Transitions to Democracy: Chapter 6

Lisa Anderson (editor)

1999

6. Myths of Moderation: Confrontation and Conflict During Democratic Transitions

In a 1970 essay that marked the beginning of a new wave of writing on democratization, Dankwart A. Rustow argued persuasively that democracy is the fruit of “choice” and “conscious decision” on the part of political elites. ₁ This perception seems widely shared, for our current literature is full of tactical insights on how elites might be induced to choose democracy over alternative political systems. This essay focuses on the tactics of the transition period, defined as the time between the breakdown of the dictatorship and the conclusion of the first democratic national elections. It examines what I call the “moderation argument”: that radical popular organizations threaten democratic transitions if they fail to moderate their demands and behavior as the moment of elite choice approaches. Implicit and explicit variations of the moderation argument are widely purveyed, and this essay assesses their merits in light of evidence from Iberia, Latin America, and Asia. It concludes by explaining when the moderation argument does and does not hold.

Popular Mobilization in the Literature

The current literature on democratization accords much less attention to popular organizations than to political elites. Thus, the role of popular organizations in the transition process remains a subject of some confusion. Many of the major theoretical works on democratization suggest that popular mobilization is important for regime change, but even this very simple proposition is not universally shared. Some scholars argue that a “popular upsurge during the transition is by no means a constant,” ₂ while others argue that “social mobilization is undoubtedly indispensable.” ₃ Although most of the generalizations regarding the role of mass publics in the transition to democracy are neither widely shared nor very clearly specified, a few related propositions nevertheless emerge.

One proposition argues that too much popular mobilization and too much pressure from below can spoil the chances for democracy, what I have called the moderation argument. Though the conclusion that popular participation sometimes harms rather than enhances democratization is politically unpalatable, it is nevertheless widely drawn. For example, Terry Karl writes that “no stable political democracy has resulted from regime transitions in which mass actors have gained control even momentarily over traditional ruling classes.” ₄ Myron Weiner writes that pressures from left-wing parties are often least effective for democratization because they frequently provoke “only... increased authoritarianism,” ₅ and Samuel Huntington argues that “democratic regimes that last have seldom, if ever, been instituted by mass popular action.” ₆ Daniel Levine concludes that the “obvious” lesson from experience is that “conservative transitions are more durable.” ₇ Even Rueschemeyer, Stephens, and Stephens, who argue that democracy is the outcome of working-class mobilization, assert that radical mass parties are more likely to “evoke strong defensive, anti-democratic reactions.” ₈

Fear of the masses lies at the root of this cautionary argument. For some, this fear emerges from the conviction that the general citizenry may not have the values a sustainable democracy requires. For example, Juan Linz and Larry Diamond argue that in many Latin American states “the choice of democracy by political elites clearly preceded... the presence of democratic values among the general public.” ₉ In other cases the argument develops from an awareness that pivotal political groups harbor a fear of the masses themselves and often act upon it. Robert Kaufman argues that “the threat from below... is the lowest common denominator and the most important bond of cohesion within bureaucratic-authoritarian coalitions.” ₁₀ Kaufman’s observation, like the others cited above, has clear tactical implications: if a transition is to be carried out successfully, the “threat from below” must somehow be moderated.

Of course, what constitutes “the threat from below” varies from one regime to another, just as “extremism” lies in the eyes of the beholder. In some cases, the most threatening pressures will come from students and other relatively well-educated urban groups. In other cases, the primary threat will come from armed opposition with a rural base, subnationalist groups, or organized labor. I use the term “popular organizations” to denote a varied set of nonelite but formally organized actors who are perceived as extremists by existing elites. Who these actors are sociologically needs to be contextually defined.

Discussions that focus on labor organizations are especially explicit on the need for moderation. Adam Przeworski argues that “complete docility and patience on the part of organized workers are needed for a democratic transformation to succeed” and reminds us that the “democratic system was solidified in Belgium, Sweden, France, and Great Britain only after organized workers were badly defeated in mass strikes and became docile as a result.” ₁₁ Samuel Valenzuela makes a similar point. He writes that labor mobilization may “act like a double edged sword and permit a reversal of the process of redemocratization” and cautions that waves of “strikes and demonstrations may lead to a protracted crisis [and] may lead employers to reconsider their commitment... to supporting the process of democratic change.” ₁₂

Valenzuela speaks directly to the question of tactics. He argues that the “ideal mix” for democratization is “high labor mobilization at certain critical moments of the breakdown of authoritarian institutions,” followed by “restraint when the political agenda shifts in favor of redemocratization.” ₁₃ This and the other readings of democratization suggest that citizen mobilization is effective, even essential, but that it is ultimately dangerous if it continues too long or with too much intensity. ₁₄

But how much is too much? Guillermo O’Donnell and Philippe Schmitter offer us some leads in their now classic study of uncertain transitions. They argue that increased activity is useful up to a point but that it will provoke a dictatorial reaction “if widespread violence recurs” or if it threatens “the vertical command structure of the armed forces, the territorial integrity of the nation-state, the country’s position in international alliances, [or]... the property rights underlying the capitalist economy.” If any of these extremes are reached, they write, “then even bland regime actors will conclude that the costs of toleration are greater than those of repression” and the movement for democracy will fail. ₁₅

This cautionary message corresponds with Robert Dahl’s argument in Polyarchy, and it is intrinsically appealing. ₁₆ It makes sense to distinguish between the amount of opposition activity and its content. It also makes sense to argue that the likelihood of democracy decreases as the costs of toleration rise. But how are the costs of toleration calculated? To answer this question we must think systematically about popular mobilization and elite response in concrete cases of successful transition.

