THfc THIRD WAVE SHORT LO Democratization in the Late Twentieth Century \ By SAMUEL P. HUNTINGTON 3> S ' > LIBRARY OK THE - C E I) CEiV'i HAl. EUROPEAN UNIVERSITY BUDAPEST UNIVERSITY OF OKLAHOMA PRESS : NORMAN AND LONDON CONTEXTUAL PROBLEMS, DISILLUSIONMENT, AND AUTHORITARIAN NOSTALGIA If new democratic regimes are to be consolidated, they have to deal in some way with transitional problems such as coping with the legacy of authoritarianism and establishing effective control of the military. More persistent challenges come from the contextual problems endemic to individual countries. In some countries these were neither numerous nor severe; in others they were both. A very rough listing of the major contextual problems confronting third wave democracies in the 1970s and 1980s and the countries in which those were most severe might go as follows: (1) major insurgencies: Philippines; El Salvador, Guatemala, Peru, 1 ( IS THE THIRD WAVE (2) ethnic/communal conflicts (apart from insurgencies): India, Nigeria, Pakistan, Romania, Sudan, Turkey; (3) extreme poverty (low per capita GNP): Bolivia, El Salvador, Guatemala, Honduras, India, Mongolia, Nigeria, Pakistan, Philippines, Sudan; (4) severe socio-economic inequality: Brazil, El Salvador, Guatemala, Honduras, India, Pakistan, Peru, Philippines; (5) chronic inflation: Argentina, Bolivia, Brazil, Nicaragua Peru; (6) substantial external debt: Argentina, Brazil, Hungary, Nigeria, Peru, Philippines, Poland, Uruguay; (7) terrorism (apart from insurgency): Spain, Turkey; (8) extensive state involvement in economy: Argentina, Brazil, Bulgaria, Czechoslovakia, East Germany, Hungary, India, Mongolia, Nicaragua, Peru, Philippines, Poland, Romania, Spain, Turkey. The eight problems listed above are a reasonable list of the major contextual problems that confronted new democracies of the third wave. The judgments as to the countries where these problems are severe argcasjjal andadhoc. If, however, the judgments have any validity at all, they suggest that these twenty-nine third wave countries could be grouped into three categories in terms of the number of severe contextual problems they face: (1) four or more major contextual problems: Brazil, India, Philippines, Peru; (2) two or three major contextual problems: Argentina, Bolivia, El Salvador, Guatemala, Honduras, Hungary, Mongolia, Nicaragua, Nigeria, Pakistan, Poland, Romania, Spain, Sudan] Turkey; (3) less than two major contextual problems: Bulgaria, Chile, Czechoslovakia, East Germany, Ecuador, Greece, Grenada, Korea, Portugal, Uruguay. Many have argued that the new democracies facing severe contextual problems have to cope successfully with those problems in order to develop the legitimacy essential to the consolidation of democracy. This general proposition has HOW LONG? been reinforced by arguments that failure to solve the country's most serious problem—be it debt, poverty, inflation, or insurgency—would mean the end of democracy in that country. If this is the case, the key question then becomes: Will the new third wave democracies confronted with severe contextual problems (which also bedeviled their authoritarian predecessors) successfully resolve those problems? In some cases new democratic regimes may deal successfully with individual problems. In the. overwhelming majority of cases, however, it seems highly likely that third wave democratic regimes will not handle these problems effectively and that they will, in all probability, be no more or less successful in doing j this than their authoritarian predecessors. Insurgencies, in-j flation, poverty, debt, inequality, and/or bloated bureaucra-1 cies will continue more or less as they have in previous de-* cades. Does this then mean an inevitably bleak future for third wave democracies? For some it may well mean that. Clearly democracy was under great stress in countries such as the Philippines, Peru, and Guatemala. The problems are numerous and severe; they will not go away, and they will not be solved. Other countries confront only slightly less challenging sets of contextual problems. Unresolved and seemingly unresolvable contextual problems reinforced tendencies toward disillusionment in the new democracies. In most countries the struggle to create democ-j racy was seen as moral, dangerous, and important. The coH lapse of authoritarianism generated enthusiasm and eupho-j ria. The political struggles in democracy, in contrast., rapidly^ came to be seen as amoral, routine, and petty. The workings.;' of democracy and the failure of new democratic governments^-to resolve the problems endemic to society generated indiffer-^ ence, frustration, and disillusionment. A short time after the inauguration of democratic government disappointment over its operation became widespread in Spam, Portugal, Argentina, Uruguay, Brazil, Peru, Turkey, 256 THE THIRD WAVE Pakistan, the Philippines, and most East European countries. This phenomenon first appeared in 1979 and 1980 in Spain, where it was labeled el desencanto (disillusionment), a term that then spread throughout Latin America. In 1984, ten years after the overthrow of the Portuguese dictatorship, the "excitement and creative enthusiasm that accompanied the transition to democracy" had disappeared and the "predominant political mood" was "one of apathy and disenchantment." By 1987, euphoria over democratization in Latin America had "given way across the restless continent to frustration and disappointment with the results so far." In 1989, it was reported that "a groundswell of public disillusionment with Brazil's political leadership and an explosive mood of social discontent have replaced the high hopes of 1985 when millions of Brazilians celebrated the restoration of democratic government after two decades of military rule." In Pakistan, less than a year after the transition, "a sense of impatience and sadness" had "replaced the euphoria that greeted the country's return to democracy." Within a year of the collapse of dictatorships in Eastern Europe, observers were speaking of the phenomenon of "post-totalitarian depression" and the mood of "disappointment and disillusion" that was sweeping the region.52 / Politically, the years after the first democratic government ( came to