Resumé

The increasing instability in the electorate and the party system forces us to reconsider the Lipset-Rokkan thesis about the frozen party system. In this approach, cleavages played a major role, but in Western Europe, this role has diminished significantly. The frequent shifts in voter allegiance call for a new understanding of the mechanisms that connect voters and parties. A reinterpretation of the concepts of a "frozen party system" or "cleavages" is not enough. We need a new theory about the strategic interaction between parties and a floating electorate.

Introduction

In Europe, national as well as local elections are open contest, multi-party elections, with a high level of participation. One key question when we interpret election results is, how stable are the ties between parties and voters? Voters may shift from one party to another if they are not pleased with the party they voted for in the last election, or if they are more attracted to another party. Voters may also abstain from voting if they do not find the alternatives attractive. Parties want two things that contradict each other: On the one hand, they want stable support from loyal voters, but on the other hand they also wish to attract new voters. However, election results are based on a zero-sum game, where the gains of one party are offset by the losses of another party, since the share of mandates given to a party is more or less related to the relative strength of the party.

Stability versus expansion is a theme that entails the possibility of considerable losses and retrogression, but also the sweetness of electoral victory. Stability implies that parties receive the same relative amount of electoral support in one election after the other. How is that accomplished? We may distinguish between surface stability, measured by net volatility, and deep structural stability, measured by gross volatility. Parties fear high

---

* Jan-Erik Lane, Geneva University, Department of Political Science, 102 Bd Carl-Vogt, 1211 Geneva 4, Switzerland. Svante Ersson, Umeå University, Department of Political Science, S-907 87 Umeå, Sweden.

levels of both gross and net volatility, but how can they protect themselves
against voter instability?

Scholars have pointed out numerous changes in European electorates and party systems during the early 1990s and speak of "new politics," "changing values," and "volatility." The analysis of the relationships between the electorate (micro) and the party system (macro) necessitates a reassessment of the Lipset-Rokkan hypothesis about "the freezing of the major party alternatives" and "the mobilization of major sections of the new reservoirs of potential supporters." (Lipset & Rokkan 1967, 50). They stated in 1967 that "the party systems of the 1960s reflect, with few but significant exceptions, the cleavage structures of the 19205" (emphasis in original) (ibid.).

It has been argued that this hypothesis is still valid today (Bartolini & Mair 1990; Mair 1993), but is this really true? Two entities are involved here, namely the political parties and the cleavages in the electorate. How stable are these two entities today? Let us begin by examining the basic Rokkan model of distinct and stable ties between political parties and the voters in terms of alignments. What is a frozen party system, and which alignments constitute cleavages?

The Rokkan Mechanism

Participation in elections establishes the relationship between the voter and
the political party. The vote may express a deep commitment or partisanship,
but it may also be a casual or a nonrecurrent deliberate choice.

In all elections, the relations between voters and parties can change. First, some people switch from one party to another. Second, some people abstain from voting, although they voted in the last election. Third, new voters participate, either because they decide to participate or because they become eligible. And fourth, some people have left the electorate because of death.

We can assess the possible changes in the voter-party relationship at two different levels, the microlevel and the macrolevel. The microlevel represents gross volatility, or the tendency of voters to vote for different parties in different elections. The macrolevel refers to net volatility, or the actual loss or gain of a political party. Extremely high gross volatility does not necessarily mean high net volatility. Figure 1 shows the four possibilities.

All four combinations are possible, but they are not equally probable. Extremely low gross volatility could coincide with extremely high net volatility. It would be possible in for example countries with plurality elections where in each single constituency one single voter could in principle change the outcome. Extremely high gross volatility could result in

zero net volatility, if at the same time no party suffered any net loss, although
no one voted for the same party. However, combinations II and 111 are most
probable.

Rokkan's theory about party system stability comprises both gross and net volatility. "Cleavages" connect voters and parties and structure these relationships. "Alignments" are the processes that result in stable cleavage patterns. Two separate model assumptions are involved here:

(1) Cleavages are the glue between voters and parties, which means that
gross volatility will be low as long as the cleavages are operating.

(2) Cleavages tend to last for generations, which means that a cleavage
such as class would still structure the electorate in countries with a tradition
for class-based politics.

