Category Archives: Presidentialism

Miguel Carreras – The Rise of Political Newcomers in Presidential Systems

This is a guest post by Miguel Carreras at UC, Riverside. It is based on a forthcoming article in European Journal of Political Research

In the wake of the euphoria generated by the Third Wave of democratization during the 1980s, a group of scholars studying Latin America were more pessimistic about the prospects for democratic consolidation of the countries in the region. These scholars argued that there were a series of “perils of presidentialism” that created obstacles for the healthy functioning of democratic regimes in countries with presidential systems. Among the main perils of presidentialism, these scholars mentioned the dual democratic legitimacy, the temporal rigidity of presidentialism, the winner take all logic of presidential elections, and the principle of non-reelection (Lijphart, 1992; Linz, 1990, 1994). Since the early 1990s, several scholars of political institutions and Latin American politics tested most of these claims. The current consensus is that these perils of presidentialism were greatly exaggerated in these early studies (Carreras, 2012). However, the critics of presidentialism also claimed that the rise of political newcomers is a peril associated with presidential systems. This issue has been neglected until recently, and the main implication –i.e. newcomers are more likely to come to power in presidential systems– has never been tested empirically.

In a forthcoming article in the European Journal of Political Research (Carreras, fortcoming), I take on that task and I analyze whether the election of political newcomers is more likely in presidential systems. In my work, I define political newcomers in national executive elections as “candidates who lack substantial political experience in the legislative or the executive branches of government.” As for the operationalization, heads of government are considered as “political newcomers” when they had at most three years of political experience before reaching office –combining executive and legislative experience.[1] Using this definition and operationalization, I have identified 73 political newcomers elected (or selected) as heads of government following national elections around the world in the period 1945-2015. The sample includes 870 democratic national elections around the world. In other words, more than 8% of national elections in democratic countries have led to the election of a political newcomer as head of government.[2]

I assessed the impact of presidentialism on the success of political newcomers in national elections by estimating a series of random effects logistic regressions (this estimator is appropriate because the dependent variable is binary –1 if the elected head of government is a newcomer, 0 otherwise–). I also controlled for several other factors that might be related with the rise of political newcomers according to previous research (party system stability, economic performance, age of democracy, quality of democracy, and compulsory voting). The results of the main empirical model in the paper are presented below.

The empirical analysis demonstrates that Linz and the other critics of presidentialism were right about this particular claim. The variable “presidentialism” is positive and statistically significant in the statistical analysis, and this result is robust under different specifications and different operationalizations of the dependent variable. It appears that the personalized nature of presidential elections indeed facilitates the rise of politically inexperienced outsiders. But how exactly can presidentialism lead to the rise of political newcomers? I postulate that there are three causal mechanisms that may explain the connection between presidentialism and the electoral success of political newcomers. First, the organizational efforts that are necessary for leaders to become contenders for the top executive position differ significantly in presidential and parliamentary democracies. Political newcomers need to create a formidable party organization and have to recruit viable legislative candidates in many districts in order to have a chance to become prime ministers. Politically inexperienced candidates in presidential elections do not face equally insurmountable obstacles. Presidential elections are much more personalized, and political newcomers may win with very little support in the legislature (and without the support of any traditional party), especially in moments of deep economic and sociopolitical crisis that create a loss of confidence in the political establishment.

The second, and related, factor is the impossibility of popular inexperienced candidates to transmit their charisma or popularity. The deep popular dissatisfaction with the political establishment tends to be embodied by one or a few political leaders. Legislative candidates may ride on the coattails of very popular political newcomers irrespective of the type of political system, but the probability of them winning is always lower than the probability the charismatic candidate has of obtaining an electoral victory. Thus, in parliamentary systems the probability of an allied legislator winning a seat is always lower than the probability of the political newcomer winning a seat. In presidential systems, a charismatic neophyte candidate may become the president even if the party represented by the newcomer obtains poor results in legislative elections.

The third factor is the possibility to split the ticket in presidential elections. In presidential systems, voters normally have the possibility to vote for a legislative candidate of one party and for the presidential candidate of another party. Sometimes, this leads to a high discrepancy between the votes received by a party in concurrent legislative and presidential elections (Ames, Baker, & Renno, 2009; Helmke, 2009). The possibility to split the vote facilitates the election of a political newcomer in presidential systems, because it allows ambitious politically inexperienced public figures to run in presidential elections with a new party or a new electoral movement. These candidates may win, even if they are not associated with a single legislative candidate.


[1] The empirical results do not change if we adopt a more restrictive operationalization of “political newcomer.”

[2] The list of newcomer presidents and prime ministers  in the period 1945-2015 is  available in the supplementary information in the EJPR website:


Ames, B., Baker, A., & Renno, L. R. (2009). Split-ticket voting as the rule: Voters and permanent divided government in Brazil. Electoral Studies, 28(1), 8-20.

Carreras, M. (2012). The Evolution of the Study of Executive-Legislative Relations in Latin America: Or How Theory Slowly Catches Up with Reality. Revista Ibero-Americana de Estudos Legislativos(2), 20-26.

Carreras, M. (fortcoming). Institutions, governmental performance and the rise of political newcomers. European Journal of Political Research.

Helmke, G. (2009). Ticket splitting as electoral insurance: The Mexico 2000 elections. Electoral Studies, 28(1), 70-78.

Lijphart, A. (1992). Introduction Parliamentary versus Presidential Government. New York: Oxford University Press.

Linz, J. J. (1990). The Perils of Presidentialism. Journal of Democracy, 1(1), 51-69.

Linz, J. J. (1994). Presidential or Parliamentary Democracy: Does it Make a Difference? In J. J. Linz & A. Valenzuela (Eds.), The Failure of Presidential Democracy. Baltimore: Johns Hopkins University Press.

New book series – Palgrave Studies in Presidential Politics: Robert Elgie and Gianluca Passarelli (series editors)

We are announcing a new book series, Palgrave Studies in Presidential Politics. The series is edited by Robert Elgie and Gianluca Passarelli and the books will be published by Palgrave Macmillan. The series will include books on all aspects of presidential politics. We are currently accepting proposals for books in the series. The first volume, authored Philipp Köker, will be published in 2017.

Palgrave Studies in Presidential Politics will include books on contemporary presidencies, including presidential powers, the administrative presidency, and presidential advisers, as well as the history of presidential offices, and presidential biographies. The series will also include books on presidential elections, including presidential party politics, and the media and presidential communication.

The series will focus on presidents throughout the world including the US, Latin America, Europe, Africa, and Asia, including both directly elected and indirectly elected presidents. The series will publish single-country and comparative studies of presidential politics. The series will also publish books on individual presidents. The series will focus primarily on empirical studies of presidential politics, but it could include volumes on conceptual or theoretical aspects, such as how to measure presidential power.