power were usually characterized by the fragmenta-) tion of the democratic coalition that had produced the tran-! sition, the decline in the effectiveness of the initial leaders / of the democratic governments, and the growing realization V that the advent of democracy would not, in itself, produce \ solutions to the major economic and social problems con-\ fronting the country. The intractability of problems, the con-i straints of the democratic process, the shortcomings of politi-\ cal leaders—these became the order of the day. The leaders of the new democracies often came to be viewed as arrogant, incompetent, or corrupt, or some combination of all three. A related response to democracy was "authoritarian nos- HOW LONG? 257 - talgia." This was not significant in countries where the authoritarian regimes had been extremely harsh, incompetent, or corrupt, or where they had been unwilling to give up power. It was more prevalent where the dictatorship had been mild, where there had been some economic success, and where the regimes were more or less voluntarily transformed by their leaders into democracies. In these countries, memories of repression faded and were in some measure replaced by images of order, prosperity, and economic growth during the authoritarian period. In Spain, for instance, the ratings of the Franco government in terms of general satisfaction, living standards, law and order, and social equity all increased between 1978 and 1984: "memories of Franco have become rosier, the farther the dictator fades into the past." This "absence-makes-the-heart-grow-fonder" effect also ap- \\ peared in Brazil. In 1989, reassessment of the rule of General Geisel was reported to be "in full swing. Today his rule is remembered fondly as a time when annual inflation was running well below 100 percent, instead of quadruple digits, and it was safe to walk the streets of Rio de Janeiro at night." In 1978, when asked which government or regime governed Portugal best, three times as many Portuguese citizens chose the Caetano dictatorship as chose the democratic regimes of Mario Soares. In 1987, seven years after the inauguration of democracy in Peru, residents of Lima picked Gen. Juan Ve-lasco, military dictator of Peru from 1968 to 1975, as the best president of the country since 1950. By 1990 the reputations of both General Zia and General Ayub Khan were on the rise in Pakistan.53 The intractability of problems and the disillusionment of I publics were pervasive characteristics of the new democra- / cies. They dramatically posed the issue of the survivability of ) the new regimes: Would they consolidate or collapse? The essence of democracy is the choosing of rulers in regular, fair, open, competitive elections in which the bulk of the population can vote. One criterion of the strength of democracy E3 It ■I 258 THE THIRD WAVE would be the extent to which political elites and publics firmly believe that rulers should be chosen this way, that is, an attitudinal test of the development of a democratic political culture in the country. A second criterion would r)e the extent to which political elites and publics do indeed choose leaders through elections, that is, a behavioral test of the institutionalization of democratic practices in the politics of the country. DEVELOPING A DEMOCRATIC POLITICAL CULTURE The democratic culture issue focuses attention on the relation between the performance or effectiveness of new democratic governments and their legitimacy—in other words, the extent to which elites and publics believe in the value of the democratic system. In an essentially pessimistic argument about this relationship, Diamond, Linz, and Lipset held that a primary reason for the instability of democratic and other regimes in the Third World, was "the combination and inter-S action of low legitimacy and low effectiveness." Regimes begin with low legitimacy and hence find it difficult to be effective, and regimes "which lack effectiveness, especially, in economic growth, tend to continue to be low in legitimacy."54 New democracies are, in effect, in a catch-22 situation: lacking legitimacy they cannot become effective; lacking effectiveness, they cannot develop legitimacy. To what extent is this pessimistic hypothesis justified? The inability of new democratic regimes to solve longstanding, severe contextual problems does not necessarily mean the collapse of those regimes. The legitimacy of authoritarian regimes (including, in the end, communist regimes) came to rest almost entirely on performance. The legitimacy of democratic regimes clearly rests in part on performance. It also rests, however, on processes and procedures. The legitimacy of particular rulers or governments may depend on what they can deliver; the legitimacy of the regime derives from the electoral processes by which the governments are constituted. Performance legitimacy plays a role in demo- HOW LONG? 2 5 9 cratic regimes, but it is nowhere near as important as the role it plays in authoritarian regimes and it is secondary to procedural legitimacy. What determines whether or not new democracies survive is not primarily the severity of the problems they face or their ability to solve those problems. It is* the way in which political leaders respond to their inability to | solve the problems confronting their country. Democratic regimes faced by extraordinarily severe contextual problems survived in the past. As Linz and Stepan have emphasized, the argument that economic crisis necessarily undermines democratic regimes is belied by the experience of the 1930s in Europe. Democratic systems survived the Great Depression in all countries except Germany and Austria, including countries that suffered much more economic hardship than those two did. They survived because, in the vyords of Ekkart Zimmerman, of "the ability of group leaders to come together, form new coalitions, sometimes on the basis of reaffirming older ones (such as Belgium), and then settle on how to steer the economy." Similarly, new democratic regimes in Colombia and Venezuela faced challenges in the 1960s fully as severe as those later confronted by third wave democracies. The lesson of their cases, as Robert Dix neatly summarized it, is that "political engineering can in substantial measure substitute for the dearth of more deterministic economic and sociological conditions of democracy in Third World nations."55 The stability of democratic regimes depends, first, on the ability of the principal political elites—party leaders, military / leaders, business leaders—to work together to deal with the } problems confronting their society and to refrain from ex- \ ploiting those problems for their own immediate material or -political advantage. New democratic regimes could not and did not rid their countries of long-standing terrorism and insurgencies. The crucial question for stability was how political elites and publics responded to this situation. In the 1960s, the elites in Colombia and Venezuela collaborated in attempt- z6o THE THIRD WAVE HOW LONG. 261 ing to cope with these problems. Similar developments occurred in the third wave democracies. Spain, for instance, confronted the continuing problem of extremist Basque terrorism. No national political party, however, attempted to exploit the issue in order "to delegitimize the democratic regime. ... no party persisted in blaming the various governments for creating the problem. No party claimed that the problem could be handled better outside of a democratic regime." In Peru, somewhat similarly, experience showed that "a guerilla movement can unite key political actors behind democracy as the only alternative to civil war."56 Second, the stability of democracy depends on the ability of publics to distinguish between the regime and the government or rulers. In 1983, for instance, twenty-five years after the inauguration of the second wave democratic regime in Venezuela, public opinion had become quite disillusioned with the performance of the elected rulers of Venezuela but not with the system of electing them. Despite, as one study reported, the "discontent with government there is nothing to indicate similar discontent with the method of selecting the government." While a substantial proportion (34.2 percent) of Venezuelans in 1983 believed the situation in their country justified a coup, only about 15 percent supported a specific alternative to a democratic regime. Fewer people than in 1973 believed that the government would be better without politicians and that politicians were indifferent to the country's problems. In 1983, "Venezuelans remained very supportive of the manner in which their governments come to office, increasingly dissatisfied with what they do once they get there, and convinced that the suffrage is the only way to improve things."57 Overall, despite the continued inability of elected governments to deal effectively with the problems confronting their country, Venezuelans were more strongly committed to democracy in 1983 than they had been in 1973. During the six years after 1983, Venezuela confronted in- tensifying economic problems stemming largely from declining oil prices. By 1989 the economic crisis had created a situation "where expectations have remained constant while the capability of the government to meet them has declined." Yet this did not pose a threat to democracy: The high level of frustration is not channeled into illegal, violent political activism, but rather in legal, peaceful system-maintaining mechanisms and processes. We find that the middle and lower class Venezuelans have turned to mainly four ways to cope psychologically with the crisis they are facing: legal protest, adaptation, resignation or emigration.50 The distinction between support for democracy and support for the governments that democratic elections produce was also manifest in Spain. Between 1978 and 1984, there was a "gradual dissociation of support for the democratic regime from satisfaction with what seems to be the mere effectiveness of democracy."59 In the last years of the Franco regime, unemployment was among the lowest in Europe (averaging 3 percent) and the economic growth rate was one of the highest in the world (averaging about 7 percent). In the first years of democracy in the late 1970s and early 1980s, unemployment rose to 20 percent and economic growth dropped to less than 2 percent. Confidence in the ability of democracy to resolve these problems varied widely. In 1978, 68 percent of the public thought democracy would allow resolution of the problems confronting the country. In 1980 and 1981, pluralities of the public thought democracy could not solve the country's problems. In late 1982 and in 1983, however, substantial majorities (55 percent and 60 percent) of the public once again had confidence in the ability of democracy to deal with Spain's problems. Yet despite these fluctuations in the public's confidence that democracy could solve their problems, support for democracy remained consistently high and even increased. In 1978, 77 percent of the Spanish public be- 262 THE THIRD WAVE HOW LON 2hy lieved that democracy was the best political system for Spain. That figure dipped to 69 percent in 1980, but rose to 81 percent in 1981 and to 85 percent in 1983.60 How can this consistent widespread support for democracy as a political system be reconciled with the variations in confidence in the ability of democratic governments to deal with problems? The answer, of course, is the electoral cycle. In 1978, voters still had confidence in the new Suarez government. By 1980 and 1981, with increasing economic hardship, they had lost confidence in that government, and in 1982 they swept Felipe Gonzalez and the Socialists into office. Having done that, their confidence in the ability of democracy to solve Spain's problems soared upward. Like Venezuelan voters, Spanish voters thus separated their support for democracy as a political system from their evaluation of the performance of the party in power. That distinction is crucial to the functioning of democracy. Under some circumstances, authoritarian nostalgia could conceivably pave the way for the "slow death" of a democratic regime, with the military or other authoritarian forces resuming power.61 Nostalgia, however, is a sentiment, not a movement. More generally, authoritarian nostalgia was further evidence of the tendency of publics to distinguish between rulers and regimes. The citizens