Presidents and the Choice of Prime Ministers in Parliamentary Democracies Cristina Bucur Department of Political Science University of Oslo cristina.bucur@stv.uio.no José Antonio Cheibub Department of Political Science University of Illinois at Urbana- Champaign cheibub@illinois.edu Work in progress. Very preliminary draft Prepared for presentation at the Institutional Determinants of Legislative Coalition Management workshop at Tel- Aviv University, 16-19 November 2015, Tel- Aviv, Israel (as part of the project The Evolution of Parliamentarism and Its Political Consequences. We thank Matt Winters for extremely helpful discussions about the project and Martin Søyland for comments on an earlier version. They do not bear any responsibility for the mistakes encountered below. We also thank Lars Sutterud and Mads Motrøen for great research assistance.
Abstract This paper aims to assess the role played by heads of state in the government formation process by investigating their ability to help their own parties gain the prime ministership. We want to evaluate whether heads of state are able to tilt the balance in the government formation process in favour of his or her party and, if they do, whether this has any consequence for the performance of the government that is formed. The analysis in this paper is based on data for 19 parliamentary and semi- presidential European democracies, plus Israel and Turkey, over the post- war period. We find that directly elected presidents do not favour their parties any more than indirectly elected presidents and that the constitutional powers of presidents to nominate the formateur do not seem to add anything to what presidents actually do when it comes to choosing the party of the prime minister. Our findings suggest that there are circumstances under which presidents do exercise discretion in choosing the party of the prime minister and that, under these circumstances, they will prefer to pick their own party over other parties. This is true for all presidents in parliamentary systems, whether they are elected by voters or by the parliament, and whether the constitution formally allows them to nominate the formateur. Importantly, presidents do not go against what the theory of parliamentarism would expect them to do, namely, to choose the party that holds the highest chance of belonging to a legislative majority.
Introduction Securing the position of prime minister is arguably the primary goal of many individuals and political parties in parliamentary democracies. Politically and institutionally, prime ministers exercise great power over the choice of governmental policies and the distribution of the spoils of office and are, thus, the most coveted position. Even though in some countries they are just the primus inter pares, they have an agenda setting advantage, which allows them to obtain policies closer to their preferences (Huber on confidence vote). But who gets to be the prime? The answer, in a sense, is unproblematic: the leader of the party most likely to obtain the support of a majority in parliament. Although usually this means the party that holds a plurality of seats, the fact that no party holds a majority of seats in most parliamentary democracies implies that multiple coalitions could potentially obtain the support of a majority in parliament. Under these circumstances, which party will fill the office of the prime minister? Consider the formation of the Romanian government after the 2004 elections. On November 28, 2004, Romania held legislative and the first round of presidential elections. The ruling Social- Democratic Party (PSD) topped both polls: on the one hand, the pre- electoral coalition it had formed with the much smaller Humanist Party (PUR) obtained almost 40 percent of the seats in the Chamber of Deputies, while the opposition centre- right alliance formed by the National Liberals (PNL) and the Democratic Party (PD) obtained 14 and 19 percent of the seats respectively; on the other hand, Adrian Năstase, the incumbent Prime Minister, had an eight- point lead in the first round of presidential elections over Traian Băsescu, the leader of the Democratic Party and the PNL- PD alliance. In the December 12 presidential run- off election, however, Traian Băsescu delivered a surprise victory against the PSD candidate. Subsequently, despite the Social- Democrats being the largest party in the newly elected parliament and leading the largest parliamentary coalition, president Băsescu went on to nominate the leader of the National Liberals as prime minister and 1
