The consequences of membership incentives: Do greater political benefits attract different kinds of members?

Abstract

In recent decades, parties in many parliamentary democracies have radically reshaped what it means to be a party member, making it easier and cheaper to join, and giving members greater direct say over party decisions. This article explores some implications of such changes, asking whether membership costs and benefits influence which supporters take the step of joining their party. In particular, it considers the impact of net membership benefits on membership demographics and on members’ ideology. The investigation examines patterns of party membership in 10 parliamentary democracies, using opinion data from the European Social Survey and data on party rules from the Political Party Database project. Our analysis shows that party supporters are more sensitive to political benefits than to financial costs, especially in terms of the ideological incongruence of who joins. As a result, parties offering higher benefits to their members have lower ideological and demographic disparities between members and other party supporters. This is a positive finding for party-based representation, in that it suggests that trends toward more inclusive decision-making processes have the potential to produce parties with memberships that are more substantively and more descriptively representative of their supporters.

Keywords

Intraparty democracy party members political parties

Introduction

In recent decades, parties in many parliamentary democracies have reshaped what it means to be a party member. To begin with, some parties have substantially increased the direct integration of members into intraparty governance, such as selecting the party leader or candidates, approving coalition agreements, or deciding party policy on controversial issues (Cross and Pilet, 2015; Gauja, 2013; Sandri et al., 2015; Scarrow, 2015). At the same time, some parties began making it much easier for their supporters to join, by reducing the required dues payments and by lowering the procedural barriers to entry (Faucher, 2015; Gauja, 2015; Scarrow, 2015). In many cases, party membership can now be purchased on the Internet in a single transaction, similar to other consumer goods. Partly as a result of these changes, contemporary political parties vary greatly in how they organize their membership, particularly in terms of the costs and benefits of affiliating as a full member.

This leads us to our question: How, if at all, do variations in the construction of party membership affect the representativeness of those who choose to join? Previous research has consistently shown that party members in parliamentary democracies are unrepresentative of party supporters and of voters in general. They are more likely to be male, older, wealthier, better educated, religiously observant, and members of a trade union (most recently shown in the party member surveys reported in van Haute and Gauja, 2015). Thus, party members as a group have never been descriptively representative of the general voting population. Party members may be unrepresentative in another sense as well: in terms of their political preferences. The classic literature on parliamentary parties articulated strong expectations that active party members would be more politically extreme than other party supporters and voters (most famously, May’s 1973 “Special Law of Curvilinear Disparity,” but also McKenzie, 1958 and others). Empirical tests have found limited or no support for the claim that either party members in general, or active party members in specific, are more politically extreme than other supporters or voters (Kitschelt, 1989; Narud and Skare, 1999; Norris, 1995; van Holsteyn et al., 2017). Yet these expectations persist, in part because political concerns seem to explain why many supporters chose to bear the costs of party membership or activism (for instance, Whiteley and Seyd, 2002).

The extent to which party members are representative of other party supporters has assumed greater importance as more parties give decision-making powers to individual members. Parties which have adopted more inclusive decision-making procedures often herald these as democratizing reforms, ones which will help them to reconnect with citizens and to better represent the profile and preferences of their supporters. Yet, institutional changes are not necessarily independent from the profile of party membership. Parties’ democratizing reforms appear to be related to the propensity of party supporters to become members, and to levels of party activism among party members (Kosiara-Pedersen et al., 2017). Such relationships may reflect the fact that differences in party member rights alter the cost/benefit equations of party membership, and of membership compared to other affiliation options. In turn, these differences could also affect the types of supporters who choose to become members, possibly changing the extent to which party members are (or are not) a representative subset of other party supporters. While the latter effects seem logically possible, so far we have little empirical evidence of whether relatively small differences in the value of the party membership “product” affect the demographic or political representativeness of those who chose to join.

This, then, is the focus of this article: we ask whether variations in the costs and benefits of party membership systematically affect who joins, making parties’ memberships more or less representative of their wider political base. In investigating representativeness, we are interested in both substantive representation (ideological congruence) and descriptive representation (demographic congruence). We answer these questions using individual-level data from the 2008 and 2010 European Social Survey (ESS) on party supporters and party members across 57 parties in 10 parliamentary democracies. We combine this with party-level data on membership procedures and members’ political rights, taken from the Political Party Database (PPDB) project (see Poguntke et al., 2016). We start by explaining what we mean by cost, benefits, and net benefits of party membership and show how we measure these concepts. We then proceed to test whether these aspects of membership construction affect who participates.

The impact of membership costs and benefits

Wherever parties offer formal party membership, politically interested citizens decide whether or not to join their favored party. Individual resources seem to play a big role in membership decisions, as with decisions about other types of political participation. On average, those who join political parties tend to be better educated and more affluent than other citizens, and they may have more leisure time, because they tend to be in the later part of their working-age years or recently retired. These patterns have held up over many countries and over many years (see, for instance, Parry et al., 1992; Scarrow and Gezgor, 2010; Verba et al., 1978; Widfeldt, 1995).

