Is parliamentarism conductive to better political outcomes than presidential systems?

Is parliamentarism conductive to better political outcomes than presidential systems?
The question of whether parliamentarism is conductive to better political outcomes than presidentialism has been a debate with immediate practical implications alongside academic ones. In analysing this question, I review Linz's (1990) critique of presidentialism, which started the academic debate on assessing constitutional type. I conclude, alongside the vast majority of the literature, that Linz's pure type approach (Elgie, 2005) is simplistic in a number of ways. Linz's argument overlooks the contingency of presidential performance on a number of institutional and partisan determinants, as well as overlooking the differing performance of presidential regimes in terms of various political outcomes that affect how democracy functions. I focus on two outcomes that follow on from each other: policy making ability and the quality of policy making. The argument I present focuses on the outcomes delivered by presidential systems, compared to Linz's views on parliamentary systems, revolving around flexibility, positive-sum outcomes, and non-personalization.
THE LINZIAN HYPOTHESIS: PRESIDENTIALISM AND DEMOCRATIC CONSOLIDATON
In his seminal work, Linz (1990) cautions against the threat posed to democratic consolidation by presidential regimes. He builds his case through the analysis of zero-sum politics and the separation of powers. Linz's ultimate concern is with the rigidity of presidential systems. Three of the core institutional features often found in presidential regimes are to blame for rigidity: the zero-sum nature of presidential politics, fixed term, and the problem of dual legitimacy. For Linz, presidential politics are zero-sum due to the emergence of one winner, whereas parliamentary politics assures that non-winners of the election in the case of a single-party, majoritarian outcome, are still represented in parliament. Coalitional presidentialism is not a possibility for Linz; he contends that "the losers must wait… without any access to executive power and patronage" till the next election (Linz, 1990: 56). Fixed term, on the other hand, hinders capabilities to deal with crises of government. While cabinets are subject to change through the vote of confidence mechanism in parliamentary systems, fixed terms make the removal of the executive nearly impossible in presidential systems, given that impeachment is only possible in the case of commitment of criminal offense. Finally, dual legitimacy brings about a lack of unity between the executive and the legislature. Interestingly, Linz discusses this in terms of the national mandate of the president, compared to the more regional mandates of congress members. Dual legitimacy only exacerbates the effects of fixed terms; the difficulty of removing a president, unlike a cabinet, may lead to the most extreme scenario of "a full-blown regime crisis in a presidential system" (Linz, 1990: 65).
Linz's predictions have been criticized from many angles (Elgie, 2005). In assuming that presidential politics is a zero-sum game in which coalitional politics is not possible, Linz delivers his discussion by reference to a "pure" understanding of presidentialism. Clearly, Linz's pure type analysis is severely limited in several ways. Firstly, there are only a handful of cases that exemplify the pure presidentialism and parliamentarism Linz identifies (Siaroff, 2003). Secondly, the high variation in the identifying features of regime type inevitably means that the political outcomes they cause will be more nuanced, and cannot be simplified to mere democratic consolidation. Thus, what is called the "second wave" of presidential and parliamentary studies by Elgie (2005) takes into account more explanatory variables, derived from the institutional variations between presidential arrangements, in accounting for the performance of presidential regimes. The increase in explanatory variables, in turn, implies that the political outcomes we will observe will be more varied than just democratic consolidation. I identify policy making ability and the quality of policy making as two ways to measure political outcomes, and show that the performance of presidentialism in these outcomes is fully contingent on the institutional arrangements that underpin the presidential system at hand.
POLICY MAKING ABILITY
Policy making is identified as less of a problem where executive-legislative unity is achieved, given that the lack of support for the cabinet's policy proposals in a parliamentary system, bringing about executive-legislative disunity, will be dealt with through a vote of no confidence. A new cabinet will be installed and executive-legislative unity will be restored. However, as highlighted by Linz, executive-legislative unity is not ensured by the presidential system. But presidents are certainly not incapable of implementing policy. This shows that policy making depends on the underlying arrangements of the presidential system. Let us distinguish between institutional and partisan arrangements, and start our discussion with institutional arrangements.
