 'high', and low and low-medium count as 'low'. The findings of our exploration of the levels of intra-executive conflict 1991— 2007 are presented in the Appendix. AN EMPIRICAL ASSESSMENT OF INTRA-EXECUTIVE CONFLICT AND CABINET STABILITY We proceed with an empirical test of our hypothesis about intra-executive conflict and cabinet survival in the semi-presidential systems of Central and Eastern Europe. Following a standard definition in the literature, we have considered a cabinet terminated once a new cabinet formation is necessary (including when new elections have been held), i.e. it does not matter whether or not the cabinet formation results in a reinstatement of the same cabinet. What is important is that new negotiations are necessary before the cabinet can be reconstituted.29 A rather straightforward and commonly used measure of duration is the number of days, weeks, months or years the cabinet remains in office after inauguration.30 Table 1 displays the length of cabinet survival - as measured by months in office, under low and high levels of intra-executive conflict - in the two varieties of semi-presidentialism in post-communist Europe, i.e. premier-presidential systems and president-parliamentary systems. The measure is in whole months from the starting month of the formal investiture. The results in Table 1 are clearly in favour of our hypothesis. Under low levels of intra-executive conflict (see Appendix), the average length of cabinet survival is 24.7 months. Under high levels of conflict, the average cabinet duration is some 17 months. Looking at the outcome for the two semi-presidential categories in Table 1, we find - quite expectedly - that the conflict level has indeed a significant impact on cabinet survival in the president-parliamentary category. Here, high levels of intra-executive conflict have apparently shortened the average cabinet survival time by more than one year (some 15 months), as compared to cabinet survival under less 29 Cf. Grofman and van Roozendaal, 'Review Article: Modelling Cabinet Durability and Termination'. 30 Cf. K. Strom, Minority Government and Majority Rule, Cambridge, Cambridge University Press, 1990; J. Blondel, World Leaders: Heads of Governments in the Postwar Period, London, Sage, 1980. © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd INTRA-EXECUTIVE CONFLICT AND CABINET INSTABILITY 517 Table 1 Cabinet Duration Under Low and High Levels of Intra-executive Conflict (average months) 1991-2007 Semi-presidential type Low level of conflict High level of conflict (n) (n) Both 24.7 17.4 (N= 54) (32) (22) Premier-presidential 24.1 20.5 (N= 34) (19) (15) President-parliamentary 25.6 10.9 (N= 20) (13) (7) Note: The total number of cabinets (54) is lower as compared to the 65 cases of intra-executive coexistence documented in Table Al. This is due to the fact that these 65 cases of intra-executive coexistence do not correspond to 65 new cabinets. For example, when a newly or re-elected president faced a cabinet formed prior to the presidential elections, this was classified as a new case of intra-executive coexistence. However, such cases do not qualify as cabinet shifts, which is why the total number of cabinets/ cases in this calculation decreases somewhat. Source: See Appendix, Table Al. conflict-ridden periods. But there is an observable effect also in the premier-presidential category. Under low levels of conflict, the average cabinet duration is just over 24 months. Under high levels of conflict, the corresponding period is 20.5 months. However, the figures in Table 1 do not tell us anything about the reasons for the cabinet shifts. It is therefore necessary to try to establish whether or not intra-executive conflicts as such have forced cabinets to leave office before regular elections. Table 2 displays the mode of cabinet resignation, by distinguishing between cabinet changes brought about by ordinary parliamentary elections, and cabinet changes brought about by pre-term resignation.31 The results are quite conclusive: under high levels of intra-executive conflict, pre-term resignation of the cabinets has occurred in 91 per cent of the cases. Under low levels of conflict, the corresponding figure is 44 per cent. Furthermore, the correlation between (high levels of) intra-executive conflict and the mode of cabinet resignation proves to be 31 G. King, J. E. Alt, N. E. Burns and M. Laver, 'A Unified Model of Cabinet Dissolution in Parliamentary Democracies', American Journal of Political Science, 34 (1990), pp. 