Yet, individual resources alone do not explain why some people become party members. In fact, in most European democracies, party membership is well under 5% of the voting-age population (van Haute et al., 2017), meaning that most people who are interested enough to vote do not join, whatever their resources. However, even though national levels of party membership are not particularly high, there is still a significant cross-party variation in enrollment levels: Within each country, some parties are much more successful than others in attracting members from their support base. In addition, individual parties can and do experience fluctuations in their membership numbers, changes that often exceed the fluctuations in their political support. These differences suggest that party-level factors may also help explain individual enrollment decisions (Kölln, 2016; Paulis et al., 2017).

The interaction of party-level factors and individual decisions comes into focus if we view party membership through a cost–benefit lens. This perspective treats membership as a consumption good and assumes that potential members are price sensitive. This does not mean that the costs or benefits are exclusively or even primarily economic. Potential members may value a wide range of benefits, including solidary and purposive rewards. It also does not mean that individuals will join a party merely because its membership is the cheapest on the market. Rather, it means that party supporters are more likely to take the additional step of acquiring membership in their preferred party when the expected benefits outweigh the costs.

In this article, we focus on the costs and benefits that parties are able to control rather than those which relate to individual circumstances or which are regulated by national party laws. These include the financial costs of annual dues.¹ Many parties have multiple dues rates, with reduced rates for those who are likely to have fewer resources (students, pensioners, unemployed). Some parties have tiered rates, with a minimum cost that is above the “reduced” rate and with higher income earners expected to pay higher amounts. In parties in parliamentary democracies, these rates tend to be quite low when compared with other uses of disposable income (cell phone plans, gym membership, a restaurant dinner). Nevertheless, it may be the case that even modest rates deter new members. Following this logic, some parties have recently experimented with special low rates for new members, sometimes labeled as “trial membership,” explicitly urging supporters to experiment with party membership at a bargain rate.

What do supporters get when they join their preferred party? Parties may place a variety of collective and selective incentives on the benefits side of the ledger (Whiteley and Seyd, 2002). Most tangibly, patronage parties offer members the possibility of receiving material selective incentives, including privileged access to state resources (e.g. school places or rent-subsidized accommodations), or gifts provided by politicians from private funds. Pure patronage parties are rare in developed democracies nowadays, both because of the relative affluence of most citizens and because of legal changes that insulate state resources from partisan distribution mechanisms. As a result, it is difficult for parties to attract large numbers of members by offering meaningful material selective incentives, although many parties have a subset of members who are motivated by personal career considerations (Kopecký et al., 2012; van Haute and Gauja, 2015). Contemporary parties may provide social benefits for those who join, particularly for those who get involved in party activities, but in doing so they compete with many other types of organizations. In contrast, the most distinctive and perhaps the most attractive incentive that contemporary parties can offer to potential members is the opportunity to make a difference in politics, that is, collective benefits.

In their general incentives model of partisan participation, Whiteley and Seyd (2002) argued just this, finding that “making a difference” is a major consideration for most who join. Subsequent studies of party membership have found strong support for the idea that political considerations motivate individual decisions about joining and remaining in a party (for instance, chapters in van Haute and Gauja, 2015; also Denmark, Pedersen et al., 2004; Germany, Laux, 2011; Ireland, Gallagher and Marsh, 2004]; Norway, Heidar and Saglie, 2003). These incentives may be in the form of opportunities to defeat an opponent or opportunities to enact favored policies. However, some of the incentives over which parties have greatest control involve intraparty political processes. In this vein, one of the most important political benefits that parties can offer to potential members is the right to directly influence party decisions, such as choosing the party leader or deciding the party’s position on policy issues. And indeed, in recent decades, an increasing number of political parties have been giving members a direct vote in party decisions. In at least some cases, parties have done so in direct response to declining membership numbers. Examples of membership surges spurred by leadership or candidate selection contests support the conclusion that supporters are sensitive to these benefits (van Haute and Gauja, 2015). Even so, parties still differ widely in the extent to which they actually offer such political benefits (Cross and Pilet, 2015; Sandri et al., 2015).

If it is true that potential members are sensitive to the costs and benefits of joining, how parties construct their membership rules should affect not only how many supporters acquire membership but also some of the characteristics of those who join. Elsewhere we have found evidence of this mechanism (Kosiara-Pedersen et al., 2017), showing that membership costs and benefits affect the likelihood that party members will be active within their parties. On average, members of parties that are relatively difficult to join are more likely to participate in partisan activities. This is particularly true if parties offer affiliation options that provide a lower-cost alternative to traditional membership. In those cases, those who are willing to pay the higher costs of traditional membership are more likely to engage in partisan activity. These findings support the argument that potential party members are attentive to membership costs and benefits, and that such incentives and barriers act as a sieve, shaping the composition of party membership in politically relevant ways. In this article, we further explore this proposition, looking more closely at how the formal construction of party membership—the costs and the benefits—affects the demography and political views of those who opt to acquire party membership.

