previous chapter
Chapter VI Beyond Transitions Why Democracy Can Deliver on Its Promises
next chapter

Chapter VI
Beyond Transitions
Why Democracy Can Deliver on Its Promises

Crafting democratic rules—crafting them collectively and expeditiously—is not an easy task. But once the rules have been attended to, democratic life assumes a significantly different, sturdier, quality.

The thesis of this chapter is simple. When an agreement on democratic rules is successfully reached, the transition is essentially over. Democracy enters a new phase in which the behavior of the actors is influenced, to an extent not seen before, by the presence of the new rules. The effect is double: (1) The new rules constitute a compelling, more civilized guide for the expression and handling of communal conflict. They cannot be easily overlooked. (2) The rules are not so rigid, however, as to prevent their adjustment over time. Nor are they intended to fully dictate the performance of democracy and the outcomes of communal conflict. Indeed, by agreeing to the rules, political actors agree that performance and outcomes will be uncertain to some degree.

Each of the two effects has a potentially positive function. On the one hand, the adoption of a civilized discourse signifies


that greater confidence is built in the stability and viability of mutual security. On the other, the pliability of democratic rules and their somewhat open effects leave room for improvement within mutual security; this is also a way to build confidence in the democratic game.

There is a familiar counterargument to my claim; to wit, just as it takes time to craft an agreement, so it takes time and habituation before the agreement is secure and any danger of failure, stemming from the transition or its antecedents, is removed. Thus, unless we were to define rule agreement in a way that suits the claim, the decision to agree on a set of rules is not sufficient to reorient collective behavior significantly and does not constitute a turning point.

This counterargument reasons from a sensible kernel: habituation, accompanied as it is by the testing and rooting of institutions, must play a role—although a difficult one to assess—in building viability and confidence. Still, it is the initial agreement on rules that is decisive for removing the danger of failure. When political actors agree on a set of democratic rules, we can take this as a sign that, for whatever reason (conversion, but more simply, and more likely, a fall in disruptive resources), those among them who were bent on "breakdown games" have now lost interest in them, or have become marginal, less capable of rallying support from sympathizers, or less likely to trigger similarly destabilizing defensive responses from adversaries equally contemptuous of democracy. Further and more to the point, the more expeditious the agreement, the better it is. Only when the agreement is reached can the new and more promising phase—buttressed by the agreement—begin in earnest. This is generally labeled a phase of consolidation. But if we follow the prevailing term, we should not attach a surfeit of meaning to the word; the decisive role in establishing democracy


belongs to the agreement phase, not to consolidation.

It should be clear that an agreement that halts breakdown games and opens a new phase of testing and adjustment may have to be something more than the agreement of a simple majority. If the object is to make sure that no significant actors are left out—that is, no actors who, by staying out, can continue to endanger the agreement—then any agreement, even if it does not formally include those actors, must at the very least be drawn with an eye to co-opting them. Or, in fact, a broader, more composite, more patently collective agreement may at times be needed.[1] Not surprisingly, this is the type of agreement whose emergence is favored by the transitional choices and tactics of garantismo and pact making. If the agreement does not include, or is not drawn with an eye to, significant actors, then single democratic institutions may well be adopted and put to work. But their effectiveness and persuasiveness—their capacity eventually to co-opt those who stayed out of the agreement—are in jeopardy.

Still, nothing more is expected of the agreement beyond the requirement that no significant actors be overlooked. Because the reasons and motives leading to the agreement are as diverse as the political actors, the agreement does not always entail the achievement of normative legitimacy. The object of negotiating democratic rules with reluctant actors is not to convert them point-blank but to affect their behavior.

True, we should be mindful of Hirschman's warning about the drawbacks of paths to development that rely on behavioral changes that precede attitudinal ones. Although such paths are more common and successful than we think, when political actors trespass, intentionally or not, into new behavior


(for instance, holding free or nearly free elections) before consonant attitudes develop, the path to development "will be more halting and circuitous," tensions between old and new attitudes may well persist, and the "development profile and experience cannot but bear the marks" of the path that was followed.[2] From our perspective, however, the important point, and Hirschman's contribution, is that development can still proceed without waiting for attitudinal change; the latter is not a requisite. Indeed, attitudinal change to remove dissonance between old attitudes and new behavior may not always occur. Meanwhile, the first and contingent choice of trespassing can induce the trespassers to yet other behavioral commitments, with further unanticipated consequences of their own. In this way, actors come to comply with the results of actions that they had taken earlier in the process with other intents and expectations.[3]

Living By The Agreement

Political actors who subscribe to democratic rules may differ in their motives but have one thing in common: they have subscribed to the rules. This has implications for the chances of reviving breakdown games and for the quality of conflict in the new phase. Once the new democratic rules have been recognized—once the prospects of emerging from the transition with some more hybrid regime, most agreeable to some political actors but least to others, have been set aside (albeit for calculus)—those prospects, and with them the prospects of reviving breakdown games, should get dimmer and roundabout. The drama of impending failure should recede. This new phase is new because it imposes previously less significant constraints on the revival of breakdown games and new incentives to focus on the more constructive democratic game.


