Direct Democracy

Abstract

Chapter 4 discusses the emergence, types and functioning of Swiss direct democracy and assesses its effects on power-sharing and the political system more broadly. Contrary to what many theorists have claimed, direct democracy is not too demanding for people in a highly developed industrial society. The fact that citizens have a direct say up to four times a year at federal level, in addition to frequent referendums at cantonal and local level, makes voting almost a routine exercise. It also allows today’s losers to become tomorrow’s winners—and vice versa, thus fostering mutual respect and cooperation. Yet contrary also to what some revolutionaries had hoped for, it has not led to radical change either. Instead, direct democracy has above all led to political conservatism, be that regarding welfare policies, defence or European integration.

4.1 Introduction: The Vote to Abolish the Swiss Army

On 27 November 1989, the New York Timesreported the following news from Switzerland:

Swiss Reject Plan to Scrap Army

Geneva. Switzerland today voted to keep its army as the best way of maintaining its neutrality. An initiative to abolish the army was turned down by a margin of almost two to one. ‘A majority of the states rejected it’, a Government spokesman said. Only in Geneva and Jura did the majority vote in favour of the proposal. The initiative, forced by a petition signed by 111,300 citizens, set off a fierce national debate on the usefulness of an army in a small neutral country.

Readers of the New York Times may well have been stunned and wondering:

• How is it that a handful of citizens can challenge the federal government to an extent of proposing such a revolutionary idea as the abolition of the entire army? And if the Swiss people can revolutionise their country at the ballot box, why is Switzerland’s government a symbol of stability and its policies so conservative?

• Does direct democracy really have an impact on policy, as this vote on the army implies, or is it just a kind of theatre with the political elite holding real power backstage?

• If—as described in Chap. 2—direct democracy is part of an old cultural tradition, has it now become obsolete? Can democracy in a modern society keep up with growing complexities if the most important decisions are taken by ordinary people?

• And if so, how does direct democracy actually work? Who participates, who does not and how do voters react when confronted with difficult questions? What are the wider effects of direct democracy on the politics of government and parliament as well as political parties?

In this chapter, we shall try to answer some of these questions.

4.2 Institutions, Historical Development and Meanings of Direct Democracy

4.2.1 Obligatory and Optional Referenda

A ‘referendum’ in this context means a popular vote on a specific parliamentary decision, with the citizens having the last word: they decide whether the proposal becomes law or is rejected. In Switzerland, there are two types of referendum. First, all proposals for constitutional amendments and important international treaties are subject to an obligatory referendum. This requires a double majority of the Swiss people and the cantons, thus offering a kind of federal participation (see Chap. 3). The obligatory referendum is relatively frequent. Since Article 3 of the Constitution leaves all powers to the cantons unless specifically delegated to the federation, the authorities have to propose an amendment for every major new responsibility undertaken at national level.

Second, most parliamentary acts and regulations are subject to an optional (or facultative) referendum. In these cases, a parliamentary decision becomes law unless 50,000 citizens or eight cantons, within 100 days, demand the holding of a popular vote. If a popular vote is held, a simple majority of the voting people decides whether the bill is approved or rejected, the wishes of the cantons being irrelevant. Since the obligatory referendum refers to constitutional amendments and the optional referendum to ordinary legislation, the two instruments are often distinguished as the ‘constitutional’ and the ‘legislative’referendum (Aubert and Mahon 2003, 1061–116) (Table 4.1).

Table 4.1 Types of referendum and popular initiative (federal level)

At cantonal and local levels, referenda occasionally go even further. Some cantons hold an obligatory referendum for most laws and important acts, and referenda may also be held on specific financial decisions (Lutz and Strohmann 1998). Direct-democratic instruments are, on the whole, more widely used in the Swiss-German cantons and communes than in the French- and Italian-speaking parts of Switzerland (Vatter 2002, 219ff.; Trechsel and Serdült 1999; Lafitte 1987; Karr 2003). In the latter, the ‘liberal’, representative idea of democracy dominates the ‘radical’, participatory one (Bühlmann et al. 2014, 404–6).

4.2.2 The Popular Initiative

One hundred thousand citizens can, by signing up to a formal proposition, demand a constitutional amendment and/or propose the revision or removal of an existing provision. The proposition can be expressed as a fully formulated text or in general terms upon which the Federal Assembly can then make a formal proposition. After signatures have been collected successfully, the initiative is discussed by the Federal Council and parliament, which then adopt formal positions on the proposed changes. This can involve drawing up an alternative proposition or, if the popular initiative is couched in general terms, formulating precise propositions. Initiatives and eventual counterproposals are presented simultaneously to the people. As with all constitutional changes, acceptance requires majorities of both individual voters and cantons.

The cantons dispose of additional instruments of direct democracy. Whereas at federal level the popular initiative is restricted to constitutional matters, it can be used to propose ordinary laws and acts at the cantonal and local level. The process leading to popular votes, notably the number of signatures required and the time allowed for their collection, varies markedly from canton to canton. One would imagine that the height of this hurdle would influence the use of the referendum and the popular initiative. However, this is not the case. There is no statistical evidence to suggest that in cantons with high hurdles, referenda or initiatives are used less often than in cantons with low hurdles (Vatter 2018, 372; Moser 1987).

4.2.3 Direct and Semi-direct Democracy: Historical Origins and Development

Swiss political rights have resulted from cultural patterns and history, political struggles and coincidences. There are some myths about direct democracy. Its protagonists in the nineteenth century claimed that it was a revival of old democratic freedoms. In reality, the Swiss confederation in medieval times had its landlords and familial oligarchies just as their neighbours had their nobility. The French Revolution ended the Ancien Régime and the privileges of old cantons over their subject regions in Switzerland. Democracy was imposed by Napoleon, not invented in old Switzerland.

Alfred Kölz (1992, 615–20), in his book on the history of the Swiss Constitution, shows that democratic institutions were directly influenced by theorists of the French Revolution, but official history in the nineteenth century declared them to be of Swiss origin. When in 1831 the progressive cantons began to establish democracy, it was under the slogan ‘sovereignty of the people’, and the constitutional framework provided for the division of power and the free election of representatives. But the representative system reminded the protagonists of democratisation too much of the old regime and its power elites. Thus, democratic forces called for full democracy, that is, law-making by the people and self-government. Whereas the holding of referenda would give the people control over parliament by ensuring they would have the last word on all important decisions, the initiative would bring citizens’ own ideas to bear on law-making. The democratic forces demanding these rights were successful. The referendum and the initiative were introduced first in the cantons, and later in the federation, whose original 1848 Constitution more resembled a parliamentary democracy.

When the legislative referendum and the initiative for partial revisions of the Constitution were introduced at federal level in 1874 and 1891, respectively, there was a second motive behind the calls for direct democracy: to prevent political and economic power being concentrated in the same few hands. As Karl Bürkli, a fervent democrat and trade union leader, wrote in 1869:

Our law-makers, elected by the people, are incapable of making good laws for the working class, even if they make excellent laws for the bourgeois class. Why? Because the representative bodies, in their majority, consist of capitalists and their servants who are hostile to social progress. Just as slave-holders are incapable of making laws in the interests of slaves, capitalist-representatives are incapable of making laws in the interest of workers. Representative democracy is not a form of government able to improve the living conditions of the working class and to resolve social problems. (Own translation)

But unlike Karl Marx, who 20 years previously had called for a revolutionary class struggle against the ‘bourgeois’ state, Bürkli put all his hopes in direct democracy as law-making by the people. If direct democracy is realised, he wrote, ‘the people will find the right way to social freedom, because they feel themselves its daily sorrows and the need for change’.

From the very beginning, this expansion of the people’s rights not only to elect its authorities but also to vote on certain issues led to another understanding of democracy. The model of pure representative democracy promotes the idea of an elected government and parliament who decide for the people. They are entitled to do so because they represent the people or its majority. Representative democracy requires trust in the parliamentary elite, and trust that the will of parliament is consistent with the preferences of the majority of citizens. Bürkli was not the only political leader distrusting the political elites. In the cantons, many bourgeois politicians, too, were unsatisfied with the politics of their government and parliament. Distrust in government for the people led to the different idea of government through the people, that is, ‘self-rule’ in the name of the ‘sovereignty of the people’.

The Swiss were aware that government through the people was not possible for every decision. However, they wanted citizens to participate in the most important ones. Democrats demanded that the people should not be excluded from participation in the most important decisions, and that there should be agreement between the authorities and the electorate on all important issues. This constitutional system involves three types of procedures (see also Table 4.2):

• The most important questions are constitutional. Here, the people—and the cantons—always participate through obligatory referenda (voting on all amendments proposed by parliament) or popular initiatives (proposing constitutional amendments).

