In 1994 there was in Uppsala a symposium on 'Democracy's Victory and Crisis' and I have the feeling that in some ways we are thinking more about the crisis than we were a few years ago.¹ Why is this so? As to the victory, there are fortunately no alternatives to democracy presently appealing to people, as there were in the 1920s and 19305: communism, fascism, authoritarian and so on. There is no political system alternative to democracy, but that does not mean that we can ignore the problems of modern democracies in much of the world. I have the feeling that in addition to the two dimensions Robert Dahl has presented so well there is a prior one that is in some ways missing. A dimension that in Western advanced democratic liberal societies we are not questioning, which is the most important one, and that is that democracy is a method to govern a state. If there is no state there can be no political democracy. If there is no loyalty, commitment or obedience to a state, with all its elements and what it means, that is, the capacity to make rules binding for all the people throughout a territory and to achieve fair implementation of those rules and a monopoly of legitimate force, you cannot have democratic processes. You cannot introduce democracy in Liberia, Somalia or the Congo, and it is not working in Colombia, part of whose territory is dominated by various armed groups that are not subdued by the state but cannot win a civil war and do not really aim to take over the power of the whole state.

So, the question of the legitimacy of the state is a fundamental prior to democracy. That legitimacy is questioned in a multinational state in which the idea is that the state should be a nation state, but many of those who are not part of the dominant nation do not recognize that state. Until this issue is confronted, democratic processes will be challenged. It is governing a state that is the purpose of democracy. Now if the state is contested, that will no doubt affect the quality of democracy.

In the past the people who added adjectives to 'democracy' - 'organic',
'basic', 'people's', 'tutelary' - were the anti-democrats, non-democrats, who

---

*Juan Linz, Sociology Department, Yale University, PO Box 208265, New Haven, CT 06520-8365, USA.

wanted to share in the aura of democracy in some way. They thought they were developing their country toward democracy, that there was an alternativeform democracy. Today, those terms have disappeared from our vocabulary but we have 'defective democracies', 'illiberal democracies', 'plebiscitarian democracies', partial democracies in part of a country but not the whole country; we have electoral democracies in which elections take place but all the other freedoms and conditions for democracy do not exist. And, indeed, some of our colleagues, particularly David Collier, especially in one very important article with Robert Adcock entitled 'Democracy and Dichotomies', try to see democracy on a continuum from the most totalitarianrule the most ideal democracy. I think we have to retain a clear notion of when we are dealing with a democracy and when we are not. Belarus is not a democracy. The Russian Federation, maybe at the center, functions as a poorly working democracy. In the case of many of the republics and units of the Russian Federation, with 89-90 percent vote for the incumbents, you wonder whether they are democratically ruled. So let us be very clear about where there is a failure of democracy. Let us call such regimes electoral authoritarian regimes. Let us call them by names that may indicate a process of liberalization, transition or transformation. But ultimatelypower not come from the will of the people at regular intervals in these countries. Another dimension of democracy is pro tempore government.When see rulers extending their mandate for seven or eight years by a plebiscite, you must wonder about their democratic commitment.

Secondly, we assume that government has the capacity to govern effectively the territory of the country: to collect taxes, pay civil servants, enforce the decisions of the courts, and so on. If those elements do not exist, we are not dealing with democracies. In many cases when we talk about dissatisfaction with democracy, for instance in Colombia, are we really talking about dissatisfaction with democracy or are we talking about dissatisfaction systems that are non-democratic? Moreover it is important to distinguish transformations in a non-democratic direction - failed transitions distorted transitions - from the failures and crises derived from the quality of the society in question. People have expectations, and governments to do much more than they actually can do, and therefore the quality of a democracy is judged by the quality of the society and we should be very careful in making that distinction. There are rulers who are bending the democratic process in an authoritarian direction and we do not have the conceptualization of these new forms of authoritarianism. The new non-democratic rulers are not going to do what Hitler and the Bolshevik Revolutionaries did (say that they do not agree with our Western type of democracy), but they are adapting it in some ways. In addition, there is 'chaocracy', and no government and no state. We should pay more attention to such situations.