convinced the Humanist Party to break their pre- election agreement with the PSD and join the new PNL- PD government. It is possible that the new president s choice of prime minister was guided by partisan concerns; such concerns may have led him to disregard considerations about the survival prospects of a government not headed by the largest parliamentary party. But it might also be the case that the president saw that at least two equally viable coalitions were possible and, given that, he chose the prime minister from his own coalition. In this case, the president exercised the discretion granted him by the constitution (art. 85), with results that were probably innocuous: the appointed prime minister, Călin Popescu- Tăriceanu, won investiture by 265 to 200 votes and governed for the next four years. This paper aims to assess the role played by heads of state in the government formation process by investigating their ability to help their own parties gain the prime ministership. We want to evaluate whether heads of state are able to tilt the balance in the government formation process in favour of his or her party and, if they do, whether this has any consequence for the performance of the government that is formed. Other researchers have asked whether prime ministers tend to be disproportionately drawn from the party of the head of state and found that they do when the head of state is indirectly elected (Kang 2009; Glasgow et al. 2011). The problem with these studies is that their findings could result from two distinct processes. In the first, prime ministers come disproportionately from the party of the indirectly elected head of state because of his/her partisan bias; in the second, prime ministers come disproportionately from the party of the party of the indirectly elected head of state because these parties are more likely to be a member of a viable coalition than any other party in parliament. Whereas in the first case the head of state is, arguably but not necessarily, exercising his discretion to favour his party, in the second he is behaving in perfect accord with what one would expect in a parliamentary democracy. The analysis in this paper is based on data for 19 parliamentary and semi- presidential European democracies, plus Israel and Turkey, over the post- war period. 2
We show that presidents who play the role of agenda- setters in government formation and face little constraints in government formation, including weak investiture regimes, tend to promote their own parties to the premiership than presidents who lack agenda- setting powers. In particular, we find that indirectly elected presidents are more likely than popularly elected presidents to designate a co- partisan as the prime minister. This is unsurprising and in line with previous studies. Our contribution is to provide better estimates of the power of heads of states. Additionally, our paper also assesses the role of monarchs in designating prime ministers and investigates whether governments that were formed in defiance of the normal parliamentary procedures, that is, governments for which we can identify the influence of the head of state, are more or less likely to last in office than those who were formed strictly on the basis of the distribution of parliamentary seats. In this version of the paper we only present a very preliminary analysis of the first issue. The role of the head of state in government formation Parliamentarism is defined as a regime in which governments are responsible to a legislative majority. This means that governments must be at least tolerated by a majority in parliament for it to exist; if there is a majority that prefers a different government or new elections to the existing government, that majority can approve a vote of no confidence, which will result in the removal of the government from office. This simple definition is profound in the sense that it is sufficient to generate a whole new theory of parliamentary governments. According to this theory, parliamentarism is a system that aligns the interests of the parliamentary majority and the government since the latter needs to be supported or at least tacitly tolerated by the former. Governments will be formed with the goal of extending its life and the life of parliament. This means that governments will be formed by a coalition of parties that has the highest chances of getting and maintaining the support of a majority in parliament. This could happen automatically i.e., the distribution of seats is such that there is one and only one coalition that could be supported by a majority and that 3
coalition forms the government or it could happen as the result of bargaining among parties that could combine into more than one majority coalition. Regardless, in the theory of parliamentarism, the government that is formed is the one that supposedly holds the highest prospects of obtaining and keeping the support of a parliamentary majority. In this sense we could say that the theory of parliamentarism presupposes the existence of a majoritarian imperative, even if, in reality, some parties under some conditions may choose to refrain from joining a government, even if they could (Strøm 1990). The role of the head of state in the theory of parliamentarism is ambiguous. On the one hand, the typical narrative about the emergence of parliamentary democracies pits a parliament that wrests power from a monarch in a long struggle for preponderance. The monarch, dependent on parliament for revenues, ultimately recognizes that the only way to govern is not to rule: fighting the majority in parliament is futile and power is transferred to the parliament. Governments are formed and supported in power strictly in accordance to the relative strength of parties in parliament and the monarch is no more than a symbolic figure. When monarchs are absent, heads of state are nothing but republican kings, figure heads that simply serve the purpose of representing the permanence of the nation- state. One the other hand, parliamentary constitutions, old and new, explicitly or implicitly