Content uploaded by Raul Gomez
Author content
All content in this area was uploaded by Raul Gomez on Aug 17, 2020
Content may be subject to copyright.
1
The limits of organizational innovation and multi-speed membership. Podemos and its
new forms of party membership
Authors:
Raul Gomez
University of Liverpool
raul.gomez@liverpool.ac.uk
Luis Ramiro
University of Leicester
luis.ramiro@leicester.ac.uk
This is the authors’ accepted version of a paper published in Party Politics
(accepted on 25 October 2017).
Please use the published article for referencing purposes.
After decades of falling membership figures (van Biezen et al. 2012) many European parties are
experiencing processes of organizational innovation. These changes notoriously include new
political rights for party members, greater incentives for affiliates, and the establishment of a
variety of forms of affiliation that have blurred the very divide between members and non-
members (Katz and Mair 2009; Fisher et al. 2014). In Scarrow’s (2015) words, we are witnessing
the emergence of multi-speed membership parties: organizations that offer individuals
opportunities to join through a panoply of affiliation modes and grant them flexibility to engage
in an increased range of political activities of diverse intensity.
The diffusion of the multi-speed membership organizational model is taking place in two parallel
ways. On the one hand, many mainstream and well-established parties have increasingly replaced
their traditional approach to party membership with a multi-speed approach that incorporates
new modes of affiliation. On the other hand, new parties born during this new phase of
organizational innovations at the beginning of the 21st century are adopting even more ambitious
approaches with regard to their members’ standing and roles. While examples of mainstream
parties moving towards a multi-speed membership model are well known (e.g. Webb 2000;
Pasquino and Valbruzzi 2013; Faucher 2015; Gauja 2015), the analysis of new parties’
innovations and their impact is still in progress. In some younger parties, new affiliation
strategies facilitated by the availability of new information and communication technologies have
produced party organizations that, to some degree, resemble the cyber party model (Margetts
2006). Pirate parties are a clear example (Hartleb 2013: 364-365), but other parties such as the
Five Star Movement in Italy (Tronconi 2015), and Podemos (‘We Can’) in Spain have also
revolutionised the notion of party membership and what it entails in terms of rights, obligations
and activities.
The implications of these changes for parties and members’ activism are still unclear. A number
of scholars (e.g. Gauja 2015; Scarrow 2015) have argued that the line between the wider
electorate, supporters, members and militants that has customarily guided the analysis of party
membership since Duverger (1954) is increasingly blurred. In this context, it is important that
we understand the way in which innovation in membership strategies might affect what affiliates
do, their characterising features, and the differences between party members and voters.
Members’ activities and levels of activism affect the very own nature of parties and their linkage
function. As Scarrow and Gezgor (2010) argue, a membership similar to the party’s constituency
can help it to reconnect with voters and better represent them. This may mitigate the public’s
2
distrust and the perception of remoteness suffered by many Western parties. Besides that, given
the increasing empowerment of party affiliates and their participation in relevant decision-
making processes such as candidate and leader selection, manifesto drafting or internal polls
(Scarrow and Gezgor 2010), analysing their social and political profiles gains a renewed
relevance.
This article contributes to the developing research field of the effects of organizational
innovations by analysing an extreme case: Podemos - a young party which deliberately rejects the
traditional notion of membership in favour of a very inclusive organizational arrangement where
all members are granted full rights and no formal differences exist between types of affiliates. We
put to test some of the expectations about the impact of organizational innovations on two of
the most relevant areas: the members-voters gap and the blurring of classical membership
categories. In particular, we aim to answer the following research questions: to what extent are
Podemos’ affiliates different from its voters? And, to what extent has the use of the repertoire of
options for activism offered by Podemos blurred its members’ differentiation in groups of
varying activism? Our findings show that the members-voters gap remains relevant in spite of
Podemos’ extremely open and inclusive notion of membership. We also find that Podemos’
innovative form of affiliation has not led to more diffused differences in members’ roles. On the
contrary, patterns of internal and external activism generate groups that resemble the classical
categories used to understand members’ involvement in more traditional parties. Therefore, we
demonstrate the limits of organizational innovations implemented by parties in the field of
membership policies and how these innovations have not substantially altered key features that
have traditionally characterized party membership.
In the next pages we first discuss the literature about multi-speed membership parties and the
voters-supporters gap, the Podemos’ notion of membership, and our expectations. We then
describe the data used in the analysis. After that, we present our findings and conclude with a
discussion on their implications for the understanding of contemporary party membership and
for future research in this area.
Members’ representativeness and multi-speed membership
The membership decline experienced by parties in most advanced democracies during the past
decades (Mair and van Biezen 2001; van Biezen et al. 2012) has had significant consequences.
Membership loss and citizens’ distrust towards parties have fostered a number of changes in
parties’ membership policies. Old and new parties alike are implementing innovative responses
that stress grassroot involvement, frequently taking advantage of ICTs (Scarrow et al. 2000;
Gibson and Ward 2009; van Biezen et al. 2012; van Biezen and Poguntke 2014). Since the end of
the 20th century, Western parties have empowered traditional members giving them increased
political rights mainly in the areas of candidate and leader selection (Scarrow et al. 2000; Cross
and Katz 2013; Fisher et al. 2014). In addition to these changes, parties have also created new
categories of adherents, or strengthened and diversified them when they already existed (van
Haute and Gauja 2015), giving these new groupings relevant political rights within party
decision-making.
These processes have resulted in the emergence of multi-speed membership parties, adopting
different forms of affiliation with sometimes overlapping political rights, and offering a wider
range of opportunities for participation (Scarrow 2015). In Scarrow’s (2015: 30-31) insightful
analysis, multi-speed membership organizations may offer up to six modes of party affiliation:
3
traditional individual membership, light membership, cyber members, financial sustainers, social
media followers and friends, and news audience. Importantly enough, Scarrow (2015) presents a
picture where, contrary to Duverger’s (1954) model, the boundaries between these categories are
increasingly blurred. Not only can party supporters move between affiliation modes in short
periods of time, but also high intensity activity is no longer confined to traditional members and
militants. On the contrary, while some traditional members may be only loosely linked to the
party (van Haute 2011; van Haute and Gauja 2015), light members and non-member supporters
may present much higher involvement in partisan activities. Empirical evidence on this is still
scarce, but Fisher et al. (2014) and Webb et al. (2017) found members of British political parties
to be more likely to engage in campaign activities than supporters (being the gap greater in the
more intensive tasks); and Gauja and Jackson (2016) found similar results for the Australian
Green Party.
