Empowering Inclusive Work
How do workers with disabilities perform in the same job as non-disabled workers, and can AI-enabled technologies such as modern text-to-speech (TTS) narrow any gaps? We examine the relative performance of deaf or hard-of-hearing (DHH) workers on a major Chinese food-delivery platform, providing the first large-scale productivity comparison of disabled and non-disabled workers, and we evaluate the effects of a TTS tool for DHH workers to call customers. Prior to the tool, DHH workers are slower and receive worse customer ratings, but work more, quit less, and generate higher platform profits. After the tool’s sudden introduction mid-sample, difference-in-differences estimates show faster deliveries; substantially fewer bad customer ratings; and higher labor supply, retention, and profits. The tool closes one-third of the disability hourly-pay gap.
-
-
Copy CitationYanyou Chen, Mitchell Hoffman, Huilan Xu, and Zhe Yuan, "Empowering Inclusive Work," NBER Working Paper 35372 (2026), https://doi.org/10.3386/w35372.Download Citation
-