The first constraint is the very collectiveness of the agreement to adopt new rules—the decision whether to enter into or exit from the agreement is influenced by what other actors do about it. Just as reluctant actors may have reasons for being included in the agreement when support for it increases, so they also may have reasons (despite their inclination to reconsider) for not being the first ones to abandon it. A move to scuttle the agreement might eventually rescind it, or may isolate the perpetrators. It may leave them out of what is emerging as the only game in town. The more, therefore, the game goes on, and the more actors practice it, the more costly it seems not to play it. If nothing else, there may be no other or safer way of attending to one's interests. So, if Dahl's costs of tolerance and cooperation are of concern to some political actors before they consider the agreement, afterward it is the cost of intolerance and repression that should worry them more.

The behavior of Communist parties in postwar Western Europe aptly illustrates this process of behavioral adaptation to first choices, an adaptation to which the parties have at times consciously contributed. Undeniably, because it has nonetheless taken some time for the Communist parties to practice the game fully and without afterthoughts, the path of democratic development has shown idiosyncracies. But a mix of subjective reasons, motives, and calculations can be found in various degrees behind practically every existing democracy. Further, the mix of reasons is not only a mix of different actors but also a mix inside single actors.

Closely connected with the constraint just discussed, the new phase also offers inducements to move away from breakdown games. The more costly it becomes to exit from the agreement, the more the partners to the agreement, even reluctant ones, will be motivated to focus their attention on the actual operation of emerging rules and institutions.


Whatever reservations they may nurture, they will formulate them in the spirit of those rules and in view of what they promise. Thus the agreement will be progressively tested, with still some drama but with fewer actual risks, on democracy's own terms and within the democratic compromise. In this sense, too, a decision to revert to breakdown games requires justification of a special sort. In other words, measuring the agreement against institutional performance is a cause of uncertainty (of voice), but not one of necessary instability (of exit).[4]

Another inducement to move away from breakdown games is that, in the new phase, the political actors and the political agenda change in ways that strengthen the relevance of democratic processes for both. In the first place, the political arena, which at first tends to be occupied by elites and selectively mobilized constituencies exploiting positional, legal, professional, charismatic, or military advantage, is now shared with social and political formations sensitive to electoral and mass appeals. In the second place, the political arena is now regulated by a set of democratic rules, therefore democratic state institutions also join in as a significant part, indeed the gatekeepers, of the arena. In the third place, the emergency tasks of the transition proper (the reestablishment of law and order, the removal of the dictatorship and its institutional residues, the search for a democratic agreement, the accommodation of vital corporate interests) are progressively overshadowed by activities concerned with group positioning, institutional routinization, and in sum the definition of how the democratic game is actually unfolding, how rules are working out, and how benefits are distributed. Thus, finally, the originally overloaded and cramped agenda gives way to a decisional process that is better timed, more normal, more informed, and more attentive to recognizable and socially sanctioned groups and institutions.


It is in good part because of these developments that the political actors' time perspectives on performance tend to become more relaxed, as we shall soon see, than they were during the transition. In the midst of the transition, when an agreement is still being sought, time seems to be of the essence, if not for all actors then at least for those who seek a democratic exit. Afterward, the justified desire to cut time short, or at least to set precise deadlines, tends to be replaced by an equally justified appreciation that, to prove itself, the agreement deserves and can afford more measured tempos in keeping with the typical give-and-take of the newly begun democratic process.

But the final and most important inducement to shift actors' attention toward the democratic game resides in what is special about its rules. Democratic rules are special in two ways. In the first place, political actors know that the rules do not concern themselves directly with outcomes, that agreeing on the rules is agreeing mostly on procedures and institutions, that each procedure and institution will impinge on outcomes only partially and probabilistically; in sum, political actors know that performance is not meant to mirror the agreement faithfully. That is why the process of entering into the agreement is surrounded by obstinate jockeying and wrangling. But that is also why, once the agreement is reached, the discovery that outcomes may occasionally vary from what the agreed-upon rules may have led one to expect should not engender necessary dismay and rejection. The agreement could still endure and thus continue to gather support.

In turn, democratic rules justify and reward patience. And this is the second way in which such rules are special. Plainly, they open the political arena to participation and the airing of political demands, they process demands through various institutions (rather than make them fall on one dictator, one


junta, one party), they let public opinion register disagreement by voting out the government rather than overthrowing the regime. If these political practices contribute to a climate of apparent political instability and democratic irresoluteness, they also function as safety valves defusing resentment and dispersing its targets. In fact, they do more. By distributing decisions among various institutions, the new practices call for a mixture of repetitive conflict-and-cooperation games.

