Democracy without Presidents[1]
Marcelo Alegre (UBA-Palermo)
Introduction
In Latin America the criticism of presidentialism is losing force. This work is a complaint against this tendency. I once thought, too, that we should abandon the proposal of a shift towards Parliamentarism. Yet as times goes by the flaws of presidentialism are even more visible. The current silence on the issue may respond to different reasons. First, maybe the whole controversy between presidentialism and Parliamentarism is out of fashion, and we should focus on other, more illuminating variables (federalism, party systems). Second, perhaps the criticisms of presidentialism were just mistaken, or some hidden virtues of the presidential scheme have surfaced. Third, even if the dichotomy remains as a relevant one, and even if the defects of presidentialism are uncontroversial, we could be warranted in indefinitely postponing the reform of our political systems given the radical change it would imply. Additionally, we could think it would not be wise to subject our societies to the trauma, confrontation and tension of new constitutional reforms.
I will reflect on these explanations in the following order: First I will refer to the current state of the discussion in the last 15 years, the new evidences about the defects of presidentialism and some of its defenses. Then, I will focus on the normative criticism of presidentialism and on how these considerations also apply to the semi-presidentialist variants. I think these are not good alternatives, and that we should go for a parliamentary solution. Lastly I aspire to convince you that academics should not keep silent in regards to a system that, like Borges´s character, “is authoritarian, but also inefficient.”[2]

1. Does the divide “Presidentialism versus Parliamentarism” remain a useful one?
A possible explanation of the abandonment of the critic to presidentialism in Latin America would be that the distinction between presidentialist and parliamentary systems is too abstract, and that therefore discussion of institutional reforms should have to center in other elements of the institutional systems. It would be more fruitful that instead of questioning the political system that we took care, for example, of the more than problematic Latin American federalism or the weak judicial independence, or that we reform the electoral and party systems. This point of view got some academic support in the literature. In effect, there have been very serious attempts to demonstrate that the dichotomy "presidentialism/Parliamentarism" had to be left.[3] Nevertheless, the debate has fortified the relevance of this divide--the attacks against the distinction, instead of eroding it, have enriched it. Of course, today we all understand the importance of taking into account the federalist features, the electoral system and of political parties, or of the number of veto players on the public arena. But the fact that a country adopts a presidentialist system or a parliamentarian one continues being the most significant institutional decision, which carries the deepest consequences in relation to the rest of the variables. In other words, the presidentialist systems usually produce very noticeable differences to the parliamentarian regarding an ample series of elements. The differences to the interior of the government systems, with all their importance, do not shadow the relevance of the differences between government systems. As David Samuels and Kent Eaton[4] have shown, there are two predominant characteristics in the presidentialist systems. On the one hand, the "unilateral powers of the executive" (veto powers, decree authority, budget initiative, etc) and on the other hand, the "separation of purpose" (that occurs when the executive and the legislators differ in their incentives, preferences, and regarding the groups, pressures, and demands to that they respond). These two elements are more frequent in the presidentialist systems, and even though they can exceptionally be present in the parliamentary systems, their impact is greater in the presidentialist schemes. From these characteristics noticeable differences arise between the presidentialist and the parliamentary systems. There are differences regarding the fragility of the regime, the probability of the emergence of minority governments, the effectiveness in the approval of legislative proposals, the incentives to propose and to integrate coalition cabinets,[5] the probabilities for the entrance of opportunists and "outsiders" to the government, the access to political information, the influence of interest groups, the vulnerability to corruption, and the exposure of the systems to steep and surprising changes of programs and governmental programs. I will review these effects, but so far we should take into account that the distinction between presidentialism and Parliamentarism deserves to continue in the center of the studies of political institutions. This does not mean, I insist, ignoring the relevance (as it perhaps happened during a time) of other variables that affect in important ways the operation of the political systems, like the legislative powers of the executive, the electoral systems, the system of political parties, the bicameralism, the federalism and the modes of judicial review.[6]
In order to illustrate the importance of the distinction between presidentialism and Parliamentarism, let me allude to the important work of George Tsebelis,[7] one of the authors who have insisted on the convenience of leaving this dichotomy. Tsebelis has proposed that the key to understand the different systems are the veto players (actors whose acquiescence is necessary to modify the status quo). In tight synthesis, the idea of Tsebelis is that the greater the amount of veto players, the greater the stability of policies and, consequently, the greater the difficulty to produce reforms. Samuels and Eaton indicate that because the number of veto players themselves could be the result of distinctive elements of both systems, these and not those are the major key. Additionally, Tsebelis himself recognizes that certain veto players can be irrelevant, given the "absorption rule": if a veto player shares the same preferences with other, its veto power loses importance. An additional problem is that usually there are smaller collective or individual actors, who are veto players, inside collective actors like political parties, unions, or legislative blocks. The conclusion is that counting veto players can be a very troublesome and not always very illuminating, task.[8] The veto players’ variable is not a good candidate to displace "presidentialism-Parliamentarism" from the center of the scene.
As for me, I believe that the relevance of the distinction "presidentialism-Parliamentarism", far from being diluted, is assumed by the Tsebelis himself when he indicates that when both systems (the presidentialist and the parliamentarian) have a high number of veto players the following difference takes place: in the parliamentarian regime there will be instability of the government, whereas in presidentialism there will be system instability.[9] Under certain circumstances, the difference between Parliamentarism and presidentialism, we learn of Tsebelis... is the one that is between democracy and dictatorship!
However, even though the distinction between presidentialism and Parliamentarism continues being relevant, there is a change in the focus of the analysis. Until fifteen years ago the predominant parameter of study was the stability of the systems, but today the attention is centered in the type of policies and results that one or another system tends to produce. [10] In other words, if before authors worried about the survival of presidentialism, today they are concerned about its life. Thus, it interests to investigate the capacity of presidentialism to produce structural reforms, to resist the attacks of the pressure groups, to control corruption, to avoid the U-turns in the directions of the governments, or to produce public goods. As we will see, presidentialism does not perform well on any of these variables.