846-71. © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd 518 GOVERNMENT AND OPPOSITION Table 2 Mode of Cabinet Resignation and Intra-executive Conflict (%) 1991-2007 Mode of cabinet resignation Intra-executive Elections Pre-term Difference in conflict (n) (n) percentage units _(rO_ 56 44 +12 (18) (14) (32) 9 91 -82 (2) (20) (22) 75 25 +50 (15) (5) (20) 13 87 -74 (2) (13) (15) 25 75 -50 (3) O) (12) 0 100 -100 (0) (7) (7)_ Note: The mode of cabinet resignation as measured in terms of whether the cabinets have stayed in office until regular parliamentary elections ('Elections') or if they have failed to survive their whole term ('Pre-term'). Source: See Appendix, Table Al. strong in both semi-presidential categories. While the presidents in premier-presidential systems do not have the formal powers to dismiss the cabinet, they have apparently used other means of influencing the destiny of governments. We may also note from Table 2 that pre-term dismissal under high levels of intra-executive conflict has been a rule without exception in the president-parliamentary systems. Recalling our tentative assumption about cabinet survival in president-parliamentary systems, we may thus conclude that these results provide empirical backing to this assumption - under high levels of intra-executive conflict, the president is likely to dismiss the prime minister. Type of semi-presidentialism Both Low (N= 54) High Premier-presidential Low (N=35) High President-parliamentary Low (N= 19) High ALTERNATIVE EXPLANATIONS? There is of course reason to assume that other variables are of significance with regard to the mode of cabinet resignation as well. For one thing, conventional wisdom suggests that minority governments © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd INTRA-EXECUTIVE CONFLICT AND CABINET INSTABILITY 519 are less durable than majority governments. Previous research has also identified the level of party system fragmentation as a determinant of cabinet survival.33 A high level of party system fragmentation is likely to increase the difficulties for cabinets to keep parliamentary support and thus to endure a full term in office. Similarly, we would probably expect cabinet turnover to be greater, all else being equal, during the early years of transition. Our initial correlation tests (not reported here) also suggest that cabinet type, party system fragmentation and post-communist period may be significant. Thus, we cannot yet determine if there really is a single and significant effect of intra-executive conflict on the mode of cabinet resignation.34 Since we are dealing with a relatively limited number of cases, the use of more advanced statistical techniques is somewhat precarious. Nevertheless, with the caveat that the results should be interpreted with some caution, we will undertake a logistic regression test of the assumption that intra-executive conflict has a significant overall effect on cabinet survival. Since the 'mode of cabinet resignation' variable is dichotomous, we cannot apply (OLS) linear regression in this case. Instead, we will have to use logistic regression, which has the advantage of making the form of the relationship linear whilst treating the relationship itself as non-linear. In Table 3, the regression coefficients are reported with the mode of cabinet resignation as dependent variable (0 = elections, 1 = pre-term) and with five independent variables: intra-executive conflict (0 = low, 1 = high), type of semi-presidentialism (0 = premier-presidentialism, 1 = president-parliamentarism), government form (0 = majority, 1 = minority), the 32 E.g. King et al., 'A Unified Model of Cabinet Dissolution in Parliamentary Democracies'; Strom, Minority Government and Majority Rule. 33 King et al., 'A Unified Model of Cabinet Dissolution in Parliamentary Democracies'; D. Sanders and H. Valentine 'The Stability and Survival of Governments in Western Democracies', Acta Politica, 3 (1977), pp. 346-77. 