The mixture, the availability of multiple decisional channels, and the opportunity to adjust and repeat decisions all improve the chances of producing decisional equity—or, perhaps more important, they increase the perception that such chances exist. Closely in keeping with the perception is the propensity to base the assessment of democratic performance on something broader and more elastic than the satisfaction of individual demands.[5] Because democratic decisions emerge out of aggregation, elimination, reformulation, deferment, give-and-take, the mixture of conflict and cooperation that often surrounds these processes is not only about which demands should be entertained and which deferred but also about how the processes should be conducted, what role the contestants located in and out of government should play, what symbolic or tangible side-rewards the losers should reap. It is the conduct of processes, as much as their outcomes, that actors will focus on in order to evaluate performance. Thus their evaluation will be broad and encompassing, both with respect to the evidence it relies on and the time span it covers. Once again, with this span the democratic agreement can continue to gather support.

There are unmistakable problems of habituation to the new tempos of democracy, as there are problems with its give-and-take. Scholars are often impressed by dramatic cases of democratic collapse, such as Spain's second republic, Italy's


pre-Fascist democracy (1919–21), or the Weimar Republic, and cite them as evidence that many political actors remain unwilling to adopt a long-range perspective on democracy's achievements, that their impatience can converge even as they may oppose each other, and that therefore democracy, with its moderate-to-reformist perspectives, can prove itself inadequate. But, before drawing such conclusions, we should take equal note of those cases where reformist perspectives prevailed.[6] These cases are often rather uneventful, and their diminished newsworthiness does not invite attention. In the folds of nonevents (or of deceptive events), we may yet discover that the new tempos demanded by democracy are well understood by those who have already entered into the agreement.

In fact, sensitivity to the investment made by entering into the agreement and reluctance to scuttle the democratic experiment prematurely are not always uneventful achievements. At times their presence is brought to light by salient episodes that, at first, suggest quite a different turn of events. I have in mind episodes, such as an attempted coup, that threaten the agreement, but by so doing may actually rally around the agreement the other political forces. The usual expectation, well illustrated by the denouement of the second Spanish republic, is that such episodes of destabilization set in dramatic motion a spiral of escalating events. If, let us say, the military threatens a coup, then the popular sector will respond in kind and democratic moderates will waffle and backslide, triggering a polarization that reveals how weakly most actors are committed to the democratic game and how ineffective the game can be. Against this plausible scenario there stands, however, the fact that not all episodes of destabilization come out this way. But, again, these are precisely the episodes that are more easily forgotten because the drama is unfulfilled.


We should also not forget that, if a coup or any other destabilizing event has greater chances of succeeding before an agreement on the rules is reached, these chances should diminish afterward. In this sense, another episode from Spain's history, the attempted coup of 1981, proved in some fundamental ways ill-timed, in that, though it came at a moment of serious internal crisis of the centrist governing party, it also came well after a broad constitutional agreement had been reached. Thus the coup had the effect of strengthening democratic resolve in the government and all other significant actors. It can be argued that the coup lacked credibility and full military support to start with. But this is part of the point I wish to make. The emergence of an agreement on democratic rules (to which the military or some of its sectors may in fact have directly or indirectly contributed) should discourage, as a starter, new breakdown games, given their diminishing chances of support.

Indeed, the record of military coups against reconstituted democratic governments illustrates well why the governments are more resilient than generally expected. Many coups are threatened, not all threats are serious, and only some of the serious ones succeed. The reason is the military's traditional reluctance to engage in coup activities unless prodded by significant civilian constituencies in a context of spreading social and political disintegration. Without such prodding, military conspiracies can prove insubstantial and ineffective.[7] Such may be the case if conspiracies resurface after a properly constituted democratic government is in place; then, the conspiracies not only may rekindle democratic support even among reluctant popular sectors but may also find little response among the military itself and its old civilian allies.

In sum, the investment placed in democracy by entering into the agreement may, to be sure, invite disenchantment


later on, as has happened even in the most successful of recent democratizations. Also, little will go without turmoil in the processes I have been describing. Because the agreement is in many ways a means to an end, and for many actors a second-best option, political actors will try to bend it in their favor, using performance as a basis for their claims. It is therefore naive to expect that political actors who managed to reach rule agreement through much struggle will suddenly convert to peaceful, orderly, and uneventful politics. Rather, the task of maintaining the agreement will still be accompanied by its own share of confrontation, tension, and animosity—a target of which is performance. Yet a charged political atmosphere, just like signs of disenchantment, should not be confused with an impending crisis of the agreement, or with what we may classically call a crisis of both performance and legitimacy. Such an atmosphere is indeed common to practically all new democracies and should be kept separate from accomplishments. The expectations vested in the agreement may produce later disenchantment, but they also counsel prudence before scuttling the agreement as defective.

