The economics of Universal Credit


Executive summary 


The aims of Universal Credit (questions 1 and 2)


The relationship of UC to the labour market (questions 6 and 7)


The aims and objectives of Universal Credit


  1. The aims of Universal Credit, and criteria for evaluating the scheme used in successive business plans, have frequently been revised. The themes stated in most documents as a priority have been:



Other objectives have included:



The design of the benefit


  1. Simplification. UC brings together disparate benefits within a common framework of rules. Pre-existing rules relating to worklessness, incapacity, and housing largely remain, and that means that UC is really a group of benefits clumped together under a shared masthead: that is, a ‘portmanteau’ benefit. There have been many other examples of portmanteau benefits. The best known is probably Supplementary Benefit. SB covered most of the same ground that five benefits did later - means-tested JSA and ESA, much of Housing Benefit, Pension Credit and the Social Fund. SB became complex because of incremental adjustments to take account of complex circumstances. It proved too big for effective policy development; governments argued that personal responsiveness was unworkable, and it had to be adapted to its “mass role” (DHSS, 1978) – a position that was highly controversial at the time.  After its transformation into Income Support, successive client groups (mainly pensioners, unemployed people and recipients of housing support) were broken away from it. Universal Credit has reunited JSA, ESA, Income Support and Housing Benefit, and added in Income Support and Tax Credits besides.


  1. Portmanteau benefits have some limited advantages over multiple benefits: they are able to rely on common rules and procedures (though that much was already true before UC, with common rules on subjects such as income and capital governing many benefits with different names). The principal categories of claim – worklessness, housing costs, child care, incapacity and so forth - still operate under distinct criteria; there is little prospect of true unification. Beyond that, a series of policy decisions have meant that UC cannot be said to have simplified rules in any sense. The complexities include, for example:


                      the absence of clear thresholds, which means that people cannot tell when they are entitled, what they are entitled to, and when they cease to be entitled;

                      ‘whole month’ assessments which mean that people cannot know what they are entitled to, even within days of the benefit falling due;

                      the problem of repaying money that people did not know they should not receive;

                      rules that tell people they must prepare for work when they are not fit to do so;

                      benefits which penalise claimants for circumstances outside their control;

                      the rules about partners and joint claims; and

                      assessments that require people to report changes across multiple dimensions.


  1. Reducing costs. The National Audit Office has told us that “We cannot be certain that Universal Credit will ever be cheaper to administer than the benefits it replacesthe extended timescales and the cost of running Universal Credit compared to the benefits it replaces cause us to conclude that the project is not value for money now, and that its future value for money is unproven.” (NAO, 2018)


  1. Reducing fraud and error. UC is error-prone by design; there are too many conditions, and too many moving parts. Estimates for fraud and error are currently greater for UC than for many of its predecessors (DWP, 2019, p 10). The main factors contributing to fraud and error are earnings and employment, income from occupational and personal pensions and living together. Others include rules about capital, household composition, housing costs and so on. All of these factors have continued to apply in the development of Universal Credit, and they are compounded by the bewildering complexity and unpredictability of the benefit. The main method which has been taken to deal with this is to make claimants personally responsible for repaying benefit - which does not reduce error, only the financial liability borne by government for that error.


  1. Improved efficiency through automation. Benefits are complicated because people’s lives are complicated. A computer can only process data as rapidly as someone can enter it, and in a world where people often do not know if they are employed, what their income is going to be, whether they have a partner or whether they are disabled, there is no way that this kind of information can be processed rapidly, accurately and consistently.  More than 50 years ago, Professor Richard Titmuss was critical of the naivety embodied in ‘computermania. (Titmuss, 1968) Its advocates were supposing that somehow the technology would provide answers to questions that human beings didn’t know how to deal with.


  1.               Personalising benefits through a tailored response. The income provided by Universal Credit is radically unstable. The All Party Parliamentary Group on UC has commented: “many claimants say they cannot understand how their UC is worked out, and it is subject to so many variations that it is far harder to budget on UC that it was on tax credits.” (APPG, 2019) Claimants cannot be certain, even a week before the due date, how much money they will receive in benefit. Incomes fluctuate; they come from multiple sources; they are commonly dependent on the behaviour of third parties, which makes them unpredictable. Liabilities (such as rent) do not necessarily square with the DWP monthly assessment period.  There are further complications for self-employed people: the Institute of Chartered Accountants warned the government that asking for returns every month was unlikely to work for self-employed people who couldn’t even make annual returns without specialist help. (Institute of Chartered Accountants, 2013)


