Modelling growth and revenue for Swedish hotel establishments

Abstract

This paper provides first estimates of the determinants of output growth of Swedish hotels based on establishment data. Growth of overnight stays is modelled as a function of initial size, age, type of accommodation, location and hotel prices measured as average revenues per guest nights. The empirical model accounts for potential endogeneity of hotel prices through the two-stage least absolute deviation model (2SLAD) and the instrumental variable quantile regression method. 2SLAD estimates show a positive and significant relationship between hotel prices and subsequent growth. The relationship is nonlinear with a decreasing impact as the price level increases. Growth of establishments is significantly higher for smaller and younger hotels. An important result is that city hotels, in comparison to tourist and other hotels – which are mainly located outside urban areas (in the mountains and at the sea) – exhibit significantly higher growth rates with a gap between 2.2 and 3.4 percentage points. Furthermore, the positive impact of hotel prices on growth is larger for high-growth establishments. Accommodation prices significantly decrease with the number of local competitors with a non-linear form and increase with size. City hotels and accommodation in the capital city have the highest revenues per guest night.