Peter McDonough, Samuel Barnes, and Antonio López Pina have reported recently that Spaniards who show sympathy for the Franco regime are not necessarily averse to the new democracy. Spaniards in general look at the two regimes as distinct experiences, the more so as time goes by, and judge them on the basis of different clusters of orientations toward government.[8] The opinion data were collected following the conclusion of the Spanish constitutional process and during a period that covered an attempted coup as well as the advent of a new Socialist government. The fact that, at least in Spain, attitudes toward the past did not hold the key to the present fits well with the thrust of our argument.

Political actors may respond to democracy, or to specific


democratic institutions and accomplishments, on their own terms—or indeed in view of the times within which a new democracy happens to fall: times of prosperity, of international support for democracy, or of lowering expectations (or their opposite). Nor does adjustment to democracy demand extended attitudinal change so that feelings toward the old dictatorship be made more consonant with feelings toward the new democracy. Even as feelings remain dissonant, I wish to suggest that—if democracy or the times have indeed made the past, past—lingering nostalgia, while cushioning those who have deserted the past, may prove to be a less than forceful guide to present behavior.

There is a note of caution in the conditional form of the last statement. Making sure that the past is past, irrelevant, and obsolete—and, similarly, making sure that the temptation to try new forms of dictatorship or guided democracy remains inoperative—is in good part the task of democratic government. Nothing I have said so far invites democratic governments to be complacent. Despite the great importance assigned to the presence of a democratic agreement, nothing ensures that democratic governments will avoid excessive complacence, boldness, or fear. It is still possible that, by their mistakes, governments will set in motion a breakdown spiral that other political actors will exploit. The cumulation of unresolved issues, especially if a new democracy has a limited capacity to address them forcefully or without sacrifices, must be a matter of constant monitoring by democratic governments.

This is not just because democratic leaders would know the real extent of the threat only after the facts but especially because (whether or not the threat exists, whether or not public opinion has a case against government) good governance, as finally judged by public opinion, is what democratic government is all about. For all the leeway tolerated


in performance, for all the starting credit democratic government may enjoy, for all the benefits it may draw from international support or the temper of the times, for all the psychological investment involved in the original agreement, democratic consent is also a reflection of performance, and democracies seek to renew that consent.

This section has listed some of the reasons why, as democracy moves beyond transition, political actors may be led to stay in the game and renew their consent. With performance in mind, I now wish to pursue one of the reasons in greater detail.

Improving The Agreement

Under democracy, the game is not foreclosed. Democratic rules, I have suggested, are pliable in two ways. Their effects are somewhat open, and the rules can themselves be subject to adjustment over time. The effects of rules are open for many, already known, reasons. To wit, rules, being about competition, do not decide directly and entirely who will exercise decisional leverage but set the standards for measuring and legalizing that leverage. The simplest example is electoral rules, which declare the criteria for victory and in so doing may make it more or less difficult for some group (or some ways of associating) to win elections, but do not handpick the winners. In addition, the rules tend to disperse decisions among multiple decisional channels. As a result of these two qualities, the rules are also open—that is, not entirely determining—with regard both to specific policy outcomes and to who stands to win or lose from the outcomes. Finally, not only are the effects of rules open but also the rules themselves are not meant to be rigid, and circumstances may not allow them to be. Rules may tolerate, more


in some cases than in others, a whole range of adaptations—from fine tuning and reinterpretation to renegotiations.

By being pliable in these various ways, democratic rules make it possible to improve both democratic performance and support. The possibility of improvement is another measure of the degree to which entering into an agreement is a turning point in democratic normalization. In contrast, (to state the obvious in order to make a point with which I shall conclude this chapter), there can be no room for improvement unless an agreement already exists.

Let us keep in mind that the agreement we are discussing is one that removes breakdown games; that, for this purpose, the agreement may have to be drawn with an eye to co-opting political actors who have an interest in those games; and that, in turn, co-optation is favored by drawing on garantismo and by pursuing expeditious pacts. For I am suggesting that agreement so fashioned is more likely to prove itself pliable in the ways I just described. Once again, the analysis takes a normative turn.

Rules Have Open Effects

A known problem with convincing reluctant or uncertain political actors to accept democratic rules is, not so much the rules themselves, but that the rules' ability to deliver a fair balance of wins and losses can be proved only in the future—past the transition. I have spoken in this sense of a borrowed, or presumptive, legitimacy. But, as we move past the transition, the future is, so to speak, with us. Democratic rules can finally give account of themselves and their promises. This is the time to prove that the decision to enter into the democratic agreement, a decision that carries an initial psychological investment and from which it is difficult to backtrack, can also pay off—despite, at times, some starting conditions that threaten the agreement's ability to deliver.


