Xiang Liu Imperial College London, London, United Kingdom, John Polak, Centre for Transport Studies Imperial College London , London, United Kingdom
On the Prevalence of Expected Utility and Non-expected Utility Decision Rules in Travellers’ Choice Amongst Uncertain Prospects (assigned to theme N3)

In recent years there has been a growing recognition that conventional models of travellers’ decision making need to be extended to accommodate the fact that travellers are often uncertain regarding the outcome of their decisions. For example, a traveller may choose to use a rail service with a scheduled arrival time that permits the timely commencement of a business meeting; yet unexpected delays en-route may cause the train to be late and the meeting to be delayed. The conventional microeconomic approach to modelling choice under such uncertain conditions, and the one used in almost all the work to date in transport, is to assume that travellers’ respond to the expected utility of alternatives (where this expectation is evaluated over the universe of possible states of the transport system). However, there is a significant body of evidence from the fields of experimental economics and economic psychology of the existence of stable, repeatable and systematic violations of the expected utility hypothesis and recent work has suggested the existence of similar phenomena in certain transport choice contexts. These findings have lead to the development of alternative approaches such as Regret and Prospect Theory, which can accommodate these violations more adequately. However, to date most of this work has been based on highly stylised laboratory experiments conducted on very small samples of respondents who were presented with rather abstract choice situations (e.g., simple lotteries involving cash pay outs). Whilst we know that under such conditions, violations of the expected utility hypothesis can be generated, we have very little evidence regarding the prevalence of such violations in more naturalistic choice contexts. Moreover, few studies to date have properly distinguished between or jointly accommodated the consequences of uncertainty affecting the decision maker regarding states of the transport system and uncertainty affecting the analysts regarding the perceptions and preferences of the decision maker. The objective of this paper is to examine these issues using data from a large, naturalistic stated preference exercise in which respondents were faced with a series of choices between alternative hypothetical unreliable train services. In the first section of the paper we present a brief overview of the literature on choice under uncertainty focusing in particular on recent attempts to integrate expected utility theory (and its derivatives) and random utility theory. The second section provides a brief description of the stated preference experiment, which collected data from a total of approximately 200 respondents. The third section formulates alternative models for the SP decision context, based on expected utility (EU) and prospect theory (PT). The prevalence of alternative decision rules in the data is investigated by assuming that each individual is either an EU or a PT decision maker and using a latent class model formulation to estimate the proportions of EU and PT within the sample. Our models also account for heterogeneity in travellers’ preferences. The fourth section presents our model estimation results and discusses their significance. The final section of the paper presents the overall conclusions.

Paper not on CD

Conference organized through conf-vienna (copyright Gunther Maier)

<