of. Spain, Portugal, Brazil, and Peru simultaneously saw Franco, Caetano, Geisel, and Velasco as effective rulers and yet also overwhelmingly supported democracy as a better system of government. Disillusionment with democratic rulers and nostalgia for authoritarian ones were an essential first step in the process of democratic consolidation. They also were a sign that elites and publics were coming down from the euphoric and ephemeral "high" of democratization and were adapting to the grubby and plodding "low" of democracy. They were learning that democracy rests on the premise that governments will fail and that hence institutionalized ways have to exist for changing them. Democracy does not mean that problems will be solved; it does mean that rulers can be removed; and the essence of democratic behavior is doing the latter because it is impossible to do the former. Disillusionment and the lowered expectations it produces are the foundation of democratic stability. Democracies become consolidated when people! \ learn that democracy is a solution to the problem of tyranny,! { but not necessarily to anything else. A striking feature of the first fifteen years of the third wave was the virtual absence of major antidemocratic movements in the new democracies. Authoritarian holdover groups (both standpatter and extremist) existed in many countries. Authoritarian nostalgia materialized in several. Enthusiasm for democracy, participation in electoral politics, and the popularity of democratic leaders all declined significantly. Yet in the first fifteen years of the third wave, in no country did a large-scale mass political movement develop challenging the legitimacy of the new democratic regime and posing an explicit authoritarian alternative to that regime. In. at least those countries that had transited to democracy early in the third wave, the consensus on the desirability of democracy seemed to be overwhelming. In Spain, as was indicated above, in five polls between 1977 and 1983 substantial majorities of the public agreed that democracy was the best political system for a country like theirs. As one study concluded, "the bases of support for the democratic regime are much more variegated-broader and more ambiguous—than is the case for the exclusionary regime that preceded it. The democratic regime is less strictly tied to particular interests; in this respect it enjoys relative autonomy." Broad support for democracy was not limited to Spain. In Peru, for instance, in four polls between 1982 and 1988 the citizens of Lima endorsed democracy by-majorities of between 66 percent to 88 percent, and 75 percent of a 1988 nationwide sample chose democracy as the most desirable system for their country.62 More fragmentary evidence suggests similar levels of support for democracy in other third wave countries. 264 THE THIRD WAVE HOW LONG? 265 I This broad consensus on democracy in the third wave j countries immediately after regime change contrasts rather Imarkedly with the relatively slow development in Germany and Japan after World War II of support for both democracy and the values and attitudes associated with democracy. In the early 1950s, over a third of Germans indicated that they would support or be indifferent to an attempt by a new Nazi party to seize power and just under a third supported restoration of the monarchy. When asked to identify the period when , Germany had been best off, 45 percent chose the pre-1914 empire, 42 percent the Third Reich, 7 percent the Weimar Republic, and 2 percent the new Federal Republic. Support for the Federal Republic rose to 42 percent in 1959 and to 81 per-cent in 1970. In 1953, 50 percent of the German public thought democracy was the best form of government for Germany; by 1972, 90 percent did. The development of support for democracy and of the attitudes of trust and civic competence that go with democracy thus occurred slowly over the course of two decades.63 In still slower and less complete fashion a somewhat similar shift in opinion toward a more prodemocratic position occurred in Japan during the 1950s and 1960s. Why was there almost instant consensus on democracy after the end of dictatorships in Spain and Peru while it took two decades for comparable consensuses to develop after the collapse of authoritarianism in Germany and Japan? In Germany and Japan, in some degree people changed their opinions, but to a much larger degree the people changed. Younger and better educated people were more prodemocratic. Support for democracy in Germany approached unanimity when the German public had come to be composed of people who had been educated and spent their adult lives in the Federal Republic.64 In Spain and Peru, in contrast, widespread support for democracy shortly after the inauguration of democratic regimes either meant that such broad support existed under the authoritarian regime, or meant that people who had supported or at least acquiesced in authoritarianism before the transition changed their minds very quickly after the transition. Neither alternative is exactly a happy one for democracy. If the first alternative holds, authoritarian regimes existed in those societies evert when there was overwhelming support for democracy. If the second holds,, people who changed their opinions very quickly in a prodemocratic direction after the transition quite conceivably could shift equally quickly in an antidemocratic direction if circum-' stances warranted. In Germany and Japan broad support for democracy was the product of generational change, and hence ( was likely to be irreversible in the short run. In Spain and) Peru, broad support for democracy was, apparently, the re- ^ suit of opinion change, and hence could be more reversible in ' the short run. INSTITUTIONALIZING DEMOCRATIC POLITICAL BEHAVIOR The disillusionment that developed in the new democratic systems manifested itself behaviorally in four ways. First, it often led to resignation, cynicism, and withdrawaFfrom politics. In most new democracies, voting levels were high., during the transition but declined, sometimes quite drasti-^, cally, in subsequent elections. Decreased political participa-1; tion may have been undesirable in terms of democratic theory, \ but it did not, in