Twitter has a reputation for linking people through interests rather than geography. But while the little blue bird lets us connect with people all around the world, the networks we form on Twitter look a lot like the airplane trips we’ve been taking for decades.
After analyzing more than 480,000 tweets, researchers at the University of Toronto and Dalhousie University found that most Twitter connections are local.* Those that aren’t local are strongly based on how often planes fly between two places. In fact, according to the report (PDF), the best way to predict long-distance Twitter connections is to look at air traffic between cities.
The first part is hardly shocking. Even if your main interest on Twitter is, say, emerging technologies and their implications for society and policy, you’ll still probably follow a lot of friends, coworkers, and other locals with shared interests in your community. It’s the second discovery that really challenges the conventional wisdom that Twitter is connecting the world in new ways.
For example, researcher Barry Wellman found that someone in Los Angeles is much more likely to be connected to someone in Toronto than someone in St. Louis is. Similarly, people in New York City are more likely to have ties with people in large cities like Los Angeles, London, São Paulo, and Tokyo, where air traffic between the cities is significantly busier than many other places around the world. So while it’s entirely possible for someone in a small town to reach a global audience, geography still plays a big role in how a tweet moves around the world.
As researchers say in the report, this could be because flights facilitate more face-to-face interactions that turn into Twitter connections, or it could be because the flight patterns themselves reflect how connected people in one city are to people in another. But either way, the study suggests an important point. Rather than building new global networks of people and places, Twitter users tend to follow pre-existing ties around the world.
Hear more from NPR.
*Correction, Feb. 22, 2011: This sentence originally suggested that all of the project researchers are from the University of Toronto. One is affiliated with Dalhousie University.
