Geography and the Internet: Is the Internet a Substitute or a Complement for Cities?

Working Paper: NBER ID: w10028

Authors: Todd Sinai; Joel Waldfogel

Abstract: We study the tendency to connect to the Internet, and the online and offline shopping behavior of connected persons, to draw inferences about whether the Internet is a substitute or a complement for cities. We document that larger markets have more locally-targeted online content and that individuals are more likely to connect in markets with more local online content, suggesting the Internet is a complement to cities. Yet, holding local online content constant, people are less likely to connect in larger markets, indicating that the Internet is also a substitute for cities. We also find that individuals connect to overcome local isolation: notwithstanding a large digital divide, blacks are more likely to connect, relative to whites, when they comprise a smaller fraction of local population, making the Internet a substitute for agglomeration of preference minorities within cities. Finally, using online and offline spending data, we find that connected persons spend more on books and clothing online, relative to their offline spending, if they are farther from offline stores. This indicates that the Internet functions as a substitute for proximity to retail outlets.

Keywords: internet; urban agglomeration; online shopping; local content; digital divide

JEL Codes: R00; R12; L81; L82

Causal Claims

Cause	Effect
larger markets (D40)	more locally-targeted online content (R23)
more locally-targeted online content (R23)	likelihood of individuals connecting to the internet (L96)
larger markets + local online content (F61)	likelihood of individuals connecting to the internet (L96)
individuals connect to overcome local isolation (F60)	likelihood of individuals connecting to the internet (L96)
black individuals (smaller fraction) (J15)	likelihood of connecting to the internet (L96)
distance from offline retail stores (L81)	online spending (L81)