The Transition to Democracy: Proceedings of a Workshop (1991)

Why do we want democracy? I can give you one of many reasons: this finding is absolutely extraordinary, but democracies do not fight one another. Democracies are not more peaceful than other political entities--it is just that they do not seem to fight one another. One could, in fact, argue that since 1816, no democracy has fought another democracy. In order to make that statement, it is necessary to presume that Germany under the Kaiser in World War I was a monarchy, that Germany under Hitler was a dictatorship, and that Lebanon in 1967 was a military government. However, if one accepts these presumptions, then one could say conclusively that since 1816 not a single democracy has fought a war against another democracy.

This finding is quite a recent discovery about wars among nations, and it is backed up by anthropological data on societies that have no political organization beyond the local community. If you look at the data from the Yale Human Relations Area Files on 186 societies, you find that the more people in individual communities within a society participate in community decisions, the less fighting there is among communities in that society. Also, the easier it is to remove community leaders, the less fighting there is among those communities. These Yale scholars get correlations of .7 between democracy and an absence of fighting, which is a very strong relationship in anthropological data. But nobody has much idea why this relationship exists.

This relatively recent discovery is a serendipitous stumbling onto a very strong relationship. When it was first discovered about 10 years ago, it sent everybody into a tizzy; nobody could believe it. Researchers have tried to figure it out and, as yet, they have not gotten very far. The Yale anthropological data are as yet unpublished; the scholars working on it are only halfway through their analysis.

Because we do not know why this relationship holds, we cannot ask what the crucial institutions are in a democracy that produce this result. What is it about democracy that leads to what I presume is for most of us

this desired end? We should be asking what specific kinds of things exist in a democracy that get us what we want. This may be a more helpful way of thinking about democracy than the metaphysical question of “What is democracy?” It may be better to ask, “What in democracy brings about the ends we want?” rather than “What is it in some platonic, ideal world?”

What is this second, equally or more important practical reason for wanting democracy? Because democracy embodies widely held ideals--freedom of expression, global equality--and because it meets deeply felt needs--such as the need to be part of larger decisions contributing to one's life, and the need to be listened to--democracy is able to produce peaceful, legitimate decisions about a larger number of matters that might otherwise end in disruptive conflict.

You could say that democracy helps us to lose peacefully. Whenever there is a conflict, it is likely that somebody is going to lose. Sometimes both parties lose, both parties give something up. But to the extent that democracy is about conflict, which it often is, it is about losing. A good democracy will help people lose well, and losing well, to my mind, includes losing peacefully.

Because we have known for a long time that democracy produces the peaceful resolution of conflict within nations, we are further along in identifying the features in democracy that help produce this end. Two of those features are fairness and participation. Even convicted criminals support the system that convicts them when they believe the process to be fair. If you try to measure support among criminals for the criminal justice system in the United States, you find that support is much higher among people who thought the system was fair than among people who simply got off. Sometimes criminals will even say, “The system's no good because it let me off.” Part of democracy's usefulness is that we can accept loss if we think losing is fair.

As for participation, let us look at the criminal justice system again. Americans prefer informal, out-of-court procedures to formal legal ones. Why? Because in the informal procedures they get to have their say instead of a lawyer speaking for them. Even when the case goes against them, they are more satisfied because they have been heard. They have had a voice. Management studies also show that when employees participate in making a tough or unpopular decision, they are more likely to accept the results. When management makes the decision alone, employees are more likely to quit or to call in sick.

Participation works this way for a while, even in cases of pseudo-participation where management has rigged all the numbers so that even after employee participation, management gets the answer that they wanted all along. But people learn. Citizens in the Soviet Union and Yugoslavia do not have much enthusiasm for the idea of participation anymore,

because they are used to it being rigged. But genuine participation and fair procedures can legitimate hard decisions. It can help people live with losing.

It is largely the ideals of democracy that help us lose gracefully. We say to ourselves that although we have lost, the fairness of the process fits our ideal of what democracy should be.

But we have, I believe, two ideals of democracy: one based on conflict, the other on commonality. And these two ideals are somewhat in tension; to some degree they are even contradictory. We need both, but we need to realize they are conflicting ideals. In fact, democracy as it is practiced in the United States and Western Europe today is a hybrid of these two ideals. In one vision of democracy, the system creates fair procedures for resolving conflicts of interest. In another vision, somewhat in tension with the first, the system encourages deliberation about how best to promote the common good. The intellectuals in the newly democratizing nations of Central Europe recognize this tension better than we do. As East Germany was democratizing, the New York Times reported a growing gap between “dissidents who had formed groups like New Forum with a notion of democracy as a process of well-meaning discussion in which the universal good was the shared goal, and political parties in West and East Germany whose primary goal was to win the elections.”

Practicing politicians in America and Western Europe tend to understand democracy in only one of these two ways, the conflict-based way. For them, democracy is an adversary system that assumes conflicting interests. The system sets up fair procedural rules under which each side attempts to win. Political scientists in the Western democracies also describe politics in this way, as who gets what in a fair fight.

But American and European philosophers who discuss democracy usually emphasize its deliberative character. In deliberative democracy, citizens talk with one another about public problems. Their talk can certainly be raucous and full of conflict; it can turn on opinion as well as fact; it can draw on emotion as well as reason. But the talk in deliberative democracy often aims at a common good. “How can we work out our disagreements?” “How can we get this thing that we all want done, done efficiently?” Every manager is familiar with this kind of deliberation. It works through persuasion, not power. And in fact, politicians in the United States practice not only the politics of power, but also the politics of persuasion for the common good in their legislative roles. Recent research in political science is uncovering the extent of this “common interest” behavior, even among politicians who, when asked directly, would probably deny it because they want to be “realists” and to see themselves as strong actors in a conflictual setting.

Believe it or not, politicians are often deliberating with the common

good as an end. It is true that rhetoric invoking the common good often masks self-interest, or the interest of a particular group or locality. But realism that stresses self-interest cannot explain how the concern with the common good--which acts as a glue for democracy--came to exist and prevail in some contexts.

