Resumé

The 1998 Swedish general election was a protest election primarily against the Social Democratic The party was hit by a debate on deceit when the unpopular financial restructuring policy was implemented. Disappointed social democratic voters from the 1994 election flocked to the Left Party. Others did not vote at all. The electoral turnout dropped to 81.4 percent; the lowest level in a parliamentary election since the election of 1958. A turnout of 81.4 percent is not particularly low from an international perspective, but it has received a great deal of attention in the political debate. Certain signs do indicate that there has been a general devaluation voting as a means of exercising political influence among large groups of voters. Other forms of influence are perceived as being more meaningful. A sense of meaninglessness, of individual and institutional powerlessness, also seems to be spreading to social groups that have traditionally had a fundamental trust in the political system.

An Historic but Undramatic Election

In the Swedish general election of 1998, the Social Democratic Party had its poorest election results since the introduction of democracy in Sweden in 1921. The party received 36.4 percent of the votes, a decline of 8.9 percent compared to the 1994 election.¹ The stunning election defeat did not result in any shift in government, but the social democratic minority government saw its parliamentary position dramatically weakened.

The outcome of the election was historic in other respects as well. The electoral turnout was the lowest since the election of 1958. In addition to the Social Democrats' catastrophic election, two other parties also recorded their worst election results ever. The Center Party and the Liberal Party continued the electoral decline they have been experiencing for a considerabletime. two centrist parties, which mobilized more than 35 percent of the voters as recently as the election of 1976, together attracted fewer than 10 percent in 1998. The Center Party garnered 5.1 percent and the Liberal Party 4.7 percent of the votes. Thus, both parties find themselves

---

* Tommy Möller, Department of Political Science, Stockholm University, SE-106 91 Stockholm, E-mail: tommy.moller(a>statsvet.su.se

very near the 4 percent threshold and risk elimination from the parliament
if the downward trend cannot be arrested.

Meanwhile, two parties achieved their best results ever: the Left Party (12 percent) and the Christian Democratic Party (11.8 percent). While the success of the Left Party was expected - the party had recorded high numbers in opinion polls throughout the entire mandate period and had also gained strong support during the EU parliament election of 1995 - the rise of the Christian Democrats, which took place mainly during the final stages of the election campaign, was sensational.

Despite a minor advance, the election was a disappointment for the Moderate Party. Expectations within the party were high. During the mandate opinion polls had shown considerably stronger support than evidenced in the election results. The year prior to the election, polls showed popular support of the party between 30 and 35 percent; it was for a brief period the most popular party in the country.

The non-socialist opposition strengthened its position somewhat - the four parties together won eleven new - seats but since the Green Party declared its intention to support a social democratic government, the Red- Green leftist block majority was never really threatened. However, had the Green Party fallen below the four percent threshold, the majority would have been jeopardized. The leftist block majority during the current mandate is secure (190 seats versus 159).

The Election Campaign

The election campaign was perceived as undramatic by many. The struggle
for governing power seemed decided before the campaign even got started,

and the issues that dominated the election debate were of traditional ilk - unemployment, the economy and social welfare. The election may be describedas traditional left-right election (see Oscarsson 1998). Competing conflict dimensions were given limited scope in the election debate. Issues aside from those that fall under the left-right dimension trailed the list of election issues considered important by voters (see Table 2).

During the mandate period of 1994-98, a comprehensive recovery of state finances took place, which came to influence the conditions of the election campaign in several ways. Tax increases and cutbacks in the public sector were both instituted. The aggregate budget reinforcement gained through the financial restructuring policy amounted during the period to SEK 125.5 billion. The justification for the unpopular policy (which was the main reason for the Social Democratic Party's election losses according to the study of their own analysis group)³ was that its ultimate purpose was to reestablish the primacy of politics and safeguard democracy. 'The borrower is servant to the lender,' preached Prime Minister Göran Persson (Persson 1997).

