A latent class approach to tourists’ length of stay

In this paper, a conditional demand function is estimated for the length of stay at a tourist destination. The microeconomic model specifies the demand for the length of stay conditional on the remaining holiday characteristics that the tourist has chosen (the destination, type of accommodation etc.). The empirical model uses a latent class truncated Poisson regression, which acknowledges the existence of groups or segments of demand with different preferences. In the case under study, a distinction is made between two segments with a preference for either a longer or shorter holiday. The statistical model also resolves a problem of multimodality, acknowledged in previous analyses to be one of the main characteristics of this variable.