Path: > Syllabus > Topics and Readings >

Types of Party Systems

Blondel Reading


From Peter Mair (ed.) The West European Party System (Oxford: Oxford University Press, 1990), pp. 302-310.


If undertaken on a world basis, the analysis of party svstems would require a consideration of the number of parties, of their strength, of their place on the ideological spectrum, of the nature of their support, and of their organization and type of leadership. In the context of Western liberal democracies, it is possible to limit the analysis to the first three of these characteristics. With very few exceptions, Western parties can be deemed to be of the 'legitimate' mass type and to have a regularized system of leadership selection. Charismatic leadership is exceptional and, at the other extreme, Communist parties are becoming increasingly similar to other parties. Western parties appeal nationally to the electorate by trying to put across a general image; differences are more to be found in the type of image than in the structural characteristics of the organization.

I. Number of Parties in the System

Two-party systems are rarely defined operationally. Yet, among Western democracies, four groups are clearly discernible (see Table 22. 1). Five countries give on average more than go per cent of their votes to the two major parties. Second, five countries give between 75 and 8o per cent of their votes to the two major parties (though Belgium fell markedly below this level at the last election). In six countries, the two major parties obtain about two-thirds of the votes (though Holland is somewhat different, as we shall see). Finally, three countries give about half their votes to the two major parties, though, for the last two elections, France would have to be moved from the fourth to the third category: but as the French situation may well alter again after the end of the de Gaulle era, it is probably more realistic to draw conclusions from the average of the whole post-war period than merely from developments in the1960s.

TABLE 22. 1. Average Two-Party Vote, 1945-66 (percentage)

United States


New Zealand




United Kingdom




West Germany




























Countries belonging to the first group can be defined as two-party systems, though there is some ambiguity in the cases of Australia and Austria; in both these countries some governments depended for their constitution or maintenance on the support of more than one party. The five countries of the second group constitute the three-party systems, Germany having arrived at this status as a result of the operation of the electoral system as well as of the Adenauer tactics. The nine countries of the last two groups are the genuine multiparty systems, in which four, five, or even six parties play a significant part in the political process.

2. Strength of Parties

Little has to be said about countries of the first group, in which go per cent or more of the electors vote for, and go per cent or more of the seats are distributed between, the two major parties. It is interesting to note that discrepancies in strength between the two major parties are remarkably small, at least if averaged over the post-war period. While it could in theory be the case that one party might be permanently much larger than the other, no two-party system (i.e., no system in which the two parties obtain 89 per cent of the votes) gives to the larger of the two parties a permanent premium of over i o per cerit of its own electorate. Despite differences in social structure, for instance, between the United States, the United Kingdom, and Austria, the electorate distributes its preferences fairly evenly between the two parties. Although it cannot be stated with assurance that such a situation cannot exist in any type of social system (there are indeed examples of uneven distribution of party support among two-party systems outside Western democracies), it seems possible to hypothesize, in the absence of contrary evidence, that, in Western democracies, two party systems show a tendency towards a relative equilibrium between the two parties.

Countries in the three-party system group also share several characteristics. First, disparities in strength between the two larger parties are generally much greater than among countries in the first group. They seem structural in that the swing of the pendulum does not, as among countries of the first group, diminish the gap, even if one considers a fairly long period: the average percentage point difference between the two major parties among the five countries of the first group is only 1.6 if all post-war elections are taken into account; it is 10.5 in countries of the second group (excluding Luxemburg where data for all elections were not available) [see Table 22.2]. The often stressed phenomenon of the German SPD, which has not been able to achieve equality of strength with the CDU, is thus a general phenomenon among countries of this group. It can therefore be stated that in countries in which about 8o per cent of the votes go to two parties the distribution of the support tends to be uneven.