• Questions of secondary importance concern ordinary laws and regulations, decided by parliament. Here, citizens can intervene if they so wish: the optional referendum permits them to challenge parliamentary decisions. However, at the federal level it is not possible to propose your own law by means of the initiative.

• Questions of less importance are settled through simple regulations or government ordinances. They are left to the government, sometimes to parliament.

Table 4.2 Constitutional selection of direct-democracy issues at federal level

This constitutional order fulfils four functions:

• 1. Selectivity: The above given constitutional order provides a selection system. Not all decisions are open to the people, but the people always have the last word on the most important issues of constitutional policies, and they have an option to control the legislation on important issues.

• 2. Securing the highestlegitimacyfor the most important and controversial political decisions: In the ideology of the ‘sovereignty of the people’, the people’s own decisions are seen as the ‘purest’ form of democracy. ‘Authentic’ decisions by the people enjoy the highest legitimacy because they constitute ‘self-rule’. That is also the reason why the Federal Supreme Court cannot invalidate Federal Laws: the people have either approved them in a referendum or decided not to contest it, which amounts to the same (qui tacet consentit).

• 3. Keeping parliament involved: Only a small part of all laws enacted by parliament are actually challenged through a referendum, and government ordinances are excluded from direct participation. Hence, most decisions in Swiss politics are taken by the parliament and the executive, just as in representative systems. That is why the Swiss system is best referred to as a ‘semi-direct democracy’,¹ which means that decision-making contains elements of both representative and direct democracy. The constitutional order tells us how this shall be done, indicating who has the last word on a political decision.

• 4. Noplebiscites: In countries which have ‘plebiscites’, it is the parliament or the president who call a referendum. France’s General de Gaulle, for instance, endorsed his project to give independence to Algeria by a plebiscite in order to have more political support and legitimacy for this historical decision. Swiss direct democracy is fundamentally different. It is not the privilege of politicians to decide if a referendum is held, but the Constitution states that all constitutional amendments have to be voted upon, and that every law must be open to an optional referendum. Sometimes constitutional lawyers have different opinions on whether an issue must be regulated by a constitutional amendment or an ordinary law, but this discretionary power is marginal. Thus, the constitutional order provides an effective guarantee of the people’s right to direct participation.

Over the last hundred years, much of the great enthusiasm for direct democracy has disappeared. Many of the hopes put on the effects of ‘people’s law-making’, as expressed by Bürkli in 1869, have been dashed by the experiences of semi-direct democracy. The political left had to learn that the people did not want revolutions. But the same people also rejected many projects of the bourgeois majority. Direct democracy has not replaced, but rather complemented parliamentary politics: both the referendum and the initiative have become the most powerful instruments of the opposition and allow for protest against the political elite.² This partly explains why the political rights have become so popular: in surveys, they regularly show up as the most precious elements of Swiss democracy and identity—even for those who belong to the losers in many votations.

4.3 A Closer Look at the Referendum and the Initiative

4.3.1 The Issues

We remember the call of democratic forces for the ‘sovereignty of the people’ when fighting for participation rights. They believed that no decision of great importance should be excluded from the direct influence of the people. This historical expectation was probably too optimistic. But when looking at the list of federal votations held in the last three years alone (Table 4.3), we can see that there is practically no kind of issue that was not subject to either a referendum or an initiative.

Table 4.3 Federal votations held 2017–2019

4.3.2 Direct Democracy’s Role in Political Agenda-Setting

Table 4.3 shows a wide variety of issues that have been put to popular vote, ranging from more bicycle lanes to less public TV and radio. We could certainly ask whether these two issues should be removed from a future list of votations—the first because it is of too little importance to merit a popular vote, the second because it is of too great an importance. Yet this would not be in line with Swiss thinking. With the optional referendum and the popular initiative, it is left to the people, political parties and other organisations to decide what they consider to be a case worth voting upon. Politicians may complain about the overloading of direct democracy with minor issues, but they would not overtly deny the right of any group to place a ‘bothersome’ problem on the agenda if it successfully attracts the required number of signatures.

As regards popular initiatives, there is first a formal control by the Federal Chancellery and then by the Federal Council of whether or not the proposal is compatible with constitutional law and certain principles of international law. On the abolition of the army, for example, some officials claimed that the proposal was unconstitutional because it would destroy the fundamental task of the Swiss federation to defend its independence and neutrality in times of war. The Federal Council, however, did not find it politically wise to follow this advice and preferred a democratic vote to be held on the issue—they were convinced that the people’s common sense would lead them to vote the ‘right’ way.

The role of the Federal Assembly is to ensure that popular initiatives are in accordance with the principle of ‘consistency of subject matter’. This means that a popular initiative cannot combine different issues; citizens must be able to express their preference on a single question at a time. If an initiative contains more than one issue, it has to be split up into separate initiatives that are voted upon individually. The Federal Assembly is reluctant to invalidate an initiative on the ground that it concerns questions that do not belong to the constitutional domain.³ The fact that practically any issue can become the subject of a popular initiative has two effects:

First, the Swiss Constitution is much less a historical document to preserve the spirit of the founding generation than an open book which every generation of people and parliament is authorised to change. The Swiss Constitution, therefore, has become a rather unsystematic charter, a collection of important fundamental principles as well as of rather unimportant and detailed regulations. This was the reason why, in 1999, the Constitution was totally revised, bringing more coherence to the constitutional text. Even so, with 32 new amendments accepted since then (Swissvotes 2019), the Constitution remains a living document. It is the written evidence on the development of Swiss politics and policies—initiated mostly by the parliament but controlled by the people and the cantons.

Second, the people have a considerable influence on the political agenda. The control of the political agenda is an unresolved problem in theories of representative democracy because, by tacit arrangement, ruling political elites can agree to circumvent questions that would impair their re-election. Some scholars go as far as to say that the ‘politics of non-issues’—that is, withholding ‘bothersome’ questions from the agenda—represents the core of a hidden power game. It limits democratic discussion to questions of conformity and suppresses issues disliked by the political elites (Bachrach and Baratz 1963). Direct democracy corrects some of these imperfections. In fact, many issues—abolition of the army, immigration policy, restrictions on genetic engineering, protection of the environment and so on—were brought forward by means of popular initiatives and sometimes against the firm convictions of almost the entire political elite. Though their direct success is limited, popular initiatives widen the horizon of what is politically conceivable. Government and parliament do not have complete control of political agenda-setting, and direct democracy enables decisions to be taken on questions which the political elite would prefer to remain ‘non-issues’.

4.3.3 The Use of Referenda and Initiatives

In the first decades of the Swiss federation, popular votes were rare. After World War II, the constant expansion of the responsibilities and expenditures of the federal state made votations much more frequent. Today, on four Sundays per year, the Federal Council organises a ballot, and the people vote on up to about 12 issues. Table 4.4 shows the number of votes held between 1848 and 2019—644 in total.

Table 4.4 National referenda and popular initiatives, 1848–2019^a

The first section refers to constitutional amendments proposed by the Federal Assembly, subject of the obligatory referendum. About one quarter of all proposals were rejected by the people and/or cantons, which reflects the rather sceptical attitude of the Swiss people towards giving the federal government new responsibilities.

The popular initiative is also widely used, but it does not always lead to a votation. In a few cases, the proposal is invalid for practical or legal reasons. More than one fourth of all popular initiatives are withdrawn, sometimes after successful negotiations with the authorities for a counterproposal. At 10%, the success rate of popular initiatives is rather low. Counterproposals by the Federal Assembly, mostly voted upon in direct confrontation with the initiative, have a considerably higher success rate.

The optional referendum is the instrument challenging the ‘ordinary’ legislative activity of the Federal Assembly. Groups contesting a bill may fail to collect the required number of signatures within the 100-day limit. From Table 4.4, we see that less than 7% of all bills passed by parliament are actually so challenged. If, however, the referendum challenge materialises into a popular vote, opponents of the bill have a 43% chance of success. This means that less than 3% of all parliamentary bills are actually rejected at the polls.

