Skip to main content

Harvard Economists: Yelp Data Can Help Predict Gentrification

Yelp data can tell us so much about the cities we live in. It can tell us the best restaurant for dinner, barbershop for a haircut, or even help find a great plumber. A study released today by three Harvard economists uses Yelp data to provide new insights into gentrification at the neighborhood level – an issue that has often stymied researchers and policymakers due to limited public data.

The authors of this study, Edward Glaeser of Harvard University, Hyunjin Kim, and Michael Luca of Harvard Business School, previously found that Yelp data can be used to measure local economic activity in real time. Their new study shows that Yelp data can help us understand how the mix of businesses changes in gentrifying areas, and how changes in the business landscape can predict which neighborhoods are gentrifying.

For their paper, “Nowcasting Gentrification: Using Yelp Data to Quantify Neighborhood Change,” the authors looked at various measures of gentrification and its manifestations – including changes to a neighborhood’s demographics, housing prices, and physical landscape (which is coded from Google Streetview). The team then compared those changes to patterns in Yelp data across four major business categories: groceries, cafes, restaurants and bars. The study focused on five cities: New York City, Boston, Chicago, Los Angeles and San Francisco. Its main findings are:

  1. Gentrifying neighborhoods – as measured by Census data on housing prices and neighborhood demographics –  tend to experience growth in their local economy, as measured by the number of Yelp-listed establishments.  As housing prices and the share of the population that is college-educated increase, so do the number of local businesses.
  2. The number of businesses in a neighborhood, as measured by Yelp, is a predictor of housing price changes. For example, when a cafe opens in a neighborhood, home prices there rise by 0.5% relative to other areas in the coming year. The number of reviews can also help to forecast price changes. Business metrics on Yelp can more generally help to measure gentrification in real time.

Yelp data gives rise to these kinds of insights in a way that other sources, such as Census data, can’t because of time lags in reporting. Yelp data provides up-to-date information that is useful for understanding when and where gentrification is occurring. This is what economists refer to as “nowcasting.” In other words, Yelp can tell you where neighborhood changes are occurring in real time, long before detailed public statistics become available.

Moreover, Yelp data gives a more nuanced look at how the local economy changes as housing prices rise. Data from online platforms like Yelp can help chart the course of a city’s neighborhoods, and can have implications for consumers as well. It’s helpful for home buyers to know that groceries, cafes, or bars opening in a neighborhood are signals of increases in home prices. If you see a new coffee shop enter a neighborhood, that can be a signal of a warming local housing market.

The study is part of Yelp’s ongoing initiative to help researchers like the Harvard team here use Yelp data for economic analyses. Our parallel internal effort, led by Carl Bialik, Yelp’s data science editor, explores how Yelp data can improve our understanding of the world around us. We hope to surface insights that add value to policymakers, businesses, and consumers.