Peer-to-peer market for tourist accommodation: characterization and implications

This doctoral thesis aims to enhance the knowledge on the functioning of the peer-to-peer (p2p) market for tourist accommodation, bringing ideas, arguments, and evidence not brought forward so far. Framed within the context of the (pseudo-)sharing economy, the starting hypothesis is that the p2p market for accommodation does not work too differently from standard markets. Thus, conventional economic theory (profit- and utility-maximization) is used to create models, explanations and predictions; and these are empirically implemented thanks to econometric techniques such as those related to hedonic pricing models, ordinary least squares, nonlinear maximum likelihood estimators for probability, differences-in-differences and quasi-experimental methods, and multilevel techniques for hierarchical data. Among the thesis' main goals, it intends to identify the recent trends that are configuring the p2p market for tourist accommodation, leading to questions such as why the percentage of hosts that activate the instant booking option is growing, or whether the concentration of supply influences prices in the p2p market for tourist accommodation; and even hot topics of debate such as whether Airbnb is bad for hotels, or what its impacts on the real estate market are. Thanks to the exploitation of natural experiments, this thesis also delves into the question of how hosts in the p2p market for tourist accommodation adapt to external shocks. And, with the help of analyses of particular regions, it also documents how the expansion of this market affects the seasonal pattern of tourist demand in islands; the dependence on the maturity of the touristic destination of the impacts of the p2p market for accommodation; and even, the peculiarities of its functioning in Latin America. Given the importance of all these issues in terms of economic policy, an additional objective has been to recommend regulatory instruments to improve public decision-making