From these statistics, we can draw some preliminary conclusions. Constitutional policies of the Federal Assembly, which is mainly concerned with providing legal bases for new federal responsibilities, suffer frequent defeats, being rejected once in every four votations. Nevertheless, groups of citizens who wish to promote new federal activities by means of the popular initiative are even less successful. Their success rate is a mere 10%. From a first glance at statistics on ordinary legislation we might think that the optional referendum is of comparatively low effect. In fact, the reverse is true for two reasons. First, the 6% referenda cases typically represent important bills of a controversial nature and, if there is a vote, the chances of the opponents of the bill are rather high. Therefore, the risk of an optional referendum defeat is taken seriously by federal authorities. Second, and as we shall discuss in the next chapter, the perceived omnipresent risk of a referendum being organised leads the federal authorities to avoid the referendum trap by two means: first, an intensive pre-parliamentary consultation phase allows ascertaining the degree of disapproval by different actors. Second, in taking into account opposing views that are dangerous enough to bring everything down, the government then presents a legislative bill to parliament that is already a compromise backed by a large coalition of interest groups and political parties.

4.3.4 ‘Braking’ Referenda and ‘Innovating’ Initiatives: Two Different Devices of Direct Democracy

Our previous discussion has shown that both initiatives and referenda in some ways ‘correct’ the policies of government and parliament. Yet apart from this, the two instruments of direct democracy fulfil fundamentally different functions. The referendum, particularly in its optional form, allows people to object to proposals by the authorities. The popular initiative, however, is conceived as an active way of shaping constitutional rules—in most cases against the will of government and parliament. From a citizen’s point of view, we could argue that the referendum has a ‘braking’ effect and the initiative an innovative one. Let us take a closer look at this distinction.

4.3.4.1 The Braking Effect of the Referendum

Democrats—the faction of the Radicals that fought for the introduction of the referendum in the nineteenth century—considered themselves ‘progressive’ and saw in the referendum a tool to promote reformist policies with the help of the people. Yet things turned out differently. From the very start, when the referendum was introduced in 1874, it was used by the Catholic-Conservative opposition to their own advantage, and projects of the radicalliberal majority were shot down as if with a machine gun (Steiner 1991, 139).

The democratic forces had to learn a lesson important to many institution-builders: the consequences of institutional designs are very hard to foresee. Decision-making devices are tools which can be used both by other actors and for purposes other than those envisaged by their protagonists. In many cases, only history can reveal the effects of institutional mechanisms.

If the referendum is used as a plebiscite to give the authorities’ policy greater backing and legitimacy, the government should have the sole power to define under what conditions and on which occasions it is held. This is the case with plebiscites in France. The president of the French Republic organises a plebiscite when, confident that he has the backing of the popular majority on an important issue, the popular vote would help him to continue the general policy of his presidential mandate. Even then this procedure is not without risks, as De Gaulle learned in 1969, when the defeat of his proposal for regionalisation and senate reform forced him to resign (Aubert 1974, 43–4). Yet if plebiscites are unsuccessful, politicians find ways to do without them. When in 2005 the French and the Dutch people said No to the European Constitution, the EU authorities proposed similar steps to integration by way of the Lisbon Treaty, on which no plebiscites in these two countries were held.⁴

Swiss politicians have the discretionary power neither to make an issue the subject of a referendum, nor to prevent a votation. As explained earlier, the Constitution says which type of parliamentary decision is linked with which type of referendum. Parliament cannot circumvent referenda, even though for some decisions it may be particularly difficult to obtain a majority. New taxes, for example, are not very popular in any state. In pure representative democracies, political leaders impose them after elections in the hope that the people will have forgotten about them by the next election. This is not possible in Swiss direct democracy, where the political authorities have to convince the people that higher taxes are necessary. Therefore, the obstacles for success are high, not only for amendments to the Constitution where the double majority of the people and the cantons are required. In ordinary law-making, parliament can never rule out the possibility that its decision will eventually be challenged by a referendum launched by a political party, an interest group, or by a spontaneous social movement able to collect 50,000 signatures. Parliament therefore is bound to be cautious in law-making. Finally, it seems that negative majorities are easier to build than coalitions in support of change (Box 4.1).

Box 4.1 Difficulties of Building Up ‘Constructive Majorities’: The Example of Taxation in Theory and Practice

If the government needs more revenue, it must theoretically encounter tax resistance from all citizens. However, it may propose a solution that obtains a majority of rational voters, for instance reducing the tax burden for a majority of modest-income households by a small amount and raising taxes for a much smaller group with higher income. By doing so, the government may expect a political majority for its project of a net fiscal gain. However, its hopes can be dashed for two main reasons. Firstly, the proposed solution may have an impact on participation behaviour: higher-income classes, highly affected by and opposed to the bill, may mobilise and participate more in the vote than people with modest salaries. Moreover, the government cannot even be sure that people with a modest income will vote for the bill by offering them only a small gain: voters of lower social or economic status sometimes do not vote according to their actual status, but according to the status to which they aspire. So they may vote no, as if they belonged to a higher-income class. Secondly, opponents may bring up the argument of ‘federalism’, stating that new taxes are much more important on the cantonal than on the federal level. The ‘federalist’ argument is strong because in many issues it is able to divide the entire electorate.

Thus, the government will end up facing a coalition of three groups opposed to the bill, albeit for different reasons:

• higher-income groups affected by higher taxes;

• lower-income groups voting as if they had higher-income status;

• all income groups preferring cantonal taxes.

On the supporting side, the government may expect one group only: lower-income groups voting in line with their current status and preferring federal taxes. The actual supporting side may thus be smaller than expected and lose against a heterogeneous opposition.

In practice, the federal authorities were able to successfully raise revenue in the past but seemed well aware of the theoretical difficulties just mentioned. Value added tax (VAT), disliked by most households, is lower than in other countries, and federal revenue relies much more on income than consumer taxes. Progression of income tax is also high—a minority of people with high income contribute more to federal revenue than all other households. The regime on federal income and consumer taxes is also limited in time—after a decade or so, the people vote again on it. Finally, a good part of federal revenue is paid back to the cantons in the form of transfers. All these factors may have helped to build ‘constructive majorities’ on a difficult issue.

For these reasons, the referendum is an instrument of the opposition and favours the status quo. Over decades, the referendum was the favourite instrument of conservative right-wing forces fighting against new competencies of the federation and the development of the welfare state. In the last 30 years, however, it has been the political left which successfully used the referendum against conservative propositions to cut social security programmes, privatise and liberalise working regulations. This illustrates that the status quo bias of the referendum is of a systemic nature and can be used against innovations from any side. The referendum appears to be a versatile vessel, comparable to a sailing ship propelled by the wind of popularity—no matter the direction from which that wind blows. But there is no doubt about the shores the crews on referenda ships are heading for: the defeat of a bill. And this means the maintenance of the status quo.

We now see the direct effect of the referendum on the political process. Its status quo bias renders ‘big innovations’ unlikely. Political elites must anticipate the risk of defeat in a future referendum and are therefore bound to incremental progress. For every political project they have to look for an oversized coalition able to defeat the veto power of possible opposition forces in a popular vote.

This leads to a second, indirect effect. The referendum has profoundly changed the Swiss way of political decision-making. When, at the end of the nineteenth century, the Radical majority realised it could be beaten time and again by a ‘destructive’ conservative minority via the referendum, it changed its policy of pure majority rule in government, allocating one of the seven seats of the Federal Council to a Catholic-Conservative (Neidhart 1970; Bolliger and Zürcher 2004). The majority thus began to strike political compromises with the minority, finding solutions that did not threaten the status quo of groups capable of challenging the bill. This integrative pressure of the referendum transformed majoritarian politics into power-sharing—an institutional effect of direct democracy discussed further in Chap. 5.

Finally, the referendum and its status quo bias had important long-term effects on the development of the federal state. Compared to other industrialised Western democracies, we note:

• The historically late development of certain activities of the Swiss central government, especially in social policy. As every new federal responsibility must obtain the double majority of people and cantons, obstacles are high. In fact, many proposals for amendments to the Constitution were rejected at first and accepted only in a second attempt.

• The low (33) percentage of total public expenditure accounted for by the central government and the modest public sector nationwide, which accounts for less than 35% of Switzerland’s GDP (EFV 2019, 5 & 97).

• The unique fact that among industrialised democracies, Switzerland’s central government is the only one which can rely on income and value added taxes only on a provisional legal basis.

• A small bureaucracy: only 11% (38,000 out of 325,000) of all Swiss administration employees work for the federation. The public sector in Switzerland amounts to only 14% of total employment (BFS 2019).

• Compared with other neutral states, a rather discreet position in international affairs. In 1992, the Swiss people and cantons refused to join the European Economic Area, and Switzerland is unlikely to become a member of the EU in the near future. Switzerland was one of the last countries to join the UN, in 2002—the first to do so via a popular vote, however. One important reason for the sceptical attitude of Swiss voters in foreign affairs is neutrality. Neutrality is more than an elite’s preference in foreign policy. It is part of the national identity, even myth, held by a majority of people that Switzerland should always stay neutral and not be committing too much in international politics.

