The economic evaluation strand of GoWell views the housing and regeneration interventions we are studying as investments in health and wellbeing.
It aims to assess whether the financial investment will provide ‘value for money’ in achieving policy aims and to inform future policies on housing and regeneration in Glasgow and elsewhere, in terms of this.
The study is designed as an economic evaluation alongside a natural experiment. All households will receive interventions but may do so many years apart. Three main evaluation arms have been identified. The two intervention arms consist of demolition/relocation; and housing improvement. The non-randomised control arm consists of households not yet in receipt of these interventions.
A longitudinal cohort (approximately 1,100 households) is nested within four cross-sectional surveys. This will be linked with routine datasets. Outcomes include both self-reported and hard outcomes, including housing amenity, empowerment, behaviours, health, employment, crime and education. Intervention costs will be identified alongside potential cost savings, including: reduction in hospitalisations, crime, and welfare claims.
A difference-in-difference approach will attribute changes in outcomes to interventions, and OLS used with baseline adjustment for possible confounders. Dose-response relationships between levels of investment and outcomes will be tested. To project longer term outcomes an interrupted time series (focussed on hospitalisations, crime, and education) will be conducted and extrapolated over a period of 60 years.
Methods of economic evaluation will then value all major impacts of the interventions, and assess value for money. Given multiple policy aims, an ‘integrated societal approach’ to economic evaluation will be taken by combining different economic evaluation methods in a coherent approach intended to tailor outputs to different policy aims.
First, a cost consequence analysis will list all costs and outcomes. Second, a cost effectiveness analyses will report on cost per housing amenity (e.g. new house, housing improvement) and cost per Quality Adjusted Life Year (QALY). Third, a stated preference discrete choice experiment (SPDCE) to elicit willingness to pay (WTP) for observed outcomes will estimate the overall impact of urban regeneration on social welfare. A sensitivity analysis will also be conducted.
The full economic analysis will be conducted once the final wave of survey data is available in 2015/16.