Correcting for Selectivity Bias in the Estimation of Tourist Spending Surveys

Abstract

Focusing on a small tourist destination, this study investigates how the conventional design and administration of surveys of tourism expenditure can result in selectivity bias. The bias arises from the assumption that day and overnight visitors possess identical spending patterns, and thus can be surveyed with the same instrument. To remove such a bias, the authors apply a two-stage procedure to distinguish the expenditure decisions made by day visitors from those made by overnight visitors. The results suggest that conventional surveys potentially underestimate expenditures by day visitors and overestimate expenditures by overnight visitors.