Membership decline may have not only affected parties’ internal functioning but also some of
their key roles. Political participation through party membership has been one of the defining
features of modern parliamentary democracies. Members play a fundamental role in the social
rooting of parties, and in the performance of parties’ linkage and representative functions (van
Biezen and Poguntke 2014; Scarrow 2015). Smaller memberships may deprive parties of key
organizational and campaign resources (Karp et al. 2008), but also weaken them when combined
with lack of members’ representativeness. As van Biezen et al. (2012: 38) argue, party members
have been frequently found to be socially unrepresentative of both their voters and the wider
electorate.
1
As becoming a party member can be relatively costly, members are a group of
disproportionately resourceful individuals who are more likely than non-members to belong to
civil society organizations, and to be male, highly educated, better off, older and working in the
public sector of the economy. Consequently, members could be understood not as part of civil
society but as part of an ‘extended political class’ (van Biezen et al. 2012: 39). This lack of
representativeness could damage parties’ legitimacy and increase public’s distrust further. The
analysis of the features of contemporary party affiliates and whether they constitute a socially
and politically isolated group from the wider electorate is also important given the increasing
power of ordinary members in very consequential internal decision-making processes. However,
innovative membership policies may help to close the gap between the social and political
profiles of voters and supporters. By offering new incentives to participate, parties implementing
new membership policies can have an impact on the type of members they attract (Whiteley and
Seyd 2002; Scarrow and Gezgor 2010: 827).
2
More inclusive, open, flexible and procedurally
cost-free membership policies might therefore help to broaden party membership and be
expected to close the members-voters gap.
A few pieces of research have directly or indirectly approached the question of how innovative
membership arrangements affect the profiles of members and supporters. Scarrow and Gezgor’s
(2010) analysis of the members-voters gap comparing the 1990s and the 2000s (when various
innovations in membership policies had already been implemented in some parties) illustrates
this issue. Although the gap still existed, they found a decrease in some key sociodemographic
differences. Party members were still older (except in Spain), more likely to be male, unionized
and religiously observant, and enjoyed higher incomes than non-members. However, the gap
with non-members had decreased substantially since the end of the 20th century, and there were
not significant differences in education. Similarly, Gauja and Jackson’s (2016) research on
Australian Green party members and supporters does not find differences in age or education,
although the former are still more likely men. Webb et al. (2017), however, find that members
1
Widfeldt (1995) reached similar conclusions using population surveys.
2
For example, Gomez and Tonge (2016) find that new members of the Democratic Unionist Party are slightly more
moderate, facilitating the party’s strategic moderation.
4
are distinct from supporters in several key sociodemographic variables (members are more
frequently male, more educated and have a higher employment status) and that supporters are
more representative of the wider electorate. Additionally, given the exhaustion of some of the
benefits provided by the parties in the past, the remaining contemporary members could be
found to be mostly ideologically motivated and more radical (Scarrow and Gezgor 2010: 828).
3
This would confirm May’s (1973) classical proposition of curvilinear disparity between members
and voters (and party elites), which has not yet found clear empirical support in the literature
(Norris 1995; Gallagher and Marsh 2004). The results showed by Webb et al. (2017) and Scarrow
and Gezgor (2010: 836) go in this direction..
To summarise, evidence regarding the effects of organizational innovations is scarce; the
magnitude of the members-supporters-voters gap in full multi-speed membership parties
remains unclear; and research into the differences between types of affiliates in multi-speed
membership parties is still very scant.
4
We therefore contribute to the badly needed accumulation
of evidence in this field by analysing those issues for a party, Podemos, which has drastically
redefined the notion of member.
Podemos, its notion of membership and our expectations
Podemos is a recently founded organization (January 2014) that became the third-largest party in
the Spanish parliament after the 2015 and 2016 general elections (20.7% and 21.2% share of the
vote, respectively). It is a populist radical-left party emphasizing left-wing socioeconomic
policies, radical democracy, citizens’ political engagement and deliberative democratic
procedures.
5
Podemos’s ideological views determine its approach to party organization and
membership strategies. Podemos firmly believes that conventional parties are in a critical crisis
that requires an organizational transformation. Consequently, it has experimented with new
forms of participation to promote mass involvement in politics. As a highly inclusive
organisation with minimal membership requirements (signing up on the party’s website and
agreeing with their political platform and organizational rules), its adherents do not have any
formal obligation but enjoy many rights without any constraint. There is no membership fee,
though adherents who wish to make donations can choose among a range of options. There is
no probation period or endorsement required to join the party. Spanish citizenship is not a
requisite, and belonging to another party is not forbidden (information about this is not even
asked). Podemos does not offer different forms of affiliation; there is only a single category of
affiliation enabling anybody who wishes to do so to participate in all the main internal decision-
making processes (leadership and candidate selection, manifesto elaboration, internal ballots on
electoral coalitions and government participation, etc.). Only for the purpose of measuring
turnout in the internal online decision-making processes, the party differentiates its members
between passive and active adherents (those that have participated in online decision-making
processes in the last 12 months). In this way, Podemos is even more inclusive and cost-free than
other new parties implementing fluid notions of memberships such as the Pirates parties
(Bolleyer et al. 2015).
3
However, this could be explained by stronger careerism as implied by van Biezen et al. (2012).
4
Bernardi et al. (2017) analyze the differences between ‘old-style’ and ‘new-style’ members in the Italian Democratic
Party; and Sandri and Seddone (2017) analyze differences between diverse types of members and supporters in this
same party.
5
On Podemos see Ramiro and Gomez (2017), Gómez-Reino and Llamazares (forthcoming) and, Rodríguez-Teruel
et al. (2016).
5
Podemos’ organizational innovation has succeeded in attracting a large contingent of adherents.
