Chapter VIII
To Craft Which Democracies?

This essay has been devoted to possibilism and to crafting. I wish to take the two one final step. To be sure, "possibilism" does not mean that democracy is the only possible outcome in the crisis of a dictatorship, or even the one that is generally favored. It simply means that, for all the reasons advanced in this essay, democracy can be crafted to be attractive. Besides, it would be foolhardy to deny that democratic prospects vary from case to case. In making reference to concrete cases, we have repeatedly contrasted countries where the crafting of a democratic agreement has proved difficult to countries where this has been less difficult. We have repeatedly advanced explanations for the difference. And even when we favored explanations that focus on the transition and its strategies, these in turn have begged for their own explanations—often of a deeper historical and structural nature. Similarly, we speculated about the chances of future democratization in different types of regimes and crises.

I wish to conclude this essay by revisiting the issue of whether crafting helps, even in difficult cases, to increase the

number of successful transitions. After all, we should not forget that in many cases the outcome of regime crises may be sufficiently indeterminate (by being, for example, stalemated) to give crafting greater leeway. Even where prospects are bleak indeed, it is not out of order to take a second look at the outer borders of what is possible and plausible.

The essay thus closes with a bottom-line argument that proceeds as follows. The society and history of some countries, or much more so their regime and system of political domination, may make their democratic prospects less promising. Thus political halfway houses may seem more likely. But their crisis may be such that status quo options are drastically curtailed and halfway houses may themselves represent costly impasses. This is where the subjective expediency of crafting can provide succor. In the process it may also dull the theoretical shine about "objective" democratic prospects.

In 1984 Samuel Huntington, whose observations on the global prospects of democracy offered the first foil to my reflections, wrote that the prospects are bleak.^[1] But in 1989 Zbigniew Brzezinski, writing about particularly nasty and tenacious dictatorships, announced in no uncertain terms: "It is almost a certainty that at some point in the relatively near future, given some major economic or political upheaval, politics as the expression of authentic social aspirations for multiparty democracy will truly return to the life of Eastern Europe."^[2] Numerically, the prospect of some Communist countries going democratic may not subtract much from Huntington's accounting; but it does raise questions about some tenets of development theory. Bearing on the latter, Huntington offered two independent explanations why "the likelihood of democratic development in Eastern Europe is virtually nil."^[3] One is the Soviet veto, and the other is that Communist systems, even when drifting from

totalitarianism toward authoritarianism, cannot renounce in principle their political control of the economy.^[4]

Given these discrepant assessments, it appears that some old certainties no longer hold. At the end of the 1970s Jeane Kirkpatrick gained public attention by pointing to a commonly held ultimate distinction between authoritarian and totalitarian regimes: the former, but not the latter, can be replaced.^[5] Now totalitarianism has been prefixed with a tantalizing "post," and Brzezinski doubts whether that significant distinction is still unshakable. If the prospects for democracy in tenacious Communist regimes can be reconsidered, why not reconsider prudently the prospects in other, and otherwise difficult, regimes? Where should prudence draw the bottom line?

What follows is not a complete account intended to review by cases and types democratic prospects worldwide. We are content to take a narrower, most-difficult-case approach, focused on the study of selected regimes. If we can show that indeterminacy and crafting still have room where, given the stubborn nature of particular regimes, the prospects are bleak, we can then make a stronger claim where the prospects are better. A source of confidence in the general prospects for democracy comes from the realization that most contemporary nondemocratic regimes should eventually experience crises stemming largely from betrayed expectations about their material performance.^[6] A source, on the other hand, of soberness comes from a theme familiar to the essay: today's crises should neither plainly topple the regimes and their personnel, nor induce the regimes, if the matter depended on them alone, to accept full political democracy.

Naturally, however, matters rarely if ever depend exclusively on regimes and their ingrained preferences. Let me thus test two types of regimes that, for different reasons, are

uncommonly resilient or, when not unmistakably routed and overthrown, singularly resistant to accepting full change. Is their resistance unshakable?

Testing Resilient Regimes

The two types ostensibly find themselves at the opposite ends of the spectrum of contemporary dictatorships. The first one is the traditional authoritarianism, infused with sultanism and personalism, that characterizes countries in the Middle East, North Africa, and Central America. I will sample out of this group the authoritarian regimes of Central America, whose military nature distinguishes them for favoring violent exists from their crises of political domination. The other type is Communist totalitarianism, whose personalism and traditionalism, though not always insignificant (North Korea, Romania, and China being only the most blatant examples), are supposedly cast within or superseded by the impersonal, pervasive, systematic, and mobilizational rule of the single party.^[7]