Let us then take, as a demanding mental test, instances of difficult transitions. Let us see some of the ways in which the openness of the democratic game can still be rescued and can overcome, early on in the course of democratic life, some of the confining conditions. All we need to do is to retrace a line of argument to which I had promised to return.

One frequently expressed concern with recent transitions is the conservative context within which, given the guarded attitudes of political actors in general and the weight of nostalgic forces among political actors, most transitions take place. The concern, as described, extends to the use of pacts as transitional devices. For, while pacts ease the transition, they also seem to give public chrism to halting further democratization in politics, economics, and society. Although the concern has ample grounds, not all the ground is hallowed. For one thing, the concern belittles the function of some pacts in rescuing openness.

In the first place, the absence of pacts is not always a good sign. True, it may reflect a less reluctant and constrained transition, for which pacts are not needed. But, for all the alleged conservatism of pacts, absence may also reflect a state of affairs with even stronger conservative implications. For example, the old regime may still rely on its ability to guide a slow and guarded transition, society may be fraught with inequalities and ingrained authoritarian values, or the popular sector may be disorganized and demobilized. Any or all of these conditions, by convincing the old regime that pacts are largely unnecessary, may foreclose further democratization much more than pacts would. After all, pacts imply the presence of opponents who cannot easily discount each other, the need to accommodate this state of affairs, some disposition to recognize the need,[9] and some ability to respond to it.[10]

In the second place, there are conservative pacts, and then


there are conservative pacts. Some of them are intended not only to fall short of immediate democratization but also in fact to relegate democratization to some unknown future at best. I have argued that such pacts, rationalized by the need for habituation to democratic practices or some similar claim, hold little promise. But the pacts to which my preference and attention have gone—pacts in which immediate socioeconomic sacrifices and the postponement of reforms in society, the economy, and the state are nevertheless compensated by official support, even assistance, in the reorganization of the popular sector—hold much greater promise. Intentions notwithstanding, they cannot, by themselves, freeze the future. Strictly speaking, their emphasis on the political reconstruction of the popular sector removes, instead, any assurance that democratic performance will continue to be constrained, let alone determined, by the conservative expectations of weighty constituencies in society or the state.

This claim is nothing but a special extension of the fact that in democracy no set of interests can be secured, and entering into a democratic agreement (the more so if it is one with strong garantista components) means accepting the need to compete for, and to share, decisional leverage with other collective interests. Thus, making the agreement part of a pact that also calls for conservative sacrifices and postponements is still no fail-safe way of securing conservative interests beyond a vital core. These are good reasons why the popular sector can place confidence in the democratic agreement, including one that calls for immediate material sacrifices.

It goes without saying that uncertainty is for everyone. There is no guarantee that democratic performance will be confined by the conservative bias, but there is no guarantee that it will go beyond the bias either. Even so, I am afraid there is a tendency to overplay the evidence that original


biases persist. Suppose that conservative policies were to continue. Shall we take this as prima facie evidence that a new democracy is the prisoner of an original bias? We should not.

Consider the very real possibility that, as democratic life unfolds, the popular sector—now with its own electoral and organizational leverage—or some parts of the sector come to accept, and indeed to negotiate, the continuation of a program of economic austerity and labor restraint. Is it because putting a stop to the program would risk the resumption of breakdown games, or because of political weaknesses of the popular sector vis-à-vis its adversary, or because the popular sector has come to look at the program as plain "good policy"?

Or consider the possibility that politico-institutional reforms designed to make democracy more democratic (good examples are regional autonomies, lifting government monopoly of the airwaves, instituting police review boards) make no progress. Is it because the reforms find resistance among nostalgic constituencies, or because the democratic forces, for electoral or other reasons of their own, are divided over them?

Consider also the possibility that, under such circumstances, political parties traditionally linked to the popular sector win the elections and replace the government under which the original democratic agreement was reached. Chances are that some of the bickering will now occur between democratic forces, and even within the governing forces and their social constituencies; chances also are that some constituencies will stand accused of caving in to conservative pressures. What shall we make of the bickering and the accusation? And what shall we make of the electoral victory?