We examine whether parties systematically affect the political complexion of their memberships by changing the cost and benefits, or the net benefits, of joining (our measurement strategy is explained below). Our two hypotheses about the impact of benefits on party membership composition start from the assumption that increases in benefits should make party membership attractive to a wider swathe of party supporters. Conversely, when membership is relatively expensive, only the most politically committed should choose to acquire it; those with fewer individual resources are likely to be deterred by the cost. There may be a net effect of costs and benefits, such that when benefits rise relative to cost, party membership should become appealing to more supporters. This effect may be unequal across various party supporters, depending on how much they value collective (political) incentives compared to other types of membership payoffs. For instance, it might be that political incentives hold greater interest for those who come from groups traditionally underrepresented within a specific party’s membership; these individuals may place comparatively less value on the solidary benefits of joining a local club whose members do not look like them, and they may place greater value on unmediated opportunities to influence party policies (because they distrust the traditional mediators). This is an example of one possible mechanism whereby an increase in net political benefits could make party membership more demographically and more substantively representative of a party’s general support base.

The last part of the preceding sentence is important because we argue that these effects are best measured by comparing the characteristics of each party’s membership relative to the universe of that party’s supporters, not relative to the entire population. Our question, then, is whether the costs and benefits of party membership at the party level affect a party’s representativeness, measured as ideological and demographic congruence between party members and other party supporters. Our hypotheses about the impact of costs and benefits can be stated as follows:

Benefits

Substantive representation (H1a): Party members will be more ideologically congruent with other party supporters when benefits are higher.

Descriptive representation (H2a): Party memberships will be more demographically congruent with other party supporters when benefits are higher.

Financial costs

Substantive representation (H1b): Party memberships will be less ideologically congruent with other party supporters when financial costs are higher.

Descriptive representation (H2b): Party memberships will be less demographically congruent with other party supporters when financial costs are higher.

Net benefits

Substantive representation (H1c): Party memberships will be more ideologically congruent with other party supporters when net benefits are higher.

Descriptive representation (H2c): Party memberships will be more demographically congruent with other party supporters when net benefits are higher.

Each of these hypotheses has important implications for democratic representation and the political system. If supported, they imply that party rules go some way to explaining the nature of the organizationally mediated linkage between party elites and party supporters.

Data and operationalization

We test these hypotheses using data from two sources. First, the PPDB round 1a data contains information about how parties organize in 19 parliamentary democracies.² This data release includes membership costs and benefits information for 122 parties, mostly reflecting rules which were in effect ca. 2011–2012. In order to assess the characteristics of those who join, we use individual-level data from the ESS rounds 4 and 5 (collected ca. 2008 and ca. 2010). These two rounds of the ESS are the most recent rounds of this particular survey which asked about party membership. We pool data from two rounds in order to increase the number of respondents who are party members. These surveys were collected close to the time for which we have PPDB data on membership costs and benefits. We expect that the relative costs and benefits will not have changed a great deal between 2008 and when the PPDB data were collected (ca. 2011–2012). Changes in leadership selection rules, for example, are relatively rare, as Cross and Pilet (2015) have shown in their longitudinal study of leadership selection rules (including more than 100 parties in 14 countries over 50 years). Chiru et al. (2015) have similarly shown that over the past 50 years, parties have on average reformed their statutes once, mainly in the direction of more inclusive processes or more rights to members.

We look only at parties in countries included in the PPDB for which we have full information on membership costs and benefits and at parties in countries for which ESS data are available in the combined 2008–2010 surveys. We include only parties with at least 28 respondents self-describing as party supporters.³ With these restrictions, the final sample contains 57 parties in 10 countries. These parties span the range of ideological party families, though they are weighted toward Social Democrats and Christian Democrats/Conservatives—the parties that traditionally have had the largest electorates and hence more declared supporters in our ESS samples (see Table 1).⁴

Table 1.

List of parties included in the analysis, by party family.

Party family	No. of parties	Percent
Christian democrats/conservatives	13	23.2
Social democrats	13	23.2
Liberals	14	25.0
Greens	7	12.5
Left socialists	6	10.7
Right wing (populist)	1	1.8
Far right (extreme right)	2	3.6
Total	56	100

Note: In addition, one party is unclassified in the PPDB data (the Centre Party in Norway). PPDB: Political Party Database.

Because our main question is how the variations in the construction of party membership affect the representativeness of those supporters who choose to join, our sample is limited to those who responded that they felt closer to a party and who said which party that was. Our dependent variable is whether a party supporter is also a party member (measured by ESS data).⁵ In our study, party supporters are those survey respondents who said they supported a party and who named the party they supported; party members are those respondents who self-describe as party members.

Our main party-level independent variable is the net benefit of party membership, which is calculated from PPDB data concerning membership costs and benefits. On the cost side, we calculate financial costs as the minimum “normal” dues levels (in other words, not counting possible lower levels for students, the unwaged, etc.). Almost all of the parties in the PPDB data set uniform national minimum dues rates. Most of the rest leave it to regional or local parties to establish their own rates; in these cases, the PPDB data provide the minimum rate set by the parties’ largest regional unit. We exclude cases with missing data. After converting all dues to a standard euro rate, we find a wide variation in the minimum annual dues rates for membership, ranging from a very affordable €5.00 (Christian Democratic and Flemish in Belgium) to €110.5 for the most expensive (People’s Party for Freedom and Democracy in the Netherlands). To take account of the wide cost-of-living disparities in the country sample, we computed a relative measure of “Financial costs” that standardizes these minimum annual dues rates in terms of each country’s average annual wage. This gives a slightly different picture of cost differences, with relative dues rates ranging from 0.01% of the annual average wage (Christian Democratic and Flemish in Belgium and the Liberal People’s Party in Sweden) to 0.29% (Democratic Convergence of Catalonia in Spain). The mean score is 0.08%, with a standard deviation of 0.06%. As an example, a score of 0.03% means that a party’s minimum dues level is equivalent to three one-hundredths of a percentage of the average annual wage of individuals in that country. This standardized measure provides a more realistic representation of how dues costs might affect individual membership decisions.