assign at least some discretion to the head of state in either directly appointing the prime minister, or designating the individual in charge of conducting negotiations to form a government (the formateur). And where there is discretion there is the potential for strategic behaviour and influence. The question, therefore, arises, as to whether heads of state in parliamentary democracies exercise more influence than their status as figure heads or mere symbols of the nation, as prescribed by the theory of parliamentarism. As a matter of fact, suspicion that this is the case has become unavoidable with the expansion since 1990 of the number of parliamentary democracies that directly elect their heads of state. The influence of the popularly elected president in government and policy formation has become the focus of a large literature, concerned with the potential conflicts between the head of state, 4
elected by the people, and the head of government, responsible to a parliamentary majority (Shugart and Carey 1992; Protsyk 2005; Elgie 2008; Schleiter and Morgan- Jones 2009a, 2009b, 2010; Sedelius and Ekman 2010; Elgie et al. 2011; Sedelius and Mashtaler 2013). Thus, given that all heads of state have at least some role to play in the process of government formation, it is fair to ask whether they are able to influence that process. Specifically, we want to know whether the head of state can choose the prime minister in such a way as to modify the kind of government that would have been formed under the strict operation of the parliamentary imperative. If we assume that heads of state have policy preferences, the question is whether these policy preferences affect whom they will choose to head the government or to lead negotiations for the formation of a government. Assuming that parties represent different policy packages, we then consider whether heads of state express a preference for co- partisans when it comes to choosing a prime minister. The party of the head of state may be correlated with the party of the prime minister for at least two reasons: first, because the head of state tends to deviate from the majoritarian imperative and chooses co- partisans as prime ministers even when, under that imperative, individuals from other parties should have been chosen; second, because the party of the head of state is more often than not the party that should be chosen to head the government under the majoritarian imperative. Recent papers about the choice of prime ministers in parliamentary democracies do not allow us to distinguish these two scenarios, thus biasing upward the estimate of the power of the head of state. When it comes to estimating the power of the head of state in the choice of prime ministers, their strategy is to simply hold the majoritarian imperative constant and then look at the effect of being the party of the head of state. The problem with this approach is that it estimates the power of the head of state on the basis of observations that are effectively uninformative: cases in which a party is the party that would have chosen the prime minister under the majoritarian imperative and is also the party of the head of state. Moreover, the assumption in this approach is that the power of the president is uniform across the entire 5
distribution of the majoritarian imperative. In other words, given that the president is able to influence the choice of prime minister, that influence is constant regardless of the party s chances as determined by the majoritarian imperative. Our approach is distinct. We start by computing for each party the likelihood that it will be part of a majority coalition. We take the relationship between this likelihood and being the party of the prime minister as a representation of the choice of prime ministers that would have been made under the so- called majoritarian imperative: that is, under this imperative, which is derived from the theory of parliamentarism, the party heading the government should be the one with the highest chances of being part of a majority coalition. We model the effect of the head of state s preferences by interacting the information about the party of the head of state with the probability of holding the premiership under the majoritarian imperative. By allowing the influence of the head of state to vary at different levels of the majoritarian imperative we are able to identify the situations when that influence really matters: when the president is able to modify what would have happened if he or she had no influence over the choice of prime ministers. Our expectations are as follows: - If the head of state plays no role in choosing the head of government, we should observe that the outcome of the choice under the majoritarian imperative is unaffected by being his or her co- partisan. - If the head of state affects the choice of prime minister, we should observe this influence manifesting itself in a reduction of the effect of the majoritarian imperative. That is to say, we should observe that the president is able to act against what would have happened under pure parliamentarism. We conduct our analysis using two separate datasets. The first one contains information for all countries in which the head of state is a president. In this dataset, we distinguish presidents along two dimensions that may affect their ability to choose the prime minister. The first is the president s mode of election: indirect or direct. Students of semi- presidentialism stress how the fact that they have been popularly 6