When (1) and (2) are both true, the result is the "frozen party system," a hypothesis launched by Rokkan & Lipset in the mid-19605. However, gross volatility is likely to increase if (1) or (2) is not true. And when gross volatility occurs, it is only a matter of time before net volatility increases.

It is up for discussion how much volatility a frozen party system could stand. It might be able to stand a high level of gross volatility, but could hardly tolerate much net volatility. When a party system collapses, as in Italy in the 19905, then clearly it cannot be called "frozen."

It has been argued that the frozen party system hypothesis is a rather weak model. It predicts only that certain cleavages will tend to be represented in one election after another, irrespective of whether the parties change. Thus, a swing in voter support from a large communist party to a small social democratic party or vice versa would not signify a party system change, as long as the entire left wing would receive about the same level of support. This aggregate level interpretation of the frozen party system hypothesis means that even if the electorate completely changed their allegiances and all the parties received a totally different level of support, the party system would still be frozen, as long as the traditional cleavages (class, ethnicity,

religion, urban-rural, etc.) were represented by the parties and both sides of the cleavage received about the usual level of electoral support. Party systems would be "un-frozen" only if one or more new cleavages emerged, or if one or two traditional cleavages disappeared.

It seems that Lipset & Rokkan must have had in mind a stronger version of the frozen party system hypothesis than this rather weak version. They apparently claimed that political parties are capable of self-preservation, which will manifest itself in the form of low net volatility as well as low gross volatility.

We should point out that the frozen party system thesis was not only a tentative answer to the question of the historical origins of party conflict but also part of a theory about the "conditions for the development of a stable system of cleavage and oppositions in national political life" (Lipset & Rokkan 1967, 1):

We hope to throw light on the origins and the "freezing" of different types of party systems,
and we seek to assemble materials for comparative analyses of the current alignments of
voters behind the historically given "packages" in the different systems (ibid., 3).

In our opinion, however, both the dealignment processes and the emergence of new politics question the thesis of the frozen party system. In support of this claim, we now examine first the rising volatility and second the shrinking core groups.

Measures of Electoral Change

Let us first look at the individual level changes, often labeled as gross volatility. Butler & Stokes (1971, 337; see also Valen 1981, 332), identify three different indicators of individual electoral change. The first indicator is "party switching" (PS), the second indicator is "overall volatility" (OV), and the third indicator is "total volatility" (TV).

Party switching (PS) covers voters who change their support from one political party in one election to another party in the next election, and who vote in both elections. When we estimate overall volatility (OV), in addition to party switching, we consider all eligible voters in the two elections and define volatile voters as voters who switch between voting and non-voting. Finally, total volatility (TV), in addition to overall volatility, also refers to the total electorate, including those who enter and those who leave the electorate. There are reasons to believe that survey estimates of party switching are more reliable than estimates of total volatility (see estimates in Heath et al. 1991, 20).

Moving to measures of aggregate level electoral change, we will
distinguish between net volatility and seat volatility. By "net volatility"

we mean net changes in electoral support for the political parties from one election to another, also known as the Pedersen Index (Pedersen 1979, 4). Estimates of net volatility might differ due to the comprehensiveness of the electoral statistics employed, or due to computing criteria, but in most cases they tend to go in the same direction. Finally, "seat volatility" refers to changes in the parliamentary representation from one election to another, measured by the same formula as net volatility.

We rely on three different kinds of data to map electoral changes in ten Western European countries: a) data from election surveys estimating individual electoral change; b) data from the aggregate level used to estimate individual level transitions in two cases (Thomsen 1987 for Denmark; and Zilliachus 1995 for Finland); and c) data from the aggregate level on aggregate level electoral change.

Voter Turnout

As a beginning, let us look at voter turnout, which is not logically connected with stability or gross and net volatility. A long-run decline in electoral turnout would be a major manifestation of growing voter instability. It would signify that more and more voters are dissatisfied with the political parties,

and that they make a negative change in the sense that they abstain from
voting. Table 1 shows the development of voter turnout in 18 Western
European countries since 1980.

