Quantifying Gentrification With Public Data

Wouldn’t it be great to have invested in Brooklyn 15 years ago? How about SoMa in San Francisco or Wynwood in Miami? In the past, investing in up and coming neighborhoods occurred as a result of gut feel and “knowing the market”. However, with advances in machine learning and user based location data (Yelp, Foursquare, Google, Twitter, etc.), it has become increasingly possible to understand the composition, quality, and evolution of a given neighborhood.

Researchers at Harvard Business School recently fused Yelp data with Census, Federal Housing data, and Streetscore (a Google product) to understand the composition of different neighborhoods. They were able to clearly identify a relationship between increased housing prices and growing numbers of local groceries, cafes, restaurants, and bars.

In a different study done at the University of Cambridge, researchers fused Twitter and Foursqaure data with the UK Index of Multiple Deprivation (IMD) (a statistical exercise conducted by the the British government which measures the relative prosperity of neighborhoods across England). They found a correlation of a “deprived” area — meaning low economic status — with “high social diversity” meant that a neighborhood was in the process of being gentrified.

Even focusing on a single type of business can highlight the changes in a given neighborhood. In the HBS study, it was discovered that the entry of Starbucks (or other similar coffee shops) were indicative of housing price growth. They also found that Yelp establishments from 2007-2011 predict changes in educations levels over the next five years.

Looking at the work done by and organic grocers like Whole Foods, the pattern is equally as visible. Whole Foods recently announced two new locations in Washington D.C., which was viewed by locals as a sign that their neighborhood was on the up and up. One location, two blocks from Howard University, “was interpreted as proof that the neighborhood’s recent gentrification is no ephermal phenomenom”.

This makes sense, though, as Whole Foods relies on “heavily on micro-demographics and complicated algorithms” to determine prime locations. On aggregate, when taking into account locations of Whole Foods and competitors like Trader Joes and Harris Teeter, you can quickly get a sense for the quality of neighborhood. These locations, paired with other types of businesses like Barry’s Bootcamp and Blue Bottle Coffee, can create a beautiful picture of neighborhoods you want to invest in.

Commercial Real Estate deals have come a long way over the past decade. With growing institutional allocations and increased foreign investment, there is a need to be more data-driven to have an edge – open data provides that.

Sources: HBS, University of Cambridge, The WashingtonianInverse.com