Resumé

Minor and new parties tend to be forgotten by analysts of parties and party systems. Good
as well as bad reasons account for this neglect. The neglect ought, however, to be
remedied, since it can be argued that minor, and especially new minor, parties play an
important role in the transformation of party systems.

On the basis of a 'broad' definition of the concept of political party it is suggested that it is worthwhile to look upon parties as mortal organizations bounded by a lifespan. The lifespan of a party can be described by means of four threshold concepts, and it is further argued that lifespan curves can be studied in terms of their modality, dispersion, flatness, and skewness.

A number of questions are derived from these concepts. A new typology of minor
parties and of party lifespans may be generated from this approach.

Introduction: Why another typology of parties?

The very fact that most of those parties which existed around 1950 are still
operating testifies to the adaptability of political parties in an era of considerable
societal change.

Recent research findings do, however, force upon us some qualifications to this statement. Thus it has been found in a survey of parties in 53 countries scattered all over the globe that 39% of those parties which were active in the 1950's had ceased to exist in 1979 (Janda 1980, 162—69). Persistence varies widely across geographical regions, but even in the relatively stable countries of Western Europe 15^—20% of incumbent parties disappeared over a period of 30 years. Equally suggestive is the fact that Kenneth Janda found 50 new parties qualifying for inclusion in his universe of parties during the years 1962^—79.' Among these new parties a considerable number emerged in Western Europe. Finally, Janda has drawn our attention to the interesting theoretical idea that parties can be characterized by means of a concept of party lifespan.2 Using an apt metaphor he is able to conclude

that the mortality of parties to some extent is a function of party age in the sense that 'infant mortality' of parties is much higher than mortality among more mature parties. Not surprisingly it turns out that the party systems of Western Europe fall on one end of a global continuum, since 'only Western Europe can be said to be conducive to party life in the sense of providing for party birth and party continuity' (Janda 1980, 169, my emphasis).

Other students of parties and party systems have observed a certain instability in European party systems, linked to the occurrence of new parties (see e.g. Smith 1978; Wolinetz 1979; Pedersen 1979; Borre 1980; Mayer 1980). Even if party systems are still stable in most ways, quite a few of them have been unstable in the sense that individual electoral volatility within the system as well as the derived aggregate volatility of the system have changed. The appearance of new parties in the electoral arena, as well as the disappearance of older parties, are linked to this development (Pedersen 1979; 1982).

Whatever one's stand in the debate on the persistence or change in Western European party systems, it is beyond doubt that parties as organizations are not 'frozen' components in the political systems. They change all the time, and they do so in many respects. The most critical change in the daily life of parties takes place at elections, when the positive or negative sanctions of the voters are translated into the crucial figures of seats won or seats lost in parliament.

These ups and downs are registered by politicians as well as by poli tical scientists. For all their intrinsic interest they do tend to obscure one crucial faet about parties, namely that parties are born, parties live, and parties die — to use again this metaphor.

With a slight exaggeration it can be said that the study of parties has concentrated on the active life of parties, whereas the two phases of ernergence/birth and disappearance/death have not been given nearly as much attention.

This author has no knowledge of any systematic and comparative effort to study the decay and disappearance of political parties. There is no parallel in political science to the studies of conditions for survival of business organizations and bureaucratic organizations (see e.g. Dill 1965; Downs 1967). Paradoxically the situation is not much better with regard to the systematic study of the emergence of parties, even if the literature is considerable ³ The main bulk of research has been carried out on the emergence of the modern mass party system in the 19th century European states, or on the formation of party systems in 20th century Third World states. Very

little has been written on the formation of new parties in contemporary
Western European politics.