When we promote democracy in other countries, we must be careful not to duplicate only the West's highly visible adversary institutions premised on conflict and designed to aggregate or sum individual preferences. We must not ignore our less visible but no less real deliberative institutions. We must encourage others to find within their own cultures traditions that may encourage a quality of citizen deliberation that surpasses that in the West.

The newly democratizing nations thus have two tasks: they must act quickly to foster the aggregative institutions that settle issues of fundamental conflict fairly on the basis of “one person one vote, ” but they also must provide what is not so common in the West, extensive forums for deliberation in which citizens have a voice in determining the common good.

To legitimate the very hard decisions that it will have to make, any newly democratizing government must first protect its new aggregative institutions from the usual forms of corruption: bribes, stuffed ballot boxes, intimidation, intentional miscounts. We know how to protect the electoral process fairly well, even though sometimes we cannot do it. We know about multiparty monitoring of elections, neutral investigative commissions, and punishments for infractions that are fast and strong. These protections help maintain confidence that the adversary procedure is fair. Such confidence is absolutely critical in a country's first elections. Faith in the electoral process is built on such confidence.

Another aspect of legitimacy in adversary democracy is more problematic. In this conflictual, counting, summing, aggregative democracy, legitimacy rests on the proposition that each citizen should count for one, and none for more than one. But, of course, every democracy admits to gross inequalities in power derived from unequal, often vastly unequal, economic and social resources. As a result, citizens on the bottom of the socioeconomic scale often feel that, as one survey in the United States put it, “people like me do not have any say about what the government does.”

In every country on the globe, citizens' political resources differ dramatically. Democratic institutions in newly democratizing nations do not have the same force of tradition behind them that is present in most Western democracies. If the newly democratizing nations cannot create institutions that consciously guard against excessive power among their new elites, if they cannot find ways to spread power, they may find the legitimacy of their decisions severely undermined.

In terms of adversary procedural fairness, we must also realize that majority rule works only in polities with many cross-cutting cleavages. Cross-cutting cleavages exist if, for example, in my relationship to you I am on your side on some issues, but against you on other issues. When majority rule results in certain groups being outvoted again and again on almost all their major points of interest, majority rule democracy will not work. It needs corrective measures, such as proportional representation, or federalism, or something called “corporate federalism,” which is devolving power to nonterritorial subgroups to legislate on matters that involve only them, or something that some political scientists have called “consociationalism,” which is dividing power and state-provided goods like school and television time in proportion to each group 's percentage of the population. Those are absolutely necessary correctives if you have a polity that is segmented and lacks cross-cutting cleavages. If you plan to use straight majority rule, you need a situation in which some may be in the minority this time, but in the majority next time, and then in the minority again, so they can feel “Well, win one, lose one.” “Lose one, lose one, lose one, lose one, lose one, and lose one” does not work. Consociational, federal, and other supplements to majority rule still do not provide equal satisfaction to ethnic and other minority groups, but they work better than winner-take-all majority rule.

Adversary democracy creates winners and losers, and therefore combines quite badly with state socialism where there is only one arena in which to lose, since the state runs everything. As state socialist systems begin to adopt adversary democratic procedures, they will need diversified political and economic systems. They will need diversified political systems so that if you lose in one arena you can turn to another arena. They will need a diversified economic system so that “apparatchiks” who lose in politics can become “entrepreneurchiks”--as they are called these days-who may win in economics.

Moving toward adversary democracy means injecting large amounts of risk into previously risk-averse systems. The new democratizing governments will have to learn to live with uncertainty. Accepting uncertainty, losing control over outcomes, and being unable to guarantee the protection of one's personal interests will require an ideological, political, and psychological breakthrough for many citizens and bureaucrats. We all try, above all else, to guarantee the little bit of security we have. Adversary democracy, where you can win one day and lose the next day, means losing that security. People who have had this security through government are not going to be very enthusiastic about losing it.

Institutionalizing continual conflict also requires tolerance of opposition parties. In cultures that are unfamiliar with the peaceful resolution of conflict, it is hard not to see one's opponents as traitors to

the state. If most people with talent and administrative experience have collaborated in some way with an old pre-democratic system, the impulse to blame may breed a rhetoric of character assassination that is bound very quickly to erode the citizens' trust in any existing system of representation. Citizens cannot be weaned from cynicism easily after decades of “facade politics” in which elites determined public policy behind a front of supposedly democratic institutions. And I refer here not only to the countries with which we are familiar, but to our own United States as well. Hungarian voters have already grown jaded. “All they do is make promises,” says one. “Those advertisements on television, it is like a cabaret, I do not believe any of them.” This is a cynicism born of facade politics.

To counter that legacy of pervasive cynicism, Western forms of aggregation through representation may have to be supplemented. The mostly symbolic device of recall is important because citizens can remove their representatives from the legislature. There are other participatory institutions such as national and local referenda. Most important, decentralizing decisions to the lowest possible level, instituting elections and referenda in schools, workplaces, villages, cities, and counties would provide experience in accepting conflict. As those who run in local elections and those who vote for them learn to lose on some issues but win on others, they should become more able to understand and bear losing nationally.

These procedural methods of adversary democracy are necessary to produce legitimate decisions and conditions of conflict. But they are insufficient to generate the individual transcendence of self-interest that hard decisions often require. Adversary democracy encourages the participants to aim at winning rather than finding a course of action that is best for the whole. It discourages listening and lends itself to short time horizons. Like an economic market, adversary democracy legitimates the pursuit of self-interest. Voters pursue their individual interests by making demands on the political system in proportion to the intensity of their feelings. And politicians pursue their own interest by adopting policies that buy them as many votes as possible.

This system of politics as a marketplace ensures accountability if it works properly, but it also mirrors, and perhaps encourages, a larger materialism. Candidates and their policies become commodities, selling themselves or being sold. The dynamic of adversary democracy has traditionally made democracies incapable of the kinds of sacrifices that many newly democratizing nations must now ask of their citizens. National unity and national sacrifice for long-run ends have instead often required a strong, even dictatorial, leader. And it would be foolish for us to think that it is just an accident that in many of these cases countries have come

together, sometimes enthusiastically, under strong dictatorial leaders. They know that this is one way to produce the kind of unity and sacrifice needed for long-run ends.

