This page details the projects that were awarded funding as part of Phase 4 of the Evaluation Accelerator Fund (EAF).
Background
In March 2025, the Evaluation Task Force (ETF) announced Phase 4 of its Evaluation Accelerator Fund (EAF). The EAF is a competitive grant fund and supports research and evaluation projects that:
- create actionable evidence in government priority areas that informs public spending or policy decisions;
- tackle evidence gaps in government priority areas; or
- provide robust evidence of financial or efficiency savings from new policies, interventions, or innovative approaches to service delivery
Phase 4 had up to £3.5 million available in the 2025/26 financial year. The government priority areas in scope for Phase 4 were:
- missions
- priority areas for public sector reform outlined at the Autumn Budget 2024
- use of technology, including AI, in the public sector (in order to digitise public service delivery, enhance productivity and improve outcomes)
Focus on prevention
Many projects awarded funding focus on evaluating the impacts of prevention (improving long-term outcomes for people through a focus on prevention, relying less on expensive crisis management). Spending Review 2025 identified prevention as a key principle of the government's approach to public sector reform. We have therefore also identified projects that relate to prevention.
Funded projects
Using quasi-experimental data to evaluate effectiveness and preparedness of NHS elective surgery Hubs
- Organisation: University of Birmingham
- Funding awarded (FY25-26): £149,852
- Government priority area(s):
- Mission: Health
This project will evaluate the impact of NHS elective surgery Hubs on surgical volume, safety, and preparedness using quasi-experimental methods and national datasets. The project will consider preparedness for external pressures including future pandemics, climate change (heatwaves), conflicts, and annual winter pressures. It will also consider adoption and readiness for robotic surgery in the Hubs. Rapid analysis will deliver an England-wide performance report and outputs for policymakers. Findings will inform decisions on operational delivery and long-term funding. This work will support a scalable, future-proof surgical model aligned to the mission of Building an NHS Fit for the Future.
Out of Court Resolutions: Impact assessment of Outcome 22
- Organisation: University of Birmingham
- Funding awarded (FY25-26): £185,935
- Government priority area(s):
- Mission: Safer Streets
- Public sector reform: Police
- Public sector reform: Prevention
This project will evaluate the impact of Out of Court Resolution (OOCR) diversion options reported as Outcome 22 on reoffending and recontact with the police. Outcome 22 is used for incidents where no further action is taken, and a diversionary intervention has been undertaken to address offending behaviour or prevent further offending. The evaluation will involve a 4-month scoping phase using a quasi-experimental design to understand how diversions that led to cases being closed as Outcome 22 affects reoffending and recontact with the police, followed by a 6-month pilot Randomised Controlled Trial (RCT). This impact analysis will be supplemented by the development of a cost effectiveness framework.
Targeting housing and debt advice to prevent homelessness: Evaluating Change Please's AI chatbot
- Organisation: Centre for Homelessness Impact
- Funding awarded (FY25-26): £255,886
- Government priority area(s):
- Mission: Opportunity
- Public sector reform: Homelessness
- Public sector reform: Prevention
- Technology/AI
The government is committed to trying to improve public services and outcomes using artificial intelligence. Large language models have the potential to have transformative impacts in the field of housing and homelessness, by providing resources to people at risk of homelessness to prevent them from facing eviction, and hence preventing homelessness downstream. However, these tools require testing. This project will test one such AI tool, developed by Change Please, using a randomised encouragement trial. It will also produce guidance for evaluating other AI tools with similar use cases.
AI-enhanced supervision of police stop and search powers: A multi-site evaluation
- Organisation: College of Policing
- Funding awarded (FY25-26): £108,423
- Government priority area(s):
- Mission: Safer Streets
- Public sector reform: Police
- Technology/AI
Police stop and search powers are less likely to reduce crime and more likely to damage public confidence when the grounds for their use are weak. Innovations that strengthen grounds for stop and search may therefore support delivery of the Government's safer streets mission. The College of Policing will evaluate an AI tool designed to enhance the supervision of the grounds for stop and search by sergeants, with the aim of improving the use of stop and search by constables. The tool's impact on supervisory effectiveness, productivity and efficiency will be tested in a quasi-experiment in up to 16 police forces.