elected may constitute a source of power or influence for presidents in these systems. No consensus has emerged as to where and how presidential influence is exercised. As a matter of fact, no consensus has emerged as to the very definition of semi- presidentialism. But it could be the case that armed with the legitimacy of having been directly elected, presidents in semi- presidential systems will meddle in government formation in ways that their indirectly elected peers will not. The second dimension along which we distinguish presidents is in their constitutionally prescribed role in the process of government formation. Because constitutions vary in how they specify the role of the president in government formation, we distinguish between presidents who are proactive and those who are passive. Proactive presidents are those who move first in nominating an individual as the head of government or the formateur (who will then become the prime minister). In some cases, once the government is formed, it begins to govern immediately and stays in office until the next election or until it faces a vote of no confidence and loses. In other cases, the person nominated has to face a vote of investiture in parliament. The target of the vote can be the individual prime minister, the government as a whole, or the government program (Rasch et al. 2015). The relevant fact is that the head of state has discretion in nominating (and in some cases appointing) the prime minister. In other cases, the head of state plays a purely formal role in the nomination of the prime minister or formateur. Whether because the parliament nominates and selects a prime minister, who is then presented to the head of state for appointment (like in Ireland and Germany), or because the president is constitutionally required to nominate a specific individual for the post (like in Bulgaria), the fact is that the head of state plays no formal role in identifying candidates for the post of prime minister or in choosing those candidates. All that he or she does is to appoint someone who is presented as the choice of the parliament or the person identified by the constitution. Our analysis, thus, distinguishes four types of presidents, depending on their mode of election and their constitutional role in government formation. Table 1 presents the cases included in our analysis classified according to these two dimensions. 7
Table 1 Presidential powers and mode of election President's role in government formation Proactive Passive Mode of presidential election Direct Indirect Austria (1951-2014) France (1965-2014) Iceland (1945-2014) Portugal (1976-2014) Finland (1992-1999) Lithuania (1993-2014) Poland (1990-1992) Poland (1997-2014) Romania (1991-2013) Slovakia (1999-2014) Turkey (2014) Austria (1945-1950) Finland (1945-1991) France (1958-1964) Greece (1975-1986) Hungary (1991-2014) Italy (1948-2014) Israel (1949-1967) Latvia (1993-2014) Poland (1989) Slovakia (1993-1998) Turkey (1983-2013) Bulgaria (1991-2014) Croatia (2001-2014) Czech Rep (2013-2014) Finland (2000-2014) Ireland (1945-2014) Poland (1993-1996) Slovenia (1992-2014) Czech Rep (1993-2012) France (1946-1958) Estonia (1992-2014) Germany (1949-2014) Greece (1987-2014) Israel (1968-2014) The second dataset is comprised of all heads of state that we find in parliamentary democracies: directly elected presidents, indirectly directly elected presidents, and monarchs. Clearly, identifying monarchs party preferences, if they have one, is impossible. We can, however, make assumptions about monarchs political preferences: they are essentially conservative. Thus, assuming that this is the case, we adapt the analysis at the level of political parties and perform it at the level of ideological blocks: left, center and right. The basic idea is that heads of state of any 8
type may influence the government formation process in order to obtain a prime minister from a block that is aligned with their own ideology. Thus, if monarchs have an independent influence over government formation, we should observe them choosing right- wing prime ministers even if these individuals would not have been the choice under the majoritarian imperative. We recognize that this analysis is relatively crude and relies on an assumption about something we cannot observe. But it may give us a glimpse of whether monarchs in contemporary democracies are as innocuous as we believe them to be. In this version of the paper we only report on the results pertaining to the first dataset, that is, the role of presidents in government formation. Data and variables We use a data set consisting of coalition governments in 23 countries included in the ParlGov database (Döring and Manow 2015) covering the post- war period (c. 1945). The countries and exact time periods included in this study are the ones listed in Table 1. Following Müller and Strøm (2000: 12), a new government is counted whenever there is a change in the party composition of a government coalition; a change in the identity of the prime minister; or a general parliamentary election. We observe 884 formation opportunities, but exclude from