But citizens will sacrifice even their lives when they believe their sacrifices are for the common good. That belief can arise not just from devotion to a charismatic leader, but also from faith in policies arrived at through deliberation that command the loyalty of those who participate in creating them. For example, throughout their past struggles, many dissident groups in Eastern Europe held together through institutions that fostered a common commitment to the national good. One of my American friends came back very surprised from talking to members of Solidarity, and told me, “The decision rule there is what is good for Poland.” Much of Solidarity, in fact, operated by de facto consensus, making decisions only after the members had worked their way through a deliberative process that tried to encompass widely different points of view. The experience produced unity in the struggle, widespread practical understanding of how to take many interests into account, and consequent willingness to live with the results of decisions.

This bottom-up practice in deliberative democracy may give Poland an edge over the other newly democratizing nations in the use of democracy to make hard decisions. Now Poland has entered into a more classic adversary process and we will see how they play out the tensions between their earlier deliberative process and their new electoral adversary process.

I would argue that whenever possible, participatory institutions should bring together citizens of opposing views in circumstances that reward mutual understanding and the accurate gathering of information. Deliberation among intellectuals, or even elected representatives, is not enough. In the United States theorists have proposed things such as referenda that require two distinct votes separated by a period of deliberation. The first vote would indicate that you favor or oppose a policy on a scale from one to ten, with space to record the various different reasons, followed 6 months later by a second, plain “yes” or “no” vote. The two-stage process would promote deliberation. Other ideas are workplace assemblies, or “policy juries,” where a representative sample of citizens meets with experts in the same manner as elected representatives would do, and comes out with policy recommendations that then inform the elected representatives. There are many other institutional means of nourishing deliberation at the citizen roots.

Obviously, each nation must work out the deliberative innovations and the mix of adversary and deliberative institutions that fit its own patterns of cleavage, history, and culture. What we need to do in each country is find a successful indigenous democratic institution and document how it works in that culture. How do the people handle their conflicts? What

are their strategies? Grass roots democracy is essential for learning how to lose, but it must be a grass roots democracy that works, that solves conflict in a way that leaves losers somewhat satisfied.

In a grass roots democracy you learn not only how to lose and how to listen to one another, you can also learn how to move from the deliberative institutions appropriate to moments of commonality to the adversary institutions appropriate to moments of conflict and back again. In the long run, deliberative processes may offer the best hope of finding ways to handle not only the class conflicts, but also the ethnic disputes that threaten to split several of the newly democratizing nations in Central Europe. While consociational and federal solutions can produce reasonably just allocations among groups, shifting citizen perspectives from class or ethnic interest to a long-run common good requires transformations of self that deliberative processes make possible.

My presentation today is about the interplay between economic and political openings and how they operate in a nation's transition to democracy. When I speak about political opening, I mean political democracy. What I mean by economic opening is not necessarily economic growth, which may result from economic opening, but a process of democracy in economic decision making. The basic theme of this presentation is that the kind of opening to come first will depend largely on national circumstances. I think that any attempt at a general rule about sequencing would lead you down a false path.

The main point here is that economic opening--for example removing bias against exports, allowing the market to make more decisions, and giving the central authorities less power to make decisions--need not lead necessarily to rapid political opening. But if, in fact, the economic opening is successful, I am convinced that the pressure will eventually grow for greater political opening. On the other hand, I am also convinced that political opening will lead quite rapidly to economic opening. To put it differently, a closed economy, dominated by state power, is incompatible in most cases with political freedom of choice. This has some policy implications, which I address later.

It is hardly startling to note that political and economic opening interact, and interact quite strongly in ways that I think are not necessarily predictable in the short term. The likely directions, if not the exact outcomes, are predictable, however. I approach these issues as an economist rather than a political scientist; what Jane Mansbridge talked about

is quite important. By “political opening” I mean such things as the ability of different groups to compete, raise funds, place items on the agenda, have an enlightened understanding of issues, access to media, voting equality at decisive stages, and peaceful transitions of power. I was helped a good deal in my understanding by Robert Dahl's book, Democracy and its Critics. Dahl talks about “polyarchy” rather than democracy, and I have drawn heavily, but not exclusively, on the political discussion in that book for this talk.

Economic opening means the ability to take private initiatives, relative freedom of imports depending on price considerations of various kinds, and a modest role for the state. No one seriously talks about eliminating the state. State enterprises are quite compatible with economic opening, I think, but not if they result in widespread state trading. Perhaps my point would be best conveyed by some examples rather than by definitions. The Soviet Union today is neither politically nor economically open. The United Kingdom is open in both areas. South Korea is mostly open economically, since despite some state intervention there is no great bias in its import or export emphasis, but the country is only slowly approaching political opening. Czechoslovakia is mostly open politically, but very far from being open economically. Again, my hypothesis is that once you are open politically, economic opening will follow. The political opening cannot survive by itself. It is a telling point that while not all market economies are democracies, I can not think of a single country that is a democracy that does not have a market economy.

Let me go through the sequence in different places and propose some general contextual rules. In Eastern Europe and the Soviet Union, political opening came before economic opening. We are seeing the struggle right now in the Soviet Union over what might happen on the economic side. We are seeing it in Czechoslovakia and Hungary as well. Part of the picture is that protest against political suppression demanded some opening in that area first. In Poland, Solidarity represents a political opening, but not yet an economic opening. Eastern Europe is experiencing that sequencing of politics first and experiencing it quite strongly. If you look at East Asia and Latin America, the sequencing is quite different. There the sequencing was economic first, while the political lagged. It lagged in Chile, South Korea, Taiwan, and Mexico. In this last case, the collapse of the economic structure has brought about a profound transformation in the economy, but political opening is lagging.

