2002 World Congress of Environmental and Resource Economists

Travel Cost Modeling is a useful tool to assess the economic impact of changes to the environment. This methodology relies on the construction of consistent travel costs for all individuals. Technological improvements in computer software for optimal highway routing create two classes of subjects for many Travel Cost Model datasets – those panelists who can be precisely located and those whose origin location can only be approximately determined. This can lead to bias in estimation of coefficients and welfare measures associated with a change to baseline conditions. This paper proposes a simple correction and examines the impact in the case of beach demand in southern California. Panelists who are not precisely located have a higher variance in the stochastic term of their underlying utility equations in a simple Multinomial Logit model, and this is shown to affect the welfare measures for two sample counterfactual scenarios.



	Paper Title:	Correcting Bias From A Partial Increase In Precision In Cost Construction In Travel Cost Modeling
	Presenting Author:	Craig Mohn (UC Berkeley)
	Coauthor 1:	Michael Hanemann
	Coauthor 2:	Linwood H Pendleton
	Coauthor 3:	David Layton
	Abstract:	Travel Cost Modeling is a useful tool to assess the economic impact of changes to the environment. This methodology relies on the construction of consistent travel costs for all individuals. Technological improvements in computer software for optimal highway routing create two classes of subjects for many Travel Cost Model datasets – those panelists who can be precisely located and those whose origin location can only be approximately determined. This can lead to bias in estimation of coefficients and welfare measures associated with a change to baseline conditions. This paper proposes a simple correction and examines the impact in the case of beach demand in southern California. Panelists who are not precisely located have a higher variance in the stochastic term of their underlying utility equations in a simple Multinomial Logit model, and this is shown to affect the welfare measures for two sample counterfactual scenarios.
	Link to paper:	Not available

	Session / Day / Time	2E / Monday / 10:15 - 11:45 am

	Please use your browser's "back" button to get to the previous page.