For decades, political scientists have been asking how political jurisdictions are formed and reshaped. Nevertheless, studies of local government jurisdictional formation are few and often plagued with endogeneity since the formation of jurisdictions cannot be separated from sorting effects. In this article, the unique case of the Danish structural reform is utilised to overcome endogeneity due to migration-related sorting by studying patterns of municipal amalgamations. In the recent Danish reform, 239 of 271 municipal entities were forced to amalgamate simultaneously, while who actually amalgamated with whom was left entirely to negotiations between the respective municipalities. Applying logistical regression to a dataset where the unit of analysis is dyads of municipalities allows the construction of a relational model for estimating the effect of different political and societal variables on the likelihood of amalgamation. Societal connectedness, population size and geography are important predictors of amalgamation patterns, while political and economic homogeneity between municipalities does not appear to matter much