New parties mostly start out as and remain minor parties. This fact may account for the relative lack of systematic analysis of these phenomena. In this respect Maurice Duverger made a start by introducing the categories of personality parties and permanent minority parties (Duverger 1959, 290_—99), but apart from some historical case studies, and a few national studies (e.g. Fischer 1974; Daalder 1966), not much has been done in this field since then. Stephen L. Fischer, who did a study of minor parties in the Federal Republic of Germany, has suggested four reasons for this neglect (Fischer 1980, 609—10):

1) by concentrating on major parties that have proven a high level of support and stability,
the analysis of party politics becomes more manageable;

2) by omitting minor parties from the analysis severe problems of data collecting are avoided;

3) minor parties are often thought to be without any significant impact on government policy
and other political phenomena; and

4) minor parties are dropped from the analysis since they are considered dysfunctional elements
in the political systems.

Right or wrong, one is tempted to concur with Fischer when he states that 'although the exclusion of minor parties from consideration in comparative studies of political parties may be understandable, it is not necessarily justifiable, and when he suggests that 'several of the conditions associated with the decline of the major parties in Western political systems appear conducive to increased minor-party activity and success' (Fischer 1980, 610).

If it is admitted that the study of minor and/or new parties deserves more attention, the problem of the proper approach arises. One may wish to speculate on the origin of new parties (see e.g. Down 1977, 127_—32); Hirschman 1970, 84_—85), or one may prefer to start collecting data for all those hundreds of parties which have been omitted by Kenneth Janda and other data collectors (Janda 1980; Mackie & Rose 1974).

It will be suggested here that the mapping of minor and new parties, and the analytical discussion of their origins, their fates, and their political impacts, can be aided by the construction of a heuristic typology which may lead to the identification of important types of these parties. The concept of party lifespan may serve as the principal building block for such a typology.

The Concept of Political Party

In order to develop a typology of parties that will make it possible to study parties as emerging as well as disappearing phenomena, one has to be particularly careful with the definition of the concept of political party. The demarcation line between those groups, movements, organizations etc. which should be considered political parties, and those which should not be called so, is of crucial importance for the subsequent steps in the argument since it determines the universe of relevant units of analyses. For reasons of expediency students of political parties are sometimes delimiting their universe of parties in such a way that minor parties are left out of analysis — even if they may be considered included in the theoretical concept used by the author (e.g. Janda 1970, 83^—84). For obvious reasons this, otherwise tempting, strategy cannot be used here. Nor can we follow another strategy, according to which the universe of theoretically interesting parties is shrunk by means of various qualifiers, like e.g. 'relevance' (Sartori 1976, 119^—30). However well founded such a strategy may be for some theoretical purposes, it is invalid in this context.

The line of demarcation between political parties and other social organizations is of course blurred; this ambiguity reflects the complexity of political life. Practical ambiguity is reflected in the very many discussions of the proper conceptualization of party which can be found in the party literature. Efforts have been made to distinguish between party and faction, party and interest organization, and, most recently, party and various popular 'grass-roots' movements. As said, this is not necessarily a sign of conceptual muddle, but rather reflects the many facets of political life. Organizations, factions, cliques, movements continually change forms, and at times they suddenly — or incrementally — appear in the form of the fullblown party. In the same way political parties sometimes disappear suddenly, and sometimes they fade away through a regression of organizational forms. The existence of a 'grey zone' between the traditional party and other organizational forms calls for concepts and for better maps.

In this article some distinctions will be made within a limited range of forms. Thus I will concentrate on organizations, which have as a minimum a declared intention to present candidates at elections for public office. This intention delimits parties from other organizations and groups.

There exists a slight difference between this definition and some of the prevailing definitions. There is no need to enter a long discussion on conceptual problems, so let it suffice to bring to the attention two among the many current definitions of the concept of political party, namely the definitions

by Gunnar Sjöblom and Giovanni Sartori, respectively.

Gunnar Sjöblom differentiates parties from other political phenomena
by defining the concept of party as

'an organization that appoints candidates at general elections to the system's representative
assembly and also to other political positions' (Sjöblom 1968, 21).

Giovanni Sartori in his magisterial discussion of the conceptual problems
in party research prefers the following definition:

'a party is any political group that presents at elections, and is capable of placing
through elections, candidates for public office' (Sartori 1976, 64).

