Research Highlights Article

June 11, 2018

There goes the neighborhood

Online platforms like Yelp can help economists learn about how gentrification happens in real time.

Chris Fleisher

A Starbucks cafe in Chicago, Illinois, in 2016.

Sorbis/Bigstock

Last November, a Denver coffee shop caused a public uproar with a sidewalk sign.

“Happily Gentrifying the Neighborhood Since 2014,” the sign at Ink! Coffee said, a joke that struck a raw nerve in a rapidly changing area of the city.

Also, it was perhaps too true. Indeed, cafes opening in a neighborhood are a reasonably good sign that an area is gentrifying, according to a paper in AEA Papers & Proceedings. So are groceries, laundromats, wine bars, and vegetarian restaurants.

Researchers Edward Glaeser, Hyunjin Kim, and Michael Luca use data from the business listing platform Yelp to study how changes in the local business environment relate to the demographics of its residents. The paper offers insights into how researchers can use digital platforms to better understand how neighborhoods evolve in real time and has implications beyond economics.

“We see multiple audiences for this," Luca said in an email to the AEA. "Policymakers benefit from having real-time snapshots of how neighborhoods and the local economy are evolving overtime. This is useful both for planning and policy purposes. Potential investors and homebuyers also benefit from understanding, for example, how housing prices are likely to change in different neighborhoods over time. Real estate investors might want to know where housing prices are likely to increase."

Gentrification has long been a concern for policymakers in economic development. An influx of wealthy, younger, and typically white residents often accompanies rising housing prices that force out longtime residents who are more invested in the area and its important institutions, such as schools. Gentrification can also perpetuate residential segregation by race and income, furthering economic inequality.

The issue has often been talked about in terms of the changing demographics of residents, but as this paper points out, gentrification also relates to change in the local business activity.

Policymakers benefit from having real-time snapshots of how neighborhoods and the local economy are evolving overtime. This is useful both for planning and policy purposes.

Michael Luca

Gentrification often occurs much faster than policymakers can respond to protect vulnerable residents and institutions. Platforms like Yelp and StreetScore, which draws on crowdsourced data to rate perceived public safety of an area, provide researchers an opportunity to study how neighborhoods change in real time.

The authors combine information from digital platforms with more traditional Census data to study rising home prices in New York City and how that relates to changing demographics. In assessing whether a neighborhood is gentrifying, they look for increases in the share of people with a college degree, residents aged between 25 and 34, and individuals who are white.

They then study whether increases in various types of businesses, like cafes, are indicators of a gentrifying neighborhood. It’s reasonable to expect that as an area’s demographics change, so would local demand for certain types of businesses. If the number of Yelp reviews increases for a Starbucks, does that mean the gentrifiers are coming?

Indeed, they find that Starbucks cafes are correlated with rising housing costs. Every new Starbucks that comes into a neighborhood is associated with a 0.5 percent increase in housing prices. Interestingly, an increase in reviews of Starbucks is an even better indicator of rising home prices. Every additional ten Starbucks reviews for a given ZIP code is associated with a 1.4 percent increase in home prices increases. However, it's unclear what comes first: the Starbucks or the gentrifiers.

Starbucks cafes aren't the only signs that wealthy educated white people are moving in. There are other businesses that have even stronger associations with gentrification. The number of groceries is significantly correlated with the share of adults with college degrees, and to a lesser extent people aged 25-35 and white residents. Laundromats, cafes, and wine bars are also correlated with these changing demographics.

Meanwhile, vegetarian restaurants were the strongest indicator of increased perceptions of street safety. Starbucks came in a close second.

Though enthusiastic for the potential for economists to use digital platforms to better understand the economy, the authors are careful to point out the limitations. Yelp and StreetScore are not substitutes for Census data, the authors say, but “they provide an important complement” by offering unique insights in close to real time. Luca hopes this paper will spur more companies -- from online platforms to credit card companies --  to think about how their data might help researchers better understand the economy. 

"The broader version of this question is how should we think about finding the right data in the digital age?" Luca said. "In our view, each data source -- which includes everything from Yelp reviews to LinkedIn connections to noise sensors being placed on trash cans and buildings -- is a piece of the puzzle. We are fascinated by the question of how researchers and policymakers might use all of these various datasets to help improve our understanding of the economy."

♦

“Nowcasting Gentrification: Using Yelp Data to Quantify Neighborhood Change” appears in AEA Papers and Proceedings.