So far, turnout is down in the 19905. In the eight countries in Table 1 that had elections in 1995-96, the average participation rate was 76.8 percent in 1990-94, compared with 74.9 percent in 1995-96. Whether this trend will result in electoral instability remains to be seen.

Gross Volatility

Moving to electoral instability, we start out at the microlevel, which can be analyzed by means of the concepts of party switching, overall volatility, and total volatility. Although most studies seem to measure party switching the same way, various estimates differ. For example, the estimates of party switching between 1971 and 1972 in the Netherlands vary between 35 percent (Irwin & Dittrich 1984, 288), 26 percent (van der Eik & Niemoller 1985, 357), and 21 percent (Daalder 1987, 230; our calculation). German estimates of party switching between 1980 and 1983 vary from 18 percent (Klingemann 1985, 241; our calculation) to 15 percent (Zelle 1995, 323). We cannot be sure that the authors have used the same methods to estimate overall volatility and total volatility.

These measurement problems notwithstanding, let us look at how these
scores are distributed between the countries (Table 2) and interpret them in
light of the Rokkan mechanism.

Although these estimates are average values, they indicate a variation
between countries. However, looking at the aggregate level, we can see that

one third of the total electorate changes its vote from one election to another. Proceeding to the variation over time, we find that the long-term trend for gross volatility is an increase from one election to another, but as Table 3 shows, that increase is not linear.

Both overall and total volatility reach 40 percent in the 19905, which must be considered a high level of gross volatility. It is not clear which voter instability measure is most appropriate, but Table 4 indicates that the three indicators have a high correlation. This is the first key to understanding electoral volatility.

In sum, these three measures of gross volatility all move in the same direction. There is a general rise in gross volatility in the long run, but there are also some fluctuations. We do not predict, however, that the high figures for the early 1990s indicate a permanent high level of gross volatility as we could well see a reversal. The next question is, what implications does this trend have for net volatility?

Net Volatility

Electoral instability at the microlevel does not necessarily translate into a
party system change at the macrolevel. The issue is the probability of a
connection between gross volatility and net volatility.

As Table 5 shows, the average level of net volatility in Western Europe is slightly less than 10 percent. The level vacillates over time, but in the long run, it is an upward trend. Net volatility increased in the early 19705, decreased in the late 19705, and peaked in the early 19905. There are now signs of a new decline in the mid-19905.

The levels of net volatility vary between the Western European countries, even if the CV scores indicate less variation in the 1980s than in the early 19505. Looking at data from the 19905, we find that in 1994, Italy exceeded a volatility level of 40 percent (compared with 14 percent in 1992) (Table 6). This could be called an "earthquake election."

A sharp rise in net volatility means a shift in voter allegiances. In Rokkan's terminology, such a dealignment process may be followed by a realignment process that reduces the net volatility scores to what they were before the earthquake election. However, if no permanent realignment takes place, a country may maintain a high level of net volatility in election after election.

Now then, what about the relationship between gross volatility and net
volatility? We have corresponding data for party switching (see Appendix 1)
and net volatility from 80 observations. The correlation r = 0.74 indicates

that the two measures covary. The fact that an increase in gross volatility is likely to raise net volatility is the second key to interpreting electoral stability or instability. Comparing the estimated level for party switching (17.7 percent) and the level of net volatility (8.6 percent), we may conclude that net volatility on an average captures some 50 percent of the gross volatility as measured by party switching.

The established parties may attempt to protect themselves against the repercussions of electoral instability by means of institutional mechanisms like election formulas and legal thresholds. Even so, Table 7 shows that seat volatility may reach very high levels in some countries while remaining low in a few others.

On average, seat volatility is actually somewhat higher than net volatility.
This is the third major key to understanding electoral instability.

Another way to examine voter instability at the microlevel is to determine whether the parties can draw on electoral core groups that remain faithful. In Rokkan's interpretation, these core groups would consist of people defined by one single cleavage, for instance class. We now move on to assumption (2) in the Rokkan mechanism with special reference to class, one example of a cleavage.

Cleavages: Class voting

In the Rokkan theory, it is the "cleavage" that establishes a relationship between political parties and voters. Stability in party-voter interaction is enhanced when core groups in the electorate remain loyal to parties. According to Rokkan's historical approach, one such core group is the working class with its traditional support for the political left, often expressed in the form of "class voting." So how has "class voting" developed? Is class voting on the decline or is it more accurate to talk about trendless fluctuations (Manza et al. 1995)?

