Tag Archives: presidentialism

Fernando Meireles – Latin American presidents and their oversized government coalitions

This is a guest post by Fernando Meireles, Ph.D candidate in Political Science at Federal University of Minas Gerais (Brazil). E-mail: fmeireles@ufmg.br

In many countries, presidents have a difficult time governing because their parties lack a legislative majority. In fact, because of the combination of separate elections for executive and legislative branches with multiparty systems, this situation is far from uncommon: during the last two decades in all 18 Latin American countries with presidential systems, only 26% of the time has the president’s party had a majority in the lower house. Due to this constraint, as a vast amount of research now highlights, minority presidents usually form multiparty government coalitions by including other parties in their cabinets. Again, only four Latin American presidential countries in the last twenty years were not governed by a multiparty coalition at some point since the 1980s.

However, the need to craft a legislative majority alone does not explain why presidents frequently include more parties in their governments than necessary to obtain a minimum winning coalition – forming what I call an oversized government coalition. The distribution of this type of coalition in Latin America is shown in the graph below. As can be seen, it is not a rare phenomenon.

If government coalitions are costly to maintain, as presidents have to keep tabs on their coalition partners to ensure they are not exploiting their portfolios to their own advantage – not to mention the fact that by splitting spoils and resources between coalition partners, the president’s own party is worse off – then why are these oversized coalitions prevalent in some Latin American countries?

In a recent article in Brazilian Political Science Review, I tackled this puzzle by analyzing the emergence of oversized government coalitions in all 18 presidential countries in Latin America[1], followed by a case study focusing on Brazil, spanning from 1979 to 2012. To this end, I gathered data on cabinet composition[2] from several sources to calculate the size of each government coalition in the sample: if a coalition had at least one party that could be removed without hampering the majority status of the government in the lower house in a given year, I classified it as an oversized coalition.

Specifically, I examined three main factors that, according to previous research, should incentivize presidents to include more parties in their coalitions than necessary to ensure majority support: 1) the motivation party leaders have to maximize votes, which would make joining the government attractive to opposition parties (vote-seeking); 2) the motivation presidents have to avoid coalition defections to implement their policy agendas (policy-seeking); and 3) the institutional context, considering the effects of bicameralism, qualified majority rules, and party system format on government coalition size.

The results support some of the hypothesis suggested by the literature. First, presidents are more prone to form oversized coalitions at the beginning of their terms, which shows that the proximity to the election affects Latin American presidents’ decision to form, and opposition parties to accept being part of, large coalitions – as others studies argue, this is mainly due to parties defecting from a coalition to present themselves as opposition when elections are approaching. Second, party fragmentation also has a positive effect on the emergence of oversized coalitions, consistent with the hypothesis that presidents might include additional parties in their coalitions anticipating legislative defections. Yet on the other hand, presidential approval, party discipline, and ideological polarization do not have the same positive effects on the probability of an oversized coalition being formed.

The factor that has the most impact on the occurrence of oversized coalitions, however, is the legislative powers of the president. As the literature points out, legislative decrees and urgency bills could be used by skilled presidents to coordinate their coalitions, facilitating horizontal bargaining between coalition partners. The comparative results show that this is the case in Latin America: the difference in the predicted probability of a president with maximum legislative powers in the sample forming an oversized coalition and another with minimum powers is about 32 percent points.

By exploring the Brazilian case in more depth, I also found that bicameralism dynamics and qualified majority rules impact the emergence of oversized coalitions. With two chambers elected through different electoral rules, parties in Brazil are often unable to secure the same seat share in both houses; to make things worse for presidents, party switching is still widespread in the country. In this context, as my results uncovered, differences in the number of seats controlled by the government in the Chamber of Deputies and the Senate positively affect the emergence of oversized coalitions. Finally, as some bills require supermajorities to be approved, such as constitutional amendments, reformist presidents also tend to form and maintain larger coalitions: the maximum value in this variable predicts increases by up to 10 percentage points on the probability of an oversized coalition being formed.

