Localized knowledge spillovers and patent citations: A distance-based approach (revised version)

We develop a new distance-based test of localized knowledge spillovers that embeds the concept of control patents. Using microgeographic data, we identify localization distance for each technology class while allowing for spillovers across geographic units. We revisit the debate by Thompson and Fox-Kean (2005a,b) and Henderson, Jaffe and Trajtenberg (2005) on the existence of localized knowledge spillovers, and find solid evidence supporting localization even when using fine-grained controls. We further relax the assumption of perfect controls, and show that our distance-based test detects localization for the majority of technology classes unless hidden biases induced by imperfect controls are extremely large.