Written evidence from Fran Bennett (University of Oxford), Dr Rita Griffiths (University of Bath), Professor Jane Millar (University of Bath),  Marsha Wood (University of Bath) (CPM0003)

Evidence from ‘Couples balancing work, money and care under the shifting landscape of Universal Credit’ ESRC ES/R004811/1

We are academic colleagues working on an ESRC-funded research project entitled Couples balancing work, money and care under the shifting landscape of Universal Credit, ESRC ES/R004811/1. The project is a three-year (2018 - 2021), two phase, longitudinal, qualitative research study of couples claiming Universal Credit, and is a collaboration between the Institute for Policy Research (IPR) at the University of Bath, and the University of Oxford.

The research is exploring how low-income couples and families with children claiming Universal Credit juggle work and care, and manage their household finances. Between June 2018 and January 2019, we conducted 123 interviews with 90 participants in 53 households, in four areas in England and Scotland. A majority of households had dependent children and just under half had at least one earner. In 31 households, Universal Credit was the main source of income. In September and October 2020, we re-interviewed over two thirds of the sample (63 participants) to find out how they had fared in the intervening two years, and how well they were managing in the context of the Covid-19 pandemic and suite of emergency measures put in place by the government, including the £20 uplift.

Our research does not directly address the issue of how to measure child poverty, but does help to shed light on questions of how and why child poverty arises. Taking this broader perspective is important because any measures of child poverty are only as useful and effective as the policies designed to help reduce its risk, incidence and impact. Below, we address four issues relevant to the terms of reference of this stage of the Committee’s inquiry, based on the findings of our research to date. In particular, we address the Committee’s concern about how child poverty can most accurately be measured and defined. We therefore distinguish between the way in which income is measured in household surveys and the way in which this is then used to construct a measure of (child) poverty (or low income, as this is termed in HBAI).

First, we would emphasise the importance of ensuring that the survey data that underpins the calculation of numbers (of children) in poverty in HBAI measures the income that people actually receive. Our research demonstrates that this may be of particular concern in relation to Universal Credit. The current practice in household surveys is to ask about income, as well as about what benefits are received by name; we would stress the importance of not only ensuring that this practice continues but also that it is as accurate as possible.

Our research evidence to date shows the importance of taking into account the amount of money that benefit claimants are actually paid. A common thread running through the testimonies of our interviewees was the often large discrepancy between Universal Credit entitlement and the actual payment they received each month. In many cases, the amount of Universal Credit families were paid was much less than what they expected or had been informed they would get, and a long way short of notional ‘entitlement’.

A key reason for this was deductions – automated amounts of money taken at source from the Universal Credit payment to repay benefit and tax credit overpayments, loans, rent and council tax arrears, together with other debts that the government, local authorities, utility providers, landlords and other third parties claim are owed to them. Deductions take priority over other debts at all legal jurisdictions and reduce the payment in strict order of priority according to a fixed set of minimum and maximum percentages, generally until the debts (of both partners in a couple) have been recovered in full. Currently, deductions can reduce the monthly payment by up to 30 per cent of the standard allowance. This percentage may be exceeded by adding ‘last resort’ deductions if there is a risk of homelessness or utility disconnection. Unlike consumer debt, which is often statute-barred, there is no time limit for the recovery of these debts.

In our research, deductions sometimes reduced household income to below the level needed by families with children to cover their rent and basic living expenses. Couples found that benefit and tax credit overpayments and third-party debts, even from before their relationship began, were aggregated and deducted automatically from Universal Credit. Those for whom Universal Credit was the main or only source of income frequently ran out of money before the end of the month. Income inadequacy, not poor budgeting skills, was the key underlying issue.

The difficulties of knowing how much benefit money they were entitled to or would get each month also affected experiences and perceptions of the temporary £20 uplift in Universal Credit[1], which varied in terms of its visibility to our interviewees. An extra £20 per week on top of the uprating of 1.7% in April 2020 increased the monthly standard allowance for couples (aged 25 or over) from £498.89 to £594.04. This difference of £95.15 per month meant a great deal to many of our families, helping them to keep their heads above water at a time of increasing food and utility bills due to the family being at home more. However, although all Universal Credit (and working tax credit) claimants are currently entitled to the £20 uplift, many working families in our research reported that, as far as they were aware, they had not received it. This is because, for those in paid work, the uplift does not automatically translate into a £20 per week increase in payment, but rather is tapered away with monthly earnings. As noted above, deductions can also reduce the amount claimants get each month; and deductions could be increased as a result of the standard allowance increasing with the uplift. Universal Credit is also taken into account for other means-tested help. The £20 uplift could therefore also sometimes mean reduced entitlement for help under local Council Tax support schemes.

The complex way in which benefit entitlement, deductions, earned and unearned income interact to produce the monthly payment, which arise largely due to Universal Credit’s hyper-sensitive system of means testing, strongly suggests that it is essential to ensure that actual household income, rather than simply notional entitlement, is the basis of any system for measuring poverty.