These definitions are not chosen at random from among the multitude of party definitions. They are typical.⁴ They are, both of them, created for special purposes and therefore stress different, if related, characteristics. For most purposes they will do — hence the relatively undisputed status of these 'electoral' definitions of the concept.

In this context it is, however, obvious that one has to broaden the denotation of the concept in order to make it possible to discuss parties as emergent and/or disappearing political phenomena, i.e. organizations, which are not yet/not any longer capable of placing candidates (cf. Sartori), and organizations which are not yet/no longer able to nominate candidates and participate in elections (cf. Sjöblom) — but which nevertheless have the will and wish to do so. Hence the definition of the party as an organization — however loosely or strongly organized — which either presents or nominates candidates for public elections, or which, at least, has the declared intention to do so.

This definition demarcates the universe of organizations that will have to
be taken into consideration, if one wants to understand the emergence, the
life, and the disappearance of parties, in short the lifespan of parties.

This universe is conceptually bounded. Unfortunately the denotation of the concept will never be known, since it is physically impossible to identify on a global scale all those organizations which qualify for inclusion.⁵ On the other hand it is perfeectly possible, if laborious, to identify these parties in most European polities.

According to Merkl (1980, B—9)8_—9) the number of parties gaining representation in 18 Western European polities around the middle 1970's amounted to 105. Of these, 55 obtained more than 10% of the seats in their respective parliaments, and 35 got access to governmental positions. But the number of parties that tried without succeeding in gaining representation was

higher. A total of 115 European parties acquired at least 1% of the vote. Many more were competing in elections without obtaining this minimal fraction of the vote, and it is beyond doubt that a considerable number of rudimentary party organizations were politically active without actually nominating candidates in these elections.

If students of parties only consider those approx. 50 European parties
which actually are, or are close to being 'relevant' parties, an important
aspect of the party systems, and especially its development, is missed.

The Party Lifespan and Its Thresholds

The notion of a party lifespan is based upon an assumption which many political scientists forget, and very few politicians like to ponder, namely that parties are mortal organizations. What normally comes to the eye are the small or large fluctuations in party strength from one election to the next, from one public opinion poll to the next. Very seldom will a party totally disappear by disintegrating or dissolving itself. In most cases of party breakdowns some activities will be carried on by some of the supporters, typically working under new and different political labels. But the histories of e.g. the French Mouvement Républicain Populaire and Le Mouvernent Poujade should remind us about the mortality of even quite large parties.

As parties increase or decrease in strength in the electorate as well as in parliament, they pass or eventually strive to pass — or avoid — some important thresholds. The concept of thresholds is a general notion which is universally applicable, — or to be more precise — applicable in all liberal democracies.

The first threshold is the threshold of declaration, the point in time when a political group declares its intention to participate in elections and thus becomes a party in the sense of party used here. The declaration of intent is usually coupled with some overt acts, like campaigning, collecting of signatures for registration, various media events etc.

It is important to note that this threshold also has a reciprocal interpretation, i.e the point when a party decides to abstain from further electoral participation, or acts to the effect of doing so. One should note that the threshold of declaration in contrast to some other thresholds (cf. below) is not necessarily related to the actual numerical size of the political group. Although in most cases these critical decisions are made in small group:» of militants, the threshold can just as well be passed — both ways — by groups or organizations of considerable size.

The second threshold is the threshold of authorization. In all countries legal regulations are set up, according to which certain requirements have to be met by a party in order to participate in the electoral campaign and the election itself. When a party fulfills these requirements, it is as a minimum authorized to nominate candidates.⁶ Often additional benefits flow from this authorization, such as financial support for candidates and/or the party, access to mass media — in particular state-controlled television — and radio networks etc. The requirements may differ in qualitative as well as quantitative terms. A certain number of signatures collected per candidate or for the party as such is a typical requirement. A deposit of a certain amount of money may complement or substitute for the collection of signatures (see e.g. Hand et al. 1979). In some cases these requirements will only have to be met by hitherto unrepresented organizations, not by those already present in the outgoing parliament. The threshold of authorization is primarily a barrier against the intrusion of new parties.

