Government and Opposition Volume 25 Issue 1 1990 [Doi 10.1111٪2Fj.1477-7053.1990.Tb00747.x] Arend Lijphart -- The Southern European Examples of Democratization

A rend Lgphart

The Southern European Examples of Democratization: Six Lessons for Latin America*

THE RETURN TO DEMOCRACY OF SPAIN, PORTUGAL, AND Greece in the 1970s is an encouraging and inspiring example to democrats everywhere - but especially to Latin American demccrats because of their regions close historical and cultural ties with two of the Southern European countries. However, apart from the general feeling of optimism that the Southern European experience legitimately engenders, are there any specific lessons and lessons specifically relevant to Latin America that can be learned from it? In this article, I shall suggest six such lessons. Some of these are positive lessons - examples to be followed, such as choosing a form of democracy that is suitable to a countrys size and to its political and social divisions; others are negative - examples to be avoided, such as Portugals and Greeces experimentation with a presidential form of government. Some lessons are based on common characteristics of the new Southern European democracies; others concern traits on which they differ.

In a recent co-authored study of these three newly democratic countries, plus Italy, the fourth Southern European democracy, I have shown that the differences among them are more important than the similarities: differences in their forms of democracy and in the process of moving from authoritarian to democratic regimes. The differences in their forms of democracy are particularly striking. As I have argued in my book Democracies,

*This article is a revised version of a paper presented at the International Seminar on Las Perspectivas de la Estabilidad Democritica en 10s Paises Andinos, organized by the Department of Political Science of the Universidad de 10s Andes, in Villa de Leyva, Colombia, August 8- 12, 1988.

Arend Lijphart, Thomas C. Bruneau, P. Nikiforos Diamandouros, and Richard Gunther, A Mediterranean Model of Democracy? The Southern European Democracies in Comparative Perspective, West European Politics, 11, 1, January 1988, pp. 7-25.

I .

Highlight

SIX LESSONS FOR LATIN AMERICA 69

two basic forms of democracy, majoritarian and consensus, differ along two dimensions.' Each of these dimensions is divided in four categories in Figure 1, creating a matrix with sixteen cells. The democracies classified in these cells are the three newly democratic Southern European countries plus the twenty-one democracies that have been continually democratic since the end of the Second World War; since France underwent a major change in democratic regime in 1958, the French Fourth and Fifth Republics are treated as two separate cases.

The executives-parties dimension is based on a cluster of five characteristics of the party and electoral systems and of the arrangement of executive power. The majoritarian and the contrasting consensual elements are 1) single-party governments versus broad coalition governments, 2) executive dominance versus executive-legislative balance, 3) two-party versus multiparty systems, 4) parties that differ mainly with regard to socio-economic issues versus parties divided by religious, cultural, foreign policy, and regime support issues as well, and 5) plurality or first-past-the-post electoral systems versus proportional representation. Each of these contrasting characteristics represents a continuum from pure majoritarianism to pure consensus democracy; they were operationalized and scores on them were assigned to the 25 democratic systems. Since these scores use different scales, they were all standardized (so as to have a mean of 0 and a standard deviation of 1). The total scores on the executives-parties dimension as a whole are the averages (again standardized) of the five individual scores. Democracies that are majoritarian on the executives-parties dimension are shown in the top half of Figure 1, the consensus democracies at the bottom.

The second dimension has to do with the three related variables of 1) government centralization, measured in terms of the central government's taxing powers, 2) constitutional flexibility, ranging from an unwritten constitution to a written constitution that is difficult to amend and protected by judicial review, and 3) unicameralism versus strong bicameralism. Since these differences are commonly associated with the contrast between federalism and unitary government, this dimension is called the federal-unitary dimension. The country scores on this dimension are again the standardized averages of the three standardized

Arend Lijphart, Democracies: Patlerns ofhfajoritarian and Consensus Government in Twenty- One Countries, New Haven, Yale University Press, 1984.

