The determinants of the hotel sector revenues: The case of French Polynesia

Abstract

This article is an empirical analysis focused on the hotel sector in French Polynesia in 2007–2017. One assesses the impact of a set of variables on the French Polynesian hotel sector monthly revenues through a gravity model. First, one specifies a basic model that embeds several potential explanatory variables (the exchange rate (both nominal and real), the rate of unemployment, the geographical distance, some specific historical events, etc.). Next, a second model is specified so as to assess the impact of hotel capacities measured by the number of bedrooms offered. Estimates rest on an unbalanced monthly panel database that embeds main countries from where tourists present in French Polynesia are originated. In order to compare results, each specification is estimated by two methods: classical panel regression (Ordinary least squares /Generalized least squares) and pseudo Poisson maximum likelihood. Both methods lead to coherent results.