With 455,932 affiliates (members/voters ratio = 0.14),
6
it has managed to be the second largest
Spanish party in number of members after the conservative Popular Party. Podemos is one of
not so many cases of exponential and rapid growth in an era of general membership crisis (van
Biezen and Poguntke 2014) and it is also a remarkable case because of its explicit attempt to
recreate the mass party model adapted to the 21st century context (Caruso 2016) and to combine
online and traditional offline activism.
In sum, Podemos is a critical case to explore the effects of organizational innovation (Seawright
and Gerring 2008). The party has implemented a radical innovation in membership policy. It has
reduced to a minimum the procedural costs of party membership going beyond to what some
other innovative parties (e.g. the Pirates) have done, and widened the scope for members’
participation in decision-making processes through the use of ICTs. If the expectations about
the capacity of more inclusive membership models to impact the type of members parties attract
are right, Podemos members and voters should be very similar to each other. Additionally,
Podemos has a loose notion of membership whereby no formal difference exists between
members (all have the same rights), and where members are offered many different engagement
opportunities with hardly any obligation. If the new membership models have the impact on
members’ engagement that is often attributed to them (Scarrow 2015), Podemos actual
membership should not only be similar to its voters, but also be in stark contrast with the
concentric membership model described by Duverger (1954).
Thus, our first research question asks about the difference between members and voters in
parties with innovative membership policies. While parties with less inclusive notions of
membership should increase the unrepresentativeness of members, loose affiliation rules and the
blurring of the voter-supporter divide should generate better representativeness. It would,
however, be unrealistic to expect innovative organisational initiatives to make the well-known
participation bias disappear completely. Political participation through party membership, even
in its loosest form, is likely to demand a considerable amount of resources (Verba et al. 1995;
Parry et al. 1992; Whiteley 2011). Therefore, our first hypothesis, in line with previous research
(Scarrow and Gezgor 2010; Gauja and Jackson 2016; Webb et al. 2017), states that there should
still be a voter-supporter gap in this new innovative membership party, with members being
both more resourceful and more ideologically committed than voters:
H1 Podemos’ members are still more resourceful than the party’s voters in a) sociodemographic
terms and b) levels of political activism.
H2 Podemos’ members will be more radical than their voters.
Podemos do not have a formal list of membership options but will very likely engender diverse
groups depending on their use of the repertoire of activities offered by the party. While we know
that most party members usually remain inactive (van Haute 2011; van Biezen and Poguntke
2014; van Haute and Gauja 2015), we do not have evidence about how affiliates choose to
engage with a new party like Podemos, which adopts such a fluid membership notion. As
mentioned earlier, in her depiction of multi-speed membership parties Scarrow (2015) presents a
fuzzy picture where the levels of activism do not correlate well with different options of
affiliation and traditional party members may often be less active than other
affiliates/supporters. However, this view, which has found some support in the literature on
6
In 2015, members/voters ratios for the other main parties were as follows: PP = 0.12, PSOE = 0.03 and Citizens
= 0.01.
6
British parties (Webb et al. 2017), raises an important question: why would anybody become a
traditional member only to very rarely engage in partisan activities? Indeed, a model like the one
depicted by Scarrow may work in established parties where the multi-speed membership model
has been adopted on top of existing traditional organizational arrangements and a formal
distinction still exists between different types of members. In such parties, existing categories
have been created to offer prospective members a mix of old and new forms of affiliation that
are not necessarily related to their desired levels of activity. For example, ‘traditional
membership’ may be the only choice available for those who are committed to the party and
wish to enjoy some of the collective and selective goods associated with full membership (Gauja
2015: 239), but do not have the resources to become a party activist. This, however, may be
different in a party adopting a fluid membership model where supporters wishing to have a
looser relationship with the party do not need to choose any specific membership category.
Therefore, we expect members in parties like Podemos to self-select based on those variables
that are known to be positively associated with higher levels of activism, such as resources and
ideological motivation (Seyd and Whiteley 2002; Whiteley and Seyd 2002). A concentric pattern
of membership modes may then emerge, where those with more resources and stronger
ideological commitment are more likely to be active in both internal and external activities. This
brings us to our last two hypotheses:
H3 Podemos membership follows a concentric model of participation based on different degrees
of external activism and internal involvement
H4. More active members will present a) more socio-economic resources; b) a higher degree of
activism; and c) stronger ideological attachments with the party
Data and methods
To test our hypotheses, we rely on data from a membership survey of Podemos conducted in
the city of Madrid in January 2017. To the best of our knowledge, this is the first membership
survey ever to be conducted in Podemos. At the time fieldwork took place, Podemos had
455,932 registered adherents nation-wide, 282,163 of whom were considered active (by
Podemos’ definition, ‘active adherents’ are those who have participated in online decision-
making processes in the past 12 months). Madrid city is the largest of Podemos’ local
organizations. It represents 10.1% (46,155) of the party’s adherents, and 12% (33,850) of its
active adherents - to put these figures in context, only 6.8% of the Spanish population resides in
Madrid city. This is not surprising given that, besides being one of its electoral strongholds
(Podemos controls the local government of Madrid since 2015), many of the party founders are
also based there. The survey was conducted online using Podemos’ own survey application. An
email was sent by Podemos to all registered adherents.
7
Fieldwork lasted for 15 days and there
was one reminder after the first week.
8
4,996 adherents answered the questionnaire, a response
rate of 10.8% of adherents and 14.0% of the ‘active’ ones. These figures are similar to turnout in
relevant internal ballots (García de Blas 2017; Sánchez 2017). The data were weighted using
known information provided by the party about the distribution of adherents’ age and joining
date.
The survey contained questions about members’ roles and participation in internal and external
party activities. Respondents were asked whether they regularly vote in internal ballots; attend
branch meetings; participate in (and/or coordinate) any of the branches’ working groups;
7
All members have an email address.
8
80% of the answers were received in the first week of fieldwork.
7
represent their branch in any of the regional bodies; and/or had been elected to membership of
any of the party’s governing bodies. Respondents were also asked about the way(s) in which they
had economically contributed to the party (if any)
9
; and the campaign activities they had
participated in (if any).