Let us begin with Central America. We have already contrasted the Western European with the Central American experience to point out why the respective dictatorships made exit toward democracy a likely occurrence in the former (even irrespective, for the countries that returned to democracy immediately after World War II, of its effects) but makes it a more distant one in the latter.^[8]

West European dictatorships had to contend with a number of state/political institutions with histories of their own, as professional carriers of public functions, that predated the advent of dictatorship. Efforts to uproot or radically politicize these institutions (bureaucracies, armies, judiciaries) were limited and halfhearted, or resulted in failure, or, once the dictatorships faced their crisis, revealed themselves as

ephemeral. As a rule, the institutions were expressly incorporated by the dictatorships to parallel new and specifically repressive or totalitarian institutions. The effect, not always or completely unintentional, was to restrain the regime's totalitarian features. In only one case (Germany) were the older institutions more clearly subordinated to the new ones. But, whatever their new place, the old state institutions never became coterminous with or were defined by dictatorship.^[9] Thus, at the moment of crisis the institutional weight of the past (a liberal and even democratic past) was central in helping European countries return to competitive politics. Not only could the structure of the state be more easily recycled to serve democracy, but also state institutions had an interest of their own in separating from dictatorship and being recycled, thus taking a step toward reaffirming their professional autonomy.

By contrast, the state institutions of the Central American countries that have experienced dictatorship are not just hostages or morose partners of their regimes. In countries where the state lacks a tradition of autonomy and impersonality, but shows one of primitive parasitism, in countries where public functions are privatized and the state operates intermittently,^[10] it can be said instead (though the expression unduly suggests the impossibility of change) that state institutions have a "natural affinity" for a form of dictatorship that is itself parasitic and predatory, and in which depredation is both personalistic and militarily organized.^[11] Why then should such state institutions, coterminous with dictatorship, be interested in recycling themselves for democracy? As we saw when discussing trade-offs, democracy has no immediate trade-offs to offer them and would instead threaten the logic of patrimonial appropriation by which they live. Why should a military, which in the region uses the state as both object and agent of violent appropriation, renounce its

privileged arrangement? Military leaders of Central America have a personalized praetorian conception of the state and of the relations they (as men of arms) have with their own military institution that is a far cry from that of, say, the Spanish military, whose reasons for installing the Spanish dictatorship had quite different roots. Thus it would seem that only a violent overthrow, not a negotiated transition, could dislodge similar regimes. Indeed, it is difficult to see how their opponents can place much trust, or rally much interest, in the willingness of the regimes to reform themselves.

A similar lack of interest in a negotiated transition characterizes in principle Communist regimes. The essay is laced with numerous, provisional, statements to this effect, but the essential point is simple, and not dissimilar to one of Huntington's points. It emerges quite well in the canonical contrast between communism and the also ambitious Fascist or semi-Fascist regimes of Western Europe. Despite fascism's ambition, communism, much more than fascism, is the master of a political program that, upon taking over the state, subjects state institutions to total guidance. Even when systematic terror subsides, the guidance does not lack justification. Indeed, by collectivizing the economy, Communist nomenklaturas, more clearly than their Fascist counterparts, build a doctrinally justified vested interest in their own survival. This holds them together and stymies lateral secessions.

Secessions are similarly stymied because a collectivized economy does not tolerate the survival and operation of influential corporate/economic interests—the type of interests (such as labor, professional, or business associations) that would be likely to look at their probable alliance with the regime as one of convenience and to act accordingly when the convenience fades. In fact, totalitarian disruption of collective

bonds supposedly travels beyond state and corporate interests to reach civil society at large and therefore the innumerable horizontal and often primary ties that allow people to act together.

Thus, it would follow from this model of communism that only limited reforms from the very core of the regime are conceivable, if at all. Further, Communist regimes have no reasons or no capacity to push reforms that jeopardize their nomenklaturas . Conversely, if the regime is interested only in limited reforms, civil society may not put much trust in those reforms and may lack the capacity (and possibly the interest) to obtain more.

If we now put the two models side by side—the Communist model and that of Central American despotism—we may understand the skepticism about how far their transformation can go. The fact that a number of Communist regimes settle for a sort of dictablanda should not imply their demise. There is similarly an impasse in Central America, where Panama is an embattled dictatorship, Guatemala and El Salvador have taken the road to democracy but cannot rein in military violence, and Honduras suffers from the same problem to a lesser degree. But we should not expect that impasses must somehow come to a felicitous end. For, in both cases, the models of dictatorship offer ample reasons to doubt foreseeable breakthroughs.