Let us take a look at the development in two class voting indexes. The first one is the Alford index of class voting, which - in Korpi's terminology - relates the political distinctiveness of the working class to the political nondistinctiveness of the upper class (Korpi 1972, 629). The other index is the Rose-Urwin index (Zuckerman & Lichbach 1977, 526), which relates the class distinctiveness of the left party to the proportion of the workers among the voters (Korpi 1972). When we calculated these measures, we relied on a number of estimates (see Appendix 2). Table 8 shows averages of class voting data from 1950 to 1994 in eight countries.

We can see that class voting declines over time, but it is doubtful if we can declare class voting dead. Rose & Urwin entered an ad hoc criteria (Rose & Urwin 1969, 11) according to which values greater than an index value of 0.17 indicate cleavage voting, while lower values indicate the absence of cleavage voting. In the 1980s and the early 19905, the Rose-Urwin index is below this critical limit, but the early 1990s may also display a small resurgence of class voting. However, there is no doubt that the shrinking working class means that class voting is less important today than it was in the 1950s or in the 1960s (Nieuwbeerta 1995).

The impact of traditional cleavages like class on voting is decreasing, but are there any new cleavages? And what is a cleavage? Is it a permanent tie between voters and parties? In relation to "new" cleavages, we note a general increase of these new orientations, but there is also a fluctuation over time. Consider for example the data on the rise of post-materialist orientations in Europe. Table 9 lists the average values for post-materialist orientations from 1973 to 1993 in eight Western European countries (Abramson & Inglehart 1995). If post-materialism is a new cleavage in Rokkan's sense, then its attraction has almost doubled in less than twenty years, which probably reflects a process of electoral dealignment and realignment.

Thus, in relation to Rokkan's second assumption, we find clear signs of a declining impact of traditional cleavages like class voting. A new cleavage, post-materialism, is on the rise, but this is a gradual, long-term development. In the short run, there are fluctuations that indicate that we may find examples of a resurrection of class voting as well as a decline in postmaterialist orientations in the 19905. It is impossible to tell whether the Rokkan mechanism with its two assumptions will be replaced by a model

which focuses on volatility or a model that harbors new cleavages. Perhaps
we will see a new theory that covers both trends.

Description or Explanation?

The central question in the debate on the Lipset-Rokkan frozen party system hypothesis is whether the changes in party support (macrolevel) or individual voting behavior (microlevel) have been so significant that we can conclude that the hypothesis is no longer valid.

Yet, we may wish to raise a much more fundamental question in relation to the hypothesis: How did Rokkan conceive of the relationship between the parties and the voters? Modeling the interaction between voters and the parties by means of the sociological conception of a cleavage entails a commitment to a few assumptions, although their applicability cannot be taken for granted.

Why would parties focus mainly on mobilization of their cleavage bases? Why would voters take cleavages into account when they choose a party? An entirely different set of assumptions that recognizes the tactics and strategy of electoral campaigning as well as voter efforts to attain reciprocity in relation to the parties might make it easier to understand electoral change

and party system developments. If we go one step further and simply do away with the concept of cleavages, we might be able to model the interaction between parties and voters in a more dynamic fashion, especially if we recognize myopia, opportunistic behavior, and the impact of the media as the most powerful filters between parties and voters. Perhaps the Rokkan mechanism, i.e. inherited cleavages molding the future, is more misleading than helpful in the interpretation of Western European politics today?

The frozen party system hypothesis is not simply an ad hoc description of a set of countries at a particular point in time. It is part of an ambitious theory modeling the relationship between parties and social groups. Theories are like webs of concepts and propositions, where it is often not clear-cut which ones are important and which ones may be revised without the theory tumbling. However, some concepts and propositions are core ones which call for a revision of the entire framework when they are shaken by empirical refutation.