Taken together, these results show a more nuanced picture of why and how presidents form multiparty government coalitions in Latin America: often, obtaining a legislative majority is not enough to implement their legislative agendas, and so they might resort to a complementary strategy: to form larger coalitions. And presidents with greater legislative power, at the beginning of their terms or facing fragmented party systems, are in the best position to pursue such a strategy. In this way, both electoral and programmatic factors, as well as the institutional context, become key to understand variations in the size and the composition of government coalitions in presidential countries.

Notes

[1] These countries are Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, El Salvador, Ecuador, Guatemala, Honduras, Mexico, Nicaragua, Panama, Paraguay, Peru, Dominican Republic, Uruguay, and Venezuela.

[2] The criteria employed to identify a government coalition is the party affiliation of the ministers of the principal ministerial portfolios in each country – taking into account that ministers are not always recruited due to their connections or their congressional influence, and that in some cases they are not recognized by their parties as legitimate representatives of the same.

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.

Notes

[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: http://onlinelibrary.wiley.com/wol1/doi/10.1111/1475-6765.12181/suppinfo

References

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.

Turkey – Erdoğan is closer than ever to his dream of a hyper-presidential system

On January 21 the Turkish parliament passed a constitutional reform package introducing a presidential system. the reform was passed with 339 votes in favour, slightly more than the minimum threshold of 330 votes. The ruling AKP party had the support of Devlet Bahçeli, the leader of the pro- nationalist MHP (Nationalist Movement Party) and some of his party’s MPs, despite the fact that considerable number of the MPs and party supporters opposed the proposal. Now the reform bill is going to be sent to the President Erdoğan’s Beştepe Palace  for promulgation. He has two choices, either send it back to the Grand National Assembly for reconsideration, or refer it to a referendum. It is expected that he will refer it to a referendum, which will take place in April.

The reform package has no provisions enhancing basic rights or correcting the defective Turkish democracy. The constitutional amendment has two important and interconnected intentions; one is to change the current semi-presidential system into a hyper-presidential system and the other is to reform the judiciary so that the president can have a major role in the formation of judicial supervisory body, the Council of Judges and Public Prosecutors.

The reform package abolishes the dual executive and replaces it with a president who is the sole executive authority. He appoints all ministers, undersecretaries, and bureaucrats without the approval of the assembly. He has the power of legislative decree. He may regulate any issues that are not enacted by the assembly in detail, except individual and political rights, and he may do so without an enabling law or any prior conditions, such as necessity or urgency. When it comes to issues enacted by the assembly, the president may claim that the parliamentary act is not detailed enough or that his decree is covering another aspect of the issue. There is no retrospective examination of decrees by the assembly either. This type of regulation is always likely to create legal chaos. The constitutional court has the power of judicial review over presidential decrees. However, the president’s power to appoint 12 of the 15 court members for a 1- year term creates certain doubts that the court may not be independent enough to actually challenge the presidential will.

Furthermore, the president may create or abolish any public legal entity, regulate the duties, powers and the structure of ministerial bodies from top to bottom, and change the whole administrative structure by decrees without needing a parliamentary act. This means that he may reorder the main principles of administrative law without a parliamentary act. This is a big change in Turkish administrative law, since one of its main principles is that administrative law has to be enacted by parliament (the legality principle). If the reform is accepted in the referendum, the person who makes the rule will be the same person who implements that rule. There will be no external oversight of the administration, making administrative courts meaningless.

In addition to above-mentioned powers, the president will also have the power to declare a state of emergency and issue emergency decrees which may infringe or suspend all constitutional rights without any judicial review. Such a powerful legislative decree authority is hard to find in any Latin American Constitutions, even though almost all the current Latin American constitutions give presidents the power of legislative decree. In this region, they either require prior enabling laws (Chilean Constitution art. 32/3), or they can only be issued if the usual law-making procedures in parliament are not working properly and when there is an urgent need for such decrees (Argentinian Constitution art.99/3, Brazilian Constitution art.62). Such power also comes with retrospective control exercised by the assemblies, which is not the case for ordinary decrees (only for emergency decrees) in the current Turkish constitutional reform proposals.

The president is also responsible for determining and implementing national security policies as well as having the power to decide to use the army. Under the current constitution, this type of decision making traditionally involved chiefs of staff, the council of ministers, and the parliamentary assembly.