Next follows the the threshold of representation, the barrier which all parties have to cross in order to obtain seats in the legislature.⁷ The 'height' of this barrier varies across polities: the minimum is the number of votes necessary to gain one seat, and the maximum is probably the one practiced in the Federal Republic of Germany, i.e. 5% of the total vote.⁸ The qualitative meaning of the threshold is, however, equivalent. It defines the 'ins' and the 'outs' in the party system. New as well as old minor parties keep their eyes fixated on the magical number of votes which constitutes the operational threshold of representation. The anticipated reaction of voters against those parties whose electoral strength is slightly above the threshold may be one of the most important strategic determinants in some party systems — witness the present Social Democratic 'dominance by default' in Denmark (Smith 1980, 79).⁹

While the thresholds of authorization and representation may differ across polities, they do share the property that they lend themselves to operationalization. As legal norms they will typically be defined in an electoral law.

Not so with the final threshold to be introduced here. This is the threshold of relevance, as it can be derived from the writings of Giovanni Sartori (especially Sartori 1976, 119^—25). Although such criteria of relevance/ irrelevance may be difficult to operationalize, the very notion makes theoretical sense, as amply demonstrated by Sartori. It does also make a lot of practical sense: leaders of very small as well as somewhat bigger parties perfectly well understand when and in what way their party belongs among the relevant, and when they are relegated to the political impotence which

often signals further decay, eventually the fading away of the party.

It is important to note the relations and the character of these four thresholds. In the first place, they form a continuum in the sense that parties strive to advance as much as possible beyond the threshold of declaration. The goal of any minor party is to pass the threshold of relevance, and, next, to become an influential, at best a ruling party. But on the other hand these thresholds are not just markers on a continuum. Quite to the contrary, they divide up the history of a party in discrete phases, each with its own dominant and different quality.

With these conceptual tools in hand we may take the next step in the argument by sketching some crude models of the lifespan of parties. Starling out with the assumption that no parties, however long-lived and stable they may look, are eternal organizations, it is possible to differentiate among political parties according to the number of thresholds they succeed in passing during their lifespan. Fig. 1 presents some stylized lifespans.

Skipping type V which, according to the definition used here, does not qualify as a party (even if a study of the potential for becoming a party among social and political movements is a fascinating topic), the four remaining categories (I^—IV) portray in the crudest of terms party histories in an inclusive way. Research on parties and party systems mostly concentrate on one or two of these types, seldom on all of them. Giovanni Sartori asks us to concentrate all our attention on relevant parties. If we don't do so, we are told that we count in an unintelligent way (Sartori 1976, 124). He is probably right, but only if the primary goal of the research is to understand how parties affect political systems at the most general level, that of societal change and political stability. For obvious reasons legislative specialists will tend to concentrate their studies on those parties that gain representation and make a legislative impact. Electoral specialists will sometimes take a special interest in political forces that shape unconventional behaviour, including voters' support for new minor parties — but as a rule these parties will only appear in tables under the heading 'Other Parties'. Stephan L. Fischer has told us why it is so, cf. above.

In a developmental perspective one has, however, to consider all these four types as worth while studying. No one — be it a politician or a political scientist — will know beforehand how the lifespan of an emerging party will look like. This is why these four thresholds and the corresponding categories of party lifespans deserve further specification, aiming at a new typology.

The Dimensions of the Party Lifespan

Lifespan-curves for political parties (hereafter Ls-curves) only in rare cases will look like the bell-shaped, unimodal curves of Fig. 1. Observed close-up, i.e. in a time perspective of a few election periods, it is often extremely difficult to discern any pattern at all. But from a distance, and when qualitative, threshold-transcending change is focused upon, it is fairly easy to identify a limited number of patterns. These patterns can be described in terms of four dimensions, namely the modality, the dispersion, the skewness, and the flatness of the Ls-curve.¹⁰

According to the Rokkan-Lipset hypothesis about the freezing of the major party alternatives in the wake of the extension of the suffrage and the mobilization of the masses, one should expect the typical Ls-curve of major Western European parties to look almost like the graph in Fig. 2.