The series will publish books that look at the reform of presidential politics, e.g. the reform of presidential elections. However, it will not publish obviously partisan, clearly normative, or personally critical studies of presidents or presidential politics. The series will have a disinterested, academic focus.

The series will normally take the form of 80,000-word monographs, or edited volumes. However, shorter books, or Palgrave Pivots, will also be considered. To submit a proposal, you should complete a proposal form. These are available from Ambra Finotello (, or from the series editors.

For further information about the series and to submit a proposal for consideration, please contact Ambra Finotello ( at Palgrave, or the series editors, Robert Elgie (, and Gianluca Passarelli (

Feel free to send an informal e-mail to the series editors if you wish to discuss a book idea prior to the formal submission of a proposal. We look forward to hearing your ideas for books and to receiving your submissions.

Victor Araújo, Thiago Silva, and Marcelo Vieira – New Perspectives on Executive Decision-Making Processes in Presidential Systems

This is a guest post by Victor Araújo, Thiago Silva, and Marcelo Vieira. It is based on their paper ‘Measuring Presidential Dominance over Cabinets in Presidential Systems: Constitutional Design and Power Sharing’ that was recently published in the Brazilian Political Science Review


The third wave of democratization approaches its 40th anniversary in Latin America in 2018. The process started with the Ecuadorian Constitution of 1978, promulgated after the withdrawal of the military power in the previous year. Like Ecuador, most Latin American countries opted for a presidential system of government in their new journey for consolidation of a democratic political regime.

From the late 1970s to the late 1980s, comparativists wrote extensively on the contrast between presidential and parliamentary systems, rushing to conclusions and predictions about the undesirable institutional choices of Latin American democracies. In the search for predictions, assumptions were taken as evidence and black boxes remained intact, undermining our ability to understand the real differences and similarities between systems of government, particularly among presidential systems.

The different criteria suggested by the literature of the 1980s and 1990s for the definition of presidential and parliamentary systems are well known. However, it is worth mentioning one aspect of it here developed by Arend Lijphart and sustained by Giovanni Sartori. For these authors, the executive power in presidential systems has a sole character: being the head of government and the head of the state, the president would be the only relevant actor in the executive decision-making process in these systems. The idea is that the president derives a dominant role in the executive decision-making process from its discretionary powers of selection and ministerial removal, in a context in which his/her tenure is independent from parliamentary confidence. The premise suggested by Lijphart, taken for granted by scholars of systems of government, is that the vertical nature of presidential cabinets contrasts with the horizontal character of most parliamentary cabinets (i.e., primus inter pares cabinets in Sartori’s definition).

In a recent article published in the Brazilian Political Science Review, we challenge the studies mentioned above, arguing that the assumption of a vertical executive decision-making process in presidential systems underrates variations that may exist regarding the degree of dominance exercised by the chief executive over cabinets. Although conventionally characterized as a non-collegial decision-making process, led by the president, we reveal that the sole executive of presidential systems is not a distinctive feature of this system of government. Instead of analyzing the process of executive decision-making in parliamentary and presidential democracies dichotomously—based on collegial and non-collegial processes, respectively—we should, as suggested by Vercesi (2012) evaluate the possibility that the power sharing of the executive decision-making process systems varies continuously across systems of government.

Building on this idea, our study develops an index to help scholars analyze how powerful the president is vis-à-vis cabinet members, according to specific constitutional rules that verticalize or horizontalize the executive decision-making process in presidential democracies. More precisely, we created a summation index of codified constitutional rules regarding executive powers raging from 0 to 5, with a value of 0 indicating the absence of presidential dominance over cabinets, and the value of 5 indicating the absolute dominance of the president over cabinets.


As we can see in Figure 1, our index reveals significant variation of presidential dominance over cabinets in Latin America presidential constitutions. We inferred at least three clear standards. Presidential democracies can have: a decentralized (e.g. Bolivia, Argentina and Uruguay), a shared (e.g. Chile, Colombia and Costa Rica), or a centralized (e.g. Brazil, Mexico and Panama) executive decision-making process. In sum, the executive decision-making processes in presidential systems are not necessarily vertical, and the presidential powers over ministers are not necessarily unrestricted or unbounded.

These results have important implications for the literature on comparative politics comparing systems of government. First, our study strengthens the argument suggested by Elgie (1997) that the power sharing within executive decision-making processes should be evaluated empirically and not assumed by scholars. Our results for Latin America reveal that, regarding executive decision-making process, there are more similarities between different systems of governments than conventionally assumed by the literature. In another study, by analyzing more than 50 countries we also show that the degree of dominance of the chief executive in presidential and parliamentary democracies can be very similar.

Second, our study reveals that the costs of presidential decision-making may vary depending on other factors such as institutional rules that restrict the “selection” or “removal” of ministers, a dimension not systematically explored in the literature yet. A more careful analysis of this topic could explore, for example, how ministerial survival rates can be affected by constitutional rules that determine the distribution of authority among the members of the cabinet.

Third, our findings can lead to exploration of new topics such as ministers as veto players in presidential systems. In contexts where the president depends on the consent of ministers to be able to propose policies, the stability of the coalition might depend on the chief executive considering the preferences of the parties of different ministers. We could ask, for example, how pivotal are minister’s parties in maintaining the presidential coalition in the legislature.

Fourth, given that the distribution of resources for policy-making in the cabinet is directly related to the degree of influence the parties can exert on the composition of the coalition, our results encourage scholars to investigate what preferences are being expressed in the policy-making process, instead of assuming that policies express only the preference or the agenda of the president.

In sum, the results from our study on how powerful the president is vis-à-vis cabinet members lead us to ask several new questions and hypotheses worthy of further empirical investigation, such as: How powerful are the ministers in presidential systems? How can cabinet veto players affect ministers reshuffle and cabinet survival? How does the distribution of authority between the chief executive and cabinet members affect the content of the executive policy agenda?


Victor Araújo is a PhD candidate in the Department of Political Science at the University of São Paulo, Brazil.

Thiago Silva is a PhD candidate in the Department of Political Science at Texas A&M University, United States.

Marcelo Vieira is a Professor in the Department of Social Sciences at the Federal University of Espírito Santo, Brazil.

