There are at least three generic “models” of democracy circulating among theorists and practitioners in contemporary Europe. Each of them places primary responsibility on different types of actors and processes of decision making. In order to guide our collective thinking on the challenges and opportunities facing these actors and processes, we propose to use a generic working definition of democracy that does not “commit” to any specific institutional format or decision rules. By leaving open the key issues of how citizens choose their representatives, what the most effective mechanisms of accountability are and how collective binding decisions are taken, this definition does not preclude the validity of what we shall later call “numerical”, “negotiative” or “deliberative” democracy.

Modern political democracy is a regime or system of governance in which rulers are held accountable for their actions in the public realm by citizens, acting indirectly through the competition and co-operation of their representatives.

This definition provides us with a tripartite division of labour. Three types of actors combine through a variety of processes to produce the sumum bonum of political democracy, namely, accountability. We have, therefore, divided our analyses of contemporary transformations and responses into those primarily affecting citizenship, representation or decision making.

Citizenship

Political discontent

Today, one of the most striking features of European democracies is an apparently widespread feeling of political discontent, disaffection, scepticism, dissatisfaction and cynicism among citizens. These reactions are not, or not only, focused on a given political party, government or public policy. They are the result of critical and even hostile perceptions of politicians, political parties, elections, parliaments and governments in general – that is across the political spectrum.

Political discontent expresses itself in opinions, attitudes and deeds. Some citizens give utterance to their political disappointments or angers through day-to-day talks with friends or relatives. Social scientists try to analyse such opinions through polls, or in-depth interviews. The more intense these opinions or attitudes, the more likely they are to lead to actual deeds. In the political sphere these deeds are often “non-deeds”. Many disappointed or angry citizens refrain from voting or from joining a political party. Others explain that they are so angry with (established) politicians and political parties that they intend to cast a vote for some outsider, protest, or radical political party. Discontented voters are thus more likely to make unstable electoral choices, which partly accounts for the unprecedented high rate of turnover in the composition of governments.

Whether expressed through talks, polls or interviews, opinions may be (more or less) fragile, volatile, dependent on context, and even artificial. That is the reason why acts are more significant than words. Even if electoral participation is affected by many factors and cannot be reduced to satisfaction or dissatisfaction with politics, its evolution may provide a rough, but nonetheless informative view of the spread and growth of political discontent.

Figure 1: Turnout at parliamentary elections in Western and Eastern Europe

Sources: Figure 1 measures the evolution of the mean yearly turnout at parliamentary elections in all Council of Europe member states since 1980. It is based on electoral data of the member states with a population of more than 1 million which were members of the Council of Europe before 1980 (Austria, Belgium, Denmark, France, Germany, Greece, Ireland, Italy, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, Turkey, the United Kingdom). Such a grouping is obviously artificial, but it illustrates the overall declining trend of electoral participation in Europe and, consequently, the seemingly growing political discontent which partly determines voter turnout. Data for Eastern and Central Europe have been processed analogously and take into account voter turnout at parliamentary elections in seven states (Bulgaria, Czech Republic, Hungary, Lithuania, Poland, Romania, Slovak Republic). These states have been chosen because they have a population of more than 1 million inhabitants and because they have been a member of the Council of Europe for more than ten years after 1990.

“European” voter turnout has decreased from 88% in 1980 to 74% in 2002, and even 70% in 2000. Electoral participation is declining – with a more or less gentle slope – in all countries except Denmark. If we extrapolate the past overall tendency, turnout will be close to 65% in 2020, and even lower if we take into account the voting-age population instead of the registered voters. The decline in electoral participation is even more marked in Central and Eastern Europe. In this region, the weighted mean voter turnout has moved from around 70% at the beginning of the 1990s to 60% just ten years later. It would decrease to around 45% at the beginning of the 2020s if we extrapolate its evolution over this brief period. These conclusions are backed up by opinion polls, showing a clear downward trend of trust in parliaments in Europe.

Attitudes towards the political realm are however more ambivalent than opinions of discontent may suggest. National parties and politicians, that is to say specialised and professionalised political actors, are much more criticised than those on the local level. In other respects, people make distinctions among various political levels or dimensions. Some of them, belonging mostly to upper middle and upper socio-cultural strata, see differences between a “politicking” component, which refers to parties, politicians, elections, rivalries and struggles for power, and a non- or less-“politicking” aspect, associated with projects, programmes, issues, ideas, principles, convictions and efforts to solve problems. When people in this category criticise politics, they usually (tacitly) think of the former dimension while refraining from criticising the latter. Some of these most censorious citizens are willing to believe in politics when a leader or a party appears “different” to them or when crucial issues (terrorism, fascism, welfare state) are at stake.

The way people perceive and criticise the political realm depends on their political investment and skills. Two different types of discontent may must be distinguished: a rather simplistic and timeless one and another more sophisticated type. The former has been around for a long time, well before politicians and political scientists began to worry about distrust in political institutions and actors. It is older than the political changes (for example globalisation, “de-localisation”, the “crisis” in nation-state capacity, rising unemployment, European integration) often singled out as explanations for political disaffection. Discontent of this type is easily – perhaps too easily – picked up by opinion polls. Respondents who share these views belong, at least statistically, to definite segments of the public which are characterised by:

• a rather low (but somewhat unequal) level of education, social status, political information and sophistication, and a feeling of personal political incompetence;
• no marked political preferences and even an inability to perceive differences between politicians and parties;
• a fear of “being had” by politicians because of this incompetence;
• a lack of interest in politics which leads to thinking and arguing that politics does not deserve their attention;
• a narrow vision of politics, mainly reduced to the above-mentioned “politicking” dimension, as a consequence and a determinant of this lack of interest;
• bad living conditions that prompt people to think that politicians do not really care about them and, therefore, that politics do not matter.

At the same time, other people express more sophisticated feelings of discontent. Contrary to those who may be subsumed within the previous category, they refer to various shifts in the political realm. They say, for instance, that “there are not many differences among political parties nowadays”, “left- and right-wing parties are presently very similar; they pursue and drive the same policies”, “politics is increasingly lukewarm”, “it's no longer important, it's economics that matter now”, “nation-states can't do much against firms' decisions to relocate”, or that “the EU decides on everything”. These opinions are held by people who add that they used to be, but are presently much less, interested in politics and that their political preferences have waned. Nevertheless, many of them still have strong negative preferences, in the sense that they are strongly opposed to some political parties. They also pay enough attention to politics to be able to criticise political actors, using informed arguments. People who may be classified in this second category have a higher (but not necessarily a very high) level of education. They are more interested in, informed about and confident in their ability to cope with politics than those in the first category, and have a more complex, diachronic, and even lofty view of it.

Causes and explanations of political discontent

Political discontent proceeds from a set of convergent factors.

Education. The higher the level of education, the higher the feeling of political competence. The higher the subjective and objective political abilities, the higher the criticising capacities and tendencies. Increasing cognitive competence among citizens increases capacity for criticism, and a greater willingness to criticise if something appears to be wrong. A more educated citizenry has a more critical mind and is potentially more demanding with its political leaders and representatives. More educated citizens also tacitly wish to be more active, even if they are not ready to invest time and energy when they are really asked to participate in something. Demands for more significant and direct forms of political participation are therefore real, although somewhat ambiguous. One of their real effects is perhaps that the significance of voting for representatives, as the most important form of democratic participation, is bound to diminish. A small but seemingly growing number of (relatively) educated citizens are more or less plainly asking for greater opportunities to express their own opinion and to decide by themselves on important subjects.

Changing values. European citizenries or, at least, large segments of them, seem to have shifted from deference to authority and authorities to scepticism of elites and institutions. But, for numerous and complex reasons, a growing pervasive permissiveness and intolerance of social norms and authority has been spreading for a long time. A growing culture of rights, equality and personal autonomy is somewhat contradictory to the deference, compliance, discipline, hierarchy and leadership that organise citizens-representatives relations in a representative democracy.

Economic shifts. Economic growth has been weak during the last three decades. Unemployment has increased. Real wages have remained stable or have grown only slowly for years. Declining trade barriers and transportation costs and improvement in communication have enhanced the role of international trade and investment in all economies. Global competition brings various advantages to some categories, but also entails relocation of firms to low-wage countries, depression of wages in advanced countries and downward competitive pressures placed on labour standards. New technologies are also eroding skilled labour and wages, even if they help to create new skilled jobs at the same time. Globalisation has challenged the capacity of the states to provide effective regulation in the economic and social domain. New institutions like the European Union or the World Trade Organisation have weakened nation-states' policy latitudes. They have also suggested that nation-states may become a less significant collective actor. Nation-states are also hollowed out by deregulations and privatisations. Governments have at the same time faced a “fiscal crisis”, and tried to balance budgets by containing public sector outlays. Social services have been reduced or their expansion has at least come to an end.

A growing number of citizens have been increasingly confronted with problems resulting from global economic competition, economic crisis and diminishing welfare protection. Those who personally, or whose relatives, endure or fear unemployment, and those who think their economic situation will worsen, are more prone to a negative perception of politics. People who already thought that politics could not improve their life and that there was nothing to expect from politicians have seen their opinion confirmed.

Dramatic and highly salient woes, whether “objective” or “imagined”, like recessions, rising immigration, loss of local control, unemployment, and insecurity, have led some segments of the public to share the conclusion that government was handling problems poorly and was failing to keep its promises, whether they were personally affected or not, and regardless of achievements in other areas.

For more sophisticated segments of the public, the level of political discontent is also linked to more complex evaluations of governments' performances. The successful economic policies of governments during the first thirty years after the Second World War and the subsequent downturn of the mid-1970s have raised and then disappointed expectations of state capacity to deal with growth, inflation and employment. Poor or poorer states' economic performances of the last decades have seemingly been assessed with reference to the economic boom of the thirty “glorious” post-war years, and also through expectations arisen from a century of expanding public interventions.

The most sophisticated segments of the public are more aware of the economic and social changes of the last decades. They do not reason as if no change has occurred. They think that the nation-state is no longer able to deal with major economic difficulties, that it cannot oppose the decisions of international companies and prevent de-localisations of plants. Their expectations of what governments are able to accomplish are diminishing. However, they still remain unsatisfied with politics because they tacitly compare present governments' performances with prior ones, or with their developed normative views of what governments should do. Both normative and ideological expectations thus provide critical resources which are activated by what appears as government failures. The conjunction of growing critical resources due to higher education and numerous political disappointments give rise to permanent critical dispositions in the politicised strata of the public. These critical leanings are activated when people face personal difficulties, whatever they may be, in their own life.

Political context. When people explain their political disappointment, they refer or allude to various elements of social and political contexts to vindicate their disillusions. One observes that dramatic revelations of political corruption and scandals in numerous countries have fostered a climate of ethical distrust.

Ideological and political distances among political parties have been reduced. In various European countries, politics was, but is no longer, regarded as a struggle between contrasting, even utopian, views of society and its future. Since the collapse of the “real existing” socialist system, almost no established party intends to overthrow the market economy, capitalism and liberal democracy. For various reasons mentioned before, governments' leeway has also been reduced. This has led some segments of the public to the conclusion that politics does not matter anymore, that it is not worth losing time to decide between similar parties, defending similar policies, and that parties and politicians are competing only to enhance their own power and privileges. Those who have kept some partisan attachments deeply regret that left-wing parties support what they regard as “neo-liberal right-wing” policies, or that right-wing parties leave unchanged “socialist left-wing” policies when in government. Some citizens feel that politics has lost authenticity, and is increasingly run by self interest and ulterior motives. They even sometimes allude to opinion polls and communication specialists as having caused these changes.

Recurrent episodes of political life that were perceived as neutral or normal in the past now fuel political distrust when this distrust has become high enough to produce a bias against politics itself. The current and frequent bashing, bad mouthing and blaming of “the government” by elected and selected representatives thus help to develop increasingly negative perceptions among segments of the public already prone to reducing politics to “politicking”.