  1. An IPPR report identifies a series of features about UC which are causing problems. They include the five-week wait; monthly payments; the two-child rule; the benefit cap; the high taper rate; the level of benefit; joint payments; sanctions and conditionality; and the management of 'debt' to the system. (IPPR, 2019) Only one of those features - monthly payments, linked to monthly assessments - is integral to the design of the benefit. But UC has other, built-in flaws - problems that exist because of what the benefit is. They include:





Universal Credit and the labour market


  1. The relationship of Universal Credit to the labour market is often represented as the central justification for the benefit’s design and operation. Benefits, however, are there for many other reasons – reasons such as social protection, meeting need, smoothing income, providing a safety net, relieving poverty, financing provision in areas such as housing and child care, insurance and financial inclusion. Most of the claimants of existing benefits – and so, most of the anticipated claimants of UC - will either not be expected to participate in the labour market, or will receive supplements to low income (such as housing support) regardless of whether or not they are in work. The central focus on getting people into work is a distortion - a misunderstanding of what low income benefits are there to do.


  1. Reducing worklessness. Many of the arguments made about worklessness while in receipt of benefit assume that people will choose to be unemployed rather than working. An example is the assertion, made by Charles Murray in Losing Ground, that rational claimants will in general choose not to work.(Murray, 1984) That bears little relationship to anything in the real world. (Atkinson, 1990) Nearly 80% of those who become unemployed after being employed return to work within a year, and most of the remainder take up work in the year after that. Another fallacy can be seen in the claim that there are generations of families who have never worked. Detailed studies have failed to identify any such families (e.g. Kolvin et al, 1990): one group of researchers described the hunt as being like the search for the Yeti.(McDonald et al, 2013)


  1. There are two main groups of people receiving working-age benefits who are less likely than others to engage in the labour market. The first group are older workers. Early retirement has become more common in society because of occupational pensions, and there is a case to argue that Employment and Support Allowance acts for many older males as a substitute for the kind of provision that is otherwise only available to salaried workers with occupational pensions. More than 40% of longer-term claimants are over 50. The second group are people who have limited capability for work. This is the largest category of people of working age who are dependent on benefits in the long term. Under the rules for ESA, it was accepted that it was “not reasonable” to expect people who were ill to work, and the assessment process was intended to identify who fell into that category (Welfare Reform Act 2007, s.1(4)). That test has gradually been supplanted by the different, somewhat ill-defined designation of ‘limited capability for work-related activity’. Treating people who are chronically ill as requiring preparation for ‘work readiness’ is the source of considerable hardship and grievance.


  1. Continuous dependency is largely related to long-term sickness rather than long-term unemployment. The figures for the duration of claims (DWP 2016b) show the extent to which people have been in receipt of benefit for cumulative periods of time, which is not the same thing. Large numbers of people move in and out of employment, and on and off benefits. This is a mark of a precarious or ‘flexible’ labour market, with irregular movement into and out of low paid employment. In theory, this is the group that UC should be best able to serve. This role has however been hampered by unpredictable income flows, assessments based on whole months, conditionality and low work allowances.


  1. The National Audit Office have explained that there is no way of telling whether or not UC does move people into work (NAO, 2018, paras 16, 3.11-3.16.) The extent of any apparent movement is likely to be confounded by other influences, primarily the state of the economy.


  1. Work incentives. The topic of ‘incentives to work’ may refer either (loosely) to the propensity to work at different wage rates, or to factors that motivate a change in economic behaviour. The incentive effects of benefits identified in the past - such as the impact of male unemployment on female labour, or delays in returning to work have been marginal, limited and conditional. This happens because any impact of changes in benefits is drowned out by economic conditions, the demand for labour, wage rates and social norms. It is, then, difficult to say what effect changing the marginal rate of deduction will have on decisions to work.


  1. Many arguments about ‘incentives to work’ are based on the supposed relationship between the rate of benefit available and low-paid employment – a relationship that has more to do with Ricardo’s ‘iron law of wages’ (see Poynter, 1969) than it does with any analysis of incentive effects. ‘Better-off’ calculations have limited relevance to practical incentives, for three reasons. First, the potential advantage of working depends on the circumstances of the individual: most people looking for employment will gauge an offer of employment against their earnings potential, not the rate of benefit. Several European systems offer earnings-related unemployment benefits, on the basis that there is an incentive if the gain from income will be higher. (Earnings-related unemployment benefits in France reduce when unemployment extends into the longer term.) Second, the motivational effect of a change in income depends on costs as well as benefits. An increase or decrease in prospective income is not necessarily an “incentive”. Third, any rational calculation looks to future gains or losses, not just to immediate gains. One of the key questions not addressed by a ‘better-off’ calculation is what happens when things change. The stability of income matters, as well as its level (Goulden, 2010); any calculation which identifies incentives only in terms of income levels is liable to mislead.


