Kris William Kobach, The Referendum: Direct Democracy in Switzerland, 1993.3 Trends in the exercise of direct democracy
A wealth of understanding regarding the impact of referendums in Switzerland can be gained by simply looking at the various issues that have been raised and at their respective outcomes. Table 3-1 lists all of the federal referendums that have been held in Switzerland from 1848 through to the end of 1992.
Participation levels
In 1964, Max Imboden warned that declining participation levels in Swiss referendums foreshadowed a coming democratic malaise'.1 By the late 1970s and early 1980s, it seemed to some observers that participation figures were in an inexorable decline, unlikely to return above the 40 percent level except in isolated instances. John Austen, David Butler and Austin Ranney pointed to a disturbing 37 percent average rate of participation from 1978 through 1986.2 The trend was widely attributed to the fact that Swiss voters were being called to the polls so many times each year. It seemed that the Swiss people had been saturated with direct democracy. Evidently, they were burned out. Yet, in the late 1980s the downswing reversed, with a relatively high turnout for a number of controversial referendums. On the polarized December 1988 ballot issue of prohibiting land speculation to reduce rents, 52.8 percent turned out to vote; and the question of whether the army should be abolished drew 68.6 percent to the polls in November 1989. Then, in December 1992 a massive turnout of 78.3 percent was recorded for the referendum on Switzerland's membership in the European Economic Area, the highest since 1947. Over the three-year period of 1987-89, participation averaged 48.2 percent, marking a significant rebound from the lackluster turnout of 1978-86. This rally was followed by two relatively apathetic years in 1990 and 1991, in which turnout averaged only 36 percent. In 1992, participation jumped again, up to 53.9 percent. These snapshot views illustrate how the trend can appear to be upward or downward, depending upon which years are taken into account. Nonetheless, if the 1987-92 period is viewed as a whole, the average participation rate of 45.8 percent still looks considerably better than that of the 1970s and the first half of the 1980s. Overall, it is fair to say that, since the early 1960s, the gradual decline of previous decades has levelled out. Average voter participation figures for the last three decades have remained fairly close to the 40 percent mark.
Of course, there has been a noticeable decline in voter turnout since the federal referendum was first introduced in 1848. Table 3-2 provides a circumspect picture of voter turnout since the 1880s, when participation levels were first routinely calculated. In the turbulent 1930s, average participation figures for the decade reached a high-water mark of 64.6 percent. In the 1933-35 period, in particular, turnout was extraordinarily high (averaging 74.6 percent). This was almost certainly a popular reaction to the perceived fragility of democratic systems at a time when democracies all around Switzerland were giving way to authoritarianism. In such an environment, taking part in order to preserve democratic institutions from decay was of paramount importance. At the same time, political extremists were keen to participate in order to usher in the sweeping changes they desired.
As Table 3-2 illustrates, the 1970s and 1980s approximately repeated the experience of the 1960s. The most dramatic decline occurred in the 1940s, 1950s, and 1960s. Decade averages slipped by 21.4 percent during this period. Since then, participation levels have remained at a plateau of just over 40 percent, with intermittent peaks for highly controversial issues and valleys for more mundane questions. It is therefore debatable whether Switzerland is in a state of democratic malaise. According to surveys conducted by the University of Bern in 1985, 31 percent of Swiss voters can be classified as regular' participants in referendums.3 These voters take part in most, if not all, federal referendums. On any given question, a number of 'occasional' participants will show up at the polls adding their numbers to the total turnout. These occasional voters, which comprise 69 percent of the electorate, participate in some referendums and abstain in others. Thus, the 40 percent or so that vote in a given referendum are not entirely the same group of people that voted in the previous one. If the question is particularly controversial, a greater number of occasional voters are likely to participate. For example, the December 1992 referendum on the government's proposal to join the European Economic Area elicited the greatest response among occasional voters in decades, with 78.3 percent of the electorate participating. Of those taking part, more than half (47 percent of the total electorate) were occasional voters.
