Context

The Financial Conduct Authority (FCA) undertakes regulatory intervention to support its strategic an operational objectives in the UK financial services industry. The FCA uses evidence to inform its decision-making around market interventions and to understand the social impacts of them.

The FCA has previously commissioned research which estimated the effects of household debt on people’s subjective wellbeing, and valued entry into and changes in debt levels monetarily for use in its appraisal and evaluation activities. This study aimed to extend and refine the previous research.

The study

The study was commissioned by the FCA to quantify and monetise the potential benefits on subjective wellbeing that are generated by its regulatory interventions, via changes in household debt status and level.

The study extends and refines the previous research by using more recent data from the Wealth and Assets Survey (WAS, GB, 2010 to 2020) and Understanding Society survey matched to credit file data (USoc, UK, 2012 to 2021). Both data sources were large-scale, nationally representative longitudinal surveys of individuals and households which capture a subjective wellbeing measure (life satisfaction), a range of debt-related variables (aggregated to the household level) and key control variables.

Econometric modelling using static fixed-effects models and dynamic panel models with generalised methods of moments (GMM) estimators were undertaken in two stages:

• First, the analysis estimated how much a change in household debt status or level affected subjective personal wellbeing both in the same survey year and with debt status in the previous survey year, in both data sets and in both model types.
• Second, the analysis attempted to convert the changes in wellbeing into equivalent monetary values using the optimal model from stage 1, and using values for low, central, and high scenarios, and under different assumptions of debt persistence.

Key findings

Statistical significance was tested, and is reported here, at the 95% level of confidence (p<.05).

• Impact of arrears: Using WAS, entering arrears was associated with a 0.41-point drop (on a scale from 0-10) in life satisfaction in the same survey year, and a 10% increase in arrears was associated with a 0.0006-point drop. Analysis of USoc returned non-significant effects.
• Impact of high-cost debt: Using WAS, entering into high-cost debt was associated with a 0.17-point drop in life satisfaction and a 10% increase in high-cost debt was associated with a 0.003-point drop. Analysis of USoc found non-significant effects. The impact of increasing high-cost debt was greater in the bottom income quartile of households.
• Impact of credit product: In both WAS and USoc, entering into and increasing their overdrafts by 10% were both associated with a drop in life satisfaction (e.g. by 0.20-points and 0.003-points respectively in WAS), and in USoc it was greater for unauthorised than authorised overdrafts. There were also small drops in the context of personal loans and home-collected high-cost loans.
• Other outcomes: None of a wide range of other, non-debt outcomes that were of interest to the FCA (such as increased suitability of products, seeking help or advice) significantly affected individuals’ subjective wellbeing.
• Estimated monetised values: The estimated monetised value of the (negative) impact on subjective wellbeing in year 1 ranged from £90 for entering into high-cost credit to £20,310 for entering into a cash loan (which reduced to £2,270 at year 4). It was £5,640 for entering into unauthorised overdraft and £12,430 for entering into arrears year 1.

Points to consider

Methodological strengths or limitations

• Although the research was conducted in the context of the desire to understand the impact of regulatory intervention, the study did not directly measure or evaluate this.
• The analysis described was complex and the report itself is highly technical, with some terms not clearly defined.
• There is nonetheless evidence of a diligent, rigorous and robust analysis of the data. For example, models and variables were selected carefully and extensive testing of model validity and reliability was undertaken.
• The authors caution the reader to consider the reasonableness of some of the assumptions made in estimating monetary values. Other assumptions are also set out in the report.
• Differential findings from WAS and USoc raises questions about the replicability and generalisability of the findings. The authors discuss possible reasons, and others may include a less sensitive response scale for financial satisfaction in USoc (1-7, compared with 0-10 in WAS).

Applicability

• The findings should be of interest to policymakers, regulators and providers within the financial services industry.

Relevance

• The findings are relevant given the increasing emphasis of policy and regulatory intervention on subjective wellbeing, particularly from a cost-benefit perspective.

Generalisability/transferability

• Although the specific findings are unlikely to generalise to other countries and context (including the impacts of other outcomes), learning from the study, particularly methodological learning, should transfer well.