Evaluation Scotland Wales
The UK Strategy for Financial Wellbeing is taking forward the work of the Financial Capability Strategy Opens in a new window


Squeezed out or opting out?

Evidence type: Insight i


Access to financial products became a policy concern in the 1990s. Since then, even effective policies have failed to benefit all societal groups equally. The Covid-19 pandemic further highlighted disparities, both as a result of increased demand for credit and in the different ways that ethnic minority groups approach and experience financial services. Ethnic minority groups were more likely to cash in investments and to have used savings to cover day-to-day expenses. Understanding why there are these differences, and how they relate to consumers’ attitudinal and behavioural perspectives, is crucial for policymaking.

The study

The study was undertaken by the Social Market Foundation and the report was sponsored by The Which? Fund, which is funded by the Consumers’ Association, to understand the scale and nature of ethnic differences in the take-up of financial products and services and the drivers of these.

The study used a multi-methods approach:

  • A review of existing literature.
  • The collection and analysis of primary survey data, using an online survey of 1,000 people run by Opinium, plus a 500-person booster sample of ethnic minority respondents, in July 2022. The data were weighted to ensure they were nationally representative.
  • Secondary analysis of three large-scale nationally representative surveys: the Financial Conduct Authority’s Financial Lives Survey; the Wealth and Assets Survey; and Understanding Society.
  • In-depth online qualitative interviews with 21 ethnic minority consumers, 11 of whom were drawn from respondents to the primary survey and 10 of whom were recruited by specialist recruitment agency Indiefield. Participants were specifically recruited on the basis of not having any one or more of a current account, a savings account, a private or workplace pensions or key insurance products.
  • Conversations with stakeholders (academics, policy researchers, charities, regulators and trade bodies) from across the financial services industry.

Key findings

The figures noted below come from different data sources used in the study and the differences were not subject to statistical significance testing.

  • Differences in product take-up: There was little difference in the prevalence of the take-up of investments and personal credit by ethnicity.
    • Lower rates of take-up for savings and pensions among ethnic minorities were largely accounted for by age and income, although a large gap in the level of average pensions savings (which was £3,350 lower for ethnic minorities) was not.
    • A lower likelihood of holding insurance products – notably home insurance (28% among ethnic minorities as a whole vs 48% in the general population) and contents insurance (37% vs 62%) – also held true regardless of age and income.
    • Low-income Asians were especially likely to be unbanked (16% vs 4% of low-income White people).
  • Drivers of differences: Age and income were the biggest drivers of the gaps in product take-up rates by ethnicity, but did not wholly explain them.
    • Other relevant factors were area-based and poverty-related premiums (on insurance and credit products), language barriers, lower awareness and different preferences (choice and judgment). For example, many ethnic minority consumers chose to ‘self-insure’ (using savings) than to buy insurance products they deemed to be poor value for money.
    • Choice is underpinned by several factors including lower self-perceived awareness, lower understanding of and confidence in financial products and services and perceptions of discrimination or poorer service for minority groups.
    • As such, ethnic minorities may still be disadvantaged by being less resilient to financial shocks, over-confident about their provision for retirement and unable to participate fully in society and the economy.
  • Recommendations for policymakers and industry: Lower the pensions auto-enrolment and employer contributions thresholds; campaign to improve trust in financial services; peer-led financial services support hubs in ethnic minority communities; support for new migrants; better data on the use of financial services among ethnic minorities.

Points to consider

  • Methodological strengths/weaknesses: The authors note that they were unable to report the primary-survey findings for some specific ethnic minority groups due to small sample sizes.
    • It is not clear if the regression analysis used in the study accounted for any factors other than age and income, for example, sex (which the authors note elsewhere was an important mediating factor). If other important mediating or confounding factors were excluded, the regression results of may tend to overstate or understate the true influence of ethnicity.
  • Relevance: The study should be of interest to researchers, policymakers and other stakeholders who wish to understand the levels and drivers of financial exclusion among ethnic minorities and what could be done to address it.
  • Applicability: Any findings which consider ethnic minorities as a whole may tend to mask important differences between ethnic minority groups, and should not be interpreted as applying to all ethnic minority groups or to all ethnic minority groups equally.
    • While the report relates to the United Kingdom, the individual data sources used do not necessarily cover the whole of the UK (the Wealth and Assets Survey does not cover Norther Ireland and the geographical coverage of the depth interviews undertaken for the study is not stated).
  • Generalisability/transferability: The authors note that the findings from their primary survey data did not always concur with those of the secondary data sources. As a result, they say they elected to focus on results that were consistent. Nonetheless, the authors go on to report finding more modest differences in product take-up than those previous estimates. This may suggest that there are unobservable biases in their survey sample which the survey weighting could not adjust for. As such, the findings from the primary survey data may not be truly representative, and should only be generalised to the population with caution.
    • The authors refer to significant differences in rates of product take-up, but it is not clear if these have been subject to statistical significance testing or if statistical confidence around point estimates have been considered. As such, precise percentages should not be generalised from the sample to the wider population without caution. The study did nonetheless use logistic regression to test for the independence of differences observed in relation to ethnicity once controlling for factors such as age and income,. The significance of any remaining differences based on this analysis is more reliable because logistic regression incorporates statistical tests.

Key info

Year of publication
United Kingdom
Contact information

Niamh O Regan, Aveek Bhattacharya, Gideon Salutin and Scott Corfe, Social Market Foundation, Third Floor, 5-6 St Matthew Street, London, SW1P 2JT