Political Parties: Some Contradictions in Their Structural Role and the Way It Is Perceived

It should be clear - and this was the point of my introduction - that we must distinguish non-democracies from reasonably institutionalized democratic states, governed by leaders committed to respect democratic-liberal rules. In addition we must separate - at least analytically - the problems faced by democratic institutions (even generated by them) from those derived from social, economic and cultural development, in so far that those problems are not directly the result of political processes - problems that in the short (and even medium) run are not amenable to solution by democratic governments.

Therefore I shall focus only on democratic regimes and a few problems faced by democratic institutions, particularly political parties - problems that should be the object of research. Our knowledge may help politicians to understand such problems and even to explain them to the citizens. Democracy involves contradictory principles and some of the dissatisfaction which Professor Dahl refers is, in a way, structurally inherent in democracy. We have data on how people prefer democracy to any other form of government and are committed to it and do not question the legitimacy democratic institutions as institutions, but also on how they question their performance, their efficacy, their incumbents. This is perfectly legitimate. There is an analogy in the history of the Church: people believed in the churches but were very convinced that some popes and many priests were very sinful. The distinction between the charisma of the office and the qualities of the individual were essential elements of the church institution as it developed from the Middle Ages. The charisma of the office, the authority of the president as president, coexists with the very negative opinion that many people may have about the particular incumbent. The two things should not be confused. That is why sometimes when data in Latin America show that the rating of a president has gone down from 70 percent to 20 percent (which is likely in presidential systems for various reasons), they conclude that democracy is in crisis. We should be careful about extrapolating from the incumbent's performance to the institution.

Let me take the central institution of modern democracies - political parties. We find general agreement that political parties are essential to a democracy; that there should be more than one party, and that the parties are in principle the instrument by which to participate in the political process. Without parties there can be no democracy. In Spain, for instance, 69 percent agree, 15.7 percent disagree, only 15 percent believe that parties do not serve for anything. As many as 74 percent agree that parties are necessary to defend the ideas and interests of different social groups. We have survey data from not only the Latinobarometer but also the Bruszt-

Simon 1991 survey of Eastern Europe showing the same pattern. However, in all those countries we find that those that believe in democracy as preferable,who that parties are necessary, still to a very large extent have very little or no confidence in political parties. But the differences between those who prefer a democratic system and those who prefer an authoritarian system in their confidence in political parties is not always that great.

In 1997 the Latinobarometer survey found that 62 percent of Latin American^ agreed that without political parties there can be no democracy, but only 28 percent had 'much' or 'some' confidence in parties, 34 percent had 'little' confidence, and 33 percent had none. As we might expect, there are significant differences between countries, with 81 percent in Costa Rica, 79 percent in Uruguay, 75 percent in Argentina and 67 percent in Chile considering parties necessary. The most negative responses - beliefs that democracy can function without political parties - are found in Ecuador (48 percent), Venezuela (43 percent), Colombia (42 percent) and Paraguay (39 percent), four democracies that have recently experienced serious crises. However, we should note that there are countries in which people believe in the need for parties but have little or no confidence in parties. One example

is Argentina, where a very large proportion consider parties necessary but only 29 percent have 'much' or 'some' trust, 32 percent have 'little' trust and 35 percent have none. Only in Uruguay do we find a congruent strong belief in the necessity of parties and a high level of confidence (45 percent 'much' or 'some', 34 percent 'little' and 17 percent none).

Venezuela is an example of a country in crisis where democracy has perhaps disappeared; among those who were democrats 12 percent had much or some confidence in parties, 25 percent had little confidence and 61 percent had none. Among those who thought that under some circumstances authoritarian system is better for a country, 9 percent had much or some confidence, 20 percent had little confidence and 70 percent had none; not a dramatic difference. A difference in the desirability of democracy not determined by the confidence or lack of confidence in political parties. If we were to construct a scale from the Latinobarometer the most favorable country would be Uruguay, where only 24 percent of the democrats had no confidence in parties. Spain is included in the survey, with 21 percent among the democrats having no confidence.