Secondly, we suggest that the way in which Universal Credit works for many people means that the time periods for measuring household income to be used to calculate (child) poverty in HBAI should be considered afresh. Our research shows that income fluctuates from one month to the next when people are receiving Universal Credit.

All the two-earner and most single-earner couples in our research said that Universal Credit fluctuated each month, sometimes significantly. Even those with fixed salaries paid monthly reported that their Universal Credit sometimes fluctuated in unpredictable and seemingly arbitrary ways. It was difficult to anticipate drops in the payment and set aside ‘surplus’ earnings; household income was generally spent in the month in which it was received. For families with children, a month with no Universal Credit meant losing that month’s work allowance for good – and once lost, this entitlement can never be recovered. Fluctuating payments and loss of entitlement to Universal Credit could also have knock-on effects for eligibility and entitlement to other forms of means-tested help, including council tax support, help with prescription charges and free school meals.

Problems often arose due to the interaction of earnings with the fixed monthly assessment period (the subject of recent legal challenges). In our research, the double counting of wages in a single assessment period not only affected people with irregular earnings, but also those paid weekly and four weekly, and even those paid a regular monthly salary. Being paid a few days early - due to a bank holiday, for example - could result in more than one wage packet being counted. Particularly badly affected were working couples whose pay days were close to the end or beginning of the fixed monthly assessment period. Receipt of a lump sum – such as holiday pay, performance bonus or back-dated salary increment - could also reduce or eliminate a couple’s monthly entitlement.

Because the monthly earnings of couples are aggregated for the purposes of calculating entitlement, in dual-earner families these effects could be amplified, as well as prolonged. The double, triple, even quadruple counting of wages in the same assessment period could cause Universal Credit entitlement to cease altogether. Even though these couples had not actually earned any more money, multiple sets of aggregated wages in a single assessment period could also trigger ‘surplus earnings’ rules, meaning that it could be many months before they became eligible to reclaim. With notification posted to the online account only a week before payment was due, families with awards that were suddenly and unexpectedly reduced or stopped could find themselves with no income for the coming month. Many families preferred the annual assessment of tax credits to the repeated monthly means test leading to fluctuating income under Universal Credit.

The Committee could therefore recommend that the government revisits the time periods for measuring income and compare different approaches to this as the basis for their (child) poverty measures, given the evidence from our research about how important this issue now is.

Thirdly, although in the relevant household surveys for HBAI information is collected from each individual about benefit receipt, currently we believe that this information is not used to examine individuals’ access to income, but rather to confirm total household income. Our research highlights the importance of independent income for each partner in couple households, whether this is earnings or individual benefits. In some households, who receives the means-tested benefit(s) for the household may also be important, in particular in relation to resources for children. Some families preferred the ring-fencing and labelling of payments for children under Child Tax Credit, finding it more reliable and predictable, as the amount did not change. Under Universal Credit, absorbing the child element into the single payment risked it being spent on general household expenditure. Whilst in our couples the woman (usually the ‘main carer’) was often either the sole or a joint signatory to the bank account into which Universal Credit was paid, we had some evidence that in some previous relationships there had been some financial control by the male partner. Thus the payee for Universal Credit can potentially be important in relation to the poverty children experience in practice. Thus, making more use of the data collected on benefit receipt by individuals could be helpful, particularly given the single payment of universal credit for most couples.

Finally,  we would agree with the Social Metrics Commission that childcare costs should be taken into account in measuring (child) poverty, in the same way as housing costs are in the ‘after housing costs’ measure.

At the moment, one poverty measure (after housing costs) takes into account what people have to spend on their housing. This is considered important, because those living on low incomes in particular often have little or no choice over their housing costs. But, whilst income to help with other unavoidable costs is taken into account as income in HBAI, no account is taken of those costs themselves on the other side of the equation. In relation to children, this means that child poverty for those with childcare costs may be under-counted.

Our research shows that for families childcare costs can be a real financial burden - even with the help of Universal Credit. With any childcare contributions part of the monthly payment and tapered with (aggregated) earnings, they also found it hard to work out what they had actually been paid. In practice, it was often significantly less than the headline ’85 per cent contribution’ families had expected to get (itself subject to maxima that have not been uprated for many years). Some parents tried their hardest to avoid incurring childcare costs, indeed, in part for these reasons and in part because of the difficulty of paying childcare costs upfront and only being reimbursed at the end of the month. We only had small numbers in our sample who were paying childcare costs, and these decreased over time. But this was partly because families found informal child care instead, or because women in particular reduced their hours or gave up their jobs. It was clear that both the cost and the administrative burden of childcare costs and how they were dealt with in Universal Credit were creating burdens for families with children in our sample. This would lend weight to the recommendation by the Social Metrics Commission that poverty measures should also take into account childcare costs.

February 21