Applied Welfare Economics with Bounded Rationality: Public Policies toward Remote Sensing

Article excerpt


Earlier experimental evidence indicates that some decision makers under-invest in information when it is costly. This insight is applied to the cost-benefit analysis of the provision of data obtained by satellites. In fields like agriculture, remotely sensed information of great precision can be provided to additional farmers at minimal extra costs if a service for farmers with similar information needs is set up. Here, the tendency to under-invest in information asks for creative solutions by the private sector: bankers may work together with rational producers in setting up the information service and promoting it. However, if producers' financial resources are insufficient public intervention in the form of a free information service may be necessary. (JEL D61, H42)


Recent years have seen substantial and diverse contributions dealing with the analysis of forms, effects, and models of bounded rationality in economics. Bounded rationality (as opposed to unbounded rationality) is the notion that at least some economic agents make sub-optimal decisions because of their limited cognitive abilities. Analysts have been keenly interested in how bounded rationality affects market functioning and policy effectiveness [see Conlisk, 1996; Bomfim and Diebold, 1997; Selten, 1998]. However, after a number of early contributions by Simon [reprinted in Simon, 1983] little has been written on the question of how bounded rationality influences the field of applied welfare economics. Consumer theory is the notable exception [see Roth, 1998]. This paper attempts an advance by taking up a classic issue of welfare economics: the question of whether information should be provided as a public service or produced privately. (1)

The starting point is the insight that with bounded rationality the benefits of information to producers differ substantially depending on how information is made available to users. Experimental evidence indicates that many individuals systematically under-invest in information that--if obtained freely--they are able to use for their own benefit. This tendency has to be taken into account when we address the questions relating to who should provide information and how an information service could be financed. A welfare analysis proceeding on the assumption of perfect rationality would go astray and leave important issues in the provision of information services undetected. In contrast, the present welfare analysis takes bounded rationality into account and, thereby, helps to identify critical issues for producers, financiers, and governments. The article is structured as follows: the next section provides background insights derived from behavioral decision research; then, remote sensing, that is, the provision of earth surface data based on satellite observation is described. So far the economics profession has given very limited attention to this subject [see Macauley and Toman, 1991 for an exception]. This paper focuses on the provision of remote sensing information in the field of agriculture. Next, the paper offers a cost-benefit analysis indicating when remotely sensed data should be made available to agricultural producers. The following section asks how welfare losses due to bounded rationality can be avoided and finally, the last section concludes.

Boundedly Rational Acquisition of Information

Research on human decision-making under uncertainty has revealed many cases where acquisition of information is individually and socially suboptimal. Both the cases of over-acquisition and under-acquisition of information have been documented in the literature [see Connolly and Thorn, 1987; Newell et al., 2004]. Much of this research has analyzed so-called stopping rules that describe the behavior of people who invest either more or less of their time before reaching a decision. Busemeyer and Rapoport [1988] show that it is cognitively demanding for people to appreciate the value of extra information when they have to anticipate future decisions that are affected by this information. …