The Realities of Transition Tactics in Portugal and Spain

Portugal and Spain provide good places to begin our inquiry. They were among the first transitions of the “third wave” and thus have had a special impact on our thinking about transitions in general. They are also useful because they are alleged to represent two radically different routes to democratic transition. Any commonalities we find in these contrasting cases are likely to be shared by a broad spectrum of others.

Portugal

The Portuguese transition, beginning in April 1974, violated most of the cautionary parameters set out by the literature on democratization. The laboring classes were far from docile. Capitalist property rights were challenged successfully on a very broad scale. The country’s position in international alliances was the subject of strenuous debate, and decolonization shattered the territorial integrity of the state. The vertical command structure of the armed forces was completely transformed. Nevertheless, democracy muddled through. Individual actors might have concluded that the costs of toleration were greater than those of repression, but no one succeeded in reversing the transition.

Because the details of the Portuguese transition are not well known, these points require some elaboration. The Portuguese working class took the world by surprise by leading the most massive seizures of property in Europe since the Russian Revolution. Workers occupied more than 23 percent of the nation’s farmland in less than twelve months and took control of more than 940 industrial enterprises. ₁₇ Two thousand houses were seized in the two weeks following the fall of the dictatorship, and in February 1975 2,500 apartments were occupied in Lisbon alone. ₁₈

Provisional governments legitimated property seizures through a variety of legal supports and eventually nationalized banks, insurance firms, and all basic industries. ₁₉ The constitution, approved by the freely elected constitutional assembly of April 1975, promised a “classless society” and the “transformation of capitalist relations of production and accumulation.” ₂₀

Just as property relations were being radically restructured, so were the territorial boundaries of the Portuguese state. The Salazar–Caetano dictatorship had gone to great lengths to convince mainland Portuguese that Portugal’s colonies were an integral part of the Portuguese nation. (The colonies were officially called “the overseas provinces.”) Yet the men who controlled the provisional governments after April 1974 granted independence to Guinea, Mozambique, Cape Verde, and Sâo Tomé–Príncipe within a year of taking power and withdrew from Angola in November 1975. Millions of dollars of property were lost, and 800,000 refugees flooded onto the Portuguese mainland. ₂₁

The vertical command structure of the armed forces was radically transformed as part of the process of ending the colonial wars. The middle-ranking officers who toppled the dictatorship engineered the most extensive purges of any democratic state in the third wave, including those of eastern Europe. The first set of purges affected the ultras, who supported the old regime, while the second set embraced top-ranking centrists. ₂₂ Even General António de Spinola, the head of the first provisional government, was forced from office in September 1974. A dramatic breakdown in discipline occurred in the lower ranks when radicalized enlisted men organized a movement for the democratization of the armed forces in September 1975. ₂₃ The intense politicization of the military caused the rise and fall of six provisional governments. ₂₄

Although official pronouncements emphasized that Portugal would not change its status as a loyal member of NATO, western powers were extremely concerned about Portugal’s loyalty to its old allies. ₂₅ The chaos within the Portuguese military, the strong Communist presence in the cabinets of Portugal’s provisional governments, and the occasional popular demonstrations against NATO led Henry Kissinger to advocate covert action against the Armed Forces Movement. How much covert action took place is still unclear, but we do know that other western powers were also fearful of Portugal’s loyalty and that the nation was expelled from NATO’s nuclear planning group. ₂₆

This brief review of the Portuguese case illustrates that democratic transitions can, indeed, survive radical pressures from below and that they can even survive radical provisional governments. But is Portugal simply an exception to the rule? A close look at other successful transitions suggests that the radical elements of civil society have a much more boisterous and enduring presence in the drama of democratization than we originally believed.

Spain

Spain is particularly instructive in understanding the limits of the moderation argument because it is so often held up as the model of peaceful and “pacted” transition. Elite pacts were certainly key to the democratization of Spain, but these pacts were forged in a situation in which extremism and moderation existed simultaneously. The law that guaranteed free elections was passed by the Francoist Cortes in November 1976 and approved by a popular referendum in December of the same year. However, 1976 was also a year of widespread violence and unceasing mobilization. Scholars have placed so much emphasis on the comparatively peaceful nature of Spain’s transition that it is easy to forget its violent elements.

In fact, violence was much more pervasive in Spain than in “revolutionary” Portugal. ETA (Euskadi Ta Askatasuna), the armed wing of the Basque separatist movement, provided a source of “widespread violence” throughout the transition period and threatened “the territorial integrity of the nation state.” Yet the Cortes passed the Law of Political Reform just eighteen months after the declaration of a state of emergency in the Basque country and just thirteen months after five separatists were publicly executed following a season of terrorist activity and bombings. The majority of Basques did not support ETA violence but were radically opposed to the central authorities’ treatment of separatists. As the year of democratization began, some two-thirds of the Basque population struck in protest against police repression, and on March 3 five workers were killed when police shot into a crowd of Basque nationalist protesters in Vitoria. ₂₇ In early October, barely a month before the approval of the Law of Political Reform, ETA murdered the head of the provincial council. On December 11, 1976, GRAPO, another revolutionary organization, kidnapped the president of the council of state and the resident of the superior council of military justice. ₂₈