In addition, the president also has the power of parliamentary dissolution, again without any prior conditions or time limits attached. The parliament would mean that an early presidential election is held as well, since the two elections have to be held concurrently to help guarantee that the party led by the president can also win a majority of parliamentary seats. The parliament may also decide to call an early election, but this would require a three-fifths majority of the whole members (360 of 600). Clearly, a single person is more likely to make such a decision than an extraordinary majority. The president may dissolve the legislature if there is a conflict with the majority, or when he is about to be impeached and right before the decision to send the case before the constitutional court, or simply at a convenient time. Dissolution power is quite rare in presidential systems. However, it is often seen in competitive or electoral authoritarian presidential systems such as Pinochet Chile before 1989, Venezuela, Syria, Guinea.

This amendment also alters one of the main principles regarding presidents, namely that they cannot lead a political party. Instead, they need to be impartial towards all political parties. With this change, presidents are no longer required to be neutral. They can be the chairman of a political party and lead this party’s parliamentary majority. Traditionally, Turkish parties are leader-oriented, and internal democracy is quite weak. The party leader decides who gets to be nominated.

As for the structure of the Council of Judges and Public Prosecutors, which is responsible for overseeing the appointment, promotion, discipline, and dismissal of judges and public prosecutors, six of the thirteen members of the Council will be appointed by the president; the rest will be selected by the parliamentary majority. Since the president will be the head of a political party, he may lead the parliamentary majority. In the light of the current conditions in Turkish politics, the president is highly likely to control the parliamentary majority, which would make him indirectly involved in the selection process of the other members of the Council of Judges and Public Prosecutors.

The council selects the members of the High Court of Appeal (yargıtay) and three-quarters of the Council of State (the rest are appointed by the president). Their term of office is four years and they can be re-elected. The head of the council is the minister of Justice and his undersecretary is a permanent member. As pointed out above, the president also appoints a majority of the members of the Constitutional Court. In short, the President may shape all the high courts and the Council which control all the courts. This would potentially affect the independence of the courts from executive authority. Article 6 of the ECHR and Art. 38 of the current constitution state that there is the right to a fair trial, which includes being tried by an independent and impartial tribunal. Independence requires being free from the executive’s influence. The European Court of Human Rights uses four criteria to define independence; “the manner of appointment, term of office, existence of guarantees against outside pressures, and appearance of independence”. Under this amendment, none of these criteria are fulfilled. Without independent judiciary there is no fair trial for anyone and no rule of law. Furthermore, the manner in which the constitutional court judges are appointed by the president breaches a universal principle in law, whereby “no one can choose her judge” as the court is responsible for impeachment trials as well as examining decrees the president issues.

Overall, the reform package creates a very strong presidency without any checks and balances. It also supports the fact that in competitive authoritarian regimes presidents opt for new constitutions that consolidate their power, such as Venezuela (1999), Bolivia (2009), and Ecuador (2008). Currently, Turkey shows the signs of being a competitive authoritarian system. There is no free and fair competition among parties. It is a clientelistic and patronal system, which punishes the opposition (tax law, criminal law, etc) and rewards political loyalty by using state wealth and facilities. Selahattin Demirtaş and Figen Yüksekdağ, the co-leaders of HDP, the third largest party as well as many MPs of the same party are currently in jail; the main opposition, CHP, works under constant treats and some of its members are in jail too. Under the state of emergency the opposition faces especially tough constraints. Organising demonstrations and rallies are severely restricted.

Despite these facts, the AKP leaders still needs the support of MHP voter in the upcoming referendum according to the latest polls. If the right is unified, possibly with the help of a highly populist discourse, the reform package is likely to be accepted by the popular vote. However, “the no front” is getting ready for a tough struggle. It is going to be very tense three months in Turkish politics.

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

araujosilvavieira-illustrativepicture

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.

araujosilvavieira-figure1

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?