A good way to analyse policy making ability in a presidential system is the veto-players concept designed by Tsebelis (2002) . For Tsebelis, veto players are the number of actors needed to agree with a proposed change. Veto players are categorized as institutional or partisan. Let us start from institutional veto players. It is important to understand that presidency is defined institutionally in a variety of ways, depending on the legislative powers granted to the president (Tsebelis, 2002: 78; Sing, 2010). This, in turn, determines both the extents to which the president and the legislature are veto players respectively. Presidents may have very little legislative power, as in the US. This raises the possibility of political deadlock, in which the president is unable to implement policy due to disunity with the legislature, and has the potential to pose a regime crisis as in Linz's argument.
However, recall that Linz assumes that coalitional politics is not possible under the presidential arrangement. Cheibub (2007), on the contrary, argues that coalition formation is approximately as likely under a presidential system than a parliamentary one. The incentive for coalition building is that the president would be unable to implement policy without the support of the legislature. To provide an incentive to the legislature to cooperate, the president uses portfolios, assigning roles to other parties in her cabinet. In turn, the recipients of portfolios have an incentive to support the president's policy bills at the legislature. Of course, whether or not portfolios generate coalitional incentive in the legislature is highly contingent on party and coalition cohesion. For example, policy making in Chile is often cited as a success, as cohesion is higher compared to other Latin American countries; thus portfolios have been more effective at generating and maintaining coalitions.
We must also consider the legislative powers of the president. It is often the case that presidents are able to introduce legislation directly, with congress able to counter them (Cox and Morgenstern, 2001: 181). Then, congress still has veto powers, but the scope in which it can exercise them is decreased, as these systems are ones in which presidents are able to make use of effective decree powers. (Siaroff, 2003). The continuous use of decree powers brings about the threat of moving into a system where presidents dominate the legislative process completely (Cheibub, 2007). Then, policy making ability ceases to be a problem, but we observe what O'Donnell (1994) calls a "delegative democracy". In delegative democracies, "whoever wins election to the presidency is thereby entitled to govern as he or she sees fit, constrained only by the hard facts of existing power relations and by a constitutionally limited term of office" (O'Donnell, 1994: 66). Thus, congress is effectively eradicated as a veto player. This is often what is observed; Latin American presidents, such as Fujimori in Peru, Collor in Brazil and Menem in Argentina bypassed congress and ruled through decree when they knew it would not generate the support necessary (Cox and Morgenstern, 2001). This has detrimental implications for the quality of policy making, as will be discussed in the next section.
Up till now, I have been touching upon the importance of congress in determining the president's policy making abilities. A factor of paramount importance that influences how congress behaves is the prevailing party system. Mainwaring (1993) finds that party systems are the determinant of policy mobility and immobility; a multiparty system causes immobility, polarization, and coalition instability. Thus Mainwaring suggests that a two-party presidential regime is ideal to ensure policy mobility. However, his findings are disproved by Power and Gasiorowski (1997), who conduct a study on Third World states, finding that presidentialism requires a higher number of parties for consolidation. Returning to the idea of coalition formation, Cheibub (2007) finds that high party pluralism is ideal as it creates a higher incentive for coalition building. Moderate party pluralism, on the other hand, is found to be more dangerous than a two-party system or a highly multiparty system.
The tension between the works of Mainwaring versus Power and Gasiorowski brings the debate to an interesting point. Instead of looking at the prevailing party system, Przeworski et al (1997) conclude that what is necessary to safeguard the president's policy making ability is not a two-party system, but a legislative majority. Clearly, two-party systems are often associated with single-party majorities in legislature (Lijphart, 2010: 81). However, it is clear that this is not the only option, and legislative majorities are often produced in multiparty systems, given that parties are able to agree along policy lines. Given that agreements and coalitions are determined by the parties that make up the legislature, policy making ability is not only contingent of institutional factors, but also partisan factors, in that who the actual parties that form congress are.
In analysing partisan factors, we can return to veto players. Partisan veto players "are generated inside institutional veto players by the political game", thus determining the way in which institutional veto players behave. Let us consider a presidential system, as the US, where the separation of powers is strict. Firstly, at the most simple level, parties' ideational standings in relation to the president determine whether or not congress acts as a veto player. To consider congress' standing as a veto player, we need to analyse the extent of diversion between the president and parties in congress. In the case of low diversion between the president and the parties that form congress, we observe interbranch hierarchy, in the sense that the president proposes policy, and the legislature approves it; policy is made at the directive of the president. Congress, although an institutional veto player, does not act as one due to its partisan composition that parallels the president. In other words, congress, with its party distribution, is absorbed by the president (Tsebelis, 2002: 26).
