Yale School of Management

Center for Customer Insights

Advancing the frontiers of consumer understanding

The Rideshare Effect

When people talk about the economic impact of Uber and Lyft, they often talk about taxis and public transportation. But what about ridesharing’s far deeper and broader implications for the labor force as a whole?

August 27, 2020

Uber and Lyft arrived in Austin, Texas, in the spring of 2014. In December of the following year, the city council passed an ordinance requiring fingerprint background checks for all rideshare drivers. Uber and Lyft fought back. Austin ultimately prevailed, and, unwilling to concede the demand, Uber and Lyft cancelled service in May of 2016.

Up the companies went, to the state level, where they lobbied aggressively for House Bill 100, “relating to the regulation of transportation network companies.” The bill passed and, among other things, scuttled requirements statewide for fingerprint background checks of rideshare employees.

Come May of 2017, Uber and Lyft were back in Austin.

For Jiwoong Shin at Yale SOM and three of his colleagues, that year of absence, from May of 2016 to May of 2017, provides a rich intellectual vein to mine—a so-called “natural experiment” that lets them compare Austin with and without the presence of major ridesharing companies.

In one new study, for example, the researchers look at the relationship between ridesharing and restaurant quality. They investigate a straightforward hypothesis: Uber and Lyft increase the workforce options for people in the service economy; this, in turn, increases turnover in restaurants; and this results in a drop in quality.

In examining more than 150,000 Yelp reviews of Austin restaurants, from May of 2014 through May of 2019, this is precisely what the researchers uncovered. Through natural language analysis, Shin and his colleagues sorted every review based on whether it classified a restaurant’s service negatively, neutrally, or positively. They found that when Uber and Lyft first entered the city, the ratio of negative reviews began to climb; when the companies left Austin, this ratio declined; upon reentry, the ratio again went up.

Notably, a parallel analysis of how customer reviews talked about food quality revealed no effect. That is, Uber and Lyft seemed to detract from service quality, but not food quality. This trend was also most pronounced in the least expensive restaurants. Taken together, these results suggest that employment with Uber and Lyft is particularly attractive to a specific segment of the restaurant labor force: frontline, not kitchen, workers in low-end establishments.

Shin and his colleagues reinforced their findings by matching the case of Austin against the comparable cities of San Antonio and Houston, where Uber and Lyft were not absent between May of 2016 and May of 2017. No similar drop in service quality was detectable in either of these ‘control’ cities, which helps rule out the possibility of some larger, statewide trend that may have influenced service quality at the time. Similarly, while restaurant turnover rates rose in Austin upon Uber and Lyft’s return, the same was not true in San Antonio.

Our findings suggest significant policy ramification of the gig economy on broader industries through the labor market, and particularly low-skill labors.

And while this project focused narrowly on the restaurant industry, the findings present profound questions for the economy as a whole. Discussions of the effect of Uber and Lyft typically focus on incumbent industries that exist in clear competition—taxis, for instance, or public transit. But this work shows how ridesharing may disrupt far deeper economic currents by reshaping the contours of the labor market. As the researchers note, “our findings suggest significant policy ramification of the gig economy on broader industries through the labor market, and particularly low-skill labors.”