An Instrumentalist Critique of "Cost-Utility Analysis"
Hildred, William, Beauvais, Fred, Journal of Economic Issues
The philosophy of instrumentalism developed by John Dewey establishes a political imperative and a criterion of truth. The political requirement seeks integrity and vitality of democratic decision making via the widest possible participation by well informed citizens in solving public problems. The criterion of truth, warranted knowledge, is satisfied when participation gives vent to all viewpoints and insights. In this process, reasoned arguments produce public validation of evidence, leading to agreement for action. The role of experts is to invigorate informed participation by using their expertise to clarify complex matters. That role is perverted when it supplants judgments of citizens about desirable public arrangements or bolsters the positions of unrepresentative elites.(1)
Unfortunately, some experts seem unaware that control over decision making in their area of expertise is not legitimately theirs, even though they may hold it simply by default. Often they frustrate the evolution of democratic policy development by erecting barriers to participation. A current deplorable manifestation of the rule of the experts is the use of Cost-Utility Analysis (CUA) of medical care, especially as presented by persons whose medical expertise is not clearly augmented by exposure to economic theory. In articles published in medical journals (and apparently not very visible to economists), they argue that their work has the potential to improve medical practice and resource allocation by embedding the values and preferences of patients in medical decisions. In fact, this practice is contrary to the instrumentalist vision. It obfuscates, rather than clarifies, that which should be open to scrutiny and debate.
The purpose of this paper is to advance the instrumentalist purpose of improving public dialogue by bringing to the attention of an audience wider than that of medical journals some of the serious deficiencies of CUA. Broader awareness of the intrinsic defects of CUA may facilitate more effective challenges to its misuse in the political process. A forthcoming companion paper extends this discussion to the abuse of more familiar Cost-Effectiveness Analysis (CEA), based on review of several articles recently published in the same medical literature. Briefly, the difference between CUA and CEA is that CEA makes cost comparisons among courses of action that are deemed to achieve the same outcome (equally "effective"), while CUA goes further in attaching weights to the outcomes that incorporate measured utility accruing to those who benefit from the outcomes.
Desire for a Rational Allocation Process
Efficiency in health care delivery is a major policy issue, but we have little ability to identify appropriate inputs and outputs of medical procedures. Some investigators erroneously measure inputs with the dollar value of charges for medical services. For example, in a hospital environment, a "charge" to a patient or insurance carrier includes relatively arbitrary, and often large, overhead allocations determined by one or another accounting algorithm. The procedure distributes costs from centers that do not deal directly with patients to centers that can assess fees for direct patient contact [Finkler 1982]. In addition, this process achieves cost shifting, whereby redress is obtained for treatment given uninsured or otherwise non-paying patients. Thus, these data bear no resemblance to those required for standard microeconomic analysis.
Efforts to relate benefits to costs are also complicated because outcomes are often indeterminate and complex. For example, consider the ramifications that follow from the mathematics when states of health cannot be characterized by a single trait.
. . . four different levels of "intensity" for each of four different "dimensions" of health would generate 256 different combinations . . . [and] over 30,000 different pair-wise comparisons of states of health! …