4.3.4.2 The Innovating Effect of the Popular Initiative

In contrast to the referendum, the popular initiative is a promoter of political innovation. Moreover, it promotes innovation against the will of the political elite. We have already discussed one such innovating effect: the initiative widens the political agenda and gives authoritative voice to problems that might remain non-issues as far as the elites are concerned. Yet agenda-setting alone does not mean gaining majority approval for a proposal. Statistics show that hopes for political change by means of the initiative are dashed in 90% of cases that come to a vote (Table 4.4). If the people are so sceptical towards grass-roots innovation, we have to ask why so many initiatives are handed in. Practice shows that the popular initiative may serve four different objectives:

• 1. Direct success against the federalauthorities: In parliament, the permanent coalition of governmental parties may constantly ignore the claims of the opposition parties. Thus, the popular initiative can be an instrument for parliamentary minority groups. They hope that their issue will be popular enough to find a majority in the votation—even against the mainstream of the political elite. For a long time, it was primarily the Social-Democrats and trade unions who used the initiative to compensate for the lack of support for social reform in parliament. They made the experience, however, that an initiative is a good instrument for political protest but less suited to realise their claims. For initiatives seeking direct success, the degree of innovation must be modest. Typical examples are the introduction of a national holiday on 1st of August in 1993 or the protection of moor landscapes in 1987. In recent years, however, some quite radical conservative initiatives (ban on minarets, ‘against mass immigration’ and life-long detention for sex-offenders) were successful.

• 2. Indirect success through negotiation with the authorities: As just mentioned, direct success in a popular vote is rare. But defeat does not always leave proponents with nothing. Sometimes the federal authorities pick up ideas from an initiative by drafting a counterproposal or fitting them into ongoing legislative projects. This way the long shots of popular initiatives are transformed into proposals that are more in line with conventional wisdom and therefore stand a better chance of being accepted. At the root of many important federal policies—from social security through the environment to equal rights—we can find a popular initiative. In this way, ideas too innovative and radical at first can later be transformed into proposals acceptable to a majority. In the long run, these indirect effects of the initiative may be even more important than rare direct success (Delley 1978; Sigg 1978; Werder 1978; Papadopoulos 1994).

• 3. Mobilisation of new issues and political tendencies: The objectives of radical groups are different. They may prefer agenda-setting and discussion of political taboos and non-issues, which is provided by the arena of a popular vote. They refuse to pay the price of negotiation and compromise. Therefore, these groups draft ‘long-shot’ propositions, even if their chances of success are minimal or even zero. The initiative on the abolition of the Swiss army is a good example. From the very beginning, its proponents were aware that they would not win a majority of the vote. Still, they used the four years’ discussion to change political attitudes on the formerly taboo subject of Swiss military and peace politics, and with considerable success.

• 4. Self-staging and mobilisation for electoral success: Finally, political parties and social movements can use the popular initiative as a platform for electoral success. Popular initiatives not only ‘create’ issues but also help to establish new political parties and rally the followers around a common cause. This is typical of the way in which several grass-roots movements of the 1970s put environmental issues onto the national agenda and finally established a new, national Green Party. The xenophobe movement of that period also regularly launched popular initiatives asking for restrictions on immigration. This helped the small anti-immigration parties not only to keep their pet issue on the political agenda, but also to survive for quite a long time.⁵

4.4 Participation in Direct Democracy

4.4.1 The Deciding Majority, Or Who Are the People?

On the evening of a popular vote, the news readers on TV and radio often say: ‘The Sovereign of Switzerland has accepted (or rejected) the following propositions …’. The allusion to the ‘Sovereign’ (der Souverän) is an old expression for the highest democratic organ or authority, but who is that? The expression ‘direct democracy’ implies that it is the people, or at least its majority. We shall see that, in practice, it is far from this.

First, the share of people who are qualified to vote in Switzerland is only about 62% of the total population. Those under the age of 18 and foreign nationals, who make up 25% of the resident population, are not allowed to vote. Then again not all those who do qualify take part in a vote—participation over past year averaged some 45%. If voters are split roughly 50:50, the deciding majority may become rather small. Using the above figures, 100% * 0.62 * 0.45 * 0.5 equals 14% of the entire population. Figure 4.1 shows the deciding majorities in federal votations as a percentage of the total Swiss population since 1866.

Fig. 4.1

Deciding majorities as a percentage of the total Swiss population in federal votations, 1866–2019. (Source: own calculations and graph based on BFS [2019; population censuses and referendum results]. After 1977 including Swiss citizens living abroad [cf. Kuenzi 2018])

It shows that during the long decades of male-only democracy, before women’s suffrage was introduced in 1971, the actual ‘Sovereign’ could be as small as 5–15%. Since then, the deciding majority has varied between 12% and 22%. Even so, the democratic majoritynever represents the majority of the population, and the ‘will of the Sovereign’ is in reality the vote of a minority. Moreover, as we shall see, the participation of the different strata of citizens is far from being equal. One could argue, therefore, that a serious survey of 30,000 people would cost less and give more accurate information on the true preferences of the people. This argument, however, misses the point.

The goal of a votation is not the most precise reproduction of public opinion, but the participation of active citizens in a collectively binding decision. This process of direct participation gives high democratic legitimation to the decision taken, for several reasons. First, the legitimating effect lies in the fact that all citizens are offered the chance to participate, and in that those who do so put time and effort into making up their minds and casting their votes. Second, a popular vote is usually accompanied by intensive campaigns for and against the proposal, including adversarial public meetings, party recommendations, lobby slogans and extensive coverage in the media and online. This process of public deliberation and decision-making may lead to changes in public opinion and individual preferences. It is a collective learning process. Third, the collective decision is authentic: people are binding themselves with the consequences of their own decision. It is their own decision, not one imposed on them by political elites. Direct participation corresponds to the idea of ‘self-rule’. All this creates double legitimacy—for the concrete decision at stake and for the democratic institutions in general.

From a normative point of view, one could still argue about two imperfections of direct democracy. First, what about foreign nationals living in Switzerland who, despite paying taxes and otherwise contributing to public welfare, are excluded from participation? Indeed, while a large part of the Swiss citizens would not be willing to change this rule, others call it an imperfect state of democracy. This reminds us that the concept of democracy continues to change. The entitlement to vote has evolved over time: once it was restricted to adult married men with some degree of wealth and social status. This restriction was later abolished. While Switzerland may have been late in granting political rights to women, there is one canton that introduced political rights for foreigners as early as in the nineteenth century: in the communes of the canton of Neuchatel, foreigners have participated in elections and votations since 1850.⁶ The canton of Jura followed in 1979, but in many other cantons proposals to extend political rights to foreigners have failed.⁷ In turn, Swiss citizens living abroad, counting for more than half a million by the end of 2019, have been given the possibility to vote since 1977. Since 1992, they can do so without having to travel to Switzerland to exercise this right (Kuenzi 2018). This illustrates that the historical process of the ‘inclusion’ by political rights is long but certainly not at its end yet.

The second question is, does low participation not discredit direct democracy despite its procedural value? Should not a turnout of, say, 40% or 50% be required, the result being invalidated if participation falls below that level—an idea that is applied in Italy, for example? In order to answer this question, let us have a closer look at individual participation.

4.4.2 Regular Voters, Occasional Participants and Abstentionists

As shown in Fig. 4.2, the level of participation in federal votations varies above and below an average of 45%, depending on the attractiveness of the issue voted. Controversial subjects of great importance to everybody attract the most voters, such as those to abolish the army, limit immigration, or joining the UN.

Fig. 4.2

Voter turnout in Swiss ballots, 1980–2019. (Source: own graph based on BFS [2019])

From post-vote surveys, Sciarini et al. (2015, table 1) and Serdült (2013, 48), we learn that Swiss voters fall into three groups which differ in behaviour and general attitude towards voting:

• The first group, comprising some 25–30% of citizens, always votes, that is, at least nine out of ten times. These ‘duty-conscious citizens’ interpret their political right to vote as being a citizen’s duty as well.

• The second group of about 20–25%, the ‘abstainers’, never participates at all. Different reasons can be found for such behaviour. Some are disillusioned; some feel incapable of dealing with the issues involved; others are simply not interested in politics.

• The third and largest group, some 50–55% of the electorate, consists of occasional voters, participating à la carte. They participate in between one and eight votes out of ten. Their selective interest in politics according to the issue at stake makes the participation rate fluctuate between 30% and 70%. Occasional voters participate when they feel they are personally involved or when concrete advantages or disadvantages for them are in play.

