We challenge the conventional approach to distinguish between professional and non-professional Airbnb hosts by solely using the number of managed listings.
We leverage the recently released platform policy that categorizes hosts' professionalism by their self-declared status. Our multinomial modeling approach predicts true host status, factoring in the number of managed listings and controlling for listing and host traits. We employ data from five major European cities collected through scraping the Airbnb webpage.
Our research reveals that relying solely on the number of listings managed falls short of accurately predicting the host type, leading to difficulties in evaluating the platform's impact on the local housing market and reducing the effectiveness of policy intervention. Moreover, we advocate using more fine-grained measures to differentiate further between semi-professional and professional hosts who exhibit heterogeneous economic behaviors.
Reliable professionalism metrics are essential to curb unethical practices, promote market transparency and ensure a level playing field for all market participants.
This work pioneers the revelation of the inadequacy of a commonly used measure for predicting host professionalism accurately.