On the other hand, where diversion is high, we observe interbranch transactions (Shugart, 2010: 355-6); congress acts as a veto player, affecting the president's policy making ability. Whether or not policy is made depends on the progression of these transactions, in which the president has to reformulate policy for it to be acceptable to congress, and congress have to take into account the president's preferences in avoiding deadlock. Making policy acceptable to congress is a function of several factors. It depends on the number of partisan veto players in congress, and the extent to which they diverge from the president, determined by the ideational standings of both the parties and the president. Clearly, the same factors also determine the prospects of coalition building.
More interestingly, the composition of congress, in terms of party characteristics, can be looked at in ways other than party ideology. In a study of Latin American presidentialism, Cox and Morgenstern (2011) classify parties based on their motives rather than their ideational standings. While their discussion of recalcitrant and subservient parties mimics the discussion above on partisan veto players, the distinction between workable and parochial-venal parties adds new insight. Workable parties are those parties that "demand a seat the policy table", while parochial-venal parties are those that are "willing to concede policy issues in exchange for access to pork or other resources" (Cox and Morgenstern, 2011: 173). While policy making with workable parties will be a function of the ideational diversion between the president and congress as above, policy making with parochial-venal parties will be a function of the pork and patronage resources the president is able and willing to offer the legislature (177). Thus, not only the ideational standings of parties in congress, but also their motives determine the president's prospects of policy making.
THE QUALITY OF POLITICS
Having analysed the ability of policy making, let us turn to the quality of policy making. I identify the quality of policy making as a function of experience and responsibility. Both experience and responsibility are a function of party cohesion and institutionalization.
Firstly, the single-head style of executive office in presidential systems is often blamed with creating conditions for the election of inexperienced political outsiders as presidents (Linz, 1994: 26; Samuels and Shugart, 2003: 36). Linz argues that this is because candidates run as individuals, rather than as parties as in a parliamentary system, and voters have no incentive to look at political experience as a determinant of who they vote for; making their minds up about the personal qualifications and positions of the candidate is sufficient (Linz, 1994: 27). Fujimori in Peru, who entered politics as an academic, and won against his opponent Llosa's highly unpopular austerity proposals, is a prime example. The danger for the quality of policy making is that such figures more than often search to maximize their personal interests through the use of office. However, Mainwaring and Shugart (1997: 456) find that this is not the case if parties are institutionalized. Where parties are institutionalized, "the election of political outsiders is the exception". Indeed, Fujimori was elected when the Peruvian party system was unravelling.

Party cohesion can have a similar effect on moderating the effects of fixed terms, which have consequences for responsible or irresponsible policy making. Fixed terms are implemented in order to limit the personalization of the presidential office. However, Cheibub (2007) argues that they limit accountability. With a fixed term, the time horizon of the incumbent president is greatly diminished, compared to an incumbent prime minister in a parliamentary system, where the time horizon is uncertain. The effects of a shortened time horizon are exacerbated when the incumbent is in her last term, or in other words, a lame duck. When the prospects for re-election are none, the president does not have an incentive to consider the preferences of the electorate and the future effects of policy. With the accountability mechanism of re-election removed, policy making is likely to be conducted more irresponsibly. It is interesting to consider whether and how party cohesion can compensate for the loss of accountability due to a fixed term. In settings where parties are cohesive, presidents are more likely to rely on their own party's legislative support in passing policy. Thus, the president will know that her party will consider prospects for re-election in supporting her policies. This may have a mediating effect on the president's behaviour. More research should be done to verify whether this is the case.

In analysing policy making ability and the quality of politics, I have demonstrated that parliamentarism is not inherently conductive to better political outcomes than presidentialism. I showed that the answer is contingent on a multitude of factors: the institutional powers of the president and congress, party cohesion, the prevailing party system, the ideational distribution of parties, the motives of parties, political experience and responsible policy making, among others. Given so many contingencies, it does not make sense to debate which constitutional type is conductive to better outcomes. Instead, we should focus on lower-level institutional arrangements such as the ones above in order to avoid hasty generalizations of a highly diverse set of political systems (Kent, 2000).