Yet excessive dependence on the models can also paint us into an intellectual corner. Both models imply that—because the leadership of the respective regimes, whether the regimes are still in existence or have been recently set aside, see no appealing trade-offs in the democratic compromise—stalling, repression, breakdown games, even military violence are therefore still among the methods that, each fitting its own circumstances, presumably pay off. The problem, then, is what to make, for instance, of the following statement issued

by the leader of the Communist group in the Polish lower house after voting for the election of a Solidarity prime minister: "Perhaps, we and our honorable colleagues from Solidarity had to live through tough years and mature separately before meeting. As to my party, maybe it was necessary to end up with our back to the wall for us to understand reality ."^[12]

But the answer to the problem (why does the regime not resist?), an answer that can take us out of the intellectual corner, is starkly contained in the statement itself: Assumptions to the contrary notwithstanding, resistance may indeed no longer pay off. It may no longer pay off in Communist regimes and in other regimes as well. It is not, in other words, unshakably true that our regimes are impervious to change, despite the way they have been canonically depicted. They may lose in the most incontrovertible way—on their own terms, because their methods do not achieve their goals. And when this happens, a new way may emerge of looking at the costs of toleration versus those of resistance and repression. Toleration may still seem costly, but if resisting change turns out, by its total ineffectiveness, to be incomparably more costly, then toleration may invite a second look. This is another way of understanding Robert Dahl's axiom about repression and toleration.

We should point out that a stalemate between regime and opposition is not the ultimate predicament it seems to be. Stalemates are by no means unusual ways of setting transitions in motion, but resemble what Dankwart Rustow believes to be the lever of democratic transitions: an inconclusive struggle, the protracted ineffectiveness of resistance, weariness, and disillusion; all of which makes outcomes indeterminate and may thus lead to final shifts in the strategies and even objectives of political actors.^[13]

The fact that a stalemate forces some of the regimes we

are focusing on to seek collective exits out of the stalemate raises serious questions, especially in the case of Communist regimes, about the value of our canonical models. Under the Communist model, a stalemate or, more so, an old regime's willingness to seek an exit that could travel all the way to democracy seem well-nigh inconceivable. If it turns out to be conceivable, this is only partly because, understandably, the model of a steady regime does not concern itself with crisis, demise, and strategic responses to them. Another and more important reason may well be that the model overstates the tightness, overbearing presence, pervasiveness, and therefore resilience of the mechanisms for regime domination.^[14] It may also be, as part of the latter problem, that the model (a handy guide through reality) is one matter and the realities another.^[15] In the first place, the model does not apply equally well to all concrete regimes. When it comes, for instance, to the pervasiveness and penetration of the Communist totalitarian model and its fit in concrete cases, the Soviet Union is one thing, but the often internally imported regimes of Eastern and east-central Europe are another. Yet another is China and its own domestic version of communism. Thus, without by any means having exhausted the list of Communist or pseudo-Communist regimes around the world—in Asia, Africa, and Latin America—we already have our hands filled with different possibilities and trends, as students of Communist models are well aware.

But picking at models and paradigms is neither my final assignment nor my forte. My final assignment is to explore the full value of a stalemate as a first step out of the most stubborn regimes; to show how an otherwise unlikely democratic exit may be brought about by a situation in which the divergent interests of stubborn actors can be satisfied jointly or (stalemate) not at all; to see how crafting may insert itself in the situation to produce joint satisfaction. Let

me carry the assignment one more step by examining how, when resistance becomes nonproductive, our two types of regimes may be forced and then enticed to reconsider their behavior. The exercise will also serve to remind us that recognizing and exploiting the appropriate conditions for a democratic exit may fall victim to human misperception and miscalculation. Therefore, optimism about the future boundaries of democracy should always be tempered by prudence.

Exiting from Central American Regimes

As to Central America, the important note of prudence is that weariness, growing from the realization that resistance and violence no longer pay, may be long in coming. The recent state of affairs in the region should suggest weariness, yet signs of readiness for a collective exit from the impasse are unclear. A predatory military with no institutional affinity for democracy may feel for a long time that it has nothing to lose by persevering with violence, even as its material returns, so crucial to the military, may stagnate, even as violence achieves no rollback. The military may therefore put up with a protracted struggle, which gets further protracted by radical armed resistance on the other side. Thus, in a region where, in addition, civil society is too thin and political formations are too poorly implanted to impose by themselves a process of peaceful reconstruction of collective life—in a region where revolt against repression, violence in response to violence, are endemic political currency—violence may as easily flare up as subside and weariness may as easily dissipate.

Still, nothing we have said declares weariness to be unproductive. Rather, the point to be stressed in relation to Central America is that, if weariness must break the spiral of violence, special timing is necessary to take advantage of

it before it dissipates. Because weariness is also relative (ineffective strategies become tiresome as concrete alternatives surface), it is also necessary that a concrete political settlement be proffered at the moment of weariness, perhaps by third parties. This may well be the only way to break the spiral in a peaceful manner. Thus, the theme of pacts reappears.