Violence did not decrease with the successful referendum on political reform in December 1976. In what became known as the Black Week, in January 1977, two students, five labor lawyers, and five police were gunned down in separate incidents on the streets of the capital. ₂₉ Between the passage of the reform laws and the nation’s first free elections, Spain experienced “a full-scale destabilization attempt from terrorists” of both the left and the right. ₃₀ Although the attempt was unsuccessful, it was unleashed in full force and is of great importance for our conclusions about the necessity of moderation. Statistics show that violence did not decrease as the transition progressed. Kidnapping rose steadily through 1979, as did the number of people wounded in political violence. Political killings rose steadily through 1980. ₃₁

The Spanish working class never challenged property relations as their counterparts did in Portugal, but they exceeded their Portuguese neighbors in their level of strike mobilization. In 1976 alone, over 3.6 million Spanish workers participated in strikes affecting all major industrial centers and public services in Madrid and Barcelona. Worker mobilization continued well into the following year. ₃₂

The existence of what reactionary forces might have deemed extremism was not confined to trade union and separatist circles. Spain’s Socialist party (PSOE) emerged from the dictatorship with a vocal radical faction. Although the party cultivated a moderate image in some settings to attract middle-class support, it also emitted strong signals which reactionaries could read as threatening to both “alliances” and “property rights.” The party’s position on international alliances was one of “active neutrality.” It sought to dissociate Spain from its bilateral ties with the United States and from its former role in the Atlantic alliance. ₃₃ Political actors who feared that democratization would threaten capitalist property relations could easily find justification in party documents. According to official party principles, formulated in December 1976, the Socialists aspired to “the possession of political power for the working class [and] the transformation of individual or corporate ownership of the instruments of labor into collective, social or common property.” ₃₄ As time passed, the radical elements in the PSOE were either transformed or overpowered, but at the time that the regime disassembled itself, there was no guarantee that this change would take place. Thus, even in Spain, the model of the controlled transition, “extremist” forces were far from absent.

These two very different transitions took place in the context of radical popular mobilization and perceived extremism, but it would be a mistake to dismiss the moderation hypothesis altogether. There are clearly cases in which extremism and mass mobilization provoked the installation of a more coercive dictatorial elite. The reversals of the liberalization in Chile in 1983 and of the milder liberalization in Argentina in 1981 illustrate that right-wing reactions are not mythical. ₃₅ Yet these negative cases do not in themselves make an argument for moderation. Such an argument can be supported only by assembling a large number of cases in which democratic regimes actually replaced dictatorships and then showing that in each case popular mobilization rose, crested, and became moderate before the successful transition was completed. Portugal and Spain clearly did not follow this pattern. Is the moderation hypothesis better sustained by cases that came later in the third wave?

Comparative Perspectives

Cross-national time series data for a phenomena as varied as “popular mobilization” are not available, but we can learn something from analyzing strike data, because organized labor is often thought to be a prime source of “immoderate” demands and destabilization.

Table 6.1 provides some suggestive evidence on the number of strikes in six states in which dictatorships gave way to formal democracies between 1977 and 1989. ₃₆ These states vary in level of development, regime longevity, and political culture. Yet worker mobilization decreased substantially prior to democratization in only one case. Aside from Ecuador, where a left-wing military regime was ousted, worker mobilization either rose or was extremely high when the first democratic elections were held. If we consider the number of workers involved in these strikes, the pattern remains the same (see table 6.2).


Table 6.1: Annual Strike Statistics: Number of Strikes/Year ^a

Year^b	Brazil	Chile	Ecuador	South Korea	Peru	Philippines

–5	534	38	NA	206	788	260
–4	843	42	61	186	570	158
–3	1,494	41	61	88	779	155
–2	2,369	81	58	98	440	282
–1	1,954	72	9	114	234	371
First Free Elections	4,189	101	7	265	364	581

^aExcludes work stoppages involving less than 10 work days. ^bYears correspond to the following calendar years: Brazil 1984–1989, first elections November 15, 1989; Chile 1984–1989, first elections December 14, 1989; Ecuador 1973–1978, first elections July 16, 1978; South Korea 1980–1985, first elections February 1985; Peru 1973–1978, first elections June 22, 1978; Philippines 1981–1986, first elections February 7, 1986; Spain 1972–1977, first elections June 15, 1977. Source: International Labour Office, “Yearbook of Labour Statistics.”


Table 6.2: Annual Strike Statistics: Workers Involved

Year^a	Brazil	Chile	Ecuador	South Korea	Peru	Philippines

–5	NA	3,595	NA	48,970	416,250	98,585
–4	5,431,000	8,532	5,948	34,586	362,740	53,824
–3	7,146,000	3,900	11,913	8,967	617,120	33,638
–2	8,304,000	9,900	7,016	11,100	258,100	65,306
–1	7,136,000	5,600	802	16,400	1,315,400	111,260
First Free Elections	14,099,000	17,900	538	28,700	1,398,400	169,480

^aYears correspond to those in Table 6.1. ^bYears correspond to the following calendar years: Brazil 1984–1989, first elections November 15, 1989; Chile 1984–1989, first elections December 14, 1989; Ecuador 1973–1978, first elections July 16, 1978; South Korea 1980–1985, first elections February 1985; Peru 1973–1978, first elections June 22, 1978; Philippines 1981–1986, first elections February 7, 1986; Spain 1972–1977, first elections June 15, 1977. Source: International Labour Office, “Yearbook of Labour Statistics.”

This evidence does not show that workers become more moderate as successful transitions are being negotiated. It suggests, instead, that a successful transition to democracy does not require moderation on the part of the working class. If we look carefully at the details of the cases, we see that there is much more leeway for “extremist” pressures than the literature on democratization leads us to believe.