Works Cited
Cheibub, J.A. (2007) Presidentialism, Parliamentarism and Democracy. Cambridge: Cambridge University Press.
Cox, G. and Morgenstern, S. (2001) 'Latin America's Reactive Assemblies and Proactive Presidents', Comparative Politics, 33(2), pp.171-189.
Elgie, R. (2005) 'From Linz to Tsebelis: Three Waves of Presidential/Parliamentary Studies?' Democratization, 12(1), pp.106-122.
Kent, E. (2000) 'Parliamentarism Versus Presidentialism in the Policy Arena', Comparative Politics, 32(3), pp.355-376.
Lijphart, A. (2010) Patterns of Democracy. Government Forms and Performance in Thirty-Six Countries. New Haven, CN: Yale University Press.
Linz, J. (1990) 'The Perils of Presidentialism', Journal of Democracy, 1(1), pp.51-69.
Mainwaring, S. (1993) 'Presidentialism, Multipartism and Democracy: The Difficult Combination', Comparative Political Studies, 26(2), pp.198-228.
Mainwaring, S. and Schully, T. (1995) 'Party Systems in Latin America', in Mainwaring, S. and T. Schully (ed.), Building Democratic Institutions: Party Systems in Latin America. Stanford, CA: Stanford University Press.
Mainwaring, S. and Shugart, M. (1997) 'Juan Linz, Presidentialism and Democracy: A Critical Appraisal', Comparative Politics, 29(4), pp.449-471.
O'Donnell, G. (1994) 'Delegative Democracy', Journal of Democracy, 5(1), pp.55-69.
Power, T. and Gasiorowski, M.S. (1997) 'Institutional Design and Democratic Consolidation in the Third World', Comparative Political Studies, 30(2), pp.123-155.
Przeworski, A., Alvarez, M., Cheibub, J.A. and Limongi, F. (1996) 'What Makes Democracies Endure?' Journal of Democracy, 7(1), pp.39-55.
Shugart, S. (2008) 'Comparative Executive-Legislative Relations' in Binder, S. and R.A.W. Rhodes (ed.) The Oxford Handbook of Political Institutions. Oxford: Oxford University Press.
Siaroff, A. (2003) 'Comparative Presidencies: The Inadequacy of the Presidential, Semi-Presidential and Parliamentary Distinction", European Journal of Political Research, 42(3), pp.287-312.
Sing, M. (2010) 'Explaining Democratic Survival Globally (1946-2002)', Journal of Politics, 72(2), pp.438-55.
Tsebelis, G. (2002) Veto Players: How Political Institutions Work. Princeton, NJ: Princeton University Press.



The same can be said of his discussion of parliamentarism, which favours a coalitional understanding. Although he argues that losing parties are still represented in parliament in single-party majority governments, issues relating to a single-party executive are not discussed.
Tsebelis is often associated with a third wave of presidential and parliamentary studies, in which regime type is not the main inquiry, but one of many institutional functions that are explained through the proposed methodology (Elgie, 2005: 115). Nonetheless, his argument relating to regime type is beneficial in underpinning arguments regarding policy making ability, given that it highlights various contingencies for the ability, distinguishing between institutional and partisan ones.
By cohesion, I refer to voting unity among party members in congress.
Tsebelis' discussion of partisan veto players focuses on parties that form a coalitional executive in parliamentary systems, given that the agreement of all coalition partners (provided that we are facing a minimal-winning coalition) is necessary, making each one a partisan veto player. However, I believe the same logic is applicable to analysing congress as a veto player.
Parties are institutionalized when interparty competition is governed by stable rules, political actors accord legitimacy to the electoral process, political parties have a degree of roots in society, and party organizations are not trivial (Mainwaring and Schully, 1995).
The separation of powers is often argued to reduce accountability, given that it becomes harder to track the origins and processes of policy making. However, "presidentialism typically allows for the possibility that voters can hold presidents and legislators accountable for different things" (Samuels and Shugart, 2003: 41).
We cannot deny that presidents care about their legacies. However, it is not possible to make predictions based on this factor in the way we can on term limits from a rational choice perspective.