Certain social groups in Switzerland are more likely than others to vote in referendums regularly. There is a definite age bias in most referendums; 42 percent of citizens in the 65-84 age group vote regularly, compared to 24 percent in the 20-29 age group. There is also an educational bias. Of those who attended gymnasium or university, 40 percent vote regularly. In contrast, 25 percent of citizens who only completed their obligatory schooling are regular voters. Of the main political parties, the Christian Democrats count the greatest proportion of their sympathizers as regular voters â€" 43 percent. The lowest percentage of regular voters is found in the Social Democratic Party â€" 35 percent. However, this is still above-average. Citizens without party sympathies are the least consistent participants. Only 24 percent of people in this category are regular voters. Certain other attributes, such as home ownership and self-employment, also correlate with habitual participation. These characteristics often reflect social integration. In general, the more socially integrated a voter is, the more likely he is to participate in referendums regularly.4
In addition to the amount of controversy surrounding an issue, the mobilization of occasional voters is also affected by the complexity of the question. The difficulty of making a decision affects participation greatly, with extremely technical or complicated issues tending to yield the lowest participation levels. The lowest participation level for any federal referendum was recorded in June 1972, when a two-question ballot concerning monetary policy and control of the construction market drew only 26.7 percent of voters to the polls. In the following year, a government-proposed constitutional amendment concerning education took second place with 27.5 percent participation. Both cases involved fairly complex issues. This tendency is confirmed by survey data. In the March 1986 referendum on UN membership, 70 percent of Swiss citizens found it rather easy' to make up their minds. Turnout was an above-average 50.7 percent. Similarly, 83 percent of citizens described the May 1978 question on whether there should be twelve Sundays a year without motor traffic as easy to decide. In this case as well, participation was above-average, at 48.8 percent. In contrast, only 32 percent of voters found it easy to reach a decision on the February 1978 economic policy amendment to the Constitution.5 As a result, more than 110,000 voters who went to the polls to cast a vote on changing the legal age of retirement (another issue on the same ballot) left their ballot blank on the economic policy question.
Straightforward questions on controversial issues are the most likely to yield high participation. For example, the highest recorded turnout on any Swiss referendum was 86.3 percent. This occurred in December 1922, when 87 percent of those voting spurned the hotly-disputed capital tax initiative. The highest turnout in the postwar period was sparked by the question of introducing a comprehensive old-age pensions scheme in July 1947, on which 79.7 percent of the electorate voted. This was a second attempt at establishing the program, and it won an impressive 80 percent majority. Clearly, the issue itself was important in stimulating participation; on the first time around in 1931 it also drew masses to the polls, attracting 78.1 percent of the electorate. The second highest postwar turnout came in December 1992 with the question of joining the European Free Trade Association; 78.3 percent of Swiss voters took part. This proposal was extremely controversial, involving Switzerland's relationship with the EC and coming on the heels of the Danish rejection of the Maastricht treaty. In all there have been only five referendums on which more than 80 percent of citizens have voted: in 1872 and 1874, over the total revision of the 1848 Constitution; in the 1922 capital tax vote; in 1933 when a law to reduce the salaries of federal civil servants was challenged; and in 1935 over a Social Democratic initiative to combat the economic crisis. None of these were particularly complicated proposals, and all generated controversy. The two examples in the 1930s occurred in a highly-charged political atmosphere, another factor which undoubtedly boosted participation.
Given that the nature of the issue affects participation, one might therefore suppose that multiple-question ballots are more likely to conjure a high voter response than are single-question ballots. The number of questions on a given ballot is essentially a matter of timing, bearing little relation to the controversy or importance of the issues. However, the setting of dates for particular ballot items, especially initiatives, can be manipulated to affect outcomes. This tactic is discussed further in Chapter Four. Nonetheless, the number of questions on each ballot is rarely, if ever, manipulated for political reasons. Therefore, it is possible to make a valid comparison between participation levels on omnibus ballots and those on single-issue ballots. It seems logical that the former would engender higher turnout. If numerous questions are presented on the same ballot, then some people will be particularly interested in issue number one, others in issue number two, and so on. Taken together, these various groups should combine to vastly exceed the turnout on most single-question ballots. Strangely, this has not been the case. In fact, there is a tendency for omnibus ballots to elicit a lower voter response, as Table 3.3 illustrates. In 8 of the 11 decades considered, participation levels on single-question ballots were higher, in most cases significantly higher. The fact that this pattern has recurred so often for more than a century means that it is probably more than coincidence. Although a relationship seems to exist, explaining it is not easy. It may be that the presence of several questions on a single ballot devalues the appeal of each individual question. More importantly, it seems that voters prefer simple decisions and that a ballot with numerous questions dissuades many from taking part. It may also be that the public debate over numerous, unrelated questions tends to blur the focus of attention that might otherwise rest upon a single issue.