  1. In principle, Universal Credit was supposed to make sure that ‘work pays’ by smoothing transitions to and from work. There has been some smoothing of marginal rates of deduction, and an assertion of the value of benefits in work, but on the whole little has been done to integrate benefit systems with the realities of a casualised and precarious labour market. Complexities arise because of the growing importance of marginal labour and sub-employment, including zero-hour contracts and pseudo ‘self-employment’, where employers use nominal self-employed status to avoid the inconvenience of administration, tax and national insurance.


  1. The effect of reducing the ‘work allowance’ is to reduce, and then to end entitlement, to benefit at lower levels of income. This has two effects. The first is to reduce the effectiveness of the benefit in supporting the incomes of people while in work. The second is to lift people off benefit at lower levels of income. That means not just that they will no longer be receiving benefit, but that they will no longer be engaged with Universal Credit. It removes the main reason a former claimant might have to remain in contact with the system; and that implies, in a position wherever incomes are precarious or fluctuating, that a major part of the purpose of UC has been negated.


  1. Reducing in-work poverty. The effect of a taper that is substantially higher than initially proposed tapers has been to reduce the amount of benefit people get while in work. It also increases the marginal rate of deduction, sometimes called the ‘poverty trap’: the combined effect of taxation, national insurance, the withdrawal of UC and the loss of council tax reduction means that people on low earnings are liable to lose up to 75p for every pound they earn. This is again commonly interpreted as a ‘disincentive’, but like other supposed incentives it has never been shown to have a clear effect on behaviour; the real objection is about fairness.




  1. The purposes of UC are defeated by its complexity, and its failure to provide a stable, predictable income. One might say that it needs to adapt to its ‘mass role’ (as in DHSS, 1978). An effective benefit system needs to be simpler, slower to change, more comprehensible and less intrusive. To achieve this, assessments should be based on previous income, not present income; payments should be made on a standard date, not an individual one. Universal Credit needs to be broken up. Adjustments to individual benefits can most effectively be made by having distinct benefits, which in different combinations will lead to income packages that make sense from the perspective of the recipient.




APPG on Universal Credit, 2019, What needs to change in Universal Credit?

A Atkinson, 1990, Institutional features of unemployment insurance and the working of the labour market, London: Suntory-Toyota Centre for Economics and Related Disciplines.

DHSS, 1978, Social Assistance.

DWP, 2010a, 21st Century Welfare, Cm 7913

DWP, 2010b, Universal Credit: welfare that works, Cm 7957

DWP, 2016a, Universal Credit Evaluation Framework.

DWP 2016b, Duration on working-age benefits: Great Britain

DWP, 2018, Universal Credit Programme Full Business Case Summary.

DWP, 2019, Fraud and error in the benefit system.

C Goulden, 2010, Cycles of poverty, unemployment and low pay, York: Joseph Rowntree Foundation.

Institute of Chartered Accountants for England and Wales, 2013, Progress towards the implementation of Universal Credit: written evidence, HC576

IPPR (Institute for Public Policy Research), 2019, Social (in)security.

I Kolvin, I, F Miller, D Scott, S Gatzanis, M Fleeting, 1990, Continuities of deprivation?, Aldershot: Avebury.

R McDonald, T Shildrick, A Furlong, 2013, In search of intergenerational cultures of worklessness, Critical Social Policy doi: 10.1177/0261018313501825

C Murray, 1984, Losing Ground, New York: Basic Books.

NAO (National Audit Office), 2018, Rolling out Universal Credit, HC 1123

J Poynter, 1969, Society and pauperism, London: Routledge and Kegan Paul.

R Titmuss, 1968, Universal and selective social services, in Commitment to Welfare, London: Allen and Unwin.



Paul Spicker is Emeritus Professor of Public Policy at the Robert Gordon University, currently working as a writer and consultant on social policy and administration. His books on social security include Poverty and social security (Routledge, 1993), How social security works (Policy Press, 2011) and What’s wrong with social security benefits? (Policy Press, 2017). In 2007 he was a special advisor to the House of Commons Work and Pensions Committee for their report on the simplification of benefits.



17 February 2020