• party organisations;
  • party membership;
  • membership decline;
  • disengagement


  1. Top of page
  2. Abstract
  3. Party membership in the twenty-first century
  4. Declining levels of party membership
  5. Implications and discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Appendix

This article offers an overview of levels of party membership in European democracies at the end of the first decade of the twenty-first century and looks also at changes in these levels over time, comparing party membership today with figures from both 1980 and the late 1990s. While relying primarily on the direct and individual membership figures as reported by the parties themselves, the fit of the data with survey data is explored and it is concluded that the two perform well in terms of convergent validity. The differences between large and small democracies are examined, as well as old and new democracies, and it is found that levels of party membership are related to both the size and age of the democratic polity in question. Finally, the implications of the patterns observed in the membership data are discussed, and it is suggested that membership has now reached such a low ebb that it may no longer constitute a relevant indicator of party organisational capacity.

This article offers an overview and initial analysis of the levels of party membership in European democracies at the end of the first decade of the twenty-first century, and discusses the implications of changes in party membership for our understanding of models of party organisation. The first two sections provide a broad comparison of the data on party membership in European democracies and analyse changes in membership over time, comparing contemporary levels with those from both 1980 and the late 1990s, and exploring variations in national patterns. In the third section, we discuss some of the implications of the patterns that we observe in the membership data and suggest that membership has now reached such a low ebb that it may no longer constitute a relevant indicator of party organisational capacity. Alternatively, if membership continues to be regarded as offering a meaningful gauge of party organisational strength, we might then conclude that party organisations have reached such a low ebb that the formal organisational level is itself no longer a relevant indicator of party capacity.

In presenting an overview of the levels of party membership in European democracies at the end of the first decade of the twenty-first century, our intentions are threefold. First, we wish to update the data reported in Mair and van Biezen (2001), which followed on from an earlier study by Katz et al. (1992). In gathering these data, we have once again been fortunate to have been able to rely on the help, advice and information generously provided by a great number of party scholars across Europe.1

Second, we have sought to extend the coverage of the data we report. Our earlier study encompassed twenty European democracies, including most of the established liberal democracies in Western Europe, the younger democracies of Greece, Portugal and Spain in Southern Europe as well as four post-communist democracies: the Czech Republic, Slovakia, Hungary and Poland. This time we have been able to collect data for an additional seven countries, adding Cyprus and the post-communist democracies of Estonia, Latvia, Lithuania, Bulgaria, Romania and Slovenia, thus bringing the total number of countries to 27. This means that our study now covers virtually all of the European democracies, with the exception of some of the smaller states such as Iceland, Luxembourg and Malta, or more recently consolidated democracies such as Croatia and Serbia. This larger and more inclusive sample will permit a more comprehensive investigation of the levels of party membership across contemporary European democracies, and will also allow a more robust exploration of the differences in the patterns of party affiliation between the longer established democracies and the more recently created ones, and the post-communist democracies in particular.

Our third intention has been to explore changes in the levels of party membership over time, and to assess the extent to which the trends observed at the turn of the twentieth century continued to persist into the twenty-first. The conclusion in the earlier research was that total party membership, expressed in both absolute numbers and as a percentage of the electorate, was then markedly in decline (Mair & van Biezen 2001: 11). Not only was there a continuation of the decline in membership relative to the size of the electorate already noted by Katz et al. (1992), but, for the first time, there was also evidence of a very strong and consistent decline in the absolute numbers of party members, suggesting that parties were struggling to hold on to their existing membership organisations and were failing to recruit new members in significant numbers. As we will show at greater length below, both trends have continued to be manifest in the early twenty-first century, thereby accentuating the sense of disengagement from party politics that appeared to gather pace in the 1990s.

Following the same approach and methodology as in the earlier studies, the aggregate data we are presenting here are based on the direct, individual membership figures as reported by the parties themselves, with all the usual caveats about their potential unreliability (for a discussion, see Mair and van Biezen 2001: 6–7). Although we have aimed to be as inclusive as possible, the data are inevitably incomplete in that minor parties are sometimes not included or membership figures are sometimes not available. As a result, the membership as percentage of the electorate (M/E) level may slightly underestimate the real membership ratio.2 At the same time, it is possible that parties are inclined, for reasons of organisational resources or legitimacy, to exaggerate their membership levels, which may have the opposite effect of overestimating the overall membership ratios. We know little, if anything, about the net effect of these tendencies. At the same time, alternative data sources, such as the evidence produced by survey research, are not always a suitable substitute as they sometimes run into similar issues of availability, reliability and validity, and cause particular problems when attempting to trace trends over time. However, one alternative data source that does appear robust is the fourth round of the European Social Survey (ESS), which included a separate question on party membership and covers 23 of the 27 countries we address through the party data. This not only offers a useful snapshot of the levels of party membership in 2008 when the survey was conducted, but also offers an important check on the validity of the aggregate data. We explore this question below. As far as our interest in comprehensive cross-national and cross-temporal analyses of party membership is concerned, however, the options are severely limited and hence we rely primarily on the aggregate party data.

Over and above the more general issue of the decline in party membership, we also anticipate that two general distinctions will be apparent from the data. The first, which follows in line with much of the expectations and hypotheses in the literature on post-communist Europe (e.g., Kopecký 1995; Biezen 2003; Lewis 2000), and which was already indicated in the membership levels recorded in the late 1990s (Mair & van Biezen 2001), is that party membership levels in the post-communist democracies will have remained substantially below those in the established Western polities. Indeed, the newness of these democracies, the weakening (if not elimination) of traditional cleavages as a result of the communist attempt to build a classless society, and the fact that the party organisations emerged in a context in which they could already avail of modern communications networks in their efforts to campaign for support, are likely to have discouraged efforts to build mass organisations, even in the longer term. This is also likely to be still the case, albeit obviously to a less developed extent, in the southern European polities that emerged from authoritarian rule in the 1970s (Morlino 1998; van Biezen 2003). In other words, we continue to anticipate that the newer the democracy, the smaller is likely to be the membership level.

