Limited Attention, Asymmetric Information, and the Hedonic Model

Abstract

The broad objective of this research is to gauge the importance of relaxing the full information assumption in revealed preference models when decisions are made in complex, public information environments. This thesis focuses on housing markets. An information acquisition process is outlined that describes why homebuyers are often less informed than sellers for some housing attributes when they face more stringent information search and processing constraints. Adapting the hedonic model for the possibility that sellers are more informed than buyers suggests that estimates of the implicit price for a housing attribute may be attenuated towards zero if there is asymmetric information about the quantity of the attribute. The importance of the asymmetric information argument is gauged by applying the quasi-random experiment methodology to three applications involving exogenous information shocks for different housing attributes.

The first of these applications describes the impact of an airport noise disclosure on housing prices. The results indicate that the disclosure reduced housing prices near the airport by 2-3 percent. This suggests that an estimate of the implicit price for airport noise would have been attenuated towards zero by approximately 36 percent prior to the disclosure. The second application described the impact of a flood plain disclosure on housing prices. The results indicate that the disclosure reduced housing prices in designated flood zones by approximately 4 percent. Thus this application reconfirms the results from the airport noise application and the conceptual framework.

The third application describes the impact of information shocks related to the locations of registered sex offenders on housing prices. The results indicate that housing prices fall by 2 percent within one tenth of a mile of a registered sex offender when a sex offender moves into a neighborhood. However, this impact was not affected by increased media attention surrounding two child-abductions committed by registered sex offenders near the study area. These results are somewhat less conclusive about the role of asymmetric information on the estimated implicit price for proximity to sex offenders.