Highlight

70 GOVERNMENT AND OPPOSITION

FIGURE 1 Twenty-Five Democracies Chsijied According to the Two Majoritnrian-Consensus Dimensions

I1 Federal-unitary dimension

I Executives- Parties Dimension

Strongly majoritarian

Majoritarian

Consensual

Strongly consensual

Strongly majoritarian Majoritarian

New Zealand United Kingdom

Iceland

Israel

Greece Ireland

France v Luxembourg Sweden

Belgium Denmark Norway Portugal

Finland France IV Netherlands

Strongly Consensual consensual

~

Austria

Spain

Australia Canada United States

Germany Japan

Switzerlan

Notes: 1. The dividing points between the four categories of each dimension are the standardized means of .75, 0. and - .75.

2. The period covered is approximately 1945-80 for most of the countries and approximately 1975 - 86 for Spain, Portugal and Greece.

Source: Adapted from Arend Lijphart, Thomas C. Bruneau, P. Nikiforos Diamandouros, and Richard Gunther, A Mediterranean Model of Democracy? The Southern European Democracies in Comparative Perspective, West European Politics, 11, 1 January 1988, p. 12.

scores. The unitary (majoritarian) democracies are shown on the left-hand side of Figure 1, the federal (consensual) democracies on the right.

As Figure 1 shows, there are relatively few democracies that occupy an extreme position on both dimensions. New Zealand and the United Kingdom are unambiguous examples of the majoritarian model and Switzerland of the consensus model. Israel is a virtually perfect example of majoritarianism on the first dimension and unitary government on the second; Australia, Canada, and the United States exhibit the exactly opposite characteristics to a high degree. Most of the other countries -

SIX LESSONS FOR LATIN AMERICA 71

including the four Southern European democracies, italicized in the figure - are more in the middle. Spain, Portugal, Italy, Greece clearly do not form a distinctive cluster: they are located in different and non-contiguous cells, and they differ from each other particularly as far as the executives-parties dimension is concerned.

Spains, Portugals, and Greeces contemporaneous democrat- ization in the 1970s should not divert our attention from the different circumstances of their return to democracy. For one thing, their previous authoritarian periods were of widely different durations: a mere seven-year interlude in Greece, compared with more than a third of a century of authoritarianism in Spain, and almost half a century in Portugal. Moreover, their authoritarian regimes came to an end as a result of very different events: the dictators death in the case of Spain, war and institutional exhaustion in Portugal, and a combination of internal crisis and foreign-policy adventurism in Greece. Finally, the transition to democracy occurred under quite dissimilar ideological leaderships: the conservatives in Greece, the left- leaning military in Portugal, and a broad coalition in Spain that practised what was referred to as the politics of consensus.

There are also important differences within Latin America to be noted, as well as differences between Southern Europe and Latin America. In the concluding part of this article I shall consider to what extent the inter-regional differences affect the relevance of the Southern European lessons to the Latin American situation. But let me first turn to the substance of these lessons.

LESSON ONE: OPPORTUNITIES FOR DEMOCRATIC ENGINEERING

In contrast with the striking differences among the Southern European democracies with regard to their forms of and transitions to democracy, there are many background characteristics that they have in common - cultural, social, economic, historical-developmental, and geographical. For instance, in addition to their shared Southern European geographical location, they are economically less developed than most other European countries. Spain, Portugal, and Italy have agricultural sectors characterized by latifundia in the south and small farms in the north - a division that has strongly affected politics in the past and present. These same three countries share

Highlight

Highlight

72 GOVERNMENT AND OPPOSITION

a Latin culture and also a common religious condition; unlike in many other European countries, there was never a Catholic- Protestant split but a deep and politically very salient clerical- anticlerical cleavage.

The contrast between these similarities in background conditions and the differences with regard to type of democratic regime and the process of democratization is theoretically very significant. It reveals the limitations of socio-economic and cultural reductionist arguments: political institutions and the basic rules of the game of democratic politics are not merely a superstructure that grows out of a socio-economic-cultural base. Politics, including democratic politics, has an independent life of its own.