Follow-up evaluation of Video First Response
- Organisation: College of Policing
- Funding awarded (FY25-26): £59,045
- Government priority area(s):
- Mission: Safer Streets
- Public sector reform: Police
- Technology/AI
This follow-up evaluation will look at the impact on criminal justice outcomes of responding to domestic abuse calls using video response (VFR). The evaluation will assess charge and arrest rates, time taken to arrest and charge, and time taken for investigations after the initial VFR call. Case file reviews and interviews will explore how evidence captured via video is used during the investigation and charging process. The findings will inform the national roll out of rapid video response (RVR) and policy and operational learning across the police service.
Evaluating the Impact of Family Safeguarding on Children's Social Care Outcomes in England
- Organisation: Foundations
- Funding awarded (FY25-26): £237,214
- Government priority area(s):
- Public sector reform: Children's Social Care
- Public sector reform: Prevention
This project will deliver both impact and implementation and process evaluations of Family Safeguarding, a model of social work practice. Family Safeguarding is a multi-disciplinary and multi-agency whole family approach to working with children and families in children's services. Family Safeguarding teams work with children aged pre-birth to 17 facing issues with abuse and neglect by their families. The model is popular and currently implemented in 24 local authorities. The impact evaluation will use a quasi-experimental difference-in-differences design, and findings will be used to inform the DfE's approach to reforming children's social care, prevention and keeping children safe.
Optimising police deployment strategies to advance the safer streets mission: a quasi-experimental evaluation of Police Uplift Programme outcomes across forces
- Organisation: University of Manchester
- Funding awarded (FY25-26): £38,368
- Government priority area(s):
- Mission: Safer Streets
- Public sector reform: Prevention
This project will evaluate how police forces deployed recruits from the Police Uplift Programme, and which strategies most effectively reduced crime. By comparing forces that adopted different recruitment and deployment approaches, the study will identify which models delivered measurable reductions in crime, particularly knife crime, violence against women and girls, and antisocial behaviour. The project plans to apply difference-in-differences and Causal Forests methods to estimate causal effects. Findings will provide police leaders with actionable guidance to maximise the impact of uplift investment, supporting strategic deployment decisions aligned with the Safer Streets mission.
Development of Machine Learning techniques for evaluation of irregular migration interventions
- Organisation: Home Office
- Funding awarded (FY25-26): £200,000
- Government priority area(s):
- Public sector reform: Asylum
Secure borders are one of the three foundations set out in the Government's Plan for Change. Reducing the number of people entering the UK irregularly (in particular, those arriving by small boat) is a key part of restoring order to the asylum system. Asylum has also been identified as a priority area for Public sector reform. A feasibility study funded by the Evaluator Accelerator Fund successfully applied machine learning techniques to the evaluation of irregular migration interventions, providing valuable insights. This project will improve the robustness of these techniques, apply improved models to evaluate various new key interventions and develop a framework for the ongoing application of these techniques.
Good Work AI Evaluation: measuring productivity, accuracy and job quality in relation to DWP's agent-assist tools
- Organisation: Institute for the Future of Work
- Funding awarded (FY25-26): £179,822
- Government priority area(s):
- Public sector reform: Civil Service
- Technology/AI
This project will deliver a multi-stream evaluation of two agent-facing AI tools deployed by the Department for Work and Pensions (DWP) to support frontline decision-making and improve service delivery. The study plans to use a quasi-experimental design to assess impacts on productivity and job quality. Mixed methods, including analysis of administrative data and interviews, will be used to evaluate AI adoption in the workplace and generate actionable findings on how to best use AI in the public sector. The evaluation framework developed through this project will also be applied to two further use cases in the planning and design phases. The project will also develop and test a Good Work AI Evaluation Playbook to support the UK Government's AI Opportunities Action Plan commitment to adopt high-performing, trustworthy AI at scale by facilitating further comprehensive evaluations of AI adoption in government.
NHS Health Checks Longitudinal Population-Wide Clinical Outcomes and Health Economic Value Assessment (NOVA): A 2015-2025 Propensity-Score-Matched, Quasi Experimental Evaluation
- Organisation: Imperial College London
- Funding awarded (FY25-26): £251,466
- Government priority area(s):
- Mission: Health
- Public sector reform: Prevention
This project will deliver the first comprehensive, longitudinal evaluation to detail the impact of the NHS Health Check programme and guide evidence-based health policy for maximising population health impact. England's NHS Health Check is the world's largest systematic population health programme. Despite local authorities administering an annual £70 million budget for delivery, attempts at evaluation have been flawed and inconclusive - principally due to a gap in large-scale, representative data assets and robust methodology for comprehensively informing longitudinal clinical and health economic impact. We will address this gap using 10+ years of granular retrospective data from 2.7 million patients reflecting England-wide diversity. A quantitative, quasi-experimental methodology using propensity-score matching will be paired with measurement of non-engagement through qualitative community engagement frameworks.
