Written evidence submitted by Dr Maria Sobolewska, University of Manchester, and

Professor Anthony Heath, Universities of Manchester and Oxford (VUK 30)

The case of ethnic minorities

Executive summary:

• The main challenge for BAME voter engagement in the UK is registration.
• Once registration differential is accounted for, turnout among minorities does not significantly depart from the national average.
• Some responsibility for lower participation among BAME groups must fall to political parties who contact minority voters less during the electoral campaign than they do white majority voters.
• The main obstacle to registration is socio-economic marginality: with youth and people living in privately rented accommodation particularly excluded, among white British as well as among BAME groups.
• BAME groups therefore experience the same reasons for under-registration as white people, but the concentration of these reasons among BAME residents is greater than among the general population.
• Self-identified BAME- specific reasons for lack of registration are knowledge of eligibility and the registration process.
• The obstacles to registration will persist and some may get (at least temporarily) worse under the new registration system.
• On-the-day registration has been shown, in the United States of America, to improve overall registration levels, so while this measure would not specifically address the low BAME registration rates, it would benefit the whole of the electorate.
• Tackling the effects of socio-economic marginality would be the best way to address under-registration of BAME groups.

Introduction

Definition of Ethnic Minorities

1.1 For the purposes of this evidence we define ethnic minorities as immigrant-origin, visible minorities of African, Asian, or Caribbean descent, otherwise described as Black and Asian Minority Ethnic (BAME) groups. This is a standard definition and is used both in academic and public debate. Since data is scarce on some of the smaller minorities of East Asian descent, we will particularly focus on the five main groups of African, Caribbean, Indian, Pakistani and Bangladeshi origin.

Why should we care about voter engagement of ethnic minorities in the UK?

1.2 Participation in elections is fundamental for the proper functioning of a representative democracy. While overall trends for falling participation are worrying on their own as they raise questions about the legitimacy of governments and the dissociation of citizens from the political process, lower levels of participation among certain social groups is particularly worrying. This means that the political opinions and interests of many electors will not be represented in the democratic process. What is of special importance is the source of this exclusion. Are the lags of participation among certain groups a result of self-exclusion, for example because of personal feelings of disinterest in a process that seems far removed from citizens’ everyday life (which could be interpreted as a result of a smooth running of government), or a direct result of structural and societal obstacles: i.e. external, rather than self-exclusion.

1.3 In this context there is particular concern about the participation of immigrant-origin minorities.  These groups most obviously suffer some structural disadvantages such as limitations on their eligibility (particularly among recently arrived migrants), distribution in mostly urban, relatively deprived and ‘safe’ constituencies which see little political mobilisation (Sobolewska, Fieldhouse and Cutts 2013), and direct as well as indirect prejudice in terms of access to political office (Durose et al 2013).

1.4 An additional danger arising from these groups’ under-participation, as Ed Fieldhouse and David Cutts so aptly put it: “Inequalities in participation may be regarded as especially problematic if a minority group has distinctive political attitudes, preferences or interests.  Low levels of participation may lead not only to a failure to represent these interests, but also undermine the legitimacy of the state in the eyes of the minority group” (Fieldhouse and Cutts 2008, p.333).’

Our source

1.5 Most of the material used in this evidence comes from the Ethnic Minority British Election Study (EMBES), which we conducted after the last General Election in 2010. EMBES is the first nationally-representative probability sample of the ethnic minority electorate in Britain since 1997, and by far the largest survey of political attitudes of British minorities. The sample size is 2787: Indian 587, Pakistani 668, Bangladeshi 270, Caribbean 597 and African 524. It also contains a core of questions directly comparable with the British Election Survey (BES) run simultaneously on a sample of the general population. As a result we are able to compare electoral engagement of minorities with the general population. In addition, for our purposes of talking about political engagement, we were able to check whether respondents actually were registered to vote against marked –up Electoral Registers, which are publicly available after the election. As a result we can look at the validated, rather than the less reliable self-reported turnout at elections. Unfortunately we did not have the resources to check the registers at other addresses to those where respondents were sampled, so we have to recognize that there will be some uncertainty about the total valid rate of registration.

