Beyond Digital Optimism: What Finland Can Learn from the UK’s Universal Credit Experiment

Garima Singh, Minna van Gerven and Aino Salmi

Levana Magnus presenting at the REPAIR+KELA event

In a rapidly digitizing world, the promise of streamlining complex social security systems through technological infrastructure appears increasingly attractive. The recent REPAIR+Kela seminar, held in Helsinki on May 23, 2025, posed a timely and consequential question: Can digitalization simplify complex problems in welfare policy? Drawing on over a decade of experience from the United Kingdom’s Universal Credit (UC) system, the seminar offered Finland both a cautionary tale and a rich repository of institutional insights as it embarks on a similarly ambitious reform.

Lessons from the UK’s Universal Credit

The United Kingdom’s Universal Credit programme described as “digital by design” has been one of the most sweeping transformations in European welfare systems since its inception in 2010. It aimed to unify six different benefits into a single, simplified system, mainly administered online, with the intention of improving efficiency, reducing fraud, and incentivizing work.

However, as Professor Emerita Jane Millar from the University of Bath observed, the digital format did not guarantee the achievement of substantive policy goals. She was of the view that, while the Department for Work and Pensions (DWP) in their report claims “modest gains” in employment for specific groups, such as lone parents and single adults, these effects are marginal and in no way transformative when viewed against the scale of investment and disruption.

Professor Millar went further, noting that the original policy architecture of Universal Credit, particularly its monthly assessment model and single-payment structure, has rendered the system unusually resistant to reform. This rigidity considerably hampers policy implementation and the achievement of the system’s policy goals. In simple words, the design logic of automation has hardened into institutional inflexibility.

The digital welfare is never just technical

The Finnish context differs, but it is not immune to the same structural risks. Kela’s strategic vision for 2035 foresees that the Finnish social security system will harness the potential of technology, automation and artificial intelligence. To support these digitalisation goals, new ICT-systems are procured from the private company, PricewaterhouseCoopers (PwC). These systems will be built on Salesforce architecture and will require a massive investment (estimated at € 589 million over the coming ten years).

In her opening speech, Professor Minna van Gerven from the University of Helsinki cautioned against viewing digitalization in purely administrative or technical terms. “It is also political, institutional, ethical transformation,” she insisted. “Digital welfare reforms are never just about efficiency or innovation. They shape human judgment, social inequality, and the very structure of institutional dependency.”

Digital welfare is not merely about digitalizing specific processes; more importantly, it is about reconfiguring the relationship between citizens and the state. As the speakers in the seminar repeatedly emphasized, the simplification that digitalization promises are contingent upon the capacity to manage and mitigate the technical and institutional harms and burdens that accompany it. Digitalisation should serve policy intent, not the other way around.

Dr Hayley Bennett's (University of Edinburgh) contribution captured the heart of this dilemma. Through empirical research on UK claimants, she shared how automated decisions, real-time income data, and monthly means-testing algorithms introduced new layers of unpredictability for low-income workers. Importantly, these burdens are not just technical glitches; they are political artefacts. When digital systems discount claimant testimony and privilege employer-provided data as inherently authoritative, they obscure accountability and shift the burden of proof onto individuals least equipped to challenge the system. These burdens appear in the form of time demands, financial costs, and emotional strain, ultimately weakening the core principles of social protection.

An exciting and notable theme was the phenomenon of what Bennett termed “algorithmic disentitlement.” Here, the digital infrastructure effectively prevents some claimants from accessing their legal entitlements, not through overt exclusion, but by eroding their capacity to navigate opaque, automated systems. For Finland, currently developing its General Social Security Benefit (yleistuki), this serves as a clear warning. A system that is over-reliant on automation but under-responsive to human contingency may meet its efficiency metrics while failing its normative mandate.

Digital reforms must be built on trust and inclusion

Dr Levana Magnus (University of Bristol) extended this critique to labour market activation policies. The Universal Credit journal, a digital communication tool between claimants and the Department for Work and Pensions (DWP) is central to monitoring job-seeking behaviour. Magnus’ research highlighted how it has become a digital blanket that enforces compliance without necessarily supporting employability. While some claimants found digital tools to be convenient or even empowering, many others were trapped in miscommunication loops, surveillance pressures, and inconsistent enforcement (e.g. of what is considered constituting as “36 hours active work search”). The reduction in human discretion was not replaced by fairness or consistency, but by algorithmic rigidity and, at worst, digital diversion.

Finland’s approach must therefore remain attentive not just to what digitalization enables, but what it displaces. Kela’s aspiration to reach a 70 to 80 per cent automation rate by 2035 is, although technologically possible, challenging from the normative aspects of social protection. Increasing the level of automation without deliberation introduces potential risks for example regarding the access to social rights. As was repeatedly emphasised at the seminar, the path to inclusive, responsive and trust-based digital welfare lies not in technocratic optimism, but in sustained critical reflection, co-designed interfaces, and accountable governance.

As Professor Minna van Gerven noted in her opening remarks, Finland is at a turning point, with several extensive social security reforms currently being developed.  “The transformation is going to rely heavily on digital infrastructure and AI-assisted services. But we cannot afford to lose sight of the human element. These reforms must be socially legible, institutionally flexible, and democratically accountable.” Designing for digital resilience, then, means designing for plurality, contingency, and trust. It is crucial to create systems that endure over time and accommodate change. As client situations and policies change, technologies must support both, not hinder their progress.

REPAIR consortium organised the seminar “Lessons from the UK’s Universal Credit System - Can digitalisation simplify complex problems?” in collaboration with KELA on May 23, 2025.