This policy led to more favorable opinion of the parties on the left and right flanks, the Left Party and the Moderate Party. The cutbacks gave the Left Party an opportunity to win voters from the governing party's left wing, which was reflected not only in the opinion polls during the mandate period, but also in the election itself. For the Moderate Party, the tax hikes and the decision to begin dismantling Swedish nuclear power plants provided the party's first opportunity to make inroads into social democratic groups. The government's policy thus engendered discontent from two directions at the same time. Voters on the left wing of the Social Democratic Party were frustrated over the cutbacks, while voters on the party's right wing were irritated over a perceived inability to achieve political In contrast to the Left Party, however, the Moderate Party was unable to fully attract discontented social democrats. As the election approached, it was instead the Christian Democrats who benefited from voter discontent. Nine percent of the Christian Democratic Party's voters came from the Social Democratic Party.

The economy recovered during 1997. Although unemployment remained high, an increasing number of curves were now pointing in the right direction. Interest rates and inflation dropped, growth accelerated, and in October 1997, the National Debt Office forecast showed that the national budget was expected to yield a surplus in 1998. A change of course was announced at the Social Democratic Party congress in September 1997. The state financial crisis was declared averted and a return to traditional social democratic welfare policy would take place. Investments in the core sectors of the welfare state - health care, education and social welfare would be initiated. These issues were later to dominate the election campaign. 'Health eare-educaiionsocialwelfare'

socialwelfare'became something of a mantra for all parties during the
campaign.

Table 2 shows that the agenda for the campaign - set by the governing party a year before the election at the Social Democratic Party congress in September 1997 - turned out well. Alongside the economy, voters were most interested in traditional welfare policies. There were, however, differences on voter party affiliation. Moderate Party voters ascribed greater significance to the tax issue for their party vote than did other voters, and correspondingly, Green Party voters considered environmental issues to be the most critical.

The issue of government was, as usual, the object of great attention, but in the 1998 election in a different way than usual. Few pundits saw a shift of government as likely. The debate thus primarily came to be about which parties might possibly support the social democrats after the election. That both the Left Party and the Green Party would support a social democratic ministry rather than a non-socialist government dominated by the Moderate was naturally considered self-evident. There were other possible alternatives for the government. Between 1995 and 1998, there had been an institutionalized collaboration between the social democrats and the Center Party. The collaboration was so extensive that it was, in practice, an informal government.

However, the Center Party's future preferences were somewhat unclear
prior to the election. Did the party, which had changed leadership just

before the election, want to be part of a non-socialist government, or did they prefer continued collaboration with the social democrats? The social democrats' preferences in the government issue were also unclear. Did the party intend to continue collaborating across block lines and, if so, with which party? Or did they prefer collaboration with the Left Party and, if so, in what form? The question was of fundamental interest, since the governmentwas in 1994 with parliamentary support from the Left Party, but governed during the mandate period with the support of the Center Party. The switch of collaborative partners was justified by saying that the Left Party was 'unreliable' and 'not ready to assume responsibility.'

The lack of clarity gave rise to a debate among political scientists during the election campaign. Leif Lewin asserted in a guest editorial in Dagens Nyheter (Lewin 1998a) that the absence of clear indications with respect to the government by the Center Party and the social democrats represented an 'elitist and authoritarian view of democracy,' with roots in the predemocratic of King Oscar II in the 1800s. Voters were kept in the dark, thus setting aside their opportunities to demand political accountability. Lewin's opinion, the culture of cooperation that characterizes Swedish politics is behind the growing distrust of politicians and politics. There is, he said, 'unambiguous' empirical evidence that distrust is less prevalent majoritarian systems than in consensual systems. Lewin therefore recommended that all parties clearly declare with whom they intended to collaborate during the subsequent mandate period (see also Lewin 1998b).

Olof Ruin argued against this interpretation. In his opinion, it was urgent that the parties make their positions on various issues clear rather than talk about the government issue. Stating preferences about the formation of the government could lead to deadlocks that would make it more difficult to reach the type of agreements necessary in minority parliamentary systems. As for the connection between cross-party collaboration and distrust of politicians, he questioned Lewin's interpretation. There is no authoritative evidence that there is greater distrust in consensual democracies according to Ruin (Ruin 1998).