10
Besides organizational questions, the survey also contained questions
regarding the respondents’ associational involvement
11
, political participation,
12
left-right self-
placement and perception of Podemos and the main other parties’ left-right position (0-10), as
well as socio-demographic variables such as age (in years), gender, employment status and
education. To measure ideological coherence with Podemos, we also calculated the Euclidean
distance between respondents’ ideology and each of the parties’ perceived position on the left-
right continuum.
Information about Podemos’ voters was extracted from two post-election surveys conducted by
the Centre for Sociological Studies between 27th May 2015 and 23rd June 2015 (after the regional
elections in Madrid) and between 7th January 2016 and 19th March 2016 (right after the 2015
Spanish general election). We merged both samples because the number of Podemos voters
interviewed in Madrid city for the general election study was too small (n=52), while, of course,
combining both files resulted in a larger sample (n=142). All results hold when the sample from
the regional election is excluded.
We use logistic regression models to investigate the differences between voters and members. In
order to classify members’ activism, we rely on a two-step cluster analysis of members’
participation in party activities. Two-step cluster analysis is a particular type of unsupervised
machine learning method that is designed to identify mutually exclusive groups of similar cases
in the data based on the variables provided. The appropriate number of clusters is then decided
following the Schwarz Bayesian criterion. Once the optimal number of clusters is identified, we
proceed to investigate the differences between them using multinomial logistic regression
models.
Findings
To test Hypotheses 1 and 2, Table 1 compares the socio-demographic and ideological
characteristics of Podemos members and voters (including both respondents’ self-declared
position on the left-right continuum, and the Euclidean distance between their position and each
of the parties’ perceived position) as well as their political engagement. As can be observed, at
first sight Podemos members and voters are significantly distinct on a number of variables, but
the differences tend to confirm the hypothesis that members are more resourceful (consistent
with the first part of H1) and ideologically more radical (consistent with H2). To start with,
voters are older than members, though differences are small and not statistically significant
(confirming previous findings by Gauja and Jackson 2016 and Webb et al. 2017). They are also
9
This includes contributing through loans for electoral campaigns, crowdfunding, one-off donations, regular fees,
regular fee paid to the party’s foundation, buying official merchandising, or helping the local branch economically to
fund individual activities.
10
This includes volunteering to organise any party activity or campaign; assisting with information tables; hanging
posters in electoral campaigns; and acting as party representative at a polling station.
11
This includes membership of trade unions, professional associations, environmentalist associations, sports
associations, parents’ associations, and neighbourhood associations. For the analysis, a summative index of
associational involvement was constructed ranging from 0 (no membership) to 6 (member of all six types of
associations).
12
This includes taking part in demonstrations, strikes, occupations or blockades, and political consumerism. For the
analysis, a summative index of political participation was constructed ranging from 0 (has not taken place in any of
those activities) to 4 (has taken place in all the four types of activities).
8
significantly less likely than members to be male (67% of Podemos members are male, compared
with 50% of voters), and more likely to be unemployed. The proportion of homemakers,
although low in both cases, is also higher among voters than it is among members. With regard
to education, members are significantly more highly educated (62% of them have a university
degree, compared with 48% of voters). They are also slightly more left-wing than voters
(although, admittedly, the difference is less than a quarter of point on a 0-10 scale), and see
themselves as being ideologically closer to Podemos (and the other radical-left party, IU, which
in 2016 run in coalition with Podemos). Conversely, ideological distance with PSOE (the
mainstream centre-left party) is smaller for voters than it is for members. Lastly, while members
are, on average, involved in a larger number of associations, voters had used more forms of
political participation than members in the past. The fact that voters have been engaged in
slightly more forms of political participation is the only piece of information contradicting the
second part of H1. Differences are, however, really small (0.2 points on a 5-point scale).
[TABLE 1 ABOUT HERE]
By and large, the multivariate analyses in Table 2 confirm these findings, with members being
more highly educated than voters, and more likely to be male, employed (Model 1) and more
left-wing (Model 2). Members are also ideologically closer to Podemos and more distant from
PSOE than voters (Model 3). Moreover, once controls are added, they are also ideologically
more distant from IU than voters. The results of the descriptive statistics concerning political
participation and associational involvement are also confirmed (Model 4). Associational
involvement is higher among members but, oddly enough, although differences are small the
opposite is the case with political participation –in line with Gauja and Jackson’s (2016: 369)
findings on Australian Green Party members and supporters.
[TABLE 2 ABOUT HERE]
Once we have established the differences between members and supporters, we turn to analyse
different types of members. According to H3, patterns of members’ involvement in internal and
external activities on behalf of the party should produce a concentric model of membership.
This model would involve the existence of a core of members who are both very strongly linked
to the party and very active, and other groups of members who are both less active and less
strongly linked to the party. To test this hypothesis, we conducted a two-step cluster analysis
using all the variables on members’ participation in (internal and external) partisan activities
mentioned in the Data and Methods section. All variables are nominal except for two continuous
summative indices measuring the breadth of activities members have participated in on behalf of
the party (0-4 scale), and the breadth of economic contributions made to the party (0-7 scale).
Cluster analysis was used to explore the existence of mutually exclusive groups of similar
members in the sample. If H3 was wrong, the analysis would not show a small number of clearly
differentiated groups ranging from very active to very inactive members but a more fuzzy and
heterogeneous picture where members can be both very active in one domain and inactive in
others.
A 3-cluster solution was automatically selected based on BIC scores – adding additional clusters
would have only marginally improved the BIC scores (see Figure 1A in the Appendix). The
silhouette measure of cohesion and separation for the 3-cluster solution is 0.6, which is above
the 0.5 threshold of good cluster quality.
13
13
To check that the cluster solution was not influenced by the order of cases, we randomly rearranged the data
matrix several times and results were always the same.