This theoretical conclusion has a significant practical relevance. It means that there is ample room for political and constitutional engineering. While politics remains the art of the possible, democratic engineers need not feel too constrained in making the necessary choices of rules and institutions that will serve their democratic systems best.

LESSON TWO: DEMOCRACY IN DIVIDED SOCIETIES

Several conclusions as well as practical lessons emerge when we try to explain why particular countries occupy particular positions in Figure 1. There are three explanations: the degree of pluralism, population size, and the influence of the majoritarian Westminster model on countries with a British heritage. The first and second contain positive practical lessons, the third a negative lesson.

Let us turn to the question of pluralism first. Figure 1 shows a connection between both dimensions of the majoritarian- consensus contrast on the one hand and the degree to which the countries are plural societies on the other. Plural societies are societies that are sharply divided along religious, ideological, linguistic, cultural, ethnic, or racial lines into virtually separate sub-societies with their own political parties, interest groups, and media of communication. No country is either completely plural or completely nonplural, of course, but we can make a rough threefold classification into plural, semiplural, and nonplural societies. As we move from the upper left-hand corner to the lower right-hand corner of Figure 1, we encounter plural and semiplural societies with increasing frequency. The four cells on

Highlight

SIX LESSONS FOR LATIN AMERICA 73

the diagonal from Israel to Australia contain nonplural societies (Denmark, Norway, Portugal, and Australia) and plural and semiplural ones in approximately equal numbers. Above and to the left of this diagonal, all countries except Austria, the French Fifth Republic, and Luxembourg are plural or semiplural; below and to the left of the diagonal, only one country -Japan - is nonplural. Of these four exceptions, Austria is probably not really deviant; my coding system probably resulted in a too high majoritarian score for this country on the first d i m e n ~ i o n . ~ The other three are true exceptions, but the important point is that there are only three such deviant cases out of a total of 25. The relationship would be even stronger if it were not for the other two factors - British political traditions and population size - that also have strong influence. I shall discuss these subjects shortly.

The general pattern is that countries with significant societal divisions tend to adopt forms of democratic government that can accommodate these divisions, namely, rules and institutions of consensus democracy. And the practical lesson for political engineers is explicitly to take such deep divisions and differences into account wherever they exist and to be creative and constructive in establishing the appropriate consensus-oriented and consensus-inducing democratic arrangements. Moreover, it is important not to think of plural societies too narrowly as countries divided by ethnic or other primordial cleavages. Consensus democracy is clearly needed by all countries that have deep divisions of any kind or that face immense problems, including countries with a recent history of military dictatorship and civil war, countries with huge socio-economic inequalities, and so on. Many Latin American countries can be described in these terms.

If we look at the placement of the four Southern European democracies in Figure 1, we see that, as expected, religiously and ideologically divided Italy and linguistically plural Spain are located to the right of nonplural Greece and Portugal. On the

One of the reasons for Austrias high majoritarian score is that the country was governed from 1949 to 1966 by a grand coalition of the large Socialist and Peoples Parties, neither of which had a majority in parliament. Technically, therefore, this coalition had to be classified as a highly majoritarian minimum winning coalition (or bare-majority coalition) in spite of the fact that together these parties enjoyed overwhelming parliamentary support. If the minimum-winning criterion could be relaxed so as to allow the Austrian grand coalition to be classified as an inclusive oversized coalition government, Austria would move down one cell in Figure 1.

74 GOVERNMENT AND OPPOSITION

other hand, it is surprising that on the executives-parties dimension Greece is among the most majoritarian regimes and that Spain is on the majoritarian side, too, given the severe political and ideological divisions in these countries in the recent past and the continuing cultural-linguistic cleavages in Spain. Greece had single-party majority cabinets between its return to democracy in 1974 and the middle of 1989, and Spain has abandoned its earlier reliance on the politics of consensus and has had one-party majority governments since 1982. This emphasis on straight majority rule has not led to a serious crisis, mainly because the ruling parties have behaved with moderation, but it has entailed a serious risk for these newly democratic countries - a risk that I would not want to recommend as an example to be followed.

