Drowning Not Waving: Navigating the choppy waters of monthly assessment in Universal Credit

Posted in: Culture and policy, Data, politics and policy, Evidence and policymaking, Housing, Public services, UK politics, Universal Credit, Welfare and social security, Young people

Dr Rita Griffiths is a Research Fellow at the University of Bath Institute for Policy Research (IPR). Dr Marsha Wood is a Research Associate at the Institute for Policy Research. This blog post is part of their work on the report ‘Coping and hoping: Navigating the ups and downs of monthly assessment in Universal Credit’ which explores how Universal Credit is affecting income security and financial well-being, month to month, in real life settings.

The Post Office Horizon scandal laid bare the human cost of doggedly adhering to the notion of digital infallibility. For decades, the Post Office and its IT contractors insisted that flawed, bug-ridden software was operating correctly, in spite of sub-postmasters’ vociferous and increasingly desperate protestations to the contrary. But IT systems do not have to be inherently defective for their operation to impact on people’s lives in damaging ways. Even when functioning as intended, they can have potentially harmful effects on the financial and emotional well-being of their so-called ‘end users’. Many of us will have experienced the frustration of trying to correct a billing error or reverse a payment wrongly taken by a bot-driven automated system. Not for good reason has the ‘computer says no’ trope become such a familiar part of the cultural zeitgeist.

Though infuriating and time-consuming, rarely do these digital encounters have serious financial consequences. But what if the automated system you were dealing with wasn’t just designed to manage your mobile phone or utility bill but was responsible for determining how much household income you and your family had to live on each month?  How would you feel if the payment you were trying to correct was the money you needed to pay that month’s rent, food or childcare bill? What would you do if you were told nothing could be done to rectify an error until the following month at the earliest. This is the context in which people whose income is low enough to entitle them to Universal Credit, must try to plan and budget their household finances.

Universal Credit and how monthly assessment works

Universal Credit (UC) is the UK’s main, means-tested, working age benefit for supplementing low household income among people both in work and out of work. It replaces six means-tested benefits and tax credits - combining elements for basic living costs and rent, together with supplements for full-time carers, people with disabilities and work-limiting health conditions and those with dependent children, within a single monthly household payment. Working families with paid childcare may also be eligible for a contribution of up to 85 per cent towards their childcare costs. This, too, is included in the monthly payment. Under the previous system, several of these elements were assessed and paid as separate benefits, with all the additional bureaucracy and potential for complexity this entailed.

Another difference is that UC is assessed and means-tested monthly ‘in real time.’  Tax credits were assessed annually using a self-completed form based on the previous year’s income. Changes in earnings and household circumstances were not generally taken into account until the end-year reconciliation, potentially giving rise to under- and overpayments and fraud. Under UC, for working claimants on PAYE, the payment reflects net monthly earnings for the month just gone, as reported by employers to HMRC’s RTI (Real Time Information) system. Any earnings captured within a claimant’s assessment period – a fixed calendar month which runs from the date the claim started - together with any reported changes of circumstance, are automatically taken into account when the monthly UC award is calculated. Self-employed claimants and those who earn under the PAYE threshold self-report their earnings online. A single taper - currently 55% - automatically reduces entitlement as earnings rise. Claimants with children or limited capacity for work may be eligible for a ‘work allowance’ which disregards an amount of earnings each month before the taper is applied. If monthly income, earnings and household circumstances remain the same, then the UC payment should too, but if they change, the payment is automatically programmed to rise or fall in response. Claimants are advised to prepare in advance and budget for a potential change in the monthly UC payment, saving or setting aside money in the months when earnings rise and using this ‘buffer’ to compensate for the months when the UC payment reduces.