The variation in turnout produced by different ballot combinations and different issues, combined with the fact that nearly 70 percent of the Swiss electorate can be classified as occasional voters, suggests that the 40 percent average participation level may not tell the whole story. Over a typical four-year period, Swiss citizens may be confronted with twelve or more federal referendums. Occasional voters pick and choose the among the issues, taking part only in those which interest them. It is not unreasonable to suggest that more than 75 percent of Swiss citizens vote at least once in a four-year period. Thus, if we count the number of citizens who vote in a referendum at least once every four years, participation levels are likely to compare favorably with election and referendum turnout in other democracies.
In any case, participation levels in Swiss referendums have undeniably fallen since the 1940s. What explanations can be given? As mentioned above, it has been suggested that Swiss voters simply got tired of direct democracy. However, there are reasons to doubt this burn-out theory. Experience with direct democracy in the United States suggests that the opposite is true. In those states with more frequent referendums (most of which place the questions on election ballots, alongside candidates), voting turnout has been considerably higher than in other states.6 It is therefore by no means inevitable that frequent referendums generate an endless spiral of increasing voter apathy. A second cause for skepticism is the timing of the drop in participation. Why did it occur when it did and not earlier? Why did the marked increase in the number of ballot issues in the 1970s not produce a further drop in participation? More to the point, participation levels did not change significantly after women were given the vote at the federal level in 1971. If one assumes that the primary cause of the falling figures is apathy generated by the burden of having to vote so frequently, then the turnout figures should have jumped with the influx of women voters. This new electorate was certainly not tired of having to vote so often. If anything, they should have been particularly keen to exercise their new-found political rights. In sum, the burn-out theory is shaky at best.
However, statistics on women's voting suggest an explanation for the slight drop in average turnout in the 1970s and 1980s, compared to that in the 1960s. Newly-enfranchised groups in virtually all democracies tend to participate in low numbers initially. For example, this was the case with early twentieth-century immigrants to the United States. A history of disenfranchisement depoliticizes such groups, and it usually takes at least a generation for political integration to occur. Similarly, women in Switzerland did not flock to the polls after winning suffrage rights. Indeed, participation among female voters was considerably less than that of male voters. This was illustrated clearly in the 1971 Federal Assembly election, the first federal election in which Swiss women took part. Overall, 56.9 percent of the electorate voted â€" a marked downturn from 65.7 percent in the previous (1967) election. Barely one out of two women went to the polls, compared to seven out of ten men.7
This phenomenon was not restricted to the 1970s. In 1969, Philip Converse's comparative observations on enfranchisement suggested that it would take at least two generations for Swiss women to reach the same level of participation as Swiss men.8 This prediction seems to be holding true, both in referendums and in elections. In the March 1977 referendum, only 47 percent of women in the electorate took part, compared to 64 percent of men â€" a gender gap of 17 percent.9 In the 1979 parliamentary elections, it narrowed to 16 percent.10 The March 1985 referendum saw a gender gap of 14 percent;11 and the elections two years later registered a gap of 10 percent.12 However, the continual narrowing trend seems to have slowed or even reversed in recent years, with a 16 percent gender gap showing in participation figures for the April 1990 referendum.13 Not surprisingly, 34 percent of men are regular referendum voters, compared to 28 percent of women.14
Assuming (conservatively) that the gender gap in participation has averaged 10 percent or more for the last two decades, then this would account for a drop in overall turnout of at least 5 percentage points since 1971. In fact, current participation figures are only 3 percentage points lower, on average, than they were in the 1960s. This is due to the fact that average participation figures for the male electorate have actually increased since 1971. The sluggish pace of the political integration of women accounts for the very slight drop in overall participation levels over the last two decades. This slight upward trend in the turnout of male voters demonstrates convincingly that the burn-out theory does not apply in the Swiss case.