Contributors

Victor Araújo is a PhD candidate in the Department of Political Science at the University of São Paulo, Brazil.
E-mail: victor.asaraujo@usp.br
Website: http://www.victor-araujo.com

Thiago Silva is a PhD candidate in the Department of Political Science at Texas A&M University, United States.
E-mail: nsthiago@tamu.edu
Website: http://people.tamu.edu/~nsthiago

Marcelo Vieira is a Professor in the Department of Social Sciences at the Federal University of Espírito Santo, Brazil.
E-mail: mmarvieira@gmail.com

Thomas O’Brien – Presidentialism and Democratisation in South Africa and South Korea

This is a guest post by Thomas O’Brien, Lecturer in Political Science at the Centre for International Security and Resilience, Cranfield University at the Defence Academy of the UK. It is a summary of an article that will appear in Government and Opposition

The regime changes in South Africa and South Korea provide interesting insights into the role of presidential leadership during democratisation. In both cases the incumbent leader was forced to choose to subject their position to a democratic vote, thereby facing the risk of defeat. Echoing the point made by Escribà-Folch and Wright (2015), the respective regime types made this option viable as there was a belief that victory was possible and the status quo was increasingly unsustainable. F.W. de Klerk in South Africa was head of the National Party and had some hope that he would be able to retain power through democratic means given the institutional base and resources of the party. Similarly, Roh Tae Woo’s military background provided an institutional base on which he could rely to ensure stability and call on for support, in spite of his move into a civilian role. The position of President and head of a formal institutional apparatus gave them authority and control, which facilitated a degree of confidence that they could make the transition to democratic leadership successfully. However, the decision to accept the need for reform was not driven by altruistic ideals. Opposition to the incumbent regime structures had been growing significantly by the time each leader came to power, limiting the space they had to operate. South Africa had seen sustained social protest against the apartheid policies and faced growing foreign pressure in the form of sanctions and boycotts. At the same time, de Klerk faced internal divisions as hardliners within the party sought to block reforms. Roh Tae Woo faced extensive social protests against continued authoritarian rule, having taken over from Chun Doo Hwan who had been forced to resign in the face of widespread and sustained social unrest.

The issue of continuity is particularly important in these two cases. Both de Klerk and Roh assumed the presidency following the inability of their predecessors to continue (due to ill health and loss of legitimacy) during periods of instability. Taking on the role at pivotal moments provided an opportunity to make a change that had not been possible for their predecessors due to their deeper association with the regime structures. While both leaders had held high-ranking posts, their profile had been less contentious enabling them to maintain control over the institutional structure as they introduced reforms (on the emergence of reforming leaders from within see O’Brien, 2007). Continuity in this sense enabled the emergent leaders to introduce what they perceived to be reforms necessary to ensure their continued control. In both cases the eventual loss of control did not disrupt the democratisation process, as the leaders had been able to initiate reform internally to safeguard against reversion to authoritarian practices and were willing to accept the outcome.

The relative success of democratisation in these two cases warrants continued consideration of the role of incumbent leaders in shaping trajectories around regime changes. Democratisation by its very nature is a period of uncertainty, as roles and institutions are contested and reconstituted. Events in the Arab Spring and the Colour Revolutions show that regime change does not necessarily lead automatically to consolidated democracy. External pressure plays a key role in creating the opportunity for democratisation or reform by introducing a degree of uncertainty, as more actors become involved and take a stake in the outcome. A leader committed to change may be able to draw on this pressure to exercise agency and challenge entrenched institutional practices and patterns. In such situations the actions of the incumbent leader are crucial in shaping the outcome, as it is ultimately the elites that determine how to manage the opportunities and threats that arise. Mainwaring and Pérez-Liñán (2013) note that elite policy preferences (moderate or radical), normative preference for democracy or authoritarianism, and the regional political environment are key in determining whether a process of democratisation will be initiated.