itself, threaten the stability of the new'' democracies. Second, disillusionment manifested itself in an anti-incum-blervTteaction. As in Spain, voters could oust the ruling party and replace it in office by an . lternative group of rulers. This is, of course, the familiar democratic response, and it occurred frequently in the new third wave democracies. Incum- < bent leaders and parties were more often than not defeated \ when they attempted to win reelection. The parties that thus came to power in the first and second turnovers after the establishment of democracy usually pursued moderate policies well within the mainstream of opinion in "their"country. In 266 THE THIRD WAVE HOW LONG? 267 particular, parties identified with the Left—the socialists in Portugal and Spain, PASOK in Greece, the Peronists in Argentina—generally adopted highly conservative and orthodox economic and financial policies when in office (the major exception was Garcia's APRA government in Peru). iThird, disillusionment with democracy at times produced an antiestablishment response. In this case, voters not only rejected the incumbent party; they also rejected the principal alternative party or group within the political establishment and threw their support to a political outsider. This response was more frequent in presidential systems, where candidates for the top office run more on an individual than a party basis; hence it tended to be more prevalent in Latin America, where it was identified as populism. Notable examples of the populist antiestablishment response were the successful candidacies of Fernando Collor in Brazil and Alberto Fujimori in Peru. The candidacy of Carlos Menem in Argentina had some populist characteristics, although he was also the candidate of what was arguably the strongest established political party in the country. Successful populist candidates won office on the basis of "outsider" antiestablishment political appeals, with little or no backing from established political parties and with broad, multiclass support from the public. Once in office, however, successful populist candidates generally did not follow populist economic policies but instead launched rigorous austerity programs designed to cut government spending, promote competition, and hold down wages. Anti-incumbent and antiestablishment responses are the classic democratic reactions to policy failure and disillusionment. Through elections one set of rulers is removed from office and another is installed in office, leading to changes if I not improvements in government policy. Democracy is con-| solidated to the extent these in-system responses become * institutionalized. One criterion for measuring this consolidation is the two-turnover test. By this test, a democracy may be viewecTas consolidated if the party or group that takes power in the initial election at the time of transition loses a subsequent election and turns over power to those election winners, and if those election winners then peacefully turn over power to the winners of a later election. Selecting rulers through elections is the heart of democracy, and democracy is real only if rulers are willing to give up power as a result of elections. The first electoral turnover often has symbolic significance. The 19891, transition in Argentina was the first turnover since 1916 from \ an elected president of one party to an elected president from [ another party. The 1985 and 1990 Peruvian elections marked [ the second and third times in the twentieth century in Peru" that one elected president has transferred power to another. A second turnover shows two things. First, two major groups of political leaders in the society are sufficiently committed to democracy to surrender office and power after losing an election. Second, both elites and publics are operating within the democratic system; when things go wrong, you change the rulers, not the regime. Two turnovers is a tough test of democracy. The United States did nofcTearly meet iPurTHTTfieTacksonian Democrats surrendered office to the Whigs in 1840. Japan was universally and properly viewed as a democratic nation after World War II, but it did not meet this test and, indeed, effectively never has had even one electoral turnover. Between 1950 and 1990, Turkey had three military interventions and several first turnovers but never a second one. In three countries (Sudan, Nigeria, Pakistan) of twenty-nine that had transition elections between 1974 and 1990, the governments installed by those elections were removed by military or executive coups. In ten other countries with transition elections in 1986 or later, no other national election was held before the end of 1990. In fifteen of the sixteen remaining countries that held one or more elections after the transition election, a first turnover had occurred, the exception being Turkey. In six of eight countries that had two or more national 268 THE THIRD WAVE HOW LONG? 26g elections after the transition election, a second turnover occurred, the exceptions being Spain and Honduras. In twenty-two of the total of twenty-eight elections in the sixteen countries incumbent candidates or parties were defeated and the opposition came to power. The democratic process, in short, was operating: voters regularly ousted the ins and the ins always yielded office to the new choices of the voters. Apart from the three cases of democratic governments overturned by coups, in terms of institutionalizing the electoral process democracy in 1990 was alive and well in third wave countries. j['A fourth bnd most extreme political manifestation of dis-cohT:eTrt-wrjuld be a response, directed not at the groups in office or at the establishment generally, but at the democratic system itself. Concrete political forces opposed to democracy included both surviving standpatter groups from the authoritarian regime and continuing extremist groups from the anti-authoritarian regime opposition. Standpatter groups included in some cases elements among the military, but, as was pointed out above, these were normally discontented middle-ranking officers opposed by the military leadership and unable to mobilize significant support from civilian groups. In formerly communist countries, elements