9
The size and nature of the three clusters identified by the procedure is shown in Figure 1. By far
the largest group identified (76.2% of the sample), which we have labelled ‘traditional members’,
is made up of those who tend to participate in online internal ballots, but do not attend branch
meetings, participate in working groups, represent their branches in regional meetings or have
been elected to Podemos’ governing bodies. Importantly, ‘traditional members’ only participate
in some activities on behalf of the party (mean = .4 on a 5-point scale), and do not engage with
the options available to contribute economically to the party’s finances (mean = 0.42 on a 7-
point scale). The second largest group in our sample (13.2%), which we have labelled ‘militants’,
comprises members who are overall very active within the party. Militants declare to participate
in internal ballots regularly (at least 96% of them do), attend branch meetings (78%), participate
in working groups at the branch level (62%), and many of them even coordinate some working
group (22%). Some of them (16%) represent their branch in regional bodies or regional working
group meetings and some have been elected to Podemos’ governing bodies (5%). These
members contribute actively to the party’s finances using several of the options available to them
(mean = 1.36), and tend to participate in a range of external activities on behalf of the party
(mean = 2.56). Lastly, the smallest group in our sample (10.5%), which we have labelled
‘audience’ or ‘friends’, are members all of whom do not regularly participate in internal ballots or
attend meetings of any kind. Their contribution to the party’s finances is low (mean = 0.13) and
so is their engagement with external activities on behalf of the party (mean = 0.19). Of course,
we do not believe the group sizes can be generalised beyond our sample. Since our response rate
is similar to turnout in some of the most relevant internal ballots, we did not expect many
‘friends/audience’ members -who are only loosely linked to the party- to have filled in the
survey. This group is, therefore, likely to be largely underrepresented in our sample.
[FIGURE 1 ABOUT HERE]
Coming back to H3, it is important to note that the clustering procedure failed to identify a
consistent pattern of party members who are active in a range of external activities but are only
loosely linked to the party. It may be argued that any segmenting procedure which is designed to
identify mutually exclusive categories will inevitably force respondents to fit within one of the
clusters. However, although outliers do indeed exist, there was little evidence in our data of
substantive numbers of ‘inactive’ traditional members or activists who are only loosely linked to
the party. The percentage of respondents who regularly attend branch meetings but do not
participate in external partisan activities is only 1.3%. In a similar vein, those who do attend
branch meetings are significantly (p<0.01) more engaged with a wider range of external activities
(mean = 2.7 on the 5-point index of activity breadth) than those who do not (mean = 0.44). This
provides strong evidence that the flexible organizational structure of Podemos has generated a
concentric model of party organization with very active members at the core and less active
members in the margins. In this vein, our results point towards a slightly more nuanced image of
party membership engagement (with three types) than the one portrayed by Lisi and Cancela
(2017) who found only two (active and passive members).
Finally, Hypothesis 4 states that more active members will have more socio-economic resources
and stronger ideological attachments with the party than less active members, and will also
present higher levels of political engagement. To test this hypothesis, we first present descriptive
statistics for the three groups of members identified through the two-step cluster analysis (Table
3). As can be seen, although there are some socio-demographic differences between ‘audience’,
‘traditional members’ and ‘militants’, most of those differences are small. Age is higher for more
active members than for less active members. However, the proportion of males and females is
similar for ‘militants’ and ‘audience’ (62%), but there are seemingly more males among
10
‘traditional members’ (69%). Differences are also small with regard to employment status,
although the presence of retired and unemployed members seems to be somewhat higher among
‘militants’. Finally, higher levels of education can be appreciated among ‘traditional members’ in
particular, followed by ‘militants’. Regarding associational involvement, there is a clear (though
not strong) positive relationship between type of member and the number of associations they
belong to. On average, ‘militants’ belong to at least one civic or political association (mean =
0.9), but this figure is smaller for ‘traditional members’ (0.8) and ‘audience’ (0.7). Differences in
political participation are really modest, with ‘militants’ being slightly more active. When it comes
to ideological differences, militants are also somewhat more left-wing (2.68) than ‘traditional
members’ (2.72) and ‘audience’ (2.84), and they also see Podemos as being ideologically closer to
them, while they perceive IU and particularly PSOE as ideologically more distant.
[TABLE 3 ABOUT HERE]
By and large, Table 3 shows very small differences between members, with little evidence of
socio-demographic differences and somewhat more robust evidence when it comes to
ideological differences and differences in associational involvement. The multivariate analyses
confirm those findings (Table 4). As can be seen, ‘traditional members’ are more likely than
‘audience’ adherents to be male, and are also more likely than them to be highly educated (Model
1). The effect of education is only significant at p<0.05, but its significance increases when we
control for ideological distance with Podemos and other parties (Models 3 and 4). There is also
very weak evidence that militants are largely distinct from ‘audience’ adherents in their socio-
economic characteristics, with the exception of unemployment as they are more likely than any
other category of adherents to be unemployed. The first part of H4 does, therefore, not find
strong support in the data, as the degree of involvement with the party does not seem to be
explained by socio-economic resources. In contrast, what is correlated with type of member is
ideology and associational involvement. The more left-wing members are, the more active they
are likely to be (Model 3). Similarly, the smaller the perceived ideological distance between
members and their party (and the larger the distance between them and other competitors), the
more active they are (Model 4). Finally, while the degree of associational involvement does not
seem to be largely different for ‘audience’ and ‘traditional members’, ‘militants’ are more likely to
be engaged with a higher number of civic/political associations (Model 5). These findings
support the last part of H4, which states that more active members have stronger ideological
attachments with the party, but also point to higher degrees of associational engagement among
them.
[TABLE 4 ABOUT HERE]
Conclusions
In the context of the rise of multi-speed membership parties, Podemos represents an extreme
case of redefinition of the membership notion which makes no formal distinction between
different types of adherent and allows anybody who wishes to do so to participate in all the key
internal decision-making processes and party activities with minimal procedural requirements.
This innovative membership approach has nevertheless not led to the disappearance of the well-
known gap between voters’ and members’ profiles. Podemos members are still significantly more
resourceful in terms of education, employment status, and activity in voluntary associations, and
are more likely than voters to be men. Moreover, members are also more ideologically extreme
than voters, and feel closer to the party and more distant from its left-wing competitors. A more
11
fluid, open and inclusive membership strategy, providing greater incentives for affiliation, has
therefore not helped eradicate the members-voters gap. Our findings confirm that, even for such
an innovative party, participation is greatly influenced by individuals’ resources and ideological
commitment (e.g. Verba et al 1978; Whiteley 2011).