We can demonstrate the low trust in political parties compared with other institutions in Latin America (Latinobarometer 1997) by giving some examplesfrom countries with now stable democracies and from two with unstable democracies. In all five countries people not surprisingly trust the church more, over 70 percent, except in secularized Uruguay (57

percent) and Argentina (59 percent). More unexpected is the trust in the armed forces in Ecuador (71 percent), Venezuela (63 percent), but particularly in Chile (48 percent), Uruguay (43 percent) and Argentina (34 percent). Only in Uruguay are the parties more trusted (45 percent), while in Argentina and Chile - in spite of the tragic role played by the armed forces - the parties are less trusted (29 percent and 35 percent). The gap between the trust in parties in Ecuador (16 percent) and Venezuela (21 percent) and the trust in the armed forces (respectively 71 percent and 63 percent), could not be more telling and worrisome. In four of the countries television is trusted by more than 50 percent and only in Uruguay is the trust in parties (45 percent) and TV (46 percent) matched. The church, armed forces and TV are probably perceived by a significant number of respondents as non-partisan, serving the people as a whole (although this is far from true for the armed forces and TV), whereas parties are seen as divisive and power seeking.

The combination of the belief that democracy requires political parties and a low trust in parties was also found in nine countries of Central Eastern Europe (Bruszt & Simon 1991). Agreement with the statement 'We need political parties if we want democratic development' ranged from 95 percent in Bulgaria to 82 percent in the Ukraine. The question on trust was: 'In order to get ahead, people need to have confidence and to feel that they can trust themselves and others. To what degree do you think you can trust the following totally, to a certain point, little or not at all?'. The respondents were given 14 institutions and groups including political parties. The maximum 'totally' for parties was 6 percent and therefore we have added those saying 'totally' and 'to a certain point'; the answers range from a maximum of 36 percent in Bulgaria to a low of 6 percent in Estonia. The most negative, 'not at all', ranged from 25 percent in Slovenia to 49 percent in Poland. Again we find a striking contrast with confidence in the media and the greater trust in the army, except in Estonia and Lithuania.

Now, why is there this lack of confidence in political parties? We really do not know much about parties. We know a lot about party systems. We know a lot about types of political parties. We know who votes for parties, but we know very little about what people really expect from parties and how they see the function of parties in a democracy. We get contradictory perceptions. Those contradictory perceptions are apparently not linked with Left/Right positioning or with voting for one party or another, as I can show with some Spanish data. Indeed, some of those opinions are exactly the same irrespective of which party people vote for.

For instance, one of the themes that we find in the literature, the public debate and the newspapers is that parties are all the same: 'they are not really very different'. On the other hand, we find the opinion that the parties create differences between people which were not really there. Both feelings

that parties are the source of disunity and feelings that they raise a false consensus in society are very widespread. It is natural that when we deal with the catch-all parties they should be perceived as undifferentiated. This is the case when we deal with parties that no longer represent any narrow social basis, kept electorates, but appeal to all voters, with parties that have no strong ideological models like those in the 1920s and 19305: the Nazis, the Communists and their democratic opponents. They certainly were not the same. We might have preferred that they had been more alike than those parties, which represented totally different visions of society, or even the kind of parties that existed after the Second World War. Think of post-war Italy, the world of the Don Camillo movie where within one village the Catholics played in one soccer team and the Communists in another and the whole society was fragmented along party ideological lines. We have lost that kind of world, which I think is a good thing.

The competition between moderate catch-all parties aspiring to govern satisfies some of the basic aspirations of democratic citizens but also generates number of dysfunctions that reflect negatively on parties. It has eliminated the ideological and social polarization that was so destructive in the inter-war years, and it has facilitated alternation of governments with considerable continuity in policies and a climate of consensus. This certainly valued by many citizens disturbed by the potential divisiveness of democratic politics.