34 Yet another factor of relevance to premier-presidentialism is cohabitation, i.e. where the president's party is not represented in the cabinet and where the prime minister and president are from different parties. It is logical to expect that the president under such periods has less to lose by criticizing the government, with possible pre-term resignation of government as a consequence. Among our cases we identified cohabitation in only eight relevant instances (in Bulgaria, Lithuania and Poland) and from this very small sample we found no support for the assumption that cohabitation would increase cabinet instability (in four instances the government endured its full term, and in the remaining four instances it stepped down earlier). © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd 520 GOVERNMENT AND OPPOSITION Table 3 Logistic Regression Test (mode of cabinet resignation) B Intra-executive conflict 3.125 Semi-presidential type 1.899 Government form 1.046 Party system fragmentation -.156 Post-communist period -.518 Nagelkerke R-square: .530 Standard error Wald Sig. Exp (B) .951 10.794 .001 22.761 .892 4.534 .033 6.680 1.065 .966 .326 2.847 .940 .028 .868 .856 .484 1.145 .285 .596 Note: Logistic method: forced entry method ('Enter'). Dependent variable: Mode of cabinet resignation (0 = elections, 1 = pre-term). The independent variables are coded as: intra-executive conflict (low/high), semi-presidential type (premier-presidential/president-parliamentary), government form (majority/minority), and party system fragmentation (low/high) and post-communist period (1990-94, 1995-99, 2000-4, 2005-7). Party system fragmentation is calculated by using the Laakso and Taagepeera effective number of parties index. The equation of the index is as follows: N = (Ep;2)"1 where N denotes the number of effective parties and P; is the share of votes or seats (in this case seats) won by the ith party. For the number of effective parties by country, see Table A2 in the Appendix. When the effective number of parties has been found to be below 3.5, the level of fragmentation has been coded as 'Low', and from 3.5 and above as 'High'. level of party system fragmentation35 (0 = low, 1 = high) and post-communist period (0 = 1990-94, 1 = 1995-99, 2 = 2000-5, 3 = 2005-7). From the logistic regression test in Table 3, we find that intra-executive conflict is by far the strongest predictor among the included variables. In probability terms we may interpret the figures to indicate that the chances of pre-term cabinet resignation - when controlling for other variables - are almost 23 times higher under high intra-executive conflict, (see Exp (B) in Table 3). We also find that the semi-presidential type in itself is quite a strong predictor, which is hardly surprising considering the results in Table 2. The presidents in the president-parliamentary systems have used their power to dismiss the prime ministers quite frequently, and accordingly, the regression 35 As measured by the Laakso and Taagepera effective number of parties index, M. Laakso and R. Taagepeera, 'Effective Number of Parties: A Measure with Application to West Europe', Comparative Political Studies, 12: 1 (1979), pp. 3-12. © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd INTRA-EXECUTIVE CONFLICT AND CABINET INSTABILITY 521 test suggests that the odds of pre-term cabinet resignation are about seven times higher in president-parliamentary systems than in premier-presidential systems (Table 3). More unexpectedly, neither cabinet type and post-communist period nor party system fragmentation prove to be nearly as strong predictors as intra-executive conflict in the overall regression model. These results clearly support our assumption that intra-executive conflicts, as such, are likely to have a significant negative effect on cabinet stability in both premier-presidential and president-parliamentary systems. So far, descriptive statistics have been backed up with a logistic regression analysis, but we still need to explore in more depth just in what ways intra-executive conflict increases the risk of pre-term termination of governments. In the next section, we will provide some empirical in-depth illustrations from our cases in order to show that intra-executive conflicts, as such, are likely to have a significant negative effect on cabinet duration. EXPLORING THE LINKS BETWEEN INTRA-EXECUTIVE CONFLICT AND CABINET INSTABILITY To what extent are the specific institutional features of premier-presidentialism and president-parliamentarism relevant to the relation between intra-executive conflict and cabinet instability? The institutional mechanisms behind cabinet dismissals in president-parliamentary systems can be illustrated by the cases of Russia and Ukraine (1991-2006). Whenever the prime ministers in Russia and Ukraine have entered into conflict with the president it has led to the subsequent dismissal of the prime minister. As for Russia, most observers would probably agree on the description of the relationship between the president and the government as 'strong president-weak government' (at least if