Besides looking at the differences between Podemos members and supporters, we also analysed
how adherents choose to engage with the party. By and large, our findings suggest that the
classic distinction between the wider electorate, supporters, members and militants outlined by
Duverger (1954) has only partially lost its analytical relevance. An analysis of members’ activity
finds their behaviour has engendered a concentric model of engagement that enabled us to
identify three groups, from less to more involvement, that we have labelled ‘audience/friends’,
‘traditional members’ and ‘militants’. These categories (particularly the existence of a group of
‘audience/friends’) partially reflect some of Scarrow’s (2015) categories of multi-speed
membership. Podemos does not offer different modes of affiliation but the unprompted
behavior of its members has de facto engendered some sort of multi-speed membership
organization.
For some parties, multi-speed membership strategies seem to have led to the blurring of
Duverger’s distinctions. In those parties, traditional party membership co-exists with new
categories of affiliates that enjoy rights and perform tasks previously reserved to core party
members. This gives raise to situations in which new affiliates (such as ‘friends’ or ‘sympathisers’)
may participate more actively in partisan activities than traditional members – thereby obscuring
the organizational and analytical distinction between those categories (Scarrow 2015). However,
the case of Podemos shows that is not the end of the story. When a new party purposely
removes any distinction between types of adherents and, through a very inclusive and cost-free
procedure, allows everyone who agrees to become a member to participate in all party activities,
individuals eventually align themselves into concentric groups where those at the centre are more
active than the rest in both internal and external partisan activities.
In this way, our findings put into question the capacity of new models of party membership to
live up to their expectations. The radically new notion of membership implemented by Podemos
has not been able to remedy the voters-members gap and, with it, one of the likely sources of
party malaise. Similarly, Podemos’ membership model has not removed the well-known gender
gap in party membership (Coffe and Bolzendahl 2010). or completely fulfilled expectations
regarding internal participation. Podemos’ members group together into concentric groups that
differ in their degrees of engagement, which could potentially create the sort of internal oligarchy
the party aimed to avoid. While our findings come from a case study, there is no reason to think
they cannot be found in other new parties too. If Podemos’ innovative approach to membership
has fell short of the expectations attributed to organizational innovations, other more modest
attempts should more clearly reproduce old patterns.
Interestingly enough, the differences in Podemos members’ levels of activism do not seem to be
mostly due to disparities in resources.
14
Traditional members are a very particular subset of
adherents who are ideologically motivated and have experience participating in a range of other
movements or associations. Many of them are best defined as militants who are driven by
ideological/motivational reasons and organizational experience rather than socio-economic
factors. In this way, Podemos’ model seems to spontaneously generate an ideologically driven
14
Further research will be needed to confirm that this finding indicates a different pattern to the one found by
Webb et al. (2017).
12
participatory ‘elite’ despite the party’s attempts to expand participation and to fight inequalities in
this regard.
15
In sum, our results question the implications of some new membership categorisations, cast
doubt about the capacity of new membership models to have consequential effects and seem to
strengthen classic typologies (Duverger 1954; Scarrow 2015). However, the introduction of new,
loose and flexible forms of affiliation is still a recent phenomenon. While this article analyses the
case of a fully new party, further research should investigate the way in which party adherents
choose from the repertoire of (formal or behavioural) affiliation opportunities available to them
in a broader range of parties. In addition, we believe future studies should rely on panel data to
address the dynamic nature of party membership -with members moving between different
forms of affiliation (Scarrow 2015)-, and the factors determining such changes. Only then will we
be able to fully understand the complexities of contemporary forms of party affiliation.
References
Bernardi L, Sandri G and Seddone A (2017) Challenges of political participation and intra-party
democracy: Bittersweet symphony from party membership and primary elections in Italy. Acta
Politica 52(2): 218-240.
Bolleyer N, Little C and von Nostitz F-C (2015) Implementing Democratic Equality in Political
Parties: Organisational Consequences in the Swedish and the German Pirate Parties. Scandinavian
Political Studies 38(2): 158-178.
Caruso L (2016) Il partito di massa del XXI secolo. il Manifesto 30-12-2016
https://ilmanifesto.it/il-partito-di-massa-del-xxi-secolo/
Cross W and Katz R (eds) (2013) The challenges of intra-party democracy. Oxford: Oxford University
Press.
Duverger, M (1954) Political Parties. New York: John Wiley & Sons.
Faucher F (2015) New Forms of Political Participation. Changing Demands or Changing
Opportunities to Participate in Political Parties? Comparative European Politics 13: 405-429.
Fisher J, Fieldhouse E, and Cutts D (2014) Members are not the only fruit: volunteer activity in
British political parties at the 2010 general election. British Journal of Politics and International
Relations 16: 75-95.
Gallagher M and Marsh M (2004) Party membership in Ireland: The members of Fine Gael. Party
Politics 10(4): 407–425.
García de Blas E (2017) Podemos no logra movilizar a sus bases en la consulta sobre la moción
de censura. El País 15-05-2017. Accessed 16-06-2017
http://politica.elpais.com/politica/2017/05/15/actualidad/1494873538_421394.html
15
Podemos is currently coming to terms with this reality and has timidly started to formally distinguish ‘militants’
from affiliates, although both groups have the same rights and obligations (Podemos 2017).
13
Gauja A (2015) The Construction of Party Membership. European Journal of Political Research 54:
232-248.
Gauja A and Jackson S (2016) Australian Greens party members and supporters: their profiles
and activities. Environmental Politics 25: 359-379.
Gibson R and Ward S (2009) Parties in the digital age: A review article. Representation 45(1): 87–
100.
Gomez R and Tonge J (2106) New members as party modernisers: The case of the Democratic
Unionist Party in Northern Ireland. Electoral Studies 42: 65-74.
Hartleb F (2013) Anti-elitist Cyber Parties? Journal of Public Affairs 13(4): 355-369.
Karp J Banducci S and Bowler S (2008) Getting out the vote: Party mobilization in a
comparative perspective. British Journal of Political Science 38(1): 91–112.
