• Chart of the Week
  • August 27, 2018

What our Facebook network reveals about us

Bigstock

It was once dismissed as a fad that would quickly be forgotten as a relic of the digital age.

But Facebook has become an essential platform for its more than 2 billion users: A way to stay in touch with family and friends, to organize events, stay on top of news and trends, and engage politically.

In some ways, it is even a mirror of life in 2018. But the reflection comes from more than the content of what we post. The ways we are connected with each other through our network of “friends” can reveal insights into how a person lives socially and economically, according to a paper that appears in the summer issue of the Journal of Economic Perspectives.

Authors Michael Bailey, Rachel Cao, Theresa Kuchler, Johannes Stroebel, and Arlene Wong found that Americans with tightly clustered social networks within a short distance of their home tended to have lower incomes, education, and life expectancies than people with more far-flung networks.

 

Figure 3 from Bailey et al. (2018)

 

Figure 3 from their paper shows the relationship between geographic distribution of social networks of a county and county-level socioeconomic measures. Panels A and B show average income and the share of population with no high school education. Counties with a lot of Facebook friends who live close by (within 100 miles) tended to have lower incomes and high school graduation rates than counties with a greater share of friends who live far away.

Of course, correlation is not causation. But the paper still provides a starting point for further research on the social, economic, and political effects of social networks.

The authors created a “Social Connectedness Index (SCI)” based on Facebook friendship links, providing the first comprehensive measure of friendship networks on a national level. They hope the index will help researchers explore topics relating to social networks that go beyond what they examined in this paper.

(Note: The authors invite researchers to submit a 1-page research proposal to sci_data@fb.com to work with these data.)