In Rokkan's theory, the core is the connection between the party system (macrolevel) and the social groups (microlevel) by means of cleavages. It may be shattered in one country like Italy, but a strong theory can always handle a few refutations as long as new ad hoc assumptions are added. In the weak version of the theory, the frozen party system hypothesis could perhaps accommodate extremely high levels of gross volatility as long as net volatility remains low. But can the theory survive, when not only social groups fail to back their parties, but also the parties themselves find it increasingly difficult to "camp on the seesaws" (Hedberg et al. 1976).

Normally, theories do not collapse because of a single refutation. It takes an overwhelming amount of counter-evidence to defeat a theoretical framework. However, in Western Europe the changes since the 1960s have been strong: Voter turnout is down, gross and net volatility are up, and the number of parties has increased. Is this enough to shatter the Lipset-Rokkan model? In any case, once we admit that things are not what they used to be concerning parties and social groups, we can proceed to the core of the model, i.e. the cleavage mechanism that builds on the two assumptions of (1) low gross volatility and (2) the persistence of class-based politics. Such a move might give us a different view of how voters and parties interact.

Conclusion

The critical question in relation to the frozen party system hypothesis is whether the various forms of volatility are empirically related. We find that net volatility as well as seat volatility are about half as large as gross volatility. Therefore, we predict that dealignment processes will sooner or

later change the party system. The frozen party system hypothesis falters, because gross, net, and seat volatility tend to coincide. The correlation between the three measures is high, which means that if a country experiences one form of instability, it will probably also experience the others. Since we know that gross volatility is on the rise, we can conclude that it is only a matter of time before net volatility also rises. Thus, neither the weak nor the strong version of the frozen party system hypothesis is valid.

Our conclusions are that in the long run, voter volatility, gross as well as net, increase simultaneously and that traditional cleavage voting, here exemplified by class voting, is declining. Such findings cast doubt on the frozen party system hypothesis. All the evidence seems to imply that assumptions (1) and (2) are no longer valid - one or both are wrong. In order to understand the floating electorate or the electorate in search of new cleavages, we need new assumptions.

ACKNOWLEDGEMENTS

The first version of this article was discussed as a seminar with the Mannheimer Zentrum für
Europäische Sozialforschung, January 1996.

REFERENCES

Aardal, B. & Valen, H. 1989. Vaelgere, Partier og Politisk Avstand. Oslo: Statistisk
Sentralbyra.

Aardal, B. & Valen, H. 1995. Konfligt og Opinion. Oslo: NKS-forlaget.

Abramson, P. R. & Inglehart, R. 1995. Value Change in Global Perspective. Ann Arbor:
University of Michigan Press.

Alford, R. A. [1963] 1973. Party and Society: The Anglo-American Democracies. Westport,
CT: Greenwood Press.

Allardt, E. & Pesonen, P. 1967. "Cleavages in Finnish Politics," in Lipset, S. M. & Rokkan, S.,
eds., Party Systems and Voter Alignments: Cross-National Perspectives. New York: Free
Press.

Anker, H. 1996. Personal Communication: data from the 1989 and 1994 Dutch election study.
University of Nijmegen.

Bartolini, S. & Mair, P. 1990. Identity, Competition, and Electoral Availability: the
Stabilization of European Electorates 1885-1985. Cambridge: Cambridge University Press.

Berger, M. et al. 1986. "Legitimierung des Regierungswechsels," in Klingemann, H.-D. &
Kaase, M., eds., Wahlen und politischer Prozess: Analysen aus Anlass der Bundestagswahl
1983. Opladen: Westdeutscher Verlag.

Butler, D. & Stokes, D. 1971. Political Change in Britain: Forces Shaping Electoral Choice.
Harmondsworth: Penguin.

Daalder, H. 1979. "The Netherlands," in Henig, S., ed., Political Parties in the European
Community. London: Allen & Unwin.

Daalder, H. 1987. "The Dutch Party System: From Segementation to Polarization - and
then?," in Daalder, H„ ed., Party Systems in Denmark, Austria, Switzerland, the
Netherlands, and Belgium. London: Pinter.

Dalton, R. J. 1989. Politics in Germany. Glenview, IL: Scott, Foresman.

Gehmacher, E. 1974. "Faktoren des Wählerverhaltens," in Fischer, H., ed., Das politische
System Österreichs. Wien: Europaverlag.