Similarly, we use the PPDB data to construct an index of membership benefits, looking here at whether party statutes authorize membership ballots for three types of decisions: selecting the party leader, selecting candidates, or adopting the election manifesto. We focus on these decisions because such rights tend to be exercised during intraparty contests, which generate high media attention. High publicity makes it more likely that nonmember supporters are aware of, and might respond to, these rights. These are also substantively important rights and therefore should be especially valued by supporters who are politically engaged. Among our sample of 57 parties, 43% give party members a direct vote on the election of the political leader, while only 17% grant members full rights (input and vote) on the formulation of the election manifesto, and 18% grant members full rights (input and vote) in the candidate selection process. For the purpose of the analysis, these plebiscitary benefits are computed in a single additive scale, the Plebiscitary Benefits scale, which combines the three statutory rights, rescaled to 0 (no rights for members) to 1 (full rights to members). Fourteen parties do not grant their members any of these rights and seven parties grant their members all three membership rights. On average, parties score 0.4 on the Plebiscitary Benefits scale.

Finally, we standardize and combine these scales in order to compute a net benefits score as the benefits minus the financial costs. Because we are standardizing across the values within our data set, the net benefits score tells us only about the relative benefits of joining one party in this group compared to other parties. It is not an absolute score and would need to be recalculated if this kind of analysis were to be repeated with a different group of parties or with these parties at a different time period. However, what it does give us is a relative measure of net benefits, set to a single metric, which allows us to investigate the “price” sensitivity of potential party members across multiple parties. Because our measure is relative, not absolute, and because the real variations are not huge, this is a stringent test, thus increasing our confidence in any findings.

Our main individual-level independent variables are classic demographic and ideological variables, measured using ESS data. We consider the impacts of age, gender, income level, religious observance, and union membership, variables commonly found to matter in resource models of partisan participation. Because education and income are highly correlated, only income is included. To measure ideology, we use the classic left–right self-placement of respondents.

Whereas many previous studies of political participation and party membership have compared the demographic characteristics of participants with those of the general population, we are interested in the distance between the characteristics of the party members and the mean characteristics of all supporters of the respondent’s party. This measurement strategy allows us to assess which demographic factors influence the decision of a supporter to become a party member, independently of the influence of demographics on support for specific parties. We take this approach because we know that support for specific parties is demographically heterogeneous, with support for some parties correlated with certain demographic characteristics, particularly income and religiosity. We want to untangle whether these demographic properties have an impact on the decision to join a party even after taking account of their impact on the propensity to support that party. This operationalization implies that, for instance, an individual’s score for religiosity is not the value on the 7-point scale, but the difference between his/her religiosity score and the mean value for all supporters of his/her preferred party. We include dummy variables for gender and trade union membership.

Similarly, we measure ideological incongruence as the difference between the respondent’s self-placement and the mean supporter on a left–right ideological continuum. Because it is a measure of incongruence and not extremeness, this variable accounts only for distance from the mean, not the direction of the deviation. The larger the value, the greater the distance between the views of party members and the mean views of other party supporters.

In addition to our main independent variables, our model includes country dummy variables to control for the contextual variation in which parties operate within each country, and a dummy variable of the ESS round to account for time variation.

The representativeness of party membership: Who joins?

We begin with a confirmatory analysis to make sure that the ESS data match previous findings about the demographic unrepresentativeness of party membership (Table 2). To do this, we fit a multilevel logit model of party membership with five factors commonly included in resource models of partisan participation: age, income level, religious observance, gender, and union membership. Our data are hierarchical, with 13,741 respondents clustered in 57 parties, and parties clustered in 10 countries.⁶ Therefore, we specify a two-level model with fixed effects by country and survey weights. This allows us to decompose the variance explained by variables at each level, identifying the role that party factors play in the individual decisions.

Table 2.

Which party supporters enroll as members? The impact of individual- and party-level characteristics.

	Model 1	Model 2	Model 3
Individual-level variables
Ideological incongruence	0.295*** (0.0530)	0.292*** (0.0529)	0.292*** (0.0529)
Religiosity	0.171*** (0.0317)	0.165*** (0.0308)	0.165*** (0.0304)
Female	−0.313*** (0.0953)	−0.307*** (0.0944)	−0.307*** (0.0944)
Union membership	0.419*** (0.162)	0.422*** (0.162)	0.423*** (0.162)
Income	0.113*** (0.0221)	0.112*** (0.0221)	0.112*** (0.0221)
Age	0.021*** (0.0065)	0.021*** (0.0065)	0.021*** (0.0065)
Party-level variables
Benefits		−0.642*** (0.205)
Costs		0.257 (0.965)
Net benefits			−0.652*** (0.202)
Intercept	−1.899*** (0.268)	−1.531*** (0.305)	−1.542*** (0.307)
Party-level variance	0.0336 (0.0376)	0.0196 (0.0287)	0.0207 (0.0298)
Country dummy	✓	✓	✓
ESS round dummy	✓	✓	✓
Observations	13,741	13,741	13,741
Number of parties	57	57	57
Number of countries	10	10	10
Log pseudo-likelihood	−2926.466	−2923.968	−2924.038
Akaike information criterion	5888.933	5887.935	5886.076
Bayesian information criterion	6024.439	6038.498	6029.111

Note: Robust standard errors in parentheses. The universe is self-described party supporters and the dependent variable is self-identified party membership. ESS: European Social Survey.