LESSON THREE: PR AND PROPORTIONALITY

The next lesson follows immediately from the above point. What has made single-party majority cabinets possible in Spain and Greece is that their parliamentary elections have resulted in many victories for one party; this is true for four of the six elections in Greece held so far (the 1989 elections are the exception) and in the last three of the five Spanish elections. It is important to note that six of these seven one-party majorities won at the polls were what Douglas W. Rae calls manufactured majorities: a party winning a majority of the legislative seats with only a minority of the popular vote.4 Such manufactured majorities are quite common in plurality systems but rare under proportional representation (PR). Spain uses PR but applies it in small districts, which discriminates against minor parties and in favour of the large parties. Until the 1989 elections, Greece used a PR system that was usually referred to as reinforced PRY, but what was being reinforced was the large parties rather than proportionality.

The lesson to democratic engineers is: if you want to encourage power-sharing coalition government instead of one-party rule, and if you want to avoid the artificial manufacturing of majorities, you should choose a PR system that is proportional in reality as well as in name - unlike the Spanish and Greek examples. This lesson is of special importance to Latin America

Douglas W. Rae, The Political Consequences of Electoral Laws, New Haven, Yale University Press, 1967, p. 74.

Highlight

SIX LESSONS FOR LATIN AMERICA 75

since many countries in this region have strong traditions of conducting their legislative elections by PR. It is worth noting that these traditions constitute an unexpected deviation from the United States model of mainly plurality and majority electoral systems; as I shall show later, the American model of democracy is very strong in Latin America in other respects, especially as far as the reliance on presidential forms of government is concerned.

It should also be pointed out, however, that presidentialism and PR elections of the legislature are related to each other: presidentialism tends to limit the operation of PR. The reason is that the presidency is the biggest political prize to be won and that only the largest parties have a chance to win it. This creates a major advantage for the large parties and a disadvantage for the smaller ones not only in the presidential elections themselves but also indirectly in the legislative elections even when the latter are conducted by PR. As Matthew Shugart has pointed out, this is especially the case when the legislative election is held at the same time or shortly after the presidential e l e ~ t i o n . ~ In addition to this indirect effect of presidentialism on limiting proportionality and minority representation, presidentialism also has a direct negative effect on proportionality, of course: the fact that a presidential election entails the election of one person necessarily means that plurality or majority methods have to be used and that PR is logically excluded.

LESSON FOUR: THE TRADITION OF PRESIDENTIALISM

The second explanation of the configuration of democracies in Figure 1 has an indirect but quite important relevance to our concerns: the countries with a British political heritage (Britain itself, New Zealand, Ireland, Australia, Canada, and the United States) are all highly majoritarian on the executives-parties dimension and are all located in the top row of the figure. The influence of the Westminster model as a normative example has tended to interfere with the need for more consensual arrangements in some of the countries. In particular, had it not been for the strong British influence, it is quite unlikely that

Matthew S. Shugart, Duvergers Rule and Presidentialism: The Effects of the Timing of Elections, paper presented at the Annual Meeting of the American Political Science Association, Washington, D.C., 1988.

Highlight

76 GOVERNMENT AND OPPOSITION

linguistically plural Canada would have developed along such strongly majoritarian lines (on the first dimension).