However, neither opening can lag forever behind the other. In Chile it took 16 years for political democracy to be restored, but the pressure was there. Modest political opening is now taking place in East Asia as well. The Mexican government is deliberately seeking to phase in the political opening slowly until the economic opening breeds results. It is

quite clear from the government's plan that part of the motivation is to hold on to political power in what is seen as an inevitable political opening. The official party does not want to lose power; hence, it is holding back political opening. If any of you have observed Mexico, however, you know that the official party has lost full control of the process. The type of activities once used to win elections are less possible in the new context. Therefore, I am convinced that Mexico will need to move faster--now that it has opened its economy--than South Korea or Taiwan had to move.

Let me now try to give some reasons behind particular examples of sequencing. One important factor is whether the initial impact came from above or from below. In Eastern Europe it came from below, and the political opening occurred first. In East Asia and Latin America, political opening came mostly from above. In Korea and Taiwan, and in Mexico, economics came first. Some of the discussion in A.I.D. 's papers on the Agency's regional democratic initiatives make the point that economic development sometimes takes place under an enlightened dictator. There are some cases where this is true, though they are rare. South Korea would not have developed the way it did without Park as leader. Chile would not have developed the way it did after the Allende regime without the Pinochet government. But there are not too many cases of enlightened and successful dictators. It seems reasonable that the particular sequencing of these cases had real impact on the shape of the outcome.

I think, also, that the degree of political suppression makes a difference. In Eastern Europe, where the political suppression was so total, once the shackles came off, a widespread political opening occurred. In Mexico, where the suppression was not as great, as long as you had economic success, the ruling party could hold back the pace of the political opening for a while. In all these cases, a good deal of the pressure came not just because people wanted democracy, although that is part of it, but because the economic system collapsed around them. That is not the case in South Korea. There you are getting, after a long transition, demands for political opening when the economy is doing quite well. Economic collapse cannot be cited as the start of the political sequence in all cases.

I believe that there is likely to be a transference of some important consequences from one kind of opening to the other. When an economy opens and nongovernmental actors make major decisions, a form of democracy is already instituted. Jane Mansbridge talked about politicians treating themselves as commodities selling ideas. It is possible to look at it another way: once the state gets out of the way, private actors have to make decisions. In the case of Latin American economies, where prices and markets increasingly determine the decisions rather than officials and governmental regulations, decision-making power has been expanded from

the center down to thousands, hundreds of thousands, in some cases millions, of decision makers making those decisions for themselves. And that, too, will have a definite effect on the political structure. A political structure may not necessarily open all the way--it may open only for certain groups within that society, such as the middle class or dominant elites. But I do not think that arrangement is likely to prove highly durable in the long run either.

Let me make a few general conclusions, and then draw some policy implications from what I have said. We are obviously now in a time of both political and economic transitions. What happens in one place is infectious. When the rest of Latin America began to move toward some sort of political opening, there was no stopping this movement in Chile, just as there is no stopping it now in Mexico. By the same token, once Chile demonstrated the success of its model of economic opening, followed by Mexico, the infection spread all over Latin America. The only point at issue is the speed of the transfer from one place to another.

I believe that the speed from political to economic opening is almost always more rapid. However, moving from economic to political opening depends on the context, the tradition, the history, the degree of political suppression that previously existed, the general level of education, and a variety of such factors.

What kind of policy implications can we draw from this? What should the U.S. government be supporting? What should A.I.D. be supporting? The first piece of advice is to conclude that any approach has to be country-specific and must depend on the context of what is going on in that country. Second, you can encourage democracy not only by directly encouraging political democracy, but you can also promote democracy by encouraging private decision making in the economic sphere. Indeed, in many countries that may be the best opening that A.I.D. has. I get a little nervous when I read in the newspapers that A.I.D. is preparing the type of democratic conditionality that must be imposed through U.S. foreign aid. I see that discussion coming up over and over again. I do not object to the conditionality; if it will work, go ahead and impose conditions, but if it is going to be counterproductive, it may actually slow down the process. On the other hand, if democratic economic conditions are imposed, it will, in a slow, progressive way, also be imposing the political conditions. The economic conditions may be within A.I.D.'s power to impose, while the political conditions may not.

I think you will find that the conditions for what I am talking about are now extremely good in Latin America because of shifts in both economic and political openings that are taking place in almost every country in the Western hemisphere. The politics have grown far more fragile than the economic opening, and therefore, I would advise A.I.D. to focus much

of its program on economic opening, as this is where the leverage will be greatest. Leverage is less great, I suspect, in Africa because of the different level of economic conditions that exist along with the greater problems of ethnic differences (compared to Latin America as a whole).

My final point is that it will be difficult in most cases, if not impossible, to encourage simultaneous opening. Push for them, but push hardest for that opening where it is clear from the country analysis that leverage is likely to be greatest.

I have been asked to address the question of what is a democracy from a societal and perhaps a sociological perspective. I propose to do so by juxtaposing two essentially contested concepts, that of democracy and that of civil society. I do not have a lot to say about democracy, thanks to Jane Mansbridge, who laid out the fundamental parameters of what we mean in her presentation. By “civil society ” I mean the presence of intermediary organizations and arrangements that lie between the primary units of society--individuals, families, extended families, clans, ethnic groups of various kinds, village units--and the ruling collective institutions and agencies of the society. All the key properties of civil society hinge on the presence of these intermediaries between the primary units and what could be called the tertiary units of the systems, that is: the governing institutions, the all-embracing and coercive institutions of the society as a whole.

It is doubtful whether either democracy or civil society can be attained, and especially sustained, without the presence of the other, although the sequencing of one with regard to the other seems to provide certain important clues for explaining the long-term outcome. Civil society seems to be a necessary, but alas, not a sufficient condition for the presence of stable democracy. That, I think, is about as general a hypothesis as one is likely to find in the sociological literature.

Inversely, without at least some even unstable democratic practices, civil society is unlikely to persist. Eastern Europe and Chile are the cases that have taught us that it is possible for elements of civil society to survive, in some cases even to flourish, under protracted autocratic and even totalitarian rule, and hence, to precede the advent of democracy. So civil society is likely to have some precedence that may play a causal role in the process of democratization. For example, we observed in a comparative study of Southern Europe and Latin America that it is frequently only after the previous authoritarian or autocratic regime has begun to

“liberalize,” that is to say to loosen up its most arbitrary and illiberal procedures, that a resurgence or revival of civil society takes place. It precedes democracy. It is the process or the phenomenon that links liberalization, the mere loosening of arbitrary aspects of autocratic rule, and pushes it into a democratic outcome. But the existence of civil society is not the cause of liberalization in and of itself.