we consider the period up until the end of Vladimir Putin's presidency in 2008). The president's strong hold over prime minister appointment and dismissal, as well as his overall influence over the cabinet's work have reduced the likelihood of cabinets challenging the president. Consequently, the frequency of intra-executive conflict in Russia has generally been quite low. The country did, however, experience rather frequent changes of prime ministers in the 1990s. A manifest example of intra-executive struggle did occur © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd 522 GOVERNMENT AND OPPOSITION under the cabinet of Yevgeny Primakov in 1998-99, under which president-prime minister relations were marked by rivalry. Primakov had close links to the old Soviet/Russian security apparatus and to the powerful ministries of defence, interior and foreign affairs, and his focus, naturally, tended to move away from economic issues towards the presidential domain of foreign and internal political affairs. Yeltsin, furthermore, disliked the popularity and independent power of Primakov and decided to replace him with a more loyal and dependent person.36 In the case of Ukraine, the relatively more frequent occurrences of intra-executive conflict reflect the high level of political tension that has continuously disrupted the country's political system. Leonid Kuchma's ten years as president (1994-2004) were marked by institutional rivalry, both between the president and the prime ministers, and between the president and the parliament. Kuchma used the dual authority structure of the president-parliamentary system to 'hide' behind the prime ministers in order to avoid popular dissatisfaction stemming from the consequences of ineffective economic reforms.37 Up until 2004, Kuchma had clashed with no less than seven prime ministers, all of whom he consequently dismissed. Apparently, institutional instability has also continued in the post-Orange Revolution era (2005 and onwards) and even after the shift from president-parliamentarism to premier-presidentialism in 2006. Under the presidency of Viktor Yuschenko, there were a number of intra-executive struggles, some resulting in stalemating clashes between the president and the cabinet, perhaps most clearly demonstrated by the tug-of-war between Yuschenko and Yanukovich in 2006-7. The still unconsolidated premier-presidential framework under which the latter conflict occurred may in fact have prolonged the political deadlock at the time, as President Yuschenko, in accordance with the premier-presidential definition, was left without the power of prime minister dismissal. Recall that the definition of premier-presidentialism stipulates that the prime minister is subjected to parliamentary confidence, i.e. only the parliament is provided with the power of prime minister dismissal. The president is therefore constitutionally destined to 36 R. Sakwa, Russian Politics and Society, 3rd edn, New York, Roudedge, 2002, p. 118. 37 E.g. A. Wilson, 'Ukraine', in Elgie, Semi-Presidentialism in Europe, p. 265. © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd INTRA-EXECUTIVE CONFLICT AND CABINET INSTABILITY 523 accept the cabinet as long as it is tolerated by the parliament. However, the results from our data show that intra-executive conflict has not only been associated with cabinet instability in the president-parliamentary systems, but also in the premier-presidential ones. These findings suggest that we have to look beyond the distribution of cabinet dismissal power factor for possible explanations. Although we find only limited theoretical guidance in the literature for interpreting this phenomenon, some factors that are related both to the specific context of post-communist transition but also more generally to the institutional feature of premier-presidentialism can be pointed out. The puzzle is such that the presidents often find that their prestige and popularity exceeds their formal constitutional powers, while prime ministers, who possess the main share of executive powers, often lack sufficient legitimacy among the population. Survey data report that popular support for presidents is consistently higher than for other political institutions (including the prime minister) all throughout Central and Eastern Europe.38 These imbalances have tended to favour a go-public strategy by the presidents under intra-executive conflict. By publicly criticizing the government through media addresses and public speeches - with negative effects on the