Peru

The transition from dictatorship in Peru in 1977–78 provides compelling evidence that formal democracies can be constructed amid violence and high levels of popular mobilization. Peruvian democracy was destroyed by a presidential coup in 1992 and thus was not consolidated, but the transition which began in the late 1970s produced a regime which held together for approximately fourteen years. Peru’s transition to democracy began on July 28, 1977, when Francisco Morales Bermúdez announced that the armed forces would hold free elections for a civilian constitutional assembly the following year. The announcement came less than ten days after “the entire nation was shut down by... the most massive [general strike] in the country’s history.” ₃₇

The promise of free elections did not curtail the growing wave of popular mobilization. On the contrary, in May 1978, less than one month before the scheduled elections, the military government declared a state of emergency after strike waves and bombings left twenty people dead and fifty injured in a single week.

The elections nevertheless took place on June 18, 1978, and showed a dramatic shift to the Marxist left. Marxist parties garnered 29 percent of the vote, far exceeding the 5 percent they had won in the past (and in spite of the fact that the government had arrested and deported many leftist leaders, shut down many politically oriented publications, and decreed that illiterates would not be allowed to vote).

The convening of the constitutional assembly also did not curtail the growing wave of popular mobilization. Teachers’, health workers’, and debilitating miners’ strikes were timed deliberately to coincide with assembly debates. Ten provinces were placed under a state of emergency in August 1978 as the assembly was just beginning. In October the nation’s cities were rocked by more rioting, but the assembly continued to meet.

Though Sendero Luminoso had no more than three hundred members when the military began to withdraw from power, it had already begun to shoot officials who tried to enter what was (even then) known as “Senderista territory.” ₃₈ In 1980, the year of the nation’s first postdictatorial presidential election, Sendero launched 505 terrorist attacks. ₃₉ Sendero Luminoso apparently did not play a central role in elite calculations at the time, ₄₀ but the fact that the group was not taken seriously lends support to the argument that the mere existence of radical forces does not threaten the transition from authoritarianism to democracy.

The Philippines

The transition from dictatorship in the Philippines shows how an electoral democracy can be constructed in a state with an even larger guerrilla presence. The “People’s Power” movement that toppled the Marcos dictatorship was a coalition of moderate forces drawn from all classes of Philippine society. But the coalition was forged at a time when a decidedly nonmoderate group, the New People’s Army (NPA), was expanding rapidly. The NPA was founded in 1968 as the armed wing of the Communist Party of the Philippines. From its inception, its ultimate goal has been the overthrow of the Philippine capitalist state. Its method, like that of Sendero Luminoso, is the armed mobilization of the rural poor.

The NPA did not waver in its commitment to armed struggle when the possibility of a peaceful ouster of Marcos presented itself. The Communist Party of the Philippines boycotted the February 1986 presidential elections, declaring that dictators could not be ousted through the polls and that in any case Corazon Aquino was not significantly different from Marcos. ₄₁ The NPA had proved remarkably successful outside of any coalition of moderate groups and had few organizational incentives to change course. It grew tenfold after martial law was declared in 1972 and won the loyalty of 23,000 rebels by 1986. ₄₂ Its fifty-odd guerrilla units penetrated 32 percent of the nation’s barangays. ₄₃ In 1985, just as the moderate movement for democracy was taking shape, the NPA launched 5,000 violent attacks. ₄₄

The NPA was clearly an “extremist” force when the Marcos regime entered its final crisis. It was also a relatively powerful force: U.S. intelligence experts believed that it could actually take armed control of the Philippine state by 1990. ₄₅ Yet moderate, prodemocratic political actors succeeded in forging an electoral democracy. The existence of uncompromising forces in the opposition did not lead to the triumph of reaction.

South Korea

The establishment of electoral democracy in South Korea offers a final illustration of successful democratization despite high levels of popular mobilization and widespread violence.

South Korea’s construction of electoral democracy began on June 29, 1987, when Roh Tae Woo, the designated successor to Chun Doo Hwan, announced that he would accept the demands of the political opposition and allow direct popular elections for the presidency. Roh’s action did not result from compromise and bargaining among political elites. On the contrary, it came as a complete surprise to the opposition and followed three weeks of serious rioting in which tens of thousands of students and middle class citizens “effectively broke government control” of important commercial areas. ₄₆

The opposition to the South Korean dictatorship was a highly heterogeneous group with important moderate elements, but opposition extremists remained unchecked throughout the months between Roh’s announcement and the holding of elections. Within the universities, anti-American groups seeking to alter the nation’s “position in international alliances” dominated the student movement. Underground study groups called chiha served as bases for vanguard demonstrations of 200 to 300 students that often involved the hurling of gasoline bombs.

Violence did not decrease with the approach of elections. As the draft of the new national constitution was being debated in August 1987, the dictatorial regime faced the most serious strikes in its history, along with massive rioting linked to the killing of a worker by a tear gas canister. ₄₇ The presidential campaigns were marked by violence as well. In Kunsan, hundreds of rioters attacked Roh’s campaign motorcade with bricks, bottles, and homemade bombs. His planned visit to Chonju was canceled because of further rioting and fears for the president’s life. Most significant, perhaps, the extremists did not confine their attacks to the dictator himself. ₄₈ Moderate opposition candidates were also physically attacked on several occasions by leftist groups, yet the transition to electoral democracy continued. ₄₉

Rethinking Moderation

The examples reviewed above provide compelling evidence that high levels of popular mobilization do not inevitably sidetrack the transition from authoritarian to democratic rule. The more recent transition in South Africa suggests that we might find supportive evidence in other regions as well. ₅₀ Moderation is not a prerequisite for the construction of democracy; the parameters of tolerable mobilization are broader than we originally anticipated. In many cases, democratization seems to have proceeded alongside weighty and even bloody popular challenges.