In order to maximise their audience, the media tend to simplify, personalise, dramatise and stress the “spectacular” aspects of political events. They cover politics rather than policies, focus on scandals, tactics and personal rivalries, and describe electoral campaigns as if they were “horse-races”. Candidates and public officials are often depicted as duplicitous and self-serving. The media tends to reinforce the fears and prejudices of those among their consumers who see all politics as merely “politicking” – if only because this makes the information more entertaining and easy to understand. This is especially the case for those “citizen-consumers” who are only slightly interested in the subject and already prone to mistrust politics due to their attitudinal predispositions, social marginality and lack of trust in institutions.

Does political discontent matter?

Is the apparent increasing level of political discontent threatening the legitimacy of European polities? First and foremost, political discontent is ambivalent and the present disenchantment is potentially reversible. A second point is that there is an ambivalent decreasing confidence in politicians, parties, elections, legislatures and governments, but apparently no pervasive distrust of other dimensions of European polities. The legitimacy of a political system depends on the existence of an alternative and competitive polity or utopia, and the struggle over different forms of governmental and societal organisation has disappeared at least since 1989. Some scholars argue that since the collapse of the socialist system, citizens' support for democracy is becoming increasingly dependent on governmental performance, especially in former socialist countries. Democratic systems seem thus more vulnerable, but also unquestionable and stronger at the same time.

For the same reasons, a high and growing level of electoral abstention is not a threat to the political system per se. But as abstention increases with lower social rank and as politicians are more eager to take voters' than non-voters' expectations into account, declining electoral participation should tend to introduce or strengthen the class bias in public policies.

The lack of confidence in political institutions raises the question of the willingness of the public to comply with laws, to pay taxes or to enter administrative careers. Several isolated acts of violence against politicians and officials perpetrated in some countries could be linked to a growing political discontent. Ethical distrust of politicians is already a serious problem since it weakens dispositions to comply with rules and laws. Young criminals say, for instance, that they do not feel ashamed of their thefts, robberies, or drug dealings, because “political leaders have stolen much more than we have”.

Cultural identity and protest

Migration, defined as the movement of persons from one region or country to another, irrespective of motivation, gives rise to important population changes which affect democratic life in Europe. Migration diversifies the composition of the European demos as it causes people with different legal status to co-habit under the same democratic roof; along with national citizens there are guest workers, long-term residents (or denizens), asylum seekers and undocumented migrants. All these groups of people, because of their legal status, are subject to different sets of rights and obligations.

Democracy, citizenship and rights

a) Levels and characteristics of migration

Since 1989 net migration has been the main factor impacting annual population change in Council of Europe member states. Figure 2 below presents the trends in the change of stock of foreign population as a percentage of the total population for fifteen countries in Europe. The total recorded stock of foreign population is approximately 21 million people in 1999 amounting to about 2.6% of the total population of all countries. The data suggest that in 1999, the highest proportion of foreigners relative to the total population was in Switzerland (19.3% with two-thirds of foreign nationals being EU citizens). The greater part of the foreign stock is resident in Western Europe while in Central and Eastern Europe, the proportion is relatively small (less than 2%). Net in-migration in both regions was relatively high in the early 1990s, with the Federal Republic of Germany experiencing the largest absolute increase. In the late 1990s, in some countries in Western Europe, the proportion either declined or stabilised. The percentage of foreigners has been increasing for most countries since 1998 albeit at lower levels for Central and Eastern Europe (the largest numbers have been in the Czech Republic among Central and Eastern European states). For Western Europe, data point to considerable diversity in terms of the origins of foreign migrants and a majority of the foreign national population comes from outside the European Economic Area (EEA) plus Switzerland. Foreign migrants from different regions select different countries as their destination. For example, Africa is an important source of migrants for France while for Spain and Portugal, Latin America is a major region of origin. Asians migrate to different European countries for various reasons; those from the Indian subcontinent usually go to United Kingdom, Filipinos mainly to Italy for temporary employment, and Greece receives immigration from the Middle East region. Germany stands out as the most common destination for nationals of non-EU European countries. Temporary and transit migrants also constitute a substantial population in Central and Eastern Europe.

Figure 2: Stock of foreign population as a percentage of total population in selected Council of Europe countries, 1980-2020

Sources: The estimations are based on data from Council of Europe yearbooks, OECD SOPEMI “Trends in International Migration, 2003”, Salt, J., 2001; “Current Trends in International Migration in Europe”, Council of Europe; Wanner, P., 2002; “Migration Trends in Europe”, European Population Paper Series No. 7. The data draws on foreign stock as a proportion of the total population since those countries with the highest number of foreign residents are not necessarily with the highest proportion of foreign residents.

In Europe, starting from the late 1950s, migrant workers were actively recruited abroad but were not expected to stay in the receiving country permanently. Foreign labour recruitment has formally ceased in Western European countries since the mid-1970s; however, the stock of foreign population has not decreased due to low return rates and family reunification. Many former guest workers have acquired the status of resident non-citizen. This category of people, often referred to as “denizens”, enjoys an intermediate status between citizens and aliens. They are incorporated into various social, economic and legal structures while not enjoying full rights of political participation. Rules for granting “denizenship” and the rights and benefits attached to that status vary from state to state. However, “denizenship” has become a salient and stable feature in all Council of Europe democracies, and has led to reconsideration of who has the right to participate in politics and how.

b) Denizenship and nationality

The continuity between people and place, nationality and demos, is a major premise of modern democracies. EU citizenship is a prominent example of how the boundaries of political membership can be enlarged and the demos can extend beyond national borders. However, even in EU member states, third country nationals are not included in the complementary status of EU citizenship as defined by the Maastricht Treaty. This indicates that so far attempts to expand citizenship rights beyond nationals do not offer a comprehensive framework for attending to issues relevant to the political participation of third country nationals. This invites innovative thinking on the composition of political constituency, citizenship and mechanisms of political participation.

In fact, civil and social rights have also been extended to third country nationals in the EU. Such a trend suggests that citizenship is no longer the exclusive way to access the benefits of state membership and become fully integrated members of a community. Yet, political rights are a prerogative of citizens only. This prerogative is an important one as, for example, the rules for the allocation of social and civic rights are made and altered by those who have and exercise political rights, namely the “native” citizens. This is particularly problematic, for example, in times of economic crises, when citizens and their representatives may decide to cut down social benefits for resident non-citizens, with the latter being excluded from the decision-making process.

However, resident non-citizens contribute substantially to the economic and social development of their country of residence, pay taxes and are expected to abide by its laws. In other words, they share the burdens and benefits of social co-operation. Denying them full political rights appears to violate one of the basic (normative) democratic principle according to which those affected by a certain set of social and political institutions, should also be granted rights that allow them to influence these institutions and their policies. Recognising that the absence of such political rights as a form of democratic deficit, some governments have endorsed various channels of political participation for denizens other than the right to vote. In some Council of Europe countries, denizens are granted opportunities for indirect influence in decision making through government funded organisations, consultative bodies and unions. Such models usually shift the focus of democratic practice from the national level to the local level. By engaging in civic practice at the local level, resident non-citizens come into contact with representative bodies, associations and lobby groups which could also give them a voice at the regional and national levels. Moreover, practice at the local level can result in skills which will enhance participation at national and supranational levels.

Granting denizens political rights, for example the right to elect a representative in municipal elections – as in Ireland, Sweden, Norway, Denmark, Netherlands, Finland and Luxembourg – introduces a significant change in the terms of political competition. In other words, candidates who win elections have to be accountable to a more diversified constituency, and be responsive to the needs of a sector of the population previously excluded from political life. Moreover, new issues (for example, in the fields of education and health) appear or gain precedence on the political agenda.

Additionally, giving voice to denizens offers an opportunity for dealing with potential ethnic and cultural conflicts through democratic procedures. In this way conflicts are not overlooked, but faced and possibly solved. This would favour a free confrontation of ideas through open dialogue and deliberation, enhancing self-reflectivity and critical multiculturalism in the receiving society.

The fact that this category of migrants is going to reside permanently in the receiving country has been regarded as justifying a project of multicultural citizenship in which political rights can be shared by national and non-nationals.

c) Minorities

Increasingly, some resident non-citizen and sub-state national groups demand collective recognition as well as individual participation in policy processes, for example as “minorities”. Favourable conditions for meeting such claims originate from their considerable numeric presence in some countries, the international conventions supporting minority recognition and the general concern for securing fair access to political, social and economic life for the previously excluded sectors of the population. On the one hand, most Council of Europe countries articulate a common commitment to recognition of group rights and their accommodation (for example, in the countries that emerged from the former Yugoslavia). On the other hand, some (such as France) contest the recognition of these groups as “minorities” and the concession of group rights even when they recognise the equality of people of different cultural and ethnic origin. Some states (such as Croatia) have introduced quotas for linguistic minorities in regional and local representation, while others maintain consultative bodies, that is a second chamber of parliament, or “veto” mechanisms for national or religious “communities”. However, the “smaller” and territorially dispersed members of national minorities, especially the Roma, remain excluded from almost all schemes.

d) Illegal migration

In recent years, illegal migration estimates have reached worrying levels, especially in Southern European countries. Undocumented migrants, officially prohibited from taking up employment, supply a significant proportion of the labour force in the “hidden” or “underground” economy in these countries. Irregular migration is advantageous for employers in the receiving country that profit from undocumented migrants’ lower pay and more flexible and longer working hours. Both the state and the juridical system are absent from this informal sector of the market. The demand for labour supply of the undocumented migrants fosters human trafficking and smuggling, and a dramatic growth in the shadow industries that treat people as commodities in this trade. Every year, thousands of people, especially women and children, fall victim to trafficking for the purposes of sexual or other types of exploitation, which can be equated to a new form of slavery.

Attitudes towards migrants

Observers have pointed out the alarming increase in anti-migrant attitudes in Council of Europe member states. The United Nations World Conference against Racism, Racial Discrimination, Xenophobia and Related Intolerance has brought this problem into focus, and international organisations such as the International Organisation for Migration, the International Labour Organisation and the European Centre for Racism and Immigration have pointed to a growing negative attitude towards non-EU migrants, reinforced by racial stereotypes diffused through the media and by some political leaders. Data from the European Monitoring Centre on Racism and Xenophobia (EUMC) show that racially motivated attacks have increased in most EU member states. For some, official statistics indicate a possible reduction of crimes in the last two years. Concerning Central and Eastern European countries, Amnesty International has reported a rise in xenophobic attitudes and racist violence in the late 1990s.

Consistently in all Council of Europe countries, the typical perpetrators of racist crimes are young (18-26) males with low levels of education. However, some NGOs (non-governmental organisations) have reported to the EUMC that worrisome numbers of racist acts of violence are also being committed by law enforcement officers. This suggests that racism finds expression even within the established institutional structures. Extreme right-wing parties – the electoral success of which has risen substantially starting from the 1980s – often appeal to the xenophobic feelings of people in their political campaigns. They also base their electoral strategies on the claim that migrants threaten the national culture and its symbols (such as crucifixes in Italy and Germany), and on a supposed link between unemployment and the number of migrants in their respective countries.

There is no actual empirical or theoretical evidence to support the claim that immigration causes unemployment. To the contrary, some studies show that citizens who are the closest substitutes for immigrant labour do not suffer as a result of increased immigration. Additionally, various studies on demographic trends have identified migration as a possible solution to overcome the “demographic deficit” of Europe and its related problems. It has been argued that in-migration may be a political option to meet the strategic economic and social goals that underpin Europe’s market economy.

A further issue often raised by leaders of xenophobic parties is that immigration constitutes a threat to political and social stability. It is also a commonly spread perception that crime rates have increased as a consequence of immigration. The fear of migrants has been exacerbated by the events of 11 September 2001 and 11 March 2004. Suspicions and fears are especially directed towards migrants coming from the Arab countries and South-East Europe.

These negative perceptions have often been reinforced by the image of migrants portrayed by the media. News programmes report crimes indicating the offenders generically identified as members of a minority group. In a similar vein, several crime fiction programmes characterise murderers and offenders as people of foreign ethnic origin.

