Incorporating Endogenous Preferences in Cost-Benefit Analysis

The conventional assumption made by cost-benefit analysts is that individual preference structures are not altered by any of the policies that are under consideration. This simplifying "exogenous preferences" assumption is not always satisfied, however, and in some instances the preference structures of a significant proportion of the people who are impacted by a policy are "endogenous" in that they are also altered by that policy. Under those endogenous preferences circumstances an important question is presented as to whether the willingness to pay-based valuations of the impacts of the policies should be calculated with regard to the pre-policy implementation preference structures, or instead with regard to the different post-policy implementation preference structures, including any transitional preference structures that may exist for a period of time, or perhaps with regard to some combination of the above.Several prominent scholars have previously addressed aspects of this inquiry, including most importantly Gary Becker, Cass Sunstein, Kenneth Dau-Schmidt, and Samuel Bowles, but the endogenous preferences valuation question has not yet been definitively resolved, and virtually all cost-benefit analysts continue to ignore the implications of the possibility of endogenous preferences in their work. In this article, I assess the merits of Becker's "extended utility function" valuation approach, and of the various valuation suggestions offered by Sunstein, Dau-Schmidt, or Bowles. I also offer my own thoughts regarding how the endogenous preferences valuation problem can be best addressed. My main conclusions are that the willingness to pay-based valuations of policies should be derived solely from the post-policy implementation preference structures, as a general matter, and that if there are any transitional preferences structures they should also be used to value those policy consequences that occur while they are in existence. These conclusions have significant implications for the valuation of environmental policies, as well as of policies of any sort that have as one of their major objectives the shaping of character and the alteration of attitudes and preferences. This theoretical endogenous preferences valuation framework may need to be modified at times to accommodate preference structure estimation difficulties and to counter potential analyst bias, particularly with regard to prospective cost-benefit analyses