The evils of presidentialism revisited.
Before analyzing some of the new developments in the study of presidentialism, let me briefly review four of the most traditional objections against presidentialism.
i)Conflicting legitimacy. Arend Lipjhart[11] has pointed out that due to the fact that both the legislature and the executive are popularly elected, the possibility arises of conflicts between the two powers, for example if they express different choices;
ii) Rigidity. If the first feature describes a problem, the second relates to the weakness in resolving the problem. As the presidential term is fixed, this is an obstacle for eventual “readjustments.”[12]
iii) Majoritarian tendencies: According to Lipjhart[13], presidentialism discourages agreements and consensus, which are especially necessary in times of crisis and in transitional periods.[14] Linz[15] attributes it to the “zero-sum game” or “winner-takes-all” character of presidentialism. A more sociological approach in the same direction is due to Guillermo O’Donnell, who criticizes the style of democracy typical of Latin American presidential systems as a delegative democracy, where presidents are “enabled to govern the country as he or she sees fit...”[16];
iv) Personalization of power: Carlos Nino[17] emphasized that in presidential systems excessive power is concentrated in one person’s hands. That makes it vulnerable, for example, to institutional ruptures. Therefore weakness in the presidential leadership implies the weakness of the system as a whole.
Studies on presidentialism of the last fifteen years reaffirm the problematic character of presidentialism. As I previously said, in the last years the center of the analyses has been in the problems of performance of presidentialism, rather than on its problems of stability. Nevertheless, although there have been some works that confirm the presidentialist vulnerability to institutional breakdowns, in this section I want to describe some of the functioning problems of presidentialism, which should be read as new variations on the classic melodies of Linz, Nino, Lipjhart and others.

v) U-turns. Susan Stokes has studied a persistent tendency in the Latin American democracies: the steep turns that the presidents carry out in their policies of government in contrast to their electoral promises.[18] This is not an isolated phenomenon, since it happens in more than 30% of the cases. That is to say, in at least 3 of each 10 presidential elections the winner breaks his or her electoral commitments and puts in place policies different from the announced ones.[19] Remarkably, in all the cases of steep turns the adopted policies are towards the right. There is no a single case of a president who governs to the left of his campaign. Even when in some cases (Fujimori, Menem) the course change was electorally validated in later elections, it is evident that U-turns damages public confidence in politics. These U-turns are much more difficult in the parliamentary systems, since it would be required that the whole party of government or the governing coalition agreed in favoring the surprise change. Besides, parliamentary systems tend to be more immune to the entrance of “outsiders”.

vi) Institutionalized Conservatism: In general the presidentialist systems originate a greater number of veto players than the parliamentary systems. (I say in general, because the parliamentary systems with pluriparty coalitions also have many veto players, and on the contrary the executives who concentrate great legislative powers, de iure or de facto, imply a diminution in the number of veto players).[20] The probability that a legislative project of the government gets approved by Congress is smaller in presidentialism, given the high frequency of minority governments. Whoever favors a progressive agenda in Latin America should take note that in a presidentialist system reforms of the status quo are 1) more improbable, 2) less deep, 3) slower, 4) more expensive. [21]
Persson and Tabellini show that parliamentary regimes usually are accompanied by governments of greater size, and tend to provide more public goods than presidentialist ones.[22] The explanation provided by these authors is that the electoral base of the parliamentary regimes tends to be greater (and, we could add, more durable, especially if coupled with consensualist mechanisms such as the constructive censorship), which favors more egalitarians policies. On the other hand, in Parliamentarism a greater legislative cohesion exists, which favors more “generous" agreements on public spending.
Latin America badly needs a reversion of the double regressivity given by tax systems that penalize the poor and by public spending oriented mainly in favor of the accommodated classes.[23] In order to face the enormous infrastructural, educational, and income inequities, public spending will have to increase or at least remain high. Of course, a high level of public cost is no guarantee of social justice, but it is a necessary condition for some progress on fairness. Presidentialism does not seem to be, after all, a neutral option when we reflect on the abysmal inequalities in our region.

vii) Corruption and inefficiency. Jana Kunicová and Susan Rose-Ackerman[24] [25] have studied the relation between diverse institutional systems, including presidentialism, and corruption. Kunicová points to two sources of risks of corruption in presidentialism: The fixed presidential mandate (which denies citizens all possibility of punishing presidents in their last years of mandate),[26] and the concentration of functions in the president vis a vis the Congress. Kunicová has studied the relation between different indexes of corruption and presidentialist systems and the results are clear: presidents irresponsible before the electorate and before the Congress are not the best guarantee of ethical transparency.[27]
The vulnerability of presidentialism to corporatist pressures has been described by Linz, Nino, and others.[28] David Vogel has explained that the type of separation of powers typical of presidentialism tends to lower the cost of access of interest groups, by creating two decision-making centers.[29] Ackerman argues that the type of competition between the Congress and the President for the control of the bureaucratic apparatus of the state tends to generate a style of excessively politicized government, transforming the executive authority into an enemy of the rule of law.[30]
It seems as if the new studies on the subject do not encourage the prevailing passivity on the system of dominant government in the region: presidentialism foments corruption, corporatism and administrative inefficiency.
3. The presidentialist counterattack.
Were we too hard with presidentialism? In this section I will analyze some reactions to the criticisms of presidentialism. In summary, the accusation against presidentialism is that it is a weak democratic system and at the same time, a weakly democratic system. It is a weak democratic system as soon as it has shown a great fragility in front of institutional ruptures. And it is a weakly democratic system because it tends to accentuate very negative characteristics in the political system. But some voices have arisen questioning the reach of the criticisms, doubting both the connection between institutional presidentialism and breakdowns, and the supposed functional deficiencies of presidentialism.