At the same time, media represent an important opportunity for the participation and integration of resident non-citizens. Through various mechanisms such as funding multicultural programmes, some countries have highlighted the positive effects that media can have on the perceptions and attitudes of the public, as well as on the self-perception of migrants.

Representation

Political parties

No democracy exists without political parties even if parties differ in organisational structure, ideology, size, functions and goals. They act as an intermediary between voters (citizens) and public authorities (rulers) By structuring the political field they help voters in making their choice and they help rulers put together governments. There exist many definitions of political parties ranging from the very broad to the extremely narrow. Often, definitions are based on one or more of the functions of political parties. The most commonly accepted criterion is that parties should compete in the political arena, try to get their candidates elected and play a role in forming the government. Parties might fulfil a wide variety of functions – although not all parties are engaged in all functions, and certainly not to the same degree. They might play a crucial role in recruiting and selecting the political elite by nominating candidates for elective offices and filling government positions, in forming and sustaining governments, and in policy making. They might also play an integrative role in society by mobilising and providing a collective identity to voters, by aggregating and articulating social interests, and by enhancing the legitimacy of the political system. In addition, they might similarly engage in voter socialisation, issue structuring and/or social representation. However, a party in a democracy cannot represent the whole of the society as the origin of its name pars (part) illustrates well. In order to avoid democratic deficits, parties in democratic systems are expected to be democratic and transparent themselves as well as establishing lasting and regulated relations between party leaders and their membership.

Party membership, size and organisation

Although party membership shows a declining trend in Europe, such a claim fits long-established democracies of Western Europe better than the recently democratised Southern, and Central and Eastern European countries. All Western European countries report a decrease in the number of party members. Countries belonging to the third wave of democratisation display more variation, with some even showing growth (Greece, Hungary, Slovak Republic and Spain). All this suggests that declining membership is a pervasive trend in well-consolidated democracies. This raises concerns over citizens’ participation in public affairs: democracy is in danger if citizens are so apathetic or disillusioned with it that they avoid joining one of its most salient institutions. The less they join parties, the less they are likely to vote, and when they do both, the less governments can be held accountable, the less individual rights can be enforced, and the less individual and group demands can be represented in the policy process. Finally, the more people choose not to be represented in decision making, the less they will recognise the actions of their democratic government as legitimate.

Figure 3: Party membership in Western and Eastern Europe

Source: Figures from Peter Mair and Ingrid van Biezen (2001). “Party membership in twenty
European democracies, 1980-2000”, Party Politics vol. 7, No.1.

However, decline in membership is not always worrying per se. It is not necessarily a sign of declining political participation in general: the decrease in party membership can, at least in part, be accounted for by the appearance of new, more individually appealing and socially acceptable forms of participation (such as signing petitions, boycotting certain products for political reasons, or demonstrating in favour or against a specific policy). Furthermore, the incentive to join parties becomes weaker as parties face various competitors that partially seem to take over the one or the other of their functions while inflicting less demands on and guaranteeing easier access to citizens. Similarly, parties’ demand for members is also waning: the changed nature of campaigning (increase in the use of the media accompanied by a decrease in traditional forms of campaigning that rely on volunteers) and the restructuring of party financing (greater reliance on public funds and, thus, less on member dues) have made it increasingly less important for parties to maintain large memberships. That is, no longer do parties have to provide a “regular or full programme of participatory events” besides those that are directly related to elections. Finally, a large party membership may not make the democratic system more responsive to citizens’ demands: the larger the membership, the smaller its ability to influence party leadership may be.

During the 1990s, scandals involving illicit party financing were frequent throughout Europe. Regardless of political or party system, party organisation or ideological orientation, political corruption related to party financing has become a persistent problem. Despite differing institutional and policy arrangements, almost every European democracy has had serious difficulty in coping with providing sufficient funds to its political parties and ensuring that these funds were equitably distributed.

a) Higher (and growing) party expenditures that outweigh legal revenues

By the 1970s, parties had built up permanent and sizeable bureaucracies. These administrative apparatuses proved to be costly to sustain, especially in non-election years when donations were not as generous. Parallel to this, the importance of national-central offices had come to outweigh that of local party branches, and this brought with it a rise in the need for (expensive) professional expertise.

Election campaigns became more and more expensive. First, campaign techniques changed due to technological advances that made volunteer labour less effective. Not only was the importance of partisan militancy diminished, but its replacement – television, radio and newspaper advertisement – were much more costly and effective in influencing voter behaviour. Second, the increasingly competitive, market-oriented, nature of electoral politics and the tendency for longer campaigns have forced the central organisations of parties to invest more money and professional resources in their effort to catch as many votes as possible, regardless of previous affiliations and class loyalties. The more that parties found it necessary to reach out beyond their traditional supporters, the more expensive each additional vote became.

Moreover, in Europe, parties have historically felt the need to sustain a level of participation in national and, sometimes, regional and local political assemblies, social organisations, study groups, partisan foundations and intellectual academies between elections, and these have had to be financed from party budgets (although sometimes with the help of public subsidies). Although it is difficult to gather reliable data on the amounts involved, in recent decades Western European parties of both the Left and the Right have engaged in supporting “sister” parties or political groups in foreign countries undergoing political liberalisation and democratisation. Again, government funds have often been channelled through party organisations for this purpose, but such trans-national activity has no doubt contributed to professionalising their permanent staffs.

Even if legislation on the legal sources of party finance varies throughout Europe, certain general trends can be observed almost everywhere. The composition and sources of income have changed substantially since the 1970s. Membership dues have become less important for party budgets. First, because of declining membership, parties simply cannot raise as much money from this source. Second, the great increase in the need for money demanded that parties look elsewhere for financial support. And, third, in the new democracies of Southern and Eastern Europe membership dues never gained central importance due, in part, to the historical circumstance of single-party systems with various forms of obligatory contributions and, in part, due to the very timing of their processes of regime change.

Another important source of party income has been donations. These may come from various entities: private individuals, business firms, trade unions and/or associations in civil society. Some domestic donations have been banned or limited by law, but subterfuges have not been difficult to find. In most European countries, donations by foreign governments, parties, firms or individuals have been prohibited, but again unknown amounts still seem to manage to get through the ban, especially, when “kick-backs” from foreign aid contracts and state firms operating abroad are involved. Needless to say, many of the party finance scandals of recent years have originated in this murky and difficult to control area of donations.

Parties have historically earned money from a wide variety of firms that were owned or closely affiliated with them, such as printing plants, newspapers, book publishers, travel agencies, consultancies, planning bureaux, research institutions, recreational societies, sports clubs and, more recently, party foundations. While it is difficult to judge the evolution of the importance of these sources of finance, the impression is that they have declined due either to the break-up of ideological solidarity or to commercial competition. For example, it is doubtful whether any partisan newspaper or publishing house in Europe currently earns a profit which is significant enough to represent an important source of party finance.

Public subsidies to parties have gained greatly in significance since the 1970s. While three decades ago such subsidies were rare, today they are a major source of party income throughout Europe. Legislation in each country determines how these subsidies should be distributed, how they can be spent and how they are supposed to be monitored. They can also be allocated directly in the form of money, indirectly in form of free access to television or other media, or some combination of both.

b) Corruption

Illicit party financing has long been a common phenomenon throughout European democracies, but only recently seems to have become a threat to their legitimacy and a source of decline in public trust. What counts as illegal is determined by national legislation; what is illegal in one country may be quite legal in another. Incomes may be illicit if they come from entities whose contribution is banned by law (such as foreign firms or governments), from organised crime, from individual contributions that exceed legal limits or circumvent the legal requirements for recording. Two time-honoured, if illicit, sources are kickbacks from public contracts and bribes – usually to the incumbent party – in exchange for some favourable policy treatment.

There is no reliable and objective way of evaluating whether, over the past thirty years, parties have become more or less corrupt. The above-mentioned gap between rising demand for funds and limited supply from traditional sources suggests that there is a greater material incentive to resort to inconfessable means of finance than in the past. What does seem clear is that public tolerance for illicit fundraising – even when not tainted by personal fraud or profiteering – has increased. Citizens seem to be applying higher standards of ethical behaviour to their representatives and rulers and they are better informed about corrupt practices, thanks to the Internet and to comparative indicators such as that produced by Transparency International. The media have become more inclined to publicise funding scandals; the judiciary more disposed to prosecute those who engage in such acts; the citizenry more likely to react by punishing even those just suspected of corruption. Regardless of what this implies for the long-term future of democracy, the immediate term consequences are serious. Contemporary regimes in Europe have a serious problem with their “internal political economy”. Democracy costs money – and more money each time its electoral game is played. Its ultimate beneficiaries, the citizens, are less inclined to pay these costs whether through voluntary private contributions or compulsory public subsides. Since parties are still the only known way of structuring both electoral competition and the formation of governments, they cannot simply go bankrupt and disappear – or democracy as we know it would disappear.

c) Parties move away from the civil society and closer to the state

This is the result of many factors, including declining party membership, changing campaign techniques and dependence on the state for financing their growing expenditures. Volunteer work for parties became outdated once parties tried to reach people primarily through television. Also, individual donations as a passive form of participation have been declining. A small number of large donations have tended to replace a large number of small contributions. At the same time, as we have seen above, parties have become increasingly financially dependent on public funds and, in some cases, upon big business firms. The state is crucial, not only because in Europe it has become the major supplier of party funds, but also because the incumbent party or parties that control it can gain access to other (often illicit) sources of revenue. On the one hand, the advantage of incumbency has been rising and, hence, the likelihood of oligarchy; on the other hand, all parties have become increasingly under-funded and, hence, the prospect of irrelevance and fragmentation when none of them succeeds in connecting with its targeted public.

d) Citizen disaffection from and apathy towards politics

The problem is not simply that the participation of people in the life of parties has been declining (other forms of political participation may replace this) or that private donations are decreasing (other sources, especially public ones, have been increasing). Cynicism about the motives and practices of politicians has increased to the extent that a large proportion of the population considers political corruption to be “business as usual”.

* * *

These problems are only in part endogenous to “real-existing” democracies. They are also closely connected to the exogenous challenges and opportunities outlined in Part I.

Globalisation and its consequences pose one of the biggest challenges for parties in Europe. First, trade liberalisation meant that money could move more and more freely across national borders which widened the range of potential sources of financial support for parties and this proved to be vital for the opposition in post-communist countries struggling for democracy in Central and Eastern Europe and the former Soviet Union. It also poses a dilemma for those established national democracies, however, which try to resist the influx of foreign funds into their domestic elections and policy processes. Second, the concentration of money that globalisation places in the hands of multinational corporations and wealthy individuals could make it easier for parties to raise money. However, this would make these parties more vulnerable to the accusation that they have become excessively dependent upon business interests. To the extent that this occurs across the political spectrum, it reinforces the popular notion that “all parties are alike” and, hence, that choosing among them is a futile exercise.

European integration poses a very similar challenge. The tendency to integrate markets, professions and policies on a regional scale blatantly contradicts one of the central premises of existing party systems, namely, that they are responsible for organising political competition on the basis of a sovereign territorial unit, the national state. For example, people living and working in a country different from that of their citizenship are often prohibited from financially supporting parties in the state in which they reside, while firms incorporated in any EU member state, but foreign owned, are free to make partisan donations.

Technological changes have literally revolutionised campaigning and, in some countries, are beginning to affect fundraising. Television has firmly established itself as the major medium for addressing the general public during elections. Depending on the mix of private and public channels and on the content of licensing arrangements, this has greatly increased the financial cost of campaigning and depreciated the importance of voluntary contributions of labour by party members. It has also enhanced the focus on the personality of candidates at the expense of the appeal of party platforms, since that is what is best projected by this exceedingly time-dependent form of mass communication.

It is too early to assess whether information and communication technology (ICT), in particular the Internet, will provoke an analogous revolution. The Internet seems to have the potential for reversing the trend towards exploding costs and, therefore, for evening out the conditions of competition between large and small, poorly and well-endowed parties. And there are strong indications that parties are experimenting extensively with this medium to reach actual members, potential donors and eventual voters. No party today can afford to be without its website. But this is also the case for individual candidates and elected representatives. Will this rapidly expanding form of direct communication (and, eventually, of electronic voting) have the effect of further undermining traditional forms of party organisation and affiliation?