TABLE22.2. Average Strength of the Two Major Parties
in Two- and Two-and-a-half Party Systems
(percentage votes cast)
Two-party systems
Two-and-a-half party systems



United States






New Zealand






Australia United















Mean disparity between the two major parties=1.6

Mean disparity between the two major parties=10.5

Three-party systems also have another, and converse, characteristic. They all have two major parties and a much smaller third party. While it would seem theoretically possible for three-party systems to exist in which all three significant parties were of about equal size, there are in fact no three-party systems of this kind among Western democracies: all the three-party systems are in the second group; all systems of the third group have more than three significant parties. It can thus be stated that, while theoretically possible, a genuine threeparty system is not a likely type of system among Western democracies: countries of the second group should therefore be more strictly labelled as 'two-and-a-half-party systems'. It may indeed be permissible to add, after considering the evolution of party systems, particularly in the early part of the twentieth century, that genuine three-party systems do not normally occur because they are essentially transitional, thus unstable, forms of party systems.

The situation arising out of the 1965 Belgian election was thus of particular interest, since the upsurge of the Liberal party and the decline of the two major parties was such that Belgium appeared to have moved into the exceptional and seemingly transitional position of a 'genuine' three party system. If the proposition advanced earlier is correct, it would seem that in the next few years Belgium will move to one of three types of further changes. The Liberals could return to their 'normal' position of small party; they could displace one of the two major parties (an unprecedented development in Western Europe since all the movements happened in the other direction); a split could occur among the supporters of one or both of the major parties and Belgium might move from the second group of countries (two-and-a-half party systems) to the third or fourth (multiparty systems). From the experience of other Western democracies the first and third of these three outcomes appears more probable than the maintenance of a genuine three-party system.

We included Holland among the countries which constituted the third group, composed of those countries in which the two largest parties obtained about two-thirds of the votes of the electorate. But if we consider the relative strength of the two major parties, the five other countries of the group differ markedly from Holland, in that they have one very large party, which might be defined as dominant as it obtains about 40 per cent of the electorate and generally gains about twice as many votes as the second party. In fact, in both Norway and Sweden, the mechanics of the electoral system have sometimes enabled this dominant party to obtain the absolute majority of seats in the Chamber. But these countries are at the same time multiparty systems as they all have four or five significant (and often well-structured) parties playing a crucial role in the operation of the political system. Four of the five Scandinavian countries are included in this group (the fifth being Italy), though in Iceland the dominant party is not the Socialist but the Conservative (Independence) party. The Fifth Republic, at least under de Gaulle, appears to be moving towards this pattern of party system, though the characteristics of the Gaullist party are such that it seems unreasonable to predict that the present party configuration is likely to survive de Gaulle. Thus, alongside two-party systems and two-and-half-party systems, the third group should be defined, not so much as a multiparty system in which two parties obtain about two-thirds of the votes of the electorate, but as a multiparty system with a dominant party which obtains about two-fifths and less than half of the votes.

Finally, the last four countries (if Holland is included and France maintained in the group in the view of the past and possible future performance of that country) constitute the genuine multiparty systems: they have no dominant body and indeed seem to show a pattern of political behaviour in which three or four parties are equally well placed to combine or form coalitions. Recent electoral movements in Holland appear to bring that country gradually nearer towards this model: the slow decline of the Labour party and the patterns of governmental formation have tended to conform fairly closely to those which have been customary of Finland.

3. Ideological Spectrum and Party Strength

Western democracies can therefore be divided into four groups, if both numbers and strengths are taken into account: five are in the two-party system group, five in the two-and-a-half-party system group, are in the group of multiparty systems with dominant party, and four in the group of multiparty systems without dominant party. But the party system can only wholly be defined if we take into account the position of parties on the ideological spectrum, particularly when the system does not have a 'symmetrical' character.

Even in Western democracies, parties are difficult to categorize accurately. The United States and Eire always had parties which do not fit in any easily recognizable typology. The French Gaullists are not ordinary Conservatives, though they may have to be lumped with Conservatives for comparative purposes. The other multi-party system countries all have parties of a somewhat peculiar type, such as the two Conservative partles in Holland, the Swedish party in Finland, and the party of the Peasants, Artisans, and Bourgeois in Switzerland. In all these cases, certain rough approximations have had to be made.

