Kris William Kobach, The Referendum: Direct Democracy in Switzerland, 1993.

Introduction

Switzerland has held more nationwide referendums since it introduced the institution in 1848 than all other countries combined, since the emergence of the modern nation-state.1 Among democracies in the postwar period, Switzerland accounts for more than two-thirds of the total.2 In fact, no other state in the world even comes close in applying direct democracy to national political questions.3 By the end of 1992, Switzerland had held a total of 398 nationwide referendums, covering virtually every sphere of government activity. Australia, in a distant second place, had held only 43. In most years, the Swiss voter is called on to decide 6-12 national questions, which are typically spread over 2-4 separate ballots. In addition, he will be asked to vote in numerous cantonal and communal referendums. When elections of national, cantonal, and communal representatives are added to these, it is no exaggeration to say that the average Swiss citizen is called to the polls more times in one year than an Englishman is in a lifetime.4 Only in California and in a few other western US states do referendums play such an important role in daily political life. However, there the referendums concern only sub-national issues and are unable to significantly affect such policy areas as foreign relations and military affairs. Furthermore, referendums in US states operate in the long shadow of national politics, where direct democracy is absent.

The dearth of referendums elsewhere does not mean that they have never been tried. On the contrary, a majority of European countries and more than a third of United Nations member states have experimented with the device. The only countries that have been democracies from a date prior to 1900 and have never held a nationwide referendum are the United

States and the Netherlands. However, most governments have been reluctant to institutionalize the referendum or take significant risks with it. They have dabbled in the use of what Gordon Smith classifies as controlled' referendums " in which the government decides whether or not to hold it, when it will take place, and how the question will be asked.5 Of course, some controlled referendums backfire, as in the French vote of 1969 concerning the reform of the Senate and local government. President Charles de Gaulle miscalculated popular support, linked his political fate to the question, lost, and resigned.

In Switzerland, all national referendums are uncontrolled. In fact, the parliament and the government are expressly prohibited from calling optional referendums. Direct democracy in Switzerland is not a weapon employed by central authority. On some questions, it is a constitutional necessity; but in most cases it is introduced by a petition of citizens. Referendums in Switzerland are a central element in the political system and a definitive aspect of the country's political culture. They may be categorized as governing' referendums. That is, they play an integral role in the day-to-day governing of the polity and occur relatively frequently. The overwhelming majority of referendums held elsewhere in the world have not been of this type. They have either been 'state-defining' referendums or deadlock-breaking' referendums. State-defining referendums are one-off projects designed to establish the legitimacy of territorial borders or to usher in a new political regime. They are not part and parcel of a country's political system. For example, the May 1980 referendum in Quebec on the question of independence from Canada fell into this category.6 This case was unusual, in that the result was not a foregone conclusion. Most state-defining referendums are relatively low-risk moves for the governments involved, and results yielding an affirmative vote of more than 90 percent are quite common.7 Deadlock-breaking referendums, the third category, usually serve as politically-expedient escape routes for governments which are divided over controversial questions. In such instances, the government may be in danger of splitting on the issue. This was the case with Britain's June 1975 referendum on EC membership.8 The deadlock-breaking referendum is also used in less urgent circumstances, when a government simply wishes to avoid taking a stand on a particularly divisive issue. Instead, it turns the issue over to the people, abnegating responsibility in an effort to dodge political flak.9

However, governments are rarely ambivalent. The fact that referendums can easily go the wrong way explains why those in authority have been so reluctant to call them. In Switzerland, the option lies not with the government, but with the voter. This is particularly evident in the employment of a certain species of referendum " the popular initiative.^ This instrument allows citizens the right to propose legislation or constitutional amendments which are put to a popular vote and must be implemented by the government if passed.11 The only other country to allow voters this right on national questions is Italy,12 although the initiative also exists in one form or another in 23 US states and in the District of Columbia.