A sense of insecurity probably has a significant, but indirect effect upon partisan organisation and activity. As we have seen above, the increased demand for money (and decreased supply of it from members) makes parties vulnerable to engaging in corrupt practices and this can enhance the influence of organised crime as a potential source for the missing funds. Indeed, some techniques of illicit party financing closely resemble those of money laundering and some of the means for soliciting contributions are hardly distinguishable from racketeering or extortion. Again, to the extent that all or a wide range of parties resort to this source of clandestine funding, this will reinforce the already existing tendency to condemn parties as such as intrinsically corrupt and incapable of combating organised crime.

No one can accurately judge the extent to which external sources of insecurity, hostile states and threatening non-states are influencing the behaviour and status of political parties in Europe. In the not-so-distant past, clandestine “contributions” by the Soviet Union were used to discredit national Communist Parties, just as happened in the interwar period with Fascist and National Socialist “transfers” across borders. In the contemporary context, trans-national financing of partisan and civil society organisations in democratising countries has become an openly recognised practice which does not seem to have brought discredit upon the recipients. Money circulating in the opposite direction, that is from autocratic governments to parties in democratic polities, is quite another matter. The “War on Terrorism” and “the War on Drugs” have focused a great deal of attention on international traffic in clandestine funding, but so far European political parties have been spared any embarrassing revelations.

A European party system: an excursus

European political parties could potentially offer a response to the declining autonomy of the national state and the parallel decline in membership in its political parties. The development of a genuine party system among EU member states would constitute an important step towards creating a European demos, with its distinctive citizenry and electorate. These parties would be unlikely to replace well-established national parties for the foreseeable future, given the asymmetry that persists between the importance and the functions of national parliaments and the European Parliament (EP) – not to mention the intrinsic difficulties in creating partisan identities on such an extensive scale for such a linguistically and culturally heterogeneous population.

The shift of economic and political competencies from the national to the European level has (so far) not been matched by a corresponding shift in democratic legitimacy. EU institutions lack the legitimacy of their national counterparts and the gap between EU citizens and European institutions seems to be growing. Public opinion surveys tell us that many people see the Union’s institutions as remote, bureaucratic and undemocratic. This democratic deficit is aggravated by national politicians who tend to use the European Union as a scapegoat, and fail to explain their own role in adopting European legislation. The resulting lack of a European demos is aptly demonstrated by the large discrepancy between the turnout for national and European parliamentary elections.

The existence of a European demos would first require an increase in the salience of European (as opposed to national) political issues. Sandwiched between the traditional distinction between domestic and foreign policies, the significance of the emerging regional dimension has not been communicated strongly enough to those affected by it. Europeans, if they were aware of how many of the issues that formerly belonged in the category of domestic politics have been shifted to the regional level, might be more inclined to combine across national borders to found and fund genuinely trans-national political parties. As it stands, they are vaguely aware that their interests are structured in European elections and within the European Parliament by “federations” of national parties that have no common platform. This merely aggregates and reproduces in a superficial fashion the different cleavages that emerged historically within each member state, rather than recognise and reflect the cleavages that transcend these national borders.

One reason for this lack of genuine European parties is related to the way that elections for the European Parliament are conducted. They are not organised on a uniform basis across member states. Although in the latest ones all countries have used roughly the same system of proportional representation, the rules for assignment of seats and definition of constituencies are still quite divergent. The elections are not held on the same day, and in some cases they do or do not coincide with local, municipal or provincial contests. The result is what has been called “second-order elections” in which the ostensible purpose is to choose representatives to the European Parliament where they will have to deal with European issues, but the actual process reflects the standing issues within each national state. Euro-citizens, needless to say, are aware of this and use these elections primarily to send a message to their national rulers – often one of discontent, since they can afford to vote for more extreme candidates and parties, knowing that they will not be governed by them. This has produced the embarrassing outcome that incumbent governments and centrist opposition parties tend to fare badly – which can have serious implications for the stability of domestic politics. Another increasingly embarrassing aspect of European-level elections is that they have been characterised by a markedly lower level of voter turnout than national ones. Each successive contest since 1979 has attracted proportionately less voters. This has been the case in virtually every member state, despite the fact that the effective powers of the European Parliament have manifestly increased during the same period.

The political groups within the European Parliament do not and cannot function as European parties. Not only is their composition heterogeneous – the parties from different member states can be very different in social composition and programme even when collected under the same label – but they also have no effective organisational infrastructure. For example, they have little or no role in the selection of candidates for EP elections. Their financing has long been a mysterious matter due to lack of transparency and insufficient monitoring. Expenditures are left exclusively in the hands of national parties that receive subsidies directly from the European Parliament to cover the costs of campaigning. These parties have virtually no incentive to focus on distinctively European issues and, as we have noted above, base their campaign efforts primarily around national ones.

Civil society

Virtually all students of contemporary democracy recognise that the presence of a viable and lively civil society “pressuring” authorities to pay attention to rights, entitlements, interests and causes contributes positively to both the persistence and the quality of modern democracy – and not just in Europe and America. Nota bene that civil society contributes to –but does not itself cause this outcome. It cannot unilaterally bring about democracy. Nor can it alone sustain or improve democratic processes once they are in place. As we shall see, civil society acts along with other institutions and practices – participation by individual citizens, competition between political parties, the legislative process, regular and fair elections for major offices, checks-and-balances between governing bodies, a free and diverse press, autonomous local and provincial governments, the rule of law and an independent judiciary – just to name the most obvious ones.

Before proceeding to an analysis of the present state of civil society in Europe, let us first define it. Civil society is a set or system of self-organised intermediary groups that: (1) are relatively independent of both public authorities and private units of production and reproduction, that is of firms and families; (2) are capable of deliberating about and taking collective actions in defence or promotion of their interests or passions; (3) but do not seek to replace either state agents or private (re)producers or to accept responsibility for governing the polity as a whole; (4) but do, however, agree to act within pre-established rules of a “civil”, meaning mutually respectful and public, nature.

The multiple and varied units of such a civil society may limit themselves by seeking to influence and not to replace elected officials, and by accepting to treat each other respectfully, but their presence in political life is not an unmitigated blessing. In other words, the mere presence of such a mixture of self-regarding interest associations and other-regarding social movements can produce both “public goods” and “public bads”. The European (and American) experience over the long run suggests, however, that the positive effects of civil society far outweigh the negative ones. What interests us is whether, given the challenges and opportunities facing the contemporary societies of Western and Eastern Europe, this will prevail in the future.

The most obvious obstacle to assessing changes in the role of civil society is the continually changing nature of the subject itself. Unlike abstention in elections or public trust in institutions, or shifts in electoral preferences or increases in the number of referendums, when it comes to civil society, the interest associations, social movements and charitable foundations that compose it do not remain fixed in either form or function. With the exception of those organisations whose membership is compulsory and whose interest domain is determined by public law, for example professional “orders”, sectoral “chambers” and some trade associations and unions, most of its units are free to choose whom they wish to represent and how they interpret their mission. This means that their material resources and organisational status are continually at the mercy of shifts in social structure, consumer preference and political purpose. Forms of association that previously played an important, even a crucial, role in political life may gradually decline – hopefully, to be replaced by other kinds of autonomous collective action. For example, an American social scientist drew dramatic negative conclusions – “there is reason to suspect that some fundamental social and cultural preconditions for effective democracy may have been eroded in recent decades” (Putnam and Goss, 2002, p.3) – from the tendency of his concitoyens to “bowl alone”, while ignoring their propensity to seek out and use other means of socialising with each other and articulating politically their shared interests and passions.

Figure 4: Union density (% of economically active population) in Europe, moving average

Source: Eurobarometer and World Value Survey (1995-97)

Let us first take the case of trade unions. There is no question that this form of collective action has had a continual and significant impact on the democratisation of European polities and subsequently upon their everyday politics. Trade unions struggled to enfranchise their members and workers in general and mobilised periodically to ensure that the benefits of public policy would be more equally distributed among citizens. No national history of civil society could possibly ignore them, or the wider democratic effect that they had upon other political parties, interest organisations and social movements.

Figure 4 above displays data on the long-term evolution of membership in trade unions as a percentage of the economically active population in Europe since 1972. All observations have been “smoothed out” by using three-year moving averages and “normalised” to reflect the differing size of countries and the changing composition of the Council of Europe’s membership. According to two alternative progressions (one linear, the other weighted by time), the density of union membership (which was 28% in 2001-4) will be c. 25% in 2010-12 and c. 22% in 2018-20, provided that the underlying socio-economic trends persist and no major changes in public policy intervene. If we take into consideration only those countries for which we have data and that were members of the Council of Europe in early 1970, the picture does not change very much. The trend is still relatively stable and membership density in 2018-20 is projected to be 23% rather than 22% of the economically active population. Comparable data for trade unions in Central and Eastern Europe and the republics of the former Soviet Union are not available, but those that do exist suggest that membership density fits within the previously established trend lines, although at the lower range of variation.

Despite alarming voices predicting the disappearance of the organised working class (or its suffocation by non-unionised workers from the East), our conclusion is more re-assuring, especially when one takes into consideration changes in the sectoral composition of employment (the relative decline of manufacturing where unionisation has historically been greater), the shifting balance of men and women in the active workforce (the former have been easier to recruit than the latter) and the growing proportion of part-time workers (ibidem). For example, the density of trade union membership in the United States fell much more dramatically – from 45% in 1970 to 18% to 1995. Nevertheless, the conclusion seems inescapable that one of the most significant and stable categories of associability within European civil societies will diminish in relative importance – but certainly not “fade away”.

Figure 4 also illustrates a second trend. At the initiation of the time series (c. 1972), the disparity in national densities of trade union membership was of the order of 48 points, from a high of 68% to a low of 20%. Thanks largely to the entry of the Southern European countries into the Council of Europe, this disparity has widened considerably. The most unionised polity had 87% in 2003; the least had 10% – a 77 point difference. Whether this is a temporary “diversion” due to the recent nature of democratisation and, with it, the sudden diffusion of freedoms of association, assembly, speech and petition after a long period of repression by single-party rule, or whether this represents a more deeply entrenched tendency towards “free-riding” and even hostility in neo-democracies to forms of collective action based on class and sectoral interest, is not yet clear. What is clear is that, if it should be the declared policy of the Council of Europe to promote greater convergence over time in the qualities of the respective civil societies of its member states and, moreover, if this convergence should be towards a higher level of performance, this will require a good deal of reform effort.

A third trend among trade unions – more difficult to document – seems to be towards a decrease in their number at each level of aggregation (largely, through mergers) and an increase in the proportion of specialised associations that are members of higher-order federations and confederations. In short, the trade union movement seems to be undergoing a process of organisational consolidation through which its base units are becoming larger in their members and more comprehensive in their scope of representation.

Figure 5: Membership in voluntary associations in Europe, moving average (in %)

Source: Eurobarometer and World Value Survey (1995-97)

So far, we have made the mistake of presuming that the evolution of membership and organisational structure in a single type of association was somehow emblematic of civil society as a whole. Granted that trade unions have historically been of a much greater significance for democracy than, say, bowling societies, it is nevertheless perfectly plausible that other types of associability have been following different patterns. Now, we are about to commit the inverse fallacy, namely to assume that all memberships in voluntary associations are of equal significance. Thanks to the regular surveys carried out by Eurobarometer since 1977 and the World Values Survey of 1995-57, data are available on the proportion of a random sample of the population in twenty-eight countries that report belonging to at least one association. They are displayed in Figure 5 above according to three-year moving averages beginning in 1975. The two points of inflection (1975-77 and 1995-97) again reflect major changes in Council of Europe membership (first, southward and, then, eastward), and in both cases they depress the proportion of those claiming to belong to an association. The summary figure for Europe as a whole (weighted by size of country) is 47% and both the linear and time-weighted projections would be in 2010, 48% (linear) and 46% (weighted) and in 2020, 48% (linear) and 45% (weighted) – ceteris paribus. If one includes only those countries already members in 1972, the corresponding figures are 50% (2003), 55% (2010) and 57% (2020). The “spread” between best and worst performers was 47 points in 1975 and an astonishing 72 points in 2003, if all countries are included, and 46 points in 1975 and 61 points in 2003, if only the 18 original member states are included.

