Facts, values, and the burden of proof.

• Author: Lewin, Peter

The very terms "efficient" and "inefficient" are terms of normative and not positive economics, so much the better: immense confusion has been sown by the pretense that we can pronounce "scientifically" on matters of "efficiency" without committing ourselves to any value judgments.

--Mark Blaug, The Methodology of Economics (2d ed.)

I examine here the connection between facts, values, and the burden of proof, and I suggest a relationship that is always present, though seldom acknowledged. We like to think of ourselves as conducting value-free scientific inquiry, yet values inevitably intrude. This inevitable intrusion does not necessarily suggest the abandonment of the attempt to keep facts and values separate, but it does suggest that we ought to make plain our prior moral commitments as an aid to those evaluating our findings. (1) This revelation is especially desirable in the design of experimental frameworks used to interpret the policy significance of the "evidence." Where one places the burden of proof in such a design is closely related to such prior moral commitments.

I begin by reviewing the fact-value divide and the concepts of efficiency that usually accompany it. I consider some well-known problems with the concept of efficiency and show that these problems necessarily imply the intrusion of values into this apparently "scientific" idea. I speculate on how this intrusion may be handled appropriately by shifting the focus from individual policies to policy regimes. I then turn to an examination of the role of placing the burden of proof in appeals to "evidence" and show what this placement implies about the fact-value divide. I conclude that values necessarily intrude here as well, perhaps more subtly and dangerously. In fact, because of the nature of "evidence," there is a remarkable connection between the problem of value neutrality at these two contextual levels. I provide an extended example by looking at the debate between Paul David, on the one hand, and Stan Liebowitz and Stephen Margolis, on the other, with regard to determining the efficiency of alternative standards. This example illustrates the general situation and forms the basis for nay conclusion.

Efficiency and All That

Positive and Normative Economics

Students of economics are told from the start that good scientists are careful to separate facts and values. Positive economics deals with "facts" about how the world works. Normative economics deals with what we ought to do with knowledge of those facts. There is a widely shared conviction that good scientists, as such, ought to keep their moral sensibilities out of the analysis. (2) In any case, we know from David Hume that one cannot get an "ought" exclusively from an "is," (3) so it behooves the good scientist to be forthcoming about the values introduced in the derivation of any policy conclusion.

This fact-value divide has received much critical scrutiny, particularly in the social sciences. The long development of welfare economics can be understood as an ongoing attempt to derive normative propositions from certain minimum accepted norms, such as agreement that the valuations (utilities) that are to count are those of the economic agents who will be affected by the policy in question. In some ways, welfare economics, with its use of the concept of efficiency, is made to appear every bit as "value free" as positive economics. As such, it has been the subject of much debate. (4) A brief overview follows.

Problems with Efficiency: What Does Efficiency Mean?

Economists have searched long and hard for a concept of efficiency that is "objective" or value free. (5) An economic outcome consists of an array of goods and services, prices, and states of being of different individuals that are in themselves incommensurable. To pronounce a particular outcome more or less efficient than another, one must overcome this "apples and oranges" problem. Commonly, one resorts to the valuations that affected individuals place on the outcomes. This approach obviously involves the presumption that individual preferences ought to be what counts when deciding efficiency issues. This procedure, in itself, however, merely defines efficiency. No value judgment is involved (beyond that of supporting a definition) unless one says something like "efficiency is good" or "a more efficient outcome is a preferred outcome." When we take this step, as we often do in policy discussions, we are saying that we believe individual preferences ought to count in deciding which economic outcomes are preferable. This approach may strike readers as eminently reasonable. If what is "efficient" is defined as what the "people prefer," how can we oppose the definition? Is it not the quintessence of "economic democracy"?

As is well known, however, numerous practical difficulties arise in deciding what "people prefer" and, indeed, what this statement means. When a change is contemplated from which all of the involved individuals would clearly gain (that is, they can be confidently said to prefer that the change be made), then there is little ambiguity, and we have a Pareto improvement. The most common difficulty arises, however, when some individuals gain and others lose. In such "mixed" situations, we have to resort to so-called compensation tests to judge whether the gains outweigh the losses. If we take this leap, we arc saying in effect that when matters of efficiency arc decided, the distribution of gains between individuals is not relevant or is something about which the economist should remain agnostic. A standard defense is that the distribution might be relevant, but is a separate issue--we ought to make the pie as large as possible before we consider how it ought to be divided up. This approach involves an additional value judgment, and clearly it is a much less plausible and easily defensible position than the one that arises in the simpler situation where no one suffers losses. To be sure, this position still represents a kind of economic democracy, claiming something like "more people prefer this outcome" or "the intensity of the preferences in favor outweighs the intensity of the preferences against" and that "losers" ought not be able to "hold out" against "winners" in the face of changes that are manifestly beneficial. (6)

This efficiency standard is widely accepted in economic discussions and has penetrated deeply into the policy and legal environments. The situation is complicated because the word efficiency has a very strong colloquial connotation, and its meaning in economic policy discussions is often confused with its meaning in the natural sciences, where inputs and outputs are much more easily identified and evaluated and no compensation criteria are necessary. The widely accepted efficiency standard gives economic policy discussions a spurious aura of "scientific" objectivity.

Difficulties in Implementing Efficiency Standards: The Importance of Dynamics

Economists encourage this impression of "scientific" objectivity even though they are aware of the insurmountable obstacles to arriving at unambiguous decisions about which changes are efficient and which are not. These obstacles include the well-known impossibility of discerning individuals' preferences, which requires the use of hypothetical market valuations and the necessity of positing unknowable counterfactuals on the basis of unknowable (even unimaginable) futures. Perhaps an even more important problem with this traditional efficiency standard has to do not so much with its theoretical conception per se as with how it is traditionally used in economics--namely, in the context of the static model of resource allocation. In this context, it has encouraged the kind of attacks, made in the name of efficiency, that we often witness on the unfettered emergence and development of products, organizations, and standards.

In a static context, in which the values of all potential resource uses are known (either with certainty or probabilistically), the technology is unchanging, the set of products is fixed, and no external effects or elements of monopoly exist, it is well known that a "competitive solution" is also Pareto optimal and therefore efficient in the sense discussed earlier. This idealized situation of neoclassical "perfect competition" has unrealistically and unreasonably served as a standard of comparison for real-world situations. For example, in the context of network effects and standards, economists have thought it relevant and meaningful to argue that the presence of such effects suggests that private markets may produce an "inefficient" result. Such arguments, however, rest on an...