3.1. Presidentialism and coup d'etats. The close relationship between presidentialism and democratic instability has an ample academic acceptance. A classic study of Alfred Stepan and Cindy Skach[31] shows that the rate of survival of presidentialist regimes in 53 countries outside the OECD, between 1973 and 1989, is of 20%, in comparison with 61% of pure parliamentary systems. Scott Mainwaring shows that of the 32 democracies with a stability of 25 years (to 1991), 23 (that is a 72%) are parliamentary.[32] Fred W. Riggs indicates that except for the U.S.A., all the presidentialist countries have undergone coup d'etats, whereas two third parts of the parliamentary regimes of the Third World have remained in democracy.[33] Linz insists that U.S.A. is the only "presidential democracy with a long history of constitutional continuity."[34] More recently Adam Przeworski and others have shown that during years 1950-1990, 54% of the presidentialist regimes have undergone breakdowns, against a 28% of the parliamentary regimes. The life expectancy of a presidentialist regime is 21 years, against 73 of a parliamentary one.[35] These authors also refute the idea that the instability of presidentialism is due to the fact that the analyses usually are centered in the Latin American cases, since the survival of the Latin American presidentialist systems is quite greater than the one of the presidentialist systems outside Latin America (10,6 years against 6,5 years). In addition they also refute the notion that it is underdevelopment and not the government regime the cause of the political instability, since they show that "the presidential democracies tend to die more than the parliamentarians in any level of development."[36]
Shugart and Carey question this relation between presidentialism and instability. These authors provocatively claim that they “find no justification for the claim of Linz and others that presidentialism is inherently more prone to crises that lead to breakdown.”[37] On the one hand, they identify twelve presidential regimes and twenty-one parliamentary regimes that have broken down in the 20th century. On the other hand, they list as stable democracies twelve presidential systems against twenty-seven parliamentary, but when they focus on Third World countries, they find that 59.1 percent of the parliamentary regimes have broken down against only 52.2 percent of the presidential.
However, I find their analysis strongly defective. First, when they list breakdowns corresponding to parliamentary regimes, they count each breakdown independently of the fact that more than one of them had occurred in the same country. Thus, they list as different cases the breakdowns of Greece in 1936 and in 1967, and the same happens with respect to Pakistan’s breakdown of 1954 and 1977. Surprisingly they don’t apply the same criterion to the list of presidential breakdowns. To speak only of the first country in the list, Argentina, the consistent application of the criterion would imply counting six breakdowns in this century, (1930 –the only one mentioned by the authors-, 1943, 1955, 1962, 1966 and 1976). All these breakdowns satisfy the peculiar conditions given by the authors to qualify as such (only those cases in which at least two consecutive general elections were hold before the breakdown). Second, the criterion used to analyze the success of the regimes is even more astonishing: it is satisfied by having had at least two democratic elections without breakdowns, as of 1991.That means that those countries with biannual elections are counted as consolidated democracies with two years of constitutional life. According to this, (like the smoker who says that he gives up smoking every Monday) Argentina has had ...seven periods of consolidated democracy in the 20th century! Linz also calls the attention to other errors and omissions of this work, like the fact that India is not included in the group of stable parliamentary regimes[38].
In a later work, Shugart and Mainwaring[39] follow a different line of argument. They don’t directly contend the connection between presidentialism and breakdowns, but, instead, try to show that some other factors may be the real causes of the disruptions, and conversely, that the prima facie relation between parliamentary systems and democratic stability can be due to elements other than the type of governmental structure. The authors claim that the correlation between parliamentary systems and consolidated democracies could be spurious, because it could be that in those cases the continuance of democracy could be the consequence of “relatively better background conditions”. On the other hand, they also doubt that Latin American breakdowns “resulted from presidentialism rather than from conditions afflicting South American democracy more generally”.[40]
First, it is important to note that no critic of presidentialism attributes to it the whole causal responsibility for the breakdowns. Rather, the arguments against presidentialism are formulated in a different, negative way: They claim that the presidential system is unable to provide solutions against breakdown forces and tendencies. But even so, the possibility suggested by Shugart and Mainwaring (that there are other factors which are decisive in explaining the success of parliamentary regimes and the failures of the presidential democracies) may itself be questionable. Consider the “widely recognized...strong correlation between British colonial heritage and democracy”. [41] The example of these former colonies is particularly welcomed by critics of presidentialism, because they interpret it as showing that even in countries with low levels of economic growth and with enormous poverty, democracy can succeed if the right governmental system is chosen. The idea suggested by Shugart and Carey is that the democratic record of those countries could be explained by factors independent of Parliamentarism. But suppose the case is proved that a certain level of tolerance and consensus is what explains the success of democracy in those countries. That is not the end of the story. Because the critics of presidentialism can reply that the existence of that culture is due (at least partially) to the existence of institutions that create and reinforce deliberative conducts and arrangements. In favor of these hypothesis counts the fact that all the British colonies that adopted a presidential system suffered breakdowns.
Shugart and Mainwaring defend the possibility that breakdowns in that region resulted from other conditions. In fact, a great number of alternative explanations have been provided to explain the failures of Latin American democracies. Some authors point to the cultural legacy of the colonial period and to the mixed influence of Spanish and catholic authoritarianism. Others think that the explanation lies in economic causes. These are perhaps the two prominent types of explanations. Now, are they fully independent from the institutional structure? The tradition of dependence on strong personal leaderships can be mitigated through a governmental structure providing incentives to the diffusion of power, or contrarily, emphasized through an institutional system that concentrates a great amount of power in a single person. This later possibility can reinforce the cultural reliance on leadership. The historical lack of tolerance in Latin American societies could also be weakened by the adoption of institutional devices which favored cooperation against confrontation. Instead, the adoption of presidential systems could operate as an institutional way of reproduction of cultural habits. A similar point can be made in relation to the economic explanations. Poverty, inequality, underdevelopment, stagnation, inflation, are indisputable causes of Latin American regime instability. But once again, how independent are they from the institutional components of these societies? For example, the tendency of presidentialism toward stalemate conspires against the enacting of good legislative norms. And recall the correlation between presidentialism and trends conducive to poverty and underdevelopment, like corporatism and corruption.[42]
Shugart and Mainwaring also recognize the impact of institutions on culture. When they enumerate probable reasons (other than regime type) to explain the correlation between British colonial heritage and democracy, they include “the tendency to train civil servants” and “the governmental practices and institutions”. Regarding the first it is widely accepted as one clear characteristic (and an advantage[43]) of parliamentary systems. And how can the second, “governmental practices and institutions”, be disentangled from the parliamentary system itself? Shugart and Mainwaring say that these “include but cannot be reduced to Parliamentarism.”[44] But if they include Parliamentarism, as the authors accept, then the point of the critics of presidentialism is right--Parliamentarism and other governmental practices and institutions help to explain the success of democratic regimes in these countries. (Recall that the standard criticism of presidentialism is not reductive. It doesn’t claim that governmental system is all that counts, only that it counts strongly).
The same line of arguments could be directed against their intent of attributing the continuance of democracy in parliamentary systems to the fact that most of them are “high-income” countries. Think in Canada, Israel and Japan, the three examples outside of Europe given by Shugart and Mainwaring. These countries are high-income ones today, but that was not precisely the case 50 years ago. Again, critics of presidentialism can read the same facts as showing that parliamentary systems favor economic growth.
We may conclude that the link between presidentialism and instability is still firm.