Katz R and Mair P (2009) The cartel party thesis: A restatement. Perspectives on Politics 7(4): 753–
766.
Gómez-Reino M and Llamazares I (forthcoming) Embracing Populist Discourse in Party
Formation and Electoral Mobilization: The Case of Podemos. In: Hawkins K, Carlin R, Littvay
L and Rovira-Kaltwasser C (eds) The Ideational Approach to Populism: Concept, Theory, and Method.
Lodon: Routledge.
Lisi M and Cancela J (2017) Types of party members and their implications: Results from a
survey of Portuguese party members Party Politics DOI: 10.1177/1354068817722445
Mair P and van Biezen I (2001) Party membership in twenty European democracies, 1980–2000.
Party Politics 7(1): 5–21.
Margetts H (2006) The Cyber Party. In: Katz R and Crotty W (eds.) Handbook of Party Politics.
London: Sage, pp. 528-535.
Parry G, Moyser G and Day N (1992) Political Participation and Democracy in Britain. Cambridge:
Cambridge University Press.
Pasquino G and M Valbruzzi (2013) Prime Minister Primaries: Candidate Selection Between
Innovation and Manipulation. Italian Politics 28(1): 112-132.
Pedersen K et al. (2004) Sleeping or active partners? Danish party members at the turn of the
millennium. Party Politics 10(4): 367–383.
Podemos. 2017. Reglamento para conformar el censo de militantes.
https://files.podemos.info/CpMByRams8.pdf [Accesed on 7 September 2017]
Ramiro L and Gomez R (2017) Radical Left Populism during the Great Recession: Podemos and
its Competition with the Established Radical Left, Political Studies 65(1): 108-126.
Rodríguez-Teruel J, Barrio A and Barberá O (2016) Fast and Furious: Podemos’ Quest for
Power in Multi-level Spain, South European Society and Politics 21(4): 561-585.
14
Saglie J and Heidar K (2004) Democracy within Norwegian political parties: Complacency or
pressure for change? Party Politics 10(4): 385–405.
Sánchez MF (2017) Baja participación en las últimas consultas de Podemos a sus inscritos.
Cuartopoder 16-05-2017 Accessed 16-06-2017
https://www.cuartopoder.es/espana/2017/05/09/las-ultimas-consultas-podemos-inscritos-
registran-una-baja-participacion/
Sandri G and Seddone A (2015) Sense or sensibility? Political attitudes and voting behaviour of
party members, voters, and supporters of the Italian centre-left, Italian Political Science Review 45(1):
25-51.
Scarrow S (2015) Beyond Party Members. Changing Approaches to Partisan Mobilization. Oxford:
Oxford University Press.
Scarrow S and Gezgor B (2010) Declining membership, changing members? European political
party members in a new era. Party Politics 16(6): 823-843.
Scarrow S, Webb P and Farrell D (2000) From Social Integration to Electoral Contestation. The
Changing Distribution of Power Within Political Parties. In: Dalton R and Wattenberg M (eds)
Parties Without Partisans. Oxford: Oxford University Press, pp. 129-153.
Seawright J and Gerring J (2008) Case Selection Techniques in Case Study Research: A Menu of
Qualitative and Quantitative Options. Political Research Quarterly 61(2): 294-308.
Seyd P and Whiteley P (2002) New Labour’s Grassroots: the Transformation of the Labour Party
Membership. London: Palgrave.
Tronconi F (2015) Beppe Grillo’s Five Star Movement. Farnham: Ashgate.
van Biezen I, Mair P and Poguntke T (2012) Going, Going… Gone? The Decline of Party
Membership in Contemporary Europe. European Journal of Political Research 51(1): 24-56.
van Biezen I and Poguntke T (2014) The decline of membership-based parties. Party Politics
20(2): 205-216.
van Haute E (ed) (2011) Party membership in Europe: exploration into the anthills of party politics.
Brussels: Editions de l’Universite de Bruxelles.
van Haute E and Gauja, A (eds) (2015) Party members and activists. London: Routledge.
Verba S, Schlozman K and Brady H (1995) Voice and Equality: Civic Voluntarism in American
Politics. Cambridge, MA: Harvard University Press.
Webb P (2000) The Modern British Party System. London: Sage.
Webb P, Poletti M and Bale T (2017) So who really does the donkey work in ‘multi-speed
membership parties’? Comparing the election campaign activity of party members and party
supporters. Electoral Studies 46: 64-74.
15
Whiteley P (2011) Is the party over? The decline of party activism and membership across the
democratic world. Party Politics 17(1): 21-44.
Whiteley P and Seyd P (2002) High-Intensity Participation. Ann Arbor: The University of Michigan
Press.
Widfeldt A (1995) Party Members and Party Representativeness, in H-D Klingemann and D
Fuchs (eds) Citizens and the State. Oxford: Oxford University Press, pp. 134–82.
16
FIGURES AND TABLES
Table 1. Descriptive statistics of members and voters with bivariate significance tests.
Members
Voters
Sig
Age
43
44
Gender (% Male)
67
50
***
Working (%)
64
54
Retired (%)
17
14
Unemployed (%)
14
23
***
Student (%)
4
6
Homemaker (%)
1
4
***
Primary education or less (%)
6
11
***
Secondary education (%)
32
41
***
Higher education (%)
62
48
***
Left-right ideology
2.7
3.0
***
Distance Podemos
0.7
1.0
***
Distance IU
0.9
1.1
***
Distance PSOE
3.7
2.8
***
Associational involvement (0-6)
0.8
0.3
***
Political participation (0-4)
2.2
2.4
***
n
4,942
141
+ p <0.1, * p<0.05, ** p<0.01, *** p <0.001
17
Table 2. Logistic regression models explaining differences between voters and members.
