Bias in Choice Experiments for Public Goods

  1. Lookup NU author(s)
  2. Professor Susan Chilton
Author(s)Carson K, Chilton S, Hutchinson G
Publication type Report
Series TitleNewcastle Discussion papers in Economics
Report Number2010-05
Full text is not currently available for this publication.
A “choice experiment” is a survey containing a sequence of choice sets of two or more options in which the levels of several attributes and the price vary. Respondents indicate their most preferred option for each choice set. Although these types of surveys have the advantage of increased statistical efficiency of parameter estimates, it is unknown whether this benefit comes at the expense of strategic or other types of non-demand revealing responses. This experiment systematically tests for the effects of both the number of options and single vs. repeated choices on non-demand revelation in a choice experiment. The results indicate that repeated choice formats are susceptible to non-demand revelation in favor of the status quo option. We propose several hypotheses about the potential causes of this bias.
InstitutionNewcastle University Business School
Place PublishedNewcastle upon Tyne, UK
NotesISSN: 1361-1837 To be submitted to Journal of Environmental Economics and Management May/June 2011
ActionsLink to this publication