3.2. Some alleged advantages of presidentialism
I will now analyze some of the supposed advantages of the presidential model as exposed by its defenders.

3.2.1. Prospective voting superiority (Identifiability)?
Shugart and Carey[45] point to the superiority of presidentialism in “the degree to which voters can identify before the election the likely alternative governments that may emerge after the election.” They recognize that in parliamentary systems with two parties “there is a great range of identifiability”, but they show that in multiparty systems, it “tend to be far lower”. Linz contests the argument, claiming that “[i]n presidential elections the voter may know much less about who will govern than the voters of a party in most parliamentary systems”. This is due to the fact that in parliamentary systems prime ministers are not as free as most presidents in selecting the cabinets.
In addition, the work of Stokes referred in the previous section shows the enormous amount of U-turns in the government plans who of in the Latin American presidentialist systems, which denies the idea of identifiability. Finally, the existence, in many cases, of intermediate legislative elections also affects identifiability, since the resulting president of the election can not have (from the beginning or later) the legislative tools necessary to govern. In these cases, the electorate not even identifies alternative governments, since he can well be, after the elections, a situation of anarchy.[46]

3.2.2. Retrospective voting superiority (Accountability)?
Persson and Tabellini affirm that "a presidential-congressional regime that works well operates improve terms of accountability, because it deals better with the problem of agency between voters and politicians."[47] According to Shugart and Carey[48] “on the principle of maximizing direct accountability (in the sense of sanctioning incumbent elected officials by voting) between voters and elected officials, presidentialism is clearly superior to Parliamentarism...” That is due to the fact that “voters vote directly for an executive that cannot be removed by shifting coalitions.”
On the contrary Stepan and Skach affirm that "the principle of accountability in presidentialism is weaker than in the Parliamentarism".[49] Linz[50] denies the existence of this advantage, based on two arguments. The first one is that the argument cannot apply in all those countries where reelection is forbidden. This consideration is less forceful today, for a good number of presidential countries amended their constitutions to allow for reelections, (v.g. Brazil, Argentina, and Peru). Of course, Linz could insist that if reelection is not indefinitely allowed there will always be a period in which the president knows that he will not be able to appear to re-election, period in which there will be no way of punishing him or her through the vote. The second problem Linz sees in the idea of accountability is that the voters have to wait until the end of the government to exercise accountability. This is true, since under Parliamentarism there are more possibilities of forcing general elections. But let us not lose sight of the fact that voters can do a lot against a president they dislike before he or she finishes his or her mandate. They can vote against the presidential party or coalition in the mid term elections that almost all presidential systems allow. The real problem in these cases is that the ire against the president is inefficiently channeled in the presidential system, because it doesn’t provoke the end of the presidential tenure but the paralysis of the system, leaving the president without majority in the legislature. It is not that the president is not accountable in this occasions, but rather that the punishment is channeled both against him or her and against the system as a whole.
Additionally, there are a couple of other problems with the possibility for citizens of exercising accountability. First, it is an historical fact that in a lot of cases that possibility never arrived, because the systems broke down before. Secondly, a conceptual difficulty arises because accountability under presidential systems is (with all the caveats and objections above mentioned) to be exercised in a previously fixed moment. It could be argued, I think, that real accountability must not be subjected to that restriction. Accordingly, an official being accountable entails the possibility of being fired at any moment. On the contrary, the possibility of demanding accountability only at a fixed moment allows for the well known habit of presidents of having good (or even demagogic) conduct the previous few months before the general elections, taking advantage of the psychological tendency to diminish the weight of past events and remember more vividly the most recent.

3.2.3) Voting choice superiority?
Shugart and Mainwaring[51] claim that presidentialism has the advantage of providing voters with two electoral choices (one for the legislative and other for the executive). They see this characteristic as the flipside of Lipjhart critique based on the competing claims of legitimacy.
But is it a double or a fragmented choice? Dual legitimacy leads to conflict and/or immobilism, and that implies no efficient legislation and no fluid government. If this is true, then this feature of presidentialism, instead of allowing more real electoral choices, seems to divide the electoral choice into two weaker choices. On the one hand, the electoral choice may be frustrated because the Congress cannot legislate against the will of the president, (due to the existence of legislative veto power in the hands of the president), and the executive choice can be turned illusory by a Congress where the opposition has a majority: the cost of providing two votes is the danger of weakening both.
Paradoxically, the only way in which in a presidential system the two votes can be reflected in legislation and government, is when they are directed to the same party or coalition, or, in other words, when the two votes... are transformed in one. But this is the characteristic of electoral choices in parliamentary systems. As a conclusion, presidentialism doesn’t multiply the electoral choice, but divide it in two mutually neutralizing choices.

