An Analysis of Visitors' Expenditures in a Tourist Destination: OLS, Quantile Regression and Instrumental Variable Estimators

Abstract

The impact of tourism on the local economy and employment has been widely reported in the literature through the use of regional or county expenditure multipliers. Indeed, the money that visitors spend in a community is likely to be the benefit that residents probably recognize most readily. Despite this, only a few studies have focused on the determinants of tourist expenditure at the individual or household level. This paper sheds further light on this issue. For this purpose, the authors use information gathered by interviewing visitors about their levels of expenditure. The analysis focuses on total expenditure and expenditures on lodging, restaurants, shopping and car rental, and uses OLS, quantile regression and instrumental variable techniques. The results suggest that the impact of sociodemographic and trip-related characteristics on tourist expenditures goes far beyond the mean effect. Furthermore, correcting for the endogeneity of some explanatory variables might be a worthwhile exercise.

Insights from analysing tourist expenditure using quantile regression

“…Following the example of Santos and Cabral Vieira (2012), an OLS regression was first used to estimate the coefficients based on the conditional mean. The quantile regression approach was then applied to take into consideration unobserved heterogeneity across quantiles and heteroscedasticity among the error terms.…”

On the Relationship between Length of Stay and Total Trip Expenditures: A Case Study of Instrumental Variable (IV) Regression Analysis

“…In this respect, it is worth reemphasizing that the goal of IV regression is to obtain a correct estimate of an endogenous explanatory variable and not to maximize the R-squared. Santos and Vieira (2012) recently used IV regression when examining how a number of determinants explained variation in micro-level tourism expenditures. Based on knowledge gained from the human capital literature, they considered (the amount of) education as a possible endogenous variable and found that the IV estimates for tertiary schooling tended to be higher than the analogue OLS estimates.…”

“…In this regard, the education level of tourists has not really been considered an important predictor of tourism expenditure at the micro level (Marcussen, 2011). Therefore, the remedy proposed by Santos and Vieira (2012) likely has small practical consequences. Nor did these authors provide any information about the instruments they used, which is the topic of the next subsection.…”

Bargaining Power in Tourist Shopping

“…The literature suggests that the most common technique for estimating tourist expenditure is ordinary least squares (OLS) regression (Thrane 2014). Apart from OLS, the quantile regression model has been found to present a more comprehensive picture by incorporating heterogeneity from expenditure distribution (e.g., Lew and Ng 2012; Santos and Cabral Vieira 2012; Hung, Shang, and Wang 2012) in order to fully understand tourist expenditure and provide more accurate information to tourism marketers. The explanatory variables of tourist expenditure can be categorized as economic, social, psychological, trip-related, and destination-related (Yi.…”