As far as I am able to judge, Ruin's interpretation of the research status is correct. Theories on the cause of growing distrust of politicians are many and conflicting. Unfortunately, none of the theories carry any general explanatoryforce. of politicians has many faces, of completely disparatenatures, thus has various explanations (Möller 1998). Lewin's argument appears reasonable to the extent that the act of voting may seem less meaningful if the voter does not know what preferences the parties have with respect to the government issue, but there is no strong empirical evidenceto of.⁵ Certainly, there is a belief that those in control have entered into a quiet conspiracy and that underneath it all they agree about most things, i.e., that there is excessive collaboration, but it is equally commonthat

monthatcitizens distrust politicians because they are thought to collaboratetoo
(Möller 1999).

The Voter Stream

There is still a great deal of flux between parties in Swedish politics, but it seems to have been less extensive in 1998 than in the two previous elections in 1991 and 1994, when party mobility was around 30 percent. However, slightly more than every fourth voter (27 percent) switched parties between 1994 and 1998. Considering that this time as - opposed to the three preceding elections - no new parties won seats in the parliament, 27 percent may be regarded as a high figure. Most party switches in this round occurred within the two blocks - the non-socialist versus the socialist (Oscarsson 1998).⁶

The Christian Democratic Party and the Moderate Party were best at mobilizing their former voters. Both parties were able to retain 80 percent of those who voted for them in 1994. The Center Party, Green Party and Liberal Party did not perform as well, succeeding in mobilizing only slightly more than half of their voters from 1994 (57, 55, and 53 percent).

Where did the parties that were exceptionally successful in 1998 - the
Christian Democratic Party and the Left Party - get their new voters?

Not surprisingly, the Left Party picked up most of its new voters from the social democrats. Just under one third (30 percent) of the party's voters came from the governing party. The greatest voter stream occurred between these parties. A more detailed analysis also shows that it was precisely those voter groups that the Left Party targeted in an attempt to win new voters that in fact switched parties: members of the Swedish Trade Union Confederation women, the unemployed and public sector employees. Among LO members, the social democrats recorded a dramatic decline: from 66 percent in 1994 to 52 percent in 1998. Every fifth LO member voted for the Left Party in 1998, which is by far the highest level in history. In 1991, only four percent of LO members voted for the Left Party (Gilljam & Holmberg 1995:103). Among the unemployed, the scales were about even between the two labor parties: 33 percent voted for the social democrats and 28 percent for the Left Party. This may be seen as one of several indicators that the 1998 Swedish election was a protest election. The governing Social Democratic Party was abandoned by its traditional social democratic voter groups, who either voted for the Left Party or did not vote at all.

Thus, the Left Party managed to break into groups that had traditionally been regarded as social democratic core groups. Those who moved from the Social Democratic Party to the Left Party were characterized by their view of themselves as standing 'clearly on the left,' by their low confidence in politicians, and by their opposition to Swedish membership in the EMU.

Another wide voter stream went from the Moderate Party to the Christian Democratic Party. Just over every fourth Christian democratic voter in 1998 (27 percent) voted for the Moderate Party in 1994. The party also attracted a considerable number of voters from the Liberal Party (10 percent), the Center Party (9 percent) and the Social Democratic Party (8 percent).

Also many among those who switched from the Moderate Party to the Christian Democrats had little trust in politicians. This indicates that the 1998 election cannot solely be characterized as a protest against the sitting government, but also against both of the major parties that have long dominated the political arena.

Oscarsson interprets the shift from 'centrist parties to flank parties' as an 'ideological protest against the centrist collaboration that had taken place' (Oscarsson 1998), i.e., the collaboration across block lines between the Social Democratic Party and the Center Party. As a description of the voter stream from the governing party to the Left Party, this is certainly correct. One can also state that the parties perceived by the electorate as being furthest to the right - the Moderate Party and the Christian Democrats both strengthened their positions. However, there was simultaneously significant voter switch between these parties, which means that it is first and foremost still relevant to speak of a protest against the two largest parties. In both cases, it seems there were ideological reasons for the protest. However, while the protest vote within the leftist block may be described as a reaction against a government policy oriented towards the center, the protest vote within the non-socialist block may more correctly be seen as a reaction to the politics of the Moderate Party, which had, in the voters' eyes, become far too oriented towards the right.