the analysis those that were for caretaker governments (93), those where a single party controlled a parliamentary majority (134), those that resulted in a non- partisan prime minister (10), those with a non- partisan president (95) and those where the head of state was a monarch (200). Because some of these overlapped, we are left with a total of 415 PM selection opportunities and 3,473 potential PM parties. In line with recent studies on government formation (Martin and Stevenson 2001; Diermeier and Merlo 2004; Bäck and Dumont 2008; Kang 2009), we model the choice of prime ministerial party as a discrete choice problem, where the units are the formation opportunities and the parties are the alternatives. Since Martin and Stevenson (2001), the standard procedure for the empirical analysis of coalition formation outcomes has been McFadden s (1974) conditional logit model, which we also use to estimate the power of the president in government formation. In this 9
section we report our basic findings. Before we do that, however, a word is necessary about our measurement of the majoritarian imperative. As we discussed earlier, the theory of parliamentarism suggests that governments will be formed by the party with the highest chances of holding a majority of seats in the parliament. We call this the majoritarian imperative of parliamentary democracies. All existing studies implicitly recognize the existence of such imperative and operationalize it as the largest party in parliament or the share of seats held in parliament. We use a different approach. We operationalize the majoritarian imperative as the likelihood that a party will be a member of a majoritarian coalition. We do so by computing, for each formation opportunity (which is characterized by a distribution of seats), the number of majoritarian coalitions party Pi would belong to, divided by the number of majoritarian coalitions that can be formed at that opportunity. Thus, majoritarian imperative ranges from 0 to 1, that is, from a situation in which a party has no chances of being in a majoritarian coalition, to a situation when a party will be part of every majoritarian coalition that is formed at a given opportunity. Compared to the standard seat share and largest party indicators, our measure presents two advantages: on the one hand it provides information about each party s legislative size relative to other parties in parliament; while on the other hand it reveals how likely each party is to form a majority government. Of the 415 formation opportunities we observe, 151 happened with a directly elected president and 264 with an indirectly elected president; 260 happened with a pro- active president and 155 with a passive president. To code the president s political affiliation we used the data provided by the World Political Leaders database (Zarate s Political Collections 2015). Heads of state who contested presidential elections as independents were coded as politically non- affiliated even if their candidature was eventually formally endorsed by one or more political parties. We start by examining the role of presidents in general, then distinguishing each of these types separately, and in combination. In estimating the power of the president in choosing the prime minister, we control by a number of factors. Following the existing literature, we include an 10
indicator variable for the party of the prime minister in the government immediately preceding the formation attempt and the party that occupies the median position in the parliament. Most of the data for the latter variable comes from the European Representative Democracy Data Archive (ERD) (Andersson et al. 2014), which identifies the left- right placement of political parties based on Benoit and Laver s (2006) expert survey on left- right positions in 47 countries. For the cases where this data is missing, we used the ParlGov party positions on the left- right dimension (which are based on various party expert surveys Castles/Mair (1983), Huber/Inglehart (1995), Ray (1999), Benoit/Laver (2006) and CHESS (2010) to determine the largest party closest to the median voter ideology following the procedure outlined by Kim and Fording (1998: 98 99). Following the ERD coding principles, we coded the median party as part of cabinet in cases where there were two median parties and one of them was in the cabinet. Using this algorithm to identify the median parties for the ERD data produced identical results, proving the robustness of our results. The literature also distinguishes between formation opportunities that happen after an election and those that happen in the middle of the legislative term. We do the same and include a variable indicating post- election formation opportunities, interacted with the majoritarian imperative. If there is a difference between formation in each of these circumstances, it should manifest itself by modifying what would have happened under the pure theory of parliamentarism. Because the chances that a party will hold the premiership may increase if that party did well in the election, we control for each party s electoral performance by including the change in seat share between two elections. Since this effect should only be observed in opportunities that occur after an election, we interact this variable with the indicator for post- election formation opportunities. Because we are interested in assessing the effect of the variables related to the president on the majoritarian imperative, the coefficients are of little use. We present the results through a series of graphs, which allow for a better understanding of what the effects are, if any. The coefficients are presented the appendix. 11
