Empirical ethics, moral philosophy, and the democracy problem

A central issue for the development of summary measures of population health (SMPH), as well as for their use in the prioritization of health care resources, is the source of the value and ethical assumptions employed. Some, not surprisingly they are often social scientists, hold that any ethical premises should be empirically grounded, derived through empirical research on the actual values and ethical beliefs of real persons. Others, and again not surprisingly they are often philosophers and other norma-tive theorists, hold that empirical research is irrelevant to establishing ethical and value judgements because they are normative not empirical claims and so require normative argument for their support. Neither of these polar positions is in my view correct. Empirical research has been used to supply ethical and value premises in several different contexts. First, in developing weights for different states of disability. Once one delineates different domains and levels of function, as in a measure like the Health Utilities Index (Torrance et al. 1992), different relative values must be assigned to those different states of disability in order to make comparative judgments of the impact on health of different states of disability and to arrive at overall assessments of health where there are multiple functional limitations. Some of the early work derived disability weights on a scale from 1 representing perfect health and 0 representing death by simply asking individuals, using standard gambles, time trade-offs, or interval scales, their preferences for life with a particular disability. Second, for health care resource priorities and trade-offs: Erik Nord and many others have used the person trade-off method to enable people to express in a quantitative form specific normative concerns for fairness (Nord 1999). Since most people care both about aggregate health benefits and how they are distributed, this early work simply asked for preferences about trade-offs, such as between producing the greatest aggregate health benefit and treating the sickest, to assign quantitative weight to these distinct moral concerns. Finally, empirical research has been done 654 Summary Measures of Population Health on some explicit assumptions of SMPH such as whether people use discount rates or age-weight health benefits. A variety of methodological concerns can be and were raised about these early empirical efforts, but ethical concerns were raised as well. One was simply the general scepticism of philosophers noted above about whether any ethical conclusions could possibly follow from this empirical research. But another …