Social Differences

Class voting is still a highly significant factor in Swedish politics. All five traditional parties were formed during the shift from an agricultural to an industrial society, and still retain their class-oriented profiles to a great extent. Two thirds of the collective working class and almost three quarters of LO members vote for either the Social Democratic Party or the Left Party; the Center Party always mobilizes nearly half of all farmers; the Liberal Party still has a strong position among white collar workers and academics (12 percent of members of the Swedish Confederation of Professional and nearly half of all business owners (44 percent) vote for the Moderate Party.

During the 19905, the gender gap has become progressively wider in
Swedish politics with respect to voting. In the 1998 election, two parties in
particular distinguished themselves through gender imbalance, the Left

Party with a clear preponderance of female voters (4- four percent) and the
Moderate Party, which has for many years primarily attracted male voters
(+seven percent).

The age variable was significant in the 1998 election. As a rule, parties that advance during elections gain disproportionately strong support from younger who are much more likely to switch parties and are more sensitive to trends than older voters (Oscarsson & Oscarsson 1994; Gilljam & Holmberg 1995, 92). In this respect, the 1998 election was an exception. Certainly, the Left Party had clearly stronger support among younger voters - particularly first-time voters, 17 percent of whom voted for the party - but the other successful party, the Christian Democrats, were strongly underrepresented the younger voter groups. Only six percent of first-time voters cast their ballots for the Christian Democratic Party. However, the party attracted older voters in a remarkable way: 17 percent of voters of retirement age (65) voted for the Christian Democratic Party in 1998. In 1994, only three percent of voters in the oldest age group voted for the party.⁷

The Christian Democratic breakthrough among older voters is historically and may possibly be seen as a trend break. Unlike young people, older voters are regarded as extremely stable and loyal to their parties. Since they rarely switch parties, they have been thought of by the parties as strategically uninteresting groups. This may change in the future.

The Impact of Party Leadership

Swedish voters had to become acquainted with three new party leaders in the 1998 election. Göran Persson had replaced Ingvar Carlsson as chairman of the Social Democratic Party and prime minister in 1996; Lars Leijonborg succeeded Maria Leissner as leader of the Liberal Party that same year; and Lennart Daléus took over leadership of the Center Party from Olof Johansson just three months prior to the election.

Previous research has shown that the significance of party leaders has grown with respect to voting patterns. Party campaign efforts and media attention are being concentrated to an ever increasing extent upon the party leaders, who have been referred to as 'the presidential candidates of parliamentary (Gilljam & Holmberg 1992, 98). This trend entails greater potential for party leader impact with respect to the party vote. While a popular party leader may function as a bobber for his or her party, an unpopular leader may function as a sinker.

Culling the voters who stated that the party leader was highly significant for their party vofe-gives us a strong indication of whether there was a party leader impact. In at least two cases, there is absolutely no doubt. The exceptionalsuccess the Christian Democratic Party may be attributed to a

great extent to the popularity that party leader Alf Svensson enjoyed among voters. 75 percent of Christian democratic voters stated that the party leader factor was an important reason for the party selection. Almost a third of the party's voters gave the party leader factor as a 'very important' reason. Carl Bildt, leader of the Moderate Party, also enjoys strong support. Every fourth Moderate Party voter stated that the party leader was of great significance in their party vote. There is also a strong indication that the Moderate Party's election results would have been poorer still had the party not had such a popular leader.

However, the chairman of the Left Party, Gudrun Schyman, seems to have been a bobber for her party to a much lesser extent than many assumed during the election campaign. Schyman was portrayed by many political analysts during the campaign as an exceptionally skillful campaigner with unusually good communication skills. A relatively large percentage of the Left Party's voters also said that the party leader was an important reason for their votes. Nevertheless, Schyman's figures here are not quite as high as some analysts probably would have guessed. One interpretation is methodological: may feel foreign to many leftist voters to refer to the party leader when asked to give the reasons for their votes, no matter how well liked that person may be. The Left Party has a critical view on the growing personification within politics, which is reflected in a skeptical view on the new individual candidate election system.

