Office of the Deputy Prime Minister

Turnout at Local Elections

CONCLUSIONS

The programme of research described in this report constitutes the most extensive analysis of local electoral procedures and turnout ever undertaken in Britain. It has used primary data drawn from a variety of sources together with a number of different analytical techniques to build on and extend knowledge obtained from previous enquiries. The research comprised four stages:

  1. A postal questionnaire sent to all local authorities in England responsible for compiling the electoral register asking for responses on a wide range of topics relating to electoral processes and local election participation;

  2. A comprehensive and rigorous statistical analysis of the relationship between local election turnout and a wide range of political, structural, socio-economic and geographical variables;

  3. Case study research in 8 local authorities displaying levels of turnout higher or lower than expected;

  4. A review of secondary material on electoral processes and patterns of participation at sub-national elections abroad.

The four stages of the research were explicitly designed not only to stand alone as individual projects but also to inform and shape each other. The survey of local authorities conducted at Stage 1 produced variables that were used in the statistical analysis at Stage 2 as well as providing an overview of English electoral practice to be compared with the international experience in Stage 4. The model of turnout at Stage 2 enabled the identification of authorities whose turnout was higher or lower than expected and these formed the basis for the selection of case studies at Stage 3. Those case studies, in turn, provided richer information on local procedures than was possible to glean from the Stage 1 postal survey and highlighted additional variables that may prove to be important in further refining a model of turnout (Stage 2) in the future.

The research had five discrete objectives:

We will describe here how each of these objectives has been met.

LOCAL AUTHORITY INITIATIVES AND PRACTICES

All local authorities in England responsible for compiling the electoral register were sent a detailed questionnaire seeking information on their current initiatives and practices aimed at enhancing electoral registration and participation. Responses were received from 71% of councils, broadly representative of the types and pattern of political control in English local government - see Appendix 1. The survey provided clear evidence that many local authorities continue to strive to improve the process of electoral registration and to facilitate voting among their electorate. There remain wide variations in the extent to which initiatives have been adopted, but more local authorities than at the time of a previous survey in 1996 (Rallings, Thrasher and Downe 1996) now appear to be committed to modernisation in this area.

The range of measures adopted by local authorities is very wide. The vast majority have tried to streamline the information available to those registering to vote. More than 80% of local authorities have made significant design changes to the voter registration form (Form A) to encourage electors to recognise and return it. A similarly large number have taken steps to publicise the absent vote facility. However, fewer authorities have thought it appropriate to introduce initiatives designed to encourage people actually to use their vote. Although more than half of authorities have taken steps to ensure that as many polling places as possible have disabled access, both the survey respondents and our case study authorities have noted an increased difficulty in finding suitable and convenient polling places. There has been an increase since 1996 in the number of local authorities that actively publicise forthcoming local elections. A good deal of such activity is conducted within multi-authority consortia and/or aimed at young voters. A list of good practice ideas drawn from the survey is contained in Appendix 2.

About one in two local authorities reported that registration and/or turnout had been the subject of formal discussion at committee or whole council meetings. Again, this is an increase on the figure for 1996. These deliberations have varied from the confirmation of electoral registration budgets to the adoption of corporate policies and targets on local election turnout. Several authorities, both in the survey and during the case study interviews, accepted that electoral registration was a function amenable to audit and that the service would be included within their 'best value' regime. There was less willingness to take responsibility for turnout, although a few authorities had committed themselves to targeted increases in participation.

There was widespread interest among local authorities in piloting new electoral arrangements. The single most popular experiments were universal postal voting and electronic voting. Both the survey and the case studies showed how the degree of enthusiasm displayed by authorities for different reforms was often related to their own socio-geographic circumstances. For example, the provision of electronic voting in a rural area would bring practical difficulties and be unlikely to increase turnout if it still involved a long journey to a polling place.

IMPACT ON REGISTRATION AND PARTICIPATION

Although local authorities found it easy to list the initiatives they had undertaken to try to enhance registration and participation, they were less clear about their impact. There was some consensus that the introduction of new procedures had initially led to increased registration. Several authorities were able to pinpoint such increases quite precisely, whether they came in the wake of a newly designed Form A or following the more rigorous chasing-up of non-respondents. Our evidence suggests that the more local authorities follow 'best practice' in this area, the more accurate will be their electoral registers. However, some respondents did point out that this process could result in diminishing returns, as the number of additional electors recruited did not justify the levels of expenditure involved. Similarly, many survey respondents could quantify the success of campaigns to encourage absent vote applications. Even taking into account the fact that simply applying for an absent vote betokens some interest in the electoral process, it does seem significant that the turnout among absent voters is often twice as high as among those obliged to vote in person. That agrees with evidence from other countries that have moved towards postal voting.

There was greater scepticism, however, about the impact of local authority based initiatives to increase turnout. Some authorities claimed that they had increased awareness about local elections even if not turnout, but little systematic analysis seems to have been undertaken. Many were equally concerned to point out that important determinants of turnout were outside their control. These included the salience of local issues and the degree of campaigning activity undertaken by political parties. Our respondents believed that such specific factors, together the general political climate, weighed more heavily with individual electors than any general exhortations to vote made by the local authority.

It was also argued, both in relation to current initiatives and to putative experiments with new procedures, that any impact would be seen only over the long term. There was certainly no evidence in the analysis in Stage 2 of a statistical relationship between those local authorities identified as most keen to encourage either wider public participation and/or local election turnout and actual turnout levels in 1999.