***p < 0.01.

Model 1 tests our demographic and ideological incongruence variables coded as described above. Unsurprisingly, it shows that all the identified individual resource variables are statistically significant and that the effects are in the direction predicted by past research. Party members in these 10 countries are, on average, older, wealthier, more religiously observant, more likely to be union members, and less likely to be female than party supporters. This model also provides support for the idea that party members’ political preferences differ from those of others who support the same party. However, because we measure distance only, not direction, we cannot say that members are necessarily more extreme, as “May’s Law” (1973) asserts, just that their views are different.

Our next models test the additive effect of party-level variables by assessing costs and benefits as separate variables (model 2),⁷ and then by using a single measure of net benefits (model 3). Adding these party-level measures does not change the size of the effect for the demographic or ideological variables as compared with model 1. Most surprisingly, in models 2 and 3, we see an inverse relation between (net) benefits and the willingness of party supporters to join their preferred parties.⁸ We believe that the unexpected direction of the relationship should not be interpreted as meaning that higher membership benefits discourage supporters from becoming members. Instead, it seems more likely that the causality flows in the opposite direction: Parties with smaller or declining memberships may be more likely to have restructured membership in an effort to make it more attractive. Unfortunately, we cannot further analyze the causal theory with our current cross-sectional data. Whatever the explanation, the focus of this article is to determine whether net benefits affect which supporters join, not to determine whether they affect how many join. Models 2 and 3 do not give us an answer about who joins. To understand this relationship, we look at the interaction between membership pricing (benefits, costs, and net benefits) and the demographic and political representativeness of those who join.

Do costs and benefits affect which supporters join?

As a reminder, we expect that increased benefits and reduced cost of membership reduce the marginal effect of individual-level variables, thereby decreasing their impact on the likelihood of a supporter to become a party member. In other words, these should increase the ideological (hypothesis 1) and demographic (hypothesis 2) congruence of party members with other party supporters.

We build on the multilevel structure of the model to identify the expected moderating effect of party characteristics (benefits, costs, and net benefits) on the ideological and demographic congruence on party membership. These models enable us to test the interactive effect of benefits (model 4 = test of hypotheses 1a and 2a), costs (model 5 = test of hypotheses 1b and 2b), and net benefits (model 6 = test of hypotheses 1c and 2c) with the demographic and ideological variables.

We display the results of these calculations in models 4 to 6 in Table 3. We first discuss our results with regard to ideological incongruence and then turn to demographic representativeness.

Table 3.

The interactive impact of individual- and party-level variables on party supporters’ propensity to join.

	Model 4	Model 5	Model 6
Individual-level variables
Ideological incongruence	0.465*** (0.103)	0.284*** (0.103)	0.406*** (0.0869)
Female	−0.476*** (0.127)	−0.261 (0.178)	−0.413*** (0.0989)
Union membership	0.314 (0.228)	0.059 (0.210)	0.422* (0.237)
Income	0.096** (0.0402)	0.039 (0.0429)	0.115*** (0.0363)
Age	0.002 (0.007)	0.028*** (0.009)	0.008* (0.0048)
Religiosity	0.224*** (0.0564)	0.120** (0.0527)	0.209*** (0.0426)
Party-level variables
Benefits	−1.299** (0.626)	−0.662*** (0.203)
Financial cost	0.251 (0.999)	0.787 (2.654)
Net benefits			−1.211* (0.643)
Cross-level interactions	A. Benefits	B. Financial cost	C. Net benefits
(A, B, or C) * Ideological incongruence	−0.481** (0.234)	0.0783 (0.759)	−0.454* (0.234)
__________ * Female	0.474 (0.349)	−0.327 (1.373)	0.434 (0.334)
__________ * Union membership	0.305 (0.417)	3.558 (2.225)	0.005 (0.505)
__________ * Income	0.053 (0.0672)	0.649 (0.446)	0.005 (0.0750)
__________ * Age	0.053** (0.0218)	−0.07 (0.0751)	0.052** (0.0207)
__________ * Religiosity	−0.149 (0.102)	0.388 (0.492)	−0.154* (0.0887)
Intercept	−1.289*** (0.422)	−1.574*** (0.323)	−1.403*** (0.377)
Party-level variance	0.021 (0.0290)	0.023 (0.0290)	0.021 (0.0314)
Country dummy	✓	✓	✓
ESS round dummy	✓	✓	✓
Observations	13,741	13,741	13,741
Number of parties	57	57	57
Number of countries	10	10	10
Log pseudo-likelihood	−2910.025	−2911.743	−2908.53
Akaike information criterion	5872.049	5875.485	5867.059
Bayesian information criterion	6067.781	6071.217	6055.263

Note: In the cross-level interactions, the condition interacting with the demographic factors varies across the model. It is either A (Benefits, Model 4), B (Financial Cost, Model 5) or C (Net Benefits, Model 6). Robust standard errors in parentheses. ESS: European Social Survey.