Clearly, Switzerland has carried the practice of direct democracy to a level which no other nation has reached. As David Butler and Austin Ranney wrote in their comparative study of referendums in various nations, 'Switzerland stands out as the only country that has become addicted to the referendum'.13 The predominant explanation for the Swiss attachment to this institution is its longstanding experience with direct government in citizen assemblies prior to the emergence of the referendum. In 1991, the Helvetian Confederation celebrated its 700th birthday. Swiss direct democracy is almost as old, the first direct vote of citizens on policy being documented in 1294 in the canton of Schwyz. Popular legislative assemblies, or Landsgemeinden, were used in several of the mountain cantons from the thirteenth century onward.14 The resilience of the Landsgemeinden was a major reason that representative, parliamentary institutions never fully took root in Switzerland.15 However, population growth in the nineteenth and twentieth centuries rendered the Landsgemeinden impractical in most cantons. Referendums and initiatives came into common usage as a way of preserving the tradition of direct legislation. Similarly, referendums in American states have their roots in New England town meetings and direct government on the frontier.

This study is not primarily concerned with the reasons underlying Switzerland's addiction to the referendum. Rather, it seeks to identify and analyze the effects of the institution upon the Swiss political system. Many characteristics of modern Swiss politics can be attributed, directly or indirectly, to the influence of referendums. The primary objective of this study is to pinpoint such attributes. A secondary objective is to draw more general conclusions which are applicable across national borders. There are clearly lessons that can be distilled from Swiss experience and applied to other nations that use, or are considering the use of, referendums.

Obviously, the referendum plays a slightly different role in every state that makes use of it. Historical circumstances and national political cultures vary, as do modern governmental environments. Operating within different political structures and tethered by different constraints, the referendum can hardly be expected to have identical repercussions everywhere. The Swiss political system swims against the prevailing international current in several ways, most notably with its federal structure involving relatively small sub-national political units. However, this international variety of systems does not render analysis meaningless. If it did, the field of comparative political science would be of little worth. Although the study of Switzerland cannot reveal universal truths about direct democracy, it can shed light on referendums elsewhere, as long as specific circumstances and local idiosyncrasies are taken into account. Thus, the impact of Swiss referendums on the behavior of Federal Assembly members may not always be relevant in Britain; but it is likely to offer numerous valuable insights into the behavior of politicians in California. In several respects, such as the involvement of mobilized interest groups in referendum campaigns, the Neinsager (nay-sayer) phenomenon, the fragmentation of parties, and the consequences of low participation, the Swiss experience offers lessons relevant to the use of referendums in any country, even if the devices are of the state-defining or deadlock-breaking type.

This study focuses primarily upon the federal, or national, referendum in Switzerland. Any serious analysis of the cantonal referendums, which operate in 26 distinct political environments with varying political structures and procedures, could fill 26 separate books. Compressing such a quantity of information into a manageable size for inclusion in this study is simply not feasible.

Before continuing, a few notes regarding terminology are in order. Although referendums are nearly as old as democracy itself, the term is not. The alternative term, plebiscite, has a tnuch older lineage. It is derived from the Latin plebiscita, describing votes of the Roman plebs in the fourth century B.C. The term was applied to popular consultations in France from 1793 onward. It was also used to describe votes resolving League of Nations boundary disputes and used in Nazi Germany to describe popular votes legitimizing the regime's policies. The term referendum can be traced to seventeenth-century Switzerland, in which members of the Confederal Diet (of the thirteen-canton Confederacy, 1513-1798) took policies back to their respective cantonal councils or populations ad referendum et instruendum (for referring back and instructions). In its modern form, the term was first used in the constitutional plebiscite attempting to legitimize the new regime of the Helvetic Republic, which was imposed by the French conquerors in 1798.16 The word first appeared in English in its current sense in the 1880s.17

Although there is no agreed line of distinction between the terms plebiscite and referendum, the former is more often associated with ad hoc popular votes to endorse a regime, a man, or a specific policy. Current usage tends to favor the word referendum, as do the Swiss in describing their own institution. I will therefore do the same. A final note concerns the plural form of the word. Many dictionaries recognize both 'referendums' and the Latin version, 'referenda'. (The plural form in Swiss German is Referenden.) However, current writings on the subject use the former most frequently.

