In response to the reset of the Universal Credit project, DWP built a highly effective policy-delivery machine. With this in place, the department has demonstrated it has the capability to make changes to the Universal Credit system quickly and competently. These interventions have attempted to mitigate issues with the original policy design, but much of the conceptual framework and political assumptions - the monthly cycle, assumptions about family life, what ‘normal work’ looks like, finances and housing situation - remain unchallenged despite clear issues.

From its origins, Universal Credit was designed to be complex - the aim was to allow policymakers to reward or punish detailed combinations of circumstances and behaviours. This complexity, along with the political assumptions of the original policy, has conspired with the variation of people’s lives to push administrative, financial and cognitive burden onto claimants. This has been done under cover of ‘personal responsibility’ and ‘mirroring the world of work’, but the reality is that it creates contradictions. Many of the desired outcomes of the policy have become harder to achieve because of the desire to means-test and to (attempt to) control people’s behaviour.

Mirroring an idealised view of the world of work

Despite the aspirations for Universal Credit to ‘mirror the world of work’, the month-long assessment period does not match many people’s experience of work or their financial needs. Emblematic of this is the growing use of advance payments and the increased demand for food banks it creates. DWP has also lobbied big employers that pay their employees every 2 or 4 weeks to change their employment practices to fit with the policy.¹

Advance payments are the current approach to mitigating the problem of claimant’s lives not fitting with the monthly payment cycle, while leaving the dogma of monthly payments in place. The fact that 60% of applicants apply for an advance means that neither the policy or the digital account are designed for the common case where most people need support sooner. It is hard to see how a system that puts the majority of new users directly into debt and reduces future payments encourages financial responsibility. Or how it meets the aim, set out in the business case, of creating a “financial safety net” for the public that can be tailored to “those who need it most”. ² Further, the account is not optimised for a world where people are paid on a more ad hoc basis, potentially from multiple employers, or have a payday that is out of sync with the Universal Credit cycle. This surely creates some friction and uncertainty for claimants wanting to take an additional job and increase their income.

Complexity and burden

In spite of DWP’s removal of ‘waiting days’, the payment cycle remains conceptually complex. The complexity of the calculation, combined with the multiple dates claimants must internalise and the fact that the statement is not available until the end of the month, mean understanding how much money to expect is no simple task. It is hard to see how this complexity and absence of information fits with the aim to encourage personal responsibility.

The Child Poverty Action Group have suggested adding extra information to the statement to help improve understanding.³ Based on reports and screenshots that appeared on the RightsNet website in December 2019, DWP does appear to be trialling new designs for the statement that explain the breakdown of payments, advances and the payment cycle in new ways.⁴ However, while adding extra information may help, it also carries the risk of confusing users, so must be done with care. Additionally there is a security and privacy dimension to adding extra information, much of which is sensitive, given these pages may be accessed via public computers in libraries. The complexity caused by the monthly assessment process and the hyper-means-tested nature of Universal Credit is something that good design can only do so much to patch over. It may be that designing a statement that can be understood without assistance, while covering all circumstances may be an intractable problem without a simplification of the conceptual framework of Universal Credit.

Hyper-means-testing creates a system where claimants must fully internalise a list of changes of circumstance that even the department is unable to collate in a publishable form. Each time an exception is added to the policy, such as the two-child limit or reductions for “extra” bedrooms, the administrative burden and the risk of errors must surely increase. Complexity is also present in the amount of guidance the department published for claimants to read and understand the totality of their obligations. Combined, these place cognitive and administrative burden onto claimants, both in terms of understanding the system and in planning their finances. Claimants with complex circumstances, more paperwork, less digital skills or less access to the internet are further penalised because the first payment date is linked to the date of the complete submission of a claim.

Given one of the objectives from the Universal Credit business case was to simplify a complex system, it begs the question: “complex for whom?”. Complex for government? Or complex for the public?

Personalisation and choices

This responsibility paradox can also be seen in Universal Credit’s approach to ‘personalisation’. Personalisation in Universal Credit is more about giving the government the ability to personalise the demands placed on the public than allowing claimants to take responsibility and personalise their interactions with the government in a way that works for them.

DWP has shown that it is possible to add flexibility to the system while keeping it manageable for the department. For example:

• the changes made to the system in Scotland show that it is technically possible to apply different payment cycles to particular claims ⁵
• the Scottish changes to allow claimants to request direct payments, and the development of the Landlord Portal and other tools, show that separating the housing element of Universal Credit is achievable
• signposting the use of advances on the homepage and making those payments faster than the default seven day wait

Rather than offering all claimants genuine opportunities to make these choices, Universal Credit remains optimised for choices the government wants to make for them based on an idea of what ‘normal’ looks like. Where choices are available, they are obfuscated rather than being designed into the service. This means that, when it comes to the process by which payment cycles are changed, payments split between couples or rent paid directly to landlords, the defaults remain inline with the original conceptual framework of Universal Credit. This is despite the evidence of the problems they cause. In turn, this creates a complex set of processes for handling exceptions.

As detailed earlier, landlords are given an explicit way to request the direct payment of rent, but claimants (outside Scotland) are not. Scottish claimants are given the option, but not upfront during the application process. Instead, they are handled via the journal, and do not take effect until after the first month. For claimants in the rest of the UK, the process of changing the payment cycle is even more obfuscated. As has also been noted, some elements of Universal Credit, such as appeals have not been digitised at all, despite a clear user need.

However, accusing DWP of not being user-focused in their approach to design and development would be unfair. The NAO reports clearly show that DWP makes heavy use of prototyping and user research to understand the needs of different types of users during the design and development process. The department can (and does) make changes to the system when it identifies a need (not something that can be taken for granted in other areas of government). The implication, therefore, is that these things represent policy choices - technically it is possible to offer more flexibility, but the desire to attempt to force the public into a particular pattern of behaviour wins out.

Taken together, the way the policy frames ‘responsibility’ along with the desire for policy complexity generate a considerable administrative burden, which is placed on claimants rather than government. They also create a system where ‘personalisation’ is one-sided, with some processes that could be digitised, not being digitised. The result of this is that the benefits of digitisation are not shared equally between the government and the public. The next section looks at some opportunities to redress that balance.