If this picture portrayed a universal pattern, there would be no need to pursue the topic of this paper. But, in the first place, the picture represents a vulgarization of the original Rokkan-Lipset hypothesis. These authors were considerably more subtle in their analysis than the oft-quoted phrase about the frozen party alternatives would suggest (Lipset & Rokkan 1967, 50^—56). Second, this picture was drawn 15 years ago, and some of the conditions of the unaffected survival of old parties, as well as the breakthrough of new party alternatives, may have changed since then. Third, as we saw in the opening paragraphs of this article, it is an empirical fad: that party alternatives have not been constant in Western Europe over the last decades. Even if some major parties in the foreseeable future will preserve their strength, and their Ls-curves accordingly will look like the model curve of Fig. 2, we will proceed on the assumption that this pattern is not necessarily the typical pattern of party lifespan.

The first important dimension of a Ls-curve is its modality. The ragged short-term profile of curves of party strength tends to blur the fact that most parties will reach only one culmination point in their lifespan. Using the cruder measuring tools applied here, we may say that parties will normally only pass the thresholds up-bound once in their lifespan. Thus the typical Ls-curve is a unimodal curve.

But whatever the most typical pattern might be, it can be argued that the most interesting point to be raised by means of the notion of modality is that some parties do have bimodal or even multimodal Ls-curves. Thus in some cases parties will cross the various thresholds several times. The most common fate in this respect is a temporal drop of a party below the threshold of relevance. Quite a few European parties have experienced this situation. More dramatic — and rare — reversible fall-backs are those that involve a temporary ousting of parliament. As an example the Ls-curve of the Danish Communist Party may serve, see Fig. 3.

The Danish Communists participated with candidates in elections from

1920 on. The threshold of representation was passed in 1932, and the culmination of the party in terms of electoral strength was reached in the election of 1945, when the party again was made legal after the end of German occupation. After a steady decline which accelerated during the late 1950'5, the party was ousted from the Folketing in 1960. The next 13 years the Communists participated in elections without ever reaching the 2% of the vote required to gain representation. In the 'earthquake-election' of 1973, the party had its come-back in legislative politics. This second period lasted 6 years, until the election of 1979, in which the party again obtained less than the required minimum of votes.¹¹

This example brings to the attention that some party lifespans deviate from the typical pattern.¹² It is well known that much can be learned from the study of deviant cases. We may therefore suggest that the recuperating party deserves more attention, since almost nothing is known about the conditions and probabilities of recuperation of those parties which have dropped below one of the crucial thresholds.

The second dimension of the Ls-curve is its dispersion. The lifespan of
parties may range over a few election periods, but it may as well range over
decades or even centuries.

Too little is known about the variations on this dimension. When an organization is considered a party as soon — and as long — as it has the intention to participate in elections, the average duration of party lifespans will of course be somewhat higher than if a 'pure' electoral definition of the party concept is being used. But even if a very crude definition is applied, as e.g. that of Kenneth Janda, when he distributed parties in his global sample according to their age in 1979 or at termination, it comes out clearly that parties vary considerably with regard to the number of years they have been operating, see Fig. 4.

Assuming that most Ls-curves will belong to the unimodal family, the third relevant dimension that characterizes lifespans will be the flatness of the Ls-curve. Statisticians use the term kurtosis for this property, distinguishing between platykurtic, mesokurtic, and leptokurtic patterns in accordance with the curve's similarity to the normal distribution (see e.g. Croxton & Cowden 1962, 232—35). The statistical analogy ought not to be drawn too far,¹³ but it may be useful to make a distinction between two polar categories of parties:

— parties with an extended lifespan, and which in a shorter time perspective appear
fairly stable and/or depicting moderate upward/downward trends;
— parties, whose lifespans stretch over only a short period, in which the party appears
on the scene, quickly reaches its apex, and subsequently disappears.l4

These are polar types, and most party lifespans will fall somewhere in between.
One should also note that these categories are only meaningful and
precise to the extent that a periodization has already been defined, and they

are only useful for comparative purposes, if this periodization is derived
from theoretical concerns and from a previously established knowledge of
the dispersion patterns.

