Functional form and the non-marginal effects of NHS expenditure
Functional form and the non-marginal effects of NHS expenditure
Theme 5: Empirical work on the marginal productivity of public expenditure in health and care
The focus of most of the work on the marginal productivity of NHS expenditure has been to estimate the marginal effects of expenditure on average across the NHS. This provides the basis of calculating a cost-per-QALY of NHS expenditure at the margin, which provides a measure of opportunity cost relevant to many national policy decisions. However, there is also a need to go beyond estimating the ‘average’ marginal effect, (i) when the scale of the budget impact of a policy is significant and not marginal; and (ii) when considering reallocations of resources between local geographies, e.g., transitioning areas from actual allocations towards target allocations based on measures of health care need.
The aim of this project is to better understand the non-marginal effects of changes in different types of NHS of expenditure and to provide estimates that are more relevant to local geographies. These estimates can better inform proposed reallocations of resources between local areas and types of NHS expenditure. This project will also fill an important empirical gap when estimating the marginal value of public expenditure on health and social care.
James Lomas, Francesco Longo, Andrea Salas Ortiz, Karl Claxton
James Lomas james.lomas@york.ac.uk
Background:
Previous work has shown that the NHS spending an extra £15,000 will generate health benefits equal to one additional year of life in full health. However, as more and more is spent, it is thought that the health benefits generated may start to fall with each additional £1 spent. For example, £15,000,000 may be expected to generate fewer than 1,000 additional years of life in full health.
Aims and objectives:
This project will estimate the health benefits of changes in NHS expenditure. It will extend current work by estimating the health effects at different levels of NHS expenditure and in areas with different levels of mortality.
Methods:
To estimate the causal effect of NHS spending on health benefits, using data on local authorities (LAs) in England, we will use an econometric model that attempts to control for other observed differences between LAs that determine health outcomes, such as measures of need and socioeconomic variables. Variables that are unrelated to health outcomes but are good predictors of expenditure will be used to control for any remaining unobserved differences to identify the causal effect of NHS expenditure. The analysis will take account of how the size of health benefit varies according to the level of NHS spending and the mortality rate of the local authority.
Policy relevance & dissemination:
The findings of this project will be useful for considerations of benchmarks of value for healthcare interventions that carry substantial costs to the NHS. In addition, it will show the benefits of a range of potential increases in overall NHS expenditure and the benefits of allocating more resources to those areas with higher mortality. We will produce a report for policy purposes as well as a paper for publication in a peer-reviewed academic journal.