1.6 We also use our existing publications and other published research, details of which are available in the bibliography.

Are minorities less engaged than the general population?

2.1 Voter engagement is a two-stage process. Firstly, eligible voters need to register to vote; secondly, they need to turn out on the Election Day to return their vote. Evidence gathered at the last elections shows that minorities face obstacles to engagement at the first, registration stage.  Before we proceed to delve into the obstacles however, it is important to put any shortages of participation in perspective: internationally, Britain’s ethnic minorities do have an extraordinary access to voting. The eligibility to take part in the electoral process in all countries rests on the condition of citizenship- in Britain it is not exclusively British citizenship, but any Commonwealth and Pakistani citizenship that carries eligibility to vote in all forms of elections in the UK. As a result most of Britain’s ethnic minorities are eligible to vote and, unlike in most other countries, eligibility will not be the main obstacle to participation for the British BAME population.

Registration

2.2 Table 1 shows the patterns of registration among those BAME respondents who were British or Commonwealth and Pakistan citizens and who thus were eligible to register.  The third column in this Table shows the proportion of those who were not registered at all.  We can see that for every minority group the figure is significantly higher than the white British one, being particularly high for the black Africans. This group contains the largest proportions of recent immigrants and so this result is not entirely surprising. What is worrying however is that these figures are remarkably close to estimates from twenty years ago:  an analysis of 1991 data by the Office of Population Censuses and Surveys- now the office for National Statistics (Smith 1993). While the methodology used by OPCS and EMBES differs, which means direct comparison is not perfect, it does seem that despite the valiant efforts of organizations like Operation Black Vote, which runs campaigns to encourage ethnic minority registration, and the Electoral Commission, the problem remains undiminished.

Table 1  Registration among those eligible (British and Commonwealth citizens)

Row percentages

Sources: BES 2010, EMBES 2010, weighted (Heath et al 2013)

Notes:  Excludes those for whom details were not sufficient to identify in the registers. Figures in bold are significantly higher at the 5% level than expectation.  Chi2 = 188.8 ( 12 df), p < 0.001.

2.3 We asked our respondents why they were not registered. The reason most often given was that they were not eligible (28 per cent) although quite a number of these were almost certainly eligible since they were Commonwealth citizens.  Next most common was ‘I have recently moved’ (24 per cent).  Other reasons were ‘I’m not interested in elections’ (10 per cent), ‘I didn’t know how to register’ (8 per cent), ‘I never received the application form’ (8 per cent), ‘I got left off by mistake’ (8 per cent) and ‘I just couldn’t be bothered’ (6 per cent). Sadly, we were only able to ask those who told us that they were not registered – for those who said that they were registered, but turned out not to be after we checked the Electoral Register, we do not have this data.  

2.4 Therefore, we also run an analysis of those who were not registered to try to explain how they differ from those who were registered. We found that for all respondents registration was linked to an individual’s age, housing tenure (people renting their accommodation privately were the least likely to register) and the sense of a duty to vote. In fact, these predictors accounted for a large proportion of the minority registration deficit. To be sure, some differences remained, especially in the first generation, although these too were explicable in fairly conventional ways.  Thus, we found that fluency in the English language was a significant predictor of the adult migrants’ likelihood of registering, while the number of years since arrival was of borderline significance (Heath et al 2013). 

2.5 Thus, every single one of the relevant predictors revealed by our statistical analysis, apart from duty to vote, involves some degree of social or economic marginality. Young people, who are more common among BAME groups, and recent arrivals in Britain are marginal in British society.  People who are not fluent in the English language are clearly marginal. People dependent on privately rented accommodation are marginal when compared with home owners or local authority tenants, as private rent offers the least security of tenure and quality of accommodation. Since more of the people of BAME origin fall into all these categories, even among seemingly successful groups such as Indians, we have to conclude that they are more vulnerable to marginalisation.

