As the roll-out of Universal Credit gets closer to completion, there is an opportunity for the government to use the capability the DWP have built up to revisit some of the core policy assumptions (now nearly a decade old) and to better share some of the benefits of digitisation more equally with the public. It can do this by giving claimants more choice, looking for opportunities to automate processes for claimants and understanding the true costs of digitisation for claimants. DWP could and should be actively looking for opportunities to reduce complexity for claimants, not just for government. This section sets out some ways it might do this.

A benefit system designed to recognise the differences in people’s lives

Firstly, the concept of ‘responsibility’ in Universal Credit should be redefined to become less about conformity with a particular worldview, and more about ensuring the design of the policy meets the desired outcomes with the grain of people’s lives, rather than against it. By helping claimants choose what works best for their circumstances, the department has the opportunity to make it easier for claimants to meet their responsibilities.

Claimants should be able to choose a 2-week cycle, then potentially transition to a monthly cycle over time, for example, if they start being paid monthly. They should also be able to realign their payment period if they take a new job and the employer’s choice of payday is causing problems calculating Universal Credit.

Following appropriate user research, choices around payment cycles, direct payments and split payments should be designed into the application process and following significant changes of circumstances. This should be the default for all claimants of Universal Credit rather than continuing to treat them as exceptions. Hopefully, this would have the additional effect of reducing the number of claimants getting into debt to DWP by claiming advances.

The same approach should be taken with splitting payments between couples and making direct payments to landlords. Given the underlying system can apparently handle them, there can be no technical reason why these cannot be provided as an option during the application process. Making these opt-in, rather than means-tested on the basis of an abusive relationship or chaotic lifestyle, would have the added benefit of no longer having to collect and hold sensitive data about people’s relationship and housing status.

Claimants should also be signposted to additional support during the application process and during the initial weeks of a claim, such as free travel passes provided by local government or support from food banks. Over time, these could become automated, so that (with permission) the digital account could actively apply for help on behalf of a claimant.

Using automation to remove administrative burden for claimants, not just from DWP

There are other opportunities to use the digital account beyond offering more choices to claimants. Despite the many documented problems with the policy, the concept of using earnings data to calculate working-age benefits automatically is the insight from the architects of Universal Credit that will likely stand the test of time. Today, DWP’s plans for automation appear to be aimed at reducing its costs in administering the welfare system. This is hardly surprising, given it is in line with the business case, but represents a missed opportunity. Automation could also be used to help claimants meet their responsibilities under Universal Credit more easily by removing administrative burden and reducing the likelihood of errors.

These include, but are not limited to:

• using the data government holds in the form of RTI to recommend the best payment cycle for a claimant - e.g. spotting when someone has started a regular second job
• using the data DWP holds about claimants, along with its risking capability to turn the reporting of changes-of-circumstances from an active one into a more passive one - prompting claimants to reconfirm, for example, when it is likely that a child’s education status may have changed
• with appropriate consent, service design and ethics oversight, using data sources not held by the department through the use of ‘verifiable proofs’, for example from child-care providers or banks to remove the need to actively report changes

Universal Credit should also become much more real-time and responsive.

One of the characteristics of today’s digital services is that they operate in real-time. If you want to buy something on Amazon, you see if it’s out-of-stock immediately; the locations of cars on Uber are displayed in real-time, and prices change based on demand; users sending money between Starling or Monzo bank accounts receive immediate notifications that the transaction has taken place. Currently, Universal Credit does not display the real-time characteristics of a modern digital service.

Despite the presence of words ‘real’ and ‘time’ in the acronym RTI, this is ‘real time’ in HMRC terms (monthly rather than annually, good enough for the calculation of tax) – there is a time-lag in retrieving and duplicating the data from HMRC and even then, it is not displayed to claimants until the end of each month. DWP and HMRC should aim to move to a world where the data is accessed directly, from a single source (with appropriate safeguards), rather than duplicated using legacy-style batch processes. This would mean that errors in the reporting of wages could be caught sooner by claimants and resolved faster by DWP staff.

VocaLink, the organisation which operates the UK’s payments infrastructure, has said that it believes the government underutilises the Faster Payments systems in making welfare payments.¹ DWP should also use a combination of the Faster Payments system and automation of the calculation to shorten and eventually eliminate the seven-day calculation and payment period. The automation aspect of this should be done cautiously to ensure accuracy of payments does not fall.

Finally, the messages that are sent when new journal entries are added, or items added to the to-do list should be revisited to see if extra context can be added to help claimants make informed decisions about the urgency of a particular message (for example, if they are currently unable to access the internet). DWP should also consider options for claimants to submit evidence of searching for work to the journal via other mechanisms such as speech-to-text over the phone, SMS, and direct from third party services.

Costs and benefits

Not all the problems of Universal Credit can be fixed with the application of more digital service design and automation. While there are clearly opportunities to improve how the service works for claimants, there will always be some needs that DWP is unable to meet or prioritise, either because of its institutional incentives or the potential cost implications.

Where there are cost implications for government (such as greater use of Faster Payments), DWP should be open about those costs. It should also be open about the costs of inaction to the rest of the public purse (such as the costs to Ministry of Justice of additional demands on the tribunal system). It should commission research on the financial cost of using aspects of the digital service for claimants (including additional mobile phone bills, charges for the use of internet-cafes, helpline scams and other costs).

The other thing that DWP should do to ensure all the needs of claimants can be met is to start to open up Universal Credit. DWP should start to think of the welfare system less as a closed system, and more as a shared platform that digital and non-digital services can be built on top of and that can cater for a broader range of use-cases. The idea of rethinking Universal Credit as a platform is considered in the next section.