of the party and state bureaucracies, including the secret police, also fought rearguard actions against democratization. In Nicaragua, the standpatter Sandinista-controlled labor unions'overtly challenged the elected democratic government, threatening to "govern from below." Extremist opposition groups also attempted to challenge the new democratic regimes. By their very nature, however, radical groups employing violence, such as the Shining Path in Peru, the NPA in the Philippines, or the FMLN (Farabundo Marti National Liberation Front) in El Salvador, were not able to mobilize extensive support among the publics of the new democracies. Extremist groups that employed more peaceful tactics also had little success. In May 1990 ii _ x>rea, for instance, radical students organized demonstrations to commemorate the tenth anniversary of the Kwangju massacre. One demonstration involved almost 100,000 people, others numbered between 2,000 and 10,000. These were the largest demonstrations since those in 1987 that impelled the ruling party to agree to elections. The 1990 demonstrations against an elected government, however, did not draw the broad support that the 1987 ones did against the authoritarian regime. Only a "tiny fraction" of Korea's large student population joined the 1990 demonstrations and the middle class abstained because of its "broad lack of confidence in the opposition's ability to form an alternative government." "The middle class," it was reported, "prefers to grumble at home in front of television sets."65 In general, holdover standpatter and extremist groups tended to be isolated on the margins of politics in the new democracies in the 1970s and 1980s. The prevalence of democratic political practices in third wave democracies reflected the absence of authoritarian alternatives. Military juntas, personal dictators, and Marxist-Leninist parties had been tried and had failed. As a result, democracy was the only alternative. The crucial question, of crJufse7_w^~'wiietiiCT"rjT hoTlhis* would remain the case, or whether new movements would appear promoting new forms of authoritarianism. The extent to which such movements did materialize and develop significant support would presumably depend on the extent to which democratic behavior, including electoral turnovers, had become institutionalized. In addition, however, there was the possibility that over time within-system democratic alternatives would become exhausted. How many times would a public be willing to replace one party or coalition with another in the hopes that one of them would resolve the problems confronting the country? How often would voters be willing to elect charismatic, populist outsiders believing they would work economic and social q icles? At some point, publics could become disillusioned not only with the failures of democratic 270 THE THIRD WAVE cesses. They might be willing to shift from anti-incumbent and antiestablishment responses to antisystem responses. If democratic options appeared to be exhausted, ambitious political leaders would have powerful incentives to produce new authoritarian alternatives. CONDITIONS FAVORING CONSOLIDATION OF NEW DEMOCRACIES What conditions promote the consolidation of democratic political institutions and of a democratic political culture in third wave countries? As of 1990, the third wave was only fifteen years old, the returns on this issue were not in, and no definitive answer was possible. Two bodies of potentially relevant evidence, however, were available. First, the experience of the consolidation of first and second wave democracies could yield lessons for the third wave. Second, as has been pointed out, the factors promoting the inauguration of democratic regimes did not necessarily promote their consolidation. Yet some may do so. In addition, it may be possible reasonably to conclude that some developments will be more supportive of democratic consolidation than others. It would be folly to attempt to predict in which countries democracy will consolidate and in which countries it will not, and no attempt will be made here to make that prediction. It may, however, be useful, if speculative, to attempt to identify variables that could affect democratic consolidation and to identify to what extent these variables were present or absent in individual third wave countries. Consolidation success could be^mfluenced by several factors. (First^as was pointed out earlier, in the twentieth century veTyTew countries created stable democratic systems on their first try. It is reasonable to conclude that prior democratic experience is more conducive than none to the stabilization of third wave democracies. Extending this proposition it may also be reasonable to hypothesize that a longer and more recent experience with democracy is more conducive to demo- HOW LONG? 271 TABLE 5.1 Post-World War II Democratic Experience of Third Wave Countries_ Years of Democracy post_World War II and Pre-Third Wave Countries 20 or more 10-19 1-9 Less than 1 Uruguay*, Philippines, India, Turkey, Chile* Greece*, Ecuador, Peru, Bolivia, Korea, Pakistan, Brazil Argentina*, Honduras, Guatemala, Hungary*, Czechoslovakia*, Grenada, Nigeria Spain*, Portugal*, El Salvador, Poland*, East Germany*, Romania, Bulgaria, Nicaragua, Sudan, Mongolia * Countries with some democratic experience before World War II. cratic consolidation than is a shorter and more distant one. As the breakdown in Table 5.1 indicates, five countries— Uruguay, the Philippines, India, Chile, and Turkey—had twenty or more years of democratic experience after World War II before their third wave democratization, although for Turkey this was broken by brief military interventions in i960 and 1971. At the other extreme, ten countries had no democratic experience after World War II; and six—El Salvador, Nicaragua, Romania, Bulgaria, Mongolia, and the Sudan— had na^democratic experience at all before the third wave. , ■''Seconals was also emphasized in chapter 2, a high corre-lafurn;exists between level of economic development andTtne existence of democratic regimes.7^ more industrialized, modern economy and the more complex society and educated populace it entails are more conducive to the inauguration of 272 THE