3.2.4) Checks and balances?
Shugart and Carey[52] suggest that the presidentialist type of separation of powers may be understood as a consensualist feature. After all, as they recall, this fact is one of the consequences of applying the Madisonian idea of “ambition against ambition.”[53]
Two situations must be distinguished. The first one is when the president’s party or coalition has not a majority in the Congress. In these cases, instead of check and balances, the picture tends to be one of stalemate and immobilism. In the second case, where the president has a majority supporting him or her in the Congress, the case for the defenders of presidentialism as anti-majoritarian is stronger. The legislators can analyze legislation on its merits, because the survival of the government is not endangered and so they are “freed from the threat of a vote of confidence”[54]. However, this advantage depends on a very contingent feature of the system, namely, the degree of party discipline. In those countries where parties are disciplined, in the case that the president has a majority in Congress, the presidential system doesn’t show any advantage over a parliamentary one in this respect. On the other hand, the power of the Congress to legislate without the consent of the President is illusory. Almost all presidential systems (roughly, 80% of them) provide the President with a legislative veto and require a supermajority of the legislature in order to override it.

3.2.5. Presidents as Arbiters?
Shugart and Carey[55] timidly refer to the possibility that “the president might serve as an above-partisan arbiter of political conflict”. This possibility, as the authors recognize, is unrealistic, and only arise in semi-presidential systems, when a president left in a minority still has a certain power in the formation of a cohabitation cabinet[56].
In standard presidential systems, the political dynamics tend to divide the society between the supporters of the president and the opposition. In this context, it is impossible for the president to place him or herself over the political disputes, simply because he or she is one of the poles of the confrontation.

3.2.6. Political Transparency.
It has been argued, on behalf of presidentialism, that this system provides more information about the preferences of relevant political actors. This is the alleged effect of “the back and forth between the branches.”[57] In parliamentary systems, by contrast, much information is less visible, and many executive preferences never get unveiled. This tendency, however, should be counterbalanced by the probability of the abuse of unilateral powers by the President, and by the negative effects of personalized power on public deliberation. Presidential regimes usually lack anything similar to the public debates in Parliament where PMs defend their policies and must respond to criticisms by the opposition.
As usual, it is important to take the U.S as an outlier, not as the paradigmatic case of presidentialism. As Eaton observes, the power of the US congressional committees is not replicated in other presidential systems. “These systems centralize policy negotiation in the executive branch despite the formal separation of powers.”[58]