When first discussing pacts, we argued that in Central America there is nothing to trade off—except, we may now interject, when weariness sets in. We have similarly seen, when discussing elections as a lever for democratization, why unilaterally called elections are a weak weapon to persuade actors engaged in breakdown games to abandon their resistance. This is because such elections are not instruments for trade-offs. They do not constitute the type of political settlement that guarantees who plays what roles, which the region preliminarily needs. We may now interject, however, that in the presence of weariness, elections may serve to nudge political actors from resistance toward settlement. Elections become a more useful weapon. The ability of moderate political actors to run fairly competitive elections and to get out the votes during stalemated violence (and both words need emphasis) is per se an achievement with a potential to co-opt dissident actors that should not be discounted.^[16] The examples of El Salvador and Guatemala come to mind, but we may also draw a lesson from Portugal.^[17]

To be sure, the new democratic institutions of Portugal were not as weak as those of Central America. Still, two regime projects, as we saw them, confronted each other: one by the civil-military left and the other by the moderate parties. As a result, democratic institutions were introduced at the sufferance of the left, the constitution was emerging as something other than a full democratic compromise, and elections were dismissed by the left. Nonetheless, what began

to break the developing stalemate and cracked the cohesion of the left by showing a failure in its strategies was the very ability of the moderate parties to use the elections effectively as a weapon against their detractors and to compel them to come finally to terms with political democracy.

But the Portuguese case offers Central America another lesson: that of the importance of negotiated exits. Finding a constitutional settlement satisfactory to all Portuguese, while removing from the settlement what the left originally aspired to, has taken political realignments, painstaking caution, and careful transactions—occasionally stronger tactics notwithstanding. In the more difficult Central American context, a careful political settlement is not only equally necessary but also, unlike Portugal, cannot wait for the ineffectiveness of violence to become glaring on its own. It bears emphasis that in order to expose its ineffectiveness, violence should be challenged by its alternative, as it is, in fact, challenged in the regional settlement proposed by the five Central American presidents.^[18] Further, settlements cannot but be formally negotiated. Unspoken concessions, temporary withdrawals, implicit accommodations, ambiguous muddling through would exercise limited attraction by leaving in doubt the nature of the democratic game and the place of the contestants.

Last in treatment, but first in importance remains the question of what precisely a predatory military that must be negotiated out of its instinctive preferences (other avenues being closed by the stalemate) can obtain in exchange for unchallenged civilian rule. Once more, the answer is stark. It can obtain what, under normal circumstances, it has never been interested in. It can obtain its survival in a radically altered institutional form: that of a professional military, carrying explicitly assigned public functions. Thus, in order to break the spiral of violence, a negotiated settlement must

contain not only an electoral pact (of importance especially, though not exclusively, for the armed opposition) but also measures that would pursue what European polities had fortunately achieved before their authoritarian interlude; that is, the depersonalization of armies and other legal institutions of the state.

In the southern hemisphere, where professionalism has been equated with counterinsurgent ideologies of national security and with views of the military as the irreplaceable institutional guarantor of the constitutional order, professionalism is often taken to be the problem rather than the solution. Nevertheless, a professional military, more than a predatory or praetorian one, can be rooted in, and protected by, corporate guarantees that alleviate its need to dominate politics and society. Also, the prospects of a professional military may still seem slight. Yet, we must remember that conversions do not happen en masse and all at once, but through bandwagon effects.

As the psychology of secessions shows, blocs come apart progressively, revealing that even the most cohesive institutions are never perfectly so, but incorporate clusters of interests that respond differently to internal crises and external solicitations. If this is true of institutions with a collective public function, it may in one special way be even truer in our case. If a military with limited professionalization fails in its chief task—the effective perpetration of domestic violence—there is little else to hold it together. Thus, few situations work better than a protracted stalemate (if exploited at the right time) in dividing such a military. In addition, if the chief object of military violence—i.e., armed civilian resistance—is also included in a general settlement, military cohesiveness has even less reason to endure.

Besides, a perfect predatory and praetorian military—totally insensitive to alternative roles, totally in command of

the society that is the object of its depredation, totally tied to equally parasitic civilian oligarchies—does not exist in reality. Our attention to regime models should not blind us to the fact that in specific countries and specific historical contexts sectors of the military in the southern hemisphere have been tempted by more dignified roles, or that the alliance of the military with civilian oligarchies has occasionally come apart.^[19]

We should nurture no excessive expectations about the democracies that might emerge from the political settlements being sought today in Central America. The settlements may give birth to democracies where the military's role is still a matter of concern; where political parties run the gamut from personalismo to clientelism, populism and Third World radicalism, pay lip-service to or shelter myths about social revolution, and embrace chauvinist and anti-imperialist creeds; and where all the while social transformations go unattended. Nonetheless, such democracies would be an unquestionably positive change—not just because they would disprove scholarly pessimism, but because they would be infinitely superior to anything their people now have.