THIRD WAVE TABLE 5.2 Levels of Economic Development of Third Wave Countries 1987 GNP ~~ per Capita (in dollars) Countries 5,000 and more 2,000-4,999 1,000-1,999 500-999 Less than 500 Spain, East Germany, Czechoslovakia, Hungary, Bulgaria Greece, Portugal, Argentina, Uruguay, Brazil, Poland, Romania, Korea Peru, Ecuador, Turkey, Grenada, Chile Guatemala, El Salvador, Honduras, Nicaragua, Bolivia, Philippines India, Pakistan, Nigeria, Sudan Sources: Non-Eastern European countries: World Bank, World Development Report 1989 (New York: Oxford University Press, ^ 1989), pp. 164-65. Eastern European countries: Estimated from 3 Central Intelligence Agency, "Eastern Europe: Long Road Ahead ^ to Economic Well-Being" (Paper presented to the Subcommittee "iZ- on Technology and National Security, Joint Economic Commit- tee, United States Congress, May 16, 1990), pp. 1-6. Note: Mongolia is omitted due to lack of data. democratic regimes than are their opposites. It seems plausible to hypothesize that they will also be more conducive to the consolidation of new democratic regimes than will non-industrialized societies. If GNP per capita (as of 1987) is taken as a rough index of socioeconomic development, third wave countries fall into relatively clear categories (see Table 5.2). Spain (with a per capita GNP of $6,010), East Germany, and probably Hungary, Czechoslovakia, and Bulgaria were in the top group, followed by Greece (per capita GNP of $4,020). Several other countries also were above the $2,000 mark, including Portugal, Uruguay, Korea, Brazil, and probably the three other East European countries. At the bottom were the HOW LONG? 273- four third wave countries with per capita GNPs less than $500. As of late 1990, two of these (Nigeria and the Sudan) had reverted to military rule, and in a third, Pakistan, the democratically elected ruler had been summarily removed from office by the head of state, reportedly at the behest of the army. In 1990, consequently, India remained the only extremely poor third wave country where democracy remained clearly-intact. (_ Third, tj)e international environment and foreigri_actors play^ecTsignificant roFes in the creation of third wave democracies. Presumably an external environment supportive of democracy should also be conducive to its consolidation (see Table 5.3). An "external environment" here means for-. TABLE 5.3 External Environment and Democratic Consolidation in Third Wave Countries External Environment of Consolidation Countries Extremely favorable Quite favorable Favorable Indifferent/ unfavorable East Germany, Spain, Portugal, Greece Czechoslovakia, Hungary, Poland, Turkey, Philippines, Guatemala, El Salvador, Honduras, Nicaragua, Grenada, Bolivia Peru, Ecuador, Uruguay, Korea, Chile Argentina, Brazil, India, Nigeria, Sudan, Romania, Bulgaria, Mongolia Note: Classifications of the external environment are based on the author's impressionistic judgments. They rest on the assumption that both the European Community and the United States will continue to be concerned with the promotion of democracy. 1 274 THE THIRD WAVE eign governments and other actors that are democratic themselves, favor the existence of democratic regimes in other countries, have close relations with the newly democratic country, and are able to exercise influence in that country. Gerrrmny^_unificaLtiou made the future of democracy in what had been East Germany identical with that of the stable democratic environment of what had been West Germany. Membership in the European Community is extremely desirable for economic reasons, and democratic governance is a condition of membership; hence third wave EC members (Spain, Portugal, and Greece) have strong incentives to maintain their democratic institutions. Other countries, such as Turkey, Hungary, Czechoslovakia, and Poland aspire to membership, and that possibility provides an incentive for them to sustain their democracy. Some countries had extremely close relationships with the United States and have been heavily influenced by the UnifH^tatesTThese include the Central American countries, Grenada, Bolivia, and the Philippines. Countries where U.S. influence was present but probably less strong included Peru, Ecuador, Uruguay, Korea, Turkey, Poland, and Chile. The influence of major democratic powers was relatively weak in Argentina, Brazil, India, Nigeria, the Sudan, Romania, Bulgaria, and Mongolia. * r^TourtHTt^e timirg_oJLa_ra within the third wave may be indicative of factors that have an impact on the consolidation of democracy in that country (see Table 5.4). Countries that began a transition to democracy earlier in the *In an analysis of the reasons why the small Caribbean countries, mostly former British colonies, have sustained democracy, Jorge I. Dominguez emphasizes the role of the international subsystem and other Caribbean states, as well as the United States, in acting to defeat coups and other threats to democracy. The Caribbean international system has given priority to "democracy over non-intervention (the opposite of what has been the more common norm in Latin America)." "The Caribbean Question: Why Has Liberal Democracy (Surprisingly) Flourished? A Rapporteur's Report" (Unpublished paper. Harvard University, Center for International Affairs, January 1991), p. 31. HOW LONG? 275 TABLE 5.4 Inauguration of Democracy in Third Wave Countries Date of Founding Election Countries Before 1980 1980-83 1984-87 1988-90 Spain, Portugal, Greece, Ecuador, India, Nigeria Peru, Argentina, Bolivia, Honduras, Turkey Uruguay, Brazil, Philippines, El Salvador, Guatemala, Korea, Grenada, Sudan Pakistan, Poland, Hungary, East Germany, Czechoslovakia, Romania, Bulgaria, Nicaragua, Chile, Mongolia Possible after 1990 Mexico, Soviet Union, South Africa, Taiwan, Nepal, Panama wave did so largely as a result of indigenous causes. External influences and snowballing tended to be_more_significant as causes^of democratization for countries that made the transi-^"tlonlateranjhe_ waveLIt seems reasonablelrjTiypcHKesiZ'ethat the prevalence of indigenous causes, present largely in early-in-the-wave transitions, were likely to be more conducive to democratic consolidation than were external influences, more present in later-in-the-wave transitions. To the extent that this was a factor, it favored consolidation in the southern European countries, India, Ecuador, and Peru. It should have favored consolidation in Nigeria, but clearly did not prevent an early return to authoritarianism. Presumably the forces responsible for their later transitions should make consolidation more difficult in the Eastern European countries, Korea, Pakistan, and Nicaragua, as well as those countries (such as Taiwan, South Africa, the Soviet Union, and Mexico) that as of 1990 were still in the process of liberalization. 