The fourth, and probably the most interesting aspect of the Ls-curve, is its skewness, or the degree of symmetry of the curve. Abstracting from thresholds, dispersion, and flatness and speaking only about unimodal Lscurves, we may distinguish between three distinct categories of lifespans:

— the suddenly appearing, but gradually disappearing party (positively skewed Lscurve);

— the gradually appearing, but suddenly disappearing party (negatively skewed Lscurve);

— parties with symmetrical lifespans.

In conclusion, it has been argued in this section of the article that the Lscurve of any party can be interpreted in terms of four statistics, i.e. modality, dispersion(range), flatness (kurtosis), and skewness. By means of these dimensions it is possible to describe the lifespan of any party, whatever its special historical pattern, from major parties with a history stretching over more than a century, to minor parties which did only exist for a few years.

Description is, however, only a beginning. More important is the heuristic value of these concepts for the generation of questions and problems, and the eventual formulation of hypotheses. Immediately four sets of questions come to mind:

1. What is the probability that a new party will survive for an extended period of time — or disappear within a few years? — Which are the factors that account for long or short life? Which are the institutional conditions that support sustained life for a new party?

2. What is the probability that a party, which at one point in time has dropped below
one of the crucial thresholds of relevance, representation, or authorization will recuperate?
Under what conditions is recuperation likely?

3. What is the probability that a newly formed party will pass one, two, or three thresholds? Which are the factors that account for a given party gaining representation or joining the crucial category of relevant parties? Which are the conditions that facilitate or create obstacles for sustained growth?

4. What is the probability that a new party will leap to representation or relevance? Is it possible to pinpoint special factors that account for a sudden rise to prominence of a new party? Which are the institutional conditions that support such dramatic growth?

These are only some of the questions that will have to be answered if we want to understand the conditions of persistence and change in a party system. No inventory of hypotheses and generalizations exists. Even a systematic mapping of the occurrence of various types of minor parties is lacking.

---

NOTES

1 Since Janda only included a party in his universe if the party won at least five percent of the seats in the lower house of the national legislature in two or more successive elections (Janda 1970, 84), it is evident that his figures underestimate the actual number of new parties.

2 The concept of party lifespan was introduced somewhat earlier by Rawson (1969). However, his ideas were not utilized by students of party systems.

3 A good bibliography on this topic is Lars Svåsand (1977). But see also La Palombara & Weiner (1966).

4 On this conceptual topic, see the discussion in Sartori (1976, 58_—64).

5 In Pedersen (1979), a conservative estimate of the number of European parties since 1945 was set to a little less than 200. Janda (1970, 84) argues that his narrow definition 'reduces the number of parties to a more manageable level of hundreds, rather than thousands of units to study'.

6 In some countries the threshold of authorization is linked to the nomination procedure, whereas in other countries registration of party and nomination of the party's candidates are regulated by two distinct sets of rules.

7 The concept of a threshold of representation was introduced by Stein Rokkan (e.g. Rokkan 1970, 79). Rokkan did, however, use the concept in a different, if related, context.

8 The German Clause is more complicated than described here, see e.g. Hand et al (1979, 72—73).

9 It is difficult to grasp the political balance in Danish politics since the 1973 election without considering the precarious situation of the three small parties (Radical Liberals, Center Democrats, and Christian People's Party). These parties, and especially the latter two, are in danger of extinction in the next election; they tend to behave as it they are afraid of such extinction.

10 On curve shape analysis, see e.g. Galtung (1967, 231^—39).

11 The fate of the Danish Communist Party presents a tricky problem, since it can be argued that the relevance of the party is not at all related to the numerical size of the party, nor to the actual representation of the party.

12 Another interesting example is the Danish Justice Party which has been represented in the Folketing 1926—60, 1973—75, 1977—81.