Impact evaluation of the use of Global Positioning System Electronic Monitoring in community sentences
- Organisation: Ministry of Justice
- Funding awarded (FY25-26): £103,000
- Government priority area(s):
- Mission: Safer Streets
- Public sector reform: Prisons
- Technology/AI
This project will deliver a retrospective impact evaluation to address a critical evidence gap on the effectiveness of Global Positioning System (GPS) Electronic Monitoring (EM) for community sentences in England and Wales. The evaluation will employ quasi-experimental methods using linked administrative justice data. This will provide robust new evidence on the impacts of this increasingly used technology to reduce reoffending - a significant priority as part of the Safer Streets Mission of the Government.
Split Standing Charge Trial: Encouraging demand flexibility to meet clean power goals
- Organisation: Ofgem
- Funding awarded (FY25-26): £386,200
- Government priority area(s):
- Mission: Clean Energy
This project will explore the impact of a daily price incentive to reduce peak energy demand through reduction of the standing charge. It will provide evidence to support a number of Ofgem workstreams on flexibility, pricing and the Cost Allocation and Recovery Review. The randomised controlled trial will collect half-hourly energy usage data over ~9 months to explore how the peak time price signal impacts consumer behaviour. The project will also collect qualitative and quantitative data on consumers' experiences.
Measuring what matters: Driving improvements to children's social care sufficiency in England
- Organisation: Ofsted
- Funding awarded (FY25-26): £220,388
- Government priority area(s):
- Mission: Opportunity
- Public sector reform: Children's Social Care
This project focuses on the government's aims to improve the sufficiency of children's social care. Where residential care is needed, it is often in the wrong place and may not cater to the specific needs of a child. The reforms aim to reduce the overall number of children in care, but in cases where children do need care, the reforms seek to ensure that children are in provision that gives them a positive experience and the best start in life. This project will inform and evaluate the implementation of children's social care reforms by quantifying sufficiency challenges across local authorities, evaluating what currently works well in the face of these challenges and sharing this good practice. The project will develop innovative metrics to assess the sufficiency landscape and identify key risk factors for sufficiency challenges, which will support future robust evaluation. Qualitative research will be used to uncover good practice and successful interventions. The findings will build the evidence base to evaluate the future success of reforms.
Evaluating the Impact of Waiting Times on Labour Market Outcomes Across Elective Procedures and Emergency Departments
- Organisation: Office for National Statistics
- Funding awarded (FY25-26): £219,813
- Government priority area(s):
- Mission: Health
- Mission: Growth
This project will evaluate the impact of NHS waiting times on patients' labour market outcomes across two critical areas of healthcare: elective treatments (including mental health services) and A&E care. By utilising data from the Waiting List Minimum Data Set (WLMDS) and Emergency Care Data Set (ECDS), and linking these datasets with Census, Hospital Episode Statistics (HES), HMRC, and DWP data, this analysis will provide valuable insights into how delays in treatment affect patients' pay, employment and benefit receipts in England.
Improving productivity in Local Planning Authorities through peer learning
- Organisation: University of Oxford
- Funding awarded (FY25-26): £55,735
- Government priority area(s):
- Mission: Growth
- Public sector reform: Local government and devolution
There are significant and stubborn backlogs in the English planning system. Peer learning has huge potential for improving productivity in Local Planning Authorities (LPAs). It is sector-led, tailored to local circumstances, and inexpensive, but there is limited evidence about its effectiveness. This study will use big data and quasi-experimental methods to evaluate the Planning Advisory Service's "Peer Challenge" programme. Using data on 30 million planning applications, and a state-of-the-art boundary regression discontinuity design, the study will test the short- and long-term effects of Peer Challenge on multiple dimensions of planning performance. Findings from this study will be compared with those for alternative improvement strategies which have already been evaluated, enabling evidence-based recommendations for optimising peer learning.