There are other examples of dominant and potentially dangerous foreign models that have been obstacles to an optimal choice of democratic type based on a countrys needs. The presidential form of government of the French Fifth Republic was such a model for Portugal and Greece; because of the fortunate circumstance of having a monarchy, Spain escaped this influence. And the US presidential model has had a powerful impact in Latin America. Without going into a full evaluation of the advantages and disadvantages of presidentialism, let me emphasize two serious problems.6 One is that it entails a strong predisposition toward majoritarian democracy: it means the concentration of all executive power in the hands of one person, which is incompatible with broad coalition government and power-sharing. Guillermo ODonnell notes the infrequent recourse to formal and explicit political and economic pacts as transitional devices in Latin America, in contrast with pactismo in Southern Europe. He tries to explain this phenomenon in terms of popular pressures that make compromises difficult and in terms of the absence of strongly institutionalized party system^.^ In addition, it seems to me, presidentialism is to blame: by its very nature it is inimical to collective and collegial decision-making and hence to compromises either on an ad hoc or regularized basis. Moreover, where pacts have been successfully established in presidential regimes, as in Colombia and Venezuela, they have entailed more drastic limitations on democratic participation and the rights of oppositions than similar pacts in more flexible parliamentary systems.

The second major problem of presidentialism is that it is based on the principle of separation and balance of executive and legislative powers but that, in practice, most presidential systems have found it impossible to achieve this balance. The United States itself is an exception, although historically it has also experienced swings between, in Woodrow Wilsons words,

For an excellent and much more extensive analysis, see Juan J. Linz, Democracy, Presidential or Parliamentary: Does It Make a Difference?, paper presented at the Annual Meeting of the American Political Science Association, Chicago, 1987.

Guillermo ODonnell, Introduction to the Latin American Cases, in Guillermo ODonnell, Philippe C. Schmitter, and Laurence Whitehead (eds), Transitions from Authoritarian Rule: Prospects for Democracy, Baltimore, Johns Hopkins University Press, vol. 2 , pp. 11-12.

SIX LESSONS FOR LATIN AMERICA 77

Congressional government and the so-called imperial presidency .* The usual interpretation of presidential power in Latin America is that is has overwhelmed the power of the legislature and that it has tended to turn into dictatorial power. One typical response to this danger has been to limit the presidents right to be re-elected. As Harry Kantor points out, however, such rules are infractions upon true democracy, which demands that voters be allowed to vote for whomever they choose.g I would add that they also conflict with the democratic assumption that the opportunity to be re-elected is a strong incentive for elected officials to remain responsive to the voters wishes. The fact that parliamentary executives do not need to be limited by such basically undemocratic rules shows that they are much safer for democracy. In addition to restrictions on presidential succession, many other limitations on presidential prerogatives have often been adopted, based on the same fear of too strong presidential power. Ironically, this has often meant that presidents have become virtually powerless and frustrated. In fact, the increasingly prevalent interpretation of the problem of presidentialism in Latin America is that presidents suffer from too little instead of too much power.

Portugal had a French-style presidential government - a strong president combined with a cabinet dependent on the legislatures confidence - from 1976 to 1982. The constitution gave the president extensive powers, and his popular election added to his political stature. However, the presidents powers were severely reduced in the 1982 constitutional revision and, although popular election was not changed, Portugal reverted to a parliamentary system - similar to the Austrian, Irish, and Icelandic parliamentary systems which have weak, albeit popularly elected, presidents, too. Greece also adopted a strong presidency, inspired by the French model, although the president was not popularly elected; President Constantine Karamanliss great personal prestige partly compensated for this lack of popular legitimation. The 1986 constitutional amendments eliminated

* See Fred W. Riggs,The Survival of Presidentialism in America: Para-Constitutional Practices, International Political Science Reuiew, 9, 4, October 1988, pp. 247 - 78.

Harry Kantor, Efforts Made by Various Latin American Countries to Limit the Power of the President, in Thomas V. DiBacco (ed.), Presidential Power in Latin American Politics, New York, Praeger, 1977, pp. 23 - 24.

See Scott Mainwaring, Presidentialism in Latin America: A Review Essay, Latin A m i c a n Research Review, forthcoming, 1990.

78 GOVERNMENT AND OPPOSITION

almost all of the special powers of the president, making the regime unambiguously parliamentary.