Accounting for the dramatic decline in referendum participation from the 1940s (when it averaged 58.8 percent) to the 1960s (when it averaged 43.2) is more difficult. Clearly, the most important political development during this period was the completion of Swiss consociational government with the magic formula of 1959. To the extent that this arrangement has minimized the number of people in society who find fault with government legislation, it has probably reduced the number wishing to express their discontent through direct democracy. Voter indifference in Switzerland may also reflect general satisfaction with the political system and with the country's impressive economic performance. Throughout the entire postwar era, Switzerland has enjoyed a relatively fair income distribution, minimal poverty and homelessness, good industry-labor relations, and none of the urban squalor that has arisen in large European cities elsewhere. This economic prosperity has been complemented by marked social and political stability. With the exception of the isolated Jura crisis of the late 1960s and early 1970s, incidents of social unrest have been relatively minor. These factors undoubtedly account for the willingness of much of the electorate to stay at home and let other citizens make referendum decisions.
Swiss parliamentary elections have always elicited greater participation, on average, than have referendums. Table 3.4 shows the voter participation level in federal elections since 1919, along with the average level of participation in referendums during the same year. In every instance, election turnout was higher than referendum turnout, usually by a margin of more than 10 percent. Although choosing among candidates for the National Council has considerably less impact on policy, the Swiss voter is generally eager to take part. In many cases, elections to national office represent little more than contests of personality, with issues playing virtually no role in the outcome. Nonetheless, these contests consistently draw crowds.
In most other countries, referendums are exceptional political events. On the whole, turnout in states with infrequent referendums has been very close to election participation levels, in some countries exceeding election turnout.15 The exceptionality of the device accentuates its importance and induces citizens to take part. In Switzerland, elections are the rare events in relative terms. It is not unreasonable to suppose that many citizens have become somewhat blase about referendums, intrigued more by the personal battles waged in elections.
Nonetheless, Swiss participation levels in postwar national elections have been the third lowest of all democracies with a continuous record of democratic elections since the late 1940s. Between 1945 and 1981, election turnout in Switzerland averaged 64.5 percent, coming in just ahead of India and the United States (58.7 percent and 58.5 percent respectively). Italy had the highest participation figure of democracies without compulsory voting laws, at 92.6 percent.16 The Swiss election of 1955 was the last in which more than 70 percent of voters turned out; and 1975 was the last in which a majority of voters took part. The most dramatic decline in participation levels occurred immediately after the entrenchment of the magic formula in 1959. This pronounced drop in turnout reflected the citizens' realization that their votes would have no effect on the composition of the national executive. The existence of the grand coalition government since 1959 is undoubtedly the primary reason for falling election participation figures.
In sum, the factors influencing voter participation in referendums are numerous and complex. One rather obvious tendency is visible â€" that controversial issues elicit higher turnout. Also, technical or complicated questions tend to produce lower participation levels. In addition, a rather surprising pattern emerges â€" multiple-question ballots generally produce lower turnout. This is probably because the apparent complexity of the ballot increases with multiple issues and because the controversy that the issues would generate individually is obscured.
Overall, voter participation levels in the 1970s and 1980s were not significantly lower than they were in the 1960s; and the slight drop is more than accounted for by the addition of women to the electorate in 1971. Participation levels among male voters have actually risen in recent decades. This fact suggests that the burn-out theory of slackening participation is inaccurate. As suggested above, the dramatic downturn from the end of the Second World War to the early 1960s was partly a reflection of Switzerland's social stability and economic prosperity and partly a consequence of the permanent grand coalition government introduced in 1959. Finally, it is worth remembering that occasional voters make up 69 percent of the Swiss electorate and that the great majority of them take part whenever a referendum subject interests them. The number of voters who participate in at least one referendum a year is undoubtedly much larger than the 40 percent of the electorate voting in any particular referendum. Thus, the impression (conveyed by individual referendum turnout figures) that the Swiss electorate is one of the most apathetic among democratic nations may be inaccurate.