This time the evidence is less preoccupying. Democratisation, Southern and Eastern, seems to have had some downward impact on “primary associability” in Europe, but the overall impression is one of exceptional stability. If nothing changes, those persons in Western Europe who are members of at least one association will even be marginally higher in 2020 than in 2003. Their eastern brothers and sisters may be less “associative”, but their net effect will depress the total by only two or three percentage points.

Let us now take a second look at this same data set by selecting out and distinguishing grosso modo between two types of organisations: first, those that directly provide services and satisfactions to their members (social); and second, those that are more likely to make demands upon authorities that indirectly benefit their members and the public at large (political). In the first, we find groups that provide social welfare, personal health, education, art, music and cultural appreciation, youth, sports, recreation and entertainment. In the second, we have included trade unions, professional associations, local community groups, political parties, movements for human rights, peace, Third World development, resource conservation, environmental protection, gender equality and so forth. At the beginning of our time-series (1974), ostensibly political organisations were proportionately slightly more important (55.0% of the European population reported membership in at least one of them) than social ones (51.2%). By our last observation (2003), the former had declined much more rapidly (to 33.2%) when compared with the latter (39.5%). According to our projections, only 22.8% of Europeans in 2010 and 13.7% in 2020 will be members of any type of political association or movement – again, ceteris paribus. Now, a lot can change during that period. We have reason to believe that participation in such organisations did increase markedly during the 1950s and 1960s, which suggests that some cyclical process may be at work within civil society. But what will provide the incentive for such a turn-around in the future? Our analysis below has failed to detect any “natural” process external to democracy that seems likely to do this. Only conscious and consequent reforms in its internal rules and practices can provide the necessary incentives.

Volunteering to work in an association is not the same thing as being a member of one. It is possible, therefore, that fewer people could be joining and more people could be working in various parties, associations and movements. The data on such “volunteering” is sporadic and subject to wide variations due to seemingly minor changes in the wording of survey questions, but they do point to a gradual increase in most of the countries in Western Europe between 1981 and 1999. Comparable data for Eastern Europe, even for a shorter time period, do not exist. However, no one would be surprised if they showed marked lower levels given the turmoil that has accompanied regime change there.

And now we come to an interesting paradox. Although the data are scattered and difficult to interpret comparatively, they indicate no tendency towards a decrease in the sheer number of associations, movements, societies and foundations. The decline in the proportion of the population reporting membership in at least one of them does not seem to be discouraging “organisational entrepreneurs” from trying to create new units of civil society. Granted that we lack reliable information on those organisations that fail and disappear, but the clear impression is one of net growth in virtually every European society. This suggests that the universe is becoming increasingly specialised. More and more associations, movements and foundations are chasing after members and funds to support ever more specific definitions of collective interest and passion.

And, as we have noted above with regard to trade unions, there is reason to believe that “traditional” organisations representing the interests of social classes, economic sectors and professional specialisations are merging and therefore decreasing in number. The dynamism, therefore, can only be coming from entrepreneurs appealing to new interests and passions – mostly, we suspect, recreational, cultural, educational and social-service oriented, but also including a wide variety of “causes” – environmentalism, human and animal rights, feminism, anti-globalism and democracy itself. It is difficult to document this shift to “new social movements” since their very nature often precludes an accurate count of their numbers or their members. Nevertheless, the increase in “unconventional” collective action by these movements – protests, petitions, boycotts and demonstrations – has become manifest and has transcended the boundaries of national polities. What is much less obvious is the relation of this activity to more traditional forms of democratic participation: voting, party identification, union membership and civic associability. Nor is it clear whether the young people who form the overwhelming bulk of participants in these network forms of organisation will eventually settle down and join the same parties and associations as their parents.

Analytical overview

As one might have expected from its intrinsic variability and constant adaptability, civil society has probably been affected more than any other aspect of democracy by all of the unprecedented challenges and opportunities discussed in Part I. Every one of them seems to be having some impact on either the membership of associations, their composition, their number, their scope or their resource base.

Globalisation. Here, the major difference has been the spread of trans-national non-governmental organisations, especially those advocating a wide range of causes from democracy and human rights to environmental and gender issues. The impact has been particularly great in the new democracies to the East where the relative importance of financial resources and conceptions of passion and interest has been more disproportionate. In the more established Western democracies, the focus of these NGOs has often been on globalisation itself and its economic, social and environmental impact upon an increasingly well-educated citizenry sensitive to the dilemmas of “complex interdependence”. There is hardly a government in Europe that has not had to face pressure from organisations whose human and material resources come from beyond their borders and whose networks of influence penetrate deeply into what had previously been a relatively autonomous realm of national politics. Whether this narrows the range of policy responses, or widens the potential resources that can be brought to bear on such complex issues, remains to be determined – but the outcome will have a significant effect on the effectiveness and legitimacy of rulers at both the national and the supranational levels.

European integration. EU directives and regulations have affected the civil societies of member, candidate and adjacent states and even led to the emergence of an embryonic European civil society. Again, the greatest impact has been on the neo-democracies to the East, especially those struggling to meet the obligations of the acquis communautaire and competing for funds from the various EU programmes. In a few policy areas, such as agriculture and regional funds, exerting influence at the European level has become imperative, whereas in most cases, associations and movements tend to work through their respective national authorities. EU policies have also opened up unprecedented opportunities for direct access to large trans-national enterprises. The overall picture is, therefore, mixed: pluralism for specialised functional interests and selected causes through a proliferation of points of access in this emerging “multilevel” and “polycentric” polity and corporatism for those at the supra- and national levels with privileged resources or special access to specific agencies. Particularly striking has been the re-emergence of national systems of policy concertation in response to the twin challenges of a single European market and monetary unification.

Inter-cultural migration. One of the major challenges to European civil societies has been their response to increasing numbers of migrants, asylum seekers and refugees from outside Europe. Historically, these “aliens” were first assimilated into national cultures and then entered into pre-existing trade unions, professional associations and other intermediary organisations. When they did form associations or movements of their own, it was usually assumed that this was merely a “way-station” to their eventual integration. What is distinctive about the contemporary situation is the presence of large numbers of foreign residents who insist on their right to remain different – and, therefore, to create their own civil societies. They demand, not only that their organisations be recognised, but also that they be accorded access and influence. Making this process even more contentious is the fact that aliens often come from countries that are deeply divided internally, if not subject to endemic violence. Whether these claims to persistent difference will provoke an “uncivil” backlash among natives in their existing parties, associations and movements or whether they will contribute to a pluralistic diversification of patterns of political competition and social tolerance is one of the most difficult things to predict in contemporary Europe.

Demographic trends. The impact of this trend is relatively easy to assess. Older people are becoming a more and more important component of existing associations, especially trade unions, or they are forming specialised organisations representing the interests of retired persons. Young people are less and less likely to join these pre-existing associations (or to participate in politics in general) and are providing more and more of the dynamism for the “unconventional” behaviour of new social movements. The result is an increasing imbalance in the distribution of organisational capacity and a less homogeneous mix of political strategies across generations and, therefore, a tendency towards public polices skewed to benefit the aged and a growing resistance to fiscal and other reforms intended to redress this imbalance. Unless the politically disaffected youth of today can find stable niches within the reformed national or even supranational civil societies of the future as they mature, regime effectiveness and eventually legitimacy is bound to suffer.

Economic performance. All European societies, even the most impoverished ones to the East, have sufficient human and material resources to sustain a large number and variety of civil society organisations. High levels of unemployment, no doubt, depress individual participation and place high demands on service organisations, but this is often compensated by increased voluntary work and contributions from those more favoured by the conjuncture. Governments, also, come to depend more on intermediary bodies for the implementation of social programmes and this can increase associational revenue. The fact that, generally speaking, European economic performance has been inferior to that of the United States during the recent decade seems not to have had much impact on their respective civil societies. If anything, this has only brought out the contrast between the quality of life in Europe and America, where the higher levels of social solidarity and community organisation tend to favour the former.

Technological change. None of the challenges/opportunities has had a greater impact upon civil society than this one. Many of its organisations have seized upon the innovations in ICT and even become agents for their diffusion throughout the rest of society. The cost and ease of contacting members and soliciting their support has been dramatically reduced. Networks tying together previously separate local efforts across large distances and especially across national borders have been formed and even proven efficacious in co-ordinating the behaviour of activists at the level of Europe as a whole. Which is not to say that ICT has been an unmitigated blessing. It is not yet clear the extent to which the time spent “surfing” the Internet detracts from the time that individuals, especially the young, spend interacting with each other. Solicitations mailed by post or sent over the Internet have been successful in creating and funding a vast number of “virtual associations and movements” whose members never meet each other and who have little or no knowledge of or control over what their leaders do in their name. Many of these organisations are dominated by their professional staff and are run similarly to profit-making firms – with “customers” receiving selective goods or services in exchange for their contributions.

State capacity. In several of the new democracies, the main issue has been whether, along with their transition from authoritarian rule, there has also come a change in geographic boundaries and collective identities. The break-up of former multinational states has brought with it the problem of a plurality of civil societies within the same political unit – and the prospect of quite “uncivil” relations between them. In some instances, this has been resolved peacefully by a process of mutually acceptable secession – but even then there usually persist serious cleavages between the new “titular” national majority and various national minorities. But for most of Europe, the issue has been quite the inverse: how can well-established national civil societies cope with a marked decline in their state’s capacity to carry out effectively and autonomously the tasks that are expected of it by citizens. Here, the problem is not national disintegration but international integration. What can the organisations of civil society do when the state they have been seeking to influence becomes part of a larger process of “pooled sovereignty”? The simple answer is to reorganise across national boundaries and expand the scale of collective action. Unfortunately, this may mean overcoming deeply entrenched differences in national culture, language and organisational format – and the resulting “European civil society” can be much less efficacious and skewed to favour particular interests and passions than were the previous national ones.

Individuation. If this challenge were very serious, there would be no opportunity for a civil society. If every individual citizen had different working conditions, living contexts, patterns of mobility and family situations, the probability of acting collectively and voluntarily with others would diminish greatly. Fortunately, this has not happened and human beings seem to have an intrinsic genius for discovering new goals that they have in common. It does mean, however, that some of the large (“encompassing”) socio-political categories based on class, race, religion, ideology and nationality have yielded to much more fragmented and personalised conceptions of self-interest and collective passion. Presumably, this helps to account for the continued pace of forming new associations and movements with more specialised objectives and to the gradual decline in more traditional forms of associability such as trade unions. One clear implication of this transformation is a diminished probability for reaching widely-embracing “social contracts” and, hence, a more unruly and less predictable pattern of bargaining between interests and passions. It also helps to explain why political parties have less and less connection with associations and movements and have lost much of their historical function of aggregating citizens under broad “ideological” labels.

Mediatisation. Previously, the units in civil society played an important role in providing their members and followers with political information and, thereby, helped to form their conceptions of interest and identity. Nowadays, the mass media – especially television – has usurped this function and whatever specialised information is offered by associations and movements usually has to compete with rival commercial sources. The “party press” has virtually disappeared and the newsletters and broadsheets from trade unions and professional groups have less and less circulation. The Internet, as we mentioned above, may be offering them novel and less expensive means to get their messages across, but the competition for attention is ferocious and the audiences are much less captive than in the past. Commercialisation may trivialise (and scandalise) information about politics, but has also contributed to liberating citizens from partisan manipulation and government propaganda. For the (unfortunately declining) number of them who wish to participate in an informed way in the processes of democratic accountability, there are many more sources than in the past and accessing them is easier and cheaper, but they do not involve the opportunity for direct, inter-personal exchange and deliberation that used to characterise the “public sphere”.