4. Presidentialism or democracy. [59]
The problems of presidentialism are not simply mechanical. Many of its deficiencies turn it not only an undesirable system for its effects, but because it offends a robust notion of democracy. This is an additional and stronger reason to makes us worry about the current indifference towards presidentialism. I will focus now on the forms in which presidentialism collides with democracy. The discussions that we have had throughout these years in SELA may serve us, as a departure point, to delineate an ideal of democracy around the ideal of equality. My suggestion is to start from a simple idea of equality that we can use in diverse contexts. Thus, instead of arguing around the content of political equality, economic equality, or social equality, we could converge in the requirements of that simple idea of equality in the dominions of politics, the economy, social relations, etc. I hope to be able to show that the effort that I propose could clarify some of our discrepancies.
The simple idea of equality that I have in mind is the moral equality of persons, under any of its multiple formulations, namely, that all we are "equally human",[60] "equally valuable", "equally worthy” or "equally respectable". The different formulations are equivalent, in the context of this discussion. This idea is rooted in diverse traditions, like the Judeo-Christian and the liberal. The foundation of the idea usually makes reference to our capacity to give and to receive justice, [61] or to evaluate, [62] or to lead a life.[63] The most obvious implication of the idea is the disqualification of discriminatory actions and policies, or those establishing or perpetuating chaste societies or (at least strongly) hierarchical orders.
In order to arrive at less obvious implications of this idea (those that generate greater disputes, for example in relation to the economic regulations and social benefits), it may be useful to notice that the foundations of the idea of moral equality (the capacity to give and receive justice, to evaluate, or to lead a life) all converge in the value of choice. If some persons, but not others, were allowed the possibility of choosing types of life or conducts, then the state would be violating the demands of equality. [64]
What is the minimal content of a political system respecting the moral equality of people? First, democracy, as the political form of equality, requires to be designed in ways that do not amplify certain voices at the cost of silencing others, or that do not privilege the satisfaction of preferences of certain people at the expense of those of others. The deliberative character of democracy, as Carlos Nino insisted, responds to an egalitarian requirement. Democracy demands a dialogue between equals, through fora open to all, and a political system that absorbs the different perspectives and guarantees the and free confrontation of ideas and proposals. The exigency of a public justification of preferences tends to protect us against the inequality that would imply reducing the political dynamics to the mere aggregation of interests, given the asymmetry in our powers of negotiation.
Second, in the heart of democracy we find the majority rule, since any other form of decision would imply to give an extra weight to certain people or groups. The majorities must influence government in a significant way and must prevail without excessive institutional obstacles.
Third, equality gets institutionalized through human rights. Of course, I mean through all rights, not only the named as “classic”, but also social and economic rights. In order to stand in a relation of equals the satisfaction of certain basic interests cannot be subject to the swings of politics, since their justification is not tied to the amount of votes endorsing them. Therefore, the liberal egalitarian promise of the bills of rights included in the majority of our constitutions require states that on the one hand, promote tolerance, freedom of expression, the greater amplitude of life projects open to all, and on the other hand, reform the structural inequalities that deny to the great majorities the access to the complete set of human rights.
Now the three requirements of equality are harmed by presidentialism. Presidentialism degrades the deliberative quality of democracy in several ways. In the first place, through personalization of power it generates a public dialogue, if at all, very asymmetric, because it is rare that presidents participate in public deliberations giving and receiving reasons. Their interventions tend to adopt an imperial tone, above the level of daily politics. When things get complicated, the president abuses his or her role of Chief of State, denigrating the opponents as conspirators against the interests of the Nation. The form in which the presidents participate in the public discussions is the fullest negation of a rational, egalitarian and civilized dialogue. In the second place, the high polarization that presidentialism tends to create further deteriorates the quality of the public discussion, strengthening intransigent positions. In the third place, the tendency to the concentration of functions in the executive displaces legislatures from the center of the stage, hiding from public scrutiny the most important measures, usually adopted in the privacy of the ministerial offices.
Presidentialism is also hardly compatible with the second egalitarian requirement, collective self-government. Ackerman notices that under presidentialism a party needs to gain elections during one decade or more before being able to control key institutions.[65] This implies too unacceptable a distance from the ideal of collective self-government. The problems that we previously reviewed, like the weakness of the electoral options (options in the legislative and executive elections may be mutually neutralized) demonstrate that presidentialism impoverishes political equality. The high vulnerability to corporatist pressures and groups of interest fortifies this conclusion, since they imply providing a much easier access to power to those who have greater pre-political power.
Regarding the third egalitarian requirement (a climate of respect for rights), presidentialism is not the best option either. The authoritarian elements of presidentialism contribute to explain the phenomenon of "illiberal democracies",[66] the tendency of Latin American societies towards deterioration of individual rights, the weakening of freedom of expression, and very especially, the erosion of socioeconomic rights, threatened by the greatest inequality of the planet. Latin America needs structural reforms to correct enormous political, social, and economic inequalities. From this perspective, presidentialism is problematic for at least two reasons. In the first place, because this system makes reforms more difficult. In presidentialist systems the probability of the emergence of minority governments is much greater than in parliamentary regimes (two times greater, more precisely).[67] In these situations, in which the president and the pivotal legislator are politically far away, blockades take place which prevent, make more difficult, increase the cost, or delay reforms of the status quo. In the second place, these blockades generate instability and consequently political and social crises. It has been abundantly demonstrated that crises hit harder the disadvantaged sectors of society. Therefore, it seems reasonable to tie presidentialism with the perpetuation of the social and economic inequality.
Presidentialism is unacceptable from a democratic conception of government, because it collides with some requirements of an acceptable theory of democracy in a direct way (through personalization of power, the distortion of preferences, its low deliberative quality, its tendency towards political instability and the deterioration of rights). But there are other, more indirect ways, in which presidentialism is undemocratic, since it requires in order to avoid paralysis and blockades, of factors that are themselves objectionable from a democratic perspective, like non-proportional electoral systems, and, perhaps, intolerably ample delegations of legislative faculties to the executive. At this point a theory of democracy provides additional reasons to reject presidentialism, by calling attention to the fact that certain adjustments that would make the system more fluid, would also make it more undemocratic. On the one hand, democracy forbids fortifying the legislative power of the presidents. On the other hand, it enhances the importance of certain components of the political system, which are hardly compatible with presidentialism:
i. More power to the presidents?
One of the temptations against the presidentialist tendency to blockades between branches of government is to solve these tensions between presidents and parliaments in favor of the executives, granting them greater legislative powers (in the style of the reformed Brazilian Constitution) by easing legislative delegation to the president, or fortifying the presidential veto powers, or by recognizing the president legislative decree authority. This is unacceptable from a democratic perspective, because it implies to debilitate even more the role of parliaments, and to accentuate concentration of powers in the hands of the president.
ii. Less electoral proportionality?
Here we have an instance of indirect incompatibility between presidentialism and democracy.[68] In order for the presidentialist system to work more or less effectively, the electoral regime effectively would have to be of the type "winner takes all", since the proportional system tends to create manifold parties (although there are some exceptions). The idea is that proportional representation favors fragmentation, which makes it difficult for a single party to control the Congress, worsening the risk of minority governments and consequently of political crises. On the contrary, electoral systems as the uninominal favor the existence of two parties, excluding other political perspectives, and are, thus, more compatible with presidentialism. But democracy requires, as an issue of principle, a proportional system. This is required by the need of amplitude in the representation of political perspectives, and diminishes the tendency to excessively ambiguous platforms aiming to seduce all opinions (making almost inevitable to change horses once reached the government).[69]

5. Semi-presidentialism?
Given these principled and instrumental objections, it is time to wonder whether semi-presidentialism (for example, French-style one) is a promising alternative to presidentialism. An old suspicion exists to the effect that this system hides a potential of conflicts, and that if in the case of France these risks have not derived in serious crises of governability this was the result of the virtues of its political elite.[70] But we cannot trust that, in the context of Latin American political culture, the blurred limits between the "cohabitant" powers will be peacefully negotiated.
Perhaps the problem of defining the jurisdiction of the President and of the Prime Minister who "cohabit" can be resolved with good constitutional technique, as suggested by the Argentine Council for the Consolidation of Democracy in the eighties.[71] But another problem remains, in my opinion an insurmountable one. When the mixed system works with a President with majority in the Congress, the risks of authoritarianism, “caudillismo” and hegemony shine with the force of pure presidentialist systems. Democracy requires collective ways of deliberation and decision that are incompatible with the president-monarchs. When in the mixed systems the president has parliamentary majorities, democracy shakes.
If the mixed systems solve (arguendo) the paralysis problems when presidents lose popularity, they leave intact the authoritarianism that accompanies the presidents with parliamentary endorsement. My impression is that when the president has a majority in Congress, then the problems arise that worried us before, in particular the risk of excessive personalization of power and the low quality of the public dialogue. The key of the reform of the presidentialist systems of government is based in moving the center of power from the executives to the parliaments, and doing it in a permanent way, not only in times of crisis. Parliamentarism should not be a desperate therapy, or a lesser evil against blockades, but a lasting and stable change of our political societies.