Gehmacher, E. 1982. "Faktoren des Wählerverhaltens," in Fischer, H., ed., Das politische
System Österreichs. 3rd ed. Wien: Europaverlag.

Gilljam, M. & Holmberg, S. 1990. Rött, Blått, Grant: en Bok om 1988-års Riksdagsval.
Stockholm: Bonniers.

Gilljam, M. & Holmberg, S. 1993. Väljarna Inför 90-talet. Stockholm: Norstedt.

Haerpfer, C. 1983. "Nationalratswahlen und Wahlverhalten seit 1945," in Gerlich, P. &
Müller, W. C, eds., Zwischen Koalition und Konkurrenz: Österreichs Parteien seit 1945.
Wien: Braumüller.

Haerpfer, C. 1985. "Austria," in Crewe, I. & Denver, D., eds., Electoral Change in Western
Democracies: Patterns and Sources of Electoral Volatility. London: Croom Helm.

Heath, A. et al. 1991. Understanding Political Change: The British Voter 1964-1987. Oxford:
Pergamon Press.

Heath, A. et al., eds. 1994. Labour's Last Chance?: The 1992 Election and Beyond. Aldershot:
Dartmouth.

Hedberg, 8., Nystrom, P. C. & Starbuck, W. H. 1976. ' "Camping on Seesaws': Prescriptions
for a Self-Designing Organization," Administrative Science Quarterly 21. 41-65.

Holmberg, S. 1981. Svenska Väljare. Stockholm: Liber.

Holmberg, S. & Gilljam, M. 1987. Väljare och Val i Sverige. Stockholm: Bonniers.

Irwin G. A. & Dittrich, K. 1984. "And the Walls Came Tumbling Down: Party Dealignment in
the Netherlands," in Dalton, R. J. et al., eds., Electoral Change in Advanced Industrial
Democracies: Realignment or Dealignment? Princeton, N.J.: Princeton University Press.

Janda, K. 1980. Political Parties: A Cross-National Sun<ey. New York: The Free Press.

Keesing's Record of World Events. Cambridge: Catermill.

Klingemann, H.-D. 1985. "West Germany," in Crewe, I. & Denver, D., eds.. Electoral Change
in Western Democracies: Patterns and Sources of Electoral Volatility. London: Croom
Helm.

Korpi, W. 1972. "Some Problems in the Measurement of Class Voting," American Journal of
Sociology 78, 627-642.

Lijphart, A. 1974. "The Netherlands: Continuity and Change in Voting Behavior," in Rose, R..
ed., Electoral Behavior: A Comparative Handbook. New York: Free Press.

Lijphart, A. 1994. Electoral systems and Party Systems. Oxford: Oxford University Press.

Linz, J. J. 1967. "Cleavages and Consensus in West German Politics: The Early Fifties," in
Lipset, S. M. & Rokkan, S., eds., Party Systems and Voter Alignments: Cross-National
Perspectives. New York: Free Press.

Lipset, S. M. & Rokkan, S. 1967. "Cleavage Structures, Party Systems and Voter Alignments:
An Introduction," in Lipset, S.M. & Rokkan. S., eds.. Party Systems and Voter Alignments:
Cross-National Perspectives. New York: Free Press.

Mackie, T. T. & Rose, R. 1991. The International Almanac of Electoral History. 3rd ed.
Basingstoke: Macmillan.

Mair, P. 1993. "Myths of Electoral Change and the Survival of Traditional Parties." European
Journal of Political Research 24, 121-133.

Manza, J. et al. 1995. "Class Voting in Capitalist Democracies Since World War II:
Dealicnmcnt, Realignment, or Trendless Fluctuation?" Annual Review of Sociology 21.
2137-2162.

Matheson, D. K. 1979. Ideology, Political Action and the Finnish Working Class: A Suney
Study of Political Behaviour. Helsinki: Societas Scientiarium Fennica.

Nieuwbeerta, P. 1995. The Democratic Class Struggle in Twenty Countires 1945/1990.
Amsterdam: Thesis Publishers.

Norton, P. 1984. "Britain: Still a Two-party System?." West European Politics 7. 3. 27-45.

Pappi, F. U. 1977. "Sozialstruktur, Wertorientierungen und Wahlabsicht." Politische
Viertalsjahrschrifl 18, 195-229.