Sense of insecurity. Here we discover another paradox. In the past, nothing was more productive of associability than the most threatening form of insecurity, namely, international war. During both the first and second world wars, membership in a wide range of political and social organisations went up dramatically and many new organisations were founded during and immediately after these episodes of large-scale violence. Now that the Cold War is over, and Europe has effectively established an “international security community” within the region, that is the countries within it have no realistic expectation that their disputes will be resolved by armed force or reason to go to war with each other, this powerful impetus for the development of civil society has been eliminated. It is only the perception of avoidable risks and of their probable consequences from one’s own neighbours that gives rise to new forms of voluntary collective action. Not only is this a weaker incentive, it is also a divisive one. Its most manifest expression in contemporary Europe is the mobilisation of natives against foreigners – and of these legal and illegal aliens to protect their persons and rights.

Decision making

“Guarding the guardians”

In contemporary European democracies, a number of non-democratic or non-majoritarian institutions play an increasingly important role. In this section, we will focus on so-called “guardian” institutions (institutions made up of experts) and on the development of network, regulatory and multi-level “governance”. By “governance” we refer to patterns of decision making involving various public and private actors whose actions are not solely co-ordinated through hierarchical and/or market mechanisms. Outside and in between these two traditional mechanisms of allocation, there are a variety of new modes of governance that make use of different steering arrangements to produce policy outcomes.

Democratic legitimacy, guardians and governance

In modern societies, political legitimacy requires that matters of public and common interest should be decided democratically. For a system of governance to be considered democratic, the opinions of the citizens must be freely represented, so as to be listened to and accommodated by rulers, who in turn need to be accountable to citizens in their actions and decisions. It is an important aspect of democratic legitimacy that citizens believe they have a fair chance of influencing the outcome of the decision-making process on issues that affect their own life chances.

This does not mean that all binding collective decisions have to be taken democratically. Although of considerable economic importance and social relevance, many decisions are considered private and as such are left to individuals, families and associations. Some of these decisions are left to voluntary contractual arrangements between the affected parties, while others are made through more automatic mechanisms of co-ordination such as the market. Moreover, much of the ordinary life of citizens takes place in organisational structures of work, profession, education, worship and recreation, which usually operate according to hierarchical principles of allocation. The reasons these areas of social decision making are insulated from democratic criteria and procedures are linked to privacy, organisational efficiency and/or complex co-ordination. The particular application of these reasons is often disputed, but nonetheless widely accepted as part of democratic life.

However, non-democratic decision making extends to many public institutions such as the legal system, the police, the army and the public administration, which are normally organised according to hierarchical criteria. This is due to functional reasons and to the complex organisational nature of these institutions. Nonetheless public institutions are not entirely autonomous, nor do they operate exclusively according to self-referential rules. In dealing with public matters and when supported by public funding, they need some form of democratic legitimation. This is guaranteed from the outside by their subordination to governments and parliaments.

In the last twenty to thirty years there has been a steady erosion of the scope of democratic decision making. This has happened from the inside and the outside of politics. Inside, constraints arise due to guardian institutions’ addressing policy and regulatory problems by relying on specialised knowledge and on experts who are insulated from partisan competition, public opinion and majoritarian decision making. Outside constraints appear as public policies are increasingly decided through agreements within complex networks of governance, comprising public and private “stakeholders”, but not the citizenry as such. As a result, there is a decrease in political responsibility and public accountability.

The general result of these developments has been a shift in the balance of public and collective decision making from politics to administration, from democracy to technocracy, de facto if not always de jure reducing the space for the voice, influence and control of citizens, whether acting directly or indirectly through their representatives. The shifts have also been promoted or sustained by a number of democracy's intrinsic tendencies, such as oligarchy, functional autonomy, corruption and professionalisation.

Tendency towards oligarchy. The “iron law” of oligarchy clearly favours the ascendancy of guardian institutions. More than political parties and representatives, they are exempt from direct public scrutiny and as such are not publicly accountable.

Tendency towards self-referential autonomy. This tendency concerns politics as much as other areas of social life, since increasingly decision making requires specialised knowledge and expertise. Instead of general rules equally binding on all, the political process becomes more and more fragmented into specific functional tasks, each with its own logic and needs. As a consequence, particular clusters of organised interests become the unique point of reference for guardians appointed to regulate their behaviour and these private groups tend to “capture” their guardian by manipulating asymmetries in information and power.

Tendency to professionalisation. This is a development that is shared between guardian rulers and the political class. Virtually by definition, guardians owe their role to the need for forms of knowledge produced exclusively within certified professions: lawyers, economists, system theorists, managers, accountants, military officers, social scientists, and so forth. As we have seen in our analysis of trends within parties, politicians are also becoming increasingly specialised or, if not themselves, increasingly dependent on those who are specialists: consultants, public-opinion pollsters, media advisors and so forth. While guardian institutions are supposed to be independent to various degrees, they are also supposed to be subject to governmental control and pressure. When both sides of the equation are more professional, the most obvious result is a tendency to exclude amateurs, that is most of the affected population, on the grounds that they are insufficiently informed or conscious of what is needed to produce “good” functional performance.

Tendency to corruption. Precisely, the fact that guardian institutions are removed from public pressures and insulated by professional expertise makes them unusually vulnerable to the influence of corruption. Some of this is intrinsic to the way in which specialised and segmented decision making is virtually designed to ignore externalities and unintended consequences. What seems rational and functional to those directly involved, seems arbitrary and exploitive to those indirectly affected. More important, however, is the fact that guardians are in the business of deriving rules and handing out licenses. This creates very tempting opportunities for rent-seeking, that is acquiring an advantage over competitors or even a monopoly status that can be converted into exceptional profits – some of which may even be returned to the specialised rule-makers/licensing authorities.

Table 1 below summarises the way in which the ten challenges and opportunities identified in Part I of this Green Paper provide a context for the development of non-democratic forms of governance. It groups the “challenges and opportunities” into four main categories. The first is concerned with the effect of globalisation of governance and the decline in state sovereignty. The second emphasises the increasing porosity between the private and public domains. The third suggests the difficulties for democratic politics when dealing with increasing levels of social differentiation. The fourth and final one describes the effect that new technologies (but also increased risk and insecurity) have on the relationship between state and private powers, on the one hand, and citizens, on the other.

Table 1: Changes in the external context: impact on governance arrangements and guardian institutions

Challenges and opportunities	Globalisation European integration State capacity	Globalisation European integration State capacity Economic performance	Inter-cultural migration Demographic trends Individuation	Technological change Mediatisation Sense of insecurity
Guardian and governance	Processes of globalisation and internationalisation have contributed to give a greater role to guardian and governance institutions. State power is diminished in so far as it needs to co-ordinate policies with other states and with powerful private corporations at both the national and international level.	The combined effect of these challenges is to alter the relationship between public and private, reducing the steering and sanctioning role of public institutions. To be effective, political institutions need more flexible policy instruments, aimed at changing behavioural patterns where they cannot use command-and-control strategies.	It is more difficult for democratic institutions to address increasingly differentiated needs and attitudes. Problem-solving institutions are considered more effective in dealing with diversity.	Changes in the context of public information and communication provide new opportunities for state and private dominance. But new technology also makes it more difficult for one single power to control the flow of information. New technology lowers transaction costs for obtaining information and for open government.

Analytical overview

Both intrinsic tendencies and external developments tend to favour the proliferation of guardian and governance institutions. The effect that these institutions have on democratic influence and accountability is similar to that produced by traditional administrative and bureaucratic institutions. They extend the chain of delegation, so that the longer the chain, the feebler the voice of the citizens. They tend to control information and act as though they had a monopoly on knowledge and expertise in a particular area. They are not directly accountable since they are not affected by electoral discipline. However, these new institutions are even more independent from political power than traditional bureaucratic agencies; they are set up to avoid politicisation and, in the case of network governance, the fragmentation of political responsibility makes them less accountable.

In the face of the increasing power of such institutions in public decision making, the future of democracy will depend on the way we deal with issues raised by the following questions:

1. Can the apparent loss of democratic legitimacy be compensated by other forms of legitimacy underlying “guardian” and “governance” institutions?

2. Can non-majoritarian institutions of guardianship/governance be reconciled with and justified by reforms in democratic practices?

a) The role of non-democratic decision making in democratic society

In addressing the first question, we need to identify the justifications normally given for delegating policy making to non-democratic institutions. As in the case of public administration, the legal system and the army, the main justification has been the need for organisational efficiency. But this is a rather broad category, which may not apply to all guardian institutions or to network governance in general.

From an analytic perspective, the reasons given in support of non-majoritarian institutions reflect the demands faced by public decision making in developed societies and the standards required for public policy making. The main demands are those of complexity and specialised knowledge. The standards are those of feasibility, effectiveness and efficiency, respect for diversity (of needs or identity), respect for diversity (in application), private autonomy and enterprise. These demands and standards shift the balance of political legitimacy from one based on democratic participation, access and accountability to one based on the superior performance of functions and satisfaction with improved output.

Table 2 below identifies the more specific grounds on which the functional and substantive legitimacy of guardian and governance institutions rests. It organises them according to the type of reasons (demands and standards) underlying their functions, and according to the kind of constraint (from the inside or the outside) that they impose on democratic politics.

This table shows that output and functional legitimacy require institutions to operate in place of citizens instead as representative of citizens. But this seems to imply that modern democracies may be faced with a trade-off between institutions that promote democratic legitimacy and institutions that promote output and functional legitimacy. As a result of this, the balance of power is now decisively tilting on the side of non-democratic (and potentially oligarchic) institutions, eroding citizens' sense that they can influence collective decision making.

Table 2: Reasons supporting non-democratic legitimacy

	Demands Complexity; Specialised knowledge.	Standards Feasibility; Effectiveness/efficiency; Diversity (of needs and identity); Diversity (in application); Private autonomy and enterprise.
From inside the political system	Institutions protecting democracy; Decision making affecting highly specialised areas.	Institutions protecting minorities; Institutions protecting individual citizens.
From outside the political system	Impartial decision making; Complex co-ordination; Promoting competition and background conditions; Supranational co-operation.	Impartial decision making; Market control; Promoting competition and background conditions; Supranational co-operation.

b) Bringing democracy back in

There is another way of looking at guardian and governance institutions, not just from the from the standpoint of their underlying reasons, but from the more specific perspective of the kind of functions they perform in relation to the political system and to citizens' interests and welfare. This is a more promising perspective from which to address our second question, that is whether it is possible to reconcile mechanisms of guardianship and governance with democratic legitimacy by reforming the practices of “real-existing” liberal democracies. Table 3 below is concerned with what these institutions do in relation to public decision making.

Table 3: Types of non-democratic decision-making institutions

Institutions inside the political system	Institutions outside the political system
Institutions for the implementation of public policies	Regulatory institutions
Institutions operating as checks on the political and administrative system	Self-regulatory institutions
Semi-autonomous institutions, operating in sectors of great public concern	Networks of decision making

Table 3 argues that the tendency towards the bureaucratisation and rationalisation of politics identified by Max Weber already at the beginning of the 20th century are no longer exclusively embodied in the traditional ministries and agencies of public administration, but increasingly to be found in the growing number of guardian institutions and the spread of networks of governance. This shift from politics to administration (from the conflict and compromise approach, to the problem-resolution and policy implementation approach) is accentuated by the need to avoid overloading the political system with legislative and regulatory tasks that have become too extensive in modern complex societies. It also reflects the “blame-shirking” attitude of politicians, who tend to delegate policy-making functions to non-democratic institutions in those areas where policy success is difficult to establish and policy results cannot be easily translated into electoral assets.

But if neither the citizens nor their representatives have control over these new institutions, the question is how to ensure that the “guardians” do not overstep their duties by exploiting their privileged position to their own advantage. Who, ultimately, guards the guardians? Quis custodiet ipsos custodes? Table 3 implies that there is no “one-best” or “all-encompassing” solution, since different guardian institutions and governance networks perform different functions and, therefore, require different strategies aimed at reconciling democratic and functional legitimacy.