Finally—although the Central American model is not appropriate for understanding the regime preferences of South American civilian and military institutions—Central America (and perhaps more so Portugal) contains some strategic lessons for the new South American democracies, where the status of the democratic agreement is ambiguous. A disturbing sign in South America comes from the contingent difficulties with adjusting state and corporate interests to democracy, and therefore from the lack of pacts in regional transitions. In the region's embattled democracies, moderate forces may have to fall back on a battle of attrition, in which democratic institutions with a capacity for mass appeals are employed, as they were in Portugal, to wear down resistance.

But, with enfeebled institutions, attrition and stronger tactics alone should be last resorts. By themselves, without accompanying evidence that the democratic agreement pays, they may be insufficient for the task—leaving our democracies where they are now, at best. In Dahl's axiom, the costs of responding to democracy with repression are also a function of what democratic cooperation can offer instead.

Exiting from Communist Regimes

What is there for a Communist regime to trade off? I will approach the answer from a distance by bringing out one point of contrast between Communist and right-wing authoritarian regimes. It may be true that the latter are less durable than the former. But, at least in the Central American context, the lesser longevity does not stem from some greater propensity of authoritarian regimes to reform themselves slowly. Rather, when they are not stubbornly resisting, their inclination has been to step down abruptly, albeit often temporarily, to avoid dealing with untreatable emerging situations. In keeping with this mode of extrication, what accompanied the extrication was, not a series of reforms initiated by the regimes, but the strained resumption of competitive practices and institutions (elections, parties, unions, parliaments, civic and local associations) that the regimes had repressed or made a mockery of, but had rarely if ever formally abolished. Recent events in some Communist countries reveal instead no stepping down, but a flurry of reforms by Communist leaders. The reforms come on the heels of, but also go well beyond, a much more limited and often reversed history of reforms dating back to the beginnings of post-totalitarianism.

One way of understanding the difference is to note that the Central (and South) American pendulum between authoritarian and pluralist regimes—with the military largely

deciding when to step in or out—is patently inconceivable in Communist regimes. Communism is not a temporary caretaking affair, a régime d'exception with a limited, self-imposed mandate to put a disrupted or disrupting democratic house in order, as many authoritarian regimes like to present/disguise/justify themselves. Communism's original ambition is to offer, socially and politically, a permanent alternative, not just domestic but above all global, to liberal democracy. In turn, this very ambition may paradoxically explain Communist reforms when the ambition carries risks. At such points, reforms are judged essential for Communist survival.

A basic qualitative distinction must, however, be drawn between the earlier Communist reforms and the present ones. Only the latter ones carry a real, if unanticipated or unavoidable, potential for breaking up the unity of Communist nomenklaturas and eroding their centrality. Thus, only the latter ones create open-ended conditions for internal disagreement on the extent to which communism should be transformed. The reason for the distinction is that all earlier reforms remained within communism's pristine ambition to replace democracy the world over. In fact, they served to rescue the ambition, against domestic resistance and international obstacles. Recent reforms, on the contrary, are designed to rescue domestic Communist regimes from that very ambition and to set it aside—as the ambition appears finally unviable, except at internal costs that neither stonewalling nor retrenching regimentation can reverse.

In other words, present reforms and their potential for eroding Communist apparats reflect the fact that for the first time the global objective for which, and the global context within which, communism came to power, kept itself in power, socialized its economies, and entrusted them to its apparats, have both ceased to hold. In essence, as it loses

international coherence, communism (especially outside the Soviet Union) is becoming more like other, domestically oriented, dictatorships. And the general lesson is that, unless we take into clear account these mutations and their international referents, we may lose sight of how far Communist transformations may now go.^[20]

As Andrew Janos describes it, the original "focus of communism was external rather than internal, for its purpose was not 'catching up' with the advanced nations, but to destroy a modern world economy that was seemingly reproducing a pattern of debilitating economic disparities."^[21] Thus the Communist dictatorships of the Soviet Union, China, and Eastern Europe were not simply developmental dictatorships each designed to close the gap of its own backwardness, but the collective harbingers and instruments of a new international order. Moreover, this global purpose, thanks to substantial tactical adaptations to permanent obstacles and changing circumstances, had in fact remained basic to communism until the first part of the 1980s. Though Stalin rejected Trotsky's vision of a global revolution through popular uprisings, global Soviet hegemony through more conventional statecraft and expansionism was still his central objective. The postwar acquisition of Eastern Europe rallied those countries to that objective. And Chinese communism, also a postwar phenomenon, displayed, in its very rift with the U.S.S.R., a similar global dedication to peasant revolutions against the "Western City." Nor did the death of Stalin signify an abandonment of the global objective, but rather a retrenchment to consolidate—under the banner of peaceful coexistence—communism's global challenge to the West. In point of fact, this relative retrenchment did not and could not go without a continuous war of attrition to sustain and test the challenge in various peripheries of the world.