Constitutional Reforms that Yielded Results?

There were two constitutional changes prior to the 1998 Swedish election.
For the first time, Swedish voters could cast their votes for individual candidates,and

dates,andit was the first time since the introduction of the unicameral parliament in 1970 that those in government encountered the electorate after a four year mandate period. The individual candidate vote was introducedin effort to vitalize Swedish democracy; the reform gave the candidates stronger incentive to seek a personal mandate from the voters. The mandate period was extended by one year in order to strengthen the power of government. The intent was that such an extension would provide greater opportunities for the government to act for the long term. This was to bring greater permanence and consistency, which would not only make it easier to resist making populist statements of opinion and conduct a responsible fiscal policy, but also - by giving the parliamentary majority greater scope to achieve results during the mandate period - make it easier for voters to evaluate policies and be able to discern the lines of demarcation between the various alternatives in the election campaigns. The voters were thus not able to call their representatives to account quite as often as before but, on the other hånd, they would - once the reform was instituted - be able to do so under other and hopefully more meaningful forms.

It is naturally difficult this early on to form a definite opinion about whether these intentions were realized and, if so, to what extent. However, certain signs indicate that the constitutional reforms have had, at least partially, the intended effect.

Extension of the Mandate Period

For methodological reasons, it is difficult to determine if and how much the extension of the mandate period may have contributed to the successful recovery of the economy. That conditions for applying an economic-poltical strategy for the long term were improved is beyond doubt. The government was de facto given a one-year respite from confronting the electorate during a period when it took unpopular measures that had direct effect upon the lives of individual citizens. It is certainly noteworthy that the change of course to a more expansive and traditionally social democratic direction took place one year prior to the election, once several of the most important curves were pointing in the right direction. It is hardly likely, but nevertheless theoretically that the government would still have sustained the unpopular financial restructuring policy for as long as it did had the mandate period been three years. However, it may be interesting to examine the analysis of the successful financial restructuring policy as reasoned by the most pivotal individual actor, Prime Minister Persson:

I believe that one of the most important explanations behind the shocks to the Swedish economy lies in the constitutional reforms of the 19705. The impaired opportunities to build the strong governments necessary to manage long-term sustainable growth are a great weakness in the new constitution. The unicameral system broke the stability that resided in the

lingering majority in the old system's first chamber. The three-year mandate period encouragedshortsightedness.
shrank back from making the difficult but necessary
1970, we have found it extremely difficult to make such decisions (Persson 1997,27-28).

For prime minister Persson, there is clearly no doubt that the extension of the mandate period contributed to the economic recovery. However, even as he asserts that the conditions for long-term action had been improved, he constantly repeats in the same book how tough it was to make unpopular but necessary decisions.

Individual Candidate Election

With respect to the second constitutional reform - the change of the election system - there are also signs that intentions have been realized. The substance the new election system is that the voters will continue to primarily cast a party vote, but that it shall also be possible within the framework of the party election to vote for individual candidates. As in Denmark, casting ballots for individual candidates is optional.

The new election system is the result of a compromise, which is important to remember when assessing the outcome of last year's election.⁸ While the Moderate Party and the Liberal Party wanted a stronger element of individual election, the Social Democratic Party and, in particular, the Left Party - which did not support the election of individual candidates at all - wanted to limit the effects of the new system. The compromise was expressed in several ways, among them through a relatively high threshold level instituted for the parliamentary election. For a candidate to be elected via individual votes, he or she must receive at least eight percent of the votes cast for the candidate's party in the constituency. (Provided, of course, that the party wins seats in the constituency).

Of the voters who participated in the parliamentary election, slightly less than 30 percent exercised the option to vote for individual candidates. The percentage was slightly higher in the municipal elections - 34.5 percent. In the county council election, which is the election that voters are generally least interested in, 28.9 percent voted for individual candidates.

