Cold Related Deaths and the Effect of Nudging the Elderly: Evidence from the Longitudinal Studies

The existing official methodology to measure cold-related deaths is crude - just a comparison of average mortality rates across seasons. Moreover extremely strong inferences have been made about the impact of cold through simple interpolation. Previous research has shown that although mortality does increase as it gets colder, temperature only explains a small amount of the variance in winter mortality, and high levels of excess winter mortality can occur during relatively mild winters. US evidence suggests that latitude plays a role - places where it is usually cold have few cold related deaths. There is a strong suggestion, although never formally tested, that the relationship is highly nonlinear and that unexpected cold is more dangerous than expected cold weather. Thus, the relationship between temperature and mortality is complex. In particular, the dynamic effects and nonlinearities in the relationship requires serious attention.

We propose to provide a detailed investigation of the impact of cold on mortality - taking account of dynamics, nonlinearities, and socio-economic characteristics. In addition we aim to investigate how winter fuel payments (WFP) affects the relationship. The idea of WFP is to provide cash to keep old people warm when it's cold - so as to keep them well. So, if WFP "works", it does so by "nudging" household spending on fuel, so raising heat levels, and reducing the adverse effects of cold on health - by more than an equivalent rise in the state pension . Evidence on the extent to which the labeling of transfers affects how they are spent is mixed but there has been much talk of policy that can "nudge" individual decision making . WFP provides an opportunity to demonstrate the effectiveness of one such idea. Indeed, it provides an unusually "clean" test because of the nature of its implementation - in particular, it is not means-tested and the take-up rate is close to 100% so there is no selection issue to take account of. Moreover, the UK provides an excellent environment to test the proposition that it affects the relationship between mortality and weather separately from a general mortality effect because it is a long county and has wide variation in weather patterns across the country.

WFP is a tax-free payment made to individuals aged 60 or over living in GB to help them with their heating costs over the winter months. Older people are targeted since they are particularly vulnerable to the effects of cold weather during the winter months and are more likely to be on fixed incomes. The basic Winter Fuel payment is made to more than 12 million people in GB (of which 1.1m are in Scotland). The total cost was over £2.7 billion last year. WFP amounts to about 10% of an average retired household total annual fuel bill. The history of WFP provides a number of opportunities to investigate the extent to which WFP has had a causal impact on the seasonal and weather patterns in mortality.

The project has the potential to contribute significantly to government social policy. We intend to use the analysis to facilitate the simulation of how deaths would be affected by restricting eligibility to the lowest income elderly (passported by the Pension Credit, for example), and restricting eligibility further by age (say, just to those 80+). We aim to complement our survival analysis with QALY estimates to allow policymakers make value for money comparisons of alternative policy ideas.

We could also use the analysis to simulate the impact of cold spells to measure the extent to which residents in each local authority might be affected - which would enable local authorities to mobilise appropriate resources. This would improve the effectiveness of some public services. We anticipate embedding our survival modeling into a website that would provide traffic light mortality risk warnings based on Met Office forecast feeds that LAs and NGOs (and commercial organisations that provide services to the elderly) could use such a monitoring tool. Even relatives and friends may wish to consult such a tool.

The work would be presented at the annual DWP "WPEG" conference (if it still exists in 2013/14) where policymakers and academic researchers meet, at a relevant Age UK conference (formerly Age Concern), a Scottish event, and at an event at Lancaster's "The Work Foundation" in central London where NGOs and policymakers would both be invited. Finally, our work will promote further use of LS/SLS/NILS for public health research by providing the socio-economic background to the geography of mortality (and fertility).