Referendum outcomes
The Swiss have shown again and again that they are willing to reject the proposals and advice of their elected representatives. Legislative compromises backed by all four major parties are routinely refused by the electorate. Although the Swiss are often characterized as being conformist, thtey are certainly not deferential. As illustrated in Table 3-5, of the 265 constitutional amendments and challenged laws, decrees, and treaties put forward by the Bern government since 1848, only 164 (or 61.9 percent) have survived the referendum. Routine laws and decrees have fared considerably worse in referendums than have constitutional amendments, even though constitutional amendments require a double majority to pass. Only 47.3 percent of the former have won popular approval, while the latter have enjoyed a 71.9 percent success rate. This difference can largely be attributed to the fact that a normal law must be challenged by petition before it can be subjected to a popular vote. Thus, only those laws which generate active opposition in society are eventually tested in a referendum. Constitutional amendments, on the other hand, must always face a referendum, even if there is little public opposition to the measure.
Although these figures indicate a high propensity for the rejection of challenged laws and decrees, it must be remembered that many acts of the Federal Assembly are never brought to a referendum. Since 1874, approximately 7 percent of laws, decrees, and treaties susceptible to challenge have had to face a referendum.17 A century ago, this figure was slightly higher. Between 1874 and 1908, the Federal Assembly passed 261 bills which were subject to possible referendums. In all, 30 (or 11.5 percent) were challenged.18 In the postwar era, the percentage has been considerably lower, with 6-7 percent of legislation challenged in most years. This is mainly due to the advent of consociational democracy and to the increased volume of legislative activity in recent decades. Much of this extra legislation is technical or uncontroversial; it therefore faces little or no opposition. There are also limits to the human and financial resources readily available for challenging laws via the referendum.
The proportion of parliamentary measures rejected by referendum has changed with time. Table 3-6 illustrates what appears to be an increasing willingness among the electorate to accept legislative proposals.19 The high rate of rejections in the nineteenth century reflects the very different nature of Swiss direct democracy before the evolution of consociational institutions. The extremely high percentage of defeated measures during the first two decades of referendum usage was in large part due to the ardent and concerted efforts of Catholic and conservative opposition groups. With the exception of the 1900-19 period, which in any event saw relatively few referendums take place mainly because of the exigencies of the First World War, there has been a gradual decrease in the percentage of measures rejected in the referendum vote. It is no coincidence that the most dramatic drop in the rejection rate came in the period after 1959 â€" after the advent of the magic formula and the completion of Konkordanz-Demokratie. The culmination of the process of making Swiss government more inclusive reduced the amount of opposition in society to most parliamentary measures. Legislation usually embodied compromises struck between the four major parties and the largest interest groups. Since 1959, the rate of rejection has consistently remained at 25-30 percent. Although party allegiance is a relatively weak predictor of referendum voting in Switzerland, defeating a government measure endorsed by all four of the major parties is usually a more difficult task than defeating one subscribed to by only two or three. The agreement of all four parties can have a fairly persuasive impact on voter decision making.
There is also clear evidence of a persistent Neinsager vote. Even the most unobjectionable measures run up against opposition at the polls. Only two referendums in Swiss history received more than 90 percent of the vote. The highest was a 1977 referendum on a statute concerning the free exercise of political rights, which won a 93.7 percent majority. The second highest was a constitutional amendment for the protection of the environment in 1971, which met with 92.7 percent approval. Every other vote has run up against opposition of at least 12 percent. The 95-100 percent approval figures seen elsewhere in the world have never set foot inside Swiss borders. Most of these concerned regime endorsement in cases where the outcome was a foregone conclusion. Only one referendum of this nature has been held in Switzerland â€" the federal vote in 1978 approving the decision of the Bernese and Jurese to create a separate canton of Jura. The final arrangement was the product of numerous regional referendums in a tediously democratic process. It also promised to end a decade of sporadic political violence in the area. Most importantly, it was clearly the solution preferred by those involved. As such, federal disapproval would have been virtually unthinkable. Nonetheless, 17.7 percent voted against the measure.
