Evaluating utility conservation programs: The BPA plan

This paper summarizes a detailed plan for evaluation of the residential energy conservation programs offered by the Bonneville Power Administration (BPA). The major purpose of the evaluation is to determine the energy savings attributable to the programs and to identify differences between program participants and nonparticipants. The focus of this paper is on the sampling issues associated with reliable estimates of energy savings. The methodology attempts to correct for the self-selection bias likely to occur in conservation program evaluation by using intentional nonrandom sampling. In addition, the paper suggests a variety of analytical approaches to analyzing the data with respect to energy savings.The proposed design, using four different groups to estimate program effects, is a comprehensive endeavor. This is appropriate for BPA because of the substantial investment it is making in conservation. However, the design is flexible and allows for use offewer groups to reduce evaluation costs.     

The effects of financial management strategies on quality of family life in remarriage

A conceptualization of financial management behavior in remarriage forwarded by Fishman (1983) is refined and used to study 91 remarried couples. Couples are categorized into three financial management strategy groups: those having only joint accounts, only separate accounts, and a combination of joint and separate accounts. The different groups' demographic characteristics and satisfaction and happiness regarding family life are compared. Overall few differences between groups are found on the demographic characteristics, and no differences are found on satisfaction or happiness. Her research interests are in remarriage and stepfamily relationships, fathering postdivorce, and stress and coping in adolescents. She received her Ed.D. from Indiana University. He received his M.S. degree in Human Development and Family Studies at Colorado State University, Ft. Collins, CO 80523. His research interests include factors affecting adjustment in remarriage and multiple remarriages and problem solving and creativity in families. before her death in December 1993. Her research interests were financial management and families in transition. She received her Ph.D. from the University of Tennessee at Knoxville.     

Folk theorems for the observable implications of repeated games

The fact that infinitely repeated games have many different equilibrium outcomes is known as the Folk Theorem. Previous versions of the Folk Theorem have characterized only the payoffs of the game. This paper shows that over a finite portion of an infinitely repeated game, the concept of perfect equilibrium imposes virtually no restrictions on observable behavior. The Prisoner's Dilemma is presented as an example and discussed in detail.I would like to thank an anonymous referee, Sushil Bikhchandani, David Hirshleifer, David Levine, Thomas Voss, and participants in the UCLA Game Theory Seminar for helpful comments.     

Effects of Increased Auto Safety Belt Use Levels on Fatalities

To reduce the annual toll of highway deaths, more than 30 states have passed laws mandating the use of safety belts. All have been effective at raising safety belt use; equipped with different provisions and enforced with unequal care, however, they have done so to different degrees. The article estimates the relationship between attained belt use and fatalities averted. Monthly 1982-1986 fatality, collision, belt use, and other data from 64 areas in nine states were collected, then analyzed using a Poisson multiple regression model. The analysis indicates that an increase in belt use from 14-40% averts about 13% of fatalities; a more pronounced increase to 50% averts about 18%. An increase from 50-75% averts about 16% of remaining fatalities. Three significant conclusions emerge. First, previous estimates appear to have understated, in general, the overall effectiveness of belt use laws. Second, the benefits of programs to boost safety belt use in this country from its current level of about 50% to up to 75%, estimated on the basis of more direct evidence from U.S. data than previously available, appear to be very large. Third, Poisson and other multiple regression models including explicit allowance for other causal factors can usefully complement other statistical approaches in traffic safety studies.     

Public opposition to genetic engineering

Abstract The extent and sources of public opposition to the use of genetic engineering in agricultural production are examined through data from telephone interviews with 220 farmers and 332 nonfarmers living in eight North Carolina counties. A model suggesting that public opposition to genetic engineering is influenced by demographic characteristics mediated by three intervening variables (awareness of genetic engineering, faith in government and industry, and moral objection to genetic engineering) is analyzed. Moral objection is the strongest predictor of opposition. Opposition is also related to lower awareness and less faith in institutions. Women are more likely to oppose genetic engineering than men. Implications of findings are noted.