But how much confidence do we have in the ‘market prices’ for medical care? There are huge distortions in this market including monopoly power, asymmetricinformation, non-marginal prices and the role of insurance companies.There are 2 broad approaches to COI. Prevalence-based estimates base cost on all people who have a condition at a given time. Incidence-based estimates base cost on new cases diagnosed – essentially, they provide the total cost over a lifetime of getting a given health condition. This is usually the kind of value we want if we are estimating the benefits of preventing illness.For short-term illnesses of less than one year these two approaches will provide the same estimate.Slide 12Cost of illness can be considered as the “Potential benefits of a health care intervention” in terms of discounted lifetime expenditures due to the illness being avoided.COI usually makes a distinction between direct costs – health care-related costs –and indirect costs associated with the loss of productivity due to the illness. Indirect costs for productive labor are generally valued at the wage rate.Pure application of cost-of-illness will also value productivity loss from mortality at the wage rate – the foregone earnings or “human capital” method that predates the VSL for cost-benefit analysis. But for BCA COI is usually reserved for cases where we don’t have WTP so the VSL is retained for the value of reduced mortality risks.We should note that there is a different approach to cost of illness that rarely makes an appearance in benefit-cost analysis. This “friction-cost” methods adopts the employer perspective and only counts productivity losses when thereis no substitute worker to step in. Economists practicing BCA instead adopt a social perspective that counts as productivity loss those workers who are unable
to be employed because of illness.Slide 13Here’s an example of what is included or not in a typical cost-of-illness measure. This is from an EPA rule that reduces viruses in drinking water.Any reductions in mortality risks we would value with a value of statistical life shown here in the furthest pathway to the right. For these non-fatal effects there are a number of dimensions to consider. These are all acute effects. That is, they are short-term health effects – a stomach bug – rather than long term or chronic effects.Lost work time is reasonably straightforward and typically valued at the wage rate, since, in principle, your wage is equal to the marginal product of your labor.Lost work productivity is typically a separate issue that is not always included. This is for the folks who come to work sick and just can’t do the job as well as they can when they are healthy. These productivity losses, if they can be measured, are a cost-of-illness. Sometimes this effect is known as “presenteeism.”The third box is challenge: lost leisure time. How do we value losses to leisure?