There has been a very dramatic fall in party membership in New Zealand: from nearly 24% of the electorate in the 1950s to only 2% in the 1990s. This spectacular collapse began in the 1960s and, despite a recovery in the late 1970s and early 1980s, has continued to decline. This has meant that the political parties in Parliament are now low-membership, cadre-type institutions, which are more reliant on the resources of the state and on business donations. This blog post describes the aggregate decline in membership numbers. [Read more below]
It is important to clarify exactly what the definition of political party membership is. This allows a better idea of the validity of party figures. For the purposes of this blog post, a distinction is made between those party members who have joined the main body of the party – direct members – and those that belong to an affiliated organisation – indirect members. This has an important implication for the membership figures: for if the affiliate trade union membership of the Labour Party is included (which currently pay fees at about ten% of the cost of ordinary membership) then in 1986, for example, the Labour Party could be said to have had about 250,000 members, whereas the party claimed only 65,000 members.
It is less common in contemporary New Zealand politics, but some Western political parties have affiliated organisations that appeal to particular groups – typically trade unions. For the purposes of this series of blog posts, only the direct members count as full party members, although indirect/affiliate members are also of some interest. The party literature generally regards indirect/affiliate members as problematic, mainly because, ‘An unknown proportion of the indirect members may, in fact, be quite ambiguous in their allegiances to the party. There may be cases of indirect members who support other parties’ (Widfeldt, 1995: p.137).
Even the figures for the direct membership of the parties are questionable. All political parties are prone to exaggerate the size of their memberships, as the parties like to ‘maintain at least the image of a mass party’ and members in this sense legitimise the parties, showing that they are ‘viable channels for political representation’ (Mair, 1997: p.148). Hence, there are also often reliability problems with the claims of the parties about their membership size. In addition, internal party records are also prone to mistakes, such as multiple recording of individuals.
A loose definition of membership may have been responsible for the incredibly high membership numbers reported by the National Party in the 1950s and 1960s and the Labour Party around the early 1980s. Especially in the case of National, there is some evidence that the simple act of making a small donation to a party canvasser was regarded as bestowing the status of membership. In this sense, there is often a blurring of the separation between the categories of supporters and members. Also in the case of National, often whole families have been enrolled as members, which might suggest a less than robust and meaningful concept of membership.
Obtaining membership number information
There are great difficulties in obtaining internal information about political party membership numbers. Because the parties operate in an obviously very political environment, they function as secret societies, on guard against the preying of their opponents. As numbers decline and remain low, the New Zealand parties have become even more guarded about their current membership statistics.
There have been times, of course, when the parties have declared their membership figures, or they have been leaked. For instance, in 1985 the Royal Commission on Electoral Reform requested and received information from all the main political parties about internal information such as membership numbers. There are also a number of newspaper reports and interviews where party officials have been quoted on membership numbers, and the Electoral Commission requests information from the parties on membership numbers every three years at election time (when allocating broadcast time and money).
The validity of these figures, however, is often questionable. Not only do parties deliberately exaggerate their figures, but some membership figures include people who are merely in affiliated organisations, while other figures include people who have ceased to pay their party dues, and some membership numbers are really only the figures of those people on a party’s general mailing list of interested or sympathetic voters.
The meaning of the term ‘party membership’ and the criteria for membership is also contested. Both the Labour and National parties have claimed that they only allow a person into their parties after they are accepted by a vote in the party branch in which they live, but the standard practice for enrolling as a member is simply to send in a signed membership form with the necessary membership fee. Furthermore, nearly all the political parties allow people to join up through the party’s website.
There have also been some academic and state surveys which enquire about party membership. Such surveys must also be taken with caution as this method usually results in inflated membership numbers. According to party scholar Anders Widfeldt, this is because, ‘it is likely that people who are not willing to co-operate in a survey are also less likely to be party members, and thus are over-reported in the sample loss; in other words, a risk that the proportion of members is exaggerated’ (Widfeldt, 1995: p.141).
There is some debate in the political party literature about how best to measure and compare membership numbers. This blog post is mainly concerned with the raw membership numbers. According to Scarrow, using ‘absolute membership figures may be best for judging the aggregate strength and political capacity of grass-roots party organizations, showing whether parties have sufficient supporters to maintain some kind of presence at the local level’ (Scarrow, 2000: p.87). However, there is sometimes a need to frame membership figures for some comparison, and so this blog post also makes use of membership measured by the ratio of members against votes received by the party (M/V). According to Scarrow, ‘if the aim is to compare particular parties’ success in providing supporters with an organizational home, it makes sense to standardize membership in terms of party vote’ (Scarrow, 2000: p.87). Some authors, such as Katz and Mair et al. (1992: p.331) ‘have dismissed the first two measures as unsatisfactory because they do not take adequate account of the context of party development and the relationship between organizational and electoral success’ (Scarrow, 2000: p.87). Richard Katz and Peter Mair suggest that membership be framed as part of the overall electorate, by measuring the ratio of members against the whole electorate (M/E). This is the third measure this blog post uses. Scarrow says that ‘If the aim is to assess the extent to which parties provide outlets for citizen political participation, it makes sense to standardize party membership in terms of the total electorate’ (Scarrow, 2000: p.87).
