Estimating dynamic localization economies: the inadvertent success of the specialization index and the location quotient

After addressing definitional issues on the concepts of concentration and specialization, the paper reviews the justifications for and the interpretation of some indicators of localization economies used in the empirical literature on agglomeration economies: specialization indexes and location quotients. A simulation exercise shows under what conditions certain specifications lead to biased estimations of dynamic localization (Marshall–Arrow–Romer – MAR) externalities. The results suggest that applied researchers can choose between the size of the local industry, the specialization index and the location quotient to proxy for these externalities as far as they also encompass a correct proxy for the size of the local economy.