On the face of it, the design of UC makes a lot of sense, but, as with all radical reforms, there are significant trade-offs. Earnings, it turns out, are not the only deduction from the monthly payment. Deductions for debts – automated amounts taken at source from the UC award before payment is made to repay benefit and tax credit overpayments and arrears of rent, council tax or utility bills that claimants may owe – are a separate subtraction. For people in work, the amounts taken can vary with earnings. Deductions were also a feature of the legacy system. However, UC’s much stricter debt repayment policies, greater technical and data sharing capabilities with third parties, together with increased automation across its debt collection procedures, have significantly enlarged the scope for the identification and recovery of debts and overpayments, sometimes dating back decades. Policies such as the two-child limit, benefit cap, ‘bedroom tax’ and non-dependent adult deduction, can also lower entitlement. Claimants who are self-employed and subject to the minimum income floor (MIF) may be subject to further reductions. Taken together, this cornucopia of deductions can potentially reduce the monthly UC payment by significant amounts. And with entitlement means-tested and recalculated monthly, claimants only receive seven day’s notice of the amount they will get paid. Tax credits, in comparison, generally remained fixed for a year.

New Research by the Institute for Policy Research at the University of Bath

A single monthly household payment which rises and falls alongside changes in earnings fits into an overarching policy narrative of benefit simplification and improved work incentives, designed to smooth and stabilise household income and make the financial impact of additional hours more visible and immediate, thereby motivating claimants to work longer and earn more. Or at least that’s the theory. In fact, there has been no systematic research into UC claimants who are working. Little is known about the way in which changes in earnings interact with UC, how stable or volatile this makes household income, nor the effects on budgeting or work-related decisions. This is what new research by the Institute for Policy Research (IPR) at the University of Bath set out to discover.

The research explored the lived experiences of Universal Credit claimants in paid work or self-employment. Qualitative and quantitative data was collected ‘in real time’ from regular monthly interviews from a sample of 61 claimants in 42 households with and without children with one or two earners, for 13 months between February 2022 and March 2023. Sufficient data was collected to track monthly income and earnings over time in 37 households. Of 50 claimants in work, 23 were paid between £9.50 and £9.99 an hour, 10 were paid between £10 and £10.99 an hour, 8 were paid between £11 and £14.99 an hour and 3 were paid over £16 an hour including one who was paid £27 an hour. 16 were not sure of their hourly rate.

Monthly household incomes among UC claimants with earnings were found to be highly volatile and unexpectedly low. Average monthly income after housing and childcare costs for the households in our sample was below the Minimum Income Standard (MIS) for all but five households. Overall, 32 of the 37 households had average incomes below the MIS despite 32 households having at least one earner throughout the study and 35 households having at least one earner for eight or more months. The lower the household income, the more volatile it was. 26 out of 37 households (70%) had monthly incomes that were highly erratic or falling (19) or erratic (7). Only 11 out of 37 households (30%) had monthly incomes that were broadly stable or rising.

It was the interaction, over time, between changes in earnings and changes in the UC payment that lay at the core of income volatility. Earnings which changed month to month were surprisingly common and unexpectedly large. Indeed, fluctuations in earnings were the norm for the majority of participants, including those with a steady job and monthly pay. For 22 of 37 households, earnings varied month to month by an average of £200 or more and for 28 of 37 households, earnings varied by £500 or more from one month to the next at least once during the period of data collection.

Variability in actual earnings – where monthly pay increased or decreased- was mainly due to changes in working hours and shift patterns, periods of unemployment between jobs and unpaid sick leave which more commonly affected participants working in low paid, temporary and agency jobs, those with zero-hour contracts and the self-employed. More unexpected was the extent of monthly earnings variability among people with higher pay and in more secure forms of employment. Here, it was mostly due to one-off, unconsolidated pay awards, performance bonuses and backdated pay increases. When added to regular earnings, these lump-sum payments could elevate monthly pay by hundreds, sometimes thousands of pounds, causing UC to plunge or fall to nil the following month.

