Ms Fran Bennett:

I thank the committee for the invitation to give evidence. I have been involved, often with Jane Millar and-or other colleagues, in analysis of universal credit since 2010. Based on that, I think it is crucial to consider not only implementation and administration issues but also structural and policy concerns. That is what I will focus on but first, I wanted to raise several more general issues.

First, social security systems should, and usually do, involve a mix of universal, insurance and means-tested benefits. The policy attention paid to universal credit over recent years, which takes the UK firmly down the means-testing route, has resulted in the neglect of non-means-tested benefits. It would be better, in my view, to retain a balance between these different kinds of benefits and be more aware of how they interact or not. In my view, universal and insurance benefits should be retained and as far as possible improved.

Second, universal credit integrates means-tested benefits, providing an income for individuals or households both in and out of employment, as well as those helping with costs, for example, for housing and childcare. This also involves integrating different benefits paid by different administrative bodies. Both the multipurpose nature of universal credit and, conversely, the concentration of power over benefit awards in one official body create significant structural and policy problems. A more limited or focused approach would help avoid many such problems. Third, discussions of amalgamating means-tested benefits should, but sometimes do not, include the fact that needs, as well as income and capital, must be assessed.

I will talk a bit about simplification and the fact that the simplification resulting from integration can cause problems for claimants. Universal credit is, of course, a misnomer. This is not a universal benefit but a means-tested benefit and the word "universal" here means comprehensive.

It is not a credit; rather, as Professor Millar said, it is paid in arrears.

Simplification for claimants and administrators is a key advantage claimed for integrating means-tested benefits and is also meant to save on administration. However, in that amalgamation, and especially in automating the calculation of the integrated benefit, the algorithmic formula employed has a rigidity that limits policy flexibility. Professor Millar and I have written about this.

Universal credit is assessed on a monthly basis, using a cash flow-based accounting approach to income and a whole-month approach to changes of circumstances. These cause numerous problems for claimants, but it appears that cannot be altered, that is, without driving a coach and horses through the core features of universal cried. Thus, all income received in the month before an assessment day, no matter which period it relates to, counts when calculating that month’s universal credit. For employees, this is done via real time information, RTI, which is an automated system employers use to report earnings. RTI was apparently the clinching argument for the treasury to support universal credit because it meant significant administrative savings.

However, weekly- or fortnightly-paid workers may have differing numbers of paydays in any month’s universal credit's assessment, or an employee may get a bonus for several months’ work in one go. Outside the RTI system, someone getting a contributory benefit may be paid fortnightly, not monthly, for example. All this income, with any work allowance and withdrawal rate through a taper, reduces that month’s universal credit, causing income volatility and financial insecurity, making budgeting trickier. Only about a week’s notice is given of that month’s universal credit payment.

The Government has recognised that monthly-paid employees may sometimes be paid twice in one month, for example, due to a bank holiday coming up, and these people can ask for their universal credit not to be affected like that. However, otherwise, the Government has refused any change, arguing that the formula must have clear, bright lines for automation to be able to work. Self-employed people, which we may say a bit more about later, must report their relevant income for universal credit monthly, which may be challenging for them.

The whole-month approach to changes of circumstances, about which there is little or less debate, means that regardless of when a change happens, it is assumed to apply for the whole month. This can be advantageous, for example, if a baby is born later in the month but before the assessment day, you get more money than you would if it was pro rata. However, if a young adult moves out of your household just before the assessment day, no money will be allowed for them that month. This formula rigidity can undermine financial security, and it must feel very arbitrary to people.However, the Government says it is impossible to change the universal credit calculation to apply such changes pro rata instead.

Therefore, the simplification necessary to integrate different kinds of benefits and make administrative savings results in rough justice and precarity of income for claimants.

I wish to say a bit about gender issues. Means-tested benefits in the UK used to be claimed by one partner in a couple, including additions for dependants. I believe that is what happens in Ireland as well. Joint claims were introduced into means-tested benefits and tax credits over time for couples, and they now apply to universal credits as well. The policy goal was activation, in other words, to be able to apply conditionality to the other partner involved. However, with joint claims, payment for couples used to vary. Working tax credit, for example, was paid to the main earner and child tax credit was paid to the main carer, as identified by the couple. This could mean that each partner got some means-tested benefit themselves, or, more accurately, they could nominate the bank account for payment for that benefit. That could be crucial in cases of financial coercion and in general could help make relationships more equal.

However, these different means-tested benefits have now been amalgamated in universal credit, which is due monthly in one lump sum. The couple is given the choice of bank account, though couples with children are now nudged to nominate the main carer for payment. However, other than by discretion, in cases of financial coercion or money mismanagement, for example, the payment cannot be split between partners.

Child benefit has remained separate and is usually paid to mothers. Means-tested council tax support, if any, is also calculated and paid separately. If either partner has an individual non-means-tested benefit, that is paid to them too - one reason why we should retain these and not merge them with means-tested benefits. A couple’s universal credit may go into one partner’s individual account or it may instead be paid into a joint account, but research shows that both partners do not necessarily have equal access to joint accounts.

Scotland is committed to exploring ways to make separate universal credit payments to each partner, ideally by default. The Government, however, argues that universal credit is an all-in-one benefit and is adjusted for other income and so on as a whole; therefore, universal credit cannot be broken down into its component parts, meaning it is impossible to, for example, pay the child element of universal credit to the main carer. It is hard to see how to decide who should get what in a couple, particularly if one partner has some other income already, such as earnings. To date, Scotland has been unable to implement its plan.

In addition, universal credit's structure gives priority to having one earner in a couple deliberately, meaning insufficient incentives to have a so-called second earner. Arrangements for recouping childcare costs can also have disincentive effects. Although costs can now be paid upfront in limited circumstances, they are usually paid in arrears. Those problems are likely to mean fewer women in particular having an independent income through their own earnings.

Professor Millar also mentioned that measuring success is difficult with universal credit. The National Audit Office suggested it would be challenging to measure universal credit's success in increasing numbers in employment. In addition, while improving take-up and thereby reducing poverty was one justification for universal credit, a method for estimating take-up universal credit has not been developed yet. In fact, annual estimates of take-up of legacy benefits for working-age means-tested benefits are also now not published - only for pensioners. Therefore, it appears that several rationales for introducing universal credit cannot be tested either.