The criticism is often made that the low level of participation, which occasionally can go down to 25%, discredits direct democracy. However, proposals to introduce a minimum participation rate for the vote to be valid have been widely rejected by politicians and the public. Indeed, such a measure would probably be more likely to punish and disappoint active voters than incite the inactive majority to take part. The strongest argument against a participation quorum is that it destroys the deliberative nature of a votation: while proponents have to argue with substantive arguments for the project, opponents can renounce on any argument by a simple call to boycott the vote. The opposition not only has better cards in the game but destroys it: if non-participation pays better than participation, the legitimation of direct democracy itself will suffer.

We also have to recognise that participation in direct democracy is very demanding. Voters in Swiss democracy are supposed to vote on issues that are sometimes very complicated. To read the official documentation on four or five proposals can take several hours. Together with votations on cantonal and local affairs, a voter is supposed to give his or her preference on up to 20 or 30 issues a year. In none of these can he or she expect to have more than an infinitesimal chance of being the one who makes the outcome decisive. The cost of participation therefore outweighs the expected individual benefit, and following the logic of rational choice, the voter would stay at home. Indeed, if many deplore the low participation rates, it could be argued that a turnout of 45% is surprisingly high, given the time and effort required. Thus, there is no satisfactory criteria to judge whether 45% of participation is low or high. Nor can we say whether this participation rate is a bad or good sign for a sound democracy and a mature civic culture.

Nevertheless, there are other reasons for worrying about low turnout. As mentioned above, many abstainers are disappointed or feel unable to participate. Direct democracy does not provide guarantees against political frustration or alienation. More importantly, international research (Kern and Hooghe 2018; Vatter et al. 2019, 173f.) into participation reveals two main findings:

• The lower overall participation, the greater the difference in voter turnout between higher and lower socio-economic groups; and

• The more demanding the form of participation, the greater the difference in voter turnout between higher and lower socio-economic groups.

In Switzerland, both factors appear together. First, especially if participation is rather low, as in the example given in Table 4.5, the choir of Swiss direct democracy sings in upper or middle-class tones.⁸ Post-vote analyses reveal that in such cases workers and lower-level employees participate less than high-level employees and independent professionals by a factor of up to three. As in other countries, well-educated people with higher income are more likely to vote than their less educated or working-class counterparts. Second, direct democracy is demanding. One should not underestimate the general capacity of ordinary citizens to understand the questions they are voting upon. But besides their personal motivation and political interest, also their capacity to understand the issue at stake varies. If the issues of a vote are complex, some citizens feel unable to cope with it. In a sort of self-censure, they refrain from voting. The second issue in Table 4.5 for instance, the referendum against a new law on enterprise taxes, was difficult to understand because it involved many technicalities. And we notice that people with lower revenue and education participated considerably less, even though the other issue voted that day was considered to be easy to decide: to restrict gun ownership or not.

Table 4.5 Typical profile of a popular vote

Source: Milic et al. (2019, 45f.)

Besides education and income, there are other socio-demographic characteristics that influence political participation: younger, female, unmarried and divorced citizens participate less. Moreover, some political characteristics make a difference: people with no party affinity and with no trust in the authorities participate considerably less, and the most important single factor that determines participation is political interest (Heidelberger 2018).

From a normative perspective, however, the most important defect of direct democracy lies in the unequal participation of the social classes. Direct democracy, if its procedures and issues become too complex, turns out to be a ‘middle-class democracy’. To avoid this, direct democracy must be simple in its procedures and in the formulation of issues on the ballot.

4.5 The People Between Knowledge, Trust and Propaganda

This section addresses further key questions about direct democracy: do citizens understand the issues they vote upon? What are their motives when saying yes or no to a proposal? What is the role of the political elites and their campaigns? And is it true that, given enough money and propaganda, any votation can be won? When discussing these questions, we can draw upon a fast-growing number of scholarly studies of voting behaviour in Switzerland.⁹

4.5.1 Example of a Vote: Should There Be Tougher Restrictions on Refugees Seeking Asylum in Switzerland?

Immigration policy has been one of the most controversial issues of Swiss politics for over 50 years. Back in the 1980s, the number of refugees seeking asylum in Switzerland grew drastically from about 3000 per year (1980) to more than 37,000 (1990), and federal and cantonal resources became strained. Many refugees had to wait several years for a decision on whether they were awarded refugee status. A negative decision meant expulsion, which was considered to be inhuman by many Swiss, who even tried to hide or protect refugees facing repatriation. On the other hand, there was a growing reluctance among a part of the population to allow too many refugees to stay in Switzerland, in addition to the more than one million other foreigners. In 1985, the federal parliament revised the asylum law. It sought to process the growing number of demands for asylum more rapidly and to undertake expulsions more efficiently. The revision of the law was a compromise: right-wing and xenophobe forces were against encouraging a ‘growing mass of refugees’, who for them were mostly ‘false asylum seekers’ coming for economic reasons and not because of political persecution. They proposed severe measures to keep refugees out of the country and a simplification of the legal procedure. Refugee organisations, the Greens and the political left, on the other hand, were opposed to changes in the existing liberal law and its procedure, which offered refugees many ways (and the time) for appealing against negative decisions. Parliament finally chose a middle way, restraining the procedure for asylum, but leaving doors open to refugees according to the standards of international law and the humanitarian tradition of Switzerland. The revised law was not to the taste of Swiss refugee organisations, which, together with the Greens and parts of the left, successfully launched a referendum challenge that was voted on 5 April 1987. The challenge failed: 1,180,082 citizens voted for the revised law, 572,330 against, giving it a majority of 67.3% (see Box 4.2). It was not the first of several referenda on Swiss refugee policy, and others followed, illustrating the salience of the issue up until today.

Box 4.2 Tougher Restrictions on Refugees: Cleavages, Motives, Interests and Voting Behaviour

• A) Cleavages

According to the VOX post-vote survey (no. 32, July 1987), voting behaviour firstly mirrored the strong divide between the right and the left, each mobilised by the slogans of political parties. For instance, voters with affinity to the Swiss People’s Party (90% yes), the Radicals and Liberals (88%) and Christian-Democrats (70%) massively supported the law. On the other hand, voters with affinity to the Social-Democrats (41%), Greens (37%) and small left-wing parties (9%) were clearly opposed to asylum restrictions. The ratio between voters of the political right and left was about 2:1. Note, however, that voters with no party affinity constitute a good majority of all voters. In the vote of April 1987, they supported the project with 72%. On questions such as this one where the traditional division between right and left is decisive, the left has a chance to win only if it can sway voters with no party affinity.

Besides the right-left divide, the VOX survey also revealed a social divide. There was higher support for tougher restrictions among lower social strata. Education, particularly, had a strong effect: the higher the level of education, the more liberal the attitude towards refugees. Citizens with only basic education massively supported the law (88% yes), whereas voters with university degrees rejected it (41% yes).

• B) Motives and Interests

Between 1970 and 2019, the Swiss voted 34 times on questions to do with migration and/or asylum (Swissvotes 2019). With 25% of non-Swiss among the resident population today, questions of integration and social conflict have persisted among the most salient and controversial political issues (Fischer et al. 2002; Schneider and Holzer 2002; Linder 1991). In the many votations on migration and refugees, one can observe a constant pattern composed of three main groups, each with different motives and interests:

• Categorical opponents of (growing)immigration: protagonists of restrictions on immigration and asylum seekers share a variety of motives that range from feeling the necessity to set limits on the proportion of the foreign population, through wishing to protect traditional Swiss values, to fear of overpopulation and loss of Swiss identity. Unskilled Swiss workers feel disadvantaged by growing immigration of unqualified workforce, while taxpayers are reluctant to accept refugees who cannot be integrated into the labour market.

• Categorical defenders of liberalimmigration: protagonists of free access for asylum seekers are mainly acting according to humanitarian and egalitarian beliefs, but they may have different reasons: congruence with political ideologies of the left and the Greens, or the fact that better educated people have been less exposed to the negative effects of immigration.

• Pragmatists: whereas the attitudes of categorical opponents and defenders rarely change and lead to a stable voting behaviour, pragmatists are more flexible. More than defending social values, the voting behaviour of pragmatists depends on utilitarian considerations. In the vote on workforce immigration, pragmatists can embrace the position of Liberals because as professionals they take advantage of foreign workers or new consumers. In questions of refugee policy such as the referendum of 1987, however, they vote with the opponents of immigration because asylum seekers imply public expenditure with no immediate benefit. A pragmatic attitude is to be expected especially among occasional voters with no party affinity.