Table 4 below suggests two general strategies that could be used to address this problem. One is more direct and is aimed at re-introducing forms of democratic control and accountability; the other is more indirect and works though checks and balances.

Table 4: Strategies to bring democracy back in

	Direct strategies	Indirect strategies
Polity-based	Subjecting guardian institutions to the direct control of democratically elected bodies	Promote a system of horizontal checks based on reciprocal vigilance between guardian and democratic institutions
Citizen-based	Devise mechanisms, other than electoral control, which guarantee popular participation and control	Promote institutions that operate a vertical check over political institutions by allowing for citizens' voice

Inter-level accountability

Experimentation with multilevel forms of governance is on the rise in Europe – in part, due to the devolution of powers to regional or provincial governments; in part, due to the European Union’s demonstration that national sovereignty can be parcelled-out and pooled to the benefit of all levels. Yet democratic ideals are challenged by these experiments in the scale of governance. How can politicians be held accountable? How is it possible to square the norm of one person, one vote with sub-units of different size demanding equal voice? How does one settle the issue of which decisions should be taken by which demos, at what geographical level – and who should decide the inevitable conflicts that arise from such a complicated system?

Multi-level governance and decentralisation challenge democratic norms of accountability of politicians and other authorities at various levels because such systems tend to blur the opportunity spaces for political choice enjoyed by each level. Measures for regaining accountability include more transparency and political contestation concerning decision makers, both with regard to their de jure powers and their de facto ranges of choice.

Analytical overview

a) On federalism and other multi-level systems of governance

“Multi-level governance” is a term often used to describe the plurality of decision making modes within the European Union. Multi-level may refer to the “vertical” dispersal of political authority from the state upward to a supranational – European Union – level and downward to sub-national/regional levels; and/or the “horizontal” dispersal that is involved when non-state actors are brought into the process. These raise different normative challenges concerning such issues as democratic representation and accountability, often because the alleged virtues of dispersed governance come at the cost of transparency, circumscribed competencies and accountable authority.

For our purposes, federal political orders can be characterised by a (quasi-) constitutional division of powers between central bodies and sub-units where each level enjoys final authority with regard to some functions and where changes in this distribution of authority requires consent. In contrast, in decentralised systems the central authorities can maintain, modify or abandon lower level authorities at their discretion. In confederal arrangements, sub-units typically can veto decisions and even leave the confederation. Starting with the European Coal and Steel Community, European Union institutions have had both federal and confederal elements.

To be sure, the EU may never become a complete federation with a comprehensive division of powers, but one effect of the EU draft Constitutional Treaty is that, if ratified, it would add more federal elements to the “mix” since member states will have signed away their right to veto decisions in more policy areas.

b) Subsidiarity

One of the most vexing issues within any federation or quasi-federation involves the formal (usually constitutional) allocation and use of competencies across its multiple levels of political aggregation. The principle of subsidiarity purports to resolve this issue by placing the burden of argument with those who seek to centralise authority. Sovereignty can be pooled in response to the loss of effective governing capacity by smaller sub-units, but higher-order authorities at the national or supranational levels can only act legitimately when they contribute to satisfying the objectives of citizens better than the sub-units. The application of this principle has had many different and competing interpretations, ranging from modern Catholic thought, an ancient tradition associated with Althusius, doctrines of “concurrent majorities” linked to the dispute between the North and South in the United States, to such contemporary sources as fiscal federalism and liberal contractualism. These interpretations differ on such fundamental matters as the proper objectives of the political order, the weighting of sub-units of different size and capability, whether these units should be defined in territorial or functional terms, and what sorts of inter-level protection and subsidisation are most effective against which risks. For example, is tyranny of the central authorities over a sub-unit worse than that of local authorities over a local minority? At what point does better performance of some functional task outweigh the threat to territorial identity and autonomy? Who should be the ultimate judge when conflicts arise over the application of standards – the rulers of sub-units or those of the central one? Most important for our purposes is the question of who should be held accountable (and how), especially when many decisions may involve more of a sharing of competencies than a separation of them.

Ultimately, much of the appeal of the principle of subsidiarity rests on our shared interest in liberty, that is on the ideal that one should not be subjected to the arbitrary will of others. Giving and protecting the veto right for all sub-units would protect liberty by ensuring that joint gains do not come at the price of despotism, but they would also leave the polity as a whole at the mercy of the single most recalcitrant sub-unit. Others argue that decentralisation to smaller groups that share policy preferences, personal values and/or material circumstances makes for more efficient decision making, but precisely because of this sharing such units may not have the requisite volume or variety of resources to tackle the problem at hand. Devolution of powers can also prevent the decision-making process from becoming overloaded; but it might also make that process more parochial and oligarchic. And for certain public goods with lots of “positive internalities” (also know as “synergies”) there may be subsets of individuals who should be allowed to form “clubs” for their provision, provided that they do not exclude minorities from the benefits or pass the costs on to non-members.

c) On accountability in multi-level systems of governance

It is often difficult to determine who bears ultimately the responsibility for a particular policy decision when more than one level is involved in making and implementing that decision and when each level of authority can pass the blame on to the other. Given such a complex structure, no formal system of “multi-level diplomacy” may be capable of satisfying the democratic requisite of “accountability for acts in the public realm” and of bringing the necessary sanctions to bear on those who operate in the interstices of the various levels. Nevertheless, from a normative perspective, it is possible to specify the generic standards for evaluating such complexity.

Transparency vs. opacity. Assisted by public media, citizens and authorities should be able to determine whether institutions and their decisions roughly match whatever normative requirements are appropriate for that complex political order. An added cause of opacity in the EU is that many of the processes thus far have proceeded without public access to government negotiations in the Council of Ministers. In addition, the shift from unanimity to (qualified) majority voting limits accountability even more, since this enables politicians to say that they were unsuccessful in voting against unpopular decisions. Their claims that “Brussels made me do it” cannot easily be checked.

Security vs. insecurity. Unanimity offers protection to citizens of each member state by ensuring that they would not be forced to take part in arrangements contrary to their own interests, and allows them some protection against one-sided agreements. Yet this decision rule also increases citizens’ uncertainty and vulnerability, since each sub-unit may block common decisions. (Qualified) majority voting on the other hand increases the need for trust and trustworthiness among individuals and among their representatives, requiring the latter from time to time to adjust or sacrifice their own interests and those of their voters for the sake of other Europeans. The majority must then be trusted to consider the plight of minorities, and to respect common decisions when they find themselves in the minority.

Autonomy vs. equality. There are tensions between respecting sub-unit autonomy and securing rough equality of living conditions across sub-units – often regarded as a condition and/or an objective of democratic politics. A central issue is therefore what range of outcomes and policies the sub-unit population should be responsible for in the sense that they should bear the full economic burden of their collective choices. The EU is supposed to “promote ... economic and social cohesion and solidarity among member states” – while respecting member state autonomy. But equalisation and solidarity may require centralisation of monetary, social and fiscal policies – according to the principle of subsidiarity – leaving little authority to the sub-unit.

One person, one vote vs. one sub-unit, one vote? Attempts to “democratise” federations with sub-units of unequal size can run afoul of democratic ideals: Should one person, one vote or one subunit, one vote prevail? That is, do democratic norms require majoritarianism or can one justify that small sub-units should be over-represented, for instance to reduce the risk that their citizens’ interests are regularly overruled? Such overrepresentation often occurs in federations, and might be defended also in EU institutions where less populous states are over-represented or their voting weight is disproportionately high compared to more populous states. It is not obvious that majoritarian decision making is appropriate when segments of the population risk being in a permanent minority, especially if the majority cannot be trusted to always modify their views in light of the impact on minorities. One person, one vote may not be appropriate under such circumstances.

Mechanisms for direct citizen consultation

Whereas various forms of representative democracy constitute the basic foundation for decision making in all European democracies – old or new – some polities have introduced mechanisms for the direct involvement of citizens into their repertoire of democratic institutions. In almost every European country and at almost every layer of government, citizens can file petitions that neither bind parliaments nor result in popular votes. Such petitions are bottom-up, superficial and non-threatening manifestations of deeper-rooted social dissatisfaction and conflict. They are usually channelled by established political organisations (parties, associations or movements), but they occasionally arise from ad hoc and informal units of collective action. Their primary goals are to attract the attention of rulers and to provoke public debate among citizens. Since the success of such petitions remains entirely at the discretion of those in power, they are merely an upward channel of communication, along with several others offered by modern liberal democracies, such as public opinion polling and public hearings. Presumably, some petitions are more effective than others, but none of them can be described as a regular and effective means for holding rulers accountable.

In the standard version of liberal democratic theory, the device par excellence for ensuring accountability is supposed to be “free and fair”, “regular and competitive” elections. Whether periodic elections alone are sufficient for this purpose has long been a matter of dispute – at least, ever since Jean-Jacques Rousseau observed caustically that the English were only free once every few years, on the day they voted. Regardless of the argument in theory, in practice most “real-existing” democracies in Europe have designed and implemented other means for constraining the behaviour (and, occasionally, the tenure) of their rulers. Today, what has come to be called “direct democracy” complements “indirect or representative democracy” in virtually every member state of the Council of Europe.

Direct democratic institutions take two fundamental forms: the referendum and the popular initiative. Both are closely related to each other. The referendum encompasses a process through which proposals by political authorities may be submitted to a popular vote. The popular initiative is a process through which a number of citizens may formulate a proposal and force the political authorities to submit it to a popular vote. In the latter case, there is a special form of initiative, the recall, that permit it to be applied not to a specific policy proposal, but to the tenure in office of a specific elected official. Thus, it is the originator of the activity that allows us to distinguish between the two basic forms, not its content or purpose.

The virtues and vices of both the referendum and the popular initiative have long been debated by political philosophers and normative democratic theorists. Arguments in favour and against have waxed and waned over the past centuries. Especially controversial has been the notion of the recall – perhaps, since it is by far the strongest in terms of its relation to accountability. As we have just seen in the case of California, it can be employed to remove a perfectly legally elected official who enjoyed majority support in the equally legally-elected parliament. Regardless of who is winning the abstract debate over direct democracy, in concrete terms it has become more and more of a reality in recent decades in Europe.

The main challenge direct democracy presents to representative democracy is that it introduces an additional potential veto to all of the checks and balances that may already be built into the usual indirect system of representation. Decisions made by legitimately elected representatives can be altered or simply abandoned. Moreover, parliament loses (or finds reduced) its traditional sovereignty within the democratic polity since a favourable decision on a popular initiative can produce a generally binding decision or compel the parliament to produce such a decision. If – to use Gordon Smith's terminology – “anti-hegemonic outcomes” of popular votes were to become the rule, democracy would run the risk of deadlock. From a normative perspective, this would certainly be Pareto sub-optimal. Everyone could run the risk of losing or, at least, it would be more likely that very few decisions benefiting everyone would get passed.

The inverse situation is hardly more appealing or likely. If it were the rule that the choices made by representatives and those made by citizens were identical – or, at least, systematically congruent, both the referendum and the initiative would be superfluous. In such an ideal world, where all the principles of democracy would be perfectly respected by all actors and where political deliberation would result in perfect information and shared political competence among all citizens, congruence in policies (and consensus in opinion) would be the natural outcome and, therefore, it would not make much sense to hold popular referendums on topics when rulers and citizens systematically and predictably come to the same conclusions.

In any case, such a perfect democracy remains a chimera. Even finding a stable coincidence between a majority of representatives and a majority of the citizenry can be difficult. Indeed, systematic congruence when it manifests itself formally is less likely to be healthy for democracy and more likely to indicate autocratic control by the ruling elite over “its” compliant and frightened subjects. For example, Ceauşescu's 1986 referendum in Romania produced a perfect 100.0% total of yes-votes with a turnout of 99.99% – and this triumphantly plebiscited ruler was overthrown to the general delight of the Romanian public only three years later.