6. The Latin American Academy and Presidentialism.
If presidentialism is indefensible, why, then, are we giving up the reform of our political systems? A general argument would be that the democratic consolidation requires abandoning the blind confidence in constitutional reforms: the hundreds of constitutional amendments and new constitutions that the history of Latin America[72] shows are a reason to think that, above a minimal threshold of decency of current constitutions, we must fulfill them before reforming them. One second reason to abandon the frontal critic of presidentialism may be the null probability of a radical change of system. Why sail against the current? But if the principled objections to presidentialism are right, the struggle against presidentialism is a democratic requirement, not a mere technical disquisition. Perhaps a constitutional reform ending with presidentialism is a chimera. But if we stop to think why this is so, we will find not only the explanation that the political interests that would be affected by the reform have at the same time a veto power against it. Part of the difficulty also lies in the indifference of the political and academic elites regarding the structural deficiencies of the presidentialist system.
Constitutional reforms do not come cheap. And I am open to accept the conclusion that, after analyzing the pros and cons, the best collective decision is to make the best of the rules we have, and to postpone to another era our desire to redraw our institutions. But at least, while analyzing the pros and cons, we ought to recognize that presidentialism degrades democracy, promotes a distorted and frequently irrational political deliberation, generates periodic crises that particularly affect the worse-off, stimulates electoral treasons, creates vulnerabilities to corruption, and amplifies corporatist pressures.
Maybe we are arriving at the point in which the institutional structure has been internalized by all the relevant actors, including the academia. Perhaps it has already become impossible to think outside the dynamics and the terms of presidentialism. But if there is any value in the academic activity in the region, it derives in part in resisting this type of endogeneity. Our vision of law must be transformative and critical. Our role is to question not rationalize current practices.