***p < 0.01; **p < 0.05; *p < 0.1.

To test hypothesis 1a, that higher membership benefits may decrease the gap between the political views of party supporters and members (the variable labeled “ideological incongruence” in our model), we focus in model 4 on the change in the average marginal effects of this variable along the range of benefits. We find that political benefits do indeed reduce the effect of ideological alignment at the upper end of the benefits range: When parties do not offer such benefits, or the benefits are very low, there is a bigger gap between the political views of party members and other supporters (increased ideological incongruence). Figure 1 (top left) illustrates that this gap shrinks and becomes statistically insignificant as benefits grow. In other words, the average marginal effects of ideological incongruence decrease as opportunities for direct political participation increase, with party members becoming politically more like other party supporters. These results provide strong support for the idea that how parties construct the benefits of membership can affect the political complexion of their memberships.

Figure 1.

Interactive effects of benefits and costs on ideological incongruence of party membership. Note: Average marginal effects with 95% confidence intervals.

In model 5, we interact the financial costs of membership with ideological incongruence to test hypothesis 1b that higher financial costs may increase the effect of ideological incongruence on the probability that a party supporter becomes a party member. In contrast to the interactive effect with benefits, the graph at the bottom of Figure 1 shows that the marginal effects of ideological incongruence remain almost constant along the range of financial costs. In other words, these results do not provide support for the idea that differences in membership costs affect the political complexion of parties’ memberships. This could reflect the relatively low costs of party membership across the parties in our sample; we cannot rule out that potential members would be more price sensitive if there were greater variation in the minimum financial costs of joining.

Finally, even though model 5 does not show an effect of cost sensitivity by itself, it could be that membership decisions are affected by costs relative to benefits, in other words, by the net benefits. Model 6 thus presents the interaction between our measurement of net benefits and ideological alignment (hypothesis 1c). This relation is shown in Figure 1 on the top right. The direction and magnitude of the coefficient show that the effects of benefits prevail over the financial costs of joining a party. As net benefits go up, party members are politically more like other party supporters.

We use a similar procedure to determine the effects of costs and benefits on demographic congruence. We thus include interactive terms of benefits (model 4 = hypothesis 2a), costs (model 5 = hypothesis 2b), and net benefits (model 6 = hypothesis 2c) with age, religiosity, union membership, gender, and income. Figures 2, 3, and 4 illustrate and summarize the effect of benefits, costs, and net benefits, respectively, on changes in demographic characteristics of party members.

Figure 2.

Interactive effects of benefits on demographic congruence of party membership. Note: Average marginal effect with 95% confidence intervals.

Figure 3.

Interactive effects of financial costs on demographic congruence of party membership. Note: Average marginal effect with 95% confidence intervals.

Figure 4.

Interactive effects of net benefits on demographic congruence of party membership. Note: Average marginal effect with 95% confidence intervals.

Figure 2 shows that as benefits increase, the demographic incongruence of members diminishes in several key ways, as shown by the slopes converging to 0 as benefits increase to their highest level. This supports our hypothesis 2a. For example, the positive effect of religiosity on the probability of becoming a party member becomes weaker as benefits increase. Moreover, the graph of the average marginal effect for being a female shows that when benefits are higher than 0.6 on a scale from 0 to 1, gender becomes statistically insignificant as a predictor of party membership. In contrast, in the case of income and union membership, we observe that the marginal effects remain almost constant as benefits increase.

Age is the only demographic variable for which we see the inverse of the hypothesized moderating effect of benefits. The graph at the bottom center shows that the effect of age is even greater in parties that offer higher benefits with party membership. Hence, the more benefits parties offer, the more likely they are to attract older members.

Figure 3 shows the interaction of financial costs and demographic characteristics. We find that in some cases higher costs increase the impact of the demographic variables, meaning that party members are less like the pool of party supporters (hypothesis 2b). When costs are highest, the average marginal effect of religiosity on party membership is almost three times greater and statistically significant compared to parties with the lowest costs. Similarly, the average marginal effect of income is almost five times greater in parties with the highest financial costs of membership compared to those at the lowest end. In contrast, and unlike the case of benefits, the wide confidence intervals in Figure 3 for gender, age, and union membership indicate that there is no evidence of cost sensitivity for these demographic categories.

As a last step, we estimate the interactive terms of demographic variables with net benefits, using the measurements from model 6. Figure 4 shows that as the net benefits increase, some demographic variables become weaker predictors of party membership (hypothesis 1c), with the exception of age and union membership. The effect of union membership as a predictor of membership does not change across the range of net benefits, while the effect of age becomes stronger as the net benefits increase. Hence, the results for net benefits are similar to the ones for benefits alone shown in Figure 2.

To summarize, our findings support most of our hypotheses about the price and benefit sensitivity of potential party members, with the effects in the directions that we posit. First, benefits offered by parties seem to matter for the social and political representativeness of their members (hypotheses 1a and 2a supported). As parties expand the benefits associated with membership, there is a decrease in the marginal effects of many demographic factors (religiosity, female, and income—but not age or union membership) on the probability of supporters joining, meaning that parties are more socially representative of their larger group of supporters. Similarly, as benefits increase, there is a decrease in the marginal effects of ideological incongruence on the probability of supporters joining, meaning that parties are more politically representative of their larger group of supporters.

