Impact of short term rentals on the rental affordability in San Francisco – the case of Airbnb

The short-term rental housing options supported by the sharing economy have now been established as disruptions in the housing and tourism sectors. With the advent and proliferation of Airbnb rentals all over the world, questions have been raised about their impacts (Meni, 2017). While one side discusses the democratization of tourism industry and how Airbnbs makes it easier for tourists to travel and experience places, the other side brings forth the skirting of hotel taxes and negative neighborhood externalities that inevitably result from these rentals. Airbnb has attracted controversy in cities all over the world, with high-profile lawsuits centered around criticism for evasion of taxes and for avoiding regulatory oversight that is otherwise enforced on hotels and providers of similar services. The presented research is an attempt to gauge the impact of Airbnb on rental affordability by using spatial econometric analyses. The study areas for the aforementioned research is the San Francisco Metropolitan Statistical Area (SF MSA). The hypothesis is that an increase in the Airbnb listings (i.e. the short-term housing stock) in the study area is correlated negatively with rental affordability, causing it to decrease. Key research questions are does Airbnb impact the rental affordability in an area? If yes, then, to what extent? To answer these questions, both cross-sectional and longitudinal analyses are undertaken. Aiming to contribute to the body of literature, revolving around the debate through quantitative analyses and regulatory policy discussion, this study finds positive and statistically significant correlation between both Airbnb variables (percent Airbnb listings as a proportion of rental housing units and weighted Airbnb listings based on listing types) and variables representing rental affordability like percent rent-burdened and overburdened households, median rents and median house prices). Various models were considered for both cross-sectional and longitudinal analyses using different combinations of the aforementioned variables. The spatial econometric analysis answers one of our key questions in the affirmative – the presence of Airbnb rentals does impact the rental affordability in an area. Having established a relationship, our second objective was to gauge this impact’s extent. Simulations were run to understand the results of the spatial econometrics models to help visualize this impact. In the case of cross sectional analysis for San Francisco MSA, these simulations showed that for a typical census tract (one with median percentage of Airbnb listings, as a fraction of the rental housing market) a 1% increase in Airbnb listings corresponded to a 0.06% rise in the rent overburdened household category. Hence, in the case of a census tract with 10,000 households, a 10% increase in percent Airbnb listings will correspond to 60 more households being added to the rent overburdened category. This effect is more pronounced for tracts with a lower number of Airbnb listings (10th or 25th percentiles). Additionally, tracts with no or a low percentage of Airbnb listings will have more households pushed to a rent-burdened category, with a similar rise in Airbnb listings. In the case of longitudinal analysis of panel data for San Francisco City for a period of four years (2013 – 2016), the simulations show that census tracts with a smaller presence of Airbnb listings (those below the 50th percentile) were more sensitive to an increase in Airbnb listings i.e. they saw a higher increase in the median house price per tract as compared to census tracts in the higher percentiles. This trend was consistent across all four years affirming the extent of the impact of Airbnb listings on the rental affordability in an area