There are, however, negative consequences. Foremost is the feeling that
parties are all alike, that there are no real issues, that it makes no difference
which is in power, and therefore that politics is only a competition for

power among politicians unresponsive to society and its cleavages. There are also indirect, somewhat undesirable consequences: competition focuses on personalities and from this arises the temptation to engage in negative campaigning.

A good example of such critical opinions about parties is the significant number of voters who agree with the statement - or stereotype - that 'parties only serve to divide the people'. Among the Spanish population 4.5 percent agree strongly and 31.6 percent agree with this statement, with 43.2 percent disagreeing, 9.3 percent disagreeing strongly and 11.5 percent having no opinion. At the same time, a significant number agree with the statement that 'parties criticize each other a lot, but in reality they are all the same' - 15 percent agree strongly, 45.2 percent agree, 25.5 percent disagree and 4.7 percent disagree strongly. One would think that the parties on the extremes, the more ideological parties, would feel this way about parties one might more easily characterize as catch-all. In Spain we can test this assumption only on the Left, with the voters of IU (Izquierda Unida). Contrary to expectation, the proportions agreeing (strongly agreeing 11.6 percent, agreeing 41.8 percent, disagreeing 34.9 percent and strongly disagreeing 7.4 percent) are not that different, except perhaps in the slightly larger number recognizing differences between parties.

Recent Spanish survey data, after an election that led to a change of the party in power, reflect this feeling that parties 'are all the same'. This opinion is not that of alienated non-voters, nor of supporters of anti-system parties, but belongs to equal proportions of the majority of the voters for the two major parties, the socialist PSOE (Partido Socialista Obrero Espanol) and the center-conservative PP (Partido Popular). It is not easy to interpret the answer, but it probably reflects the lukewarm ideological conflict, catch-all appeals, and moderate policy differences within a limited range that characterize democratic politics in stable Western democracies. It may, however, reflect something else, that parties in a stable democracy, although representing different policies, do as parties share certain common characteristics.

Another major theme is that the parties are too disciplined and that there is too much unanimity within the parties. In the parliamentary democracies this is the requirement for votes of confidence and stability of government. We would have to go back to earlier parliamentary politics when parties did not preclude an open debate and vote by separating issues on which there is a vote of confidence and issues on which there is freedom to vote by the parliamentarians. Sir Ralf Dahrendorf, in a recent article, 'Traurige Pariamente', in a German newspaper, writes about the changing role of parliamentarians and how parliament has been lost as an arena for real debate. There are a significant number of people in democracies, for instance in Spain 36 percent, who agree with the statement 'In the parties

there is too much unanimity'. The interesting thing is that 36 percent of Conservatives and 34 percent of Socialists say so. There is therefore agreement across parties. This reflects an image not of one party but of parties in general.

There is also a perception of excessive discipline. On the other hand, as soon as there are debates in a party in which there is disagreement between various leaders, the media immediately say that this party is not capable of governing because they are not united, they cannot 'get their act together'. result is that the voters punish parties in which there is a considerable amount of conflict. Now, you cannot have it both ways. Should members of parliament follow the directives of the party or should they follow their own criteria?

One might think that party discipline would be considered more important voters on the Left, associated with the more ideological parties, those linked to a tradition of mass-membership parties, than by voters on the Right. The responses of Spanish voters do not fit that hypothesis. It is possible that, given the anticentripetal tendencies of the Spanish electorate and the catch-all character of all major parties, our hypothesis would not hold. The fact is that the electorate of all Spanish parties splits almost the same way on the question of party discipline. The national average was 34.1 percent favoring party discipline and 48.3 percent saying that the deputies should follow their own criteria, with 67 percent without an opinion. What is significant is that 35.4 percent of PSOE voters and 37.6 percent of PP voters opt for party discipline, with, respectively, 46.9 percent and 47.7 percent the independence of MPs. Only among the voters for the Catalan CiU and the Basque PNV were those favorable to party discipline slightly more numerous (42.1 percent and 40.9 percent, respectively), although number expecting an independent attitude among CiU was also higher.