In one exemplary case to start with, that of post-Franco


Spain, these or similar developments do not suggest a frozen democratic performance or an ensuing waning of democratic support. In some ways, they suggest the opposite.[11] The Spanish transition was conservative in many of the meanings we have employed here. But, judging by the spectrum of the forces to which it paid attention, it was also open and consensual. The Pactos de la Moncloa and the constitution-making process combined these elements quite clearly. Economic sacrifices, labor restraint, and politico-institutional and economic reforms limited by the desire to preserve vital private and state corporate interests[12] went hand in hand with a negotiated constitution strong on garantismo , and with an approved reconstruction of the left in its party and labor components. The presence of the latter guarantees meant in turn that the presence of conservative sacrifices, the effort to co-opt the left in the process of democratization (sacrifices included), and the success of the first elected Spanish government in pursuing these strategies all proved insufficient to contain the growth of the left.[13] Less than seven years after the death of Franco, the Socialist party was in power, where it remains.

Of all the signs that a new democracy is consolidating its support, perhaps the strongest and also one of the most difficult to come by[14] is a peaceful government turnover. By this criterion, Spain has done well, and according to some analysts surprisingly well. So has the left. It could be argued that the success of the left, if success it is, carried a cost in democratic performance: to make their victory acceptable, the Socialists had to adopt the conservative bias. But the adduced evidence is improper, and the argument specious. True, the policies of the Socialist government, compared with those of the previous government, are remarkable not for their innovation but for their incrementalism. True, the realities


of the international and domestic economies of the last decade have defined the terrain on which the left can move, in Spain as elsewhere. True, radical reformism has lost its historical cachet.[15] But does it all mean that if the Socialist government were to act otherwise, breakdown games would resume? More precisely, does it mean that fear of such resumption is what guides Socialist policies and that, without the fear, government choices would be different? Does it mean that the sacrifices extracted from the left during the transition are still operative and unshakable? None of the evidence warrants such conclusions.

Pacts and negotiations of the type involved in the Spanish transition extract sacrifices for a precise purpose that is tied to the transition. They serve to reach mutual security more successfully. It is a measure of their success that they speed up the resumption of a normally competitive game—including government turnover—and that, with its resumption, some of those sacrifices lose their original constitutional purpose. Of course, part of the game (and part of what makes the principle of alternation in government acceptable) is that governments adhere to certain norms of restraints before deciding to undo each other's work. Behaving otherwise would betray the expectations that are built in the democratic bargain. Thus democratic restraint should not be taken as succumbing to continuing blackmail. Succumbing implies that democracy and its performance continue to be at issue. It is possible that certain historical junctures will push governments (any government) toward a conservative version of the democratic bargain and that efforts to change it will produce meager policy results. But are democracy and its performance thereby at issue?

The Spanish case advises prudence when choosing evidence that should support such notions. Indeed, Spain is a


good example, if not an easy one to replicate, of how confining conditions linked to the transition can be lifted, and how entering into the democratic agreement can pay off.

I have stressed that the occurrence of a peaceful change of government, the consolidation of democratic support, and the freeing of democratic performance from confining conditions were all facilitated, in Spain, by reliance on garantismo and pact making. But even without such reliance, and always provided that a democratic choice has been made, similar developments are not necessarily foreclosed by an otherwise difficult transition.[16] To take another southern European example from the mid-1970s, such a democratic choice was firmly present in the Greek transition to democracy.

The Greek case is an interesting combination of promises and emerging difficulties. On one side, the achievement of full-fledged political democracy, following the military's devolution of power to a civilian government headed by Konstantinos Karamanlis, met with little effective recalcitrance, either from the military (although a number of leaders of the old junta received hefty prison sentences), or from other nostalgics (although the monarchy was repealed in a national referendum), or from the left. Despite rumblings and an attempted coup by sectors of the military, the choice of democracy was expeditious and in some ways less controversial or belabored than elsewhere in southern Europe. On the other side (most likely, in retrospect, precisely because the lesser original recalcitrance made such measures less important) something was missing; there was no pact making, no consensual constitution making, and no special emphasis on garantismo . The constitution was drafted and adopted unilaterally by a conservative majority over the resistance of the left to its neo-Gaullist semipresidential features and to an electoral law assigning bonus seats to larger parties. Such


constitutional features exhibited qualities of confining conditions. They seemed designed to extend the rule of Karamanlis and his party, also by eroding traditional parties of the political center, and to quarantine the strongest party of the left, the Panhellenic Socialist movement (PASOK). But as in Spain, the Socialists found themselves in power only seven years after the fall of the dictatorship.

It seemed at the time of constitution making that Karamanlis, by cutting the left from the constitutional process and trying to reduce through the electoral law the chances of the left to govern, made a Socialist victory not only less likely but also more traumatic.[17] But the latter forecast proved to be as incorrect as the former. It overlooked the fact that the Socialist movement's acceptance of democracy, by the time of its victory, counted for more than its dissatisfaction with the strong majoritarian slant given to Greek democracy.[18] In fact, the advent of the Socialist government generated a climate of greater confidence in the new democracy. Because the Socialists won by the existing rules, the change in government showed that the confining conditions introduced by the constitutional choices of the transitional government were not insurmountable; it showed further that the rules could be adapted to the new governing context; and thus it showed that the right could coexist with a governing left.