The referendum case of restrictions on asylum seekers gives us some first insights into voters’ behaviour. First, we notice that the rationality of voting has different roots: social background and corresponding experience, moral values or political beliefs can be important for some groups of voters, while others behave in more pragmatic ways. Political scientists, for a long time, have led a debate on whether political behaviour depends on individually defined self-interest or shared social values. Evidence from other studies on direct democracy (e.g. Vatter 1994; Vatter and Heidelberger 2014, Mueller et al. 2016) confirms what is illustrated in our case: both models of behaviour, self-interest and shared social values up to solidarity, do exist. Second, voters’ behaviour is influenced by the voting campaign: to a large degree, they follow the recommendations and slogans of political parties or other actors, but they may and do change their mind based on the different arguments (Colombo 2018, 799). This brings us to the next point: the campaign.

4.5.2 Shaping Opinions in a Voting Campaign: The Actors

Citizens cast their votes individually and secretly, but they make up their minds during public discussions. Votations are preceded by intense political campaigns. Different actors provide information, try to convince, praise or denounce, to mobilise and attempt to lead voters to approve or reject. Even the most complex issues must in the end result in a simple yes or no. Therefore, especially at the end of a campaign, the issue has to be treated as a simple message. Let us first consider the actors involved in a voting campaign, and then, in Sect. 4.5.3, evaluate their impact on voting behaviour.

Citizens and their predispositions: In the political asylum case, many people would have had first-hand experience with the question on the ballot. They might have had a job where their colleagues or customers were foreigners. Many may have liked foreigners and refugees because they were good customers or willing and cheap workers doing jobs the Swiss had refused. But even if people liked foreigners for these reasons, they may have said that there were already too many of them in Switzerland. They may have felt like strangers themselves because their colleagues at work all come from Portugal, Bosnia, Turkey or Germany. They may have feared that their children would learn less in school because the majority of their classmates were foreigners speaking perhaps seven different languages but only rudimentary German, French or Italian. In this case, people have firm attitudes based on first-hand experience. If a popular vote on the issue comes up, they feel able to decide the question on the basis of their own, personal experience. The voting campaign may mobilise voters and confirm their own preference for a yes or no, but it does not change their minds because they are pre-dispositioned.

Yet there are other issues more difficult to decide. Tax reforms, for instance, may be complex affairs. In some cases, even specialists are not able to predict their consequences. Voters cannot infer from their first-hand experience if the proposed reform will improve or worsen their own or the general situation. They must rely on the information and recommendations of campaign actors they trust. In this case, the campaign becomes very important because the issue is not pre-dispositioned. Good arguments, recommendations by political parties, clues and catchy propaganda slogans are able to influence voters in shaping their opinion. The campaign, in such cases, may have a decisive effect on the outcome of the vote.

The Federal Council: The executive plays an important role. It decides the date and issues of each ballot. The Federal Council provides the official information on the proposals at stake. In a booklet sent to every voter, it describes each proposition, gives an account on the arguments of parliament and repeats the official recommendation for the vote. Part of the booklet is reserved for the position of the opponents. This and the generally sober account of the issue at stake may be two reasons why voters pay much attention to the Federal Council’s booklet; it is one of the prime sources of information they consult when voting. The Federal Council also takes part in the campaign by promoting and defending the position of the parliamentary majority.

Political parties: Parties engage strongly during a campaign. Popular votations are an opportunity to highlight themselves, reflecting on concrete issue against the background of their basic ideologies and programmes, and pointing out the presumed interests of their voters and their affinity to interest groups. Thus, in their slogans and recommendations, parties often emphasise basic cleavages such as left vs. right, urban vs. rural, or ecology vs. economy on which they are permanently positioned to attract and keep their clientele. The ways political parties engage in the campaign have fundamentally changed over time. In earlier periods, local and cantonal party assemblies were at the centre of opinion-shaping and mobilisation. Today, parties mostly rely on the media and the web. Their politicians take part in public debates, organise rallies, try to have their positions published in print, use social networks, seek face-to-face communication in shopping areas, without however forgetting about some of the old instruments of political propaganda: posters and newspaper ads.

Pressure groups:Vested interests of industry, employers’ organisations and trade unions, social movements and other non-governmental organisations become active if one of their core issues is at stake. Their means of campaigning vary to a great deal. Some of them, like trade unions or social movements, primarily try to mobilise their own members through their personal networks. Others, such as business associations, also launch public propaganda campaigns, sometimes spending big money.

The media: Radio, television and print media strongly engage in the campaign. They explain and comment on the issue, provide platforms to politicians and political parties, give background information and undertake fact-checking. Not only do they investigate people’s opinions and air the views of government and its opponents, they also present their own thoughts on the issue. There is a public TV and radio service in each linguistic region, bound to observe a balance between pro- and contra-sides. In earlier times a great number of newspapers were affiliated to specific parties and therefore represented their views. These newspapers have largely disappeared. Today, the press has become as commercial as almost any other product—yet its positions are not ‘neutral’; instead they reflect the preference of editors or what is presumed to be the preferences of their readers. Since the 2000s, websites, blogs, e-mail lists and online multimedia have become new elements of voting campaigns (see also below, Sect. 4.6.3).

Producers ofpropaganda: Marketing and Public Relations (PR) agencies are not independent actors in the process but offer their service to any actor willing to pay. This may be the organisation of an entire campaign for one side, or simple voting propaganda defined as information whose only objective is to forge the majority desired by those who pay for it. By its very nature, propaganda need not tell the whole truth about an issue, and sometimes it has little to do with the issue and nothing with the truth. Political advertisement in newspapers and on posters, propaganda flyers and pamphlets are dominated by slogans, photographs, images or cartoons. Their message is aimed at mobilising good or bad feelings, emotions and cues about the controversial issue. Campaigning has become highly professional, and short-term propaganda is not its only means. Today, actors with big interest and big money sometimes hire marketing agencies to launch long-term PR campaigns. The first example dates back to the 1970s when, following a major scandal, the Social-Democrats launched a popular initiative for tougher restrictions on banks. To counter this proposition, one of the big Swiss banks began a PR campaign, regularly taking out entire pages in newspapers to describe banking activities and their importance to Switzerland’s economy. Just occasionally there was a mention of the popular initiative. By 1984, the banks had succeeded in positively changing their image. In the last months of the campaign on the initiative, the banks even deemed it unnecessary to run a propaganda campaign on their own since their earlier PR-efforts had achieved its objective. The initiative failed (73% no).

Pollsters: When the Schweizerische Volkspartei (SVP)’s popular initiative ‘against mass immigration’ was narrowly accepted in February 2014, everybody was very surprised because the country’s leading—and in fact largely monopolistic—polling firm, gfs.bern, had predicted a clear ‘no’. The same had happened five years before, with the SVP’s anti-minaret initiative (e.g. Kovic 2014). The company subsequently lost the public contract for post-vote analyses, ending the VOX-series which had begun in 1977. The new post-vote analyses are called VOTO.¹⁰ More generally, the last decade has seen a number of new kids on the polling block, notably companies using online-only, opt-in surveys or betting markets.

4.5.3 Are Voters Capable to Decide on High Policy? Theory and Swiss Experience

Democratic theory is profoundly divided on whetherordinary citizens are capable of rationally deciding political issues. On the one hand, adherents of elitist, liberal or representative models of democracy argue that the mass citizenry is not qualified to decide about high politics. Therefore, their influence should be restricted to electing those who decide for them. Sartori (1987, 120), the Italian theorist, went so far as to say that direct democracy ‘would quickly and disastrously founder on the reefs of cognitive incompetence’ (see also Budge 1996, 69). Adherents of the model of participatory or radical democracy, on the other hand, argue that direct-democratic choice is not only desirable from a normative point of view but also feasible. It is not necessary that all citizens decide all questions fully informed and on a systematic appreciation of all arguments. If capacities and motivation are lacking, they can resort to simplifying strategies. Using shortcuts and cues, they can delegate the search for information to others and accept recommendations by authorities they trust to be competent (Lupia and McCubbins 1998; Kriesi 2005, 9).

This notion of simplifying strategies is important and needs some explanation. Just as in daily life, when we are at the limits of our knowledge, we begin to rely on trust. To drive a car safely we need some instructions on how to handle it, but we need not know how the engine works. Nobody knows precisely how all the complex components of a nuclear power plant operate. It is designed, built and run by specialists who each trust in the professional knowledge of others. In politics, we can make similar observations. MPs specialise in some preferred policy areas, and an expert on social policy, for example, may rely on the advice of colleagues when it comes to fiscal policy. She then decides based on cues or heuristics.