The significance of referendums in the world today

Since the early years of the twentieth century, referendums have increased in frequency. As Table 0-1 illustrates, this has been true not only in Switzerland, but also throughout the rest of Europe. In the early 1970s, worldwide interest in and usage of the referendum surged dramatically.18 The enlarging of the European Community gave rise to five referendums. In April 1972, France let its people voice their opinion on whether the newcomers (Ireland, Denmark, Norway, and Britain) should be accepted; and 67.7 percent of the French voters approved. Each of the four countries intending to join put the membership issue to their citizens eventually, and all but Norway received popular endorsement of the idea. The Norwegians narrowly rejected membership in September 1972, with 53.5 percent voting no. In Britain, the June 1975 EC vote was preceded by a March 1973 vote in Northern Ireland on the question of withdrawal from the United Kingdom (although the result was rendered largely meaningless by separatist boycotts); and in March 1979, the question of devolution was put to voters in Scotland and Wales.19 The referendum also began to play a critical role in facilitating the transition to democracy in previously authoritarian nations. In Greece, citizens voted in December 1974 to close the curtain on the monarchy, with 69.2 percent voting in favor. Two years later, Spanish voters registered overwhelming support for their own political reform package, with 94.2 percent approving.

In the 1980s and early 1990s, the surge of referendums continued unabated. A number of ballot issues during this period were particularly notable. .The questions of divorce and abortion were put to the predominantly Catholic voters of Ireland. Italy had placed the same questions on the ballot in May 1974 and May 1981, respectively. The Italians opted to allow both practices.20 The Irish voters went the other way, rejecting abortion in September 1983 and turning down the legalization of divorce in June 1986.21 Reconsidering the abortion issue in November 1992, they rejected a proposition that it should be allowed only when the mother's life is threatened. However, they approved the right to travel to another EC country for an abortion and the right to receive information about abortions performed abroad.

In March 1986, Spanish voters endorsed their government's decision to remain in NATO, with 56.9 percent in favor. This referendum drew wide international attention and even attracted direct campaign involvement from other countries. The most dramatic surge in referendums outside of Switzerland occurred in Italy, where fifteen ballot issues were decided between 1981 and 1991. In contrast, there have been none in Britain since 1979. However, the referendum has been repeatedly suggested as a means of addressing such controversial issues as capital punishment and the adoption of proportional representation.

From 1989 onward, fledgling democracies in Eastern Europe looked to the referendum for the legitimization of political change. For example, in November 1989, Hungarian citizens voted overwhelmingly to disband the worker's militia, ban political activity in the workplace, and force the former Communist Party to disclose its financial assets. On a fourth question, they narrowly approved parliamentary election of the nation's first President, rejecting popular election of the office.22 Similarly, in a December 1991 referendum, more than three-quarters of Romanian voters approved the country's new, democratic constitution.23 In 1990 and 1991, numerous republics of the Soviet Union held referendums on the issue of independence, increasing the centrifugal pressure which ultimately led to the disintegration of the country. The evolution of the EC in the early 1990s further heightened the importance of referendums in Europe. The ballot issues of 1992 left a particularly deep imprint on the organization. In early June, 50.7 percent of the Danish people spurned their government's advice and rejected the painstakingly crafted Maastricht treaty for the closer economic and political union of the EC countries. In doing so, this bare majority of Danish voters derailed the steam engine of European monetary union and placed a sizeable obstacle in the way of a unified European foreign and defense policy. Although Irish voters approved the Maastricht treaty by nearly a 70 percent majority in late June, the victory was not enough to fully resuscitate the agreement. Amendments to the Treaty of Rome, such as those contained in the Maastricht document, require the approval of all member nations of the EC. Nonetheless, supporters of the treaty hoped to put it solidly back on the tracks with a resounding affirmative vote in France in September. Although the Danish popular veto limited the legally possible options under Maastricht, President Mitterand hoped to deliver a much-needed dose of approval for what remained of the treaty. When he had called the optional referendum in June, French opinion polls showed a two-to-one majority in favor. However, when the vote finally came, the French offered only a halfhearted endorsement. The treaty passed with a slim 51.1 percent majority.24 It was a pyrrhic victory in light of widespread expectations that France would strike the spark to restart the Maastricht engines. In the end, the result only reinforced doubts on the part of other EC governments. The Danish and the French referendums of 1992 produced reverberations impossible for the rest of Europe to ignore.