The second distinction that is likely to be apparent is that between large and small democracies. The relationship between size and democracy was first theorised by Dahl and Tufte (1973: 43), who hypothesised that ‘the larger the citizen body . . . the weaker the incentive to participate’– a proposition which has obvious implications for party membership.3 Existing empirical evidence also seems to point to a relationship between membership levels, whether of trade unions or parties, and polity size. Bartolini and Mair (1990: 235), for example, noted that the larger European polities tended to lag behind their smaller neighbours in levels of organisational (party and union) density, and suggested that this might well have reflected a systemic property. The same pattern was noted elsewhere by Mair (1991), and was discussed most explicitly by Weldon (2006: 475), who concluded that the increasing size of a polity ‘clearly diminishes two types of participation – party membership enrolment and member activism within parties’. Although important, however, the distinction between large and small polities is also somewhat problematic, since it is unclear whether the relevant differences are categorical or gradual, or against which standard they should be measured. Nonetheless, in line with the literature, and in line with patterns that were also evident in the late 1990s, we anticipate that ‘smaller’ polities will have relatively larger memberships.

Party membership in the twenty-first century

  1. Top of page
  2. Abstract
  3. Party membership in the twenty-first century
  4. Declining levels of party membership
  5. Implications and discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Appendix

The first set of data we present summarises the overall levels of party membership in contemporary European democracies, based on the most recent data available. These figures are presented under two headings: the total party membership expressed in raw numbers, and the party membership expressed as a percentage of the electorate (M/E) – an indicator that is more suitable for cross-national comparisons. The data are reported in Table 1, which presents the countries ranked in descending order of M/E level. A breakdown by individual parties can be found in the Appendix.

Table 1.  National levels of party membership
CountryYearTotal party membershipTotal party membership as percentage of electorate (M/E)
Cyprus (Greek)200981,43316.25
Czech Republic2008165,4251.99
United Kingdom2008534,6641.21
Mean (N = 27)  4.65

Taking all 27 countries together, the evidence of our aggregate data shows that the average membership ratio is just 4.7 per cent. This is somewhat lower than the 20-country mean of 5.0 reported for the late 1990s (Mair & van Biezen 2001), and considerably lower than the levels recorded for an even more limited group of countries in the 1980s (Katz et al. 1992). This clearly points to a continuation of the downward trend into the new century. This becomes even more apparent when we note that the average levels of party membership of the twenty countries originally included in our earlier study stands at just 4.3 per cent, almost half a percentage point lower than the 27-country mean. In little less than a decade, the average membership level in these countries has therefore fallen by another 0.7 per cent. The overall 27-country mean, on the other hand, is pushed slightly higher than the 20-country mean as a result of the inclusion of the high-membership case of Cyprus (16.3 per cent). Also noteworthy are the continuing high levels of membership in Austria. Both Cyprus and Austria are evident outliers, and excluding both would reduce the overall country mean to just 3.9 per cent. Located at the other extreme are countries such as Latvia and Poland, where the levels of party membership fail to reach even 1 per cent, as well as Britain, Hungary, France and the Czech Republic, where they register under 2 per cent.4

As noted above, the analysis of party membership levels at the end of the 1990s suggested that the post-communist polities stand out from the other democracies, and we anticipated that these differences continue to be marked in the more recent data. This is clearly the case, notwithstanding two decades of democratic development since the fall of the Berlin Wall. Taking all the newer democracies together, it can be seen that the average level of party membership totals 3.5 per cent, as against an average of 5.7 per cent for the longer established democracies. This discrepancy is entirely due to the post-communist democracies, in that, and contrary to our expectations, the average level of party membership in the three Southern European democracies is now actually higher (4.9 per cent) than the overall 27-country mean. Indeed, almost all the post-communist democracies (the exceptions are Slovenia and Estonia) fall below the overall mean. The Eastern European democracies tend to cluster towards the bottom of the list, with three of the five lowest scoring positions being occupied by former communist polities. Taken together, moreover, the average level of party membership in post-communist Eastern Europe is just 3.0 per cent, little more than half the 5.6 per cent average for the Western and Southern European democracies combined.

Table 2 presents the results of the effects of the newness of democracy on the level of party membership, showing that party membership is significantly lower in the post-communist democracies established after 1989. The effect largely remains when the two outliers are excluded from the analysis. Interestingly, the effect of polity size (discussed below) remains intact under both scenarios. When the type of democracy is included, its magnitude in fact seems to increase (columns 3 and 4).

Table 2.  The effect of size and the newness of democracy on party membership
 All countriesWithout Austria, CyprusAll countriesWithout Austria, Cyprus
  1. Notes: Figures are for 2009. Entries are OLS coefficients, bootstrapped standard errors (250 replications) in parentheses.

Small2.48 (1.14)1.27 (0.734)3.23 (1.20)1.73 (0.074)
New  −3.43 (1.29)−1.57 (0.702)

We also anticipated a continuing difference between small and large polities, with the larger countries expected to have lower membership ratios. The more recent data confirm this pattern. The larger polities of France, the United Kingdom and Germany are among the lowest scoring in terms of M/E levels, while smaller countries such as Austria, Cyprus, Finland and Belgium are among the highest ranking polities. The relationship is not entirely unequivocal, of course, in that smaller democracies such as Latvia and Ireland are characterised by relatively low membership ratios, whereas Italy has a relatively high level of membership. The correlation between the size of the electorate and levels of party membership is −0.284, and, with a small n, is significant statistically only at the 0.10 level. This large versus small dichotomy offers further evidence that membership levels may reflect a systemic bias that might also be apparent in other forms of association and organisation.

That said, and given the small number of observations, any investigation of this relationship without making distributional assumptions is relatively risky. The uneven distribution of the size of the electorate across countries suggests that, if size matters, it may not exert a linear effect on the levels of party membership (see also Weldon 2006). For this reason, we have recoded the ratio variable ‘size’ into a dichotomous variable, thus creating a group of ‘large countries’ with more than 20 million eligible voters (including, in ascending order, Poland, Spain, France, United Kingdom, Italy and Germany) and a group of ‘small countries’ (all others). In order to assess the mean difference between small and large countries in their levels of party membership, we have simply regressed the latter on the former. The results indicate that this effect is 2.48, which means that small countries register on average approximately 2.5 per cent more party members than large countries (see Table 2). The effect is not exceptionally high, but it does achieve statistical significance at conventional levels. When we repeat the analysis excluding the two (small) outliers Austria and Cyprus, the corresponding coefficient diminishes by more than half (to 1.27), while the associated measure of uncertainty is reduced as well (0.734 compared to 1.14). The effect is thus clearly exacerbated by the presence of the two outliers, although even in their absence there still seems to be a significant relation between size of the electorate and levels of party membership.5

