Update: This post previously included an table with incorrect figures, which I have now replaced. Sorry about that.
Well into the eighteenth century, ships at sea had a great deal of trouble ascertaining their locations. For centuries, mariners knew how to reckon their latitude. They used sextants to measure the angle of the sun. Longitude did not give up its secrets so easily, but a combination of clocks that remained accurate at sea and painstaking astronomical almanacs finally yielded them up. They called this the Longitude Problem. Dava Sobel wrote a wonderful book about it, titled simply Longitude.
But I told you that story to tell you this one:
When I started Slavery by the Numbers, I thought that I could build a statistical picture of slavery both on the eve of war and down through time out of the census aggregates available online. Further digging left me with scans of the original tabulations for many years as well. Unfortunately, much of the original tabulations for the earlier censuses vanished sometime before 1830. Thus ended my hope to eventually create parallel listings for each state just the same way as I did for the 1860 census and present them as apples-to-apples comparisons. I especially wanted to look at the size of the slaveholding class over time, but that data did not survive. For example, I can query the 1790 census for slaveholding families, but half the states and territories listed have no data. Unlike eighteenth century mariners, I would need a time machine to go back and get that data. Much of it probably went up in smoke when the British burned Washington. Some future censuses met the same fate when they caught fire in the basement of the Commerce Department in the 1920s. I would also love it if more robust colonial censuses existed, but of course no national authority required a uniform, decennial census before 1790.
But frustrated by the lack of data I wanted and unsure how to present the other data I do have from later censuses, I put Slavery by the Numbers up and busied myself with the Road to War, Samuel Cartwright, and lately the Crafts. I wanted to get back to it, but desire never translated to action. About a week ago my blogfather asked about the series and I told him my main problem with the extant data: with so much, how to organize and present it?
“How about you do posts on individual topics instead of trying to cram it all into one big thing?”
I admit, I had not thought of that. In my defense, I don’t know how I could see that forest with all the trees in the way. As I’ve said before, these aggregates all involve certain assumptions that cannot hold in every case and do not resolve down to pinpoint the details of individuals.
Rolling all the data together into a nationwide aggregate obscures the fact that many states we call free had considerable numbers of slaves for decades after their various acts of emancipation. Slaves also traveled with their owners on personal business and followed them to military postings. For some years the census may catch some of them as well as the residual slaves that gradual emancipation schemes left behind. Without the microdata, I cannot easily separate residual slavery from slaves in transit or sojourning with owners on census day. But the data available does allow the tracing of slavery’s decline in the North and reveal some surprising side trips it took along the way.
I plan future posts dealing with those topics, both on their own and comparatively to tease out the many nuances and textures of the times. The nation did not always neatly divide at the Mason-Dixon Line and the numbers help tell those stories.