It is difficult to divine precisely the source from which the Neinsager sentiment springs. This dispensation to reject government proposals or popular initiatives for seemingly arbitrary reasons is a complex phenomenon, undoubtedly reflecting a variety of motivations. As noted previously, the referendum offers a channel for venting deep dissatisfaction with the state. However, this explanation seems insufficient in the Swiss case, given that the Helvetic political system registers such high levels of citizen approval, as demonstrated in the Sidjansky study.20
A more salient explanation in the Swiss case is that the referendum facilitates the blanket rejection of political leaders and their behavior. This manifestation of distaste for the office-holders of the day has long been an important facet of Swiss referendums. One prominent example is the referendum of 1884, popularly dubbed the four-humped camel'. In an atmosphere of widespread dissatisfaction with activities in Bern, referendums were demanded by nearly 100,000 citizens on four federal bills at once, none of which were particularly unreasonable or controversial. The issues ranged from an additional yearly credit of $2,000 for the Swiss embassy in Washington to a revision of the criminal code. All four measures were soundly defeated. As Rappard recorded in 1912, the vote went the way it did primarily because so many people felt that 'those people in Bern need a lesson'.21 The six negative outcomes of 1 April 1990 in the wake of the Fischenskandal reflected similar public discontent. It is quite likely that the concoction of the 'magic formula' in 1959 served to augment the use of the referendum as a Neinsager device. Because of the difficulty of associating government behavior with any particular party in the Swiss consociational arrangement, it is not easy to express disapproval through elections. If the government consists of all four major parties, and a vote for one of the smaller parties seems unpalatable, then the only way to cast a meaningful vote against the government may be to renounce its policies arbitrarily in a referendum.
A third explanation is simply that there exists a gut-level Swiss skepticism toward change. For example, in June 1982 a revision of the penal code concerning violent crimes passed in a referendum, with 63.7 percent voting in favor. However, a survey conducted immediately thereafter by the Research Center for Swiss Politics at the University of Bern revealed that fully 31 percent of those voting against the measure could not specify a primary reason for their vote.22 It seems that for many Swiss citizens the best policy is 'When in doubt, stick with the status quo'.
Although there is inevitably a core group of no-voters on every issue, there also seems to be a consistent basis of regime support. Even the most widely-opposed government measures win a significant percentage of the vote. Only three have emerged with less than 20 percent support â€" a 1952 law concerning the construction of anti-aircraft shelters (15.5 percent), the 1923 free trade convention (18.5 percent), and an 1896 law regarding disciplinary punishments in the army (19.9 percent). Since the creation of full proportionality on the Federal Council in 1959, the core of regime support has been even broader. Only two measures have received less than 30 percent of the vote â€" the 1986 decision on UN membership (24.3 percent), and the 1987 proposal for sickness and motherhood insurance (28.7 percent). To match the Neinsager vote, there also seems to be a sizable group of voters who believe that government knows best.
The referendal defeat of government legislation or a constitutional proposal is, logically speaking, a conservative action â€" 'conservative' in the sense of preserving the status quo. This implies a politically conservative effect as well. The facultative and obligatory referendums tend to benefit parties on the right of the Swiss political spectrum, which oppose radical social and economic changes, as well as the taxes to finance them. However this is not always the case. It sometimes happens that measures favored by the Right are tripped up by referendal challenges. In 1985, for example, a government proposal supported by the Radical Democratic Party (FDP), the Swiss People's Party (SVP), and the Christian Democrats (CVP) would have ended federal education subsidies, transferring the responsibility back to the cantonal level. The Social Democratic Party (SPS) was firmly against the move. In the referendum, the measure was narrowly defeated, with 52.4 percent voting no. The FDP, SVP, and CVP were unable to keep their ranks marshalled behind their position. A University of Bern post-election survey estimated that 33 percent, 32 percent, and 44 percent of voters belonging to the respective parties came down against the measure.23 This was a conservative vote, in the sense that change was rejected. Yet, it was not conservative in the political sense, since it favored redistribution and central oversight. The electorate will reject any change with which it feels uncomfortable, whether it comes from the Right or the Left.