But the reference to the adaptability of democratic rules makes the Greek case interesting in another way. Greece will be one illustration that democratic rules are pliable for yet another reason: not only because their effects are open but also because they can be reinterpreted and redefined.

Rules Are Adaptable

No matter how detailed the terms of the democratic agreement are—and often they should not or need not be—it will


look one way on paper and another way once implemented. Translating the terms of the democratic settlement into institutional roles and routines is a prolonged process, and its outcome is somewhat open. Eventually we may arrive at a point when the operation of democratic institutions and processes becomes routinized and predictable, as well as weightier, and therefore more difficult to change. But that point comes well after the democratic agreement is first reached. Meanwhile, because institutions and processes are not yet fixed, it is possible even to change and improve—at times explicitly, at times implicitly or surreptitiously—some of the agreement's original terms.

That rules can be adapted and even changed is fairly clear when the agreement involves mainly political actors with a democratic bias.[19] But the possibility also exists, more importantly, when the agreement co-opts reluctant actors. In the latter case, reluctance may surround the agreement and its terms with substantial fuzziness. Actors may haggle over those terms and their implementation in processes and institutions, thus revealing the constrictions of the agreement as well as the contrasting expectations and divergent interpretations it engenders. But such difficulties do not imply that a modus vivendi is ruled out. The fact that an agreement exists, fuzzy though it is, and the commitment to mutual adjustments that this implies, help a modus vivendi. As illustrated by democratizations in southern Europe, a number of possibilities are conceivable, from perfecting the agreement to making a virtue out of its imperfection.

One possibility is that some important part of the original constitutional settlement is explicitly amended before it takes full hold. Supervening developments—a significant electoral victory, the rise of a new party, and similar events that alter the balance of power and influence—may offer the opportunity. One example is Greece, as it went from a conservative


to a Socialist government following the Socialist's decisive electoral victory. What appeared to be a democratic system slanted in a presidential direction that was unappealing to the Socialists of PASOK has evolved into a fully accepted, more genuinely parliamentary system, allowing also for a more proportional electoral law. True, this required institutional reforms at the sufferance of the conservatives; nevertheless, neither the reform nor its aftermath has the makings of a serious constitutional conflict. Both conservatives and socialists have converged on a middle ground.

Another possibility is that, at a minimum, contrasting expectations and interpretations and uncertainties about the concrete shape of the agreement, will disappear in due time more or less of their own accord. This will happen not with explicit decisions or dogged renegotiations, but probably by a process of informal accommodations. As the democratic agreement is implemented, routinization socializes political actors, clarifies norms, and reshapes roles beyond original prospects. Again, and in addition, supervening political developments may contribute to such clarification.

Spain is a good example. Once the transition was over, especially after the Socialists' victory, Spain has settled into a sort of majoritarian government that is quite removed from the garantismo and the pact making of the transition, with their emphasis on consensus and cooperation.[20] The constitutional features that, in keeping with the spirit of the transition, adumbrated garantismo , have been reconciled with, and toned down by, majority rule. But this evolution was not clear at the time of the transition. Would garantismo and broad interparty cooperation become a durable feature of Spanish democracy, as they became in Italy? Would they be the instrument by which the governing UCD (Union of the Democratic Center) would extend its soft hegemony over Spanish politics? Would they become dated, or should they


maintain relevance? Different political forces had different expectations. Retrospectively, the outcome proved that the política de consenso had succeeded and, by succeeding, had made its painstaking pursuit no longer necessary. And the passage to a majoritarian style of government demanded not constitutional rethinking, but an adaptation of the democratic agreement to the improved circumstances.

A third possibility, exemplified by Italy, is to improve on an imperfect and uncertain agreement by learning to live with its defects, preserving uncertainty in fact and using it for coexistence. In postwar Italy the terms of the constitutional agreement were actually couched broadly so as to increase their attractiveness and the collective interest in entering the democratic game. Heavy on garantismo but otherwise sketchy and noncommittal on the touchier matter of governance, the agreement could still have evolved as it did in Spain: in a majoritarian direction. But although it alleviated the effects produced by reluctance of the extremes, garantismo could not fully remove, contrary to Spain, reluctance itself. Thus, when the Communist party emerged as the second political force in the country, the response to the apparent stalemate (an intriguing response in its beneficial effects on coexistence) has been to accept the uncertainty and fuzziness of the original agreement, to agree to disagree on its terms, and to continue to argue—in particular on the roles of government and opposition. The result is a continuous stretching of the agreement that, while ostensibly unsettling, leaves nobody out.