Thus, interest in voting for individual candidates is not particularly fervent.In where there has been a similar system in place for a very long time, the proportion of individual candidate votes is about 50 percent in the parliamentary election and 75 percent in the local and regional elections. On the other hand, this was the first election to take place in Sweden under the new rules. Both the parties and the individual candidates in many cases took a tentative position and were not really sure how they should approach the new election system. Many candidates did not actually campaign, in part because either they or their parties were skeptical towards individual election

campaigns and in part because they believed chances of being elected on the strength of individual votes were slim. There was also a widespread notion that the very idea of individual candidate elections was foreign to Swedish political culture. This belief was expressed, not least, in the media.

It is difficult to dismiss the notion. Swedish politics is exceptionally party centered. There are constant and palpable doubts about the new election system within several parties. Several of the candidates who chose to run individual in 1998 felt that they had been subject to opposition.⁹

The most apparent immediate effect of individual candidate voting was that twelve members of parliament were elected on the basis of individual votes, i.e., candidates who would not otherwise have won a seat through their placement on the ballot list. In total, 87 members of parliament were elected on the basis of individual votes, but 75 of them would have been elected even without individual votes. However, as many as 178 candidates in the parliamentary election who exceeded the individual vote threshold of eight percent. Of these, however, six candidates exceeded the threshold in several constituencies - the election system allows one candidate to stand in as many constituencies as he or she would like - which means that these 178 candidates equal only 139 physical persons.¹⁰

The actual effect in the municipal elections was somewhat lower percentage-wise in the parliamentary election, despite a lower threshold. In all, 144 candidates elected via individual votes would not have been elected through their placement on the party list.

Thus, one cannot say that the new election system had any revolutionary impact. However, the intention of the reform is of a long-term nature: to achieve vitalization of democracy through improved dialogue between candidates voters. The process is successive, rather than immediate. Positive signs can already be discerned: More candidates became actively involved in the election campaign than before; the candidates themselves, like the party representatives, felt that contact with the voters had been improved; campaign methods were renewed. At the same time, voter awareness of candidates is still low: Only four in ten voters know the name of any parliamentary in their own constituency. Remarkably, the introduction of individual candidate election seems not to have affected this low level.

It is thus too early to say that Sweden is on the way towards a new political It is safe to say, however, that the extremely party-dominated culture seems to be heading for a change. Most party representatives see this as unavoidable and believe that experiences from the 1998 election are already signaling the launch of a new political culture. Once the new regulations had 'time to settle,' and parties, candidates, the media and, not least, the electorate find it natural to see candidates for political bodies stepping forward and saying 'vote for me,' rather than 'vote for my party,' there will probably be a much more apparent impact.

A Lukewarm Democracy?

The electoral turnout dropped to 81.4 percent in the 1998 election, the lowest level in a parliamentary election since the second chamber election of 1958. The turnout has declined by slightly more than 10 percent compared to the record set in 1976, when 91.8 percent of all enfranchised persons voted.

Electoral participation is usually described as the most fundamental form of civic influence in a democracy. In Swedish politics, a high level of electoral has always been considered a worthy democratic aim, since voting is the most egalitarian form of political participation. High electoral turnout has also been seen as evidence that democracy is, in fact, working. Abstention from voting - or election truancy, to use a more pejorative - is not compatible with civic virtue as defined by Swedish public attitudes. Such behavior may be seen as a system anomaly, an expression of political alienation.

However, a turnout of 81.4 percent is not particularly low from an international For example, fewer than 70 percent of those entitled to vote did so in the Finnish parliamentary election of March 1999, and the turnout is, as a rule, clearly lower in both Norway and Denmark than in Sweden. For Western Europe as a whole, since World War 11, average turnout 83 percent, but voting is mandatory in several countries (Topf 1995, 27-51). Thus, 81.4 percent may be viewed as a rather normal level upon reasonable comparison.

Nevertheless, the low electoral turnout has received a great deal of attention the political debate. The government has appointed several committees investigate the causes of declining turnout. By studying public election statistics, we can already say that there are vast differences between different electoral districts. The turnout has declined most in districts with high unemployment and a large immigrant community. The widening gulfs in political participation thus seem to be an expression of the segregation and perceived alienation brought by the wave of refugee immigration, economic and unemployment of recent years.