13 The various dimensions are usually applied to static distributions, not to time series of a given variable.

14 This formulation may some as a working definition of the concept of 'flash party', a phenoman as which it often spoken about, but which hos not yet been given an adequate analytic treatment in the party literature.

REFERENCES

Borre, O. 1980 'Electoral Instability in Four Nordic Countries 1950—77', Comparative Polical
Studies 13, 141—71.

Croxton, F. E. and Dudley J. C. 1955. Applied General Statistics, London: Sir Isaac Pitman
& Sons Ltd.

Daalder, H. 1966. 'De kleine politieke partijen — een voorlopige poging tot inventarisatie',
Acta Politico I, 172—96.

Dill, W. R. 1965. 'Business Organizations', pp. 1071—1114 in March, James G. (ed.) Handbook
of Organizations, Chicago: RandMcNally.

Downs, A. 1957. An Economic Theory of Democracy, New York: Harper and Row.

Fisher, S. L. 1974. The Minor Parties of the Federal Republic of Germany: Towards a Comparative
Theory of Minor Parties, The Hague: Martinus Nijhoof.

Fisher, S. L. 1980. 'The "Decline" of Parties' Thesis and the Role of Minor Parties', pp. 609-13
in Merkl, Peter H. (ed.) Western European Party Systems, New York: The Free Press.

Galtung, J. 1967. Theory and Methods of Social Research. Oslo: Universitetsforlaget.

Hirschman, A. O. 1970. Exit, Voice, and Loyalty, Cambridge, Mass.: Harvard University
Press.

Janda, K. 1970. 'A Conceptual Framework for the Comparative Analysis of Political Parties',
Sage Professional Papers, Comparative Politics Series No. 01—002.

Janda, K. 1980. Political Parties: A Cross-National Survey, New York: MacMillan.

LaPalombara, J. and Weiner, M. (eds.) 1966. Political Parties and Political Development,
Princeton: Princeton University Press.

Lipset, S. M. and Rokkan, S. (1967). "Cleavage Structures, party Systems, and Voter Alignments:
An Introduction", pp. 1—641—64 in Lipset S. M. and Rokkan, S. (eds.) Party Systems
and Voter Alignments, New York: The Free Press.

Mackie, T. T. and Rose, R. (eds.) 1974. The International Almanac of Electoral History,
London: MacMillan.

Mayer, L. 1980. 'A Note on the Aggregation of Party Systems', pp. 515—52 in Merki, Peter
H. (ed). Western European Party Systems, New York: The Free Press.

Merkl, P. H. 1980. 'Introduction: The Study of Party Systems', pp. I—l3 in Merkl, Peter H.
(ed.) Western European Party Systems, New York: The Free Press.

Pedersen, M. N. 1979. 'The Dynamics of European Party Systems: Changing Patterns of
Electoral Volatility', European Journal of Political Research 7, 1—26.1—26.

Pedersen, M. N. 1982. 'Changing Patterns of Electoral Volatility in European Party Systems
1948—77: Explorations in Explanation', In Daalder, Hans and Peter Mair, (eds.) Working
Papers on Western European Party Systems, Beverly Hills: Sage Publications.

Rawson, D. W. 1970. 'The Life-Span of Labour Parties', Political Studies, 17, 313—33.

Sartori, G. 1976. Parties and Party Systems: A Framework for Analysis, vol. I, Cambridge:
Cambridge University Press.

Sjöblom, G. 1968. Party Strategies in a Multiparty System, Lund: Studentlitteratur.

Smith, G. 1978. 'Trends in Western European Party Systems? Parliamentary Affairs XXI,
37—51.

Smith, G. 1980. Politics in Western Europe, 3rd.ed. London: Heinemann Educational Books.

S våsand, L. 1977. 'Fremveksten av politiske Partier: en bibliografisk översikt', Statsvetenskaplig
Tidsskrift 80, 293—301.

Wolinetz, S. B. 1979. 'The Transformation of Western European Party Systems Revisited',
West European Politics 2, 4—28.