cabinet's credibility - the presidents can make it more or less impossible for prime ministers to remain in office. This strategy has proved to be effectively used by several presidents in the post-communist context. For instance, in 1999, Lithuanian President Valdas Adamkus openly voiced his criticism against Prime Minister Gediminas Vagnorius. Public opinion turned out to be strongly on the side of Adamkus, and Vagnorius ultimately had little choice but to resign.39 Similarly, Poland's first post-communist president, Lech Walesa (1990-95), repeatedly used the option of public critique directed against the government. On several occasions, he threatened to assume the prime minister post in order to expedite reforms, even though this would have violated the constitutional order severely. Fortunately, in most cases 38 Cf. data from the New Europe Barometer, 1991-2005 and New Russia Barometer, 1992-2008, Centre for the Study of Public Policy (CSPP), University of Strathclyde (reports are available at: www.abdn.ac.uk). 39 K. Duvold and M. Jurkynas, 'Lithuania', in Berglund etal., The Handbook of Political Change in Eastern Europe, East European Constitutional Review, 'Constitution Watch'. © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd 524 GOVERNMENT AND OPPOSITION Walesa's public statements, although arguably damaging to a constitutional culture, were not followed by action. Rather, this was mainly rhetorical tactics aimed at putting pressure on the government - tactics that proved to be quite 'successful' in so far as all prime ministers that entered into conflict with Walesa stepped down as a consequence. The president in premier-presidential systems may also use formal instruments to make life more difficult for the cabinet, such as veto power and remittances of legislative bills to the constitutional court. In the post-communist countries, we can point out many such cases where institutional prerogatives have exacerbated intra-executive conflict and ultimately resulted in the resignation of government. The cohabitation between the centre-liberal oriented Bulgarian President Zheliyu Zhelev and the left-wing cabinet of Zhan Videnov in 1994-96 is illustrative. The president used his rather limited prerogatives to challenge the government. An important asset in this respect was his power to appoint and dismiss high-ranking military officers as well as his claimed constitutional responsibility for shaping defence policy and overall defence strategy. Zhelev thus blocked several military appointments proposed by the government, and played out the judicial card and filed a number of government petitions to the constitutional court for review. Finally, Zhelev combined the institutional blocking strategy with the go-public option by bringing his case to the public. In a series of publicized and televised announcements, Zhelev unleashed harsh criticisms against the government and the left-wing majority in parliament, claiming that they were solely responsible for the downturn of the Bulgarian economy and the deep socio-economic crisis that the country went through at the time. The president's strategies contributed to an already growing public dissatisfaction with the government, which ultimately led to the resignation of the Videnov cabinet in December 1996.40 In a nutshell, the outcome of intra-executive conflict in terms of cabinet instability relates to the dual executive structure built into the premier-presidential system. Among factors that have been identified by scholars as particularly disadvantageous to the prime minister vis-a-vis the president in the post-communist countries are a general lack of party system coherence, prime minister responsibility for unpopular policies, and the fact that several of the prime ministers 40 Cf. V. I. Ganev, 'Bulgaria', in Elgie, Semi-Presidentialism in Europe. © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd INTRA-EXECUTIVE CONFLICT AND CABINET INSTABILITY 525 have brought with them a technocratic mindset, often being less inclined to pursue grassroots organization and party building. In addition, relatively weak administrative capacity of the prime ministers' offices (often competing with quite extensive presidential offices) has partly impeded effective reforms.41 In the post-communist countries, such factors have strongly favoured the side of the presidents under intra-executive conflict, and, accordingly, forced the prime ministers and their cabinets to resign. CONCLUSIONS The aim of this article has been to contribute to