In Eastern Europe the concept of civil society serves a double function. It indicates a somewhat elusive set of “theoretical” or “abstract” conditions that seem to be necessary but not sufficient for democracy to flourish. That seems to be an accepted central postulate of the Eastern European experience. But the concept also serves a second, much more obviously political function. The concept of civil society that is used in Eastern Europe also identifies a set of more concrete historical properties that serve to differentiate European from non- or less-European developmental patterns. In other words, for Eastern Europeans, it resolves in the mind of its users the rather difficult issue of delimiting the eastern border of European civilization. Poles, Hungarians, Czechs, and Balts in particular, are sure they are on the “right,” that is, the “civil society” side of that divide. They know that Russians are not and suspect that they will never be. And they are suspicious about their neighbors Romania and Bulgaria. Unfortunately for Yugoslavia, it is possible that the boundary of civil society runs through the middle of the country.

When we talk about civil society and democracy, we must try to grasp both what is being generically referred to and the diversity of possible types within those generic labels. My major hypothesis is that different types of civil society will be, in the long run, associated with different types of democracy. This presentation, in combination with Terry Karl's presentation about types of transition, lays out a fundamental conceptual map of the problem. In other words, there are different types of civil society, and these civil societies undergo differing types of transition, often due to fortuitous events, including contagion from an adjacent country and even from one region of the world to another. Finally, as a dependent variable, one has types of democracy. My central theme is that the countries of Latin America, Southern Europe, Eastern Europe, Asia, Africa, and perhaps even the Middle East are not just undergoing a transition to “democracy,” but that they are undergoing a transition toward various and different types of democracy. The nature of their civil societies, I believe, is one of the key determinants of the likely outcome.

The intermediary organizations and arrangements that comprise civil society have three general characteristics. The first is autonomy from both the state and from primary groups. In Eastern Europe, the emphasis is obviously on autonomy from the state. Elsewhere, in Latin America, Asia, and especially Africa, the main problem is often the autonomy of civil

society from primary groups, from family groups, or from what we would call in the jargon of various social sciences “clientelistic ” relationships that link them to primary groups.

The second characteristic is that the units of civil society have some degree of organizational continuity. They have a “corporateness ” that rests on rights and entitlements protected by the state; hence there is no such thing as civil society without the state. It requires state recognition and protection of that corporate status and also the emergence of explicit tolerance between the units of civil society.

There are two basic hypotheses regarding the limits of organizational continuity, especially affecting the central problems of mutual recognition among competing units within civil society and recognition by the state. The first is the problem of social inequality. Is it possible to sustain a civil society when there are gross inequalities, either based simply on material distribution of rewards, or on traditional distinctions of caste or race within the society? With how much inequality between the primary units of the society is it possible for there to be a civil society? The reason some Eastern Asian or Asian societies may have very substantial advantages, not merely economically, but also politically, is the previous existence of land reform in these countries that has reduced some of the grotesque inequalities one tends to find, for example, in Latin America.

The second major hypothesis that comes out of the work of Barrington Moore and others, is that it is very difficult, and one is tempted to say impossible, to imagine a civil society in which coercive force is a major element in the constitution of the productive units of the society. This is particularly true with regard to agriculture. If you have an agricultural system based on semi-serfdom, not to mention slavery, the possibility of developing norms of reciprocal tolerance between competing interest groups or competing intermediaries seems to be severely limited. We do not know exactly what the thresholds for either of these hypotheses are, but we think we know that two factors that contribute to the development of civil society are: (1) the elimination of grotesque inequalities, and (2) the elimination of coercion in production.

Another characteristic of civil society is the capacity for self-governance. These intermediary organizations are political units that more than just aggregate the preferences of their members. They not only represent their interests and make collective demands on others, especially on the state, but they should also be capable of controlling and governing the behavior of their own members. In other words, if you have a civil society, you have units of private governance and of private implementation of policy. This has very important implications for a range of possible developmental policies within those societies.

What are the major types of civil society? In the theoretical literature,

some of which is very old, people have identified the emergence of the notion of civil society with developments within the Church around the years 1000 to 1200. It is a very Eurocentric conception, but there are two major themes in this literature. One emerges out of the Scottish Enlightenment and is best expressed by Adam Smith. The author or originator is in fact a man named Adam Ferguson who wrote the first book on the history of civil society. I call this the “Anglo-American” theme, and I think it is cultural as well as geographic. This is a liberal conception of civil society in which the intermediary units are essentially voluntary associations of individuals. They are quite similar to market forces; people choose the intermediaries that by personal will or interest they prefer.

Opposed to that idea is a concept, much more associated with Hegel, and eventually Marx and Durkheim, and a number of continental European thinkers, that I will call the “corporatist” conception of civil society. It stresses collective units that are frequently involuntary. The locus classicus for this is the guild in European cities, particularly continental cities. These are units created and very often sustained by the political authority of the state, that individuals do not choose to join, and that have an involuntary or semi-voluntary, in many cases an outright compulsory, nature. People are either born into them as sons of guild members or must become members in order to practice various occupations.

In the first conception, the idea is that, with the development of differentiated social and occupational structures, multiple, overlapping, and dispersed units will emerge spontaneously from the civil society to compete with each other in highly specialized categories of self-interest. The second conception emphasizes the emergence of singular, monopolistic, hierarchically-ordered organizations that usually emerge in collusion with the state to structure this intermediary space. The code word in political science jargon for the former is “pluralism,” and the code word for the latter is “corporatism.” The United States, Canada, the United Kingdom, and in a rather different way, Italy and France, are frequently cited as those with more pluralistic, overlapping, multiple structures. The Scandinavian countries, Austria, Germany, certainly Switzerland, and, interestingly enough, contemporary Spain are countries that have adopted, or rather conformed to, the second model. Both, obviously are ideal types and of course there are mixtures in all of these societies.