7. Conclusion.
In this work I vindicated the present relevance of the critic of presidentialism. In spite of the relative silence that the subject has deserved in the last years, presidentialism continues being a key problem for Latin American democracies. The evidence continues to accumulate: presidentialism exacerbates instability, foments paralysis situations, debilitates the reflexes of democracy against corruption, stimulates steep changes of government programs, and favors corporatist tendencies and inequality.
Constitutional democracy promises us self-government, but presidentialism only delivers imposition or paralysis. Democracy offers dialogue, but presidentialism condemns us to the hypocrisy of broken electoral mandates and to the monologue of imperial presidents. Democracy proposes equality of rights, but presidentialism foments an authoritarian culture, is ineffective in the provision of public goods, and lives from crisis to crisis, impoverishing and debilitating even still the disadvantaged groups.
In terms of the discussions of SELA, there are good reasons to think that presidentialism, as we always suspected, produce violence, threaten the rule of law, erodes fundamental rights and makes significant contributions to increase poverty and inequality.
Why should we accept this inferior form of democracy?
[1] I am grateful for discussions with Lucas Arrimada, Paola Bergallo, Juan Bertomeu, Gabriel Bouzat, Alberto Föhrig, Roberto Gargarella, Jaime Malamud, and the members of the course "Current Constitutional Issues" of the University of Palermo, 2005.
[2] J. L. Borges, “El Aleph”.
[3] For example, George Tsebelis, Veto Players, Princeton, 2002.
[4] David Samuels & Kent Eaton, "Presidentialism and, or, and Versus Parliamentarism: The State of the Literature and an Agenda Future Research ", 2002.
[5] Jose Cheibub and Fernando Limongi, in "Modes of Government Formation and the Survival of Democratic Regimes: Presidentialism and Parliamentarism Reconsidered "Annual Review of Political Science 5: 151-179, deny this difference, but there are many empirical studies that confirm it. See Samuels and Eaton and its references, p. 16-7.
[6] Kent Eaton, "Parliamentarism versus Presidentialism in the Policy Arena", Comparative Politics, 2000, pp. 355-376.
[7] Op. cit.
[8] “A messy and subjective affair" writes Eaton in "Parliamentarism..." p. 362
[9] Veto Players, Introduction, and Chapter 3.
[10] Eaton, "Parliamentarism...", p. 355-6
[11] Arend Lipjhart, "Presidentialism and Majoritarian Democracy: Theoretical Observations, in Juan J. Linz and Arturo Valenzuela (eds.), The Failure of Presidential Democracy, J. Hopkins, 1994, p. 100.
[12] Linz, “Presidential or Parliamentary Democracy: Does It Make a Difference?” in Linz and Valenzuela, op. cit., p.6.
[13] Lipjhart, op. cit., p. 97
[14] Gabriel Bouzat shows in his work the dramatic example of Argentina in 2001.
[15] Linz, op. cit., p. 14.
[16] O’Donnell, Guillermo, “Delegative Democracy”, Journal of Democracy, 1994.
[17] Nino, Carlos S., “Presidentialism vs. Parliamentarism”, in Presidentialism and Parliamentarism, Eudeba, Buenos Aires, 1988, (my translation).
[18] Susan C. Stokes, Mandates and Democracy. Neoliberalism by Surprise in Latin America, Cambridge, 2001.
[19] Stokes, p. 3, 14-15.
[20] See Kent Eaton, "Parliamentarism versus Presidentialism in the Policy Arena", Comparative Politics, April 2000, pp. 355-76, at 361.
[21] Samuels and Eaton, op. cit., p. 11 and ff.
[22] Torsten Persson y Guido Tabellini, Political Economics, MIT, 2000, p. 252: “Separation of powers in the presidential-congressional regime produces a smaller government with less waste and less redistribution but also inefficiently low spending on public goods… [l]egislative cohesion in the parliamentary regime, on the other hand, leads to a larger government, with more taxation and more waste, but also more spending on public goods and redistribution benefiting a broader group of voters.”
[23] World Bank Report on Inequality in Latin America, 2004, in worldbank.org.
[24] Jana Kunicová, "Political Corruption: Another Peril of Presidentialism ", available in http://www.hss.caltech.edu/~jana/perilous%20presidentialism%20feb%2006.pdf (last checked March 30, 2006)
[25] Susan Rose-Ackerman and Jana Kunicová, "Electoral Rules as Constraints in Corruption" British Journal of Political Science 35 (4), 573-606.
[26] The limit to the number of admissible presidential re-elections seems to be another hard data in Latin America. This leads to a dilemma: in order for the electorate to preserve its capacity to punish the president, indefinite reelections should be admitted.
[27] Kunicová, op. cit.. These conclusions should counterbalance the suggestion that presidentialism increases transparency.
[28] Linz, p. 63; Nino, Fundamentos..., p. 604.
[29] David Vogel, “Representing Diffuse Interests in Environmental Policymaking”, in Bert Rockman and R. Kent Weaver, Do Institutions Matter? Brookings Institution Press, Washington D.C., 1993, p. 268.
[30] Bruce Ackerman, “The New Separation of Powers”, Harvard Law Review, 113: 3, January 2000, p. 641.
[31] Stepan, A., and Skach, C, “Presidentialism and Parliamentarism in Comparative Perspective”, in The Failure of Presidential Democracy, vol. 1, 1994, Linz, Juan J. and Valenzuela, Arturo (Eds.), The Johns Hopkins University Press.
[32] Mainwaring, Scott, “Presidentialism, Multiparty Systems, and Democracy: The Difficult Equation”, cited in Shugart and Carey, op. cit, p. 38
[33] Riggs, Fred W., “Presidentialism: A Problematic Regime Type”, in Lipjhart, A. Presidential vs. Parliamentary Government (ed.), Oxford, 1992, p. 219.
[34] In Lipjhart, A., (ed.), op. cit, p. 118.
[35] Adam Przeworski, Michael A. Alvarez, José Antonio Cheibub, Fernando Limongi, Democracy and Development, Cambridge, 2000.
[36] Przeworski et al, Democracy…, pp. 128-31.
[37] Presidents and Assemblies, p.42.
[38] Linz, J.J., op. cit, p. 73.
[39] Shugart, M. S., and Mainwaring, S., “Presidentialism and Democracy in Latin America: Rethinking the Terms of the Debate”, in Shugart, M. S., and Mainwaring, S. (Eds.) Presidentialism and Democracy in Latin America, Cambridge University Press, 1997.
[40] Op. cit., p. 23.
[41] Op. cit., p. 23.
[42] Waisman, Carlos, “Reversal of Development in Argentina. Postwar counterrevolutionary policies and their Structural Consequences”, Princeton University Press, New Jersey, 1987. Similarly, Nino, Carlos, “The Constitution of Deliberative Democracy”, Yale University Press, Conn., 1996.
[43] See for example, Linz, J, op. cit., p. 32.
[44] Op. cit., p. 23.
[45] Op. cit, p. 45.
[46] An extreme case (once again) of denial of identifiability is the case of Argentina in 2001, when the person who assumed the Presidency was the one who had lost the last presidential elections.
[47] Torsten Persson and Guido Tabellini, op. cit., p. 253.
[48] Op. cit., p. 44/5.
[49] Op. cit. p. 136.
[50] Op. cit., pp. 10 and ff.
[51] Op. cit., p. 33.
[52] Op. cit., p. 46.
[53] Federalist 51, in Madison, Hamilton, Jay: The Federalist papers, (Clinton Rossiter, Ed.), Penguin books, c.1998, p. 322
[54] Shugart and Carey, op. cit., p. 46.
[55] Op. cit., p. 48.
[56] However, this role of arbiter has been attributed by Julio Faundez to Chilean presidents before 1970, (“In Defense of Presidentialism: The Case of Chile, 1932-1970”, in Mainwaring and Shugart, eds. op. cit., p. 300/20).
[57] Samuels e Eaton, p. 34
[58] “Parliamentarism...” at 365. And we may ask what transparency is displayed by the hundreds of “need and urgency decrees” never discussed by the Argentine Congress?
[59] This normative critique of presidentialism, though with a greater emphasis on equality, has been influenced by the treatment of the subject by Carlos Nino and Bruce Ackerman. Carlos Nino, Fundamentos de Derecho Constitucional, Astrea, 1992, pp. 569-619. Bruce Ackerman, “The New Separation of Powers”, Harvard Law Review, 113: 3, Jan. 2000.
[60] Nino, ...
[61] Kant, Rawls.
[62] Gregory Vlastos
[63] Gregory Vlastos
[64] Because our possibilities of choosing are bound to those of others, equality is at the same time relational and comparative. Material inequalities may generate asymmetrical relations, and that is a reason to attenuate them, in order to make it true that we, and not others, are in control of our lives. But material discrepancies may not have that consequence (a strong sense of self-esteem or a purely ascetic conviction may make us avoid feeling that our life is in the hands of others). Even in those cases, our possibilities of choosing would be unjustly limited if the inequalities were imposed on us without reference to our decisions. This is the conviction that animates the idea of the social and natural lotteries. Both the class in which we are born and the genes that we bring to the world lack enough moral force to validate inequalities. The only things that can validate them are our decisions. This is the positive idea behind what is named luck-egalitarianism. The object of neutralization or equalization, luck, is just the negation of the positive aspect to be honored, the autonomous election of life projects. Therefore, it would be better to abandon the label of luck egalitarianism and adopt a different, and clearer, one: choice egalitarianism. The undeniable difficulties -metaphysical and practical- that surround the idea of luck neutralization do not have to move us away from what is important: that only our decisions must determine our position in the community.
[65] Ackerman, op. cit., p. 650.
[66] Peter H. Smith, Democracy in Latin America, Oxford, 2005.
[67] Samuels & Eaton, at 10.
[68] Cfr. Ackerman, “The New Separation…” p. 655.
[69] The abandonment of proportional representation is not only undesirable, but very unlikely, because it needs the acquiescence of the same actors that would be wiped out from the political scene after such reform.
[70] See, for example, Ackerman, p. 648.
[71] Consejo para la Consolidación de la Democracia, Dictamen sobre la Reforma Constitucional, Buenos Aires, Eudeba, 1988
[72] Keith Rosenn, “The Success of Constitutionalism in the United States and its Failure in Latin America: An Explanation,” University of Miami Inter-American Law Review 22. 1, 1990.