Why do members of the old dictatorial coalition give in to demands for democracy when immoderate forces elevate its risks? In the Portuguese case, we might argue that the old dictatorial coalition had no choice. But this argument would not apply to nonrevolutionary transitions, and in Portugal, as elsewhere, a new dictatorial coalition might have emerged during the transition period. Why do actors who might be harmed by extremism risk an expansion of liberties that may work to their disadvantage? The key to the puzzle lies in how risks are calculated. These actors do not calculate their risks in the manner we initially surmised.

Robert Dahl’s seminal work, Polyarchy, has encouraged us to think that the likelihood of democratization increases as the cost of suppression rises and the cost of toleration declines. None of the empirical materials presented here contradict these general assumptions. Yet the comparative study of various transitions suggests that we should think more systematically about how elites calculate these costs.

The axiom related to the costs of suppression helps explain why we have found so many regimes that democratized despite extremism. The cost of suppression increases as the level of popular mobilization rises and the number of attacks on the dictatorship increases. Rustow argued that democratization requires “a hot family feud” and “a prolonged and inconclusive struggle.” ₅₁ If antiregime forces of any sort succeed in elevating the sense of struggle and raising the costs of suppression, they affect half the decision calculus and win half the battle.

Yet the battle can easily be lost if the costs of toleration remain too high. For the costs of toleration to seem bearable, pivotal elites must believe that they will not be ruined by reform. Dictatorial regimes will pay extremely high suppression costs if the costs of toleration leave them no other choice. The logic of this argument explains why calls for moderation emerge so frequently in the democratization literature.

However, we are wrong to think that the costs of toleration are a simple function of the presence or absence of extremism. Neither the presence nor the scope of extremist activities is as important an element in elite calculations as their estimates of what the effects of extremism will be. We can resolve the seemingly dramatic differences between the cautionary arguments in the literature and the many cases reviewed here if we recognize that elite projections about the effects of extremism can take at least three forms.

In a first, pivotal elites, defined as those who have the capacity to reverse the transition, project that extremism will have powerful and wholly pernicious effects. They predict that democratization will lead to the triumph of extremist groups and to the disastrous end of the pivotal elites themselves. If this belief is widely shared, the likelihood of an antidemocratic reaction is extremely high. Democracy is therefore seen as intolerable and is rejected. The moderation argument cautions us against this scenario. As the 1989 mobilization in China illustrates, it is far from mythical.

But let us suppose that pivotal elites forecast that extremism will have other effects. They are convinced that extremists will not win the democratic game. Under these conditions, democracy may prove acceptable, and two more desirable scenarios emerge.

In the first, pivotal elites opt for democratization because they have been unable to control extremism themselves and are no longer willing to pay the high institutional costs of failing to provide political order. They forecast that democratic elections will be won by nonextremists and that ceding control to moderate actors in an electoral democracy is less risky than continuing with the status quo. Peru’s transition from authoritarianism in 1977–78 provides an example of this scenario. In it, democracy is seen as a means of escape.

In the second, pivotal elites opt for democracy despite extremism because they predict that they will win the transition elections themselves. In this case, they see democracy as a formula for their own legitimation. This scenario has two variants. In the renovating variant, the pivotal elites are associated with the old dictatorship. They accede to electoral democracy because they believe it will provide them with a new formula for legitimating their rule. In the revolutionary variant, democracy becomes the legitimating formula for a new elite. Revolution has forced the old guard into political inactivity, and the pivotal elites are the military officers and civilian politicians who oppose both the threatening extremists and the previous dictatorship. Portugal provides a classic example of this revolutionary scenario. Spain provides an example of the renovating variant. ₅₂

The existence of extremist groups is not an insurmountable obstacle for democratic forces in either the escape or the legitimation scenarios. Extremism always raises the costs of governance. When the known costs of governance rise, the projected costs of democracy will seem relatively low if pivotal elites predict that nonextremist forces will take control of the new democratic regime. If pivotal elites can exclude certain groups from electoral competition, if they can shape electoral laws for their own purposes, and if radical forces sometimes refuse for their own reasons to participate in elections, these predictions may not be at all unreasonable. Table 6.3 provides a summary of the projected effects of extremism and the different scenarios these projections are likely to produce.


Table 6.3 Elite forecasts of the Effects of Extremism and the Resulting Scenarios for Transitions to Democracy

Scenario	If pivotal elites forecast...	they...	Because they see democracy as...	As happened in...

I	Extremist victory	will reject democracy	an intolerable threat	China, 1989
II	Extremist defeat and moderate victory	may accept democracy	a means of escape	Peru, 1977 Greece, 1975
III	Extremist defeat and their own victory	may accept democracy	a form of legitimation	Portugal, 1974 Spain, 1976

Real life cases may not fit squarely in a single scenario because the pivotal elites in many transitions are heterogeneous. There may be differences in how sectors of the pivotal elite view the effects of extremism and thus in how they calculate their own particular risks. Nevertheless, if we use these categories as a guide in thinking about elite calculations, we can understand better why particular transitions turned out as they did.