These Portuguese and Greek examples of constitutional adaptation should be regarded as positive and highly instructive models to democrats, especially in the Western hemisphere, who are predisposed to think in presidential terms. Shifting from presidential to parliamentary government entails a drastic and difficult, but clearly not impossible, regime change. On the other hand, there are two reasons why such a change was easier for Portugal and Greece than it would be for the Latin American democracies. One is that the French presidential model already contains some parliamentary features - the model is, in fact, often called merely semi-presidential - and hence that it can be turned into parliamentarism more easily than the US presidential model. Secondly, presidential government in Greece and Portugal did not last long enough to become a firm tradition, while the problem in Latin America is not just the influence of the US presidential model but also the fact that presidentialism has become a strong Latin American tradition, too.

LESSON FIVE: POPULATION SIZE, DECENTRALIZATION, AND FEDERALISM

Let us now turn to the third and last explanation of the distribution of the 25 democratic systems in Figure 1 : the respective sizes of the countries populations. Figure 1 shows that this variable is correlated with the federal-unitary contrast, although the relationship ir: obviously not a perfectly monotonic one. As we move from strong majoritarianism (unitary government) on the left to strong consensus (federalism) on the right, we find that population size tends to go up. The most striking exceptions are the United Kingdom which, in terms of this explanation, is placed too far to the left in the figure, and Switzerland which is similarly much too far on the right-hand side. The four Southern European democracies display roughly the differences that we would expect on the basis of their different population sizes: the two smaller countries, Greece and Portugal, with populations of about ten million each, are located higher than Spain and Italy, which have respectively about four and six times larger populations.

It is not surprising that we find this link between population size and type of democracy: larger countries need more provisions

SIX LESSONS FOR LATIN AMERICA 79

for regional autonomy and more of the supporting federal institutions than small countries. The practical lesson for democratic engineers is that the form of democracy chosen should be appropriate to their particular country in this respect. Democracies do not have to be formally federal in order to be rated as federal (consensual) in Figure 1. The criteria for classification on the federal-unitary dimension are the degree of government centralization, the organization of the national legislature (unicameralism versus strong bicameralism), and constitutional flexibility; in other words, whether or not a country is formally federal is not a criterion, And, in fact, of the nine countries classified in the consensual and strongly consensual cells on the federal-unitary dimension in Figure 1, three are formally unitary states: Japan and our two more federal Southern European countries, Spain and Italy. On the other hand, formal federalism does appear to be a factor of considerable importance: the six countries that have explicit federal constitutions are all on the right-hand side of the figure, and five of the six are on the far right-hand side. In Latin America, there is a similar relationship between population size and formal federalism. Most of the larger countries have federal constitutions - Brazil, Mexico, Argentina, and Venezuela - but there is no strong tradition of actual government decentralization. For instance, Daniel J . Elazar states that in Mexico the states exercise limited autonomy [under] a very strong federal government, and he describes Argentina as a federation in which most power is lodged in the federal capital. O n the other hand, Colombia is a decentralized unitary state in which decentralization is maintained by the strong demands for autonomy that exist in some of the major provinces.

In the previous section, I emphasized the negative influence of the United States model of presidential government. I should now state with equal emphasis that in two other respects the US model should be regarded as beneficial for the Latin American countries, since the model also includes federalism and a strong written constitution; I shall discuss the latter in the next section. As far as the federal model is concerned, Sir Arthur Lewiss analysis of the failure of democracy in West Africa is highly pertinent. He attributes a large part of the blame to the pernicious

I Daniel J. Elazar, Arrangements for Self-Rule and Autonomy in Various Countries of the World: A Preliminary Inventory, in Daniel J. Elazar (ed.), Federalism and Political Znfegration, Lanham, Maryland, University Press of America, 1984, pp. 230 - 31.

80 GOVERNMENT AND OPPOSITION

influence of the majoritarian model, including its unitary and centralist bias. These states would have fared better if they had not had to assume that British and French constitutional ideas were superior to all others. With an American heritage, they would have taken the federal idea for granted, and it would have been the centralizers who were arguing an unpopular case. He concludes that the West African countries will need much un- brainwashin before they grasp their problems in true perspective? Lewis would undoubtedly regard the Latin American countries as very fortunate for not having to undergo the same un-brainwashing due to the strong influence of the US model.