27° THE THIRD WAVE (^Tnfth^a crucial question obviously concerns the relation between transition processes and consolidation. Does it make a _ difference for consolidation whether or not a country transits.., to democracy through transformation, replacement, trans-"" placement, or intervention? Plausible arguments can be made for and against the helpfulness of each of these processes as far as consolidation is concerned. A related issue concerns the role of violence in the transition and presents similar problems. On the one hand, it can be argued that a peaceful, con-' sensual transition favors democratic consolidation. On the other hand, it could also be argued that a violent transition is likely to develop among most population groups a deep aversion to bloodshed and hence to generate a deeper commitment to democratic institutions and values. Overall, it seems more plausible to hypothesize that a consensual, less violent transition provides a better basis for consolidating democracy than do conflict and violence. If this is the case, negotiated transplacements may be most supportive of consolidation; transformations would be next; and replacements and interventions would provide the least support for consolidation (see Table 3.1 above). It might also be hypothesized that ; whatever the nature of the process, the less violence involved 1 in it, the more favorable are the conditions for democratic \ consolidation. To the extent this is the case, it could create problems for consolidation in El Salvador, Guatemala, Nica-ragua^renada, Panama, Romania, and South Africa. Sixth, jt was earlier argued that consolidation of democra-cieTwas not simply a function of the number and severity of the contextual problems they confronted. The heart of the 1 matter was instead how political elites and publics responded |.to those problems and to the inability of the new democratic governments to solve those problems. This was not to say, however, that the problems confronting a new democracy (were totally irrelevant to its consolidation. The number and / nature of severe contextual problems may be one variable HOW LONG? 277 along with the others that will affect democratic consolidation (see pp. 253-55 above). Other factors in addition to these six unquestionably affect the success or failure of consolidation. The extent and direction of those influences, however, is not always easy to estimate. One would suppose, for instance, that the nature and success of the authoritarian regime might affect the consolidation prospects of its democratic successor. Is the prospect of democratic consolidation affected by whether the authoritarian regime was a military regime, a one-party system, a personal dictatorship, or a racial oligarchy? A variety of conflicting hypotheses and arguments are possible, including the argument that the nature of the predecessor authoritarian system does not have any significant implications for the consolidation of its democratic successor. Similarly, is democratic consolidation more likely in the wake of what might.be termed relatively successful authoritarian regimes (e.g., Spain, Brazil, Taiwan, Korea, Chile) or relatively unsuccessful ones (e.g., Argentina, Philippines, Portugal, Bolivia, Romania). This distinction is obviously related to differences in transition processes, but it could also be an independent variable on its own. But in which direction? It could be argued that the reactions of elites and publics to the manifest failures of unsuccessful authoritarian regimes should be a positive force for democratic consolidation. It could also, however, be argued / that nations may differ in their political capacities and that a ) people who made a success of authoritarianism (e.g., the | Spanish) will do the same with democracy, while a people \ who were unable to create a successful authoritarian system (e.g., the Argentines) are likely to have no more success in consolidating a democratic one. Democratic consolidation may also be affected by the nature of the democratic institutions jthat_are established^lau-/' sible arguments, for instance, have been made lihat a parliar/ mentary system is more likely than a presidential system tof 278 THE THIRD WAVE HOW LONG? 279 contribute to the success of new democracies because it reduces the "all or nothing" aspect of politics, usually requires a coalition of parties to form a government, and provides an opportunity for balance between a chief of state and a chief of government.66 These arguments are suggestive, and the desirability of shifting to a parliamentary system has been raised by several Latin Americans, including Raul Alfonsin. Evidence that parliamentary regimes contribute to democratic consolidation, however, is still scanty. A similar issue comes up with respect to the nature of the party systems in new democracies. Is democracy better served By "many parties each representing a particular economic, social, regional, communal, or ideological interest? Or is it better served by two comprehensive parties, each of which will provide a plausible and responsible alternative government to the other and whose leaderships could more easily cooperate in dealing with severe economic crises, drug mafias, and threatening insurgencies? Again the evidence is lacking for a judgment one way or another. If the factors discussed above are relevant to the consolidation of new democracies and if one makes the dubious assumption that they are equally relevant, some broad judgments emerge as to where the conditions were most favorable and least favorable for consolidation. The conclusions are not surprising. Overall, the conditions for consolidation were mosttavgr