In initiating reform the leader’s ability to manage the process and the likelihood of playing a role in the post-transitional context is arguably shaped by four structural factors: authority, institutions, opposition and continuity. Authority refers to the source of the leader’s power and in such regimes is generally derived from performance or personal charisma (Brooker, 2000). The robustness of the leader’s authority will determine their ability to maintain loyalty and exercise agency in shaping political developments.  While the reasons for the decision to relinquish power or at least allow reform of the system vary, legitimacy can be identified as an important factor. Where a regime loses support and legitimacy among the wider population it is possible to continue, but internal divisions may emerge as other actors perceive their own positions to be threatened. Institutional patterns play a key role in ‘structuring the nature of political competition’ (Elgie, 1995: 23), as they provide a base from which the leader can operate. If these have been neglected or degraded, they are less useful in times of crisis (see O’Brien, 2007 on Boris Yeltsin). As noted above, opposition is significant in pressing for reform, but the location (internal versus external) and strength of this opposition will determine the space the leader has to operate. The accretion of custom and practice over time ties actors into the system, thereby reducing the chances of defection from within, but potentially limiting the agency of the leader by encouraging pressure to maintain the status quo.

The institutional form of the regime plays an important role in the decision-making of incumbent leaders. Examining the ability of foreign pressure to force change in non-democratic regimes, Escribà-Folch and Wright (2015) find that personalist regimes are more resistant, as the stakes are higher for the leader without a formalised base. In military and party regimes the existence of a formal support base provides more opportunities in the event of systemic threats. Military leaders are able to return to barracks and exercise some degree of control over the democratising regime, through the threat of force. Party based regimes have less direct control, but possess the ability to participate (possibly under a new name) in the reconstituted system and return incumbent leaders to office. The corporate form of military and party regimes also enables the leader to rely on the hierarchy to ensure loyalty of followers and limit chances of defection, as failure would be costly for the whole of the collective. As noted, the institutional form played a role in both South Africa and South Korea, ensuring stability and a chance that the incumbent leaders may be able to secure a degree of influence over the regime trajectory.

Decisions of a leader are central in shaping the likelihood of a move towards democracy, but this does not guarantee that a fully realised democratic system will result, as structural constraints and internal opposition may stall or reverse progress made. Elite preferences determine what tools and direction the leader may choose (Mainwaring and Pérez-Liñán, 2013), but these preferences exist within a social and institutional framework that enables or constrains their actions. F.W. de Klerk and Roh Tae Woo demonstrated through their actions a preference towards greater democracy, reinforced by social instability and external pressure, but it was their control of the institutions of government that enabled this preference to be acted on. The cases also reiterate the importance of the perceived likelihood of post-transition success, maintaining a degree of control over the process. As Escribà-Folch and Wright (2015) argue, in the absence of a post-transition future a turn to repression may be a more viable option. Preferences are not absolute, contextual factors and likely future outcomes condition the ability and willingness of leaders to act on their preferences.

References:

  • Paul Brooker (2000) Non-Democratic Regimes: Theory, Government and Politics. Basingstoke: Palgrave Macmillan.
  • Robert Elgie (1995) Political Leadership in Liberal Democracies. Basingstoke: Palgrave Macmillan.
  • Abel Escribà-Folch and Joseph Wright (2015) Foreign Pressure and the Politics of Autocratic Survival. Oxford: Oxford University Press.
  • Scott Mainwaring and Aníbal Pérez-Liñán (2013) Democracies and Dictatorships in Latin America: Emergence, Survival and Fall. Cambridge: Cambridge University Press.
  • Thomas O’Brien (2007) ‘The Role of the Transitional Leader: A Comparative Analysis of Adolfo Suárez and Boris Yeltsin’, Leadership, 3(4): 419-32.

Thomas O’Brien is a lecturer in the Centre for International Security and Resilience, Cranfield University at the Defence Academy of the United Kingdom. His research examines leadership, democratization, environmental politics, human security, protest and New Zealand. Previous work has appeared in the British Journal of Sociology, Conflict, Security and Development, Contemporary Politics, Democratization, and Political Studies. @TomOB_NZ

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.

Turkey – After a period of violence and threats of political instability Erdoğan’s party wins back its dominant position in the parliament

Turkish voters went to the polls once again on the first of November, only six months after the June 7 general election. Eventually, 49 per cent voted for the ruling AKP, thus reinstating the AKP’s single party rule and its dominant status once again. The main opposition party, CHP, sustained its votes, whereas the nationalist MHP and the pro-Kurdish HDP saw a decline in their support even though they passed the ten per cent national threshold. This result came as a surprise for many as even the pro-government polls failed to predict such a strong result for the AKP.