Declining levels of party membership

  1. Top of page
  2. Abstract
  3. Party membership in the twenty-first century
  4. Declining levels of party membership
  5. Implications and discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Appendix

The second set of data we present concerns the changes in the levels of party membership over time. First of all, we have assessed the change in membership ratios over the course of the last decade. For this purpose, we have taken the data for the late 1990s reported in Mair and van Biezen (2001) and compared them with the most recent data available. In addition, benefiting from evidence recently made available, we have been able to include an additional three countries (Bulgaria, Estonia and Slovenia) in the overall comparison, such that the longitudinal comparison now encompasses 23 countries. The summary measures are presented in Table 3, which reports changes in the M/E levels as well as in the absolute number of members, ranked in descending order according to the percentage change in the latter. It should be noted that some of the caveats about the validity of comparisons – most notably the shorter time span of party political development and the legacy of communist mass mobilisation – still apply for the more recently established democracies, and the post-communist ones in particular. At the same time, however, the new democratic polities have now had an additional ten years of experience with democracy. Approximately two decades after consolidation (more than three for the Southern European countries), the newly established democracies have had considerable time for the institutionalisation of their parties and the party systems, and the figures are less likely to be distorted than they may have been before. Any volatility in the party formations or their weak organisational presence on the ground, therefore, is more likely to be a reflection of structural and systemic properties rather than a symptom of the transitional status of these polities. Furthermore, at this point, the post-communist figures are less likely to be misrepresented as a result of the organisational legacy of the former ruling communist parties and their satellites. The relative impact of these successor organisations was important primarily in the first years after the transition and is unlikely to be of the same consequence today, albeit not entirely irrelevant, as is demonstrated by the Czech case (see below).

Table 3.  Party membership change since late 1990s
CountryPeriodChange in M/E ratioChange in number of members% change in number of members
Czech Republic1999–2008−1.45−113,560−40.70
United Kingdom1998–2008−0.71−305,336−36.35

The evidence in Table 3 shows that, since the turn of the century, the large majority of European democracies have experienced a further decline in their membership levels, both in absolute numbers and as a percentage of the electorate. This can be seen, first of all, in the widespread and steady decline in M/E levels across the board. Inevitably, the fall is more muted because of the shorter time span under investigation. This is also why a certain degree of fluctuation is to be expected. It may also be simply that the trend is now bottoming out. It is no less unequivocal for that, however.

With only a handful of exceptions, membership ratios have fallen everywhere. Moreover, there does not seem to be any relationship between decline and whether the democracy in question is long-established or relatively new. Most of the older democracies have experienced a decline, although party membership in France and Italy has increased by 0.3 and 1.5 per cent, respectively.6 Membership has fallen in the two younger Southern European democracies of Greece and Portugal, even though it continues to be on the increase in Spain. Even in post-communist Central and Eastern Europe membership levels have fallen substantially, thus defying any expectations that the low levels of partisan affiliation might simply be a reflection of the newness of their democratic polities, although Estonia appears to offers a significant exception to this pattern, albeit with respect only to 2002.

These recent data also underline the evidence of continued decline in the absolute number of members, with party membership expressed in raw numbers falling almost everywhere in the last decade. In some cases, the decline has been substantial. Party membership in Slovakia nearly halved in just seven years, while membership in the Czech Republic reduced by 40 per cent compared to the late 1990s. Indeed, the decline in raw numbers actually seems to be most pronounced in Central and Eastern Europe, with four of the five sharpest drops being recorded by post-communist democracies. In some cases, as in the Czech Republic, this can be attributed to the sustained membership loss suffered by the Communist Party and its satellite. Given that the membership of the former ruling parties tended to erode rapidly following the collapse of the communist power monopoly, a decline in this context is perhaps predictable. In other post-communist countries, however, membership is falling in significant numbers among other parties as well, suggesting that the organisational erosion of the former ruling parties is not the only cause for the loss of party members. Many of the parties in these newly established democracies are now struggling to retain whatever remains of an already limited organisational presence on the ground.

The last set of data is shown in Tables 4 and 5, which analyse the changes in the membership ratios and the raw number of members in European democracies today, but this time marking the change since 1980 (early 1990s for the post-communist democracies). What emerges from these tables is that the scale of membership loss is quite staggering. A decline in M/E ratios is evident in each of the long-established democracies, ranging from a fall of more than 10 percentage points in the cases of Austria and Norway to more moderate decreases in Germany (−2.2) and the Netherlands (−1.8). Taking the 13 long-established democracies together, the average membership ratio has fallen by nearly 5 percentage points in the last thirty years. The raw numbers of members have also fallen dramatically. In the United Kingdom and France, the parties have lost around 1 million members over the course of the last three decades, equivalent to approximately two-thirds of the memberships recorded around 1980. Italian parties today have 1.5 million fewer members than their counterparts of the First Republic, corresponding to a fall of more than one-third of the earlier memberships. The Scandinavian countries – and Norway and Sweden in particular – have also suffered severe losses, with the raw numbers falling by over 60 and nearly 50 per cent, respectively. Although the losses appear more muted in some countries, in none of the long-established democracies have the raw memberships fallen by less than 25 per cent. On average, across all these polities, membership levels in terms of absolute numbers have been nearly halved since 1980.

Table 4.  Party membership change, 1980–2009
CountryPeriodChange in M/E ratioChange in number of members% change in number of members
Czech Republic1993–2008−5.05−379,575−69.65
United Kingdom1980–2008−2.91−1,158,492−68.42
Table 5.  The effect of size and the newness of democracy on party membership change
 Change since the 1990sChange since the 1980s
  1. Notes: Membership change measured as change in percentage points. Entries are OLS coefficients, bootstrapped standard errors in parentheses.

Small−1.26 (0.351)−2.11 (1.78)
New−0.179 (0.649)2.12 (1.53)

Even the parties in the post-communist democracies, despite their relatively short existence, record significant membership losses. As noted earlier, this is not simply due to the weakening of the former communist parties and their satellites, as is the case in the Czech Republic, where the absolute membership is down by nearly 70 per cent since the early 1990s. Hungarian parties have lost over a quarter of their original memberships since the transition to democracy, while party membership in Slovakia has been reduced by nearly a third. Taken together, these three post-communist parties have lost over 42 per cent of their original memberships and have seen their M/E ratios decline by an average of 2.3 per cent.