Seychelles – President Michel announces resignation after first election victory for the opposition in 40 years

After having dominated Seychellois politics for nearly 40 years, on September 10th the Seychelles People’s Party (Parti Lepep in Creole) lost the parliamentary elections in the small Indian Ocean archipelago. The election victory of the opposition (the Seychelles Democratic Alliance – Linyon Demokratik Seselwa) constitutes a dramatic political shift in the country, and can be regarded as a major step in its development towards a consolidated democracy. The recent parliamentary elections were preceded by a presidential election in December 2015, which produced a razor-thin victory (50,15% of votes) for the incumbent president, James Alix Michel. After his party’s defeat in the parliamentary elections, Michel – who has been president of Seychelles since 2004 – announced his resignation, and his intention to hand over powers to the vice president, Danny Faure. The resignation of the 74-year old president – announced via a surprise address on national television – came as a major shock to the island nation.

The People’s Party and its predecessor, the Seychelles Peoples Patriotic Front (SPPF), have controlled Seychellois politics for virtually the entire post-independence period. Seychelles obtained independence from the United Kingdom in 1976, but within a year after the attainment of sovereignty, the Marxist SPPF, headed by its charismatic leader France-Albert René, staged a successful coup d’état with help from socialist African countries like Libya and Tanzania. Between 1977 and 1993, Seychelles was ruled as a single-party dictatorship, and human rights abuses were commonplace. The fall of the Berlin Wall and the collapse of the Soviet Union instigated a transition to multi-party democracy, and Seychelles has been an electoral democracy since the early 1990s. However, the SPPF and René continued to win all subsequent elections, and international observers remained very critical of Seychellois democracy, highlighting the political fusion between the state and the ruling party as particularly problematic. Under René’s guidance Seychelles reached an impressive level of economic growth, as a result of which it is now one of the wealthiest countries of Africa. This economic success undoubtedly contributed to the continuing support for the SPPF, also after René transferred powers to Michel in 2004, and the SPPF was transformed into the People’s Party.

While the political opposition had boycotted the penultimate parliamentary elections in 2011 due to concerns about unequal campaign finance, in advance to last September’s election four opposition parties joined forces in a political alliance (the Seychelles Democratic Alliance), which managed to win 19 of the 33 parliamentary seats. Since the presidency of the country remains in the hands of the People’s Party, the tiny archipelago now enters a period of divided government. Both the newly designated president and the political leadership of Seychelles Democratic Alliance have declared their intention to cooperate with each other, but how this cooperation will play out in the coming years remains to be seen. While the two political parties are ideologically similar and do not have many disagreements regarding economic policies, the relationship between them has been marked by strong and sometimes personal antagonism.

Sun Woo Lee – The Politics of Prosecution Service Reform in New Presidential Democracies

This is a guest post by Sun-Woo Lee, a post-doctoral researcher of the College of Social Science, Seoul National University. It is a summary of his paper ‘The politics of prosecution service reform in new presidential democracies: The South Korea and Russia cases in comparative perspective’ that is being published by the Journal of Eurasian Studies.

Recently, the political distortion of criminal justice by prosecutors has arisen as a serious dilemma in several new democracies, though this issue has received somewhat less attention than the judicialization of politics by courts. Indeed, prosecutors of civil-law countries must receive more attention in that they generally initiate criminal investigations, command police officers during the investigations, terminate the criminal cases at their disposal, and indict the criminal suspects for trials in the centralized and inquisitorial criminal justice system (e.g. Germany, Japan, Taiwan and South Korea). Thus civil-law prosecutors can exercise enough power to manipulate pre-trial criminal proceedings, not only in order to stigmatize a certain political faction as immoral or criminal suspect, but also to grant immunity to another.

Of course it cannot be concluded that civil-law prosecutors always have chances to distort criminal proceedings for political purposes. Under consensual forms of government in Continental Europe, a suprapartisan coalition has been required to select top-ranking judicial officers. Moreover, the composition of an incumbent government unpredictedly changes when an assembly dissolves, regardless of regular elections, or when a shift of power occurs within a ruling party. Hence, civil-law prosecutors also tend to exercise their far-reaching power not in favor of a particular political faction, but in a depoliticized manner, for their career development.

By contrast, in several young democracies adopting a presidential system, which gave the president almost exclusive control over high-ranking prosecutors’ career, along with the Third Wave of democratization, civil-law prosecutors have a strong incentive to exercise their extensive power in favor of an incumbent president during most of his or her fixed tenure, but to betray him or her at his or her last phase, for their career progress. In practice, Alberto Fujimori in Peru, Chen Shui-bian in Taiwan, Young-Sam Kim and Dae-Jung Kim in South Korea and Boris Yeltsin in Russia dominated that office during most of their tenure but experienced prosecutorial defection at their final phase. However, an interesting point is that a civil-law prosecution system could hardly be reformed, although there were several attempts to correct the dilemma of the politicized prosecutors, in the new presidential democracies. Why?

In presidential democracies, civil-law prosecutors can enjoy a favorable political opportunity for deflecting reform against them and protecting their powers and position, even though politicians actually attempt such a change. An incumbent president will at times pursue a long-term interest in reform against the prosecutors, particularly owing to the possibility of their habitual betrayal before and after his or her retirement. Nonetheless, the prosecutors can easily defeat the president’s reform either by lobbying or placating the president, given that he or she is bound to fear political losses resulting from their defection. In fact, unless the prosecutors guard an incumbent president, he or she may become the most vulnerable political actor to the politicization of criminal justice, considering the probability that the president would be involved in corruption is much higher than for any other politician. Thus an incumbent president is likely to abandon prosecution service reform readily or be satisfied only with small achievements, but to seek his or her short-term interest in forming alliance with this power agency. In this usual circumstance, opposition parties would more often have the long-term interest in prosecution service reform than the president. Nevertheless, as long as the alliance between the president and prosecutors is sustained, the opposition forces will be persecuted and therefore cannot accomplish the reform.

However, if an incumbent president dares to launch major reform against civil-law prosecutors repeatedly, they are likely to resist his or her attempts vigorously. This is clearly a rational option for the prosecutors, because the cost his or her reform would impose on them is heavy, whereas the effect their resistance could provide is great, because of the political opportunity structure favorable for them. Since curtailment of the prosecutors’ great powers as prerogatives runs sharply against their collective interest, all the members can be motivated to engage in collective action. Moreover, the high-ranking prosecutors’ loss, when becoming a traitor to their own organization, is no less than their individual benefit from career advancement in exchange for their loyalty to the reformist president. On the other side, an incumbent president in fact has no way to protect his or her faction members from the prosecutors’ willful investigation or indictment. In addition, opposition parties which have pursued to reform the prosecution system would also reverse their previous position in order to take advantage of the conflict between the president and prosecutors. That is, they would no longer have a strong incentive to support prosecution service reform, given that political scandals involving the president’s faction are being intentionally disclosed by the agency. Then vigorous competition among the media to get the first reports of political scandals could also enhance the prosecutors’ political influence.