Table 3-6 registers a noticeable jump in the 1980s in the percentage of government measures rejected by voters. This is almost certainly related to a trend of declining confidence in officeholders over recent years. The background of fundamental trust in the government's handling of issues, against which the voter decides his own position, is increasingly disappearing. Survey data from the University of Bern reveals a definite drop from 1979 to 1986 in the percentage of voters expressing 'general trust' in the federal government. In 1979, 58 percent of voters expressed this confidence. By 1982, the figure was down to 44 percent; and in 1986, only 38 percent trusted the government. The number of voters expressing mistrust did not increase proportionally; it remained fairly constant, between 33 and 38 percent. Rather, the corresponding increase was seen in the percentage of voters who were unsure of whether or not they trusted the government. In 1979, 9 percent felt this way; in 1982, 22 percent were unsure; and in 1986, 24 percent were unsure. Between 1985 and 1986, the number of voters expressing trust in their government fell 6 percent.24 The executive's behavior in the campaign leading up to the March 1986 UN referendum (discussed in detail in Chapter Seven) undoubtedly accelerated this erosion of trust. Indeed, survey data at the time indicated specific dissatisfaction with the Federal Council.25 Public confidence received another blow with the Fichenskandal of 1989-90. Thus, while the confidence of Swiss citizens in their political system remains high, trust in the officeholders of the day has been waning. Should this trend continue, the government's success rate in referendums will almost certainly drop further in the future.
The same survey data revealed a growing tendency among voters to consider long-term factors when making referendum decisions.26 Combined with the healthy skepticism regarding the capabilities of officeholders described above, this trend suggests that Swiss voters are becoming more competent and confident in the task of self-government. As such, it is a development which offers encouragement to other political systems that make frequent use of direct democracy.
In conclusion, neither the referendum nor the initiative has been uniformly beneficial to one side of the political spectrum in Switzerland. The initiative has been employed effectively by interest groups and parties of both the Right and the Left. However, the referendum has tended to help the Right more often, with a few exceptions. It is an inherently conservative device because it offers electors a final opportunity to reject the decisions and proposals of their legislators. In Swiss experience, it has tended to preserve the status quo, presenting a difficult hurdle in the path of proposed changes and reforms. It has also reflected the basic conservatism of the Swiss electorate.
In his 1897 essays on popular government, Sir Henry Maine warned against allowing the citizenry a more direct role in lawmaking. He maintained that, far from unleashing radical political forces, it 'would produce a mischievous form of Conservatism, and drug society with a potion compared with which Eldonine would be a salutary draught'. This prediction stemmed from his belief that policy would inevitably reach 'a dead level of commonplace opinion'.27 Maine argued that progressive reforms would rarely survive the process of making legislation acceptable to a majority of voters. All the rough edges would be smoothed out of laws to create a smooth and uninspiring endorsement of the status quo. Stagnation was the primary threat posed by referendums and, indeed, by democracy in general.28 Although subsequent history has shown that Maine's predictions were overstated, his basic argument that requiring broad popular approval for legislation tends to impede change is a valid one. Swiss experience with the obligatory and facultative referendum devices affirms this notion.
Because the referendum in Switzerland tends to reflect the conservatism and cautiousness of the Swiss electorate, one cannot generalize and assume that the referendum has an equally conservative impact in all other polities. Other than creating a slight bias in favor of the status quo, all that can be said for certain is that the device ultimately reflects the political leanings of the voting population.
Notes 1. Max Imboden, Helvetisches Malaise (Zurich: 1964), p. 8; cited in Benjamin Barber, 'Participation and Swiss Democracy', Government and Opposition, 23, 1 (Winter 1988), p. 40.
2. John Austen, David Butler and Austin Ranney, 'Referendums, 1978-1986', Electoral Studies, 6, 2 (Aug. 1987), 14345, p. 139.
3. Analyse der eidgendssischen Abstimmung vom 16. Miirz 1986, VOX Survey No. 29 (Bern: Forschungzentrum fur Schweizerische Politik and the Schweizerische Gesellschaft fur Praktische Sozialforschung, 1986), p. 6.
