Abstract

The legacies of slavery have shaped nearly all aspects of American politics. Acharya, Blackwell, and Sen's Deep Roots: How Slavery Shapes Southern Politics deploys sophisticated methods of causal inference to empirically identify one of these legacies: the enduring impact that slavery has had on white southerners' racial attitudes. An important part of their causal argument is the role of the Civil War and Reconstruction, which they argue was the critical juncture when the distribution of white racial attitudes in the South began to diverge based on the prior pervasiveness of slavery. Before the secession crisis, they claim, white racial attitudes were unrelated to the level of enslavement within the South. We reconsider the evidence for this antebellum "homogeneity in racial attitudes" claim, which is critical to Acharya et al.'s strategy for identifying the causal effect of slavery and for their broader historical argument. Using multiple sources of data from different moments in southern history, we find that the pervasiveness of slavery was a systematically strong predictor of voting on slavery, secession, and the rights extended to free persons of color well-before the Civil War. Our findings suggest that there was no discrete moment at which the connection between the geographic pervasiveness of slavery and revealed commitments to white supremacy was established, raising questions about the causal identification strategies used by the authors. More generally, the findings point to some unappreciated limitations of design-based approaches when assignment to treatment is the product of a slow-moving complex historical process.