Indeed, this logical development may not spontaneously be applied to the prosecutorial reform cases of consensual parliamentary countries. A coalition government relying on parliamentary majority can more easily make an agreement between various political forces, on the curtailment of extensive power of the prosecution service, and therefore enforce the agency to abandon its prerogatives, while the government would essentially have relatively less incentive to attempt such a reform, as noted above. By contrast, an incumbent president would be unable to maintain the momentum for large-scale reform against the prosecutors, encountering their strong resistance, and correspondingly the latter could succeed in protecting their privileges from the former.

Even in presidential countries, nonetheless, the political opportunity enabling the prosecutors to make effective resistance to a reformist president would be gone, if the president becomes no longer vulnerable to the political manipulation of criminal justice. We can assume that a hegemonic party would dominate domestic politics, or a semi-authoritarian regime would emerge in new democracies. Then political competition would be significantly weakened for quite a while. But an important point is that even a semi-authoritarian president would pursue the long-term interest in curtailing the power of civil-law prosecutors and consequently in making them politically impartial, for fear of the time whenever he or she loses his or her own hegemony, as long as regular elections are formally held. Indeed, the more power an authoritarian ruler holds in his or her hands, the more interests he or she has in undermining any power base for future challengers. Under the weak political competition in a semi-authoritarian regime, however, an incumbent president can succeed in the prosecution service reform, differently from the case of competitive democracies. This is because the president would become much freer from losing moral foundations and public support than in competitive electoral democracies, even if the prosecutors intentionally institute criminal proceedings against his or her faction members. As a consequence, the prosecutors are likely to abandon their resistance against the president, and correspondingly to accept even unfavorable changes imposed on them. In June 2007, this occurred in Russia.

Although more empirical research on other countries would be beneficial, this means that a civil-law or inquisitorial country adopting presidentialism along with its democratization, “ironically,” may not be successful in correcting the dilemma of the politicization of the prosecution service until a semi-authoritarian regime reappears.

Sun-Woo Lee obtained PhD (Politics) from University of Glasgow (UK) in 2014. And he is currently a post-doctoral researcher of the College of Social Science, Seoul National University. His research interests are comparative politics, Russian politics, Korean politics and Northeast Asian international relations. He has published several papers in many academic journals, such as Development and Society, and Journal of Politics and Law

Erdoğan’s Long-Standing Struggle for a Turkish Type of Presidential System

Constitutionally Turkey is a semi-presidential country with a president whose constitutional powers are more than ceremonial but less than executive. Despite having few constitutional powers with which to check and balance the Council of Ministers, in reality President Erdoğan is an executive president who can control foreign and internal policy choices. Being the founder and the real leader of the ruling AKP, President Erdoğan has managed to compensate for what he lacks constitutionally by his de facto position. Despite the opposition’s reminders that according to the Constitution President Erdoğan should act impartially and that he has no legal powers to involve himself in day-to-day politics or to decide Turkish foreign policy, Erdoğan seems fully in control of his party and the government.

Yet President Erdoğan is still campaigning for a new presidential constitution. Since the AKP’s overwhelming win in the November general election that consolidated its predominant position in the system, Erdoğan has returned to his campaign for a so-called Turkish type of presidential system. This raises three questions. Firstly, why does the president insist on a constitutional change to a presidential system since he can already control every aspect of the government? Secondly, what is a Turkish type of presidential system? And is a new constitution going to come about in Turkey’s intensely polarised political climate?

The answers to the first question differ greatly depending on who is being asked. The president himself claims that a presidential system would stop the double-headedness within the executive which he often complains about despite the fact that he handpicked Prime Minister Davutoğlu. In a meeting with NGOs supporting his campaign, the President argued that an elected president cannot work with an elected Prime Minister, especially if they are from different political backgrounds. He thinks that a prime minister from a different background  might be elected in the future and that this would create tremendous discord within the government. For that reason, precautions should be taken against it now in the form of a new presidential constitution .

This should be an argument against all forms of semi-presidentialism, but President Erdoğan says that it is an argument against a parliamentary system. At one point he even talked of the “French Model” being a positive example, even though the French experience would seem to contradict his argument.

The second argument that the president uses is related to the first one. He claims that a presidential system would create “absolute stability” and prevent a “bureaucratic oligarchy” from implementing legislation and regulations. He says that Turkey needs restructuring, that laws and regulations would prevent it, so he has to be brave and set them aside .In order to completely restructure the system, Turkey must adopt presidential system which would bring absolute stability.

President also emphasises that it is not in favour of a separation of powers. He describes the system he defends as a Turkish type of presidentialism with a harmony of powers, rather than a system of checks and balances. He often complains that the current system is based on a conflict between the judiciary and government (meaning the executive and legislative majority). He argues that this system should be replaced by a system in which powers support each other. He perceives judicial review auto be an impediment, so often he refers to judicial review as being a problem that stems from a parliamentary system. Even though this is not an accurate, it illustrates what the president expects from or means by a Turkish type of presidential system.

In fact, any response given to the first question of why Erdoğan insists on a new presidential constitution also indirectly answers the second question of what Turkish type of presidential system he wants. Often opposition leaders or MPs express their fear that President Erdoğan wishes to become a super-president merging all state powers in a single office and eliminating any constitutional checks and balances as well as the alternation in power between political parties. For the opposition this is not a democratic model. All opposition parties oppose Erdoğan’s arguments for a presidential system and state that they are in favour of keeping alive the country’s parliamentary heritage, which goes back to 1908 albeit with certain changes to improve its efficiency as well as democracy and rule of law.

On the other hand, neither the AKP nor the President has so far produced a text showing the details of the system that they defend, except for the short text presented by the AKP to the former ad hoc parliamentary Commission of Constitutional Consensus which was dissolved in 2013 due to a failure to reach a consensus among participating political parties over the governmental system. This draft text gave strong legislative powers, like the power of decree, veto, initiating budget laws to the president and curbed judicial review. (See Şule Özsoy Boyunsuz, ‘The AKP’S proposal for a “Turkish type of presidentialism” in comparative context’, Turkish Studies’, DOI 10.1080/14683849.2015.1135064.

A new Constitutional Consensus Commission was formed a month ago in parliament under the chairmanship of the Speaker. It comprises three members from each of the four parliamentary groups and has been charged with penning a new constitution. After three meetings in February 2016 this ad hoc commission was dissolved by the speaker due to the disagreements over the presidential system, just like the previous constitutional consensus commission which was formed for the same purpose in 2011 and which was dissolved in 2013. The CHP, the main opposition party, declared that they will not discuss a presidential system as a viable alternative. The HDP and MHP, the other two opposition parties, refuse to form another commission without the participation of the CHP. So the answer to the question of how it is going to be possible to make a new constitution altering the regime remains largely unknown.

