![]()
1Party membership as linkage
Knut Heidar and Bram Wauters
In recent years, parties and party organizations have been put under pressure: party membership is declining, voters are more volatile and the party leadership is increasingly taking their cues from public (and social) media debates (e.g. Dalton and Weldon 2005; Drummond 2006; van Biezen and Poguntke 2014). Debates on the decline of mass parties, the rise of âcartel partiesâ and the changing types of party affiliation are indicative of this change (Katz and Mair 1995; Scarrow 2015). In this book, we ask whether the linkage mechanism as offered by the party organizations today â through its members â is as weak as suggested by recent literature. We investigate whether party organizations in Western democracies are (still) providing linkage between voters and political elites. Party members are supposed to channel voter opinions into political institutions offering a broader and more continuous policy base than the link provided by elections alone. But the question arises to what extent they fulfil this function and to what extent all layers of society are included in this process. In other words, we will dig deeper into the question of whether party membership is a way to connect diverse groups in society to politics and government.
The decline in membership figures has been extensively documented in several studies (e.g. van Biezen et al. 2012). But one aspect that has been underexposed in this kind of research is to what extent party membership decline has also affected partiesâ representative capacity (for an exception, see e.g. Scarrow and Gezgor 2010). One would expect party members and parties to become less representative of the population at large with low membership ratio and/or declining bases of members over time. Party members would to a lesser degree share the characteristics of those they intend to represent. In other words, the expectation is that the extent to which women, blue-collar workers, young people, ethnic minorities, etc. are present within the party will have declined. More generally, a crisis of representativeness is â as indicated by the political participation literature (e.g. Hibbing and Theiss-Morse 2002; Marien et al. 2010) â intrinsic to all high-intensity participation modes (including party membership). However, the few studies that have been published question the general assumption of an automatic link between declining party membership figures and declining representativeness (Scarrow and Gezgor 2010; Allern et al. 2016). Therefore, this volume aims to give more insights in the extent to which parties represent voters. At the same time, we bridge two research fields that are usually not in conversation with each other, the literature on party membership decline on the one hand and the literature on group representation on the other.
Representativeness is a central issue in discussions on how parties contribute to democracy. The macro-debates on party democracy look both at the governability of democratic systems (Schattschneider 1942; Cox and McCubbins 1993) and at the democratic qualities of party systems (Duverger 1954/1972; Sartori 1976). A weakening of the linkage function carries a threat to party-based democracy in general if parties rule more than they represent (Mair 2013). This development could have far-reaching consequences for the political system in terms of legitimacy as weaker links with society could lead to less-informed and less-accepted government decisions. And it could be a danger to parties themselves, who risk being out of tune with the opinion and visions of the population, resulting in unpopular decisions, electoral defeats and being displaced by new parties.
Our main point of interest is descriptive representativeness (i.e. the extent to which party members mirror (party) voters on a number of relevant characteristics), not representativeness in general. The importance of party member congruence rests on the claim that descriptive representation has a number of consequences, more in particular for party policies, candidate selection and democratic legitimacy (Kittilson 2013). Factors generating representative party membership are hardly touched upon in the literature, but is seems evident that the number of party members is a central factor. It is not possible for a one-man party to be representative of his/her voters, and it is a valid assumption that the higher the member/voter ratio, the stronger the likelihood of a representative membership.
Therefore, we will focus on three questions in this volume:
1 To what extent is the party membership representative of â or congruent with â the party voters in terms of descriptive background?
2 Is there (following the general membership decline) a decline in memberâvoter congruence over time? And do parties with declining or low memberâ voter ratios show a lower memberâvoter congruence than parties with stable or high memberâvoter ratios?
3 What are the consequences of low descriptive representativeness for substantive representation, candidate selection, political trust and legitimacy?
These questions will be answered in eight country chapters. The chapters include parties from countries with high and low levels of party membership, as well as parties that have experienced membership decline and rise. The countries are all democracies but they differ in their constitutional arrangements and in their party systems. Mapping the descriptive congruence between members and party voters is the empirical core of the book. We aim to test the universality of the claim that low levels of party membership make for low representative capacity by looking at very diverse countries. The consequences that are investigated differ from one chapter to another (depending on data availability). In some chapters, we study the impact of differing degrees of descriptive congruence on who is actually selected (or not selected) as candidate for political office, while in other chapters the focus is on the impact on substantive representation (the partyâs policies) and/or on levels of legitimacy/political trust in political institutions. Given the data availability, however, the discussions on causes and consequences will mostly be suggestive as only a few alternative causal relations can be analysed in-depth.
In the following, we will take stock of the general debate and spell out the analytical and methodological issues faced in this research. We start by discussing the relevance of party members in contemporary party organizations. Next, we discuss the importance of descriptive representativeness of political parties. We also summarize some existing empirical studies of memberâvoter congruence in Western European parties. We end with an outline of the chapters that follow.
Party members and their relevance
In the European tradition, the notion of the âmass partyâ has always had a central place. As presented by Maurice Duverger in his classic study, the âmass partyâ is a political instrument at the service of the voters (Duverger 1954/1972). Parties organize the political will of engaged citizens, alias members, and bring forward policies and candidates at elections. Parties provide linkages between voters and elites, and the members function as intermediaries in this respect. Parties serve democracy by supplementing elections in bringing voter preferences into the institutions. In this way, party members add to electoral democracy by bringing the voters into the party chain of government. Thus, parties are there for the members.
