The economics of 'more research is needed'

Results from epidemiology and other health research affect millions of life-years and billions of dollars, and the research directly consumes millions of dollars. Yet we do little to assess the value of research projects for future policy, even amid the ubiquitous assertions that 'more research is necessary' on a given topic. This methodological proposal outlines the arguments for why and how ex ante assessments can inform us about the value of a particular piece of further research on a topic.

Economics and decision theory concepts-cost-benefit analysis and probability-weighted predictions of outcomes-allow us to calculate the payoff from applied health research based on resulting decisions. Starting with our probability distribution for the parameters of interest, a Monte Carlo simulation generates the distribution of outcomes from a particular new study. Each true value and outcome are associated with a policy decision, and improved decisions are valued to give us the study's contribution as applied research.

The analysis demonstrates how to calculate the expected value of further research, for a simplified case, and assess whether it is really warranted. Perhaps more important, it points out what the measure of the value of a further study ought to be.

It is quite possible to improve our technology for assessing the value of particular pieces of further research on a topic. However, this will only happen if the need and possibility are recognized by methodologists and applied researchers.