KEY DRIVERS OF LOCAL ELECTION TURNOUT

A core concern of the research was to determine precisely how we could account for different levels of election turnout in different local authorities. A statistical analysis of voter turnout was conducted using a dataset containing a range of election data together with socio-economic variables for wards in England during the period 1973-1999 inclusive. The dataset included variables on local authority electoral process initiatives drawn from the Stage 1 survey and on techniques used to encourage wider public participation drawn from a 1998 DETR survey (Lowndes et al. 1998). The technique of multiple regression analysis was used to explore the relative impact of structural, socio-economic, political and geographical factors on voter turnout at the ward level.

Initial analysis of the data revealed the cyclical pattern over time in voter turnout to be broadly similar across types of local authority. In other words, the reduction in turnout seen in recent years was a characteristic of all types of authority and places. Within this pattern, a number of political and structural variables were shown to have a positive influence on turnout. These include:

In addition to these factors, an equation based solely on 1991 data showed a strong negative correlation between voter turnout and population migration. In other words, the more stable a ward's population base, the higher its turnout. There was, however, very little sign of a positive association between a wide range of local authority participation initiatives and electoral practices with voter turnout.

The influence of socio-economic variables on turnout appeared to vary depending on the regression method adopted. In one set of analyses it did appear that the proportion of pensioners and of ethnic minorities had a strong positive impact on voter turnout. Positive but smaller coefficients of association were also found for professionals, owner-occupiers, council tenants, students and the self employed. Conversely, the proportion of unemployed, skilled and manual workers, households in crowded housing and with no bathroom was negatively associated with voter turnout. The strongest of these socio-economic effects was that wards with higher proportions of young people had lower turnouts.

However, using a different regression technique, much of the impact of socio-economic effects on voter turnout appears to have less influence than previous studies have suggested. This is because our preferred statistical approach indicates that previous estimates of socio-economic effects may have been statistically biased. Moreover, it is also the case that we did not have actual socio-economic data and that most of the socio-economic observations in the dataset were based on extrapolations or interpolations from the 1981 and 1991 Census. Thus, some caution must be taken with the interpretation of the socio-economic results reported in the Stage 2 study.

By applying the model of turnout to all wards in the dataset it was possible to generate league tables of wards which were high/low performers, even after taking into account their political, structural and socio-economic condition. These tables, and the similar ranking of local authorities, were used to select our case studies for the Stage 3 analysis. We wished to explore why the turnout in these authorities was higher or lower than expected and to build up a picture of the political culture of each authority and its local area which appeared to behave atypically.

It is interesting to note that both the survey and the case studies highlighted factors which local authorities themselves thought had had an influence on turnout, but which could not be operationalised in the statistical model. Some of these drivers were particular to individual years or wards and therefore likely to be discounted in the fixed effects model. The importance of electoral marginality was acknowledged, but this was tied to party political activity. In other words, voters are unlikely to know how marginal their ward is unless they receive a (surrogate) indication through the campaigning activities of the political parties. Parties will have different levels of activity in different places regardless of the objective electoral situation. In the case studies there was a unanimous perception that party activity had decreased in recent years and that campaign effort was focused on only a few wards.

There also appeared to be differences between case study authorities in terms of the role of the local media and the existence of a strong local identity. In some case study authorities the continued existence of a well-read, community based local paper which gave coverage to local issues and to elections was felt to boost public interest and participation. In other authorities the media were less localist in orientation and penetration. Similarly, the findings for 1991 on the positive relationship between population stability and turnout were supported by the case studies. Authorities whose population had become more mobile, and those that had reason to think that levels of identity with the local council were weak, believed that such factors had adversely affected turnout.

LESSONS FROM ELECTORAL PRACTICES AND TURNOUT ABROAD

International comparisons show that significant differences exist between countries in the areas of electoral registration, official responses to low turnout and overall levels of turnout. Making electoral registration voluntary has a major impact on the coverage of the subsequent register. Those reluctant to register, whether the system is voluntary or compulsory, appear to be similar across different countries. The problems are greatest amongst the poor and the young alike.

Various initiatives to improve electoral participation have been introduced. The most extreme method for raising levels of turnout is to make voting compulsory but in practice very few liberal democracies resort to this policy. A number of countries have sought to introduce measures designed to facilitate voting in sub-national elections. Most of the initiatives we identified can be described under the broad heading of 'bringing the electoral process closer to the people'. Some countries, for example, Norway and some states in North America, now have early or advanced voting where voters may cast a ballot in the weeks leading to the election. The practice of absent voting, principally postal voting, has proved an important and successful initiative in raising levels of turnout. The evidence shows that, in Britain, the United States and elsewhere, electors that apply for a postal vote are more likely to vote than those that are required to vote at a polling station. There have also been recent, though limited initiatives in the field of electronic voting in such countries as Belgium, New Zealand and the United States. Allowing people to vote at remote computer terminals, whether situated in public places or indeed within their own home, seems destined to be a significant development in making the electoral process more 'user friendly'.

The comparative analysis of sub-national electoral turnout showed it to be in general decline, regardless, it seems, of structural, functional, social and cultural issues. Few countries have been immune to the process of growing voter apathy. The fact that other countries have been similarly affected, however, means that declining turnout is not solely a British malaise. Policies that are designed and adopted to tackle low turnout need to be aware of the wider context to this problem.

POLICY IMPLICATIONS FOR LOCAL AUTHORITIES AND CENTRAL GOVERNMENT

The detailed material reviewed here contains a number of lessons about how local authorities and central government may continue to pursue a goal of greater participation in local elections. It also, however, contains a warning that poor levels of turnout are a reflection of wider attitudes in society and that solutions to the underlying problem are unlikely to be found simply through changes in procedures and systems.

[ Previous ] [ Contents ] [ Next ]

Published 3 May 2000
Return to Local and Regional Government Index
Return to Homepage
Web site terms