Pedersen, M. 1979. "The Dynamcis of European Party Systems: Changing Patterns of
Electoral Volatility," European Journal of Political Research 7, 1-26.

Petersson, O. 1978. "Valet 1976," in Statistiska Centralbyrån, Allmänna Valen 1976: Del 3:
Specialundersökningar. Stockholm: SCB.

Petersson, O. & Särlvik, B. 1975. "Valet 1973," in Statistiska Centralbyrån, Allmänna Valen
1973: Del 3: Specialundersökningar. Stockholm: SCB.

Plasser, F. & Ulram, M. 1988. "Grossparteien in der Defensive: Die Österreichische Parteienund
Wählerlandschft nach der Nationalratswahl 1986," in Pelinka, A. & Plasser, F., eds.,
Das Österreichische Parteiensystem. Wien: Böhlau.

Plasser, F. & Ulram, M. 1995. "Konstanz und Wandel im Österreischischen Wählerverhalten,"
in Müller, W. C. et al., eds., Wählerverhalten und Parteien-wett-bewerb: Analysen zur
Nationalratswahl 1994. Wien: Signum-Verlag.

Railings, C. S. & Andeweg, R.B. 1979. "The Changing Class Structure and Political
Behaviour: A Comparative Analysis of Lower Middle-class Politics in Britain and the
Netherlands," European Journal of Political Research 7, 27-47.

Rose, R. & Urwin, D. 1969. "Social Cohesion, Political Parties and Strains in Regimes,"
Comparative Political Studies 2, 7-67.

Sänkiaho, R. 1993. "The Structure of the Party Support," Paper Presented to the Nordic
Political Science Association Meeting, Oslo.

Thomas, A. H 1973. Parliamentary Parties in Denmark, 1945-1972. University of Strathclyde,
Survey Research Center: Occasional Paper no 13.

Thomsen, S. R. 1987. Danish Elections 1920-79: A Logit Approach to Ecological Analysis and
Inference. Aarhus: Politica.

Thomsen, S. R. 1995. "The 1994 Parliamentary Election in Denmark," Electoral Studies 14,
315-322.

Valen, H. 1972. "Partiforskyvninger ved Stortingsvalget 1965," in Valen, H. & Martinussen,
W., eds., Velgere og Politiske Frontlinjer: Stemmegivning og Stridsspoersmål 1957-1969.
Oslo: Gyldendal.

Valen, H. 1981. Valg og Politikk: et Samfunn i Endring. Oslo: NKS-forlaget.

Valen, H. & Aardal, B. 1983. Et Valg i Perspektiv: En Studie av Stortingsvalget 1981. Oslo:
Statistisk Sentralbyra.

Valen, H. et al. 1990. Endring og Kontinuitet: Stortingsvalget 1989. Oslo: Statistisk
Sentralbyra.

van der Eik, C. and Niemoller, B. 1985. "The Netherlands," in Crewe, I. & Denver, D., eds.,
Electoral Change in Western Democracies: Patterns and Sources of Electoral Volatility.
London: Croom Helm.

Wolinetz, S. 1977. "The Dutch Labour Party: A Social Democratic Party in transition," in
Paterson, W. E. & Thomas, A. H., eds., Social Democratic Parties in Western Europe.
London: Croom Helm.

Worre, T. 1976. "Social Bakgrund og Partivalg," in Borre, O. et al. Vaeigere i 70'erne:
Resultater fra Interviewundersoegelser ved Folketingsvalgene i 1971, 1973 og 1975.
Koebenhavn: Akademisk Forlag.

Worre, T. 1978. "Folketingsvalget 15. Februar 1977," Økonomi og Politik 52, 18-47.

Zelle, C. 1995. "Social Dealignment versus Political Frustration: Contrasting Explanations of
the Floating Vote in Germany," European Journal of Political Research 27, 319-345.

Zilliachus, K. O. 1995. Finlandsk Kommunism i Ljuset av Väljarstöd 1945-1991. Helsingfors:
Finska Vetenskaps-Societeten.

Zuckerman, A. & Lichbach, M. I. 1977. "Stability and Change in European Electorates,"
World Politics 29, 523-551.