Shortly thereafter, in October 1992, direct democracy took center stage on the Canadian political scene. The last nationwide referendum in Canada (on conscription) had been in 1942. In the 1992 vote, 54 percent of Canadians (and six of the ten provinces) rejected the Charlottetown accord, a package of constitutional reforms designed primarily to quell demands in Quebec for greater provincial autonomy. Coming at the end of twelve years of constitutional haggling, the agreement would have recognized the predominantly French-speaking province as a 'distinct society'. In Quebec, separatists seeking complete autonomy combined with federalists, unwilling to concede any special treatment for the province, to produce a no vote of 55 percent. Because a rejection by any one of the provinces would have been sufficient to defeat the accord, the six-province thumping dispatched it convincingly. As a result, this brief episode of direct democracy created an important reference point for subsequent discussions of the country's future.

Although the United States is one of the two long-established democracies never to have held a nationwide referendum, a constitutional amendment to introduce the practice was widely discussed in the 1970s. In 1977, the Senate held hearings on a proposed national initiative. Surveys by the Center for Political Studies had indicated popular support for the idea. In 1970, 1972, and 1974, respondents were asked whether more issues should be decided directly by voters at the polls. In each case, 77 percent or more answered yes.25 A later Gallup poll in 1978 concerning the specific proposal to introduce a national initiative showed a smaller majority of 58 percent in favor.26 Despite the fact that Congressional consideration of the device came to naught, public support for direct democracy at the national level has remained fairly high. A 1987 survey found 58 percent in favor of a national advisory referendum.27

Although worldwide interest in and usage of the referendum has increased monumentally, it is important to bear in mind that, as Butler and Ranney conclude, there are effectively two worlds of referendums.28 The first includes Switzerland, California, a few other US states, and increasingly Italy, where direct democracy is a central element of daily political life. The potential referendum or initiative is a constant factor that legislators and other political actors must always consider. The second world is one in which referendums are used infrequently, on an ad hoc basis. Most of the rest of the world falls into this category. However, some countries, such as Australia, New Zealand, and Ireland, fall somewhere in between. They have held an unusually large number of referendums, but the device remains tangential to the basic governing of the polity.

Any country considering the use of the referendum for state-defining, deadlock-breaking, or governing purposes would be well advised to consider Swiss experience with direct democracy first. The use of the referendum in Switzerland goes far beyond territorial and constitutional matters or the isolated controversial policy. It has become a commonplace facet of the legislative process. The Swiss have applied the referendum to virtually every sphere of government activity. In fact, one is hard-pressed to find a question put to referendum elsewhere that the Swiss people have not already considered. Nearly 11/2 centuries of referendums at the national level have profoundly influenced Swiss political culture and institutions. Clearly, the consequences in Switzerland are unlikely to be duplicated precisely in nations that use referendums only infrequently. However, even in these countries, numerous tendencies exhibited in Switzerland still

apply-

Furthermore, the Swiss themselves could benefit from a closer analysis of the institution. Surprisingly, the Swiss have produced relatively little thorough research regarding the impact of direct democracy upon their political system. There is a tendency to take direct democracy for granted, which further discourages critical and extensive examination of the subject. Many aspects of governmental, parliamentary, and interest group behavior are directly or indirectly the result of operating in an environment marked by frequent national referendums. However, this causal relationship is often overlooked.

