CEU Electronic Theses and Dissertations, 2020
| Author | Tóth, Benedek |
|---|---|
| Title | Does host gender affect the prices of Airbnb listings? ??? A double machine learning approach |
| Summary | This thesis investigates how the host’s gender affects the price of listings on Airbnb. For this purpose, the double machine learning method is used to account for potentially nonlinear relationships between price, gender and other listing attributes. A standardized analysis is presented for 61 different locations, and the results are examined both together and individually. Gender effects are estimated for female hosts and couples in comparison to male hosts, and the relationships between these effects, host professionality and guest-host cohabitation in the listed apartment are also analyzed. These steps are necessary to disentangle different potential sources of gender effects, arising from either gendered pricing behavior or gender-sensitive demand. The thesis does not find evidence for the general presence of gender effects. Several individual coefficients in individual locations are significant but correcting for multiple comparisons invalidates this in the majority of cases. Only the effect of a professional, cohabiting hosting couple has a mean statistically different from zero (positive 3%) in the entire sample of results; this is also the group most likely to experience a positive gender effect based on the hypotheses presented in the thesis. The thesis also compares double machine learning estimates to linear regression and lasso coefficients and finds that the former is generally closer to zero and have a smaller variance |
| Supervisor | Matyas Laszlo |
| Department | Economics MA |
| Full text | https://www.etd.ceu.edu/2020/toth_benedek1.pdf |
Visit the CEU Library.
© 2007-2021, Central European University