The same mechanism, substituting trust for knowledge, works with voters. They rely on recommendations from other people who are supposed to know more about the tax or nuclear issues at stake. In fact, relying on the expertise of trusted professionals may even be more rational than trying to fully understand an issue oneself. So we should not blame voters for knowing too little about the subject of a vote; substituting trust in heuristics and cues from others for one’s own knowledge is not behaviour specific to direct democracy.

Thus, since Swiss direct democracy, after more than a century, has not ‘disastrously foundered on the reefs of cognitive incompetence’, we may reject Sartori’s proposition and concentrate on another question: to which degree do voters rely on a systematic appreciation of arguments or on cues, using simplifyingstrategies?

Kriesi’s (2005) extensive study on direct democracy, using VOX survey data from 148 votes between 1981 and 1999, provides interesting empirical evidence and insights. First of all, voters’ capacity should not be underestimated. In their majority, voters decide based on a systematic evaluation of pro and contra arguments. As expected, these are mainly the well informed, motivated and politically interested voters. Moreover, voters decide on arguments if they have strong preferences for an issue, based on personal knowledge (cf. also Colombo 2018). In contrast, heuristic voting is prevalent among voters with weak opinions, ambivalent towards or ignorant of the issue at stake. But the study also shows that differences between systematic and heuristic voting are not absolute: Systematic voting strongly relies on arguments provided by the political elites, many of which do not differ very much from heuristics and cues. Furthermore, voters seem to make intelligent use of heuristics; they do not take them mechanically but look at the context and actors who provide them. Cues as such do not work—they need to be credible. In sum, the study comes to the conclusion that voters do not exhibit the ‘rational ignorance’ advanced by elitist theory, and that heuristic voting in general does not lead to irrational choices (see also Steenbergen and Colombo 2018). These findings, though, depend on one essential other factor: the campaign and the quality of arguments offered by the political elites.

4.5.4 The Role of Political Parties and Their Campaign

Political parties play a crucial role for the outcome of the vote. In the ideal referendum case, when all of them support the project unanimously, success is practically guaranteed, and this is not surprising. The compromise proposed anticipates possible opposition and presents a Pareto-optimal solution in which nobody is losing compared to the status quo ante. Interest groups also back the proposition. Therefore, opposition in the campaign is weak and cannot convincingly propose a more attractive solution.

This ideal situation is relatively rare. More frequently, some groups feel as losers and the political elites are split: one or more of the four governmental parties defects and plays the game of an issue-specific opposition. This may happen already during parliamentary proceedings, or later by decision of the party rank-and-file, which not always back the position of their parliamentary delegation. In all these cases the risk of defeat for the government increases considerably. In earlier times the centre-right coalition, as a natural majority after all, was able to win two out of three votations against left-wing opposition (Papadopoulos 1994, 137). With the Swiss People’s Party seeking a stronger right-wing profile by way of issue-specific opposition, the centre-right coalition is often split, putting government projects at risk. If two parties leave the grand coalition, a defeat of the governmental project is highly probable.

Many votations are located somewhere between these highly predictable extremes of government success or failure. If the outcome is predicted to be tight, two factors play an important role: the composition of the party coalitions of government and opposition and their campaigning (Kriesi 2005, 82–3). This is astonishing, as campaign money comes largely from interest groups. Yet to make propaganda trustworthy, it must be embedded in the campaign strategies of parties. Moreover, the intensity of campaigns itself—and the amount of money spent for propaganda—varies a lot, depending on the closeness of the vote as expected by the elites (see also Hermann 2012, 16).

In all these cases, the outcome of a popular vote is characterised by high uncertainty. The outcome of campaigns, as tennis matches between two equally strong players, cannot be predicted from the past. Actors adapt and learn from past failure. Models of scientific research are able to analyse outcomes ex post, but they cannot predict the outcome of upcoming votations—which may even be beneficial for direct democracy. The main conclusions, however, are the following: political elites, their coalitions and campaign efforts play an important role for the outcome of a popular vote. Even so, they do not control direct democracy. The government coalition sometimes loses, and opposition success sometimes comes as a big surprise. The government and political parties have learned to live with it.

4.5.5 Can Money Buy Votes?

After a votation, the losing side often complains that the other side has won because it had more money to spend on propaganda. Indeed, it happens that the antipodes in a votation have vastly different resources at hand: the propaganda budget of one side may exceed that of its opponents by a factor of 20. The question whether money and propaganda can buy votes is therefore of practical importance. In an early study on the subject, Hertig (1983) found a strong statistical correlation between success and propaganda in all 41 federal votations between 1977 and 1981. An even stronger correlation was found in 20 cases where the propaganda effort was very lopsided; that is, when the propaganda of one side dominated the other by a ratio of at least three to one. Predominant ‘yes’-propaganda won in 12 out of 13 cases, whereas predominant ‘no’-propaganda was successful in all seven cases.

These statistical correlations, however, do not provide proof that votes can be bought. It is possible that some votations would also have been won without money being spent on propaganda, or that one-sided propaganda expenditure results from existing one-sided preferences. But the study gave rise to a public debate. How much money should be allowed to be spent by a single actor on a campaign, and to which degree is it tolerable that one side may spend a lot more than the other? Swiss law guarantees voters a constitutional right for fair conditions to express their undistorted preferences. Critics have argued that fair conditions of voting have become an illusion because of the influence of powerful private actors and unequal, largely opaque campaign budgets.¹¹ Bourgeois parties, the main beneficiaries of campaign money, have been hostile to any idea of regulating political propaganda as it exists in US states such as California and Colorado (Cronin 1989, 99–113).

The Hertig study was not the last word on the question. Further studies showed that the effect of propaganda was not the same for all issues: it was weaker on pre-dispositioned issues and when voters were confronted with ‘simple’ questions such as abortion or speed limits, which they can evaluate against the background of their own experience. Non-dispositioned issues and complex questions, however, are like empty labels on which propaganda can inscribe its clues because voters cannot decide on the basis of their personal experience (Hirter 1989; Longchamp 1991).

Kriesi (2009, 83–106; see also Chap. 5) also demonstrates that there is no simple equation between propaganda and success. As already mentioned, the amount of propaganda money spent will depend on the expectations of the outcome. If a tight outcome is expected, more money is spent, and in these cases money may indeed be the deciding factor. In other situations, propaganda is of less influence. Moreover, campaign money does not play the same role for the government and the opposition camps. In the hands of the latter, it is worth more. In the end, according to Kriesi, truth is in the middle: money buys votes neither ever nor never, but sometimes it can be decisive.

4.6 Conclusions

4.6.1 Semi-direct Democracy: An Exceptional System

The Swiss system is at odds with mainstream political thought. It contradicts theories of representative democracy that consider the people’s capacity too limited for rational direct policy choices. The Swiss case provides evidence that intensive political participation beyond the occasional election of MPs is possible and, as a complement to the parliamentary process, can play an important role. It shows that a substantial share of the population is willing to discuss and express their political preferences regarding even the most complex issues. And if there are shortcomings in the system of semi-direct democracy, Switzerland has neither suffered anarchy, as some have feared from the nineteenth century up to our days, nor has it experienced the political revolutions others had dreamed of.

Direct democracy and the complexity of modern society are not mutually exclusive. On the contrary, direct democracy is an important device for social learning processes which make people politically aware and able to deal with political complexity. The federation, the cantons and the communes fulfil their responsibilities and functions just as well as political authorities in other countries—optimists might even say they do a better job under the constant watch of their citizens.

Moreover, direct democracy has changed the entire political system. Use of the referendum was an important factor that has led to the institutional system of Konkordanz or consensus democracy (Neidhart 1970). Chapter 5 describes this historical process in which the referendum became an institutional constraint that induced cooperation among all major political parties and led to negotiated legislation and mutual adjustment among interest groups. In other words, power-sharing is an institutional arrangement to reduce the risks of defeat of government policies by referendum. These indirect effects of the referendum on the legislation process have become as important as the direct impact on specific policies.

4.6.2 Direct Democracy Between Integration and Polarisation

Does direct democracy polarise or integrate the people? There are good arguments for both views. On the one hand, direct popular choice amounts to the final word in a political conflict. For a certain time, all quarrels have ended. As a verdict, the popular vote is respected by the authorities and losers alike. The Federal Council, if defeated, would never say that the people’s decision was wrong. On the other hand, the campaign before a vote heats up conflict. The articulation of social and economic antagonisms, sometimes in polemic and populist ways, are a reliable means of mobilising the voters.

