4. Measurement and Welfare Economic Theory
4.4. Some Institutionally Oriented Viewpoints
Considering the complexity of the social decision-making process and the number of problems related to individual preferences and their revelation, it is not surprising that monetary valuation has been seen in many occasions as an inadequate approach to provide valid information for the purposes of policy-making Some social scientists even challenge the rationality of the use of individual preferences as the basis for the assessment of social welfare change.
They either strongly oppose the use of people's preferences as a legitimate starting point of social decision-making or favor institutional structures that eliminate the necessity of making interpersonal welfare comparisons by a third party.
If preferences are regarded as unreliable and/or invalid, a possible solution is to reject the idea that preferences (whatever they are like) should count in the social decision-making which deals with environmental amenities and services possessing public good characteristics. For instance, Sagoff (1994) promotes this rather non-economic view. He argues that choices should count instead of preferences, and that preference-satisfaction does not necessarily lead to im-proved well-being of the society as a whole. According to him, the core of the problem is that economists cannot make a difference between preferences as psychological phenomena and preferences as logical constructs derived from certain behavioral assumptions.
Empirical psychologists explaining behavior in terms of its causal conditions do not give preferences a prominent position. Some of them regard preferences as private mental states that are not observable and thus not testable in the scientific sense. If an attempt is made to explain a choice of an individual in terms of preferences, any explanation will inevitably have an ad hoc nature.
There can be a dozen of potential explanations, some of them more likely than the others, but it is not possible to detect which of them, if any, is the correct one. Thus, preferences as psychological phenomena represent unidentified behavioral motives behind actual choices.
The social choice theory, in turn, presents preferences as formal constructs derived logically from stipulated descriptions of individual behavior. The behavioral model is prescribed, and the social choice theory does not speculate on the psychic, environmental, genetic, or other causes of behavior. Usually the applied behavioral concept is rationality. When this approach is broadened to the societal level, social choice theorists begin with "given" or "assumed"
descriptions of altemative social states and then construct preference schedules as rankings among stipulated behavioral descriptions. Thus, the social choice theory is concemed with the logical, not the psychological, properties of prefer-ence orderings. Therefore, the relation between preferprefer-ence and choice in this context becomes logical, not psychological, or causal.
Now, Sagoff (1994) argues that welfare economists follow social choice theorists by constructing preferences logically from stipulated descriptions of behavior. Yet, welfare economists assert that these logical constructs function as the psychological causes of this behavior. In this connection, they make themselves guilty of the use of ad hoc explanations by supposing that theoreti-cal terms logitheoreti-cally constructed from ways of describing behavior simultane-ously cause that behavior. In other words, welfare economists misleadingly combine an epistemological program of clarifying the logical structure of pref-erences and a psychological program of identifying the causal mechanisms behind them. Consequently, there will be a mix-up of epistemological and psychological elements, which leads welfare economists to move casually be-tween the logical analysis of rationality and psychological claims about behavioral motives.
In practical applications welfare effects of preference satisfaction are ex-pressed through money metric measures of welfare change, i.e. willingness to pay. If social welfare is defined as a function of willingness to pay for thing rather than a function of "true welfare" originating from having some-thing, the allocation of natural resources takes place in favor of those who can afford to pay the most for the use. The pivotal question is whether the WTP approach really measures changes in personal welfare associated with the con-sumption of different goods and services. The problem is that welfare econom-ics does not offer any other method for measuring changes in personal well-being. The ramification of this is that the relation between willingness to pay and welfare is a stipulative or definitional rather than a contingent one. This is why the proposition that the more people will pay for things, the more welfare they will derive from having them, must be treated as a tautological definition and not as a fact. Therefore, the claim that preference satisfaction promotes well-being in essence merely assumes that the truth is what actually should be proved to be the truth (Sagoff 1994).
Conceming the utilization of CVM results in the social decision-making context, Sagoff's (1994) claims are extremely discouraging. If stated prefer-
ences are not proper indicators of actual future choices, the CVM is hardly applicable. In addition, if money metric welfare measures cannot be applied when changes in social well-being are assessed and evaluated, the Kaldor-Hicks compensation criterion is hardly an appropriate criterion to be used in the ranking of alternative policy proposals. However, SagofFs (ibid.) purpose is not to rebut the concept of consumer sovereignty. He considers that, although markets and democratic processes may fail to produce an optimal resource allocation, they should be regarded as the most appropriate settings for individu-als to make choices and to pursue their visions of the personal and social good.