The main underlying message in terms of individual countries is not to attempt to force upon a given country a mode or conception of civil society that is antithetic to how its basic institutions have emerged. A very interesting example of this comes out of the American occupation of Germany. The Americans arrived to occupy Germany and discovered a corporatist civil society. They mistakenly thought it was Nazi and therefore tried to dismantle it. It took a while before they discovered that

the roots of these institutions were several hundred years older than the Nazi regime, although the Nazis did a magnificent job of taking them over and using them for their own purposes. Here was a classic confrontation between a well-intentioned, zealous occupying power determined to bring pluralism and democracy to a country that was at its mercy. The end result was that the Americans were unsuccessful and these German roots of corporate civil society took precedence. The Federal Republic is very different from the United States if you compare their interest group structure.

The central problem is whether, and how far, we can take this Eurocentric conception of what we suspect to be a requisite for democracy and apply it outside the European area. We have seen that it serves a convenient function in Eastern Europe in distinguishing the visible, or barely visible, Eastern European frontier. What will this concept have to take on to provide the same “functional equivalent” for stable democracy in sites that are far removed from those in which it originated?

Democracy is obviously a capacious concept that seems at times almost formless and certainly contentless. It has been over-conceptualized, misunderstood, and “under-understood.” In the past there has been an incredible proliferation of suspicious adjectives stuck in front of it: guided democracy, tutelary democracy, popular democracy, people's democracy, unitary democracy, consensual democracy, even African democracy, Latin American democracy, and Asian democracy. Usually these have been very thinly-disguised attempts to justify something that was not at all or only remotely democratic. The interesting thing about discussions now is that those adjectives have disappeared. That seems to be absolutely central. I think there is a rather extraordinary consensus about what are called the threshold conditions or the minimal conditions of democracy.

Another thing that has disappeared is something that Europeans had the luxury of pursuing throughout the nineteenth century, what could be called “partial” democracy. One must not forget that Europeans practiced democracy in this somewhat cautious manner and at times had notions that are much more restrictive, particularly of the definition of the eligible citizenry. The French had a term, démocratie cencitaire, that meant democracies that were limited to taxpayers. You had wonderful democracies like Great Britain with 2 to 5 percent of the population eligible to vote. At the time nobody argued that this was undemocratic, and eventually the percentage of voters increased over time. There was also a French term, démocratie capacitaire, that referred to a democracy in which you became eligible as a citizen once you became literate or met various other criteria.

These are not options available to contemporary democracies. They cannot just say that only people over 40 years old will be allowed to vote,

or only men, or any of those other criteria that Europeans manipulated, especially in the nineteenth century. Today, crossing that threshold involves the elimination of a wide range of restrictions that Europeans once practiced. There is a Mexican social thinker, who put it very well in a book advocating “Democracia sin adjectivos,” democracy without adjectives. He was referring to some rather unpleasant practices of the Mexican regime that put not just adjectives, but unsavory practices in front to limit the possible uncertainties of outcome.

I do not have time here to go into the factors and conditions that are discussed in the paper that Terry Karl and I have written. I think that there is broad consensus on seven criteria defining democracy set out by Robert Dahl in his book Dilemmas of Pluralist Democracy:

1. Control over government decisions about policy is constitutionally vested in elected officials.

2. Elected officials are chosen in frequent and fairly conducted elections in which coercion is comparatively uncommon.

3. Practically all adults have the right to vote in the election of officials.

4. Practically all adults have the right to run for elective offices in the government.

5. Citizens have a right to express themselves without the danger of severe punishment on political matters broadly defined.

6. Citizens have a right to seek out alternative sources of information. Moreover, alternative sources of information exist and are protected by law.

7. Citizens also have the right to form relatively independent associations or organizations, including independent political parties and interest groups.

We have added two other conditions to correct one of the problems we find with discussions about the criteria of democracy, namely the concentration on the institutions of democracy itself without regard to the international and broader national context in which it is set. First, you cannot have a democracy in a country that does not control to some significant degree the content and deliberation of its collective decisions. One could have only quasi-democracies in colonies in which the outside colonial power controls the basic parameters and leaves the “natives” to deliberate and to decide minor points after the colonial power has fixed the essential ones.

Second, most definitions of democracy do not pay much attention to what the Spaniards like to call los poderes fácticos: the military, the civil service, the church, the various kinds of institutions that may condition the

possible range of deliberation and the possible range of outcomes. Clearly an adequate definition of democracy implies that the minimal procedural rules of fairness that Jane Mansbridge referred to are respected, but also that they are not conditioned by, or limited to, those spheres that the military or some other socioeconomic institutions will tolerate.

There now is relative agreement on the defining conditions of democracy. The adjectives have disappeared and, at least in terms of the definitions of democracy that emphasize procedure rather than deliberation, I think there is a fairly substantial agreement on what they are. Then the question becomes twofold: first, will democracy get over that threshold? Will those conditions be consolidated? Even more interesting in terms of my present research, what type of democracy can one expect to emerge?

Let me say something about consolidation simply to lay out the alternatives, because I think one of them is unfortunately not recognized enough in the literature. The most probable outcome, if you simply project previous experiences into the future, would be reversion to autocracy. If you simply look at the data and mindlessly say that there is no change in these countries, and the probability of Latin America remaining democratic is the same today as it was in the 1950s and 1960s, you feel pretty hopeless. From such a narrow, positivistic perspective, you have to predict probable reversion to autocracy. There are a few countries that have done this practically like clockwork; Turkey, for example, was on a ten-year cycle that you can almost get down to the month. Bolivia was another case, as was Ecuador. Obviously, if you take the past as your example, that is the probable outcome.

Second, I tend to discard, although it is probably important for some in terms of their immediate situations, the persistence of some sort of political hybrid that does not cross the minimal threshold, like the various restricted democracies that the Europeans practiced in the past. In a book I co-authored with Guillermo O'Donnell, we stole one term and invented another, to refer to these hybrids. We called one “dictablanda”: “soft” dictatorship or liberalized authoritarian rule. And we invented the term “democradura” or “hard” democracies. These are democracies in which the military, the civil service, or whatever the previous ruling power was, severely control such things as access to the ballot box or the agenda of public choice. For us, these are interim forms. It is very unlikely in the present context that this will be a stable, self-reproducing form of government.