Swiss direct democracy has also evolved considerably in recent decades with little serious analysis of these changes. For example, the number of constitutional initiatives each year has been growing, more or less constantly, since 1969. The initiative is now the most frequently occurring type of ballot issue at the federal level in Switzerland. This change has had a pronounced impact on the political system, which Chapter Four explores in detail. A more general development has been the post-1975 polarization of party competition. As I argue in Chapter Six, the referendum has strongly affected the extent and consequences of this shift. Another phenomenon to emerge in recent decades is the problem of extremely low voter participation in some referendums. As discussed in Chapter Five, this can lead to distortion in referendum outcomes and to the exploitation of apathy' by highly mobilized special interests. The analysis coming out of Switzerland on this problem is surprisingly slim, although the Research Center for Swiss Politics at the University of Bern has fattened the available data considerably by producing case histories of every federal referendum from the late 1970s to the present.

Aside from a handful of isolated articles, virtually nothing has been written in English on the subject of Swiss direct democracy. Furthermore, astonishingly little has been written about the referendum device in general. Most of what exists was produced shortly after the turn of the century when the subject was in vogue. There is only one comparative analysis of the modern device, which was written in 1978.29 There have also been a few studies of the referendum at state level in the United States;30 but these have made little effort to draw conclusions applicable in other political systems. There has also been scant research on electoral behavior in referendums. This is partly due to the fact that the countries in which electoral research is most advanced have had few nationwide referendums.31 The rising importance and usage of direct democracy around the world has made the need to understand the broad impact of referendums on political systems all the more pressing. This study represents an attempt to meet that need, as well as to fill the void in research concerning Swiss referendums.

This investigation takes an historical approach in analyzing the Swiss referendum, attempting to glean useful information from a close scrutiny of past referendums and from a broader look at historical trends. Wherever possible, statistical evidence has been presented to support any conclusions. A wide variety of sources were used in conducting the research necessary for this project. One of the most important was a survey of the members of the Federal Assembly carried out in April-June 1990. Most of the results of this polling are discussed in Chapter Six. I also conducted numerous personal interviews with politicians, government officials, interest group representatives, and academics in Switzerland. In addition, contemporary Swiss newspaper articles proved essential in weaving together a complete (and more objective) account of events. As mentioned above, secondary sources were few and far between; I have incorporated their conclusions into this study where possible. All translations into English in this work are my own, except where otherwise noted.

The questions posed

It is not within the scope of this study to explore every aspect of the referendum's impact upon the Swiss political system. Instead, I focus on nine specific issues and attempt to offer solid conclusions on each. They are as follows.

2. What changes have occurred since 1848 in the operation and effects of the Swiss referendum ?

The referendum of today functions very differently from the institution that was first introduced at the federal level in 1848. In particular, the evolution of consensual democracy completed in 1959 was both influenced by, and in turn exerted an influence upon, the use of the referendum. This question is addressed at two points; Chapter One provides an historical account of the development of the device, and Chapter Three identifies relevant trends in voting outcomes and participation levels.

2. Is there a conflict between the goals of consociational democracy and the operation of referendums; and if so, is this a flaw in the Swiss system?

The referendum is fundamentally majoritarian in nature. There is no middle ground and no means by which parties or groups may compromise when an issue is subjected to popular vote. The majority of the voting electorate wins, and the minority must accept the consequences. This fact would seem to diminish the spirit of amicable agreement that is assumed to permeate Swiss consociational democracy. Is this a contradiction inherent in the Swiss system which impedes the functioning of both approaches? Or, is one needed to offset the defects of the other? I attempt to answer these questions in Chapters Six and Nine.

3. To what degree is the referendum responsible for the weakness of Swiss political parties?

Parties in Switzerland are notoriously impotent, lacking several of the vital roles played by parties in other democracies. Generally, parties in representative democracies enjoy a central position in channelling political competition and conveying public sentiments at all stages of the legislative process. Parties link voters to legislatures and executives by nominating political candidates and by packaging issues into bundles presented to voters at election time. They also attempt to set the agenda and guide the deliberations of legislative assemblies. Adding direct democracy to this setting dramatically changes the effectiveness of parties in each of these capacities. In Chapter Five, I look at the weakness of Swiss parties and explore the ways in which referendums may be responsible. In Chapter Six, I focus specifically on the legislative process.