A glaring example of the ambivalence of voters about parties is provided
by the responses to the following question (asked by the CIS, Centro de
Investigaciones Sociologicas, Study 2240, April 1997):

Think of two persons arguing about political parties in Spain. I would like you to tell me with which of these opinions expressed by them you agree more. (1) In (inside of) the parties there should be greater unity. (2) In the parties there is too much unanimity.

Among all respondents 40.4 percent agreed with the first statement and 36.8 percent with the second, while 21.1 percent did not know and 1.7 percent not answer. What is more striking is that the socialist voters (PSOE) divided 44.7 percent versus 35.2 percent and the conservative voters (PP) divided 45.0 percent versus 34.7 percent. Non-voters were more likely not to have an opinion (27.9 percent), but when they had one they split again, although a few more felt that there was too much unanimity (33.1 percent versus 37.1 percent). With the media alternately highlighting the internal tensions and imposed unanimity, we can see how across party lines people are likely, for opposite reasons, to be critical of parties.

There is another kind of constant disagreement inherent in parties. We want politicians who have experience, who know the issues, who know how to work together, and they should be dedicated full-time to their public duties. Moreover, they should give up all their connections with private business, their professions and whatever else. MPs should cease involvement any other activity. In Spain 58 percent say so and 27 percent say no. Among socialists the figures are 58 percent versus 26 percent, and among conservatives 58.5 percent and 27 percent.

We have on the one hand the image and criticism of the professionalization politics in a society that, paradoxically, values professionalization highly. The politician, then, should be an amateur, but an amateur who spends the whole day in committees, who spends time participating in the meetings of the executive of the party, who has to campaign not only in the national elections but in every regional election in a federal country, who has to attend to his or her constituents and who has to be active in party matters. Such a politician cannot, in modern life, do what the parliamentarians in the nineteenth century. They were doctors or medical professors in the morning, went to the hospitals, did their rounds, maybe taught a lecture, and were distinguished scientists, and then in the afternoon, for a short period of the year, sat in parliament, debated the major issues of the country and contributed sometimes to the making of legislation. This ideal is impossible in the modern world. We do not like professors who come back after five years in politics; even the three-year leave for public service is not very much liked by the universities.

There is disagreement among citizens about the professionalization of politicians, reflected in the desire on the one hand for continuity and efficacy and experience and on the other hand for term limits. Who wants to go into politics for five years, just learning the ropes, and never run again for that

office? We want accountability and at the same time those we entrust with public office should not be eligible again: how can we make them accountablefor well or how poorly they have done if they do not run again? This is an interesting issue in presidential systems. Again there is a tension between two contradictory principles.

Another question is that the people are ambivalent about the basic issue that 'my member of parliament should represent my interests'. That means the interests of my neighbors, the people of my community when there is a company closure, and so my MP should fight for those interests. Many voters feel that their representative does not really care for their interests. At the same time, we hear that parties and members of parliament serve all kinds of special interests. The voters of Seattle working at Boeing have 'special' interests different from those of another district, and so on. There is an interesting difference between America, where there is a certain legitimacy for interest representation, and the European system, in which the general policy of the party and the prime minister impose the party line independently of which district interests you are supposedly representing.

A special case are parties in multinational societies with a federal system and a nationalist dominant government in one of the component units, like Catalonia. In this case the party gives its support to the central government in exchange for policies benefiting that region. It has an imperative mandate from the regional government. It cannot be punished by the Spanish voters for what it does. It is rewarded by a local electorate only. Again, you may say the parties are, in a way, not representing the general interest.