Because the federal component of the US model must be judged positively but the presidential component negatively, it is important to stress that the two are not logically linked with each other. Klaus von Beyme points out that right into the twentieth century the prejudice persisted vigorously that federalism and parliamentary majority government could not be combined. Citing the Argentinian, Brazilian, Mexican, and Venezuelan examples, he adds that this message was taken especially to heart in Latin America. Among Europeans, the examples of federal- cum-parliamentary government in Australia and Canada were long ignored, and it is only after the adoption of the similar constitutional mixes in Austria and West Germany that this prejudice has died out - at least in Europe.13 To the extent that the myth persists in Latin America, it also deserves to die out there.

LESSON SIX: CONSTITUTIONAL PROTECTION

While the new Southern European democracies have different positions on the federal-unitary dimension, they are strikingly alike with regard to one of the variables that goes into this dimension: the rigidity of their constitutions. All three have written constitutions that can be amended only by extraordinary majorities and are protected by judicial review. Their emergence from dictatorial rule accounts for much of this similarity. Of the other democracies in Figure 1 only seven have equally firm and

W. Arthur Lewis, Politics in West Africa, London, Allen & Unwin, 1965, p. 55. l 3 Klaus von Beyme, America as a Model: The Impact ofAmrican Democracy in the World,

New York, St Martins Press, 1987, p. 76 (italics omitted).

SIX LESSONS FOR LATIN AMERICA 81

protected constitutions, and three - Austria, Germany, and Japan - have similar backgrounds of authoritarianism.

Since this background is unfortunately also common in Latin America, the Southern European model of strong constitutional protection is a positive example. Flexible constitutions, such as the unwritten constitutions of Great Britain and New Zealand and the surprisingly many written constitutions that can be amended by majority rule and/or are not protected by the courts ability to test the constitutionality of laws, are a luxury that some older and completely self-confident democracies may be able to afford. It is a luxury that cannot be recommended for new and less firmly established democratic systems.

As indicated earlier, constitutional rigidity and judicial review are important parts of the US model of democracy; Karl Loewenstein even called them Americas most important export. l 4 In fact, it can be argued that judicial review is a logical corollary of both separation of powers (presidential government) and division of powers (federalism) since an impartial authority is needed to decide where exactly these powers should be separated and divided.I5 O n the other hand, there is no logical reason why judicial review and rigid constitutions cannot be combined with parliamentary government or, for that matter, with unitary and centralized systems. In practice, as von Beyme points out, today more parliamentary systems than presidential ones have established a variation on the theme of constitutional jurisdiction. l6 Very clearly, the combination of parliamentary government, strong constitutional protection, and, for the larger countries, federalism and decentralization is both possible and advisable.

CONCLUSION

I believe that these six lessons are both valid and valuable in spite of the obviously many and considerable differences among the Southern European countries, among the Latin American countries, and also between Southern Europe and Latin America. I have already repeatedly referred to the first two sets of

I Cited in von Beyme, America as a Model, op. cit., p. 85. l 5 See K. C . Wheare, Federal Government, 4th ed., New York, Oxford University Press,

l 6 Von Beyme, America m a Model, op. cit., p. 85. 1984, pp. 53 - 74.

Highlight

82 GOVERNMENT AND OPPOSITION

differences, those within Southern Europe and within Latin America. However, the most important reason for doubting the relevance of the Southern European examples for Latin America would be the difference between the two regions. It is precisely this difference that several scholars have recently emphasized. For instance, Guillermo ODonnell writes that the prospects for political democracy in Latin America are not very favorable, certainly less so than in Southern Europe. ODonnells co- author of the authoritative four-volume work Transitions from Authoritarian Rule, Philippe C . Schmitter, is equally pessimistic. In my opinion, both this forecast and the grounds for it are exaggerated.

