Research Note: Seasonal Preferences and Survey Design — Tourism Expenditure Models with Household Budget Surveys Revisited

Abstract

There is an increasing strand of literature on tourism expenditure which draws microdata from national household budget surveys (HBS). In this paper the authors discuss the fact that, owing to tourists' seasonal preferences, papers based on HBS that do not use spending information covering the whole year for each household are seasonally biased. The authors show that this bias occurs as a result of either the design of the HBS or inadequate treatment of the data by researchers. When the bias occurs, tourism expenditure estimates are misleading. Analysing the issue in detail, the authors argue that it underestimates the participation rate and exerts a downward bias on income elasticity values. These hypotheses are tested empirically with Spanish microdata, showing that the seasonal bias severely affects the results obtained. In this regard, previous literature on tourism expenditure affected by this seasonal bias should be revisited.