Earnings variability was also due to the way in which earnings data reported by employers to HMRC’s RTI system interacted with a claimant’s assessment period. At certain times of the year, participants paid weekly, fortnightly or four-weekly would have more than one pay packet captured in their assessment period. Even though underlying earnings remained unchanged, a reduced or nil UC payment would result. Monthly-paid employees could be subject to the multiple counting of wages, too, if their regular pay date was close to the start or end of their UC assessment period, or if they were paid early, due to a bank holiday, for example. Affected participants often received less UC than they would otherwise have been entitled to if the equivalent earnings had been paid monthly or spread more evenly throughout the year. Late and incorrect RTI earnings submissions, together with payroll and tax code errors, were other reasons for variability in reported earnings. Under- or over-reporting of earnings by employers meant an under- or overpayment of UC which would later need adjusting. Employers seemed mostly unaware of, or unable to mitigate, the detrimental financial effects these issues could cause for employees on UC.

Changes in earnings, in turn, drove changes in the UC payment. Monthly fluctuations in UC were ubiquitous, frequent and sometimes very large. For 23 of the 37 households, UC payments varied month to month by an average or £100 or more and for 20 of the 37 households, UC payments varied by £400 or more at least once during the year. The way in which the UC payment rose and fell, moreover, did not necessarily reflect changes in earnings in ‘real time’ or in a timely way that helped smooth peaks and troughs in household income. With both wages and UC paid in arrears, it was not uncommon for there to be a time lag of two months between working more or fewer hours and receipt of a lower or higher UC award. Higher earnings could therefore sometimes coincide with an increased UC payment, while lower pay could correspond with a lower or even nil UC payment, amplifying rather than dampening the effects of pay changes. In two-earner couples working variable hours and with different pay frequencies, the effects could be multiplied. A change of circumstance added to the mix could increase income uncertainty further.

Large oscillations in the award from one payment cycle to the next made it hard for claimants to predict when, or by how much, the UC payment would rise or fall, or to know if the amount received was correct. Raising earnings disputes with employers and/or UC was burdensome for working claimants and challenging payment decisions using the online platform was time-consuming. Financial losses caused by pay-related errors and benefit calculations claimants suspected were incorrect would sometimes be absorbed through lack of time and attrition. For working parents required to pay childcare fees in advance, but whose childcare costs were tapered with earnings and refunded in arrears, variability in the UC payment sometimes became unmanageable.

A significant reduction in UC was not easily absorbed by the lowest-income households. Only a little over a quarter of households had any savings they could draw on in the event of a domestic emergency or to cover assessment periods when the UC payment reduced or was nil. Higher earnings from extra hours or lump sum pay awards were often spent on household essentials soon after receipt, leaving little or nothing to set aside. A reduction in or loss of UC could also have serious knock-on effects in terms of the loss of entitlement for work allowances and for other means-tested and discretionary support including reduced council tax, free prescriptions, the Government’s cost of living payments and, in Scotland, the Scottish Child Payment, potentially reducing household income by hundreds of pounds over the year. Small increases in earnings actually made some participants financially worse off. Contrary to the policy intent, this could discourage, rather than incentivise longer hours and higher earnings, particularly among ‘second earners’, most of whom were women.

The regular monthly amounts taken from the UC award in deductions, together with the high rate of benefit withdrawal as earnings rose and other reductions in entitlement, such as the two-child limit, were other factors contributing to low household income. UC payments were also sometimes lower than they should have been as a result of errors and underpayments. Several participants were missing UC elements they were actually entitled to or had had their award erroneously reduced. Bridging the often sizeable gap between the contribution they received towards their housing costs and their rent (or mortgage) and council tax, and between the UC childcare element and their actual childcare costs, were other factors which reduced monthly disposable income.