Trying to clarify and pin down the terms of a democratic agreement once and for all, or trying to alter them, may have costs that are not always palatable or necessary. Somebody could be required to make sacrifices that were previously buried in the agreement. In addition, as happened with the sudden emergence of the Italian Communist party as the


main opposition, the clarification may be impeded by a stalemated balance of forces. Instead, if properly employed, a fuzzy agreement has merits. Learning to live with it means giving the reasons for its fuzziness a political dignity of their own. It means recognizing the essential role of all actors, even if they entered the agreement from different perspectives, and adjusting the role to circumstances. It means calling for a degree of muddling through and accommodation in the institutional implementation of the agreement, and in performance, that makes that recognition operative. And it means making room for political minorities and the opposition—when it comes, for instance, to their control and influence upon government—that can compensate for such costs of a political stalemate as the opposition's inability to head the government. All of this is of substantial constitutional value because, given their initial subjective and objective imperfection, the agreement and its implementation invite at the same time a more demanding and continuous testing by political actors.

I have dealt at greater length with the third way of improving upon the original agreement for one good reason. Learning to live with an imperfect agreement may seem a peculiar method of improvement. Yet, more than the other ways, it reveals how pervasive and effective readiness to adjust, and to coexist in diversity, can be. This readiness, often invisible to the skeptical analyst, makes muddling through an effective fallback position when more direct and explicit ways of reworking the agreement fail. Thus, living with an imperfect agreement may not yet constitute a consensus on fundamentals, but, because it makes room for political actors with contrasting expectations, it is a useful approximation of it. Need I add that, even in some well-established democracies, what often passes for a consensus on fundamentals is, in point of fact, only an approximation?[21]


The thrust of this section, and of the entire chapter, is that reaching democratic agreement is sufficient to usher in a fruitful period of implementation and institutionalization, with all that the period holds in store for democratic stabilization. But I have asserted repeatedly that the agreement is a watershed in another sense. It is also necessary: without it, democratic institutions lead a meager existence, democratic life cannot resume in full, and in sum, to put it starkly, a democracy does not yet exist. It is time to conclude with this second aspect. Showing the cramping effects of its absence is another way of showing the potentials of a democratic agreement.


There have been many cases in which single democratic institutions and procedures have been adopted and tested after an authoritarian crisis but before a collective democratic agreement is struck and before breakdown games have lost significance. Some of the best examples come from Central American countries like Guatemala and El Salvador, though what I have to say might also apply to a number of countries in South America. For a number of years now, moderate forces in these countries have experimented, even in earnestness and not without bravery, with free elections, parliaments, party competition, and other standard features of political democracy. Yet the problem of the extremes, reluctant to join the democratic fray to say the least, remains unresolved, and breakdown games continue to condition democratic governance. In the absence of a collective agreement, democratic institutions are weakened by dissenters and the uncommitted and are unable to emerge as the sites for processing conflict.


The institutional feebleness of these would-be democracies has deeper historical and structural reasons, but we may also understand it contextually with regard to the circumstances that accompany and motivate the appeal to democratic institutions. Such institutions are introduced into would-be democracies, not so much to fulfill and crown a collective constitutional choice as to serve as a strategic tool, a weapon to employ, during periods of uncertainty and protracted transitions, in the contest between dissenting political forces. For example, elections ostensibly held within democracy are really about democracy. They are favored by some political actors, who call them at the sufferance or against the resistance of other actors. Assuming that democratic intentions are present, assuming that elections are revived to preempt, in a fuite en avant, the prejudice against coexistence and to force the issue, it is still intuitively the case that even a string of elections and a string of parliaments will find preempting the prejudice exceedingly difficult.[22]

True, we have seen earlier that calling for elections means that civilian political parties should at last emerge as the central political actors; that their attention should shift to the more orderly and constructive business of building diverse national support and defining or implementing the rules of competition; that even a prospective opposition should have an added incentive to sacrifice the support of more radical groups and reluctant players in the interest of securing wide representation from the start. All of this is eminently sensible, but it is based on an insufficiently stressed assumption. The assumption is that, in one way or another, explicitly or implicitly, willy-nilly, the significant parties have already come to an understanding, before entering the elections, that the electoral contest will offer tolerable chances of representation to all and that the newly elected body will act to constitutionalize


the rules of contestation. Otherwise, if a previous understanding does not exist, the practice of elections alone has difficulties in building that understanding over time.

Chapter VIII will consider the circumstances that can nevertheless bring about, even in such difficult cases, the missing understanding—perhaps with the very help of still-marred elections. The task of these conclusions was only to reiterate the importance of reaching an understanding.


previous chapter
Chapter VI Beyond Transitions Why Democracy Can Deliver on Its Promises
next chapter