Thus, what I have labeled post-totalitarian reforms were

in effect institutional adjustments to tactical shifts in a global objective. Even in Eastern Europe, where the end of Stalinism allowed some of the local Communist regimes to pay greater attention to the problem of their domestic economic performances, Brezhnev's doctrine of limited sovereignty brought brutally home the permanent superiority of the global objective.^[22] And as long as this has been the case, as long as communism's socialized economies have been functional to the global objective, the integrity and indispensability of the apparats in charge of the economies have also stayed unchallenged. In fact, post-totalitarianism removed lingering menaces to party and bureaucratic integrity, security, and status that stemmed from Stalin's extralegal despotism,^[23] or from the populist subversion practiced by Mao during the Cultural Revolution. Thus, in a way, it is precisely when the global objective of communism became routinized that the apparats' staying power reached its peak.

But party and bureaucratic integrity lost much of its endurance and raison d'être when—after Brezhnev in the Soviet Union and Mao in China—the global objective receded, making room in the Soviet Union, Eastern Europe, and China for more routine concerns with domestic backwardness and development. According to prevailing analyses, this momentous shift may be explained by the devastating drain made by socialized economies on domestic resources, worsened when the economies had to be geared primarily to sustain the global objective. The inability of socialized economies to carry domestic growth, let alone the global objective, and the great deprivations that this state of affairs has cost Communist societies, reached crisis proportions as Communist societies became more exposed to Western political and lifestyle models, and as the price tag for maintaining global parity with the West became unbearable.

Thus, neither the more recent lowering of existing barriers

to communication and exchange between East and West nor the post-totalitarian emergence of an implicit "social contract" between Communist governments and their societies, could reabsorb the crisis.^[24] On the contrary, by their timidity, these socially ameliorative steps, when consciously compared to their Western counterparts, both announced and intensified the crisis. We should make no mistake. In regimes that, for all their retrenchment from goals, still attach tremendous importance to a historical and popular mission, the protracted failure to rally a sullen population to new but still less than believable performance goals signifies a major dispiriting setback.

Whatever the reasons for the crisis, the fact that the global objective has receded and that growth and social issues have become central and yet most difficult to treat, signifies, as anticipated, that present Communist nomenklaturas have become dysfunctional: no longer the solution, they are now the problem. Thus, their cohesiveness and monopoly of power, which are so much part of the functioning totalitarian model (and, in some ways, even more of the post-totalitarian one), are seriously shaken by reforms. For reforms that touch upon the political classes now reflect and demand not adjustment but a shift in the objectives—for which nomenklaturas are hardly equipped. It follows that reforms undermine the vertical and horizontal coherence of the political classes. If the latter rally against reforms, as the conventional view has it, loyalty to the leadership comes into question. And if reforms trigger latent, and previously less consequential, horizontal differences—tied to specific institutional functions—the ability to give reformed communism a reconstituted administrative backbone comes into question.

To be sure, the core leadership (from which, and only from which, change originates) wishes reforms but also wishes

to save the political class, no matter how transformed and trimmed. But it is hardly clear to anybody how the double feat should be achieved because of the unprecedented shift in goals and reforms and the inseparability of socioeconomic and institutional reforms. At what point do reforms intended both to revitalize society and save its political class instead contribute to the latter's disunity and even demise? If, in addition, civil society resurfaces, as it has in many Communist countries, with a decisiveness that takes the political classes by surprise, communism is faced with previously untested dilemmas on how far to proceed, if at all.

Communist leaders have given different responses to the dilemmas. But in all cases their responses have proved incapable of reabsorbing the crisis of their political classes. We may distinguish two types of responses. In the first one, exemplified by China, leaders may give precedence to the new objective of economic development through more market-oriented policies while keeping a tight political lid on civil society (and on the political class). In this way, the unity and centrality of the political class should presumably be reconstituted around a core that is politically conservative yet institutionally in charge of economic reforms within a political command context, and all the while shielded from civil society. In point of fact, however, backing away from political liberalization and trying in other ways (stern if necessary) to reconstitute the loyalty and unity of the political class appear to be only troublesome stopgap measures hiding unresolved tensions within the class—tensions that both reflect and intensify those in civil society.^[25]

The second (Soviet and Eastern European) response fares no better at reconstituting a shaken political class and may instead accelerate its visible coming apart. At the same time, it has a potential of its own for inducting a Communist class, wholly or in part, and willy-nilly, into competitive politics.