There are some interesting debates about the internal democratization of parties and what we want parties to do to assure internal party democracy. do we really want 383,000 members of a party to decide matters that affect 10 million voters for that party - especially if only 54 percent of the members participate (192,000 people), as was recently the case in the PSOE internal primaries. In turn, of those voting, 105,000 (55 percent) voted for the candidate who would be the party's candidate for the primeministership the general election. Or should this be the decision of the parliamentarians who, after all, are elected by the whole electorate? So, we have a contradictory principle here: some people prefer one view and others prefer another view, and so there is bound to be some dissatisfaction.

At the same time that we find people agreeing with the statement 'Parties are only interested in my vote, not in my opinion', we find growing concern about 'poll-driven democracy'. Criticism is growing about democracy that sacrifices 'responsibility' and leadership in favor of 'responsiveness'. Appeals and policies are formulated in response to public opinion surveys and focus groups conducted to find out what people think. The emphasis on 'responsiveness' would seem to question the assertion that politicians do not care about 'my opinion'. Moreover, is not the vote an expression of a

preference, an opinion about a candidate and/or a party available to all
citizens, with consequences for the politicians which require them to take
interest in the voters' concerns?

The slogan 'The only cure for the failings of democracy is more democracy' is profoundly misleading. This is so because in most societies, except perhaps the United States, it can be translated into 'more presence and power for political parties'. A largely apathetic, fundamentally privatized, sometimes overworked citizenry will leave the filling of the many positions opened by 'democratization' on all kinds of executive or advisory boards (of public institutions or corporations, savings banks, universities, the mass media, etc.) to those nominated by the parties and will vote for them (if they care to do so) following their party affinities. The worrisome question is then: to what extent do the parties have people of independence, qualification and motivation to take over those functions? The result is_{partitocrazia} and often opportunities for corruption. It would be interesting to study the Italian, Venezuelan and Austrian crises of party democracy from this perspective.

We need to understand much better how the structure of political parties and their functions in modern democracies lead some people to be critical irrespective of which party they vote for. It is not that they think the other party is wrong. Parties in general, even their own party, are seen as wrong on various dimensions. Half of the population feel one way, half another, and this explains some of the dissatisfaction and distrust of parties. It also helps to explain why the institutions in society which are not democratic, like the monarchy, army, etc., are trusted. Even in Latin America, after all that has happened in some countries, the army is trusted more, perhaps because it is not such a complex institution as a party in a democracy.

Another question about which we know very little is how the image of parties - and politicians - across the board, and the problems derived from the nature of party politics and parliamentary life, affect the vocation of politics. We hear journalistic accounts of distinguished politicians leaving politics, of people saying they would never consider going into politics, but we do not know how widespread such feelings are and even less how they affect the recruitment of party politicians and MPs. Do the negative image, the full-time demands of political activity, the need to give up one's professional or business, the subjection to party discipline limiting the freedom to dissent, and so on, affect the decision to go into politics or to stay in politics?

---

NOTE

1. I want to thank Pilar del Castillo, the Director of the Centro de Investigaciones Sociologicas, including in the CIS surveys some questions relevant to this paper and making available some of the tabulations. I am also grateful to Marta Lagos, Director of the Latinobarometer, for making available her data and some tables I requested. Naturally, the responsibility for the interpretation is mine alone.

REFERENCES

Bruszt, L. & Jånos S. 1991. The Codebook of the International Survey of the 'Political Culture,
Political and Economic Orientations in Central and Eastern Europe during the Transition to
Democracy', 1990-1991. Budapest.

Collier, D. & Adcock, R. 1999. 'Democracy and Dichotomies. A Pragmatic Approach to
Choices about Concepts', Annual Review of Political Science 2, 537-65.

Delgado, 1., Martinez, A. & Oiiate, P. 1998. 'Parlamento y Opinion Publica en Espafia,
CIS, Centro de Investigaciones Sociologicas', Opiniones y Actitudes 19, 50-1.

Diamond, L. 1999. Developing Democracy. Toward Consolidation. Baltimore: Johns Hopkins
University Press.

Hadenius, A., ed. 1997. Democracy's Victory and Crisis. Cambridge: Cambridge University
Press.