Mattia Fochesato and Samuel Bowles

Paper #: 2017-02-004

Identifying the effect of economic institutions on the distribution of wealth and
income is challenging because exogenous differences or changes in institutions rarely are observed. We investigate the effects of a particular “institution shock” -- the 1865 abolition of slavery throughout the U.S. -- on the distribution of material wealth. The PUMS 1860 and 1870 US wealth censuses allow us to use a difference in difference strategy to identify the effect of the Thirteenth Amendment of the U.S. Constitution (ratified in 1865) on the distribution of (non slave) wealth at a county, state and regional level. (We are currently working on the county level analysis and provide the results soon.) We find that changes in wealth inequality between 1860 and 1870 (as measured by the Gini coefficient) diverged significantly between Confederate states where it declined significantly and Union states where it rose somewhat. The fact that the post abolition decline in wealth inequality was greater where slaves played a more important role in a state’s economy is prima facie evidence consistent with this being an “abolition effect.” But it could have been in part the result of the Confederate states’ defeat in Civil War rather than of the abolition of slavery per se. But this seems unlikely in light of two facts. First, confining attention to states that were not part of the Confederacy (so that this Civil War damage confound is absent) including those with an appreciable number of slaves, we find a similar albeit unsurprisingly smaller “abolition effect”. Second, the post-abolition decline in the Gini coefficient was not greater in the Confederate states than would be predicted based on the extent of slavery in those states and the evidence from slave states that did not join the Confederacy, suggesting that being on the losing side in the war is not the cause of the decline in inequality.