Despite an ideology that stresses the congruence between ruler choices and citizen preferences, that presumes that elected agents are unambiguous reflections of voting principals, the practice of all “real-existing” democracies suggests that it is normal (or, certainly quite frequent) that the two sets of democratic actors are not in synchrony. Partly, this is a product of differences in time horizon or in the definition of constituencies; mostly, however, it is the result of the inevitable tradeoffs, compromises, “log-rolls” and “package deals” that are an integral part of how representative democracy functions. Moreover, as politicians become more professional and remain in office longer, they are bound to learn better how to make these deals (and to remember more clearly past ones). They are also likely to become more skilled at explaining to their constituents and voters the rationale behind such incongruities.

It seems safe to assume that, between the professionalisation of politicians and the complexity of collective choices in contemporary multi-level democracies, the incongruities have grown and that the sentiment that this generates among citizens is one of the many elements contributing to widespread discontent with rulers and distrust of representatives. Inserting forms of direct democracy as a complement to representative democracy is considered one way – perhaps, the best way – of filling that gap. One of its more subtle virtues is that both the referendum and the initiative can have a considerable impact even when they are not formally utilised. Just the anticipation that either can be initiated – by the government, the opposition or a sufficient number of citizens – may be enough to deter rulers from passing measures that they know to be incongruent with the preferences of a heretofore passive majority.

In many polities around the world, the initiation of the referendum process is strictly controlled by the head of state or government and used (and abused) to call voters to the polls when there can be no question of their voting “No” – out of fear or manipulation. In many post-colonial, single-party African states the referendum has been wielded as a tool by rulers to acquire democratic legitimisation for their decisions – if and when it pleases them. There have even been applications of this “plebiscitary” instrument by governments in Western Europe in the not so distant past. Such an ad hoc (and, at times, ad hominem) version of direct democracy is in stark contrast with its more formalised and predictable counterpart elsewhere. In such European countries as Switzerland, Liechtenstein, Italy, Ireland, Denmark, the post-communist democracies of Central and Eastern Europe and most of the former republics of the Soviet Union, the government in office cannot control the initiation of a referendum. For example, if their constitutions are amended or if they decide to join the European Union or other major regional or global organisations, a referendum has to take place and the rules for holding it are fixed well in advance. Initiatives require a pre-established number and distribution of signatures by citizens, but this cannot be manipulated ex ante or voided ex post by those in power.

Even a casual observer of recent European politics is likely to be familiar with several national referendums that had a major impact on such issues as abortion, atomic energy, joining the EU and so forth. What is often overlooked is the crucial political role played by referendums (and, more rarely, initiatives) at the local and regional levels in well-established democracies. The normal pattern in Western Europe has been for direct democratic devices to emerge and be experimented with at these levels and only subsequently to be transposed to the national level. In Central and Eastern Europe and in the republics of the former Soviet Union, however, the abrupt regime change from communist autocracy to liberal democracy brought about an immediate expansion of popular rights at the national level, before any prior experimentation in local or regional governments.

Popular votes in Europe: assessing the evidence

Between 1960 and 2003 in all current Council of Europe member states, citizens at the national level were asked to make 628 direct democratic decisions. These data show that over the past four decades, national referendums and initiatives took place in thirty-nine of the forty-five member states. If we were to include data from the local and regional level, Luxembourg would be the only European country never to have consulted its people directly through these devices. However, even Luxembourg held three national referendums before 1960 and its government has recently announced that that a referendum will be held on the draft EU Constitution Treaty. All European democracies have been exposed, at least sporadically, to popular votes – even if the frequency, form and effect of these consultations have varied greatly.

More than half of all referendums and initiatives in Europe since 1960 were held in Switzerland. This country is the world's champion par excellence of direct democracy and, in order to acquire this well-deserved reputation, it has followed a very distinctive political trajectory. For these reasons, we have chosen to exclude this outlier from most of the subsequent analysis.

Aside from Switzerland, Liechtenstein and Italy, only five European countries have had ten referendums or more over the period considered. Excluding Switzerland, the average per country has been seven. Only one quarter of all Council of Europe member states have exceeded this average. In other words, most member states have had relatively few referendums at the national level.

Figure 6 below displays the evolution of the frequency of popular consultations since 1960, allowing us to observe their cross-temporal dynamics within Europe. The 2000-09 data point is a linear projection based on the data for 2000-03.

Figure 6: Overall tendency of direct democratic votes in all Council of Europe countries and with Switzerland excluded

Sources: Research and Documentation Centre on Direct Democracy (c2d) at the University of Geneva and “Suchmaschine für direkte Demokratie” developed by Beat Müller at the Swiss Federal Institute of Technology in Zurich.

The two lines in Figure 6 show similar patterns. Including Switzerland merely doubles the numbers without significantly affecting the S-shape of the basic curve. Both show a dramatic increase in direct citizen consultations during the 1990s. The overall frequency tripled during this period. Presumably, this captures a critical juncture in European politics, during which referendums and initiatives became especially attractive as a conflict-resolving and legitimating device. Since then, however, the frequency has levelled off and our projections suggest that, for the first decade of the third millennium, their number will probably not increase. One hypothesis is that Europe may have reached a “saturation point” with regard to direct democratic decision-making, at which some equilibrium between it and representative democracy is established.

Figure 7: Overall tendency of direct democratic votes in Western and Eastern and Central Europe (all Council of Europe countries without Switzerland)

Sources: Research and Documentation Centre on Direct Democracy (c2d) at the University of Geneva and “Suchmaschine für direkte Demokratie” developed by Beat Müller at the Swiss Federal Institute of Technology in Zurich.

One potential proof of this concerns the pattern of evolution between the old and new European democracies. In Figure 7 above we can observe a clear convergence in the frequency of holding referendums in Western and Eastern Europe since the end of the Cold War. For the West, the massive increase during the 1990s can be primarily explained by a series of EU integration referendums. In the East, it was popular consultations with regard to national sovereignty and constitutions that explain the increase. According to our decennial projections, we expect to have about as many referendums during the first decade of the third millennium in both regions of Europe. One reason for this continues to be European integration. As long as its member states continue to agree upon new treaties, their ratification will inexorably promote the referendum experience. For example, the introduction of the euro triggered popular votes in Denmark and Sweden (and may eventually do so in the United Kingdom). In 2003 alone, the enlargement from fifteen to twenty-five members compelled nine out of the ten newcomers to hold referendums, the only exception being Cyprus. Now that the Council of Ministers has agreed upon the text of a new “constitutional treaty”, it is to be expected that an unknown, but nevertheless large, number of member states will have to allow their respective citizens to vote directly upon its ratification.

But European integration by successive treaties is not the only factor promoting referendums in Europe. Liechtenstein, Italy and Ireland (as well, obviously as Switzerland) continue to hold a large number of popular consultations on non-EU-related issues. They are followed closely by such newcomers as Slovenia, Azerbaijan and Ukraine. The reason for this is more generic. When choosing their institutions after the regime changes in 1989-90, a substantial proportion of the new democracies inserted provisions for direct popular consultations into their constitutions. For example, Hungary, Latvia, Lithuania, Poland, Romania, Slovenia, Slovak Republic and Ukraine have all gone beyond the convocation of referendums by governments and introduced the possibility of consultations provoked by citizens, that is, initiatives. In Western Europe, this form exists only in Switzerland, Liechtenstein and San Marino. Since 1990, popular initiatives – excluding the Swiss case – have accounted for only 13% of all popular consultations held in the “old” democracies; whereas, in the “new” ones of Eastern Europe and the former Soviet Union, the proportion rises significantly to 24%. This reveals that, not only do these countries have the necessary formal provisions for holding initiatives, but their citizens have quickly discovered how to use them. Such bottom-up efforts are almost twice as frequent there as in the old democracies. This clearly marks a difference with Western Europe where most of the experimentation with citizen-initiated votes has come almost exclusively from the local or regional level – if at all.

Finally, in order to measure the impact of direct democracy, we should try to assess how successful popular consultations have been. “Success” can take various forms. The first question to arise concerns who is successful. By definition, the authors of referendum texts are the political authorities (whether parliament and/or government). If voters approve the referendum, the assumption is that these authorities have been successful and that their policy proposal has been democratically legitimised. The opposite is the case for popular initiatives since rulers are usually opposed to them and recommend their rejection. The second question focuses on the aftermath of a popular consultation. Even referendums accepted by a majority of voters may not necessarily be taken into account by the government if they are non-binding in nature (as they usually are). Favourable referendums can also be voided when they fail to reach a sufficient quota of participation by eligible citizens. In both of these cases, the effect will simply be a prolongation of the status quo rather than a change in policy.

From a dataset on direct democratic experiences assembled at the University of Geneva, it is possible to measure the “net effect” of referendums and initiatives, that is, whether they have led to relevant policy changes. With regard to Western Europe, there has been a trend towards higher rates of acceptance. For example, since 1990, roughly three out of four government initiated referendums have passed successfully. One can also observe a relatively consistent (if slightly lower) rate of acceptance with regard to citizen-promoted initiatives. Roughly one out of two of them was accepted by the voters. Most of the popular consultations – referendums and initiatives – were binding and, yet, the rate of referendum-induced policy changes diminished. More recently, only one out of three referendums led to a directly related change in policy, despite an acceptance rate of 74%. This means that only every second proposal accepted by a majority of voters provokes a policy change. This rather surprising finding can be attributed to two factors: first, the rate of non-binding referendums has slightly increased and, second, quorum rules were less and less frequently met, meaning that citizen turnout was insufficient to make the result legally valid. This is a worrisome trend that could lead to a vicious political circle. If popular consultations are held and accepted, but subsequently ignored either because they were non-binding or because they failed to reach a quorum, voter apathy will likely increase when such opportunities arise in the future. Voting in elections is already an irrational act for the individual citizen in so far as the probability of his or her vote changing the outcome is minimal. Despite this, large numbers of people do go out and cast their ballot. But if we add to this the probability that the results will be ignored or voided by the rulers, then the conclusion is virtually inescapable that citizens will become increasingly apathetic and not bother to vote – first, in popular consultations and, perhaps, later in regular elections.

The non-binding character of certain referendums does not seem to be the most problematic issue. For example, when in 1994 a majority of Norwegian voters refused EU membership in a consultative (non-binding) referendum, it was politically inconceivable that the Norwegian government would simply ignore this outcome and proceed with joining the EU. The same political logic applied to the 2003 consultative referendum in Sweden on accepting the euro. From this perspective, there would seem to be little to gain by changing from non-binding to binding consultations – except for their potential impact upon turnout. Presumably, more citizens will vote if they are assured that the collective decision, whatever it is, will be implemented. Which brings us to the real issue: which is that of imposing minimal quorums for referendums and initiatives. Sometimes, the outcome can be determined by a very close margin. For example, in a referendum on 18 April 1999, 91.52% of Italian voters accepted a proposal changing the mode of calculation for attributing parliamentary seats to make them better respect the principle of proportional representation. However, turnout was only 49.58%, missing the 50% quorum by only 0.42% of the eligible voters. This rejected an outcome that was massively approved by Italian citizens on the basis of a more or less arbitrary threshold.

The situation in Eastern Europe since 1990 has been quite different. Here, a very large majority of referendums and popular initiatives have been accepted. Figures prior to the collapse of the Berlin Wall showed an acceptance rate of 100%. However, under communist regimes, referendums were hardly held in a democratic manner and the apparent congruence between rulers and citizens was illusory – as the world discovered after 1989-90. Popular initiatives, of course, did not exist. Since democratisation, not only has the acceptance rate been higher than in the “old” democracies in Western Europe, but they have also resulted in much more reliable and immediate changes in public policy. Over the past few years in Eastern Europe and the former Soviet Union, the ratio between ruler initiated referendums and citizen approval of them has fast been approaching that sustained under communism, that is 100%. One could ask whether, if governments always win, could it be because the process of direct democracy itself is biased in favour of incumbents? Evidence on the fate of citizens’ initiatives suggests that this is not the case in these new democracies. Here, when faced with choices that typically oppose existing policies and when urged by their rulers to vote “No”, Eastern and Central Europe’s newly enfranchised citizens have voted “Yes” – about twice as frequently as in Western Europe.