An empirical study on direct democracy between 1874 and 2006 gives evidence on whether political parties tried to mobilise or attenuate the basic cleavages in Switzerland in every one of the 537 votations, and how the cleavages were perceived by the participating citizenry (Linder et al. 2008). It reveals that in a historical perspective, two of the cleavages—dealing with religion and language—have cooled out during the twentieth century also amongst citizens. Quite astonishingly, however, we observe a rising polarisation of the citizenry along the cleavages of urban-rural and labour-capital sides in the final four decades. Is this conclusive proof that power-sharing by the political elites is in vain?

Against this inference speaks the fact that the evolution of cleavages is itself dependent on the conflict-laden modernisation of economy and society. Institutional politics can only fuel or attenuate them. In Switzerland’s semi-direct democracy, the political elites are forced to seek compromises and thus generally attenuate basic societal cleavages. Popular votations, however, are the arena of issue-specific opposition, and political parties use this arena not only for attenuation but also for fuelling basic cleavages. Thus, parliamentary and direct democracy represent two different arenas. This does not mean a clear distinction of a parliamentary theatre of integration and a direct-democratic domain of polarisation. But political parties, in regularly bringing up cleavages and using them to position themselves in their campaigns, string up the underlying conflict rather than attenuate it.

Thus, the role of direct democracy for societal conflict is ambiguous. On the one hand, we find integration. The vanishing of religious and linguistic cleavages is evident, and it corresponds with the fact that political parties are trying to bridge these divides not only ahead of popular votes but also in seeking electoral gains throughout the country. On the other hand, in Swiss society there is evidence of deepening cleavages between rural and urban areas as well as between capital and labour. On many issues, the salience of these basic societal conflicts is regularly emphasised. Over the last two decades, the Swiss People’s Party launched a series of popular initiatives on supposedly unresolved immigration problems. Two of them, the prohibition of constructing minarets (2009) and ‘against mass immigration’ (2014), even succeeded against all odds. Uneasy feelings towards Muslim practices and fear of Islamic fundamentalism as well as job market related worries, respectively, are part of the explanation. The 300,000 Muslims—some 35% of which are Swiss citizens—and foreign residents more generally, however, had reasons to feel discriminated against. As in other cases, it was also controversial whether or not the minaretinitiative violated constitutional or international law.

The risk of direct democracy is therefore twofold. One, the popular initiative can be exploited for electoral purposes, which is nothing new. But it makes a difference whether this is done by a marginal or a governmental party. In the latter case, it is detrimental for the functioning of the governmental coalition. Two, initiatives can be discriminating against minorities, especially if they become part of a permanent electoral campaign and cannot themselves vote (Christmann and Danaci 2012). Thus, in dealing with social conflict in direct democracy, the political elites have a great responsibility for the quality of campaigns, which corresponds with the findings of Kriesi mentioned above. Direct democracy, in the twentieth century, was able to deal with salient conflicts thanks to political parties that renounced populism and sought broadly acceptable solutions. The hope is that this will last into the twenty-first century as well (see however Papadopoulos 2009).

4.6.3 Digitalisation: Opportunity, Risk—Or Both?

Digitalisation has profoundly changed political processes. Politicians use personalised websites and social media as a most effective device to mobilise voters and bring their intents and messages immediately to the public, and this not only before elections. Citizens, in turn, use Twitter, Facebook or Instagram as swift, cheap and reliable communication channels open to all. These tools permit even groups lacking financial resources to articulate and debate their claims without intermediaries, to mobilise supporters in great number and to address politicians in a direct, public way. Finally, a great number of independent online newspapers present alternative issues, viewpoints and opinions—even those you would not normally find in the mainstream media.

No wonder that enthusiasts have celebrated digitalisation as a democratic revolution. And in a way, it really is one. But meanwhile it has become clear that this disruption is not always beneficial for democracy. Online interaction, rather than stimulating mutual respect and understanding, can lead to growing polarisation (Bail et al. 2018). It can also involve automated bots and anonymous, even malicious trolls which deliberately undermine deliberative standards. Tech giants such as Apple, Alphabet, Facebook or Twitter as well as parties and campaigners increasingly rely on algorithms to display targeted news and ads. This accelerates processes of individualisation, fosters so called ‘eco chambers’ of like-minded users and creates diffidence in the polity: if the virtual is the new real, the real must be fake.

Swiss politicians were, for a long time, fascinated by electronic voting, for elections as well as for votations in direct democracy. Their hope was to boost participation and rely on a new channel of communication: e-communication which voters use in their daily life. Some 20 years ago, several cantons began to introduce e-voting on an experimental base, allowing part of the citizenry to cast their vote using computers or smartphones. The results, after 15 years of experience in 15 cantons in more than 300 votations, were mixed. A comparison of two groups, one using e-voting, the other conventional voting (in person or via post), showed no difference in voting behaviour. Thus a ‘digital divide’ which many feared did not occur. But neither was voting turnout higher, nor did e-voting attract new groups of voters. E-voting also proved more complicated for voters than conventional forms of voting (Germann and Serdült 2017).¹²

When the federal government decided to extend e-voting to the federal level, opposition arose from an unexpected side, namely from digitalisation experts. They showed that existing e-voting software could not completely exclude the risk of being hacked. While irregularities at a single conventional polling station are usually negligible, they warned of a systemic risk in e-voting which could lead to the distortion of an entire election or votation. The Federal Council was unwilling to accept this risk and stopped the whole e-voting project in 2019.¹³

But how about collecting signatures for a popular initiative or a referendum, in which digital tools can play their strengths of quick and massive mobilisation? In contrast to e-voting, we do not have systematic evidence on the effects of e-collecting (Bisaz and Serdült 2017). Few experiences of private actors such as wecollect.ch show that e-collecting is promising: in several cases, they succeeded in collecting in shorter time a greater part of the required signatures for a referendum (50,000) or a popular initiative (100,000) than actors on the street. In its present form, e-collecting platforms, on request, mail an official form which the voter has to print, sign by hand and then mail back (postage is covered).

This is an acceleration, but does not use all possibilities of digitalisation: using an app or electronic signature, the voter could just click a ‘like’ or ‘dislike’ button to sign a referendum or popular initiative. Examples from US-States and the Netherlands show that in this way the required number of signatures can be collected in no time. As a consequence, authorities think about restricting or even banning e-collecting (Nuspliger 2018). The same could happen in Switzerland’s semi-direct democracy. If parliament passes a law for higher gasoline prices, for instance, any well-known platform for used cars could easily play a key role. Since some platforms are visited 50,000 times a day, they could collect the 50,000 ‘dislikes’ almost instantly, which makes the existing time limit of 100 days completely obsolete. At the same time, the decision to sign up to a referendum by a simple ‘dislike’ button resembles more an emotional reaction than a deliberative weighing up of the pros and cons of a collective decision.

In sum, digitalisation seems to play an ambiguous role especially for direct democracy. Its many advantages of mobilisation and extension of participative opinion-formation stand in contrast to an uncontrollable concentration of power in the hands of internet giants and a loss of quality deliberation. While optimists still ask: how can we use the innovations of digitalisation for democracy? pessimists worry: how can we protect democracy from being undermined by the digital disruption?’

4.6.4 The Political Culture of Direct Democracy: Particularities and Limits

Some Swiss may criticise their politicians, parliament, the courts, the Federal Council, federalism or power-sharing. There is one thing, however, which almost nobody would criticise: the political rights of citizens and the institutions of direct democracy. In surveys, direct democracy regularly shows up as the most precious element of political institutions, and only few interviewees agree with the idea of restricting it in favour of more parliamentary power. The fear that some of the people’s political rights may be lost if Switzerland joins the EU is one of the most important obstacles for those few who are advocating membership. For many Swiss, ‘democracy’ simply means ‘direct democracy’, and some even find it difficult to accept decisions of parliament or the Federal Council as truly democratic.

Against the background of the high esteem for such ‘self-rule’, one would expect the Swiss to be particularly participative in economic and social life. An unbiased outside observer, however, would probably be astonished that the values of direct democracy have not had more impact on Swiss society beyond politics. He would find no evidence that Swiss schools are more participative than those in the Netherlands or Italy. Moreover, our observer might be stunned to realise that workers and employees in Switzerland have fewer formal rights of codetermination at the workplace than their colleagues in Germany or Sweden, despite the fact that Swiss employers and unions have been practising social partnership for over 80 years, since 1937.

We may conclude that direct political participation has had little influence on Swiss economic and social life. Rather it is conceived as the specific Swiss culture of institutional democracy. With such a perspective, we can better understand the popularity of people’s political rights. They are valued as embodying the self-rule of citizens and ensuring control over the political elite. At the same time, direct democracy is considered to be one of the most important particularities distinguishing Switzerland from other countries.