The combination of low earnings, low benefit payments, high housing costs and ever-rising living expenses meant that, even though someone in the household was working, few participants in this research had levels of disposable income that stayed high enough for long enough to allow them to save or set aside funds. Months in which both earnings and the UC payment fell could sometimes lead to a deficit between income and essential outgoings, generating arrears and fuelling borrowing and debt. Households with regular monthly income deficits found themselves in a perpetual cycle of ‘robbing Peter to pay Paul’ from which many were unable to extricate themselves.

What can be done?

Our main recommendation is that working claimants should be allowed to keep more of their earnings. This could be achieved by reducing the taper rate, increasing the work allowance or extending it to a wide group of claimants (for example, second earners in couples with children). These policies don’t come cheap (although some costs would be offset by an increase in tax revenues) so further research is needed into the relative costs and benefits. Urgent reform is also needed to the way in which UC interacts with ‘passported’ benefits and other means-tested help, particularly as earnings rise - including council tax support, free school meals and free prescriptions. Cliff edges should be removed, earnings thresholds increased and greater standardisation introduced to reduce the inconsistency and unfairness that can arise from having a myriad of different, often discretionary, schemes operating at national and local levels.

Findings also support calls from the JRF and the Trussell Trust for an ‘essentials guarantee’ and for the annual uprating of working-age benefits in line with earnings, as suggested by the Resolution Foundation. Restoring benefits to previous historical levels would help claimants to save and generate a buffer – essential for managing dips in monthly household income and when earnings and the UC payment falls. Without a meaningful savings buffer, monthly assessment simply doesn’t work for claimants whose earnings fluctuate.

Monthly assessment also needs reforming to increase income security for working claimants. One possibility would be to convert the earnings of claimants paid non-monthly into a monthly equivalent before calculating the award. Another option would be to provide longer-term fixed UC awards of three or six months. This could potentially encourage claimants to increase the number of hours worked as they would not face an immediate reduction in benefit. The rigidity of the monthly assessment period also needs challenging. More upfront information and advice should be communicated to new UC applicants about the significance of the date a claim is made, giving them the option to defer the start date if their assessment period does not align well with pay dates - particularly important as the process of managed migration to move claimants of working tax credit across to UC begins to accelerate. Claimants already in work, or who move to a job with a different pay frequency or pay date, should also be given the option of changing their assessment period.

The maximum amount of deductions allowed should be further reduced and historical overpayments and social fund debts should be written off.  Higher and variable rates of deductions for households with earnings should also end. Ensuring UC claimants receive all the elements and exceptions to which they are entitled to is another important priority. Whereas deductions for overpayments and debts are automated, refunding underpayments of UC are not. The onus is rather placed on the claimant to identify errors and have them corrected. Data matching technologies which underpin UC’s automated processes operate highly efficiently in the recovery of benefit overpayments and collection of third-party debts. A similar zeal should drive efforts to ensure that claimants receive all the financial support to which they are legally entitled. Notwithstanding recent changes in the Government’s childcare offer, the childcare element of UC also requires further review. Finally, the research has shone a spotlight on the way in which payroll systems, RTI submissions and remuneration policies and practices can affect UC entitlement. The DWP should make efforts to raise awareness of these issues with employers and engage them in helping to mitigate adverse financial effects for employees on UC.

A further lesson can be drawn which has echoes of the Horizon scandal. In spite of years of research findings, media reporting and lobbying for change from across the political spectrum, successive governments, along with the DWP, have simply ignored the mounting evidence about the aspects of UC that are not working well, or as intended. It’s all too easy to ‘blame the computer’ but If there is one important lesson to take away it’s this. Universal Credit’s biggest failure is not its IT system but the dogged refusal of politicians and policymakers to take legitimate account of the experiences of the people in whose interests these systems are meant to serve.

All articles posted on this blog give the views of the author(s), and not the position of the IPR, nor of the University of Bath.

Posted in: Culture and policy, Data, politics and policy, Evidence and policymaking, Housing, Public services, UK politics, Universal Credit, Welfare and social security, Young people


  • (we won't publish this)

Write a response