EPJ Data Sci. (2020) 9: 17
https://doi.org/10.1140/epjds/s13688-020-00234-x

Regular article

Economic outcomes predicted by diversity in cities

¹ The Media Lab, Massachusetts Institute of Technology, Cambridge, USA
² MIT Connection Science, Institute for Data, Systems, and Society, Massachusetts Institute of Technology, Cambridge, USA
³ School of Management, Sabanci University, Istanbul, Turkey
⁴ School of Economics and Resource Management, Beijing Normal University, Beijing, China
⁵ Faculty of Engineering and Natural Sciences, Sabanci University, Istanbul, Turkey
⁶ New College of Florida, Sarasota, USA

^* e-mail: bahrami@mit.edu

Received: 26 November 2019
Accepted: 10 June 2020
Published online: 24 June 2020

Abstract

Much recent work has illuminated the growth, innovation, and prosperity of entire cities, but there is relatively less evidence concerning the growth and prosperity of individual neighborhoods. In this paper we show that diversity of amenities within a city neighborhood, computed from openly available points of interest on digital maps, accurately predicts human mobility (“flows”) between city neighborhoods and that these flows accurately predict neighborhood economic productivity. Additionally, the diversity of consumption behaviour or the diversity of flows together with geographic centrality and population density accurately predicts neighborhood economic growth, even after controlling for standard factors such as population, etc. We develop our models using geo-located purchase data from Istanbul, and then validate the relationships using openly available data from Beijing and several U.S. cities. Our results suggest that the diversity of goods and services within a city neighborhood is the largest single factor driving both human mobility and economic growth.