our understanding of the effects of semi-presidentialism in transitional countries and we have applied a narrow focus on intra-executive conflict in relation to survival of cabinets. The dual executive structure of semi-presidentialism carries the risk of intra-executive struggles between the president and the prime minister. Previous studies have quite convincingly demonstrated that intra-executive relations in semi-presidential systems are affected by the nature and extent of the cabinet's support in parliament as well as by the degree of presidential control over the cabinet. However, the actual effects of intra-executive conflicts in the two types of semi-presidential systems have so far not been systematically explored. The present empirical investigation has supported the assumption that intra-executive conflict generates cabinet instability. The data revealed that pre-term resignation of cabinets in both premier-presidential and president-parliamentary systems have clearly been more frequent under periods of intra-executive conflict than under periods of peaceful relations within the executive. In addition, our logistic regression test indicated that intra-executive conflict is a more powerful predictor of pre-term cabinet resignation than any of the other factors included, i.e. the type of semi-presidential system, the form of government, the level of party system fragmentation, and the post-communist period. Although limited in scope, the empirical analysis provides a certain basis for arguing that the occurrence of intra-executive 41 T. A. Baylis, 'Embattled Executives: Prime Ministerial Weakness in East Central Europe', Communist and Post-Communist Studies, 40 (2007), pp. 81-106. © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd 526 GOVERNMENT AND OPPOSITION conflict in semi-presidential systems is likely to produce relatively high rates of cabinet turnover. In president-parliamentary systems, the logic behind this relationship is quite obvious. Since the president in these systems possesses unilateral power to dismiss the prime minister, this is a most likely outcome of any serious conflict between the president and the prime minister. In premier-presidential systems, however, cabinet instability is not linked to dismissal powers (since these are not provided to the president). Although further research is required, we have pointed out some possible explanations in terms of both formal and informal means. A key factor seems to be the higher status of the presidency as compared to the prime minister position. The president can put considerable constraints on the prime minister and the cabinet by leaning on his popular mandate. As illustrated by our cases, a particularly rewarding strategy is for the president to go public and display distrust in the cabinet, thereby indirectly forcing the cabinet to step down. In a transitional context of limited party system institutionalization and lack of parliamentary coherence, the imbalance is strongly to the advantage of the president. Admittedly, a more substantial account for party system factors can probably contribute to our further understanding of these mechanisms. For one thing, presidential party influence and intra-party negotiations are crucial factors not sufficiently explored here. It may be argued that our empirical findings - the relationship between the different types of semi-presidentialism and cabinet instability - could be attributed primarily to the fact that some of our countries have more semi-authoritarian features than others, and that this negatively affects the possibility of making generalizations. However, our research task has not been to control for democratic/ authoritarian features, but more modestly to establish empirically the link between intra-executive conflict and cabinet instability, as a base for discussion and further research. The wider question is whether one should also consider cabinet stability to be a measure of regime stability? Some scholars have argued quite strongly against such an assumption, since the disadvantages of frequent cabinet shifts may not be as serious as they appear42 while others have argued that cabinet instability is indeed 42 E.g. A. Lijphart, Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries, New Haven, CT, Yale University Press, 1999; M. Dogan, 'Irremovable © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd INTRA-EXECUTIVE CONFLICT AND CABINET INSTABILITY 527 related to regime instability. One of their main arguments is that short-lived