Access to high speed internet has become a crucial resource in the 21st century, decisive for access to jobs and economic mobility, information and education, and participation in the social and political spheres. Yet, as with access to other types of infrastructure such as the public transit system, or streets, access to the internet, too, is characterized by socio-spatial disparities. Access to high-speed internet is differentiated based on geographies, class, race, and ethnicity, perpetuating existing forms of exclusion. But since the physical infrastructures that the internet relies on remain hidden in our everyday lives, buried below our feet, so do the inequities in its access.

This map aims to uncover the digital divide in 21st century New York by mapping the "invisible" internet infrastructure that is required for access to the internet and relating it to demographic and internet use data. Due to the limited data available (read more in the data section), the map is only a first step in doing so. However, I hope that it contributes to ongoing efforts of making visible the invisible (such as the crowdsourced New Cloud Atlas).

How to use this map? The map lets you compare and look for relationships among the various data layers by un/selecting them. For an explenation of the main findings and trends, please see the discussion section.

This map combines a multitude of data sets relating to the following three themes:

¹ PHYSICAL INTERNET INFRASTRUCTURE
Information on the physical internet infrastructure includes all free available WiFi hotspots (using data from NYC's Open Data Portal) and available broadband internet connections in commercial buildings throuhgout NYC (using data from NYC's Broadband Map).

² COMPUTER AND INTERNET USE
Information on access to and use of computers and the internet includes the percentage of households without computers at home per congressional district (using the Census Bureau's American Community Survey 2015) and the percentage of fixed residential internet connections with at least 10 Mbps downstream and 1 Mbps upstream capacity per census tract (using data from the Federal Communications Commission 2015).

³ SOCIAL CHARACTERISTICS
And finally, the first two data layers can be compared to demographic information, including the median household incomes at the census tract level in US$ and a race-dot layer showing the ethnic and racial composition of NYC with 1 dot representing 100 people (both data sets from the Census Bureau's American Community Survey 2015). .

The map reveals striking disparities in access to high-speed internet.

WiFi hotspots underserve lower-income populations: Out of the 2061 WiFi hotspots in the city, 1090 (or 53%) are located in higher-income areas with an above average median household income (MHHI that already have greater internet access. The remaining 971 hotspots (or 47%) are located in census tracts with an below average MHHI and generally lower internet access, and only 220 (or 11%) serve very low-income populations with an MHHI below $20,000. Put differently, while city-wide MHHI amounts to $59,000, the average MHHI of census tracts that provide free WiFi hotspots is $72,000. WiFi hotspots also disproportionally serve white neighborhoods: Out of the 2062 WiFi hotspots, 909 (or 44%) are located in majority white neighborhoods, and only 700 (or 35%) in majority non-white neighborhoods (thereof, 335 or 16% in majority African American areas, 309 or 15% in majority Hispanic, and 56 or 3% in majority Asian).

Availability of commercial broadband is also concentrated in higher-income and majority white neighborhoods. However, since the other layers on internet and computer use, and demographic characteristics are based on where people live, comparing thise layers has limited reliability. It shows, however, that commercial districts are concentrated in or near wealthier and whiter neighborhoods.

The availability of computers at home, too, correlates with income levels and the ethnic and racial composition of neighborhoods. Congressional districts with high percentages of households that do not own a computer (that is 14% and higher) are mainly located in the South Bronx, Northern Manhattan, and Central Brooklyn. These areas have below average MHHIs and high concentrations of minority populations.

The more granular data on access to broadband internet subscriptions show the digital divide along lines of class and race even more clearly. Census tracts with below average percentages of households with fixed internet subscriptions (that is 75% or lower) are concentrated in areas with below average MHHIs and high concentrations of minority populations.

Do you have ideas on additional data sets that could improve this map, have other comments or feedback, want to use the map, or have questions? Please shoot me an email!