4. To what degree is the referendum responsible for the strength of interest groups in the Swiss political system?

While Swiss parties are relatively weak, Swiss interest groups are unusually powerful. The dominance of a handful of Verbdnde has even prompted some observers to describe Switzerland's institutional framework as corporatist. Swiss special interests have clearly taken advantage of the opportunities afforded by direct democracy, assuming a pre-eminent role in the challenging of laws and launching of initiatives. The extent to which their impressive systemic influence is attributable to the existence of direct democracy is assessed in Chapters Five, Six and Eight.

5. How does the referendum alter legislative functions and behavior?

Direct democracy, as manifest in Switzerland, undercuts any pretensions of parliamentary sovereignty. The people have the final verdict on all laws, whether they exercise this prerogative passively or actively. This potential veto looms over every stage of the Swiss legislative process. Chapter Six investigates the various ways in which this affects parliamentary functions and behavior.

6. What effect does the referendum have on political competition?

This is a question to which advocates and opponents of direct democracy continually claim to know the answer. It is often said that referendums are introduced by liberal, progressive interests in most political systems only to benefit conservative, reactionary interests. I address this supposition at several points in succeeding chapters. There are also numerous other aspects of political competition which are affected by the referendum. It influences the formation of parliamentary alliances, the stability of governments, and the resolution of particularly divisive or explosive issues. These effects are considered in Chapters One, Three, Four, and Eight.

7. Which influences affect voter decisions on ballot issues most strongly?

A wide variety of actors and interests compete for voter attention in the run-up to a referendum. Parties attempt to sway their faithful, as do interest groups. Massive expenditures are devoted to advertising competing claims. This is a huge area of study, and a thorough exploration would require a great deal more time and space than that offered by this project. However, a number of basic observations can be made by looking at the historical record in Switzerland and by comparing Swiss results with outcomes in other countries. Such observations are offered throughout this study, in Chapters One, Three, Four, Five, Seven, and Eight.

8. Does the referendum serve to protect or endanger minorities?

As mentioned above, the referendum is fundamentally majoritarian in its operation. On the other hand, it also serves as a check on government power; and most institutional checks work to protect political and social minorities. Switzerland is an extremely heterogeneous country, with deep linguistic, cultural, religious, and regional cleavages. Thus, the opportunities for the oppression of minorities are numerous. However, since the Sonderbund War of 1847, there has been little evident tyranny of the majority. Whether this has been because of, or in spite of, referendums is considered in Chapters One and Six.

9. How might the mechanisms of direct democracy be improved?

Different polities use different procedures in employing direct democracy. The Swiss procedures are detailed in Chapter One and summarized in Table 1-2. Chapter Eight adds to the examination of Swiss referendums a short, comparative analysis of referendums in Australia, Italy, and California. It also offers a few observations regarding the effectiveness of various rules and procedures.

Before undertaking these matters, one final point must be made. There are three definitive elements of modern Swiss political institutions and political culture: direct democracy, consociationalism, and federalism. These elements are interlinked in various complex and subtle ways which I attempt to illustrate in this study. I devote particular attention to the manner in which direct democracy serves to reinforce the other two elements, especially consociationalism. The referendum promotes consensual behavior in political actors, safeguards consociational decision making from usurpation by the government versus opposition mode of lawmaking,- and compensates for the defects of consociationalism. This relationship is discussed throughout the following investigation. I begin with a short history of the referendum in Switzerland. It is only in this context that one can truly come to grips with its role; and I believe that, to a large degree, the events speak for themselves. I then turn to the 'political science' of the issue, briefly outlining the common arguments and assumptions about direct democracy in Chapter Two. In the succeeding chapters, the specific effects of the referendum upon the Swiss political system are addressed.


Notes

1. Between 1793 and 1978, there were 257 nationwide referendums in all nations combined, except Switzerland. In Switzerland in the 1866-1978 period, there were 296 referendums. The gap has widened further since 1978. See David Butler and Austin Ranney, eds., Referendums: A Comparative Study of Practice and Theory (Washington, D.C.: American Enterprise Institute, 1978), pp. 227-36.