Second, costs set by parties affect the social congruence of their members, but not their ideological congruence (hypothesis 1b supported, hypothesis 2b not supported). Contrary to our expectations, financial costs set by parties do not affect the substantive representativeness of those who join. But they do affect descriptive representation: As parties increase the costs associated with membership, we see increasing marginal effects of many demographic factors (religiosity, female, and income—but not age or union membership) on the probability of supporters joining parties, making parties less socially representative of their larger group of supporters.

Overall, net benefits matter for social and political representativeness of party members (hypotheses 1c and 2c supported). As parties increase the net benefits of membership, ideological incongruence decreases and their members become socially closer to their pool of supporters.

Conclusion

This article has investigated whether and how party memberships may vary depending on how parties set the costs and benefits of membership. We have been particularly interested in establishing how interparty differences in members’ political rights might affect political and demographic differences between party members and other party supporters. This is a politically important question, because where parties use intraparty ballots to make important decisions, the quality of party-based representation may depend on the extent to which party members are appropriate proxies for a party’s pool of supporters.

Our findings are largely reassuring on this score. We have shown that party members in our 10 parliamentary democracies tend to differ ideologically from party supporters; they also tend to be demographically different in regard to religiosity, age, gender, income, and union levels.

However, most of these differences tend to be lower in parties in which membership confers higher absolute or net political benefits. In other words, how parties construct membership—the mix of costs and benefits—seemingly affects the political and social mix among members and the representativeness of the larger pool of party supporters. Our cross-sectional data do not allow us to judge whether rule changes would have similar effects within a single party; to test this, we would need panel data. However, in an era when many parties are expanding their use of intraparty ballots, these findings do suggest that enhanced inclusiveness could work in a favorable direction as far as representation is concerned: low-cost parties that offer high political benefits attract more socially and ideologically representative members than costly parties that offer few benefits.

That said, we stress that our findings are an average snapshot, and opportunities for intraparty voting will not necessarily have these effects in every case. For instance, if party primary rules allow new members to immediately join and participate in a vote, competing candidates and factions have incentives to recruit members during the intraparty campaign. In those cases, who joins the party may be less about individual cost/benefit calculations and more about different engagement levels of the intraparty campaigners who seek to enlist new members. If the backers of more ideologically extreme candidates are the most active campaigners, their efforts could swamp the moderating effects of the ballot’s reduction in the cost/benefit ratio. Arguably, this helps to explain what happened in the British Labour Party’s two recent leadership campaigns in 2015 and 2016. Both campaigns brought a surge in formal membership (and in 2015, in nonmember registered supporters). The new members’ disproportionate support for the leftist Jeremy Corbyn does not fit the picture we present above of ideological convergence as net benefits rise: they were not more like the party’s supporters, at least if “supporters” are defined in terms of the party’s electorate in spring 2015, or by the position of the party’s MPs. Possibly this is a result of the effects of changes in net membership benefits being outweighed by the impact of Momentum, a strongly organized pro-Corbyn support group that operated outside the Labour Party. It is also possible that these ballots merely highlighted a divergence between committed Labour Party “supporters” and centrist and strategic voters who might vote Labour because it was the best anti-Conservative option: Under the latter interpretation, the membership surge may indeed have increased the congruence between loyal supporters and members, thus improving “representativeness,” but at a cost of reducing the party’s reach to those who were potential party voters but not party loyalists. Regardless of how congruence should be measured in this case, it is notable that the new Labour Party members were more like the general population in terms of income and gender (Poletti et al., 2016). This demographic convergence is consonant with the findings presented above in regard to the demographic effects of higher benefits.

More generally, while the effects we show above may not apply in every case, these findings provide evidence that enhancing political rights for members will not necessarily produce “May’s Law” scenarios, ones under which party activists steer parties away from the preferences of the party’s less active supporters. On the contrary, in at least some circumstances, greater intraparty rights might increase the congruence between party members and other supporters. The ways that intraparty democracy is implemented may condition this effect, but that is a subject beyond the scope of the current investigation. Last but not least, the findings presented here seem to confirm the utility of treating party membership as a consumption good. They suggest that there is a scope for further exploration of party-level effects on political participation, and especially of how parties’ experiments with party membership rules affect who joins and, indirectly, who participates in intraparty decisions.

Supplemental material

Supplemental_material - The consequences of membership incentives: Do greater political benefits attract different kinds of members?

Supplemental_material for The consequences of membership incentives: Do greater political benefits attract different kinds of members? by Gabriela Borz, Kenneth Janda, Susan Achury, Susan E Scarrow, Karina Kosiara-Pedersen and Emilie van Haute in Party Politics

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Portions of this work have been supported by an Open Research Area (ORA) project grant funded by Deutsche Forschungsgemeinschaft PO 370/11-1, Economic and Social Research Council (UK) ES/L016613/1, and National Science Foundation (USA) 106498.

Supplemental material

Supplemental material for this article is available online.