President Erdoğan announced that a new presidential constitution will be produced even if opposition parties do not sit in the Constitutional Consensus Commission and that it will then be submitted to a referendum for the public approval. However, This would require at least 14 votes in parliament from the opposition. That would mean fishing for opposition votes using any kind of methods or calling for an early election, which would be another way of changing the composition of parliament albeit one that runs the risk of losing more seats too.

Aside from legal and technical issues related to amending or making a new constitution, changing to a presidential system is a politically divisive topic in today’s highly polarised Turkish society. There is a climate of ongoing conflict between the PKK ( Kurdish separatist terrorist organisation) and the security forces in certain South Eastern cities that has claimed many lives on both sides and this is on top of the government’s increasing involvement in the Syrian war. The co-chair of the pro-Kurdish HDP, Yüksekdağ, has accused President Erdoğan of “opening the door to a very big war and chaos in the region (Syria) in order to become an executive president by becoming chief commander through a declaration of mobilization and martial law” . Indeed, when Turkish jets shot down a Russian warplane on the Syrian border, the polls showed the highest public support (53.5%) for presidential system. Yet it remains to be seen if the war will help the President realise his dream of a presidential constitution.

Gianluca Passarelli – The presidentialization of political parties: A new perspective

This is a guest post by Gianluca Passarelli, who is an Associate Professor in Political Science at the Department of Political Sciences, Sapienza University, Rome. He is the editor of a new book, The Presidentialization of Political Parties: Organizations, Institutions and Leaders, published by Palgrave Macmillan


What actually makes a president of the Republic a leader in (semi-)presidential regimes? And when, if ever, is it possible that a party leader, once he or she has become the head of government in a parliamentary regime, can come close to the style of leadership in similar cases in which the separation of powers exists? If the institutions influence the behavior of politicians, and thus of the parties, it is necessary to understand and explain if and in what way it is possible to refer to ‘presidentialized’ party organizations outside of the institutional context that defines its characteristics: the presidential regime.

The presidentialization of politics is a relatively new and important phenomenon. However, the term presidentialization has become highly debatable. In particular, the more contentious side is offered by the suggestion that presidentialization of politics could make (semi) presidential regimes and parliamentary ones more similar to presidentialism.

The present state of affairs, in very brief terms, sees those on the one side who maintain that it almost exclusively the institutions that influence, condition and make possible (or not) party ( and therefore also political) presidentialization. On the other, we find those who insist that political presidentialization – intended as a centralization of governmental, elective and party functions – is a verifiable “tendency” in practically all western democracies. For various reasons, we would argue that to this dichotomy can be added a variable, the component connected to the “nature” of the parties analyzed, which can contribute to spotlighting a phenomenon that is being widely discussed throughout the (not only academic) world. In light of these different research hypotheses, we seek to approach our analysis by flanking the party variable with the institutional one.

The aim of our research is to explain why the level of party presidentialization varies from one country to another. If it is true that that constitutional structures affect the level of party presidentialization, we are adding the party’s original features and arguing that the degree of party presidentialization varies as a function of the party’s genetics, that are the original organizational characteristics of each party.

The literature has mainly focused on the general process of personalization that has been detected in recent modern politics, especially in western democracies. Depending on the cases studied, on the research fields and the data availability (and reliability), the studies conducted have had different foci. The role of institutions, the characteristics of leadership and leaders, as well as the electoral process or the mass media influence, have been the main explanatory variables analyzed in order to explain the phenomenon of «presidentialization». Thus, the weakening of party loyalties, the kind of electoral system, the influence of mass media, and the form of government have in turn been considered as the independent variable, the factor that justifies the above mentioned phenomenon of a «presidentialization of politics». We place greater stress both on the concept of presidentialization, and on demonstrating empirical evidence of the phenomenon, if any such evidence exists. Indeed, the presidentialization of politics in our view means the presidentialization of parties, or better yet, a phenomenon that arises from the political parties’ behavior.

The causal trajectory is summarized as follows: constitutional structures affect party presidentialization through the medium of endogenous party factors. We consider the genetic model of organizational development (penetration v. diffusion), the characteristics of the dominant coalition (factions v. tendencies), as well the balance of power in the dominant coalition (central office v. public office). There are three factors concerning party genetic features: 1) the organization’s construction and development; 2) the presence or the absence (at the party’s origin) of an external “sponsor”; 3) the role of charisma in the party’s formation (Panebianco 1988, pp. 50-2).

We basically claim that differences between «personalization of politics» and «presidentialization of politics» essentially refer the fact that: a) the previous implies mainly considering a sort of personal «capital» in terms of skills, characteristics, attitudes, etc., while b) the latter consider primarily institutional resources, constraints and opportunities.

The results tend to confirm both of the starting hypotheses: a) the presidentialization of political parties is a phenomenon (and a concept with clear empirical “aspects”) that must inevitably arise in presidential regimes (and also in semi-presidential ones, under certain circumstances); b) the genetic characteristics of political parties function as an intervening variable capable of accentuating the opportunities offered by the institutions from a presidentialization perspective, or rather they circumvent restrictions and produce presidential effects, even if limited in time and space. Moreover, the results indicate an increasing level of presidentialization in parties (and thus indirectly in political systems) where the parties have original, genetic features capable of accentuating the institutional opportunities for such a development. Independent of the distinctions between regimes, «centralized» parties that are cohesive, disciplined, without factions and with a leadership that is «independent» from the organization (for extra-political or statutory resources) will be more suitable to increasing levels of presidentialization. Vice versa, political parties that have a fragmented and divided leadership, and are permeated by fractions, factions and ideological conflicts as well as by “decentralized” and diffused organizational structures, will have considerably greater difficulty in concentrating resources around a single individual. General results confirm our hypothesis about the importance and influence of the genetic characteristics of the parties and equally about the inescapable influence of the constitutional structure.

Gianluca Passarelli ( is Associate Professor in Political Science at the Department of Political Sciences, Sapienza University, Roma. He earned his PhD in Comparative and European Politics at the University of Siena, and he has been research fellow at the University of Bologna. His research interests lie in presidents of the Republic, political parties, electoral systems, elections and electoral behaviour. He has authored, co-authored, or edited books on presidents, political parties, and constitutional regimes. He has published in journals such as French Politics, South European Society and Politics, Contemporary Italian Politics and Political Geography.

