Administrative Burdens and Digital Welfare: Are Technical Features of Universal Credit Problematic?

This blog is based on an article in the Journal of Social Policy by Hayley Bennett, Morgan Currie, and Lena Podoletz. Click here to access article.

Social policy scholars have long captured the frustrations of citizens navigating byzantine welfare benefit systems. In 2015, academics David Moynihan, Pamela Herd and Hope Harvey proposed that the complexities found in getting access to social security may be intentional – politically motivated hurdles, or simply clunky design, can make the effort off receiving an entitlement too hard, putting people off of the process altogether. Moynihan et al. call these obstacles ‘administrative burdens’. In recent years numerous governments and public agencies have proposed investment in digital welfare as a solution to administrative costs and inefficiencies. But does a move to digital welfare reduce, remake, or create more administrative burdens?  

Administrative burdens: Examining technical features and the systems layer

In our study of Universal Credit, the UK’s largest social security payment and a supposed international leader of digital welfare, we explore this concept of administrative burdens and find that the complexity and burdens of a social security system can be a function of its technical design. Rather than reducing complexity for claimants, digital welfare involves the creation of new administrative processes that lead to new burdens.

Universal credit is largely digital – claimants interact with UC staff through an online platform and receive a payment that is automatically calculated anew each month, after government systems ingest data about an individual claimant or a couple. Most of this data will be static – such as whether you have dependents, your age bracket, relationship status, and whether you have a disability – but for those who are in work (especially workers in precarious employment positions or those who work variable hours), monthly wages can lead to fluctuations in the amount of Universal Credit deposited in bank accounts each month. 

We propose that this automated payment system – based on a monthly means-tested calculation – introduces specific burdens for Universal Credit claimants. These difficulties may not be intentionally there to put people off receiving the benefit, but they illustrate the political dimensions of Universal Credit’s technical features, which, we found, affect paths to accountability and legal entitlements. We argue that these technical features should be given more consideration by social policy scholars and Universal Credit’s designers.

Conceptualising costs associated with administrative burdens

We found that the administrative burdens of this systems layer of Universal Credit entail three costs: temporal, financial, and emotional. Our findings are based on a qualitative study of 44 claimants in Scotland; 25 of them reported to us monthly or bi-monthly for six to 12 months in 2022-2023 over text or phone calls, allowing us to hear first-hand what people experienced when they received their monthly payment. We detail two design features of UC that participants identified as burdensome:

1. The complexity of Universal Credit’s dynamic payment calculation, which often changes month to month for working claimants, prevents people from anticipating their income and, in some cases, understanding if their entitlement is correct. The resulting confusion causes anxiety and affects household budgeting.

2. The dynamic payment calculation system can introduce errors originating from inaccurate earnings reported by HMRC. For Universal Credit recipients who are in-work, the adjustments to Universal Credit income based on monthly earnings happen automatically through a data exchange with the HMRC’s Real Time Information (RTI) system. Employers report their employees’ earnings using payroll software to RTI for tax purposes, and the DWP accesses this stream of data daily. We found these RTI errors took months to resolve, if they ever were, and could only be rectified by the employers. Some claimants struggled to cover bills during this period. 

Our research shows that low-income workers, whose payments are dynamically adjusted each month and are prone to erroneous and unpredictable payments, experience these financial, temporal and emotional costs of Universal Credit the most. This is especially troubling as single mothers (often working in unstable employment or with variable hours) make up the largest group of workers on Universal Credit. We also find that technical features have bearings on claimants’ rights, affecting accountability procedures when payment are incorrect – as happens sometimes with the RTI earnings data – which can lead to a loss of legal entitlements. We call for a reconsideration of these systems level features of Universal Credit that affect our lowest-income citizens.

What next? 

We argue that the Department for Work and Pensions could reduce costs experienced by claimants in three key ways. First, by supporting claimants’ understanding of their monthly pay by offering a more granular breakdown of their calculation in their online account. Second, by increasing human contact in the system to enable claimants to talk to empowered staff who can rectify errors. Third, by creating a faster disputes process that allows claimants to submit evidence of earnings and to enter a mandatory reconsideration request if pay is not adjusted in a timely way. While such administrative fixes may reduce some of the main administrative burdens and related issues, they do not address other concerns raised by social policy scholars about Universal Credit design, including conditionality instruments or low benefit levels. Nor would they address fundamental limitations to the algorithmic calculation causing monthly variable payments. 

We also argue that social policy scholars have long examined the ways in which policymaking and street-level implementation are political processes, producing inequalities and resource disparities. But do we give enough attention to the technical level in social policy analysis? The details of technical design are not apolitical delivery features. Left unchecked and underexamined automation and digital reforms can enhance state surveillance and control which can have substantial effects on social security recipients and potentially undermine the values and purpose of our social security policies. It is therefore necessary for social policy scholars to increasingly examine the rise of digital welfare. 

---

About the authors

Hayley Bennett is Lecturer at the University of Edinburgh.

Morgan Currie is Senior Lecturer at the University of Edinburgh.

Lena Podoletz is Lecturer at Lancaster University.