Ethnic Minority British Election Study pilot

Lead Research Organisation: University of Manchester
Department Name: Social Sciences

Abstract

Since 2010 there has not been a single representative survey dedicated to political and social attitudes fielded among the UK's fast-growing ethnic minority population. Existing social science surveys with representative ethnic minority samples invariably lack space to ask many crucial political attitudes questions (e.g. Understanding Society) or have too small a sub-sample of minorities for detailed analysis (e.g. British Social Attitudes and the British Election Study), and the British Election Study Internet Panel, despite including a large number of ethnic minority respondents, is largely unrepresentative of minorities. In that time there have been three General Elections, and two referendums (EU and Scottish) that have reshaped politics across Britain. There have also been huge political events which have particular importance to minorities including the Black Lives Matter protests, racial disparities in experiencing the global Covid pandemic, and the Windrush scandal. Nonetheless, we know relatively little about how minorities have reacted to the huge political events of the last decade or how long-standing issues for ethnic minority populations in Britain such as political under-representation and under-registration to vote have changed. In short, in a crucial period of British political history, we have enormous data gaps in understanding how minorities perceive and understand politics broadly, and political inequality and exclusion more specifically.
These data gaps are caused predominantly by the difficult and expensive processes needed to achieve a representative sample of ethnic minorities. This already expensive and difficult task of surveying minority populations is further compounded by the ongoing pandemic, which directly prevents face to face survey research. As all surveys struggle to respond to the pandemic, many, including the British Election Study have moved to use a push-to-web approach to continue surveying the general British population and has found this approach largely successful.
This proposal is for a pilot study to study whether the surveying techniques adopted by the British Election Study in the pandemic - push-to-web methods - can be successfully deployed in surveying ethnic minorities. The answer to whether and how these can be successfully used raises the possibility that these can be employed in other contexts and at other times, without a pandemic, as a way to deliver high quality value-for-money representative surveys of minority populations.
The push-to-web approach-as used by the British Election Study-is generally cheaper as it limits staff time as first contact with sampled addresses is done by mail and the interview itself is conducted online. Given the additional benefits of conducting the interview online, such as reducing social disability bias, ability to randomise and introduce survey experiments, if successful in recruiting a new representative sample of ethnic minorities this approach would be a game changer for surveying ethnic minority public opinion.

Publications

10 25 50
 
Description Following the outbreak of the coronavirus pandemic in 2020, data collection for BES 2019 moved to a mixed-mode, push to web approach during fieldwork because face-to-face interviewing was suspended. Participants were invited to take part in an online survey through a letter to their home, with the option of completing a paper survey offered to those without internet access (or those respondents that simply preferred this method of participation). The success of this approach (see BES 2019) raised the question of whether push to web data collection might be successfully replicated for EMBES in future. Fieldwork was carried out in Jan-March 2021 in areas of high ethnic minority population. Initially, to save wastage on incentives for non-eligible (White British) respondents, sampled addresses were invited to fill out a shorter screening questionnaire before being invited to participate in the full survey. However, the achieved sample size was much lower than expected. When it transpired in February 2021 that the engagement with the survey was much slower than for the main BES push-to-web, we hypothesised it might be due to the relative complexity of the 2-stage recruitment that was responsible. To test this proposition, we simplified the process into a single mailing, with no screening. We felt that the risk of including more white British respondents in the final sample (and paying them £25 rather than the screening incentive of £5) might outweigh the costs of the survey which does not recruit enough respondents, and we have had money available at this stage, due to such low take-up of the survey. We therefore issued a new sample, in areas with similar demography, of 1000 addresses. The design change did not increase the proportion of engagement with the survey significantly, so we can conclude that it is unlikely to be the original 2-stage design that is responsible for the small sample size ultimately achieved. We believe that the most likely explanation is that the areas of 20% and above ethnic minority population are systematically different to those included in standard surveys (i.e. harder to reach), and might yield lower response rates. We intend to check this by comparing response rates from the main BES when similar areas were sampled (the BES overall response rate does not take into account any geographical differences in response rates). We will explore this and other possibilities in one methodological publication, a draft of which we ae hoping to present at a ESRA conference in July 2023. Although relatively small, this sample was pooled with the main BES and we are currently producing a substantive research article exploring ethnic minority voters political attitudes. We have now presented a draft of this paper twice, at the EPOP conference and at a Departmental Seminar at Stockholm University and are preparing it for publication. The preliminary finds are that minorities are more liberal than the white British majority, but only on "identity liberalism" or ethnocentrism not on cultural issues. Higher education attainment partially explain this, but "necessity" attitudes (favouring minority rights/equalities) also matter. Higher "necessity" liberalism does explain why minorities voted Remain more than White British but does not explain the voting gap in general elections.
Exploitation Route This pilot project highlighted the limitations of a push-to-web approach that is thought to work with general population samples. It also raises questions of differential non-response rates in these samples, which might make them less representative of certain neighborhoods than previously thought. The take-away is that so far, face to face remains the best way to recruit a new high quality sample of ethnic minorities.
Sectors Communities and Social Services/Policy,Government, Democracy and Justice