Results Figure 1 presents the results of the conditional logit estimate of the impact of being the party of the president at different points in the majoritarian imperative. The first thing to observe is that, almost for the entire range of the majoritarian imperative, a party receives a boost in the probability that it will hold the premiership when it is the party of the president. However, this boost is indistinguishable from zero when the probability of being in a majority coalition is higher than about 0.30. In other words, for parties that have a small to moderate chance of being in a majority coalition, being the party of the president makes a difference for its chances of holding the premiership. After this, there is little that the president can do to help his or her co- partisans beyond what they would already accomplish under pure parliamentarism. Figure 1 Probability that a Party Will Be the Party of the PM Presidential and Non- Presidential Parties Predictive Margins of prparty2 with 95% CIs Pr(Prime_Minister Fixed Effect Is 0).6.7.8.9 1 0.05.1.15.2.25.3.35.4.45.5.55.6.65.7.75.8.85.9.95 1 pm_party_probability prparty2=0 prparty2=1 12
Thus, presidents attenuate the operation of parliamentarism by reducing the effect of the majoritarian imperative on the choice of the party of the prime minister. For example, under the majoritarian imperative, a party with a 15% chance of being in a majority coalition has a 64% chance of holding the premiership. If the chance of being in a majority coalition increases to 35%, that party will have a 76% of being the party of the prime minister, an increase of 12 percentage points. If that party were the party of the president, the increase would have been of only 3 percentage points. Figure 2: Separating the cases of Direct Presidential Election Predictive Margins of prparty_dir with 95% CIs Pr(Prime_Minister Fixed Effect Is 0).2.4.6.8 1 0.05.1.15.2.25.3.35.4.45.5.55.6.65.7.75.8.85.9.95 1 pm_party_probability prparty_dir=0 prparty_dir=1 13
Figure 3: Separating the cases of Active Presidents Predictive Margins of prpart_proact with 95% CIs Pr(Prime_Minister Fixed Effect Is 0).5.6.7.8.9 1 0.05.1.15.2.25.3.35.4.45.5.55.6.65.7.75.8.85.9.95 1 pm_party_probability prpart_proact=0 prpart_proact=1 Figure 4: Separating Directly Elected Active Presidents Predictive Margins of prparty_dir_proact with 95% CIs Pr(Prime_Minister Fixed Effect Is 0) 0.2.4.6.8 1 0.05.1.15.2.25.3.35.4.45.5.55.6.65.7.75.8.85.9.95 1 pm_party_probability prparty_dir_proact=0 prparty_dir_proact=1 14
Conclusions Presidents, of any type, provide a boost to their own party in giving them a chance of holding the premiership when that party s chances of holding it under the pure operation of parliamentarism is small to moderate. Those situations are the ones where there are at least two coalitions that could reasonably hold a majority and the president chooses according to his or her own political preferences. This is a significant alteration of the majoritarian imperative under which parliamentarism allegedly operates, but it does not represent, in our view, a distortion of that imperative. That would be the case if presidents were to choose their own party to hold the premiership, even if some other party were clearly better positioned to do so by virtue of its likelihood of being in a majoritarian coalition. According to our data, however, this does not happen. Directly elected presidents do not favour their parties any more than indirectly elected presidents. The idea that presidents who are elected by voters will have a stronger base from which to influence the choice of prime minister is not supported by our data. This idea assumes, to begin with, that presidents will necessarily want to nominate prime ministers from his or her own party. While this is a sensible assumption, it is also plausible that presidents will be constrained by the parties that participated in the coalition that successfully helped conduct them to power, and thus may feel compelled to reward an allied party and not his or her own with the premiership. Further analysis should seek to uncover whether there is some truth to this possibility. Additionally, further analysis should investigate the relationship between the timing of the government formation and the presidential elections. But it can also be simply that directly elected presidents behave in the same way as their indirectly elected colleagues and our findings constitute just another piece of evidence that semi- presidential systems are just like regular parliamentary ones. Finally, the constitutional powers of presidents to nominate the formateur do not seem to add anything to what presidents actually do when it comes to choosing the party of the prime minister. We caution that our finding may be due to coding problems with the variable, something that we are working to fix. But it could also be 15