Portugal is clearly a case in which a new political elite used democracy as a legitimating formula for a new regime. The pivotal elites in Portugal were the moderate military officers, who constituted a majority in the military establishment, and the democratic but not radical civilian politicians who headed political parties and occupied various cabinet positions in the provisional governments. After a period of disorientation provoked by the surprise of the revolution itself, both groups became convinced that democratic processes would work against the extremist forces of the PCP and the ultra-left. Socialist Party leader Mario Soares was certain that extremists were “not even close to having the popular support” that they required to seize control of the transition. He was also convinced that “the armed forces—even the [radical] armed forces in Lisbon—would never agree to such a suicidal act.” ₅₃ He publicly stated that there was “very little possibility” of a Communist coup if for no other reason than that the Communists knew it would be immediately reversed. ₅₄ Despite the highest levels of radical mobilization in Portuguese history, a broad spectrum of nonradical elites consistently advocated continued democratization. Elections were consistently purveyed as a solution to disorder because pivotal elites were convinced that they would gain legitimized control of the state if democracy proceeded. ₅₅

Spain provides us with an example of the legitimation scenario in its renovating rather than revolutionary variant. The most visible pivotal elites in the Spanish transition were King Juan Carlos and Adolfo Suárez. They were directly associated with the old regime and clearly needed a new formula for the legitimation of a renovated elite. Suárez explicitly referred to this legitimation function when he presented the Law for Political Reform in September 1976: free elections were the means through which “respectable political groups who had no popular mandate could become representative of the public.” ₅₆

Suárez and Juan Carlos were able to take the risks of democratization despite extremism because they forecast that continued democratization would not lead to the triumph of radical forces. Elite discourse on the nature of the Spanish people had changed dramatically by the death of Franco. ₅₇ On the eve of Suárez’s reforms, ABC, one of Spain’s most important conservative newspapers, was full of references to the moderate and centrist nature of the people of Spain. If the Spanish people were indeed “prepared for democracy” and “ready to use liberty responsibly,” as these sources claimed, then democracy despite extremism was not so threatening. ₅₈

In both Spain and Portugal, pivotal elites were able to formulate their projections about the effects of extremism with highly credible information about popular preferences. In Portugal this information came from the elections for the constitutional assembly in April 1975. With a 91.7 percent turnout, these free elections were a key indicator of how the Portuguese people were grouped on the ideological spectrum. The Communist Party and the ultra-left attracted approximately one vote in five but were electorally overwhelmed by the Socialists and the centrist Popular Democrats, who together attracted over 64 percent of the vote. ₅₉ Moderate civilian and military leaders recognized that, if they could ensure a fair series of elections, they would gain power, and extremism would be held in check. This prediction held true.

In Spain pivotal elites formulated their projections about the effects of extremism with two sorts of information. One came from the public opinion polls conducted before and after the transition to democracy began. These polls indicated that Spanish society was ideologically moderate, so moderate in fact that the PSOE would have to move toward the ideological center if it were ever to be elected. Communism, even in its Eurocommunist variant, attracted surprisingly little enthusiasm, and popular support for violence as a means of solving subnationalist questions was overwhelmingly rejected, even in the regions in which subnationalist sentiments were strongest. In Spain, where the legislature of the dictatorship literally voted itself out of office, this information on popular preferences was absolutely essential in calculating the risks. ₆₀

Voting results provided a second source of information for Spain’s pivotal elites. The December 1976 referendum on political reform showed that 94 percent of the voters approved the transition to democracy and thus that the costs of reversing the transition might be very high. The June 1977 general elections confirmed the findings of the opinion polls. Neither the extreme right nor the extreme left won any parliamentary seats. The PCE attracted only 9 percent of the votes, and regionalist parties failed to attract the support of more than 7 percent of voters. ₆₁ Centrists overwhelmingly dominated the Spanish electorate, and knowledge of this fact diminished the risks of continued democratization.

The puzzle of why two very different transitions could proceed despite extremist mobilization is thus resolved if we consider how pivotal elites perceived the effects of extremism. In both Portugal and Spain pivotal elites saw democracy as a solution to the problem of extremism rather than a problem in itself.

Scholars who have seen extremism as a barrier to democratization did not misread the drama of regime transition. They simply focused on only one of several scenarios. A broader array of cases shows that democracy can be created despite so-called “extremist” demands and despite high levels of mobilization in civil society. In many cases, a “hot family feud” may, indeed, provide the proper environment for the forging of a new democracy.

Endnotes

Note 1: Dankwart A. Rustow, “Transitions to Democracy,” Comparative Politics 2 (April 1970), pp. 337–63. Back.

Note 2: Guillermo O’Donnell and Philippe Schmitter, “Tentative Conclusions about Uncertain Democracies,” in Guillermo O’Donnell, Philippe Schmitter, and Laurence Whitehead, eds., Transitions from Authoritarian Rule (Baltimore: The Johns Hopkins University Press, 1986), p. 56. Back.

Note 3: Manuel Antonio Garreton, “Popular Mobilization and the Military Regime in Chile: The Complexities of the Invisible Transition,” in Susan Eckstein, ed., Power and Popular Protest: Latin American Social Movements (Berkeley: University of California Press, 1989), p. 275. Back.

Note 4: Terry Lynn Karl, “Dilemmas of Democratization in Latin America,” Comparative Politics 23 (October 1990), p. 8. Back.

Note 5: Myron Weiner, “Empirical Democratic Theory,” in Myron Weiner and Ergun Özbudun, eds., Competitive Elections in Developing Countries (Durham: Duke University Press, 1987), p. 26. Back.

Note 6: Samuel Huntington, “Will More Countries Become Democratic?,” Political Science Quarterly 99 (Summer 1984), p. 212. Back.

Note 7: Daniel Levine, “Paradigm Lost: Dependence to Democracy,” World Politics 40 (April 1988), p. 392. Back.