Only the three Southern European countries seem to be bucking the overall trend: the average M/E level has gone up by nearly 2 per cent since 1980 and the membership in raw numbers by an impressive 175 per cent, although this change should be qualified by underlining that the three polities started from an extremely low base in the years immediately following the democratic transitions. Significance tests of the group means confirm that there is indeed a statistically significant difference between Western, Eastern and Southern Europe in terms of both the change in M/E ratios (Anova sig. 0.007) and the percentage change in absolute members (Anova sig. 0.000) since 1980. Portugal records only a small rise in the raw number of members and this increase has been unable to keep up with the increase in the size of the electorate, causing the overall M/E level to drop. Greece and Spain record an increase in both the absolute and relative levels of membership since the 1980s. In fact, Spain is the only one among the newer democracies where party membership has grown consistently since the transition to democracy. Spanish parties now have considerably more members than they did ten or even thirty years ago, both in relative and absolute terms. Despite this constant growth, however, the membership level in Spain is still comparatively low and actually falls below the European mean.

As is shown in the first column of Table 5, the newness of democracy does not seem to produce any changes in the levels of partisanship in the period since the 1990s. This implies that the difference between old and new democracies helps to explain a more permanent and structural variation in the levels of party membership but does not account for changes in these initial levels over time. Size, on the other hand, exerts a substantial negative effect. Among small countries (with an electorate smaller than 20 million) we observe a further decline in the levels of party membership by 1.26 per cent. To the extent that this pattern is not an artifact of potential bottom effects (countries with greater initial levels have more room for higher rates of decline than cases with lower initial levels of party membership), it indicates that during recent decades the process of partisan de-alignment has been more exacerbated among smaller countries. Finally, the last column of Table 5 shows the equivalent differences over time but now comparing the current levels of party membership with those in the 1980s. To be sure, the limited sample for the post-communist countries (we only have data for Hungary, Slovakia and the Czech Republic) makes it virtually impossible to draw inferences from these findings, which are therefore only displayed for indicative purposes. It is interesting to observe, however, that the effect of size in predicting membership change in the established Western democracies since the 1980s, being statistically insignificant, is effectively zero.

As noted earlier, the potential unreliability of self-reported membership figures makes it pertinent, albeit very difficult, to assess the external validity of these figures. Before we proceed to discuss the implications of the observed patterns of membership decline, therefore, it is worth discussing our findings in light of the evidence available from mass surveys. Survey-based data appear to point to similar patterns of disengagement from party politics (e.g., Whiteley 2011; Scarrow & Gezgor 2010). However, both sampling error and the differential response rates of the group of party members in surveys are likely to raise problems for the external validity of survey-based data on party membership. Nonetheless, by comparing our membership data with the evidence from surveys we can engage in convergent validity, ‘the comparison of a measure against one or more measures that are also measures of the same concept’ (McDonald 2005: 944). For this purpose, we have used the results from the fourth round of the European Social Survey (ESS).7 We have chosen the ESS for three important reasons. First, it includes a separate question on party membership. Second, it includes 23 out of the 27 countries in our sample.8 Third, the most recent round of the ESS took place in 2008, which corresponds with the time period for which our membership data were collected.

With that in mind, we proceed to the results by turning to Figure 1, which presents the relationship between the aggregate party membership data we have collected with the average figures of survey responses in each country. The first panel of the figure presents the pattern for all countries, whereas the second excludes Cyprus, the only outlier (Austria was not included in the 2008 ESS). The observed pattern is reassuring: the two measures seem to co-vary substantially. The short-dashed line which presents the linear fit of the ESS measure on our membership data is very close to the 45-degree line that serves to denote the points of perfect convergence between the two measures (the long-dashed line). Even when a non-parametric fit is used, the local regression curve shown with the solid line in both graphs, the pattern is mostly monotonic. The correlation between the two measures is 0.811 (and 0.695 when Cyprus is excluded). Clearly, the two measures closely track one another in assessing the level of party membership across our sample of European countries (see also Scarrow & Gezgor 2010). The fact that most observations fall below the 45-degree line indicates that the survey measure tends to be higher on average than the membership levels reported by the parties, which might be accounted for either by the uncertainties that surround survey respondents' understanding of what party membership actually entails, or by a greater than average likelihood that party members will respond to surveys. Finally, it should be noted that, due to the high correspondence between the two measures of party membership, a replication of our analyses on the relevance of size and the newness of democracy confirm that all relationships identified hold when the party-reported data are replaced with their survey analogues, which suggests that the ESS data, coming from a very robust survey with a separate question on party membership as such, can be used very effectively in future research to probe the correlates of party membership.


Figure 1. The relationship between party and survey measures of party membership. Notes: The solid line presents a locally weighted regression curve, fitted in the scatterplot. The short-dashed line denotes the linear fit of the ESS measure on the self-reported membership data, and the long-dashed 45-degree line indicates the points of perfect correspondence between the two measures.

Download figure to PowerPoint

Implications and discussion

  1. Top of page
  2. Abstract
  3. Party membership in the twenty-first century
  4. Declining levels of party membership
  5. Implications and discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Appendix

There is scarcely any other indicator relating to mass politics in Europe that reveals such a strong and consistent trend as that which we now see with respect to the decline of party membership. It bears repeating that with the exceptions of Greece and Spain, democratising in the 1970s, all of the 19 European polities listed in Table 4 record a long-term decline in the ratio of party members to the electorate. In extreme cases (Austria, Norway), the decline is greater than 10 per cent; in others, it is around 5 per cent. All cases, with the exception this time of Portugal as well as Greece and Spain, also record a major long-term decline in the absolute numbers of members, a drop of 1 million or more in Britain, France and Italy, around half a million in Germany, and close to that in Austria. Britain, Norway and France have lost well over half their party members since1980, while Sweden, Ireland, Switzerland and Finland have lost close to half. These are genuinely striking figures, and suggest that party membership as such has, in general terms, changed in both its character and its significance.