The one that unfortunately looks persistent is the possibility of protracted unconsolidated democracy. Some countries are likely to be condemned to democracy without being able to enjoy it. They are condemned to democracy because the alternative forms of domination are so utterly discredited that they are simply not available given the current

distribution of values and power. But these societies cannot, or have not yet been able to, come up with those famous rules of fairness that Jane Mansbridge referred to, that is, with mutually acceptable conditions for practicing what we call “contingent consent” as the central feature of any viable democracy.

The country that jumps into my mind every time people start this discussion is poor Argentina. And in fact, in European jargon, people in Poland and Hungary talk about the dangers of the “Argentinization ” of their transition or consolidation. It is possible that some of these countries now in transition will get over the threshold, but the country will still be a mess. It does not have consensual rule. People do not settle into the routine of an adversarial democracy and they certainly never get around to very much of a deliberative one.

Finally, there is consolidated democracy. The important point here is what type. What I offer you as a first approximation is the “ property space” for understanding types of democracy. The literature in political science on types of democracy is generally quite unsatisfactory because it focuses on single types and does not really try to lay out the full range of possibilities. It seems to me that there are two abstract properties to consider in charting the types of democracy. The first is something that Jane Mansbridge stressed: the dominant principle of aggregation or decision making rule. At one end, you have majoritarianism. The idea here is that democracy is a system that relies on equal and fair counting of votes, whether this is the electorate, or the parliament, or the committee room. The inverse, which is much more practiced in Europe, is a form of democracy that Americans might not even recognize, in which you weight the intensities of citizens ' preferences rather than simply count their equal votes. Switzerland would be a model of this. Voting makes virtually no difference; as a matter of fact, the Swiss are just as bad about not voting as the Americans. Switzerland is also the only country that I know of in which the turnout is greater for local elections than for national elections. The Swiss are not stupid; the only place where their vote counts is in communal elections, it counts less at the canton, and virtually not at all at the national, so they do not bother to vote. In Switzerland, it is the intensities that are weighted and aggregated, and that makes Swiss democracy the ultra-stable system that it is.

The other dimension is civil society. What is the balance in the system between the state as a source of initiatives and structuring as opposed to a bottom-up conception of democracy based on the complete predominance of civil society over the state? Each country has a different historical mix. Some countries are simply more statist; France jumps to mind if you are thinking about Europe. Switzerland, the United States, and Great Britain to a certain extent, come to mind as countries that are

fundamentally oriented around the institution of “privatism” and civil society.

Let me conclude with two “bottom lines.” Countries transit to democracy. Countries consolidate into different types of democracy. Moreover, the type of democracy for any given country is, in most cases, likely to be the outcome of a compromise or an extremely complex set of compromises, and it is likely to be the type of democracy nobody wanted in the first place, that is to say, not the original preferences of any of the actors. Early in the transition some may want an ultra-majoritarian form, others may be preoccupied with the protection of minority rights. And if things work out well, and democracy is consolidated, you will get a compromise. Frequently, the outcome is often a second-best solution, a compromise nobody wanted in the first place, but that people are willing to live with and that they subsequently come to define as fair, even though at the beginning they would have all said, “no, that particular set of institutions and rules is unfair.”

The second bottom line concerns the fit between the type of civil society and democracy. You are wasting your time if you try to promote a type of democracy that is fundamentally at odds with the nature of civil society in a given country. I recommend starting with civil society, trying to understand whether there is the possibility for one, and if so what its units are, what the distribution of various kinds of resources across these intermediary organizations is going to be, no matter whether they are unions, business associations, professional groups, or religious brotherhoods.

There is an important distinction between the literature on democracy and the literature on democratization. The literature on democracy fills a library, the literature on democratization fills a shelf. We have libraries full of books about how more or less stable democracies function, reproduce themselves in fairly regular ways, and occasionally change through realigning elections. Sensible ideas about how countries got where they are, even well-established and settled ones, are extremely scarce.

As we work on the problem of democratization, trying to understand the dynamics of becoming a democracy, there is a growing suspicion among many who work not merely in many different countries but even in different areas of the world, that the particular characteristics of institutions within the United States do not provide a viable model for most transitional cases. It is interesting that if you work in countries that are in the midst of a transition, and you talk to people who are making choices, there are two countries to which they are paying much more attention. The one institutional setting that interests people in Eastern Europe, Southern Europe, and even in Latin America the most is the German constitution. There are certain features--I will not call it a model

democracy--about the mix of institutions of the Federal Republic that is very appealing. Moreover, the Germans are out there promoting that model, too, so it is not entirely just a demand phenomenon.

The other country they pay a lot of attention to, and which in some ways provides the dominant model for regime transition in contemporary terms, is Spain. Spain is emerging, in terms of its reputation--and I think it is a deserved one--as the model transition. Latin Americans, Venezuela for example, were pioneers in the use of social pacts. But if you are looking for shortcuts to figure out what people are thinking about, then look into the German constitution and the Spanish transition. If you are looking for a “crash course” in finding out how this relationship between civil society and democracy has worked out and what kind of institutions it is likely to produce, Germany and Spain are the two examples that I recommend.

In the time remaining after formal presentations, the three speakers responded to questions from the chair and from the audience.

Charles Tilly suggested that democracy could be conceptualized either the way one thinks of a skyscraper or as one thinks of the weather. The “skyscraper” model of democracy assumes that the phenomenon of democracy has very clear, recognizable characteristics that vary within certain limited parameters. A skyscraper (or democracy) is easily recognizable, whether in Manhattan, Nairobi, or Cairo; one knows a skyscraper (or a democracy) when one sees it. There are only a limited number of ways to build a skyscraper, and a general set of rules for correct construction can be specified. In addition, there are certain conditions that make a given place unsuited for a skyscraper. Democracy, in this view, can be readily identified and promoted.