Notes

References

Chiru

Gherghina

Rodriguez-Teruel

, et al. (2015) Explaining change in party leadership selection rules. In: Cross

Pilet

(eds) The Politics of Party Leadership: A Cross-National Comparison. Oxford: Oxford University Press, pp. 31–49.

Cross

Pilet

(eds) (2015) The Politics of Party Leadership: A Cross-National Perspective. Oxford: Oxford University Press.

ESS Round 5: European Social Survey Round 5 Data (2010) Data file edition 3.2. Norwegian Social Science Data Services, Norway—Data Archive and distributor of ESS data for ESS ERIC.

ESS Round 4: European Social Survey Round 4 Data (2008) Data file edition 4.3. Norwegian Social Science Data Services, Norway—Data Archive and distributor of ESS data for ESS ERIC.

Faucher

(2015) New forms of political participation. Changing demands or changing opportunities to participate in political parties? Comparative European Politics 13: 409–459.

Gallagher

Marsh

(2004) Party membership in Ireland: the members of Fine Gael. Party Politics 10(4): 407–425.

Gauja

(2013) The Politics of Party Policy: From Members to Legislators. Basingstoke: Palgrave Macmillan.

Gauja

(2015) The construction of party membership. European Journal of Political Research 54(2): 232–248.

Heidar

Saglie

(2003) A decline of linkage? Intra-party participation in Norway 1991–2000. European Journal of Political Research 42(6): 761–786.

10.

Kitschelt

(1989) The internal politics of parties: the law of curvilinear disparity revisited. Political Studies 37(3): 400–421.

11.

Kölln

(2016) Party membership in Europe: testing party-level explanations of decline. Party Politics 22(4): 465–477.

12.

Kopecký

Mair

Spirova

(2012) Party Patronage and Party Government in European Democracies. Oxford: Oxford University Press.

13.

Kosiara-Pedersen

Scarrow

van Haute

(2017) Rules of engagement? Party membership costs, new forms of party affiliation, and partisan participation. In: Scarrow

Webb

Poguntke

(eds) Organizing Political Parties: Representation, Participation, and Power. Oxford: Oxford University Press, pp. 234–258.

14.

Laux

(2011) Was motiviert Parteimitglieder zum Beitritt? In: Tim

Spier

(ed), Parteimitglieder in Deutschland: Die Potsdamer Parteimitgliederstudie. Wiesbaden, Germany: VS Verlag, pp. 61–78.

15.

May

(1973) Opinion structure of political parties: the special law of curvilinear disparity. Political Studies 21(2): 135–151.

16.

McKenzie

(1958) Parties, pressure groups and the British political process. The Political Quarterly 29(1): 5–16.

17.

Narud

Skare

(1999) Are party activists the party extremists? The structure of opinion in political parties. Scandinavian Political Studies 22(1): 45–65.

18.

Norris

(1995) May’s law of curvilinear disparity revisited: leaders, officers, members and voters in British political parties. Party Politics 1(1): 29–47.

19.

Parry

Moyser

Day

(1992) Political Participation and Democracy in Britain. Cambridge: Cambridge University Press.

20.

Paulis

Sierens

van Haute

(2017) Jumping on the bandwagon? Explaining fluctuations in party membership levels. Paper Presented at the Conference of the Political Studies Association, Glasgow.

21.

Pedersen

Bille

Buch

, et al. (2004) Sleeping or active partners? Danish party members at the turn of the millennium. Party Politics 10(4): 367–384.

22.

Poguntke

Scarrow

Webb

, et al. (2016) Party rules, party resources and the politics of parliamentary democracies: how parties organize in the 21st century. Party Politics 22(6): 661–678.

23.

Poletti

Bale

and Webb

(2016) Explaining the pro-Corbyn surge in Labour’s membership, November 16. ESCR Party Members Project Web Page. Available at: https://esrcpartymembersproject.org/2016/11/21/explaining-the-pro-corbyn-surge-in-labours-membership/#more-1222 (accessed 31 March 2017).

24.

Sandri

Seddone

Venturino

(eds) (2015) Party Primaries in Comparative Perspective. Farnham, Surrey: Ashgate.

25.

Scarrow

Gezgor

(2010) Declining memberships, changing members? European political party members in a new era. Party Politics 16(6): 823–843.

26.

Scarrow

(2015) Beyond Party Members: Changing Approaches to Partisan Mobilisation. Oxford: Oxford University Press.

27.

van Haute

Gauja

(eds) (2015) Party Members and Activists. London: Routledge.

28.

van Haute

Paulis

Sierens

(2017) Assessing party membership figures: the map dataset. European Political Science 15(58): 1–12.

29.

van Holsteyn

JJM

Den Ridder

Koole

(2017) From May’s laws to May’s legacy: on the opinion structure within political parties. Party Politics 23(5): 471–486.

30.

Verba

Nie

Kim

(1978) Participation and Political Equality: A Seven-Nation Comparison. Chicago: University of Chicago Press.

31.

Whiteley

Seyd

(2002) High-Intensity Participation: The Dynamics of Party Activism in Britain. Ann Arbor: University of Michigan Press.

32.

Widfeldt

(1995) Party membership and party representativeness. In: Klingemann

Fuchs

(eds) Citizens and the State. Oxford: Oxford University Press, pp. 134–183.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.28 MB