Who would become or remain a member of a political party in the twenty-first century? From what we know of the members themselves – and there is now a growing literature based on professional intra-party surveys of members in many European polities (e.g., Koole & van Holsteyn 1999; Seyd & Whiteley 2004; Saglie & Heidar 2004; Spier et al. 2011; Gallagher & Marsh 2004; Pedersen et al. 2004; Bardi et al. 2008; Bovens & Wille 2011: 33–34) – they seem a relatively unrepresentative group of citizens, socially and professionally if not ideologically. The large majority, of course, are inactive – 6 out of 7 in the Danish case, for example (Pedersen et al. 2004) – and are members on paper only. In general, they also tend to be older and better-off than the average citizen, more highly educated, more likely to be associated with collateral organisations such as churches or unions (see also below), and more likely to be male than female. However, it is also important to recognise that they now seem less likely than before to be ideologically more extreme (Scarrow & Gezgor 2010), implying that a shrinking membership has developed into something that is more politically representative than was originally imputed by May (1973). The memberships also include a number of political professionals (public office holders, party workers, would-be party careerists, etc.), while in addition it seems that a large minority of younger members have clear political-professional ambitions (Bruter & Harrison 2009).

Scattered evidence further suggests that members are more likely than non-members to be connected in some way to the public sector and to state service. In the Seyd and Whiteley (2002: 40–45) survey of New Labour members in Britain, for example, 52 per cent of Labour members worked in the public or voluntary sectors, as against just 31 per cent of Labour voters. That said, this figure was actually lower than that recorded in their earlier ‘old’ Labour survey, which reported 63 per cent in the public or voluntary sectors. In the Dutch case in the 1990s, 23 per cent of the active members across all parties (and 17 per cent of the non-active members) were employed in the public sector, as against some 12 per cent of voters (Koole & van Holsteyn 1999: Table 1). Similarly, members of the public sector made up 35 per cent of German party members in 2009, while their share of the workforce was 7 per cent (Spier et al. 2011: 50). Data on this issue are relatively sparse, however, and while anecdotal evidence also suggests that the state sector accounts for a large proportion of party members, particularly on the left (teachers, professors, civil servants, social workers, etc.), precise data are not always available. In the EES 2008 data cited earlier, for example, 5.3 per cent of those with public-sector occupations claimed to be party members, as against 4.3 per cent in the private sector.9

More generally, the question is whether, having declined in numbers, these memberships have also become sufficiently distinct in terms of profile and activities that it might be reasonable to regard them not as constituting part of civil society – with which party membership has traditionally been associated – but rather as constituting the outer ring of an extended political class. In terms of the three faces of party organisation distinguished by Katz and Mair (1993), membership of this type – in terms of social profile, education and sectoral employment – might have more in common with the party in central office and even the party in public office than with the party on the ground. This would suggest, in turn, that the real party on the ground, to the extent it exists at all, would now stand outside the confines of the formal party and would be made up of the myriads of supporters, adherents and sympathisers.

The increasingly frequent adoption of broadly based primaries and membership ballots to select party leaders and candidates may also reflect recognition by the parties themselves that their active memberships are no longer representative. In the past, leadership selection was largely an elite process, and, even when more democratised, was largely the preserve of congress delegates and activist party members. Increasingly, however, parties are opening up these selection processes to party members at large or even, as in the case of the Italian left, to sympathisers in general (e.g., Kenig 2009; Lisi 2010; Seddone & Valbruzzi 2010; Cross & Blais 2011). The logic here seems to be that the more widely based the selectorate, the greater will be the chances that the candidates and leaders who are chosen will have a far-reaching appeal. In other words, if primaries are intended to broaden the base of leadership support, it makes much more sense to extend the opportunity for participation in these primaries beyond the party itself, and certainly beyond the narrow reach of the active members.

In this sense also, formal party membership becomes less and less important. Yet although some new parties try to eschew membership completely – as is the case of Geert Wilders's Freedom Party in the Netherlands – there are others that claim pride in building or at least reporting what looks like a mass membership. Moreover, there are also some established parties that remain highly committed to having real memberships (as opposed to donor lists) and that seek to engage these memberships in policy formulation, leadership selection and so on. The Italian Lega Nord offers a powerful counter-example to the general trends in this regard, deliberately seeking to build a tight and highly disciplined mass organisation on the model of the former Italian Communist Party, and, in the process, marking up one electoral success after another (e.g., Albertazzi & McDonnell 2010). Even though they are not very successful in this regard, German parties also remain committed to high levels of party membership because state subsidies are tied to their ability to generate membership contributions. The story in Europe, and even in the individual polities, is in this sense far from uniform.

Nonetheless, as is evident when one tries to gather data on membership levels, the large majority of parties seem relatively unconcerned about their memberships and are instead much more focused on reaching out to the wider public through professional campaigning and marketing techniques. That party membership has declined so enormously seems in many cases a matter of indifference to many of the party organisations concerned. This also suggests a more far-reaching conclusion: party membership levels have now fallen to such a low level that membership itself no longer offers a meaningful indicator of party organisational capacity. Alternatively, we might conclude that party organisations more generally have now reached such a low ebb that the formal organisational level is itself often no longer a relevant indicator of party capacity.

Such a conclusion inevitably calls into question our dominant way of thinking about party as a powerful organisational linkage between the mass public and the institutions of government. For many years the normative, theoretical and empirical literature on parties as organisations has been inspired by a concept of parties that was dominated by the mass party model. This model consisted of two organisational components – namely a powerful membership organisation, which is now clearly withering away, and an array of collateral organisations linked formally or informally to the party proper which provided for additional anchorage, strength and control of relevant segments of the electorate (see Poguntke 2005; Allern & Bale 2011). From this perspective, a party's own membership organisation is but a special variant of the linkage between parties and the wider society, and hence when we wish to assess party organisational strength as it has worked for most of democratic mass politics, we should also ideally complement our analysis with an investigation of how the membership of these collateral organisations has developed over time. Although this is clearly beyond the scope of this article, the evidence from existing literature suggests that here also decline is evident. If we take the two principal sets of collateral organisations that have proved most relevant for the mass party – the trade unions and the traditional churches – then the evidence of membership decline seems almost as dramatic as that in the parties.