Ryan E. Carlin and Shane P. Singh – Executive Power and Economic Accountability

This is a guest post by Ryan E. Carlin and Shane P. Singh from Georgia State University and the University of Georgia respectively

Accountability’s centrality to democratic theory makes it “one of the holy grails of empirical political science research” (Samuels and Hellwig 2010, 393). It refers to the public’s ability to “discern whether governments are acting in their interest and sanction them appropriately” (Manin, Przeworski, and Stokes 1999, 40) through elections, civil society, the media, protests, and public opinion. Are some executives held more accountable than others by design? Can they shape their own accountability by the decisions they make?

In a forthcoming article in the Journal of Politics, we posit that both presidents’ legislative powers and how frequently they resort to legislation by decree shape the extent to which the public holds them accountable for economic outcomes. Two forms of responsibility attributions connect presidential powers to accountability. “Role responsibility” concerns the obligations citizens expects presidents to fulfill and the consequences for failing to do so. Presidents endowed with vast lawmaking powers accrue great role responsibility for outcomes on their watch. “Causal responsibility” is doled out when presidents’ intentional actions are seen as proximate causes of economic outcomes. In proportion to the decree powers presidents use to legislate unilaterally, they will incur causal responsibility for the outcomes.

To test our expectations, we gathered data from the Latinobarometer on presidential approval, sociotropic economic retrospections, and several common controls from 94,861 respondents in 120 surveys from 18 Latin American presidential systems over the years 2002-2009. We tested whether the degree to which socioeconomic retrospections were related to presidential approval was a function of two measures of presidential power: an index of the legislative powers afforded to the president by the constitution and the total number of decrees and decree laws the president issued in the previous year, as measured with the Global Legal Information Database.

As predicted, economic evaluations influence the approval of powerful presidents more than their weaker counterparts. That is, economic accountability is calibrated to account for presidents’ legislative powers and prerogatives—their role responsibility for outcomes. Accountability also keeps pace with decree usage, a finding consistent with a mechanism of causal responsibility attribution. We can see this graphically through the marginal effects of economic evaluations on the probability of approving of the president according to legislative powers and the use of decrees as captured in Figure 1.

Screen Shot 2015-07-22 at 15.35.23

To put some flesh on these results, imagine an individual whose president has Legislative Powers that register in the 10th percentile of our sample—Honduras since 1982. Results illustrated in the left-hand panel of the figure suggest that a unit increase in retrospective economic evaluations corresponds with, on average, roughly a 12.5 percentage-point increase in the likelihood that she will approve of the president. If this same individual were transplanted to a country where the president’s Legislative Powers ranked in the 90th percentile—Brazil since 2001—a unit increase in economic retrospections should boost the likelihood that she will approve of the president by around 16.4 percentage points.

Turning to the link with decrees, illustrated in the right-hand panel of the figure, for a citizen whose president falls in the 10th percentile for decree issuance, a unit increase in economic evaluations corresponds with, on average, about a 13.5 percentage-point increase in the likelihood that she will approve of the president. If that citizen’s president falls in the 90th percentile for issuing decrees, the same increase precipitates around a 15.9 percentage-point rise in the chances of approving of the president.

So presidents who enjoy strong legislative power and who use it extensively invite greater economic accountability. Not only do these measures of role responsibility and causal responsibility influence economic accountability independently but, as Figure 2 shows, their effects are even greater when analyzed in interaction.

Screen Shot 2015-07-22 at 15.36.07

The public, thus, holds presidents to account for economic outcomes to a greater degree where their governing role responsibilities are vast and their causal responsibility is made more obvious by frequent use of such power. In other words, accountability for economic outcomes is strongest where presidents are strong legislators both de jure and de facto.

In presidential systems, executive power can be a blessing or a curse. When the public perceives a humming economy, strong presidents who use their legislative powers habitually can expect their popularity to grow multiplicatively. But when they feel the economy slumping, vast presidential powers compound the problem, sending support in a downward spiral. This may seem unfair, given the difficulty even strong presidents face in advancing their agendas (Saiegh 2011) and delivering outcomes citizens prefer (Mainwaring and Shugart 1997)—and the fact that congress may ask the president to act unilaterally in order to get things done (Carey and Shugart 1998). Nevertheless, we conclude that the more lawmaking power constitutions grant presidents, and the more they use these powers, the more responsibility they assume for economic policy outcomes in the public’s eye.

Our results weigh in on second-generation debates about presidentialism and democracy. In a thoughtful essay on why the “difficult combination” of presidentialism and multipartyism (Mainwaring 1993) has outstripped scholars’ predictions for durability and governability, Chaisty, Cheeseman, and Power (2014) credit presidents’ mastery of various power tools to build and manage multiparty coalitions. Yet the authors are quick to warn:

“The very same presidential tools that enhance governability may also undermine accountability – a tradeoff that has been largely overlooked even as analysts celebrate the survival of multiparty presidential democracy.” (86).

Our findings, however, suggest no tradeoff between the powers that facilitate governability and the vertical accountability mechanism of public opinion between elections. In fact, citizens appear able to calibrate responsibility attributions according to presidents’ possession and use (or even abuse) of lawmaking power. Thus, our preliminary diagnosis is that presidential powers pose no barrier to accountability at the level of mass politics.

But viewing these findings in light of our related research, blogged at Presidential Power earlier this year, renders a less sanguine conclusion. Democratic attitudes, the basis of regime legitimacy, are most positive under a middling—neither too high nor too low—degree of presidential powers (Singh and Carlin 2015). In other words, crafting an institutional framework to optimize both of these core democratic ideals under presidentialism may, therefore, prove rather challenging.


Carey, John M., and Shugart, Matthew Soberg, eds. 1998. Executive Decree Authority. New York: Cambridge University Press.

Chaisty, Paul, Cheeseman, Nic, and Power, Timothy. 2014. Rethinking the ‘Presidentialism Debate’: Conceptualizing Coalitional Politics in Cross-Regional Perspective. Democratization 21 (1):72-94.

Mainwaring, Scott. 1993. Presidentialism, Multipartism, and Democracy: The Difficult Combination. Comparative Political Studies 26 (2):198-228.

Mainwaring, Scott, and Shugart, Matthew Soberg, eds. 1997. Presidentialism and Democracy in Latin America. New York: Cambridge University Press.

Manin, Bernard, Przeworski, Adam, and Stokes, Susan Carol. 1999. Elections and Representation. In Democracy, Accountability, and Representation, eds. Przeworski, Adam, Stokes, Susan Carol and Manin, Bernard. Cambridge, U.K.; New York: Cambridge University Press, 29-54.

Saiegh, Sebastian M. 2011. Ruling by Statute: How Uncertainty and Vote-Buying Shape Lawmaking. New York: Cambridge University Press.

Samuels, David, and Hellwig, Timothy. 2010. Electoral Accountability and the Clarity of Responsibility: A Conceptual and Empirical Reassessment. Journal of Elections, Public Opinion, and Parties 20 (4):393-414.