Note 8: Dietrich Rueschemeyer, Evelyne Huber Stephens, and John D. Stephens, Capitalist Dcvelopment and Democracy (Chicago: University of Chicago Press, 1992), pp. 271, 223. Their comparison is with moderate clientelistic parties. Back.

Note 9: Larry Diamond and Juan Linz, “Introduction,” in Larry Diamond, Juan Linz, and Seymour M. Lipset, eds., Democracy in Developing Countries, Volume 4: Latin America (Boulder: Lynne Rienner, 1989), p. 12. Back.

Note 10: Robert Kaufman, “Liberalization and Democratization in South America: Perspectives from the 1970s,” in O’Donnell, Schmitter, and Whitehead, eds., Transitions from Authoritarian Rule, p. 88. Back.

Note 11: Adam Przeworski, “Some Problems in the Study of the Transition to Democracy,” in ibid., p. 63. Back.

Note 12: J. Samuel Valenzuela, “Labor Movements in Transitions to Democracy,” Comparative Politics 21 (July 1989), p. 449. Back.

Note 13: Ibid., p. 450. Back.

Note 14: These same thoughts about the need for moderation affected the Brazilian opposition. Margaret E. Keck, The Workers’ Party and Democratization in Brazil (New Haven: Yale University Press, 1992), p. 34. Back.

Note 15: O’Donnell and Schmitter, Transitions from Authoritarian Rule, p. 27. Back.

Note 16: Robert Dahl, Polyarchy: Participation and Opposition (New Haven: Yale University Press, 1971), p. 15. Back.

Note 17: Nancy Bermeo, The Revolution within the Revolution (Princeton: Princeton University Press, 1986), p. 5, and “Worker Management in Industry: Reconciling Representative Government and Industrial Democracy in a Polarized Society,” in Lawrence S. Graham and Douglas L. Wheeler, eds., In Search of Modern Portugal (Madison: University of Wisconsin Press, 1983), pp. 183–84. Back.

Note 18: Charles Downs, “Residents’ Commissions and Urban Struggles in Revolutionary Portugal,” in ibid., pp. 151–63. Back.

Note 19: Since the banks owned most of the nation’s newspapers, they were nationalized, too. Maria Belmira Martins et al., O Grupo Estado (Lisbon: Ediçôes Jornal Expresso, 1979). Back.

Note 20: Reinaldo Caldeira and Maria do Ceu Silva, Constituiçâo Política da República Portuguesa (Lisbon: Livraria Bertrand, 1976). Back.

Note 21: Because the mainland population was just over 8 million, this influx was especially disruptive. Back.

Note 22: Lawrence S. Graham, “The Military in Politics: The Politicization of the Portuguese Armed Forces,” in Lawrence S. Graham and Harry M. Makler, eds., Contemporary Portugal: The Revolution and Its Antecedents (Austin: University of Texas Press, 1979), p. 237. Back.

Note 23: John L. Hammond, Building Popular Power (New York: Monthly Review Press, 1988), pp. 233–35. Back.

Note 24: The first, second, third, and sixth provisional governments were considered politically moderate. The fourth and fifth were radical. Back.

Note 25: Nancy Bermeo, “Regime Change and Its Impact on Foreign Policy: The Portuguese Case,” Journal of Modern Greek Studies 6 (May 1988), p. 13. Back.

Note 26: Ålvaro de Vasconçelos, “Conclusion,” in Kenneth Maxwell, ed., Portuguese Defense and Foreign Policy since Democratization (New York: Camôes Center Special Report No. 3, Columbia University, 1991), p. 83. Back.

Note 27: Donald Share, The Making of Spanish Democracy (New York: Praeger, 1986), p. 77. Back.

Note 28: On the Grupos de Resistencia Antifascista Primero de Octubre, see Paul Preston, The Triumph of Democracy in Spain (London: Methuen, 1986), pp. 106–7. Back.

Note 29: José Maria Maravall, La Política de la Transición (Madrid: Taurus, 1981), pp. 24–25. Back.

Note 30: Share, p. 121. Back.

Note 31: José Maria Maravall and Julian Santamaria, “Crisis del Franquismo, Transición Política y Consolidación de la Democracia en España,” Sistema 68–69 (November 1985), p. 105. Back.

Note 32: Nancy Bermeo, “Sacrifice, Sequence, and Strength in Successful Dual Transitions: Lessons from Spain,” Journal of Politics, 56 (August 1994). Back.

Note 33: Eusébio Mujal-León, “Foreign Policy of the Socialist Government,” in Stanley Payne, ed., The Politics of Democratic Spain (Chicago: Council on Foreign Relations, 1986), p. 197. See also Fernando Morán, España en Su Sítio (Barcelona: Plaza and Janes, 1990). Back.

Note 34: Felipe Gonzalez and Alfonso Guerra, PSOE (Bilbao: Ediciones Alba, 1977), pp. 23–24. Back.

Note 35: In Chile in 1983 copperworkers led a cross-class alliance in school and store boycotts and caceroleos (intensive pot-bangings) against Pinochet’s regime. The protests continued monthly until Pinochet declared a state of siege in November 1984. In Argentina a mild liberalization was sponsored by General Viola, then reversed by General Galtieri. Back.

Note 36: “Suggestive” because different methods in recording strikes cross-nationally make the accuracy and comparability of strike data problematic. Interesting here is the similar direction of historical trends. Back.

Note 37: Cynthia McClintock, “Peru: Precarious Regimes, Authoritarian and Democratic,” in Diamond, Linz, and Lipset, eds., Democracy in Developing Countries, p. 351. Back.

