Matching and network effects∗

Individuals form teams to produce an output. The quality and quantity of a team’s output depends on the ability and the effort level of its members. The ability of individuals as well as their work ethic is however imperfectly known. In such an environment, individuals wishing to form a new team can get information about each other from the network of past collaborators. The paper carries out an empirical examination of the existence and magnitude of this network effect. This empirical work looks at the formation of coauthor relations among economists over a twenty year period. Our principal finding is that a new collaboration emerges faster among two researchers if they are “closer” in the existing coauthor network among economists. This proximity effect on collaboration is strong and robust but only affects new collaboration: it has no effect on subsequent collaboration among the two individuals. Thus the coauthor network acts as a “referral” mechanism among economists. ∗A previous version of this paper was circulated under the title, Scientific Networks and Coauthorship. We thank Michele Belot, Jordi Blanes-i-Vidal, Sebi Buhai, Jean Ensminger, Joseph Harrington, Vernon Henderson, Matthew Jackson, Jeff Johnson, Markus Möbius, Tom Snijders, Manuel Trajtenberg, Fernando Vega-Redondo, and seminar participants at Alicante, Bristol, Harvard, Oxford, Utrecht and Saint-Etienne for useful comments. Marco would like to thank the Vereniging Trustfonds Erasmus Universiteit Rotterdam for supporting a visit to Oxford University in May 2005. †Department of Economics, University of Oxford. Email: [email protected]. ‡Faculty of Economics, University of Cambridge. E-mail: [email protected] §Tinbergen Institute, Erasmus University Rotterdam. E-mail: [email protected]