2. Arend Lijphart, Democracies: Patterns of Majoritarian and Consensus Government in Twenty-One Countries (New Haven: Yale University Press, 1984), p. 201.

3. In this study, I often use the phrase 'direct democracy' to mean 'referendums and initiatives', even though, technically, the Swiss devices exemplify semi-direct democracy. This practice of referring to referendums and initiatives as 'direct democracy' is fairly common. Some writers also use the term to denote the recall device and primary elections. In this study, the term is not intended to encompass these devices.

4. Christopher Hughes, Switzerland (New York: Praeger, 1975), p. 144.

5. Gordon Smith, The Functional Properties of the Referendum', European Journal of Political Research, 4, 1 (Mar. 1976) 6.

6. The referendum failed, with 40.4 percent opting for secession.

7. Butler and Ranney, pp. 227-46.

8. 67.2 percent of voters opted to stay in the EC.

9. Obviously, the distinctions in this tripartite typology are not perfect. Some referendums in various countries combine characteristics of two, or even all three, types.

10. In this study, I often use the word 'referendum' in its more general sense, meaning 'the referendum and the initiative' or 'any direct popular vote on a ballot issue'. The initiative, although a distinct and separate mechanism, is treated as a subcategory of referendums. Where possible, I use the term 'referendums' instead of 'referendums and initiatives' for the sake of brevity.

11. Only the constitutional initiative exists at the federal level in Switzerland. However, both constitutional and legislative initiatives are allowed in many Swiss cantons.

12. Technically, the Italian device must be classified as a referendum, because it only allows citizens to abrogate existing laws (partially or entirely); and it does not permit the specification of what is to replace the defeated legislation. However, in practice it operates as an initiative, because virtually any law may be challenged at any time (rather than only immediately after the law's passage). Furthermore, it is de facto possible for campaigns to dictate what the replacement legislation will be.

13. Butler and Ranney, p. 9.

14. Ibid., p. 6.

15. Jonathan Steinberg, Why Switzerland? (Cambridge: Cambridge University Press, 1978), pp. 73-74.

16. Hughes, p. 128.

17. Butler and Ranney, p. 4.

18. From 1970 to mid-1978, inclusive, 63 nationwide referendums were held outside of Switzerland. Over the previous 81/2 years, there had only been 38. This surge continued into the 1980s, as well. Over the 8 1/2 period from mid-1978 to 1986,53 were held.

19. 51.6 percent of Scottish voters who went to the polls voted for devolution; but the government had stipulated that 40 percent of the entire electorate must assent. The yes voters only represented 32.9 percent of the electorate; so the referendum failed to satisfy the requirement. In Wales, only 20.9 percent of those who voted favored devolution.

20. 59.1 percent voted in favor of allowing divorce, and 67.9 percent voted to allow abortion.

21. 66.9 percent voted to ban abortion, and 63.5 percent rejected the legalization of divorce.

22. Ernest Beck, 'New President will be elected by Parliament', The Times, 28 Nov. 1989.

23. 'And so to Winter', The Economist, 14 Dec. 1991, p. 66.

24. 'A Vote too Far', The Economist, 26 Sept. 1992, p. 48.

25. David Magleby, Direct Legislation: Voting on Ballot Propositions in the United States (Baltimore: Johns Hopkins University Press, 1984), p. 10.

26. Ibid., p. 13.

27. The referendum stipulated in the survey would be held on certain proposed laws every two years at the time of national elections. See Thomas E. Cronin, Direct Democracy: The Politics of Initiative, Referendum, and Recall (Cambridge, Massachusetts: Harvard University Press, 1989), p. 4.

28. Butler and Ranney, p. 221.

29. Ibid.

30. For example, David Magleby, Direct Legislation: Voting on Ballot Propositions in the United States (Baltimore: Johns Hopkins University Press, 1984) and Thomas E. Cronin, Direct Democracy: The Politics of Initiative, Referendum, and Recall (Cambridge, Massachusetts: Harvard University Press, 1989).

31. Butler and Ranney, p. 15.