At the other end of a spectrum, Tilly proposed an analogy with the weather as a model for grasping the phenomenon of democracy. This model looks more skeptically at the prospects for outside intervention to promote democracy. We can more or less understand the weather and its wide variation in different times and places, perhaps even affect it in small ways. However, the most one can reasonably hope to accomplish is to show some ways of coping with variations; no one seriously talks about promoting good weather. It is possible to regard the development of democracy as similarly unresponsive to outside intervention, as something that develops based on tremendously complex and largely internal factors. From this model, the logical conclusion is that outside intervention makes little sense.

Tilly then attempted to rank the three speakers based on his sense of how each would answer the following questions: “Can democracy reasonably be produced short of transforming everything else in the society? Or, in other words, how realistic is it to expect results from outside intervention?” He interpreted the Weintraub presentation as the most optimistic, since a policy promoting economic opening could lead to demands for political participation, protection of minorities, and other political opportunities. Tilly saw Mansbridge 's comments as the most cautious. Her distinction between adversarial and deliberative democracy raised a number of potential pitfalls for outside agents trying to promote democracy and highlighted the difference between adopting democratic forms and actually producing democratic participation. Tilly placed Schmitter between the other two, as more ambivalent, keenly aware of how the specific history of a given country's civil society affects prospects for and the shape of democracy. At the same time, however, Tilly found an implication that carefully constructed outside intervention could promote traditions of civil society that would in turn promote democracy.

Tilly then ranked the three panelists based on his perception of their varying answers to another question: “To what extent do we believe that there are many different forms of democracy?” If democracy has essentially only one form, it is considerably easier to decide which countries are moving toward democracy and which are not. A variety of interventions could be devised to promote movement toward the goal. If, on the other hand, there are numerous models of democracy, it becomes more difficult even to identify countries approaching democracy, let alone promote democracy as a goal. Tilly commented that Schmitter seemed to propose “one country, one form of democracy; ” that Mansbridge apparently believes in some well-defined conditions recognizable as democratic; and that Weintraub was essentially skeptical, not only of the idea that one universal model could explain democracy everywhere, but also of the idea that democracy truly exists in myriad forms.

Mansbridge responded first to Tilly's comments and clarified a number of points. She largely agreed with Tilly's characterization of her position, but stressed that it was inadvisable to come to a place with a ready-made pattern for democracy and to treat it as “the” plan. For example, she commented that the adversarial institution of elections seemed to tap into a very basic, even “ pan-cultural” understanding of fairness. She noted how people throughout the world would often risk death to vote. However, she cautioned that once new elites assume power after elections, it becomes necessary to find ways of continuing to promote a perception of fairness. Her advocacy of various “deliberative” or “consociational” solutions was meant to address this problem. In response to another question from Tilly, she was less optimistic about using area specialists to come up with

specific programs to promote democracy. She worried that such endeavors could simply produce “thousand-page brochures” for how best to promote democracy in each country. In response to a later question from the audience, Mansbridge stressed that grass roots solutions ought to be favored over state-imposed ones and thought she detected an implicit bias toward state-run solutions in the question.

Weintraub, too, essentially agreed with Tilly's characterizations of his presentation. Noting that his task was to focus on the economic aspects of democracy and A.I.D.'s role in the economic arena, he emphasized that in his view, meaningful political participation of any kind--leaving aside entirely the finer distinction of adversarial versus deliberative democracy--would be very unlikely with a state-dominated economy. As for the likelihood of successful outside efforts, he stated that the United States could definitely exert influence at critical moments, but he expressed doubt about the ability of the United States to determine outcomes.

Schmitter's comments first stressed the tremendous dynamism of civil society and the complexity of mechanisms of collective action. He underscored that he did not mean to convey a static model for civil society; the reality was vibrant and always changing. Civil society responds to a variety of internal and outside forces and is in constant motion. Second, civil society influences the state, but the nature of the state, to a very considerable extent, affects the nature of civil society as well. He expressed skepticism about the ability of outsiders to determine outcomes, stating that the majority of choices are mainly endogenous during transitions. Furthermore, exogenous variables are frequently “endogenized,” that is, focused through the lens of local conditions.

Schmitter commented that to him it was clear that more than one type of democracy exists, but that it is generally accepted that some limited number of basic characteristics can be identified. However, he noted these minimal conditions tend to be procedural and adversarial; there is far less agreement on the common deliberative elements of democracies. He expressed optimism about the ability of countries to learn from one another and noted the existence of numerous “clusters ” of new democracies as evidence of this ability to translate and share experience from country to country. Finally, Schmitter noted that a remarkably common language about democracy is now being shared around the world in areas undergoing quite different transitions to democracy.

One participant commented that the general advice to select intermediary groups with care in order not to force an inappropriate model of civil society onto the host country was good in principle, but extremely unrealistic in practice. She observed that conditions in many countries where A.I.D. works are now in a state of extreme flux. It is often next to impossible to identify fundamental, unchanging societal elements in such

upheaval. Next, she noted a special type of unthinking cultural bias in all support decisions. For instance, as Americans raised on the “pluralist” conception of civil society, we have an inherent tendency to support these types of organizations, regardless of whether a more “corporatist ” model might offer a better fit in the host country. Schmitter agreed with this observation.

Another participant noted that much of the theoretical literature on the importance of “cross-cutting cleavages” in divided societies had not been written for areas as deeply divided as those where A.I.D. is now working. He cited African tribalism, Middle Eastern confessional differences, and Asian ethnic divisions as extremely deep, vertical divisions, and questioned the validity of applying theoretical literature largely written about northern Europe to such cases. Mansbridge largely agreed that it was correct to question the validity of such approaches, but noted the necessity of first attempting to apply the theory to see how well it fits a given case. Schmitter then noted that the literature originally written for the Netherlands no longer even applies there today. He stated that no model could realistically expect to remain valid for many decades, precisely because interests in democracies change over time.

A participant made a final comment in the session, urging everyone present to be sensitive to the language being used to discuss democracy. Much of the language used in this session would largely mean the same thing to elites in host countries as it means to the people at the workshop. She cautioned that the same words might have very different meanings for people at the bottom of those societies, however.