Table 22.3 shows the panorama of party strengths within the ideological spectrum. Group I, except for the United States, is fairly homogeneous: the four other countries in the group have a large Socialist party and a large Conservative or Christian party, together with a small other group, whose position in the centre is perhaps sometimes problematic. Countries of group 2 have two types of party systems: Belgium, Germany, and Luxemburg resemble countries of group I, except that the centre party is stronger and, as we noted earlier, this increased strength is mainly at the expense of one only of the two major parties (in fact the Socialist party). The other two countries of the group (Canada and Eire) are different: the small party is not the centre, but the left-wing, party and, probably not quite accidentally, the right-wing party Is not Christian but Conservative. In the early part of the twentieth century, it would probably have been argued that both these countries were still in a transitional stage: Canada has a party system not unlike that of Britain in igo6. But it must by now be recognized both that the two- and -a-h alf- party system is stable (Belgium and Germany have had for long periods the British party system of the 192os and do not appear to move further towards a two-party system) and that the two-and-a-half-party system is stable even if the left-wing party is the small party. Neither the Irish Labour party nor the Canadian NDP appear to be in a position to overtake the centre party in the near future: their position of small party seems stable. The reasons for the stability of the Canadian model of two-and-a-half-party system are probably to be found along lines similar to those which account for the stability of the American parties, which have remained in existence despite earlier predictions to the contrary.

Two types, and not more than two types, of multiparty systems with a dominant party can be found. Three Scandinavian countries have a dominant Socialist party; Iceland and Italy have a dominant Conservative or Christian party. In the latter two cases, we encounter for the first time countries with a really large Communist party also in these countries you can order cialis online. In Iceland and Italy (as in Finland and even France) the Communist party remained fairly stable throughout the whole period; the Socialist party is therefore correspondingly weak and the left is divided; as is well known, the converse occurs in Norway, Denmark, and Sweden. In these three countries, though the agrarian party is not placed symmetrically in relation to the Communists on the political spectrum, it probably plays the same divisive part. The absence of religious divisions might have led one to expect the three Scandinavian countries which do not have a large Communist party to have two-and-a-half-party systems, but the apparent feelings of identity of the agricultural community have created permanent cleavages on the right which have produced a party system in many ways 'symmetrical' to that of Iceland and Italy.

Countries of the fourth group have no dominant party: party strengths are therefore fairly evenly spread across the ideological spectrum, though Holland and Switzerland, with weak Communist parties, display somewhat less spread than France and Finland. This latter country combines the splintering characteristics of left and right of the Scandinavian countries: the divisions of Iceland are superimposed on those of Sweden.

There are thus six types of party systems in Western democracies. At one extreme are the broadly based parties of the two-party system countries: the United States is the most perfect case of this type, but four other countries closely approximate this model and they only diverge inasmuch as they have a small centre party and are divided ideologically between conservatives and socialists. At the other extreme, the votes of the electors are spread fairly evenly, in groups of not much more than 25 per cent and in many cases much less than 25 per cent over the whole ideological spectrum, as in Holland, Switzerland, France, and Finland. Between these two poles, one finds four types of party systems: five countries have two-and-a-half-party systems: among them, three have a smaller centre party, while the other two have a smaller left-wing party. The five remaining countries are multiparty systems with a dominant party, three of them having a dominant socialist party opposed by a divided right, largely because of the presence of an agrarian sentiment in the countries concerned, while the other two have a strong right-wing party opposed by a divided left, largely because of the presence of a substantial Communist party.

A number of types of party systems, which are theoretically possible, do not appear to exist. There are no three-party systems, as we saw; there are no 'unbalanced' two-and-a-half party systems, out of the three which might have existed. Dominant parties in multiparty systems tend to be of the right or left, not of the centre. Patterns of party systems in Western democracies are thus limited in number. There are discrete points at which party systems are to be found: not only social structures but balance and equilibrium have surely to be taken into account in an analysis of the real world distribution of Western party systems.

*excerpted from 'Party Systems and Patterns of'Government in Western Democracies', Canadian journal of Political Science, 1 /2 (1968), 180-203. Reprinted by permission of the author and The Canadian Political Science Association.