Gabriele Mari, from Erasmus University Rotterdam, reflects on how survey data from the UK Household Longitudinal Study might shine another light on the economic insecurity felt by unpaid carers.
Many benefit programs depend on the collection and handling of people’s data by state agencies. Some programs require filling in multiple forms, many ask for timely updates if financial or other circumstances change, and more and more decisions about eligibility, errors, and sanctions have been automated or offloaded to predictive algorithms.
These processes may leave many in dire financial need and cause much distress, and have now prompted public debate and policy scrutiny in several countries.
This has motivated some recent research I’ve been contributing to. Should we treat these events as anomalies in the everyday operations of the social security system? Or do they highlight and deepen the cracks in its design? And how could we approach ‘data’ that can provide some answers to these questions?
The case of Carer’s Allowance
In this working paper, I examined the case of Carer’s Allowance, a cash benefit aimed at unpaid carers in the UK.
Entitlement to Carer’s Allowance stops past a certain level of earnings. Yet this earnings limit has not been enforced in a timely, transparent or consistent fashion.
As a result, hundreds of thousands of carers unwittingly accumulated overpayment debts totalling more than £251m in 2024 (PDF).
Reclaims have taken carers by surprise, plunging many into hardship. The aftermaths have been in the public eye for a few years now, and have been met with widespread condemnation, official reviews, and piecemeal attempts at reform and redress.
But administrative failures leading to overpayments are only one of the ways in which the earnings limit might impact carers.
Once access to Carer’s Allowance is actually revoked past the limit, carers lose a source of income. The limit might also shape people’s needs and choices when it comes to paid work.
Thinking along those lines, I examine how, even over and beyond overpayments, the earnings limit might lead to economic insecurity among unpaid carers, how large and persistent its effects are, and who’s most affected.
Which data, whose data?
This work was made possible by the continuous, voluntary, and anonymous participation of carers in the UK Household Longitudinal Study (UKHLS) hosted by the UK Data Service.
In the survey, carers self-identify as such when asked about the help and support they provide to people (co-resident or not) who are ill, elderly or disabled. These and other follow-up questions are wider in scope than legal definitions.
This is justified, though, as the letter of the law often excludes or marginalises some unpaid carers. Besides, while some people consistently report on their caregiving involvement over many interviews, others do so more sparingly.
The longitudinal nature of the survey is therefore crucial to better appreciate the ebbs and flows of caregiving.
In the end, I could pool together more than 118,000 interviews from around 17,900 people, who reported caregiving at any point from 2009 to 2024. A subset of these interviews allowed for additional longitudinal analyses.
As for people’s financial circumstances, these survey data are not collected for the purpose of administering benefit programs.
Even if perhaps less precise or detailed compared to the data held by tax and other authorities, the data are of high quality and their breadth is much richer in at least one respect.
Alongside material indicators, and albeit succinctly, the survey asks people to report on how they feel about their budgets and incomes. Related worries and dissatisfaction are key aspects of economic insecurity, and often track mental health more closely than crude monetary measures.
Interviews also occur one or a few years apart, meaning that short-term changes in material conditions cannot be easily detected. This is undoubtedly a shortcoming, as incomes and benefit receipt might vary widely over the course of months or even weeks, especially for those striving to juggle paid and unpaid work.
In my analyses, some earnings changes around the limit of Carer’s Allowance will inevitably be missed. On the other hand, though, a more intensive data collection effort might have put an undue burden on carers and survey respondents more generally.
The data have limits, in sum, but limits are also warranted. The definitions, scope, and frequency of data like UKHLS is still well-suited to reflect (on) the every-day realities of caregiving relationships, and their relation to paid work and the benefit system.
The gap in insecurity past the earnings limit of Carer’s Allowance
Figure 1 summarises the main message of my paper. It shows how I examine levels of reported economic insecurity across a range of indicators, from arrears to savings to the aforementioned worries and dissatisfaction. A flexible statistical model is used to estimate the percentage of people reporting economic insecurity according to each indicator (vertical axis) and depending on their weekly earnings (horizontal axis).
I single out and compare two groups of unpaid carers, one who has reported access to Carer’s Allowance at any point during the survey (in yellow) and one who has never reported it (in black).
The logic is that, as earnings grow, the first group will earn more than the limit, while the other ‘simply’ earns more. Whilst imperfect, this comparison allows me to grasp some of the effects of brushing against the Carer’s Allowance limit, independently from the underlying changes in earnings.
Three things are noticeable about Figure 1.
On the left-hand side of each graph, we find carers with little or no earnings. They report the highest levels of economic insecurity across the board. As we move on to carers with higher earnings, from left to right, insecurity declines, at times a tad more steeply for carers who might have gotten some financial relief from Carer’s Allowance.
Yet, when labour incomes exceed the earnings limit of the allowance, set at around £150 per week in the period studied (represented by the vertical dashed lines), reports of insecurity start to diverge.
Only among carers who might have been affected by the earnings limit, we see a sizeable bump in insecurity across many indicators, and most notably for housing arrears. Leveraging the longitudinal data, further estimates show that these gaps in insecurity open up once carers first cross the earnings limit and remain fairly stable after that.
Unpaid carers in middle- and working-class occupations or without paid work, and especially renters, report relatively more insecurity once the earnings limit is breached.
And, although the data is limited to the household rather than encompassing the broader set of people’s relationships, evidence suggests that insecurity spreads beyond unpaid carers themselves and touches other people, especially those they live with.
Figure 1: Estimated proportion of respondents (%) reporting economic insecurity by history of benefit receipt and weekly earnings relative to the Carer’s Allowance (CA) limit.
Note: Earnings are top-coded at £300 above the limit. Outcomes are averaged by means of restricted cubic splines with 95% confidence intervals (shaded areas) and knots at -80, 0, 40, 80, and 160. Pooled cross-sectional survey weights are applied (UKHLS waves 1-15, 2009-2024).
Data and policy alternatives
These preliminary findings echo and might complement the accounts provided by unpaid carers in the UK, from qualitative research to advocacy. Compared to those, a survey only scratches the surface and can be far less participatory.
At the same time, large, longitudinal, and comprehensive data like UKHLS can be interrogated to add another piece of the puzzle. Some of the weaknesses of survey data, as discussed, can be reframed as strengths or taken as a reference point to think about limitations and other options.
What has transpired about the Carer’s Allowance system might question governments’ need for and management of data when implementing policies whose stated intention is to support people.
Findings in this paper might add to a wealth of alternative ‘data’ that can stimulate change. They suggest, for example, that unpaid carers who are renters might benefit from a financial buffer even when their earnings are increasing.
More broadly, a taper rate could be introduced to prevent entitlements falling over the current ‘cliff edge’ in Carer’s Allowance.
Yet both reforms would rest on (more) data from and about unpaid carers.
Part of a more ambitious vision, getting rid of the earnings limit altogether could be the less ‘data-hungry’ and perhaps fairer alternative for unpaid carers.
Meet the author
Gabriele Mari is an economic and family sociologist working as Assistant Professor at Erasmus University Rotterdam.
He conducts research on how different kinds of money and work are entangled with people’s relationships. Lately he has focused on the intergenerational consequences of economic crises and changes to social security systems.
Comment or question about this blog post?
Please email us!