Health economics. A bridge over troubled water.

In a recent issue of the European Journal of Public Health Barendregt and Bonneux 1 argue that 'the trouble with health economies', in particular economic evaluations of health care programmes, is that outcomes 'depend to a large extent on arbitrary decisions in study design', while in more theoretical discussions economists would 'display an unseemly rigidness'. Barendregt and Bonneux conclude that although standardisation of methodology is needed, it is not sufficient and health economists will 'have to be more aware of the value judgements underlying their methods, and be prepared to adjust their methods to reflect empirically measured preferences'. In this reply we want to address the issues raised by Barendregt and Bonneux, placing them in a broader context of economic literature and also correcting some inaccuracies in Barendregt and Bonneux's paper. Specifically, we will address the topics of which costs to include in an economic evaluation, how to value 'life years', discounting of costs and effects and health economics as an aid to policy makers. We will make clear that it is not health economics as such that is troublesome, but the task which health economists have taken on board. Assisting policy makers to make informed choices in health care, necessarily involves making value judgements that may be supported by empirical research, but cannot fully be based on empirical observations. Indeed, it is the task of normative economics (in contrast to descriptive or positive economics) to inform about the relative desirability of possible strategies in health policy. THE ECONOMIC ROOTS OF ECONOMIC EVALUATION Economic evaluation may be viewed as applied welfare economics. This important field in the economic discipline is concerned with assessing whether changes in society can be viewed as enhancing total welfare (or as economists would put it: utility). From most textbooks on welfare economics and some on economic evaluation it becomes immediately clear that normative value judgements are necessary in such analysis. Like Boadway and Bruce 3 state in their textbook on welfare economics: 'In order to make statements about the consequences for economic welfare of an event we must go beyond the study of positive economics, which is concerned with the effects of an event on objectively measurable economic variables, such as price and quantity. That is, the welfare economist wishes to determine the desirability of a particular policy — not in terms of his or her own values, but in terms of some explicidy stated …