cabinets do not have sufficient time to develop sound and coherent policies and reforms, which may endanger the viability of democracy and regime stability, especially in new democracies.43 In the context of recent regime transitions there is typically a need for far-reaching political reforms, and short-lived cabinets obviously have less time to develop coherent policies, which may ultimately constitute a threat to the democratization process and regime stability. Poland under Walesa's presidency is an illustrative case, where important policy reforms (relating to privatization and decentralization) were delayed and/or abandoned as a consequence of frequent shifts of cabinets. At the same time, we should be careful to point out a causal relationship between cabinet instability and policy ineffectiveness. In several of the post-communist countries, cabinets have resigned as a consequence of already failed policy reforms, and it would thus be misleading to implicate that ineffective policy reforms may be attributed to cabinet instability alone. But then again, failure to provide clear directions of the economic reforms - typical of several post-communist countries in the early 1990s - tended to be further complicated by frequent rotation of cabinets and changes of prime ministers. At the end of the day, cabinet instability tends to interrupt or delay policy reforms, something that can have negative long-term effects on transitional societies as well as on new democracies. Leaders and Ministerial Instability in European Democracies', in M. Dogan (ed.), Pathways to Power: Selecting Rulers in Pluralist Democracies, Boulder, CO, Westview Press, 1989. 43 J. J. Linz and A. Stepan, Problems of Democratic Transition and Consolidation: Southern Europe, South America, and Post-Communist Europe, Baltimore, Johns Hopkins University Press, 1996; D. L. Horowitz, 'Constitutional Design: Proposals Versus Processes', in A. Reynolds (ed.), The Architecture of Democracy: Constitutional Design, Conflict Management, and Democracy, Oxford, Oxford University Press, 2002. © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd 528 GOVERNMENT AND OPPOSITION APPENDIX Table Al Level of Intra-executive Conflict (1991-2007) President and term in office Prime minister and term in office Cabinet and base of support Intra-executive conflict Bulgaria Z Zhelev (Jan 92-Jan 97) P Stoyanov (Jan 97-Jan 02) G Parvanov (Jan 02-) Croatia F Tudjman (May 92-June 97) F Tudjman (June 97-Nov 99) S Mesic (Feb 00-Jan 05) Lithuania A Brazauskas (Mar 93-Mar 98) V Adamkus (Mar 98-Feb 03) R Paksas (Feb 03-Apr 04) V Adamkus (July 04-July 09) Moldova M Snegur (Dec 91-Jan 97) P Lucinschi (Jan 97-Dec 00) Shifted from premier-presidentialism to parliamentarism in 2000 Poland L Walesa (Dec 90-Dec 95) A Kwasniewski (Dec 95-Nov 00) F Dimitrov (Nov 91-Dec 92) L Berov (Dec 92-Oct 94) Z Videnov (Jan 95-Feb 97) I Rostov (May 97-July 01) S Saxe-Coburg-Gotha (July 01- ) - " - (-July 05) H Sarinic (Aug 92-Apr 93) N Valentic (Apr 93-Nov 95) Z Matesa (Nov 95-Jan 00) IRacan (Jan 00-July 02) - " - (July 02-Dec 03) / Sanader (Dec 03-July 09) A Slezevicius (Mar 93-Feb 96) L Stankevicius (Feb 96-Nov 96) G Vagnorius (Nov 96-Dec 98) - " - (Jan 99-May 99) R Paksas (May 99-Oct 99) A Kubilius (Nov 99-Oct 00) R Paksas (Oct 00-June 01) A Brazauskas (June 01-) - " - (-June 06) A Sangheli (July 94-Nov 96) / Ciubuc (Apr 98-Feb 99) / Stürza (Mar 99-Dec 99) D Braghis (Dec 99-Apr 01) J Bielecki (Jan 91-Dec 91) J Olszewski (Dec 91-June 92) H Suchocka (June 92-Oct 93) WPawlak (Oct 93-Mar 95) / Oleksy (Mar 95- ) W Cimoszewicz (Feb 96-Oct 97) JBuzek (Oct 97- ) Centre-right minority Technocratic minority Left majority Centre-right majority Centre-right majority Centre-right minority Right-wing majority Right-wing majority Right-wing majority Left-cen tre-righ t maj ori ty Left-centre minority Right-wing minority Left-wing majority Left-wing majority Right-wing majority Right-wing majority Right-wing majority Centre-right minority Centre-left majority Centre-left majority Centre majority Centre-techn. majority Centre-left majority Right-wing minority Centre-right minority Right-wing minority Left-wing majority Left-wing majority Left-wing majority Right-wing majority High Med-High High Low-Med Low Low Low Low Low Low Low-Med Low-Med Low-Med Low Low High Low Low Med-High Low-Med Med-High Medium Med-High Low Low Low Low High Low High High Low Low-Med © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd INTRA-EXECUTIVE CONFLICT AND CABINET INSTABILITY 529 Table Al Continued