Turnout

2.6 The second stage of voter engagement is returning the vote on Election Day. Table 2 shows the percentage of those registered electors who returned their vote and shows that unlike in the case of registration, turnout rates were pretty similar among different BAME groups and the majority.  Only the group of mixed black and white heritage had a statistically significantly lower rate of turnout than the white majority. While previous research already showed that South Asians turn out to elections in rates equal to the white majority, we must highlight here the equally high turnout among the black Caribbean group, who were identified by previous research as having lower rates of turnout.

Table 2  Validated turnout (validated against a marked-up Electoral Register)

Percentages

Sources: BES2010 and EMBES 2010, weighted (Heath et al 2013)

Notes: Measure of validated turnout includes those issued with postal/proxy votes but excludes those not registered at the sample address.  Figures significantly different from expectation at the 5% level are emboldened.  Chi2 (validated turnout) = 10.6 (6 df), p > 0.05;  Chi2 (self-reported turnout) = 65.4 (6 df), p < 0.001; Chi2 (over-claiming) = 34.5 (6 df), p < 0.001;

2.7 We analysed all the BES and EMBES respondents who were registered to vote to see what processes played a role in ethnic minority turnout.  We found that strength of party identification and a sense of duty to vote were associated with increasing turnout among ethnic minorities just as they are among the white British. However, we also saw some differences: political interest and political knowledge were highly significant among the white British but not significant at all among minorities. We also found that people of all ethnic origins who told us that they had been canvassed by one of the political parties showed a significantly higher likelihood of going to the polling booth (Heath et al 2013). This is of extreme importance, as we also found that political parties were LESS likely to canvas among minorities (Table 3). Our other research shows that, even after accounting for electoral competitiveness of the seat, relative deprivation of the area and other predictors of canvassing activity, this under-mobilisation among minorities- especially those who do not live in high co-ethnic concentration- persists (Sobolewska, Fieldhouse and Cutts 2013).

Table 3:  Canvassing (all respondents)

Column percentages

Sources: BES 2010, EMBES 2010, weighted (Heath et al 2013).

Notes: percentages do not sum to 100 since respondents could be canvassed by more than one party.  Chi2 (contrasting not contacted, contacted by Labour and contacted by other party) = 351.2 (4 df), p < 0.001.

What expected impact will the new registration system have on ethnic minorities?

3.1 The government’s own impact assessment for the new Individual Electoral Registration (IER) reforms showed that some marginalised groups may be negatively affected by the move away from the existing household registration. The main worries about the further marginalisation of the already under-registered groups centred around the main issues of lack of knowledge and information, cultural issues, and socio-economic marginalisation.

3.2 The new rules may have a negative effect on knowledge of eligibility and how to register. Lack of knowledge of eligibility is the main self-stated reason for under-registration for many groups and it seems to affect particularly the more recent immigrants from Africa. A change in rules may harm what little knowledge some groups have. Only 48% of BAME respondents of the recent Electoral Commission poll from December 2012 said they were confident in how to register- a proportion much below that of the white British population (Electoral Commission 2013).

Will the information campaign be sufficient to correct this? Government proposals to deal with this are that people will be automatically rolled over from the existing registers, if they can be verified, but it has also been shown that this process achieves only 78% accuracy (Electoral Commission, private communication). On the basis of the BES and EMBES registration verification exercise from 2010 we can see that minorities were harder to track down due to insufficient or wrong details (3% EMBES versus 1% BES) and also more likely to say they were registered at a different address (5% EMBES, 3% BES).

3.3 Also, for minorities, especially South Asian ones, there may be cultural reasons why the new reform may depress registration rates. Household registration may have meant that the head of the household registered their dependents, including wives, who may be excluded from the labour market, command poorer English and/or may be for cultural reasons expected not to take an active interest in public life. In EMBES17% of BAME women reported that someone else filled in their registration forms, in contrast to 11% of men. The move away from household registration may also have an impact on those young people who were hitherto registered by their parents. Remembering that the BAME population is younger than average, this may have a disproportionate impact on their levels of registration.

