Who uses geolocated social media in the United States?

Abstract

There has been increasing interest from both city planning scholars and practitioners in using geolocated social media posts as a way to learn about urban activity patterns: which public spaces are most popular at different times, what people’s travel patterns look like, and so forth. Information about residential demographics, employment, and motor vehicle traffic is often already available from census agencies or roadway sensors, but widespread smartphone adoption now provides novel information about people’s actual uses of urban space. However, not all residents of cities have access to smartphones, and those who do may not use the type of apps like Twitter, Instagram, Flickr, or Foursquare that create public traces of location and activity context. From large-scale surveys, we know that users of these services in the United States tend to be younger and wealthier than average residents, and are less likely to be from disadvantaged racial or ethnic groups. This leads to concern that city planners utilizing this data may inadvertently privilege some members of the community over others. This paper demonstrates a methodology to assess the demographic characteristics of geolocated social media data for metropolitan areas in the United States, using data form the U.S. Census. We will report findings for the entire U.S. if possible, and otherwise will focus on the San Francisco Bay Area, a region of about 9 million people in the state of California.