For Scarrow’s compilation of the New Zealand figures for this last method (M/E), see Figure 6.1 (which is based on the membership numbers in Table 6.1, which shows the combined membership figures for all the parties for the last five decades). This data suggests a very dramatic fall in party membership in New Zealand, from nearly 24% of the electorate in the 1950s to only 2% in the 1990s. According to Scarrow’s international comparisons, New Zealand parties have been afflicted by a greater membership decline than in any other of the 14 OECD countries she studied.
Table 1.1: Aggregate Party Membership (Scarrow Data)
Year Aggregate Numbers
Source: Scarrow (2000: p.89), using membership information from Chapman et al. (1962), Levine (1979), Milne (1966), and Mulgan (1994).
A slightly less pronounced and steady decline is plotted by Jack Vowles (2002), seen in Figure 6.2 (based on the membership numbers in Table 6.2). This data shows that about 22% of the electorate was enrolled in parties in 1954, and this dropped by about half in the early 1970s, and then declined further in the 1980s and 1990s. Vowles’ membership figures are more detailed than Scarrow’s, and generally they are very credible, except for the 1996 and 1999 figures, supplied by the Electoral Commission which are probably over-stated. The 20-odd parties registered with the Commission together claimed 153,000 members in 1996 and then 132,890 members in 1999. Unfortunately, these totals include Labour’s affiliate membership, which is not included in the previous figures, and probably should not be compared. Therefore included in the Electoral Commission’s figures of 153,000 and 132,890 were about 40,000 unionists who were ‘members’ of the Labour Party by way of union affiliation rather than being direct members. Consequently, the number of New Zealanders that were direct branch members of political parties in 1999 was probably more like 113,000 in 1996 and 93,000 in 1999. As a proportion of 1999 eligible voters, therefore, party membership density was probably about 3.3% instead of 4.8%, as claimed by Vowles.
Table 1.2: Aggregate Party Membership (Vowles Data)
Year Aggregate Numbers
Source: Vowles (2002: p.416), using membership information from Gustafson (1986; 1997), Miller (1989), RCES (1986), Vowles (1994), and Harris (1997).
Survey research also confirms these trends, and suggests some similar levels of party membership. Figure 6.3 includes the results of surveys carried out by (1) Austin Mitchell in 1966, (2) Stephen Levine and Nigel S Roberts over the 1970s, 1980s and early 1990s, and (3) by the Electoral Commission in 1994 and 1997. Mitchell’s 1966 survey suggested a membership of 28% of the electorate. According to the survey work of Levine and Roberts, about 20% of the electorate belonged to parties during the 1970s and first half of the 1980s, but by 1990 this had dropped to only 14% (Levine and Roberts, 1992c: p.71).
Since then the Electoral Commission’s research has indicated numbers to have declined further to 6.5% in 1994, and then 4% in 1997. The survey also indicated that 19.5% of the voting-age population had previously been members of parties (Harris, 1997a: pp.15-16). The survey data is anomalous with that of Scarrow and Vowles because the decline in numbers appears to occur much later in time. This could be the result of a time lag expressed by the survey method, as many of those surveyed may have claimed to be party members long after their membership had lapsed.
The downward trend in membership numbers during the late 1990s (and through to 2002) is also confirmed by aggregate membership figures released by the Electoral Commission for Labour, National, the Greens, the Alliance, Christian Heritage, United Future, and the Aotearoa Legalise Cannabis Party. Since 1996 when the Electoral Commission first began allocating election broadcast funding, all these parties have supplied membership information to the Commission, and thus it is possible to compare the aggregate figure for these seven parties over the three election period. The Commission includes the Greens in this three election comparison on the basis that its membership was combined in the Alliance’s 1996 figure, and it includes United Future on the basis of including the memberships of its component parties in the 1996 and 1999 figures. Excluded from the comparison are parties such as Act, New Zealand First, and the Progressive Coalition Party, all of whom did not submit membership numbers for all three elections. As shown in Table 1.3 below, the aggregate membership for all seven parties has declined significantly over the three elections, from 115,093 in 1996 to 83,944 in 2002. It should be noted that these figures include Labour’s affiliate members.
Paul Harris, the chief executive of the Electoral Commission, comments further on these figures:
Of the 7 parties for which we have information for both 2002 and 1999, 4 had falling memberships and 3 had increasing memberships. Of the 6 parties (i.e. excluding the Greens) for which we have information for both 1999 and 1996, 4 had falling memberships and 2 had increasing memberships. Of the 6 parties for which we have information for 2002 and 1996, 4 had falling memberships and 2 had increasing memberships (Harris, 2003).
Table 1.3: Aggregate Membership for Seven Parties – Electoral Commission
Year Aggregate Numbers
Source: Harris (2003).
The following blog posts detail the quantitative and qualitative changes in the memberships for the individual parties.