3.4 Socio-economic marginality was a major explanation for under-registration of BAME population. The proposed used of National Insurance Numbers as identity verifiers in the new registration process may further deteriorate the situation with recent arrivals and those economically inactive minorities (of which we have higher rates than among white British because of young age and greater incidence of female economic inactivity among South Asian groups being further excluded. There are also particularly high rates of NEET (not in employment, education or training) among groups such as the Black Caribbean (see Heath and Martin 2013). The Electoral Commission’s poll also suggested BAME respondents may find it harder to find their NI number (only 57% BAME respondents said it would be very easy for them to find it, in comparison to 75% of white respondents). The government has indicated that alternative means of identity verification will be available, but thus far the details are unclear. As a result this increased marginality may further depress registration rates.

3.5 However, we must not exclude the possibility that some of the changes introduced to the registration system may in fact deal with some elements of marginality. For example, people in multi-occupier households may not have- under the previous system- received registration forms addressed to ‘the present occupier’. Now, under the IER rules they would be individually identified from their NI numbers, if they have one, and the reach may in fact be improved. How the current reform will work on a practical level for those currently marginalised in the British society will largely depend on the- thus far unclear- implementation of the reforms.

What registration system is best for increasing turnout of ethnic minorities?

4.1 It is difficult to compare the impact that different registration systems have on minority engagement as majority of comparisons would necessarily be across different countries with very different electoral contexts, socio-demographic characteristics of their minority populations and different origins of these populations. The most useful comparative research done in this area has been in the USA where different states have different registration systems, but share similar electoral contexts and (to some extent) make-up of their minority populations. However, the most useful evidence is to be found from the impact of CHANGES to the registration systems within the same country. Here, again, the research from the USA which observes the impact of the many changes to the particular States’ registration changes is most informative.

4.2 In the first US research on this topic, the impact of registration rules on turnout were under question. Initially, just as it is the case now in the UK, the US literature assumed that the impact of less liberal registration laws will be greater on poorer, less educated people and blacks, but in their empirical study Wolfinger and Rosenstone (1980) showed this not to be the case (most recent studies such as Hanmer 2007 confirm this). While more liberal rules (such as registration on the day) increased registration, it did so across the board, for all social groups. Thus, the differential impact on the poor and racial minorities has not been empirically proven. The research on the impact of registration rules on registration itself is even more under-developed. Mitchell and Wlezien (1995) is an example- but they broadly confirm Wolfinger and Rosenstone’s findings as to the impact on the composition of electorate: more of all social groups registered under more liberal rules, but ethnic minorities or the poorly educated were not special in this respect.

4.3 As a result, it is important to keep in mind that while it is likely that BAME groups will benefit greatly from more liberal registration rules, such as on the day registration, in fact all marginalized section of the electorate would benefit equally.

24 February 2014

Bibliography

Durose, Catherine, Richardson, Liz, Combs, Ryan, Eason, Christina and Gains, Francesca (2013) ‘Acceptable Difference’: Diversity, Representation and Pathways to UK Politics’ Parliamentary Affairs 66(2): 246-267.

Fieldhouse Edward and Cutts David (2008) ‘Mobilisation or Marginalisation? Neighbourhood effects on Muslim electoral registration in Britain in 2001’ Political Studies 56(2): 333-54.

Hanmer Michael J. (2007) ‘An alternative approach to estimating who is most

likely to respond to changes in registration laws’ Political Behavior 29(1):1-30.

Heath, Anthony and Jean Martin (2013) ‘Can religious affiliation explain ‘ethnic’

inequalities in the labour market?’  Ethnic and Racial Studies 36 (6):1005-1027.

Mitchell and Wlezien (1995) ‘The impact of legal constraints on voter registration, turnout, and the composition of the American electorate’ Political Behavior, 17:179-202.

Smith, Stephen (1993) Electoral Registration in 1991, OPCS Social Survey Division, London HMSO, London.

Sobolewska, Maria, Fieldhouse, Edward and Cutts, David (2013) ‘Taking Minorities for Granted? Ethnic Density, Party Campaigning and Targeting Minority Voters in 2010 British General Elections’ Parliamentary Affairs 66(2): 329-44.

Wolfinger and Rosenstone 1980 Who Votes? New Haven: Yale University Press.