Using Wave 4 of the UK Household Longitudinal Study, it compared self-reported debt with Credit Reference Agency (CRA) records. Consenters to CRA linkage were more likely than refusers to be older, partnered, employed, homeowners and UK-born. Across eight debt types, misclassification was lowest for overdrafts, personal loans and retail finance. Prevalence gaps were largest for credit cards and other running debts. In Bayesian hierarchical models, the source used changed estimated associations between debt and mental health. The data came from a single survey wave, and the linked sample was selective because consenters differed from refusers.
Why it is interesting: Linking survey answers to credit records showed, debt type by debt type, where self-reports and records disagree. The estimated link between debt and mental health depended on which record was used.
Abstract — the authors' own words, CC BY
Abstract
Self-reported and administratively recorded debt measures are susceptible to measurement error, while selection bias may arise if individuals who consent to data linkage differ systematically from those who do not. This study evaluates the quality of self-reported and administrative debt measures by assessing selection into linkage, agreement between data sources, and the implications of measurement differences for associations with mental health. Using Wave 4 of the UK Household Longitudinal Study linked to Credit Reference Agency (CRA) records, we compared 8 debt types across 3 analytical samples: the full analytical sample, nonconsenters sample, and linked consenters sample. Agreement between self-reported and CRA-recorded debt was assessed using misclassification, sensitivity, specificity, positive predictive value, and negative predictive value. Mental health associations were estimated using Bayesian hierarchical regression models. Respondents who consented to linkage were more likely to be older, partnered, employed, homeowners, and UK-born. Misclassification varied across debt types and was lowest for overdrafts, personal loans, and retail finance. Substantial discrepancies were observed between self-reported and CRA-recorded debt prevalence, particularly for credit cards and other running debts. Differences between data sources also influenced estimated debt-mental health associations. These findings underscore the need to account for selection and measurement error in linked debt data.
Cite this item
BibTeX
Free sign-up. A digest of the new papers twice a week, grouped by domain.
Other drivers in Data, models and systems
-
Forecasts and their record
How the forecasts are made, how well they explained the past, and how well mortality can be forecast — and by whom.
3 this week -
What linked data reveals
What linked registries make it possible to know about whole populations
9 this week -
AI in medicine
Whether machine learning predicts, detects and decides better than the tools we have, and whether that has yet changed an outcome for a population.
4 this week -
Health systems and access
What the health system can provide and who gets it.
4 this week -
Value in health
What a treatment, health programme or policy is worth against what it costs, and how far the models that decide it can be trusted.
3 this week -
Revisions and new sources
What the figures are built from, what is newly available, and how much the numbers move when they are revised.
2 this week