President and term in office Prime minister and Cabinet and base Intra-executive term in office of support conflict A Kwasniewski (Nov - " - (Oct 01) - " - Meet-High 00-Sep 05) L Miller (Oct 01-May 04) Left-wing majority - 03 Meet-High Left-wing minority 03- Meet-High - " - MBelka (May 04-Sep 05) Left-wing minority Low L Kaczynski (Oct 05-Apr Kaczynski (July 06-July Right majority Low 10) 07) Romania Illiescu (May 90-Nov 92) P Roman (Jan 90-Oct 91) Left-nationalist maj ority High T Stolojan (Oct 91-Nov 92) Left-technocrat majority Low N Vacariou (Nov 92-Dec Left-technocrat minority Low 96) E Constantinescu (Nov V Ciorbea (Dec 96-Mar 98) Centre-right majority Low 96-Dec 00) R Vasile (Apr 98-Dec 99) Centre-right majority High M Isarescu (Dec 99-Dec Centre-right majority Low 00) Illiescu (Dec 00-Nov 04) A Nastase (Dec 00-Nov 04) Left minority Low-Med T Basescu (Dec 04— ) C Popescu-Tariceanu (Dec Centre-right majority High 04-Dec 08) Russia B Yeltsin (July 91-July 96) Y Gaidar (June 92-Dec 92) Technocrat minority Low-Med - " - V Chernomyrdin (Dec 92-) Technocrat minority Low B Yeltsin (July 96-Mar 00) - " - (Mar 98) - " - Low-Med S Kirienko (Mar 98-Aug Low 98) Y Primakov (Sep 98-May Left-cen tre-righ t maj ori ty High 99) S Stepashin (May 99-Aug Technocrat minority Medium 99) V Putin (Aug 99-May 00) Low V Putin (Apr 00-Mar 04) M Kasyanov (May 00-Feb Technocrat majority Low 04) V Putin (Mar 04-May 08) M Fradkov (Mar 04-Sep Technocrat majority Low 07) Ukraine L Kravchuk (Dec 91-July L Kuchma (Oct 92-Sep 93) Technocrat minority High 94) L Kuchma (July 94-Oct VMasol (June 94-Mar 95) Left minority Low 99) - " - Y Marchuk (Mar 95-May Technocrat minority High 96) P Lazarenko (May 96-July High 97) - " - V Pustovochenko (July - " - Low 97-Dec 99) L Kuchma (Nov 99-Oct V Yushchenko (Dec 99-May - " - High 04) 01) - " - A Kinakh (May 01-Nov 02) Low V Yanukovich (Nov 02-) Centre-left majority Low V Yushchenko (Jan 05-Feb Y Tymoshenko (Jan 05-Sep Centre-left majority High 10) 05) V Yanukovich (Aug 06-Sep Centre-left majority High 07) Note: Cases of short-lived governments, which have held office less than a year, as well as cases where we have been unable to obtain sufficient information about intra-executive relations, have not been assessed and are excluded from the list. © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd 530 GOVERNMENT AND OPPOSITION Table A2 The Effective Number of Parties in the Lower House of Parliament 1st election 2nd election 3rd election 4th election 5th election Bulgaria 2.4 (90) 2.4 (91) 2.7 (94) 2.5 (97) 2.9 (01) Croatia 2.5 (92) 2.3 (95) 2.9 (00) 3.2 (03) Moldova - 2.6 (94) 3.4 (98) 1.8 (01) Lithuania 3.0 (92) 3.3 (96) 4.2 (00) 4.7 (04) Poland 10.9 (91) 3.9 (93) 3.0 (97) 3.6 (01) Romania 2.2 (90) 4.4 (92) 4.3 (96) 3.5 (00) 3.2 (04) Russia - 14.9 (93) 6.2 (95) 8.0 (99) 3.7 (03) Ukraine - 19.2 (94) 9.6 (98) 6.7 (02) Note: Based on the proportion of seats (Ns). The equation of the index is as follows: N = (Ep;2)"1 where N denotes the number of effective parties and P; is the share of votes or seats won by the ith party. Sources: Electoral data from the International Foundation for Electoral Systems (IFES), www.ifes.org; and K. Armingeon and C. Romana, Comparative Data Set for 28 Post-Communist Countries, 1989-2004, University of Berne, Institute of Political Science, 2004. Table A3 Collinearity Diagnostics of Logistic Regression, Mode of Cabinet Termination Tolerance VLF Eigenvalue Lntra-executive conflict .947 1.056 .821 Type of semi-presidentialism .777 1.287 .570 Cabinet type .613 1.631 .343 Party system fragmentation .671 1.491 .208 Post-communist period .847 1.180 .140 Note: The logistic regression model (Table 3), calls for a test of multicol-linearity. Table A3 reports the collinearity coefficients on the variables included in the logistic regression test. The VIF indicators show whether a predictor has a strong relationship with the other predictors, and values in the range between 5 and 10 are to worry about. In this case, there are no such high values, and we can also conclude from the other two types of values, that there is little reason to be concerned about multicollinearity. The tolerance values in the first column should be above 0.1, and the eigenvalues in the third column, should not vary too much among the variables if we want to rule out collinearity problems. Both conditions are apparently satisfying in this model. Dependent variable: Mode of cabinet termination. © The Authors 2010. Government and Opposition © 2010 Government and Opposition Ltd