Background There is a healthy public policy agenda investigating the health impacts of improving living conditions. However, there are few economic evaluations, to date, assessing value for money. We conducted the first cost-effectiveness analysis of a nationwide intervention transferring social and private tenants to new-build social housing, in Scotland.

Methods A quasi-experimental prospective study was undertaken involving 205 intervention households and 246 comparison households, over 2 years. A cost-utility analysis assessed the average cost per change in health utility (a single score summarising overall health-related quality of life), generated via the SF-6D algorithm. Construction costs for new builds were included. Analysis was conducted for all households, and by family, adult and elderly households; with estimates adjusted for baseline confounders. Outcomes were annuitised and discounted at 3.5%.

Results The average discounted cost was £18 708 per household, at a national programme cost of £28.4 million. The average change in health utility scores in the intervention group attributable to the intervention were +0.001 for all households, +0.001 for family households, −0.04 for adult households and −0.03 for elderly households. All estimates were statistically insignificant.

Conclusions At face value, the interventions were not value for money in health terms. However, because the policy rationale was the amenity provision of housing for disadvantaged groups, impacts extend beyond health and may be fully realised over the long term. Before making general value-for-money inferences, economic evaluation should attempt to estimate the full social value of interventions, model long-term impacts and explicitly incorporate equity considerations.