Examining credibility of current evidence for welfare as a public health measure: pathways, causation and cost-benefit of further research

Research question: What is the value of improved confidence in evidence on cash-based welfare as a prospective public health policy? Sub-questions: What is the value of a) fully eliminating uncertainty (Expected Value of Perfect Information), b) eliminating uncertainty around targeted relationships (Expected Value of Partially Perfect Information), and c) the potential benefit to a specific research design (Expected Value of Sample Information)? Background: Growing evidence indicates that welfare reform can reduce health inequalities by mitigating ‘upstream’ socioeconomic determinants. Despite government endorsement of upstream interventions in Levelling Up and the NHS Prevention Programme, there remains uncertainty around which interventions effectively reduce inequalities, in part because policymakers lack sufficient evidence and debate interpretation of causality within existing studies Aim: To enable policymakers to make informed decisions on welfare as a public health policy Objectives 1 use an Expert Panel and PPI Lived Experience Panel to interrogate causal relationships within the welfare health impact (WHI) logic model 2 use new evidence from observational and experimental studies on causal relationships within the logic model to increase confidence in modelling 3 conduct and present new microsimulation of health, wellbeing, health economic and economic outcomes via updated logic model 4 identify value of further research to establish certainty 5 identify costs of not conducting further research or implementing reforms for which there is evidence of beneficial impact Methods: We will work through a diverse, interdisciplinary Expert Panel to establish consensus on evidence required for policymakers to decide whether to implement a welfare policy as a public health measure without a Randomised Controlled Trial or to implement a trial. We will: 1 use evidence from new studies and statistical methods to estimate how much, and with what confidence, each part of our WHI logic model (e.g. secure income) affects the next (e.g. stress) 2 employ Value of Information (VoI) analyses to estimate costs/benefits of gathering more evidence compared with immediately implementing policies 3 use marginal analysis to understand relative value of interventions at different life stages 4 deploy scenario analysis of four reforms endorsed by policymakers as well as incremental reforms to establish positive and negative health impacts 5 update microsimulation model to provide evidence on health, health economic and economic impacts 6 use conjoint surveys to test how public views of policies change according to availability of better evidence on impact 7 incorporate findings into the TriplePC model to estimate instantaneously policy impacts and public preferences 8 use dimensional analysis to identify a single measure for each part of the WHI logic model for use in studies Timelines 01/25: Commencement 02/25-03/26: Reviews of evidence 02/25-10/26: Expert Panel and PPI meetings 04-10/26: model and microsimulation update 05-12/26: Journal articles, updated TriplePC and other outputs 12/26: Final report and conference Impact: By engaging with our network of policymakers, stakeholders and funders, we anticipate impact on a) policy design and implementation and b) commissioned research by breaking a position of equipoise and equivocation among academics and policymakers on the economic interventions best suited to address determinants of health