In the literature, an âAtlantic divideâ is often pointed to in views on what role parties play in democracies (Wright 1971). In the European tradition, parties are considered to be supplementary channels of voter influence. Voters cannot express their wishes on all issues and at all times, and party members must therefore do it for them in the meantime. The main US parties do not have members in the European meaning, i.e. dues-paying participants allowed to have a say in internal decision-making on candidates and policies. US parties are only vehicles for the candidatesâ electoral ambitions: registered voters have the option to take part in party primaries and some political activists also engage in a candidateâs campaign organization. Therefore, in the US, parties are not there for the members, but âmembersâ are there for the parties. This does not mean that the US polity is not democratic, but it is not a party democracy. US voters also wire their Congress representative (today probably by email or social media) but this is a linkage outside the organized party context. Also other supplementary âcivil society linkagesâ through individual actions, voluntary organizations and lobby activities do not include an organizational linkage through party member organizations.
Clearly, much has changed since the early post-Second World War resurgence of the mass parties, also in Europe (Katz and Mair 1995; Allern et al. 2016). The way West European parties are organized today, however, as well as how they are conceived as democratic instruments, still owes much to the mass parties evolving during the late 1940s and the 1950s. Member representativeness does, however, not have the same importance or meaning in âcandidate-drivenâ parties. Such candidate parties are also present in Europe; the Dutch Freedom party is one example. This âpartyâ has only one member, its leader Geert Wilders (van der Pas et al. 2013; Mazzoleni and Voerman 2017). Other parties may also be more candidate driven (or leader-run) than most parties used to be in the past. Parties created around figures such as Macron (in France) and Berlusconi and Grillo (in Italy) provide cases in point here. These new kinds of parties with no members or with only a limited role for members have been labelled âbusiness firm partiesâ or âpersonalized partiesâ (Hopkin and Paolucci 1999; Mazzoleni and Voerman 2017). Also in more established parties, we can refer to Roberto Michels with his âiron law of oligarchyâ, who more than a century ago drew attention to the fact that parties over time move along a continuum from the member-centred to the elite-centred types (Michels 1915).
Yet another threat to party membership is that some parties have introduced options for âsofterâ types of affiliation compared to full membership (Scarrow 2015). Voters may register as âsupportersâ and be allowed to vote in internal leadership contests and/or in the selection of party candidates for (the highest) public office (Bernardi et al. 2017). This reduces the relative value of full membership. Non-member affiliates can participate in many of the same tasks as full members â and arguably in the most important ones. Moreover, the argument in favour of these lighter forms of affiliation is that it improves the link with the voters. Knowing the numbers and the representativeness of âproperâ members then becomes only one side of general party representativeness. However, this would be a relevant point only in countries where such âmulti-speed partiesâ actually have some relevance, like the UK and possibly France and Italy, but overall these types of membership options â and in particular the extent to which they are actually practised â is far too limited to make descriptions of (full) member representativeness redundant. Moreover, it seems that many parties approach these types of non-member affiliations as a first step towards more permanent affiliations such as party membership.
Our argument here is that the practice in most European parties is still mostly member-based, both in terms of organizational structures and with an official party ethos emphasizing membership democracy. Many parties prescribe internal party democracy in their statutes and as part of their ethos, with party members as their main demos. In these parties, the leadership cannot politically afford to disregard the members, and even if the party is prone to oligarchic tendencies, the votersâ trust may still depend on membersâ representativeness. If parties say they practise member democracy, the composition of the membership will be relevant to the voters (as well as to the media). The influence and relevance of party members has even increased in the last few decades, testified by members having a larger say in the selection of the party leader and the candidate lists for parliamentary elections, and in the policy of the party (Loxbo 2013; Pilet and Cross 2014). This relevance of party members, however, does not exclude the possibility that middle-level elites, e.g. conference delegates, also may generate congruent groups compared to voters.
We do know that certain groups are more likely than others to be politically active: men, the middle-aged, the highly educated, etc. (Whiteley et al. 1994). In the old Liberal Party in Norway, the saying was that it was a party of school-teachers â even if the voters were mostly peasants, fishermen, craftsmen and workers (Heidar 2001). Political resources are unevenly distributed at the individual level and this is likely to create new representative imbalances (or corrections to imbalances at a lower level in the party). Similarly, parties are different in origin and organizational structure. Social-democratic or Christian-democratic parties that were founded by trade unions and religious organizations will most likely recruit members disproportionally from these voter segments â even if (over time) they attract a broader voter segment. Therefore, one would be well advised to search for social, organizational and institutional differences to explain representative variance. In the broad sense, however, this general approach to explain differences and changes in partyâvoter congruence does not define our task here, even if we expect declining memberships to create stronger imbalances in congruence.
Is descriptive representation in parties important?
Our main point of interest is the representativeness of party members. Before indicating how representativeness could be important in relation to party membership, we first have to define this concept. In order to do so, we start with the classic work of Hanna Pitkin (1967), who defines political representation as âmaking present in some sense of something which is nevertheless not present literally or in factâ. Within this broad definition, four different (but interconnected) dimensions can be distinguished. Formal representation focuses on the rules and procedures through which representatives are authorized to act as representatives (mostly by means of free and fair elections). Descriptive representation refers to the extent to which the socio-demographic characteristics of the representatives correspond to those of the people they represent. Substantive representation, or responsiveness, is defined as acting in the interests of the represented in a manner responsive to them. Finally, symbolic representation refers to feelings of being fairly and effectively represented.
As already indicated, the main focus of this book lies on descriptive representation. If the representatives and the represented share some personal characteristics (e.g. gender), it can be said that the former represents the latter in a descriptive manner. There exists a very rich literature on the importance of descriptive representatives in parliament (see e.g. Phillips 1995; Dovi 2002; Wängnerud 2009; Saalfeld and Bischof 2013). We will argue here that descriptive representativeness is also of utmost importance when it concerns party members since it has implications for party policies, candidate selection and democratic legitimacy (Kittilson 2013).
Parties have, first of all, the power to include and prioritize certain issues of importance to societal groups (or marginalize and even igno...