In the classic case of trade union membership, for example, traditionally the strongest and most obvious of the party collateral organisations, there has been a substantial decline in membership over the past decades (Ebbinghaus & Visser 2000). While the figures from the 1980s still yield a somewhat mixed picture with a substantial element of stability, decline becomes the dominant mode in the 1990s. When we look at trade union density, which controls for changes in the size of the workforce, the downward trend is even more visible. Union density rates – comparable in this case to the M/E ratios for party membership – have been lower at the beginning of the new millennium than in 1970 in twenty countries, while in only four smaller European countries (Finland, Sweden, Denmark and Belgium) has there been a deviation from this otherwise universal pattern (Visser 2006: 45). Indeed, by 2005–2006 trade union density was lower than ten years earlier in all EU-27 countries except Malta (Eurofound 2008). Survey analyses validate these organisational data, with ESS data (rounds 1–4) showing declining trade union membership between 2002 and 2008 for 13 out of the 17 countries for which we report party membership data in Table 1.

The reach of the organised Christian religions has also declined substantially across Europe (Crouch 2008: 35; Girvin 2000: 23), although in this case the growth in Muslim populations in Europe has tended to compensate in holding the overall levels of religiosity fairly stable (see also De Vreese & Boomgaarden 2009). For the parties, however, it was the links to traditional Christian religions that mattered, particularly insofar as the churches could act as conduits between the parties and the religious communities.


  1. Top of page
  2. Abstract
  3. Party membership in the twenty-first century
  4. Declining levels of party membership
  5. Implications and discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Appendix

While political parties continue to play a major role in the elections and institutions of modern European democracies, it seems that they have all but abandoned any pretensions to being mass organisations. There are some parties, to be sure, that continue to emphasise the need for a strong membership and which cultivate close organisational links to local communities and constituencies. The Dutch Socialist Party is one such example, the right-wing Italian Lega Nord is another. In both Austria and Cyprus, moreover, political parties in general tend to maintain very large memberships, setting both polities ever more evidently as being at one remove from the normal patterns of party organisational development in contemporary Europe. Even taking account of these exceptions, however, what we see in these membership data is very concrete evidence of the sheer extent of party transformation in Europe since the 1980s. When data on party membership first began to be systematically collected and compared, the phenomenon itself was believed to matter. Members were believed to provide mass parties with a large proportion of their income and other organisational resources. They offered a valuable input into party policy making, not least in the lead-up to and usual conduct of national party congresses. By allowing the party to maintain a presence on the ground, they helped to legitimise party organisations and party campaigning. Finally, by offering commitment and loyalty, they constituted a fairly inexhaustible reservoir for candidate recruitment and office-holding obligations. The benefits provided by members were also evident in the scale of membership. In the early 1960s, according to the data collected by Katz et al. (1992), party membership constituted an average of almost 15 per cent of the electorate in the ten polities researched. In West Germany, the lowest ranking country, the figure was then just 2.5 per cent, and in Belgium, the next lowest, 7.8 per cent. The remaining cases ranged from over 9 per cent in the United Kingdom and the Netherlands, to 19 per cent in Finland and over 20 per cent in Sweden and Austria. This undoubtedly made for a very strong party organisational presence in European societies.

Today, the figures look wholly different. The average M/E ratio across Europe is just 4.7 per cent. Only two polities – Cyprus and Austria – score above 10 per cent. The figure of 7.8 per cent for Belgium in the 1960s, which placed it as second from last in the list of countries for which data were then available, would now warrant the fourth highest place in a list of 27 countries (after Cyprus and Austria, only Finland still reaches 8 per cent). What is more, the decline of mass organisations is not something that affects parties alone. Rather, as the data sources indicated above suggest, the other traditional pillars of organised mass society – the Christian churches and the unions – are also losing membership and clout, and are also shedding their once broad-based connection to the wider society. This also means that the world of collateral organisations may no longer be capable of offering a refuge to parties, and that it offers little potential for the parties to make up for their own declining memberships. Not only has the age of the mass party passed, but the conditions that fostered its development are also ceasing to prevail.


  1. Top of page
  2. Abstract
  3. Party membership in the twenty-first century
  4. Declining levels of party membership
  5. Implications and discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Appendix

A first version of this article was presented at the ECPR Joint Sessions in Lisbon in 2009. In addition to the participants of the workshop, we are grateful to George Mavrogordatos and Wolfgang C. Müller for their helpful comments, and to the editors as well as five anonymous referees of this journal for their constructive criticisms and suggestions. Elias Dinas offered invaluable help with the data analysis, while many colleagues in the profession proved very generous in providing data or ways to data. We alone are responsible for any surviving errors of data, analysis or interpretation.

  • 1

    In collecting the data for this article, we gratefully acknowledge the help of Thierry Barboni, Lars Bille, Gabriela Borz, Christophoros Christophorou, Fabrizio Di Mascio, Zsolt Enyedi, Wojciech Gagatek, Elin Haugsgjerd Allern, Jānis Ikstens, Alenka Krašovec, Evangelos Kyzirakos, Andreas Ladner, Lukáš Linek, Marco Lisi, Wolfgang C. Müller, Eoin O'Malley, Takis Pappas, Vello Pettai, Ainė Ramonaitė, Lamprini Rori, Marek Rybář, Maria Spirova, Jan Sundberg, Emilie Van Haute, Tània Verge, Liam Weeks and Anders Widfeldt.

  • 2

    This could be corrected by reducing the size of the electorate according to the electoral share of those parties for which membership data could not be included. A longitudinal analysis of the data reported in Katz et al. (1992) shows that the difference is not very large and declines over time as the data get more complete (Poguntke 2000: 222). However, as the share of minor parties has begun to grow in some countries this effect may have become larger in recent years.

  • 3

    See also Katzenstein (1985), who famously argued that there was a relationship between the size (and economic openness) of a polity, and the degree of corporatist concertation.

  • 4

    On the particular difficulties faced when looking for reliable data for France, see Billordo (2003). The remarkably low British figure (which excludes Northern Ireland) is probably more accurate and comprehensive than that reported in Mair and van Biezen (2001). Some of the parties have begun to report membership figures to the new Electoral Commission, and our data now also include the Scottish and Welsh nationalists and the far right BNP. The figure for the Conservative party in the United Kingdom, which strikes us as a rough estimate on the part of party headquarters, may not be so reliable.

  • 5

    Note that the effect of size also holds with other thresholds, such as when ‘small countries’ are defined as having fewer than 15 or 10 million eligible voters. These results can be made available on request.

  • 6

    Note that the Italian figure from 1996, against which this increase is measured, was remarkably low and that the current figure is little more than half that recorded in 1987 (see Appendix Table 1).

  • 7
  • 8

    The countries not available in the fourth round of the ESS are Austria, Italy, Ireland and Lithuania.

