A note on the welfare effects of omitting substitute prices and qualities from travel cost models.

The bias to welfare estimates from omitting substitute prices or qualities in travel cost models is examined. The presence of bias depends on whether single or multiple site changes are examined and the degree of correlation between the omitted and included prices. For a multiple site price or quality change, even when omitted price is uncorrelated with included price, welfare estimates may be biased. Alternatively, if omitted price is perfectly correlated, the welfare measure may not be biased. The presence of the bias depends equally on both the degree of correlation and whether a single or multiple change occurs.