Certain signs do indicate, however, that there has been a general devaluation voting as a means of exercising political influence among large groups of voters. Other forms of influence are perceived as being more meaningful (Petersson et al. 1998). A sense of meaninglessness, of individual and institutional powerlessness, seems to be spreading also to social groups that have traditionally had a fundamental trust in the political system (Moller 1998).

These deeper explanations may be joined by a political one: the election was not perceived as exciting, since the government issue was considered decided before the election campaign had even begun (Oscarsson 1998). This explanation is also characterized by a general sense of meaninglessness.

The state of Swedish democracy has been - and continues to be - an object of rapt attention among Swedish political scientists. The Democratic Audit of Sweden (Center for Business and Policy Studies) has been presenting annual reports on the state of democracy in the nation since 1995. The image presented is mixed. On the one hand, there is growing powerlessness and alienation among certain groups with limited resources. On the other hand, confidence in the system and in one self is growing among the citizenry in general (Petersson et al. 1998, 148-49). Researchers at the Institute for Democratic Communication in Sundsvall say that interest in politics is low: half of all citizens are politically indifferent. They speak of a 'lukewarm democracy' (Nord et al 1998). Another way of saying the same thing is that we are entering a 'post-democratic' state in the sense that we take democracy for granted. At the same time, there is strong consensus on the principles that constitute democracy, but the practice of democracy indicates that we are slipping ever further away from these principles (Jacobsson 1997, 316).

Perhaps the foremost expression of concern about democracy may be found in the long list of government-sponsored studies of the issue. Two committees on democracy were appointed after the 1994 election: Demokratiutvecklingskommittén Demokratiutredningen. In addition, a great many committees have been put to the task of illuminating various issues related to democracy. Demokratiutredningen has also been enjoined not only to study the state of democracy, but to stimulate debate about democracy. special 'Minister of Democracy' was also appointed following the 1998 election.

A fundamentally interesting reflection is that, on the threshold of a new century, it is the state that is worried about declining political engagement among the citizenry. The debate is about how one can stimulate increased engagement from above in order to vitalize democracy. The situation was quite different at the turn of the last century.

Towards a New Political Landscape?

In summary, the 1998 Swedish general election may be characterized as a protest election against both dominating parties, the Social Democratic Party and the Moderate Party. Voters who abandoned one of the big two for either the Left Party or the Christian Democratic Party exhibited greater distrust of politicians than other voters.

Naturally, social democracy was the primary target of discontent. The party was hit by a debate on deceit when the unpopular financial restructuringpolicy implemented. Prior to the 1994 election, many of the party's voters had been given the impression that the social democrats intended to pull the country out of the economic crisis by means of traditionalsocial

tionalsocialdemocratic policies. When this did not occur, disappointed
voters flocked to the Left Party. Others did not vote at all.

The election was a minor success in terms of percentages and mandates for the Moderate Party. But the party lost 40,000 votes from 1994 and expectations a strong upturn were great. The voter steam from the Moderate Party to the Christian Democratic Party that occurred during the election campaign must be interpreted as an expression of discontent with the policies of the Moderate Party. Precisely as with the voter stream within the leftist block, the shift of the center of gravity between the non-socialist parties may be a sign that a new political landscape is emerging. The common for both successful parties is that each represents a clear ideological That may be why it was not solely a general discontent of a more volatile sort that lay behind the widespread voter migrations.

In a wider sense, the 1998 election may also be seen as a protest election against the traditional center in Swedish politics. Both centrist parties, the Center Party and the Liberal Party, suffered devastating defeats and risk being eliminated from the parliament. Many voters who voted in the social democratic government in 1994 and then in 1998 chose to either to abstain from voting or vote for the Left Party also perceived the social democrats as a centrist party. In any case, the collaboration across block lines paved the way for the advance of the Left Party.

---

NOTES

1. I would like to thank Henrik Oscarsson, Department of Political Science at Göteborg University, for allowing me to read his analysis of the Swedish election, which was yet unpublished when this article was written. The majority of the figures presented in this article are basedon Swedish Television's exit poll. If not otherwise stated, the material was taken from that poll.

2. The Left Party received 12.9 percent of the votes in this election, which was the best election result ever achieved by the party.

3. The report of the Social Democratic analysis group can be read on the Internet at http./ /www.sap.se/analysgruppen (in Swedish).