  • 9

    Public-sector occupations coded as those in central and local government, other public-sector or state-owned enterprises; private-sector occupations coded as those in private firms or self-employment.


  1. Top of page
  2. Abstract
  3. Party membership in the twenty-first century
  4. Declining levels of party membership
  5. Implications and discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Appendix
  • Albertazzi, D. & McDonnell, D. (2010). The Lega Nord back in government. West European Politics 33(6): 13181340.
  • Allern, E. & Bale, T. (2011). Parties and interest groups. Special issue of Party Politics, forthcoming.
  • Bardi, L., Ignazi, P. & Massari, O. (eds) (2008). I partiti italiani. Iscritti, dirigenti, eletti. Milan: Università Bocconi Editore.
  • Bartolini, S. & Mair, P. (1990). Identity, competition and electoral availability: The stabilisation of European electorates, 1885–1985. Cambridge: Cambridge University Press.
  • Biezen, I. van (2003). Political parties in new democracies: Party organization in Southern and East-Central Europe. London: Palgrave Macmillan.
  • Billordo, L. (2003). Party membership in France: Measures and data collection. French Politics 1(1): 137151.
  • Borz, G. (2009). Romania. In D.J. Sagar (ed.), Political parties of the world. London: John Harper.
  • Bovens, M. & Wille, A. (2011). Diplomademocratie: Over de spanning tussen meritocratie en democratie. Amsterdam: Bert Bakker.
  • Bruter, M. & Harrison, S. (2009). Tomorrow's leaders? Understanding the involvement of young party members in six European democracies. Comparative Political Studies 42(10): 12591291.
  • Christophorou, C. (2006). Party change and development in Cyprus (1995–2005). South European Society and Politics 11(3–4): 513542.
  • Cross, W. & Blais, A. (2011). Who selects the party leader? Party Politics, forthcoming.
  • Crouch, C. (2008). Change in European societies since the 1970s. West European Politics 31(1–2): 1439.
  • Dahl, R.A. & Tufte, E.R. (1973). Size and democracy. Stanford, CA: Stanford University Press.
  • De Vreese, C.H. & Boomgaarden, H.G. (eds) (2009). Symposium: Religion and the European Union. West European Politics 32(6): 11821283.
  • Ebbinghaus, B. & Visser, J. (2000). Trade unions in Western Europe since 1945. London: Macmillan/Palgrave.
  • European Foundation for the Improvement of Living and Working Conditions (Eurofound). (2008). Working time in the EU and other global economy: Industrial relations in the EU and other global economies, 2006–2007. Available online at:
  • Gallagher, M. & Marsh, M. (2004). Party membership in Ireland: The members of Fine Gael. Party Politics 10(4): 407425.
  • Girvin, B. (2000). The political culture of secularisation: European trends and comparative perspectives. In D. Broughton & H.-M. ten Napel (eds), Religion and mass electoral behaviour. London: Routledge.
  • Gunzinger, M. (2008). Analyse comparative des ressources financières des partis politiques suisses. Cahier de l'IDHEAP 240/2008. Lausanne: Swiss Graduate School of Public Administration.
  • House of Commons Library. (2009). Membership of UK political parties. Standard note SN/SG/5125, 17 August. Available online at:
  • Katz, R.S. & Mair, P. (1993). The evolution of party organizations in Europe: The three faces of party organization. American Review of Politics 14(4): 593617.
  • Katz, R.S. et al. (1992). The membership of political parties in European democracies, 1960–1990. European Journal of Political Research 22(3): 329345.
  • Katzenstein, P.J. (1985). Small states in world markets: Industrial policy in Europe. Ithaca, NY: Cornell University Press.
  • Kenig, O. (2009). Classifying party leaders' selection methods in parliamentary democracies. Journal of Elections Public Opinion and Parties 19(4): 433447.
  • Koole, R.A. & Holsteyn, J.J.M. van. (1999). Partijleden in perspectief. Kiezers en leden vergeleken. Jaarboek Documentatiecentrum Nederlandse Politieke Partijen 1999. Groningen: Universiteitsdrukkerij Rijksuniversiteit Groningen.
  • Kopecký, P. (1995). Developing party organizations in East-Central Europe: What type of party is likely to emerge? Party Politics 1(4): 515534.
  • Lewis, P.G. (2000). Political parties in post-communist Eastern Europe. London: Routledge.
  • Lisi, M. (2010). The renewal of the socialist majority: The 2009 Portuguese legislative elections. West European Politics 33(2): 381388.
  • Mair, P. (1991). Electoral markets and stable states. In M. Moran & M. Wright (eds), The market and the state: Studies in interdependence. Basingstoke: Macmillan.
  • Mair, P. & Biezen, I. van. (2001). Party membership in twenty European democracies, 1980–2000. Party Politics 7(1): 521.
  • May, J.D. (1973). Opinion structure of political parties: The special law of curvilinear disparity. Political Studies 21(2): 135151.
  • McDonald, M.P. (2005). Validity, data sources. In K. Kempf-Leonard (ed.), Encyclopedia of social measurement, vol. 3. Amsterdam: Elsevier.
  • Morlino, L. (1998). Democracy between consolidation and crisis: Parties, groups and citizens in Southern Europe. Oxford: Oxford University Press.
  • Niedermayer, O. (2008). Parteimitglieder in Deutschland: Version 2008. Arbeitshefte 13. Berlin: Otto Stammer Zentrum.
  • Pappas, T.S. & Dinas, E. (2006). From opposition to power: Greek conservatism reinvented. South European Society and Politics 11(3–4): 477495.
  • Pedersen, K. et al. (2004). Sleeping or active partners? Danish party members at the turn of the millennium. Party Politics 10(4): 367383.
  • Poguntke, T. (2000). Parteiorganisation im Wandel. Gesellschaftliche Verankerung und organisatorische Anpassung im europäischen Vergleich. Wiesbaden: Westdeutscher Verlag.
  • Poguntke, T. (2005). Political parties and other organizations. In R.S. Katz & W. Crotty (eds), Handbook of party politics. London: Sage.
  • Ramonaitė, A. (2006). The development of the Lithuanian party system. In S. Jungerstam-Mulders (ed.), Post-communist EU Member States: Parties and party systems. Aldershot: Ashgate.
  • Rozenvalds, J. (ed.) (2005). How democratic is Latvia? Audit of democracy. Rīga: University of Latvia.
  • Saglie, J. & Heidar, K. (2004). Democracy within Norwegian political parties: Complacency or pressure for change? Party Politics 10(4): 385405.
  • Scarrow, S.E. & Gezgor, B. (2010). Declining memberships, changing members? European political party members in a new era. Party Politics 16(6): 823843.
  • Seddone, A. & Valbruzzi, M. (2010). Primary elections, parties and participation: A comparative analysis of Bologna and Florence. Polis: Ricerche e studi su società e politica in Italia 2: 195224.
  • Seyd, P. & Whiteley, P. (2002). New Labour's grassroots: The transformation of the Labour Party membership. Basingstoke: Palgrave.
  • Seyd, P. & Whiteley, P. (2004). British party members: An overview. Party Politics 10(4): 355366.
  • Spier, T. et al. (eds) (2011). Parteimitglieder in Deutschland. Wiesbaden: VS Verlag.
  • Spirova, M. (2005). Political parties in Bulgaria: Organizational trends in comparative perspective. Party Politics 11(5): 601622.
  • Szczerbiak, A. (2006). Power without love? Patterns of party politics in post-1989 Poland. In S. Jungerstam-Mulders (ed.), Post-communist EU Member States: Parties and party systems. Aldershot: Ashgate.
  • Verge, T. (2009). Els partits polítics Catalans. In ICPS, Anuari Polític de Catalunya. Barcelona: Institut de Ciències Polítiques i Socials.
  • Visser, J. (2006). Union membership statistics in 24 countries. Monthly Labour Review, January, pp. 3849.
  • Weldon, S. (2006). Downsize my polity? The impact of size on party membership and member activism. Party Politics 12(4): 467481.
  • Whiteley, P.F. (2011). Is the party over? The decline of party activism and membership across the democratic world. Party Politics 17(1): 2144.


