Guest Talk "Spatial social network patterns inside cities"
Date/Time: 09.11.2021, 10:00 am
Location: MS Teams*
Complex interaction networks of individuals in urban areas enabled by population density, co-location, and easy access together made cities the global engines of technological and economic progress. However, cities are also known for high levels of segregation, where disparate neighborhoods are separated from each other in the urban space. Furthermore, spatial segregation by income also fragments social networks, which can hinder progress and can deepen inequalities. Studying such phenomena in detail requires individual level spatial social network data with very high resolution. In this talk I will present two projects based on a unique, geolocalized Twitter dataset to study the spatial social network patterns of individuals inside the 50 largest metropolitan areas of the US.
The first project demonstrates that social capital measured by connection patterns of individuals is spatially more concentrated for residents of lower-income neighborhoods than for residents of higher-income neighborhoods. We map the micro-geography of individual online social connections in the 50 largest metropolitan areas of the US to analyze the spatial dimension of individual social capital by the share of friends, closed triads, and share of supported ties within short distances (1, 5, and 10 km) around users’ home location. We compare residents from below-median income neighborhoods with above-median income neighborhoods, and find that users living in relatively poorer neighborhoods have a significantly higher share of connections in close proximity. Moreover, their network is more cohesive and supported within a short distance from their home. These patterns prevail across the 50 largest US metropolitan areas with only a few exceptions. The found disparities in the micro-geographic concentration of social capital can feed segregation and income inequality within cities harming social circles of low-income residents.
The second project aims to better understand how mixing in urban social networks is facilitated by commuting. We use the same Twitter dataset to track the home and work locations as well as the mutual followership ties of 348,850 users in the 50 largest metropolitan areas of the US. An above‐median commuting distance in cities is linked to more diverse individual networks, moreover, we find that longer commutes are associated with a nearly uniform, moderate reduction of overall social tie assortativity across all cities. This suggests a universal relation between long‐distance commutes and the integration of social networks. Our results inform policy that facilitating access across distant neighborhoods can advance the social inclusion of low‐income groups.
* If you are interested in the talk, please contact email@example.com.