One of the reasons for their pessimism, with which I find myself largely in agreement, concerns the considerable socio- economic differences between the two regions. What is important here is not the overall level of economic development. Italy and Spain may be economically more developed than any of the Latin American countries, but Portugal and Greece are clearly not. The crucial difference is that socio-economic inequality is substantially greater in Latin America than in Southern Europe. ODonnell speaks of acute, pervasive, and blatant inequalities in the latter region. l8 But ODonnells diagnosis should yield two conclusions, instead of one: this major cleavage makes viable democracy less likely, but the likelihood can be improved by using the appropriate consensual instruments.

Another reason why ODonnell and Schmitter are pessimistic is that the Latin American countries do not share the Southern European advantage of the relatively modest, not to say minor, role played by the armed forces in the defunct regime. But Greece is at least a partial exception to this pattern and, as Schmitter himself points out, there is the countervailing disadvantage that the Southern European countries have . . . experienced bureaucratic-authoritarian rule more continuously and for a longer period of time. Another difference is that the international context has been more favourable to Southern Europe than to Latin America. For one thing, the United States, whose policies toward democratization in Latin America have been ambiguous and variant from one case to another, has consistently supported it in Southern Europe. But Schmitter himself concludes that such external forces have been of

l 7 Guillermo ODonnell, Introduction to the Latin American Cases, op. cit., p. 14. ibid., p. 11.

Highlight

SIX LESSONS FOR LATIN AMERICA 83

minor importance and that the transitions from authoritarian rule and the prospects for democracy must be largely explained in terms of national forces and calculations. l9

Schmitter also emphasizes the differential role of dominant models: The Southern European countries may have also benefited from their distance from the American system of government. Western Europe may seem to be monolithically democratic in the contemporary period, but beneath that overall similarity lie many differences in institutional configuration [such as] pure parliamentarism, semipresidentialism, and consociation- alism, coupled with a wide range of party systems, electoral arrangements, and territorial distributions of authority. The Latin American political engineers, by contrast, may be compelled to choose from a more restricted menu. The hegemony of the United States as a model of political democracy, not to mention the legacy of their own nineteenth-century constitutions, makes it less likely that they will deviate from the presidentialist, bicameral, formal checks-and-balances, first-past-the-post ideal with its implied two-party system. After all that I have said myself about the importance of normative models of democracy, I have to endorse much of what Schmitter is arguing, but I would add three qualifications. First, as I have emphasized several times, parts of the US model are positive and valuable for Latin America. Secondly, Schmitters statement that the Latin Americans may be compelled to choose from the US model and their own political traditions is too strong and too deterministic; in fact, there has been considerable interest in studying the European, and especially Southern European, examples. My third, relatively minor, point concerns the choice of electoral systems. Here the variation in Europe is much more limited - PR is the nearly universal norm - and the Latin American experience is much more varied - including a strong PR tradition - than Schmitter claims.

Finally, in spite of all his pessimism, ODonnell argues that the chances for democracy in Latin America may not be so small after all because of the widespread revulsion against the excesses of recent authoritarian regimes. As a result, never has the

Philippe C. Schmitter,An Introduction to Southern European Transitions from Authoritarian Rule: Italy, Greece, Portugal, Spain, and Turkey, in ODonnell, Schmitter, and Whitehead (eds), Transitions frum Authoritarian Rule, op. cit., vol. 1 , pp. 4 - 5.

ibid., p. 9.

Highlight

Highlight

84 GOVERNMENT AND OPPOSITION

ideological prestige of political democracy been higher in Latin America than now - offering an unprecedented opportunity for the establishment and consolidation of democracy. This is the terrain , he continues, where that unpredictable combination of virtzi on the part of leaders, and fortuna in the combination of circumstances, may make the crucial difference . The message of this article has been that the political engineers virtzi should include not only a strong commitment to democracy but also the willingness and ability to be creative, to be not over constrained by existing political traditions, to examine all of the available options, and to learn from the positive and negative examples of other democracies.

ODonnell, Introduction to the Latin American Cases, op. cit., p. 11

Highlight