Many claim that the IT revolution has reduced the importance of geographical proximity. People no longer need to meet each other physically, and with the emergence of the new technologies it would be possible to think of a world with no geographical barriers (Green and Ruhleder 1995, Farazmand 1999). Recent evidence has, however, shown that this is not always the case. For instance, using an extensive dataset of 100,000 Facebook users, Goldberg and Levy (2009) find that the volume of email traffic decreases with the geographical distance. Kaltenbrunner et al. (2012) find similar results when analysing data collected from Tuenti, a Spanish social service containing information about social links and messages exchanged. They show indeed that spatial proximity has a great impact on how users establish their connections on online social platforms. They also show that more active users tend to preferentially interact over short-range connections.
New theoretical research
To understand how geographical distance affects social interactions, in our recent paper (Picard et al. 2015) we develop a theoretical model where we consider a population of agents who entertain social interactions and develop social networks in a city. Each agent decides the frequency of her visits (social interactions) to every other agent in the city where the value of each interaction depends on the social network of the visited agents. We define the value of such interactions as the social capital of the agent (Putman 2000). Social capital is thus defined in a recursive fashion; it is higher the larger the volume of interactions with highly social individuals. When deciding how much to interact with others, agents face the following trade-off. Each agent can increase her social capital by entertaining more frequent interactions with agents who also entertain many interactions. However, entertaining social interactions requires costly travelling to the other agents so that spatial centrality may help in building up each agent’s social capital. We show the following results:
- Lower travel costs increase social capital for all agents;
- The intensity of social relationships between two individuals increases with the social capital of each individual and decreases with the geographical distance between these two individuals; and
- Cities that are more ‘spread’ in terms of the geographical distribution of their residents have lower social capital for all agents.
New empirical research
In the empirical section, we mainly test the second prediction of the model, that is, the relationship between the intensity of social interactions and the social capital of the agents and also their geographical distance. For that, we use a unique database on friendship networks from the National Longitudinal Survey of Adolescent Health (AddHealth). It is indeed extremely difficult to find detailed data on social contacts as a function of geographical distance between agents together with information on relevant socio-economic characteristics. Three features of the AddHealth data set are unique and central to our analysis:
- The nomination-based friendship information, which allows us to reconstruct the precise geometry of social contacts;
- The detailed information about the intensity of social interactions between each of two friends in the network; and
- The geo-coded information on residential locations, which allows us to measure the geographical distance between individuals.
Our results show that students residing far away from each other tend to interact less, and more central students in their friendship network tend to contribute more to social capital than less central students. This means that geographical distance is indeed a hinder to social interactions. We also show that similarities in gender or race are greater determinants of friendship interactions.
Our analysis suggests that better urban transport facilities are likely to enhance social capital in cities because it reduces the cost of the geographical distance between agents residing in the same city. These types of policies may be particularly important under the view that social interactions could promote economic growth (Glaeser 2000).
Farazmand, A (1999), “Globalization and public administration”, Public Administration Review 59, 509-522.
Glaeser, E L (2000), “The future of urban economics: Non-market interactions”, Brookings-Wharton Papers on Urban Affairs 1, 101-150.
Goldenberg, J and M Levy (2009), “Distance is not dead: Social interaction and geographical distance in the Internet era”, Computers and Society 2, 1-22.
Green, C, and K Ruhleder (1995), “Globalization, borderless worlds, and the Tower of Babel”, Journal of Organizational Change Management 8, 55-68.
Kaltenbrunner, A, S Scellato, Y Volkovich, D Laniado, D Currie, E J Jutemar, and C Mascolo (2012), “Far from the eyes, close on the web: Impact of geographic distance on online social interactions”, Proceedings of the 2012 ACM workshop on Online Social Networks, pp. 19-24.
Patacchini, E, P M Picard, and Y Zenou (2015), “Urban social structure, social capital and spatial proximity”, CEPR Discussion Paper 10501.
Putnam, R D (2000), Bowling Alone: The Collapse and Revival of American Community, New York: Simon & Schuster.