The AKP’s nine percent gain came after a period of increasing political violence, threats of instability, and authoritarian pressures over free press and atmosphere of fear. After losing their parliamentary majority in the June election – which was turned into an informal referendum for a presidential system by the President – the AKP continued to govern the country. Parliament stayed closed and opposition parties failed to come together to form a legislative or executive coalition. Meanwhile, President Erdoğan continued exercising de facto powers despite the fact that his recent aggressive campaign for a type of hyper-presidential system failed.

The rising star of the June 2015 election was Selahattin Demirtaş the leader of the HDP pro-Kurdish party, who famously declared that his party would not allow Erdoğan to form a presidential system. He led his party to crossing the ten per cent national threshold for the first time, and thus prevented President Erdoğan and his party from realising their goal of a presidential system by simply taking their fair share of parliamentary seats. As votes for parties which are unable to pass the electoral threshold are assigned to the biggest party, giving them a significant overrepresentation under the Turkish D’Hondt system, votes for the HDP in previous elections often translated into an increased seat share for the AKP.

Four parties entered parliament following the June 7 elections: the AKP, CHP, MHP and HDP. However, none of them had a clear single majority. In a highly polarised political climate this meant stalemate. Prime Minister Davutoğlu, the new “official” leader of the AKP was given the mandate to form the government but returned it unsuccessfully to President Erdoğan. The president also made it clear that he was in favour of a snap election rather than forming a coalition.

The six months period in which Turkey first discussed coalition formation, and later the possibility of snap election, coincided with the end of peace talks and a ceasefire agreement between government forces and the PKK. Bloody clashes between the PKK and security forces took place in civilian occupied town centres as well as mountains resulting in heavy civilian, military and PKK losses. Furthermore ISIL suicide bombers attacked two different political demonstrations in Suruç and Ankara, killing 136 people.

It was not only the increasing threat of political violence that contributed to the political instability of the country. Within this climate fears of economic crisis have been rising together with threat of political instability. In addition, there were attacks on newspapers and journalists opposed to a government run solely by the AKP members and MPs. Some of the opposing newspapers and TV channels have been seized, sparking reactions from journalists all over the world. Many of the TV channels’ and newspapers’ coverage have been pro-government and opposition parties were unsuccessful in voicing their opinion in a free, equal or fair election atmosphere.

The AKP’s election strategy was formed on the idea of stability. Single party rule against coalition governments, peace against violence, economic growth against economic crisis -propagating that coalition meant instability, political violence and economic crisis.
Furthermore, President Erdoğan was overall less visible as part of AKP campaigns and plans for the introduction of a presidential system were not mentioned this time around. This campaign strategy seemed to have worked well as the AKP regained the votes that it lost six months ago. It has been claimed that Erdoğan new strategy after June 7 election was reinstituting single party rule by the AKP which would enable his de facto presidential rule. In other words, a type of semi-presidential system without being forced to cohabit.

Meanwhile the HDP and its rising star Selahattin Demirtaş could not campaign after the Suruç and Ankara bombings which mainly targeted the party and its supporters. Campaign events had to be cancelled in fear of more violence. The CHP partly followed the same path and decided not to lead an aggressive campaign. The pro-nationalist MHP and its leader Bahçeli, who blocked any possibility for a coalition with the AKP or opposition after the June election, led an unsuccessful campaign trying to explain why he refused to form a coalition. In the end, the MHP lost more than 4 per cent of its votes to the AKP.

With this result Turkey’s chances for re-establishing a parliamentary system are significantly slimmer. President Erdoğan now has a free hand to control executive, legislative and judicial powers, resulting in a strong form of semi-presidentialism. There is no doubt that he will increase the pressure on the political opposition, free press or any force that opposes his neo-patrimonial rule. It is also highly likely that he will seek to change the constitution – even though his party lacks the necessary three-fifths majority with a referendum – to establish a so called “Turkish Type” of presidentialism.