Singh, Shane P., and Carlin, Ryan E. 2015. Happy Medium, Happy Citizens: Presidential Power and Democratic Regime Support. Political Research Quarterly 68 (1):3-17.

Ryan Carlinpolitical scienceRyan E. Carlin is an Associate Professor of Political Science and Director of the Center for Human Rights and Democracy at Georgia State University. He received his Ph.D. in 2008 from the University of North Carolina at Chapel Hill. His research interests are comparative political behavior and public opinion, with a regional emphasis on Latin America. His work has been funded by the National Science Foundation, USAID, the Mellon and Ford Foundations, and the Latin American Studies Association. He is co-editor of The Latin American Voter (University of Michigan Press, 2015) and his work has appeared in numerous academic journals and volumes. His website is found at

SinghShane P. Singh is an Associate Professor in the School of Public and International Affairs at the University of Georgia. He received his Ph.D. in 2009 from Michigan State University. His research focuses on the institutional and contextual foundations of political behavior and attitudes. His work has been funded by the Social Sciences and Humanities Research Council of Canada, and his research has appeared in many academic journals and edited volumes. His website is found at

Jean Blondel – The presidential idea

This is a guest post by Professor Jean Blondel, Professor Emeritus, EUI, Florence


In the course of the last few years, I became increasingly concerned with the apparent contradiction between the rapid development of regimes in which the role of the president is dominant and the impression that these regimes were at least very often giving way, at any rate temporarily, to what is conventionally regarded as ‘usurpation’. The ‘presidential republics’ which had emerged in the newly independent Spanish American countries in the first decade of the nineteenth century could be regarded as providing evidence for the view which was put forward by Linz and Valenzuala in 1994 in their two volume study, namely that, on the whole, ‘presidential democracy’ was a ‘failure’. The case of France’s ‘Second Republic’ of 1848-51 was also an example of such a ‘failure’ as the model was the America: thus only the successful duration of the US Constitution of 1787-9 made it impossible to adopt such a view as a ‘universal’ proposition. As a matter of fact, the fate of subsequent ‘presidential experiments’ in Europe in the interwar period seemed to confirm the validity of ‘pessimistic’ views, Finland having been the only European exception among them.

Yet this state of affairs did not prevent the multiplication of ‘presidential republics’ in Africa from the late 1950s to the 1970s and beyond; there was more reluctance in Asia and indeed in Europe as well, except for the fact that Gaullist France, alone in Western Europe, adopted presidentialism in 1958-62. Indeed, in the 1990s, a further boost for the model of the ‘presidential republic’ resulted from the collapse of the Soviet Union, as Yeltsin adopted the ‘model’ for the new Russia he created, a move which was closely followed by ten of the other eleven ex ’Soviet’ republics which did not join the European Union in 2004. A majority of the countries of the world had come to be presidential as a result (see Table 1 below)!

Table 1

World regimes in 2013 (countries of 100,000 inhabitants or more only)

Region Total Pres Parl. Rep Monarchies Usurp Communist Decentralised Unclassified
West/W. Europe 23 2 9 11 1
E. Europe in EU 11 2 9
E. Europe not in EU 7 3 3 1
Asia 39 11 10 13 1 3 1
Pacific 7 2 3 1 1
Africa 53 45 2 3 3
Amer. (not West) 30 21 1 7 1
Ex Soviet Union 11 11
Total 181 95 36 37 2 4 3 4

Details are given in the volume about the regimes adopted by individual countries in the various parts of the globe.

Even if many of the countries concerned (except in Europe) were affected by ‘coups’, the fact that the ‘presidential idea’ had spread so widely in the twentieth century in particular suggested that, on the basis of what had been truly an institutional ‘invention’ in the United States in 1789, a new ‘constitutional’ formula was being adopted in a context in which ‘new’ countries emerging from colonialism, with difficulty, admittedly, but with also some successes, particularly over time, as Latin American experience seemed to be showing especially from the 1990s and indeed even in some ‘new’ African countries as well, while there might otherwise have been a universal spread of ‘usurpation’ in the ‘new’ ex-colonised countries which were appearing on the scene.

My new book on the ‘Presidential Republic’ is thus an attempt at mapping out the difficult historical development of the ‘presidential republic’ since it was invented in the United States. The ‘presidential republic’ in its various forms is indeed, it seems to me, a genuine success, once we take into account the fact that decolonisation produced a large number of countries in which the legitimacy of the nation was, to say the least, very limited: what may be the case is that presidents in charge of the executive and in office for a number of years, as the United States constitution stipulated for the first time in the history of mankind, might be gradually the key instrument as a result of which usurpation may no longer periodically prevail.

Jean Blondel is a political scientist in the field of comparative politics. He became Professor of Political Science at the EUI in 1985 and was an External Professor from 1994 to 2000. Prof. Blondel set up the Department of Government at the University of Essex in 1964 and co-founded the European Consortium of Political Research. He was the winner of the Johan Skytte Prize in Political Science 2004. He has been awarded honoris causa doctorates from the University of Salford, the University of Essex, the University of Louvain-la-Neuve, the University of Turku, the University of Macerata (2007) and the University of Siena (2008).

Tyson Roberts – Do executive election rules matter in authoritarian regimes?

This is a guest post by Tyson Roberts, lecturer at UCLA and UC Irvine. It is based on his article ‘The Durability of Presidential and Parliament-Based Dictatorships’ in Comparative Political Studies, June 2015, vol. 48, no. 7, pp. 915-948. An earlier version of this post appeared at The Monkey Cage.


Many of the recent presidential elections covered in this blog – e.g., in Togo, Kazakhstan, and Sudan – occurred in states that many political scientists do not consider to be presidential regimes. The category “presidential regimes” is often restricted to presidential democracies, and in such data sets, democracies are states in which incumbent parties lose elections.

In nondemocracies such as Togo, Kazakhstan, and Sudan, many political scientists care little if the leader calls himself president, prime minister, or chairman of some military council. Nor do they care if the executive, president or otherwise, is elected directly or indirectly. Although recent cross-national research has recently paid attention to parties and elections in authoritarian regime legislatures (see here, here, and here for examples), election rules at the executive level in dictatorships are generally ignored in cross-national research. If the ruler cannot lose the election, the conventional wisdom assumes, then the rules for that election are unimportant.

In a forthcoming article, however, I find that election rules for the executive in dictatorships affect a number of outcomes, including regime durability and economic growth. Even if the executive has no intention of losing the election, the rules by which the election is held helps shape inter-elite dynamics in ways that have both economic and political consequences. These issues are particularly salient in recent years, since the majority of authoritarian regimes do hold multiparty elections for the executive office – in most cases, for the office of president.

