This study aims to explore the impacts of Airbnb listings on land values in the Austin, Texas, USA area, particularly on single-family homes. The goal of the analysis is to shed light on how greatly and in what direction Airbnb is affecting the housing market, with an emphasis on the spatial distribution of its effects.
The analysis in this paper is performed using three distinct models on a data set of land parcel data within Travis County: an ordinary least squares regression model, a geographically weighted regression (GWR) aimed at detecting the influence of variables at the census tract level, and a Bayesian approach, which describes spatial and temporal effects on the data.
The findings of the analysis indicate that across the years 2013 to 2019, higher numbers of Airbnb listings were associated with lower percentage increases in land value in certain tracts in the northern and eastern parts of the city. Additionally, the results of the Bayesian model indicated that much of the change in land value can be attributed to unobserved factors within census tracts.
The contribution of this study to the existing literature is its analysis of the spatial and temporal analysis of the effects of Airbnb listings on land value using a GWR and a Bayesian model. Also, as the negative correlation found in the study departs from previous research, this paper may provide policymakers insight into the complex spatial distribution and conflicting effects of Airbnb listings across distinct parts of cities.