The second response consists in going precisely in the opposite direction to the first by giving priority to a modicum of politico-institutional liberalization. It does so in order to sell difficult economic reforms that will, if adopted, have a high cost for Communist societies.^[26] The response relies on a closer rapport between reforming leaders and civil society. As in other transformations of dictatorship that are guided from the top, reformers calculate that the revival of civil society, stimulated by the very transformations in regime performance and purposes and by international demonstration effects, is irreversible beyond a certain point. They thus use such revival, and in part direct and favor it, to pressure reluctant political classes into supporting limited reforms. Still, support may not be easy to gain. Reforms may instead change institutional roles and thereby introduce conflict among them.

With a leadership pushing for some institutional diversification and divestment in a context of political liberalization, and with a political class now unable or unwilling to resummon communism's global purpose without fostering (especially in Eastern Europe) further institutional divisions, the unity of the political class may now be as threatened as that of its counterparts in other dictablandas similarly perched between past and future. As in the case of the other dictablandas , its unity (and in fact the unity of the leadership) behind limited reforms intended to prevent collapse becomes contingent on the credibility of that intent—an intent that is already shaken by institutional diversification and divestment. Indeed, measures to decentralize public tasks, to divest institutions of excessive powers, to create new bodies for legal or public control—all to loosen the stranglehold of nomenklaturas and revive performance—seriously test the vertical and horizontal solidarity of the political class. With what consequences for class behavior?

One possibility is an increase in class fears of being placed

at risk, or even jettisoned, by reforming leaders; and this, of course, may induce conservative foot-dragging by the political class. Another possibility of greater interest to us is that measures of institutional reform will offer fractions of the political class glimpses of how they may survive in a more competitive political order—for instance, if previously shielded party cadres are put in the position to hold office in elected bodies; if labor fronts are freed of party controls; or if local officials are charged with local grievances. It is thus possible that glimpses and anticipations of alternative institutional roles may at long last offer those fractions that incentive to secede that, under normal circumstances, Communist regimes are least likely to offer. So an environment of choice may emerge that is closer to that which is typical of other transitions.

Whether secessions do occur; whether they have a sufficient impact to cause bandwagon effects; whether regime reformists and opposition moderates are both present in any specific case so that a convergence between them is concretely possible; whether, finally, convergence may acquire sufficient weight to offset resistance to a democratic agreement; these are questions that cannot be fully answered before we know the facts.

In the Soviet Union, at any rate, the size of the country, the size of its political class, the size and diversity of its problems (among which ethnic and nationality ones are paramount) mean that positive answers to all these questions are not at all certain. The difficulties may no longer lie in the special, pristine nature of the regime and the role that its political class occupied within it, nor in the presumed thinness of Soviet society and its groupings—a thinness that domestic turmoil largely belies. But the difficulties nonetheless exist. They reside in the immensity of the domestic reconversion problems with which the regime is confronted.

Unlike other dictatorships that just prefer to step down and wash their hands, the Soviet regime cannot turn away from these problems, especially because, given the size and diversity of the country's problems and the way they bring to the fore old and new societal cleavages and conflicts, a moderate opposition has not surfaced. Still, even in the Soviet Union, as we will see in the next chapter, the country's international standing, its new global responsibilities, and the very immensity of the problems it must address domestically and internationally may also work to encourage innovation by the leadership.

As to Eastern Europe (Romania excepted), one potential advantage is that steadier moderate oppositions have a greater chance to develop as a counterpart in prospective transitions. The chance is greater because Eastern Europe's societal problems are less divisive, and civil society in some of the countries has demonstrated greater capacity to refuse encroachment by an imported Communist model. The presence of a moderate opposition has a double advantage. To paraphrase a previous quote, it can, as it has in Poland with the aid of semicompetitive elections, place the regime "with its back to the wall." But, by thus creating or anticipating a stalemate (Hungary), it can also induce the regime to tolerate a more competitive system.

Last but not least, what are the trade-offs? The answer, again, is stark. One trade-off is the survival of a reformed Communist party as a significant player in a prospective multiparty system. The more the regime can anticipate the stalemate, can take the initiative in dealing with the opposition and thus preserve popular support, the more likely its survival.^[27] This is perhaps the most important trade-off and the most promising it its far-ranging effects on the political class. For regimes that aspired to create new Socialist societies and assigned the single party the central role in the

endeavor, and for parties born from a labor tradition of mass organization deeply rooted in twentieth-century European politics, recovering that tradition and revamping the active role of the party are not insignificant targets.^[28] And because, for all the emphasis on returning to the market, Eastern European democracies that seek a place in a common European home would be reformist and interventionist in policy matters, this interventionism is another aspect that should favor the evolution of reformed communism.