  1. Top of page
  2. Abstract
  3. Party membership in the twenty-first century
  4. Declining levels of party membership
  5. Implications and discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Appendix
Table Appendix Table1.. Summary data, by country
Country, yearElectorateTotal party membershipMembership as % of electorate (M/E)
  1. Note: Figures in italics are updated from Mair and van Biezen (2001).

  2. Sources: See Appendix Table 2.

19805,186,735 (79)1,477,26128.48
20086,107,892 (06)1,054,60017.27
19806,878,141 (81)617,1868.97
19897,039,250 (87)644,1109.15
20087,720,796 (07)426,0535.52
2002–20036,916,151 (01)443,6006.41
20087,129,965 (09)399,1215.60
Cyprus (Greek)   
2009501,024 (06)81,43316.25
Czech Republic   
19937,738,981 (92)545,0007.04
20088,333,305 (06)165,4251.99
19803,776,333 (81)275,7677.30
19893,941,499 (90)231,8465.88
20084,022,920 (07)166,3004.13
2002859,714 (03)28,7333.34
2008897,243 (07)43,7324.87
19803,858,533 (79)607,26115.74
19894,018,248 (87)543,41913.52
19984,152,430 (99)400,6159.65
20064,292,436 (07)347,0008.08
199939,215,743 (97)615,2191.57
200943,888,483 (07)813,5591.85
1980 (West)43,231,7411,955,1404.52
1989 (West)48,099,2511,873,0533.89
200761,870,711 (05)1,423,2842.30
19807,059,778 (81)225,0003.19
19988,862,014 (96)600,0006.77
20088,500,000 (07)560,0006.59
19998,062,708 (98)173,6002.15
20088,043,961 (06)123,9321.54
19802,275,450 (81)113,8565.00
19902,471,308 (89)120,2284.86
19982,741,262 (97)86,0003.14
20083,110,914 (07)63,0002.03
198042,181,664 (79)4,073,9279.66
198945,583,499 (87)4,150,0719.10
199848,744,846 (96)1,974,0404.05
200747,098,181 (06)2,623,3045.57
2003–20041,490,636 (06)10,9850.74
198010,040,121 (81)430,9284.29
200011,755,132 (98)294,4692.51
200912,264,503 (06)304,4692.48
19803,003,093 (81)460,91315.35
19903,190,311 (89)418,95313.13
20083,421,741 (05)172,3595.04
200028,409,054 (97)326,5001.15
200930,615,471 (07)304,4650.99
20008,673,822 (99)384, 4054.43
20088,944,508 (05)341,7213.82
200718,464,274 (08)675,4743.66
20004,023,191 (98)165,2774.11
20074,272,517 (06)86,2962.02
19981,588,528 (00)156,7019.86
20081,720,481 (07)108,0016.28
198026,836,500 (79)322,5451.20
199029,603,700 (89)611,9982.07
19806,040,461 (79)508,1218.41
19896,330,023 (88)506,3378.00
20086,892,009 (06)266,9913.87
19773,863,169 (79)411,80010.66
19974,593,772 (95)293,0006.38
United Kingdom   
198041,095,490 (79)1,693,1564.12
198943,180,573 (87)1,136,7232.63
199843,818,324 (97)840,0001.92
200844,245,939 (05)534,6641.21
Table Appendix Table2.. Membership levels, individual parties
200695,28283,18072,47558,44330,68628,35325,09010,4116,2845,4975,4624,890 (07)
2008191,00095,621 (06)10,00050,000 (09)37,500 (09)15,000
200940,42419,30413,941 (05)7,000 (07)764
Czech Republic
2000128,34655,30655,000 (96)18,43217,0792,9331,889
200877,11538,31229,429 (07)17,569 (07)3,000
2009277,000232,000 (08)130,00060,00050,00022,000 (06)15,000 (06)10,000 (06)9,1238,436
2008350,000 (05)210,000
2007615,414 (06)605,344448,567 (06)401,014231,260 (06)148,38387,63651,03634,650
200413,00012,90012,269 (03)5,9004,5004,5001,500
200968,10256,45650,44436,15926,96220,32426,74512,3056,972 (08)
Poland (continued)
2000131,504124,61187,29040,0001,000 (99)
2008725,670 (07)552,689 (06)58,000 (07)29,748 (07)51,76212,35076,127*18,993*5,464*
2008100,63954,85847,866 (06)17,79922,91910,500 (07)9,1103,300
United Kingdom
YearConLabLib/LDSNPUKIPBNPGreensPlaid Cymru