4. Swedish Television's exit poll was carried out in cooperation with Statistics Sweden in Örebro and the Department of Political Science at Göteborg University. During the final week before the election and on election day, the study questionnaire was answered by almost 9,000 voters in 20 post offices (where absentee ballots were cast) and 80 polling places.

5. Lewin's conclusions are based, in part, on a student essay (Karlsson 1998) in which a re-analysis was made of two questions of trust from the World Values Survev of 1981 and in part upon an essay by Schmitt & Holmberg (1995) on party identification. There is a very weak connection in the former essay in favor of Lewin's hypothesis. Schmitt & Holmberg's study shows that party identification is affected positively by ideological polarization (which occurs, as a rule, in majoritarian systems), but the connection is weak and the issue of trust is not addressed. However, one cannot, according to Schmitt & Holmberg, exclude that aggregate trust in politicians can be increasing if differences between parties are wide, making it easier for the electorate to identify emotionally with one of the parties.

---

6. The 1998 figure is not fully comparable with the other figures, since it was taken from the exit poll; the other two (30 percent party mobility in 1991 and 29 percent in 1994) were taken from election studies carried out at Goteborg University.

7. This information, taken from Gilljam & Holmberg (1995, 94), is not immediately comparable with the exit poll data from 1998. The age group division differs somewhat. The oldest group in the 1994 election study is 71-80 years. However, even if the next oldest group (61-70) is included, the difference is stunning: five percent of this group voted for the Christian Democratic Party in 1994.

8. A comprehensive analysis of the new election system is taking place within the framework of Rådet for utvärdering av 1998 års val. (See Holmberg & Möller 1999). Reasoning and judgements in this section are basedon that project.

9. This emerged during a public seminar about the individual candidate election arranged by Rådet for utvdrdering av 1998 års val in the Swedish Parliament on February 3, 1999.

10. Alf Svensson, chairman of the Christian Democratic Party, was a candidate in all 29 parliamentary constituencies and exceeded the threshold by a good margin in all of them. My thanks to Mikael Gilljam, Department of Political Science at Göteborg University the inspiration for this section.

REFERENCES

Gilljam, M. & Holmberg, S. 1992. Vdljarna infor 90-talet. Stockholm: Norstedts Juridik.

Jacobsson.K. 1997.5å gott som demokrati. Omdemokratifrågani EU-debatten. Umeå: Borea.

Karlsson, F. 1998. 'Majoritetsdemokrati, samarbetsdemokrati och politiskt fortroende. En
jåmforande studie av hur valsystem och politisk modell påverkar fortroendet for de politiska
Uppsala Universitet: Statsvetenskapliga institutionen.

Lewin, L. 1998a. 'Samarbetet leder till politikerforakt', Dagens Nyheter, July 14.

Moller, T. 1998. Politikerforakt eller mogen misstro. En oversikt. Stockholm: Kommentus
forlag.

Moller, T. 1999. Forthcoming.

Nord, L. 1998. Min i opinionen 1998. Sundsvall: Demokratiinstitutet, december.

Oscarsson, M. & Oscarsson, H. 1994. 'De unga våljarna 1968-91. Dunungar eller stormsvalor?'
In Ungdomars vålfdrdoch vdrderingar. SOU 1994:73.

Persson, G. 1997. Den som år satt iskulddr ickefri. Stockholm: Atlas forlag.

Petersson, 0., Hermansson, J., Micheletti, M., Teorell, J. & Westholm, A. 1998. Demokrati
och medborgarskap. Demokratirådets rapport 1998. Stockholm: SNS Forlag.

Ruin., O. 1998. 'Låt partierna gora upp efter valet', Dagens Nyheter, July 27.

Schmitt, H. & Holmberg, S. 1995. 'Political Parties in Decline', i Klingemann, H.-D. &
Fuchs, D., eds., Citizens and the State. New York: Oxford University Press.

Topf, R. 1995. 'Electoral Participation', in Klingemann, H.-D. & Fuchs, D., eds., Citizens
and the State. New York: Oxford University Press.

Internet: http://www.sap.se/analysgruppen