I as reminded, and so, surely, are Eastern European Communist parties, of the path followed by the French and especially the Italian Communist party at the end of the war. Being unable to rule on their own and having reconciled themselves to a state of necessity, both parties became (for the first time in Communist history) important players in democratic politics. The Italian party in particular, by pursuing a more pliable strategy of adjustment to democratic politics and to changes in the Communist world, has managed to transform itself into a progressive party with, almost half a century after the war, a still significant national role.^[29]

The advantage of party reforms that come from the very core of a Communist regime is that, although difficult, they carry greater legitimacy. It is one thing when radical reforms are sought by specific regime fractions and institutions that are trying, so to speak, to secede laterally; it is another when the same is done by the core leadership—that is, by the leadership of the very party to be reformed. The core cannot be held disloyal or incompetent as easily as a group of secessionists. In fact, writing about the Soviet Union, Jerry Hough argues to the contrary that the strong centralization of the Communist party gives the core relative autonomy from its political base, as well as the ability to manipulate and mobilize supporters.^[30] This, plus the core's broader perspective on national interests, places it in a better position to push

for policies that would transform the party. It follows that party reforms that originate from the core have a greater potential bandwagon effect on other sectors of the Communist political class. I have in mind sectors that, because of their functions, are presumably more likely to resist institutional innovations.

Left without the ideological guidance and political protection of an unreformed party, and hardly in the position to attack the leadership's new behavior openly, the less visible nomenklaturas (those who regulate and administer in behalf of the party; those who have a vested interest, material as much as programmatic, in the preservation of their pervasive and uncontested presence) are hardly likely to turn their foot-dragging and resistance into what I have described at various points as breakdown games.^[31] After a stalemate and party reforms that foreshadow the adoption of a competitive game, it is more likely that such nomenklaturas would lose collective direction and resolve. This in turn, as other avenues seem to be closing, favors their recycling (and thinning out) in the service of a qualitatively different regime. Similar developments are not new in the history of European bureaucracies caught between fading dictatorship and emerging democracy.

For all intents and purposes, what we have at this point is the emergence of another trade-off, now involving the administrative classes. The trade-off at this point is not much different from that offered to Western European bureaucracies coming out of dictatorship: the reconstitution of their corporate autonomy as legally bound institutions. Such autonomy should not be unpalatable to administrative classes that—for all the overlapping of party and state and the attendant internal crypto-politics—must have come to see themselves over the years, especially in the post-totalitarian era, as performers of collective tasks. Timing, in particular,

may help to make the trade-off more palatable. As typical of the way that dictatorships approach reforms in response to regime crisis, reforms of the administrative apparats will presumably have been adopted by our regimes (and in some cases they already have) before their transitions are completed. Thus, these reforms may offer the apparats those not necessarily unpleasant glimpses of how they would operate within new institutional roles that I discuss above.

Incidentally, the roles assign a new weight and value to legal accountability, thus introducing external checks and internal restraints on the pseudo-compliance with which the administrative classes (even as they are reformed) will continue to plague any new political system.

Conclusions

We are nearing the end of our journey. Because it is natural that readers tend to remember, and hold the author accountable for, what they last read, I will close the chapter, at the cost of inducing tedium by repetition, with a final reminder. My excursus on Central America and the Communist world was not meant to demonstrate that democracy is near. These are still among the trickiest places for any turn toward democracy to take place. It is no accident that I have shown greater confidence in democratic outcomes where, as in Poland, the process is already further along, and have been somewhat more guarded in assessing evolutions elsewhere. My intent, here and in the essay, has been more modest. I have not undertaken global predictions. I have pursued instead a narrower range of observations: how and why repressive regimes may have come to accept mutual tolerance over repression when trapped in a no-win prisoner's dilemma. And I have pursued prescriptions for

brightening prospects, namely, what political actors can do collectively to converge toward tolerance.

But, in pursuing this line of analysis, I have aimed for something more. I have pointed out that many political actors already have those prescriptions in hand. Thus, if not near, democracy must be judged as standing a better chance. Thus, also, many set ways of understanding the prospects of democratizing stubborn dictatorships and stubborn societies may well gain from being supplemented and amended by these and similar considerations about the role of human choice.

However, to paraphrase the opening sentence of the essay, holding such a concern is not the same thing as invoking a new paradigm. Monitoring the progress of democracy needs instruments that are more sensitive to the material at hand.