that pro- active presidents do what they are supposed to do under the majoritarian imperative. In the end, our findings suggest that there are circumstances under which presidents do exercise discretion in choosing the party of the prime minister; under these circumstances, they will prefer to pick their own party over other parties. This is true for all presidents in parliamentary systems, whether they are elected by voters or by the parliament, and whether the constitution formally allows them to nominate the formateur. Importantly, presidents do not go against what the theory of parliamentarism would expect them to do, namely, to choose the party that holds the highest chance of belonging to a legislative majority. 16
References Andersson, Staffan, Torbjörn Bergman, and Svante Ersson (2014). The European Representative Democracy Data Archive, Release 3. Main sponsor: Riksbankens Jubileumsfond (In2007-0149:1- E). [www.erdda.se].". Bäck, Hanna, and Patrick Dumont (2008). "Making the first move: A two- stage analysis of the role of formateurs in parliamentary government formation", Public Choice, 135, 353 73. Benoit, Kenneth, and Michael Laver (2006). Party Policy in Modern Democracies. London: Routledge. Castles, Frances G., and Peter Mair (1983). "Left- right political scales: some expert judgements", European Journal of Political Research, 12:1, 147 57. Diermeier, Daniel, and Antonio Merlo (2004). "An empirical investigation of coalitional bargaining procedures", Journal of Public Economics, 88:3-4, 783 97. Döring, Holger, and Philip Manow (2015). Parliaments and governments database (ParlGov): Information on parties, elections and cabinets in modern democracies. Development version. Elgie, Robert (2008). "The Perils of Semi- Presidentialism. Are They Exaggerated?", Democratization, 15:1, 49 66. Elgie, Robert, Sophia Moestrup, and Yu- Shan Wu, eds. (2011). Semi- Presidentialism and Democracy. London: Palgrave Macmillan. Glasgow, Garrett, Matt Golder, and Sona N. Golder (2011). "Who Wins? Determining the Party of the Prime Minister", American Journal of Political Science, 55:4, 937 54. Hooghe, Liesbet, Ryan Bakker, Anna Brigevich, Catherine de Vries, Erica Edwards, Gary Marks, Jan Rovny, and Marco Steenbergen (2010). "Reliability and Validity of Measuring Party Positions: The Chapel Hill Expert Surveys of 2002 and 2006", European Journal of Political Research, 49, 684 703. Huber, J., and R. Inglehart (1995). "Expert interpretations of party space and party locations in 42 societies", Party Politics, 1, 73 111. Kang, Shin- Goo (2009). "The influence of presidential heads of state on government formation in European democracies: Empirical evidence", European Journal of Political Research, 48:4, 543 72. Kim, Hee Min, and Richard C Fording (1998). "Voter ideology in western democracies, 1946-1989", European Journal of Political Research, 33:1, 73 97. Martin, Lanny W., and Randolph T. Stevenson (2001). "Government Formation Parliamentary Democracies", American Journal of Political Science, 45:1, 33 50. McFadden, Daniel (1974). "Conditional Logit Analysis of Qualitative Choice Behavior", in P. Zarembka (ed.), Frontiers in Econometrics. New York: Academic Press, 105 42. Müller, Wolfgang C., and Kaare Strøm (2000). "Coalition Governments in Western Europe: 17
An Introduction", in Wolfgang C. Müller and Kaare Strøm (eds.), Coalition Governments in Western Europeovernments in Western Europe. New York: Oxford University Press, 1 31. Protsyk, Oleh (2005). "Prime ministers identity in semi- presidential regimes: Constitutional norms and cabinet formation outcomes", European Journal of Political Research, 44, 721 48. Rasch, Bjørn Erik, Shane Martin, and José Antonio Cheibub, eds. (2015). Parliaments and Government Formation. Unpacking Investiture Rules. Oxford, New York: Oxford University Press. Ray, Leonard (1999). "Measuring party orientations toward European integration: Results from an expert survey", European Journal of Political Research, 36:2, 283 306. Schleiter, Petra, and Edward Morgan- Jones (2009a). "Constitutional Power and Competing Risks: Monarchs, Presidents, Prime Ministers, and the Termination of East and West European Cabinets", American Political Science Review, 103:3, 496 512. Schleiter, Petra, and Edward Morgan- Jones (2009b). "Party government in Europe? Parliamentary and semi- presidential democracies compared", European Journal of Political Research, 48:5, 665 93. Schleiter, Petra, and Edward Morgan- Jones (2010). "Who s in Charge? Presidents, Assemblies, and the Political Control of Semipresidential Cabinets", Comparative Political Studies, 43, 1415 41. Sedelius, Thomas, and Joakim Ekman (2010). "Intra- executive Conflict and Cabinet Instability: Effects of Semi- presidentialism in Central and Eastern Europe", Government and Opposition, 45:4, 505 30. Sedelius, Thomas, and Olga Mashtaler (2013). "Two decades of semi- presidentialism : issues of intra- executive conflict in Central and Eastern Europe 1991 2011", East European Politics, 29:2, 109 34. Shugart, Matthew Søberg, and John M. Carey (1992). Presidents and assemblies: constitutional design and electoral dynamics. Cambridge: Cambridge University Press. Strøm, Kaare (1990). Minority Government and Majority Rule. Cambridge, New York: Cambridge University Press. Zarate s Political Collections (2015). World Political Leaders, 1945-2015. http://zarate.eu/countries.htm. 18
19