The moral properties related to choices, like accountability, responsibility, com-mitment, and consent, are not dependent upon preferences that can be assumed to be behind different choices. Thus, as long as social decision-making is based on information gathered from our actual choices and not from our preferences, we need not be worried about decisions related to the allocation of natural and environmental resources.
When the CVM is used, the idea of the market failure is implicitly accepted.
If the market functioned properly, there would not be any necessity to search for people' s preferences beyond actual market transactions. In essence, the applica-tion of the CVM is based on thinking that lies behind the prevailing centralized approach to environmental problems. Because of the externality nature of many environmental effects, the society is forced to take policy actions to internalize benefits and costs produced by these effects. Independent of the nature of the chosen policy instrument, a Pigouvian tax, a command-and-control regulation, or informative guidance, the point of the centralized approach is that the society identifies the amount of harm or benefit and employs some policy instrument in order to achieve an optimum where social welfare is maximized as well as possible. The CVM is an excellent tool for this kind of approach. It can be used for the estimation of nonmarket benefits created by public good or externality type environmental commodities and impacts.
However, ali economists do not feel comfortable with the idea of inevitable market failure. In his famous article Coase (1960) sketched an approach that emphasized the importance of property rights and rejected intervention by the government in favor of market bargaining in order to achieve the socially optimal level of internalization of externalities. Given the existence of an appro-priate property rights system (guaranteed ownership of resources via enforce-ment by the society) and certain other assumptions, he argued that a harm producer and a harm sufferer should be left in an unregulated situation. A bargaining process would then develop on an automatic basis. The so-called Coase theorems states that if transaction costs are zero, the initial assignment of
5 The Coase theorem is often abbreviated into the form that property rights do not matter, but this is not an appropriate interpretation as Schmid (1987, p. 219) rigorously points out.
property rights will not affect the efficiency with which resources are allocated (Posner 1993).
The Coase theorem has been questioned in many occasions in the economic literature but it has succeeded in maintaining most of its relevance. The major problem with the applicability of the Coase theorem is that in real-life situations the role of transaction costs becomes significant. It is quite conceivable that the trade between two parties always involves some kind of transaction costs.
Positive transaction costs would reduce the effective value of any offer or payment. The amount paid would he equal to the amount received minus the transaction costs. Consequently, if the transaction costs were larger than the potential gains from the trade, no trade would take place.
It is difficult to say to what extent the Coasian approach could really replace methods like the CVM. Pearce and Turner (1990) have concluded that the Coase theorem has been important in forcing advocates of public intervention to define their terms and justify their case more carefully than •they might have done otherwise. Randall (1987) emphasizes that the Coasian analysis has con-tributed to the understanding of extemalities and market fimctions. In essence, the theorem has successfully suggested that market phenomena may be more pervasive than has been generally recognized, and that the market behavior is likely to occur whenever gains from trade exist, even in settings customarily considered to be outside the market. Moreover, as Posner (1993) notes, the Coasian approach has vividly promoted the point that intelligent public inter-vention requires more information than govemments are usually likely to have.
In the CVM context the explicit or even implicit assignment of property rights matters. It may affect response behavior. When a respondent is asked, for instance, how much he is willing to pay for a conversion from conventional agriculture to pro-environmental farming, the scenario indirectly implicates that the farmers own environmental benefits related to the rural environment. It is not surprising that some respondents oppose this kind of definition, because property rights related to the rural environment are far from being defined in a nonattenuated way. The result can he a serious motive conflict. On the one hand, the respondent wants to have environmental benefits produced by pro-environmental farming. On the other hand, he is not willing to pay for them because, according to what he considers "fair" property rights, he should he entitled to them anyway. Of course, it is difficult to say if this kind of respond-ent would respond-enter any bargaining process either.
Property rights and other social institutions seem to matter more than is usually recogmized. However, it is not realistic to assume that institutional innovations and developments could remarkably broaden the applicability of the market mechanism when environmental commodities and services are con-cemed. Problems related to transaction costs will persist. This guarantees that
monetary valuation methods of nonmarket beneflts like the CVM will also be needed to assist social-decision making in the future.