Drug Diversion for Safer Streets (DDSS): extending useful knowledge on police drug diversion schemes
- Organisation: University of Sheffield
- Funding awarded (FY25-26): £197,202
- Government priority area(s):
- Mission: Safer Streets
- Public sector reform: Police
- Public sector reform: Prevention
This project builds on the work of a previous evaluation funded by the Evaluation Accelerator Fund of police-led drug diversion (PDD) schemes. That previous project built large quantitative and qualitative datasets and analysed them over a one-year follow-up period. This project will add value to that work by: extending the analysis of effects (and cost consequences) of PDD on re-offending and entry to drug treatment over two years; expanding the analysis to the effects of other forms of diversion, including drug-testing on arrest (DToA) and cautions; and creating a menu of different types and components of PDD and other diversion schemes, for use in designing and implementing effective out-of-court-resolutions. This project will create useful knowledge for the reform of the police and the creation of safer streets.
Long-term effects of childcare entitlements for parents' labour supply
- Organisation: University College London
- Funding awarded (FY25-26): £191,230
- Government priority area(s):
- Mission: Growth
England is expanding its funded childcare entitlements, based partly on research suggesting that access to full-time childcare increases mothers' labour supply. Modelling the long-term growth impact of this reform, though, will depend on whether these labour supply responses are sustained - but there is little evidence available on this. This project will fill this gap. The project will use quasi-experimental methods to identify the shorter-term effects on parents' labour supply of access to government-funded 30-hour childcare for 3- and 4-year-olds and the longer-term effects of children starting school.
Impact of SEND provision in mainstream compared to specialist secondary schools on outcomes among adolescents with neurodisability: a target trial emulation using ECHILD
- Organisation: UCL (University College London) Great Ormond Street Institute of Child Health
- Funding awarded (FY25-26): £120,906
- Government priority area(s):
- Mission: Opportunity
- Public sector reform: SEND
This project will examine the impact of SEND provision in mainstream compared to specialist secondary schools on outcomes among adolescents with neurodisability. The study will use the ECHILD database, which links national health and education data for all pupils in England. It will estimate the causal effect of mainstream compared to specialist education settings for children with specific types of neurodisability, including Down syndrome, learning disability, autism and cerebral palsy. The project will also assess how children with neurodisability move between school settings during secondary school.
Longer-term and heterogenous impacts of the Health-Led Trials
- Organisation: University of Westminster
- Funding awarded (FY25-26): £178,315
- Government priority area(s):
- Mission: Growth
- Mission: Health
A previous DWP-funded study of Health-Led Trials tested the 12-month impacts of Individual Placement and Support on the employment, health and well-being of people whose health condition presented an obstacle to finding or remaining in employment. This new study expands the original evaluation in two ways: evaluating new outcomes (using updated HMRC/DWP data to estimate longer-term impacts on employment and earnings, as well as impacts on self-employment) and assessing impact heterogeneity (using machine learning techniques to explore how impacts vary across individuals, as well as examining the scope to target the programme to maximise returns).
Evaluation of a forensic marking intervention for domestic abuse - extension project
- Organisation: Durham University
- Funding awarded (FY25-26): £85,051
- Government priority area(s):
- Mission: Safer Streets
- Public sector reform: Police
- Public sector reform: Prevention
This is a follow-on evaluation of a forensic marking intervention for domestic abuse victims. The initial evaluation, funded by the Evaluation Accelerator Fund, won the 2025 Government Social Research Award. It found encouraging preliminary evidence of a reduction in repeat offending for victims provided with forensic marking products compared to a matched group, although this was not statistically significant. The intervention was cost effective and victims reported feeling safer, freer, and having more trust and confidence in police. This follow-up evaluation will extend the number of victims in the intervention group and focus on the deterrence effect and use in investigations. Findings will be incorporated in an implementation guide to support police forces.
Impact of local public transport schemes on economic growth
- Organisation: What Works Centre for Local Economic Growth
- Funding awarded (FY25-26): £69,054
- Government priority area(s):
- Mission: Growth
- Public sector reform: Transport
This project will evaluate the economic impacts of local public transport schemes. It will focus on understanding the effects of local schemes on the number of businesses, employment, turnover, productivity, and house prices, using granular data from 1997 to 2019 and quasi-experimental methods. The study will also provide evidence on whether impacts differ by type of transport schemes (trams, buses, and accessibility projects) and whether these schemes lead to displacement effects in local economies.
