Is Uber Strangling the Restaurant Business?

Jiwoong Shin, professor of marketing at Yale SOM, likes to keep in touch with his former students. They tell him about what’s going on in the world and give him ideas for new research.

In December 2019, one of those former students, now a restaurateur in New York and San Francisco, came back to New Haven and met up with Shin for dinner. Shin asked how things were going. There were difficulties, the former student said: “Uber is killing my business.”

Shin was confused. Was Uber affecting people’s ability to get to his restaurants? Or was it competition from Uber Eats, Uber’s food delivery service? Neither, the student said. It was Uber itself. Driving for Uber, like working as an entry-level server, requires no special experience or qualifications, but the hours are more flexible and the working conditions and wages are potentially better. So every time a server was fully trained, a process that took six months or so, they would suddenly quit to drive for Uber.

“It was really shocking,” Shin says now. “Because of Uber, he was short of experienced workers constantly, and customers complained about the service.”

Was this a widespread issue, Shin wondered, or specific only to this particular group of restaurants? He consulted Google and learned that within the restaurant industry—one of the largest private industries in United States—the loss of relatively inexperienced workers to Uber was the number-one topic of conversation.

Shin wondered if this impression reflected a broader phenomenon. He and his Yale colleague Soheil Ghili, Minkyu Shin of the City University of Hong Kong, and Jaehwan Kim of Korea University, Seoul, set out to study how newer businesses like Uber and Airbnb are disrupting labor markets.

As it happened, Uber and its main competitor, Lyft, had both left Austin, Texas, in May 2016 because of new local regulations, and then returned a year later, after the state passed statewide regulations that overrode city ordinances—an elegant natural experiment. The researchers hypothesized that the entry of the rideshare apps into the local economy provided another employment option for low-skilled workers and would therefore lead to an increase in turnover at the city’s restaurants and a decrease in the quality of service.

They worked from two sets of data covering the period from May 2014 to May 2019, including the two years preceding and the two years immediately following the rideshare apps’ absence from Austin. The first data set, which would help them determine the level of employee turnover, contained month-by-month information on restaurant employees—where they worked, their job titles, and how much they were paid. The second was a set of Yelp customer reviews, which helped them determine restaurant quality, at least as perceived by customers. The researchers went beyond the star ratings and did a textual analysis of what customers had written in order to determine whether their main complaint was the service, the food, the ambience, or some combination. As a control, they compared the data sets with the same information from Dallas for the same time period.

Their findings were consistent with their hypothesis: In Austin, after the return of Uber and Lyft, the number of restaurant employees decreased, while the number of customer complaints related to service increased by roughly 10%. In Dallas, the numbers remained consistent.

Interestingly, the change in customer satisfaction was more pronounced in lower-end restaurants, coded on Yelp by a single dollar sign (as opposed to two or three dollar signs for more expensive restaurants). For these restaurants, the complaints about service doubled. This tracks with the hypothesis, too, says Shin: “In high-end restaurants, people get paid well. They don’t want to be Uber drivers.”

Neither do chefs or managers, at any level; these are career jobs, Shin observes. The restaurant workers most likely to move around or quit and take another gig job are servers. These also happen to be restaurant workers who have the most direct interaction with the public, so their lack of experience would be more noticeable to customers.

Now restaurants in Austin and elsewhere find themselves working harder to keep their servers from defecting to rideshare apps and other gig jobs. The most obvious solution is increasing worker salaries, though this is not always easy to do in an industry that was already running on narrow profit margins before the inflation drove up the cost of food. Some restaurants have decided to do without servers altogether by automating the serving process with ordering kiosks, counter service, grab-and-go options, and even robots.

The ramifications of this study extend beyond the restaurant industry. The gig economy will continue to expand, likely accompanied by increasing calls for regulation. Some cities are already regulating Airbnb; in New York City, lawmakers have imposed stringent limitations that have effectively eliminated most Airbnb rentals.

“Policymakers need to understand the full scope of the gig economy’s impacts to design effective regulations that balance the benefits of gig economy and innovation against the potential for disruption in traditional labor markets,” Shin says. “Also, local government officials and economic development planners can use these findings to anticipate and manage the effects of the gig economy on local labor markets and economies. This could include initiatives to support sectors, probably low-skill labor intensive sectors, most affected by gig economy disruptions.”