Short-term rentals in Portuguese metropolitan areas and the housing market

In the last decade, Portugal has witnessed a significant surge in tourism. This phenomenon, coupled with the proliferation of temporary accommodation services and short-term rental platforms such as Airbnb, has sparked an intense debate due to their perceived role in driving up house prices and displacing residents, shaping the urban landscape across the country,  especially in highly touristic areas. 

This paper studies the substantial growth of short-term rentals in Portugal (see the figure) driven by a pivotal 2014 policy change that created a streamlined process for hosts to register their short-term accommodation services, commonly known as local lodging establishments, to assess its impact on housing prices. Employing a novel dataset encompassing property transaction records between 2010 and 2017 for the metropolitan areas of Lisbon and Porto, the paper estimates a robust two-way fixed effects model at the quarterly level, which accounts for property-specific attributes, location variations, and time-fixed effects. 

The paper reveals various noteworthy findings. First, it documents a significant relationship between the number of local lodging establishments and house prices, with a ten-unit increase in the number of local lodging establishments translating into a 1.4%-2.7% house price increase. Second, the results show significant differences in effect between Lisbon and Porto, with the effect being felt through the entire municipality of Porto, while in Lisbon it was constricted to its historical centre. Third, the paper documents more pronounced effects on properties on larger (with four bedrooms) and more expensive (at the upper quantiles of the market) houses. Fourth, the research reveals a decrease in the number of transactions involving new constructions, coupled with a positive influence on the value of stores substantiating the influence of short-term rentals in driving demand for consumption amenities. The reliability of the findings is validated through various robustness tests.

Click here to go to the paper by Francisco Nobre, Jardim Gonçalves and Ronize Cruz.