Turkey – No Presidential System but ‘Cohabitation’ for Erdoğan

The results of the 7 June parliamentary election change many things in the political scene of Turkey. It not only ended thirteen years of single party rule by the Justice and Development Party (AKP), but also thwarted President Erdoğan’s desire for “Turkish Type of Presidential System”. The President had actively campaigned for a governmental system change in favour of a Turkish type of hyper presidentialism for quite some time and turned this parliamentary election into a referendum on it, despite the constitutional clause obliging him to be unbiased and above party politics. Now the election results show that the AKP lost almost twenty per cent of its previous electors and its parliamentary majority seats, even though it remained the first party with forty per cent of the votes.

Many commentators believe that this is an outright rejection of presidential system and a endorsement of parliamentary practice. Even Prime Minister Davutoğlu agrees that voters did not endorse the idea of a presidential system. Research company Ipsos’ polls conducted right after the 2015 election show that 53 per cent of the electorate agree with this conclusion.

The new parliament is composed of four parties, none controlling a majority (276 seats are required to form a single party government). The AKP enjoys 258 seats. It is followed by the Republican People’s party (CHP) with 132 seats, Peoples’ Democratic Party’s (the HDP) with 80 seats, and Nationalist Movement Party (MHP) with 80 seats. Thus, Turkey is entering a era of coalition government, which is associated with political and economic instability by many people due to not very successful previous examples.

The political climate is still highly polarised and is not quite prepared for a stable coalition as the MHP has already ruled out any coalition scenario with the AKP or HDP. The HDP has also ruled out a coalition with the AKP. The CHP as a left wing opposition party has a long history of disagreement with the AKP. Even if parties agree on some kind of coalition formula there is another actor whose reactions are determining: President Erdoğan.

As the first directly elected President of Turkey, Erdoğan not only enjoys democratic legitimacy but also the constitutional power to appoint the Prime Minister. If the council of ministers cannot be formed or fails to receive a vote of confidence within 45 days starting from the formation of the Bureau of newly elected Turkish Grand National Assembly (TGNA), the president can call new elections. The president may choose to obstruct the formation of a coalition behind closed doors and force new elections or may never give the responsibility to form a government to a leader other than one from his own party’s.

However President Erdoğan’s first reaction three days after the election was to call on the country’s political parties to “leave egos aside” and form a government as soon as possible. He rejected the possibility of an immediate early poll by saying that he was not opposed to any coalition possibility and that leaders should try their best to form a government before new elections were due. He also invited the eldest deputy in the TGNA, the former leader of CHP Deniz Baykal, to discuss the election results. This meeting might be an indication that President is going to be active in the coalition formation process.

Even if a coalition government is formed and survived a vote of confidence in the TGNA, President Erdoğan will still have a weight in the executive branch and would potentially make it very difficult for any government to work with him. The Turkish Constitution grants more than a symbolic, but less than a policy-making role to the president. Before his election as president Erdoğan famously declared  that he would not be a traditional president hinting that he would interpret the constitutional rules outside the parliamentary tradition. He later pushed constitutional limits, chaired the cabinet regularly, interfered with the daily business of the Council of Ministers, created intra-executive conflict, and also directly violated his constitutional obligation of being impartial towards political parties.

The election results will not make a strong leader like Erdoğan  act more symbolically overnight, but it does mean a type of ‘cohabitation’ for him. He will no longer be able to dictate his policy choices directly. As for future governments it means that the president may meddle in the list of possible candidates for ministers, major executive appointments, executive decrees etc. by just simply refusing to sign them. The president may also impede the cabinet’s program to a degree. Indeed, a weak coalition or a minority government might potentially increase the president’s power or influence within the system.

Furthermore Erdoğan might turn any possible political instability or crisis into an opportunity to press again for a Turkish type of presidential system, pointing out the apparent shortcomings of the current system. He already argued before the election that coalition means disaster and there cannot be a coalition government under a presidential system, even though this is false.

How the current semi-presidential system will cope this difficult cohabitation is to be seen in the future, but one thing for sure is that it is not going to be easy for any of the actors.

Jean Blondel – The presidential idea

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

blondel

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.

TysonRoberts

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.

Tyson

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.

A