According to the data used for the paper, most non-monarchy dictatorships prior to 1994 did not hold contested elections at the executive level (the paper does not consider monarchs, since this type of executive is never elected). The most common “executive selection system” during the Cold War was “Unelected”; either no elections were held or elections were uncontested plebiscites. With the end of the Cold War, a surge of contested Presidential systems (in which opposition parties can compete on the direct ballot for president) emerged in dictatorships, particularly in sub-Saharan Africa and former Soviet Republics in Eastern Europe, and Central Asia. By 1995, contested Presidential systems were the most common form of executive selection system in authoritarian regimes worldwide.

Meanwhile, approximately 5-10 countries per year, both during and after the Cold War, used what I call a Parliament-based system, in which contested elections are held for the legislature, and the legislature elects the executive. In some cases, the executive is a Prime Minister (for example, in Singapore, Malaysia, and Cambodia). In other cases (such as Botswana and pre-2005 Egypt), the executive is a President that is elected indirectly through the legislature rather than directly by the voters. In some cases, the executive of a Parliament-based system can be voted out at any time by a majority of the legislature, in others, the executive remains in office until the end of his term, at which time he must be confirmed by the legislature. Pre-2005 Egypt was an interesting case: the multiparty Parliament was empowered to elect the president with a two-thirds majority, and then voters confirmed that election in a referendum.


Figure 1. Frequency of non-monarchy authoritarian regimes with Unelected (including single party) and multiparty Elected (Parliament-based and Presidential) executive selection systems, 1975-2012. PRES = Presidential; PARL = Parliament-based.

If authoritarian leaders cannot be removed from office by election, then why should the executive selection system matter in authoritarian regimes? The answer begins with the fact that authoritarian regimes centered on an institutionalized ruling party tend to be more durable than regimes based upon the military or upon a personalistic leader and his clique. A party-based regime tends to share power with a broader set of elites; these elites have an incentive to work together to keep the party in power. In military regimes, on the other hand, many officers prefer to hand power to civilians in order to preserve the unity and reputation of the armed forces, and they can enforce an attractive exit agreement with the threat of a future coup, leading to brief spells in power. In personalistic regimes, the basis of support is narrow, which makes holding executive office more lucrative but also less secure than is the case for a party-based regime. Using the data from my analysis, I calculate the average risk of regime failure by dividing the number of failures by the number of country-years: a military regime has a 12% chance of failure (meaning democratization or transition to a non-military authoritarian regime) in a given year, a personalist regime has a 6% chance of failure, and party-based regimes have just a 3% chance of failure in a given year.

If a party-based regime has a Parliament-based executive selection system, the power-sharing and regime-sustaining effects of a party-based regime are amplified. The executive relies upon party members to win legislative office and then support his candidacy. In a Presidential system, on the other hand, the executive runs for office directly, and so relies less on party members, and may rely relatively more on the armed forces to repress voting in opposition areas.

These dynamics help explain why many of the most durable dictatorships are party-based regimes that maintained a Parliament-based executive system: for example, Malaysia, Singapore, and Botswana. For party-based regimes with a Parliament-based system generally, the average risk of regime failure is less than 1% in a given year, while the risk of regime failure for party-based regimes with a Presidential system is 5%. In other words, the durability of Party-based systems is in large part the result of regimes with a Parliament-based executive selection system.

However, it may be misleading to treat election rules in dictatorships as an independent cause of regime failure. Whereas democracies seldom change the rules to elect the executive, the executive selection system in authoritarian regimes can be changed by ruling elites. Weak regimes may be more likely to choose certain election rules, making such rules a marker but not an influential factor.

One way to test whether the executive selection system has an effect on regime survival is to consider cases where an ongoing regime changed its system, and compare those cases to regimes that kept the same system. Rulers of regimes with a Parliament-based system often wish to change the rules to enhance the visibility and legitimacy of their post, and to strengthen the executive office relative to the legislature. Regimes who changed from a Parliament-based to a Presidential system have higher failure rates than those that kept the Parliament-based system. There are nine cases since 1975 in which the system was changed from Parliament-based to Presidential, and in all but two cases (Togo and Zimbabwe), the regime later failed. In Egypt, Mubarak’s predecessor Sadat replaced single-party elections with contested elections at the parliamentary level in 1976. This Parliament-based system survived for nearly 30 years until 2005, when Mubarak adopted direct elections for the presidency. In Yemen, the constitution was amended in 1994 to shift the system from a multiparty Parliament-based system to a direct-vote presidential system. In Mongolia (1993) and Serbia and Montenegro (2000), the ruling party won re-election in a Parliament-based system, changed the rules to a Presidential system, and then lost power within a year by losing the first presidential election. More generally, regimes that maintained a Parliament-based system have a 2% chance of failure in a given year, while regimes that switched from a Parliament-based to a Presidential system have a 6% chance of failure in a given year.

However, this analysis does not preclude the possibility that regimes headed toward failure adopt a particular set of rules along the way. To make an apples-to-apples comparison between regimes with different rules, I use statistical analysis to understand which dictatorships are most likely to adopt a Parliament-based system rather than a Presidential or Unelected system. I find that contested election systems are more common in dictatorships with relatively high-income levels and low reliance on foreign aid and oil exports. Among these, personalist regimes are more likely to use a Presidential system, while party-based regimes are relatively more likely to use a Parliament-based system.

After controlling for these predictors of executive selection system, the Parliament-based system remains effective in promoting dictatorship survival. In other words, dictatorships that share power among many party elites are more likely to maintain a Parliament-based system that institutionalizes their power against the executive, and such power-sharing in turn promotes the ability of those elites to remain in power. Dictatorships in which more power rests with the executive are more likely to adopt a Presidential system that centers power in his hands; this concentration of power promotes survival of the president as long as the regime endures, but undermines the survival of the regime more generally.

To understand further how the Parliament-based system extends the lifetime of party-based dictatorships, I have found a number of intriguing patterns that fit with this story. Economic growth is particularly high in such regimes; Singapore, Malaysia, and Botswana are just a few examples. These regimes are also stable politically, with few cabinet shake-ups or military coups.

In short, the benefits for citizens from multiparty elections in dictatorships are particularly common in those with certain institutional features: a Parliament-based executive selection system and a strong institutionalized party. However, once an authoritarian ruler frees himself from the constraints of a Parliament-based system, citizens have little hope of implementing such constraints. I found no cases during the years covered of a Parliament-based system replacing a Presidential system in a surviving dictatorship.

Tyson Roberts is a lecturer at UCLA and UC Irvine. His research interests include comparative political institutions, democratization, international political economy, and the politics of economic development.