Second, this paper contributes to the large literature that studies the benefits of agglomeration. Ellison and Glaeser (1997), among others, show that there is significant geographic clusteringof industries, supporting various theories of agglomeration. To date, the empirical literature onagglomeration spillovers has mostly focused on the expansion of agglomeration economies throughentry decisions (e.g., Rosenthal and Strange (2003); Greenstone et al. (2010)).6 In contrast, ourstudy examines the disruption of agglomeration economies. By focusing on disruptions that breakagglomeration linkages through liquidation, and through the reliance on random variation in theassignment of bankruptcy judges, we show how agglomerations can propagate negative shocks thatimpose negative externalities on other firms within the cluster.7 Further, our detailed micro-leveldata and identification scheme allow us to examine more closely the various channels of agglomeration spillovers across a wide range of industries and with more precise geographic locations. Thisis in contrast to most of the literature on agglomeration that focuses on the manufacturing sectoronly, which is typically found disproportionately in small- and medium-size cities, as well as theirrural fringes, rather than in dense cities (Kolko (2000)).