(1)
(2)
(3)
(4)
Voters
v members
Voters
v members
Voters
v members
Voters
v members
age
-0.037+
-0.037+
-0.037+
-0.029
(0.021)
(0.021)
(0.021)
(0.021)
Secondary education
-0.527***
-0.493***
-0.450**
-0.392*
(0.098)
(0.108)
(0.140)
(0.159)
Higher education
-0.982***
-0.947***
-1.057***
-0.919***
(0.103)
(0.110)
(0.143)
(0.176)
male
-0.708***
-0.718***
-0.824***
-0.805***
(0.040)
(0.045)
(0.070)
(0.070)
retired
0.697
0.702
0.748
0.669
(0.460)
(0.462)
(0.472)
(0.452)
unemployed
0.634***
0.667***
0.752***
0.603***
(0.073)
(0.066)
(0.088)
(0.075)
student
-0.309
-0.308
-0.518
-0.588
(0.531)
(0.534)
(0.545)
(0.537)
homemaker
1.833***
1.870***
2.136***
2.285***
(0.191)
(0.189)
(0.220)
(0.229)
ideology
0.215***
(0.029)
distance Podemos
0.388***
0.391***
(0.037)
(0.039)
distance IU
-0.071*
-0.076**
(0.027)
(0.027)
distance PSOE
-0.400***
-0.444***
(0.026)
(0.022)
political participation index
0.358***
(0.024)
association index
-0.699***
(0.058)
intercept
-0.915
-1.553
0.195
-0.585
(1.423)
(1.419)
(1.456)
(1.474)
N
4,641
4,641
4,641
4,641
Standard errors in parentheses. Reference categories: education – less than
secondary education; employment relation – currently employed.
+ p<.1; * p<.05; ** p<.01; *** p<.001
18
Figure 1. Description of clusters.
Legend: Item 1 – Index of activity intensity; Item 2 – do you attend branch meetings?; Item 3 – do you normally vote in internal
ballots?; Item 4 – do you participate in any of your branch’s working groups?; Item 5 – do you coordinate any of your branch’s
working groups/areas?; Item 6 – do you represent your branch in regional bodies or working areas?; Item 7 – Index of economic
intensity; Item 8 – are you a member of Podemos’ governing bodies? Within each box, bars on the left represent negative answers
and bars on the right represent positive answers.
19
Table 3. Descriptive statistics of different clusters of adherents with bivariate significance tests.
Audience/friends
Traditional
members
Militants
Age
41
43
45
***
Gender (% Male)
62
69
62
***
Working (%)
67
65
57
Retired (%)
13
17
23
Unemployed (%)
15
13
17
*
Student (%)
4
4
3
Homemaker (%)
1
1
1
Primary education or less (%)
8
5
8
+
Secondary education (%)
33
32
29
Higher education (%)
59
62
63
Left-right ideology
2.84
2.72
2.68
**
Distance Podemos
0.73
0.66
0.6
+
Distance IU
0.89
0.88
0.94
Distance PSOE
3.36
3.69
3.84
***
Associational involvement (0-6)
0.7
0.8
0.9
***
Political participation (0-4)
2.2
2.2
2.3
*
n
526
3,801
661
+ p <0.1, * p<0.05, ** p<0.01, *** p <0.01
20
Table 4. Multinomial logistic regression with different clusters of adherents.
Model 1
Model 2
Model 3
Model 4
Traditional
member v
audience
Militant
v audience
Traditional
member v
audience
Militant
v audience
Traditional
member v
audience
Militant
v audience
Traditional
member v
audience
Militant
v audience
Age
0.0113
0.0317
0.0112
0.0315
0.0107
0.0313
0.0101
0.0307
(0.020)
(0.026)
(0.020)
(0.025)
(0.020)
(0.025)
(0.020)
(0.026)
Secondary education
0.431*
0.154
0.405*
0.128
0.442**
0.190
0.434**
0.153
(0.160)
(0.294)
(0.156)
(0.289)
(0.151)
(0.289)
(0.152)
(0.287)
Higher education
0.530*
0.385
0.507*
0.362
0.617**
0.527+
0.595**
0.407
(0.211)
(0.294)
(0.207)
(0.290)
(0.200)
(0.290)
(0.209)
(0.289)
Male
0.286**
0.0280
0.292**
0.0339
0.336**
0.0868
0.332**
0.0728
(0.092)
(0.135)
(0.094)
(0.135)
(0.100)
(0.143)
(0.098)
(0.143)
Retired
0.0720
0.0981
0.0689
0.0957
0.0719
0.100
0.0913
0.246
(0.507)
(0.646)
(0.506)
(0.644)
(0.504)
(0.643)
(0.494)
(0.636)
Unemployed
-0.0220
0.342+
-0.0423
0.318+
-0.0713
0.275
-0.0521
0.405*
(0.111)
(0.174)
(0.114)
(0.183)
(0.117)
(0.181)
(0.117)
(0.183)
Student
0.229
0.756
0.220
0.743
0.283
0.817
0.290
0.940
(0.619)
(0.835)
(0.609)
(0.822)
(0.601)
(0.811)
(0.592)
(0.830)
Homemaker
0.474
0.626
0.463
0.613
0.367
0.454
0.360
0.527
(0.492)
(1.061)
(0.500)
(1.057)
(0.506)
(1.100)
(0.515)
(1.074)
Ideology
-0.159**
-0.183***
(0.044)
(0.044)
distance Podemos
-0.174*
-0.295**
-0.175*
-0.315**
(0.069)
(0.082)
(0.068)
(0.086)
Distance IU
0.0762
0.234**
0.0801
0.257**
21
(0.049)
(0.072)
(0.048)
(0.074)
Distance PSOE
0.160***
0.212***
0.161***
0.210***
(0.032)
(0.034)
(0.034)
(0.037)
political participation index
-0.00450
0.0838+
(0.039)
(0.048)
association index
0.0537
0.265***
(0.054)
(0.060)
Intercept
0.799
-1.764
1.271
-1.222
0.230
-2.649*
0.228
-2.990*
(0.974)
(1.198)
(0.948)
(1.160)
(0.908)
(1.103)
(0.955)
(1.163)
N
4,516
4,516
4,516
4,516
4,516
4,516
4,516
4,516
Standard errors in parentheses. Reference categories: education – less than secondary education; employment relation – currently employed.
+ p<.1, *p<.05, **p <.01, *** p<.001
22
APPENDIX
Figure 1A. Two-step cluster analysis. BIC scores for different numbers of clusters.
0
5000 10000 15000 20000
BIC
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Number of clusters
