Le temps et la valeur hédonique des hébergements locatifs entre pairs : l’exemple d’Airbnb

A growing amount of economic literature explores paid online peer-to-peer accommodation pricing, a phenomenon that has grown dramatically worldwide over the past decade. Thus, academics have identified accommodation characteristics and estimated their implicit prices using the hedonic method. However, the influence of the length of stay on willingness to pay attributes is rarely considered. Using data from 47,756 Airbnb listings in 31 French cities, we examine the moderating effects of the city's average length of stay on the implicit price of main attributes (centrality, neighborhood, size, facilities). Moderating effects are estimated through OLS and corrected heteroskedastic models with multiplicative interactions per accommodation type (entire homes, private rooms, and shared rooms). In addition, we used quantile regressions to explore the significant effects of quality level on these moderating effects. We provide evidence of significant moderating effects by accommodation type and quality level that shed new light on the valuation of these accommodations. According to accommodation type and price level, the length of stay moderates the effect of centrality and neighborhood on prices. The positive influence of the accommodation capacity of the entire home decreases with the length of stay, suggesting that the value of the surface area per occupant increases with time. The length of stay may significantly moderate the willingness to pay for specific facilities. The results may be helpful in host pricing strategies, the tourism accommodation industry, and urban politics. Moreover, considering the time effect, we discuss what we know about urban key concepts (centrality, neighborhood, surface).