Note 38: Cynthia McClintock, “Why Peasants Rebel: The Case of Sendero Luminoso,” World Politics 37 (October 1984), pp. 52, 81. Back.

Note 39: Ibid., p. 51. Back.

Note 40: I owe this insight to Catherine Conaghan of Queens University. Back.

Note 41: Carolina Hernández, “Reconstituting the Political Order,” in John Bresnan, ed., Crisis in the Philippines: The Marcos Era and Beyond (Princeton: Princeton University Press, 1986), p. 184. Back.

Note 42: Robert Manning, “The Philippines in Crisis,” Foreign Affairs 63 (Winter 1984–1985), p. 392. Back.

Note 43: B. M. Villegas, “The Philippines in 1986: Democratic Reconstruction in the Post-Marcos Era,” Asian Survey 27 (Fall 1987), p. 198. Back.

Note 44: David Rosenberg, “The Changing Structure of the Philippine Government from Marcos to Aquino,” in Carl Lande, ed., Rebuilding a Nation: Philippine Challenges and American Policy (Washington, D.C.: Washington Institute Press, 1987), p. 304. Back.

Note 45: Manning, p. 404. Back.

Note 46: “DJP’s No Tae-U Advocates Direct Election,” Seoul Hanguk Ilbo, June 29, 1987, p. 1, in FBIS Daily Report-East Asia, 124, June 29, 1987. Back.

Note 47: “Daewoo Shipyard Worker Dies from Injuries” Seoul YONHAP, Aug. 22, 1981, in FBIS Daily Report-East Asia, 163, Aug. 24, 1987. Back.

Note 48: “No Tae-U’s Election Rally in Chonju Aborted,” Seoul Domestic Service (in Korean) Dec. 10, 1987, in FBIS Daily Report-East Asia, 237, Dec. 10, 1987. Back.

Note 49: Kim Young Sam was attacked with bottles and rocks during a campaign address in Kwangju. John McBeth, “South Korea to Curb Campaign Violence,” Far Eastern Economic Review, Nov. 26, 1987, p. 10. Kim Dae Jung was attacked on December 6, 1987. South Korea actually experienced more radical group activity as democracy approached. Lo Shiu Hing, “Political Participation in Hong Kong, South Korea, and Taiwan,” Journal of Contemporary Asia, 20 (1990), 245. Back.

Note 50: See Glenn Adler and Eddie Webster, “Challenging Transition Theory: The Labor Movement, Radical Reform and Transition to Democracy in South Africa,” Politics and Society, 23 (March 1995), 75–106. In eastern Europe, Poland may be the only case in which the debate about the requirement for moderation is relevant. The collapse of dictatorship was so rapid in the other East European cases that the nature of the tactical discussions was very different. Back.

Note 51: Rustow, p. 352. Back.

Note 52: The transition from authoritarianism in the Philippines began as a renovating version of the legitimation scenario. Faced with pressure from the NPLA but still confident that he could win national elections, Marcos used a United States television interview to announce surprise elections. He was surprised by his overwhelming defeat, but his intention was to use a popular mandate to renovate a severely troubled regime. Back.

Note 53: Mario Soares, Entre Militantes: PS, Frontiera da Liberdade (Lisbon: Portugal Socialista, 1979), p. 18; and “Aviso aos interessados,” in PS, Fronteira da Liberdade (Lisbon: Portugal Socialista, 1979), p. 69. Back.

Note 54: Mario Soares, Democratizaçâo e Descolonizaçâo (Lisbon: Dom Quixote, 1975), p. 194, also 17, 156, 179, 190, 215, and 275. Back.

Note 55: Francisco Sa Carneiro, leader of the center-right PPD, shared Soares’ faith in the outcome of a free poll. See his interviews and speeches in Por uma Social-Democracia Portuguesa (Lisbon: Dom Quixote, 1975), pp. 154, 176. Back.

Note 56: Adolfo Suárez Gonzalez, Un Nuevo Horizonte para España: Discursos del Presidents del Gobierno 1976–1978 (Madrid: Colección Informe, 1978), p. 20. Back.

Note 57: For more evidence see Victor Pérez-Díaz, The Return of Civil Society (Cambridge, Mass.: Harvard University Press, 1993). Back.

Note 58: ABC, May 1975, p. 51; ABC, June 10, 1975. J. M. Areilza stated: “Between the Bunker and subversion there is an enormous segment of the Spanish population that is not inclined toward either of these formulas.” ABC, Aug. 1, 1975, p. 109. Another editorial stated that “there is an enormous sector of society between socialismo and integrismo.” ABC, Feb. 1976, p. 53. Not all of Spain’s political elites shared this perspective. Within the Spanish military, there were clearly actors who were much less certain that extremism could be held in check. They tolerated democracy by default because of institutional links to the king and out of a sense of inevitability. This point shows how the Spanish transition had elements of other scenarios, too. Back.

Note 59: Thomas C. Bruneau, “Popular Support for Democracy in Postrevolutionary Portugal: Results from a Survey,” in Graham and Wheeler, eds., In Search of Modern Portugal, pp. 21–42. Back.

Note 60: Paul Preston reports that the opinion polls “carried out by the government [before the law on political reform] convinced them that a center-right party not too tainted by Franco and backed by Suárez would have a healthy electoral future.” Ironically, polls conducted by the Socialists and Communists concluded that the conservative AP was more popular than it actually proved to be. Preston, The Triumph of Democracy in Spain, p. 108. Back.

Note 61: Maravall and Santamaria, “Crisis del Franquismo,” p. 96. Back.