Download PDF

Abstract

Social behaviours emerge from the exchange of information among individuals—constrained by and reciprocally influencing the structure of information flows. The Internet radically transformed communication by democratizing broadcast capabilities and enabling easy and borderless formation of new acquaintances. However, actual information flows are heterogeneous and confined to self-organized echo-chambers. Of central importance to the future of society is understanding how existing physical segregation affects online social fragmentation. Here, we show that the virtual space is a reflection of the geographical space where physical interactions and proximity-based social learning are the main transmitters of ideas. We show that online interactions are segregated by income just as physical interactions are, and that physical separation reflects polarized behaviours beyond culture or politics. Our analysis is consistent with theoretical concepts suggesting polarization is associated with social exposure that reinforces within-group homogenization and between-group differentiation, and they together promote social fragmentation in mirrored physical and virtual spaces.

Virtual spaces mirror income inequality

CAMBRIDGE (October 24, 2019) — Income inequality drives social segregation and polarization not just in urban neighborhoods, but in online communities as well. That is the conclusion of a new paper by the New England Complex Systems Institute (NECSI) published in Royal Society Open Science. Importantly, this societal fragmentation is more than just the top one percent versus the bottom 99: it exists between every economic class.

The Internet democratized the exchange of information, but the evolution of online social networks has mirrored the segregation of urban neighborhoods in real cities, according to NECSI’s analysis of millions of tweets. Social media users have organized themselves into economically segregated echo-chambers. This breakup of information reinforces the fragmentation and polarization of communities.

By examining where people tweet and with whom they chat, NECSI researchers were able to map the networks of social mobility and communication in Istanbul, New York City, and several other U.S. cities. The networks of tweets were then compared to census data on neighborhood income.

The results show that people primarily interact with their own socio-economic group. Different income groups are distant both in the physical space and online. They are neither found in the same places, nor discussing similar issues. This divide exists not just between the wealthy and the poor, but more granularly between socio-economic classes.

Many U.S. cities have a history of racial segregation tied to economic class, but social fragmentation can arise autonomously in any community. Individuals share information and imitate the social norms of the people most familiar to them, self-reinforcing group identities.

Analysis of hashtags reveals the divergent topics being discussed in rich and poor neighborhoods. In American cities, lifestyle hashtags abound in richer areas, while sports, zodiac signs and horoscopes seem to be more popular in poorer areas.

For most cities, social segregation and polarization is driven more by the lack of mobility between neighborhoods, than the geographic distances between them. This means that urban planning policies can influence the culture of neighborhoods. Desegregating the places where people live, work and shop may foster more interactions and communication, reducing polarization and conflict.

Figure 5. Polarization of discussion topics across neighbourhoods. Spatial distribution of topics on Twitter for Istanbul (a) and New York City (b). (a(i),b(i)) Dominant topics in wealthier areas (Topic-1). (a(ii),b(ii)) Dominant topics in poorer areas (Topic-2). Dots indicate geolocated neighbourhoods and colours indicate the normalized intensity of the cluster in each neighbourhood (colour scale shown in figure). (a(iii),b(iii)) Scatter plots in which each neighbourhood is represented by a dot and the coordinates represent the extent of conversation of Topic-1 (vertical axis) and Topic-2 (horizontal axis) in that neighbourhood. Dot colour indicates normalized median income (scale inset). Results for more cities are presented in figure 6.