4. Ibid., pp. 6-7.
5. Ibid., p. 8.
6. James Jones, in Austin Ranney, ed., The Referendum Device (Washington: American Enterprise Institute, 1981), p. 15.
7. Henry H. Kerr, 'Swiss Electoral Polities', in Howard R. Penniman, Switzerland at the Polls: The National Elections of 1979 (Washington: American Enterprise Institute, 1983), p. 74.
8. Philip E. Converse, 'Of Time and Partisan Stability', Comparative Political Studies, 2 (1969), 139-71.
9. Analyse der eidgendssischen Abstimmung vom 12/13. März 77, VOX Survey No. 1 (Bern: Forschungzentrum fur Schweizerische Politik and the Schweizerische Gesellschaft fur Praktische Sozialforschung, 1977), p. 4.
10. Hans-Peter Hertig, Analyse der Nationalratswahlen 1979 (Bern: Forschungzentrum fur Schweizerische Politik and the Schweizerische Gesellschaft fur Praktische Sozialforschung, 1980), p. 6.
11. Analyse der eidgendssischen Abstimmung vom 10. März 1985, VOX Survey No. 25 (Bern: Forschungzentrum fur Schweizerische Politik and the Schweizerische Gesellschaft fur Praktische Sozialforschung, 1985), p. 5.
12. Claude Longchamp, Analyse der Nationalratswahlen 1987 (Bern: Forschungzentrum fur Schweizerische Politik and the Schweizerische Gesellschaft fiir Praktische Sozialforschung, 1988), p. 6.
13. Florence Passy, Pascal Sciarini, Simon Hug, and Hanspeter Kriesi, Analyse der eidgenossischen Abstimmung vom 1. April 1990 , VOX Survey No. 39 (Bern: Forschungzentrum fur Schweizerische Politik and the Schweizerische Gesellschaft für Praktische Sozialforschung, 1990), p. 9.
14. Analyse der eidgenossischen Abstimmung vom 16. März 1986, p. 7.
15. 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.
16. Ivor Crewe, 'Electoral Participation', in David Butler, Howard R. Penniman, and Austin Ranney, eds., Democracy at the Polls: A Comparative Study of Competitive National Elections (Washington: American Enterprise Institute, 1981), pp. 232-37. See also G. Bingham Powell, Jr., "Voting Turnout in Thirty Democracies: Partisan, Legal, and Socio-Economic Influences', in Richard Rose, ed., Electoral Participation: A Comparative Analysis (London: Sage, 1980), p. 6.
17. Jean-Franjois Aubert, 'Switzerland', in David Butler and Austin Ranney, eds., Referendums: A Comparative Study of Practice and Theory (Washington, D.C.: American Enterprise Institute, 1978), p. 43.
18. William Rappard, The Initiative and the Referendum in Switzerland', The American Political Science Review, VI, 3 (Aug. 1912), 357.
19. This trend has been missed by many authors. Others have presented inaccurate statistics on the question. For example, Katzenstein (Corporatism and Change, 1984, p. 122) argues that the rejection rate has climbed from 38% in the 1960s to 51% in the 1980s. This is incorrect.
20. Dusan Sidjansky, et al., Les Suisses et la politique (Bern: Lang, 1975), p. 36.
21. Rappard, p. 362.
22. Analyse der eidgenossischen Abstimmung vom 6. Juni 1982,VOX Survey No. 17 (Bern: Forschungszentrum fiir Schweizerische Politik and the Schweizerische Gesellschaft fiir praktische Sozialforschung, 1982), p. 6.
23. Analyse der eidgenossischen Abstimmung vom 10. März 1985,VOX Survey No. 25 (Bern: Forschungszentrum fur Schweizerische Politik and the Schweizerische Gesellschaft fiir praktische Sozialforschung, 1985), p. 11.
24. Analyse der eidgenossischen Abstimmung vom 16. März 1986, VOX Survey No. 29 (Bern: Forschungzentrum fiir Schweizerische Politik and the Schweizerische Gesellschaft fiir Praktische Sozialforschung, 1986), p. 23.
25. Ibid., p. 24.
26. Ibid.
27. Sir Henry Sumner Maine, Popular Government: Four Essays (London: John Murray, 1897), pp. 35,41.
28. Ibid., pp. 35-36.