["\n\nObjective\nIt is widely argued that a primary source of legislative deadlock in America is the combination of a secular increase in polarization, combined with constitutional provisions that divide law‐making power across branches. We argue that polarization affects productivity, but only given a particular pattern of divided government. We distinguish between split branches, where a president from one party faces a Congress controlled by the other, and split chambers, where each party controls one house of Congress.\n\n\nMethods\nMultivariate analysis of enactment data from post‐War Congresses, augmented by data on House and Senate Uncovered Sets.\n\n\nResults\nEnactments of major legislation are less likely given split chambers compared to the other options and polarization has no impact after controlling for these factors.\n\n\nConclusion\nThese results redefine the conditions under which polarization drives deadlock. They also explain why the increase in polarization over the last two decades has until recently had little impact on major enactments.\n\n", "Social Science Quarterly, Volume 101, Issue 5, Page 1712-1727, September 2020. "]