Spillovers in Networks of User Generated Content - Evidence from 23 Natural Experiments on Wikipedia
Abstract: How do networks generate externalities, such as spillovers or peer effects? Quantifying these externalities is challenging due to the endogeneity in network formation. I tackle this problem by exploiting local exogenous shocks on a small number of nodes in the network and investigate spillovers of attention on the German Wikipedia. I show how the link network between articles influences the attention that articles receive and how the additional attention is converted into content. Exogenous variation is generated by natural and technical disasters or by articles being advertised on the German Wikipedia’s start page. The effects on neighboring pages are substantial: They generate an increase in views of almost 100 percent and content generation is affected similarly. Aggregated over all neighbors, a view on a treated article converts one for one into a view on a neighboring article. My approach applies even if, absent network data, identification through partial overlaps in the network structure fails. It thus helps to bridge the gap between the experimental and social network literatures on peer effects.
Bio: Michael Kummer focuses on the production of User Generated Content in Networks, the structure of online markets as well as the interaction of firms and consumers online. He studied economics at the University of Vienna and obtained his master's degree in mathematical economics from Toulouse School of Economics. Before joining Mannheim's Centre for European Economic Research (ZEW) in September 2009, he was a research assistant at the Johannes Kepler University, Linz. He expects to defend his PhD-Thesis at University of Mannheim this summer.
Dana Research Center, 110 Forsyth St., Boston, MA 02115 on the 5th Fl – Large Elevator