The Army of Northern Virginia and Slavery: By the Numbers

The Confederacy’s latter-day partisans have no shortage of arguments, making up for the dearth of quality with a surfeit of quantity. One must use the tools one has. I’ve taken a swipe or two before at the idea that ordinary soldiers had no stake in slavery and therefore the Civil War and the Confederacy had nothing to do with it, as well as its slightly more plausible variant that we should not operate under a presumption of proslavery intent in understanding military service with the Confederacy. I think the case against the proposition that the average men and occasional woman in a gray or “gray” uniform doesn’t require much further development and planned to leave it be.

Robert E. Lee, Virginia aristocrat, military officer, and future confederate general

Robert E. Lee

Plans changed this week when I remembered Joseph Glatthaar’s statistical study, Soldiering in the Army of Northern Virginia. It concerns the Confederacy’s principal field army. For most of its existence, Robert E. Lee had command of the force. The ANV fought in all of the battles most laypeople have heard of, against the familiar rotating cast of United States generals that ended with Ulysses S. Grant. I took an interest in Glathaar’s study when it first came out, but flinched at the price tag and its distance from my usual interests. I don’t mind straight military history, but have a much stronger interest in the politics that produce it. Likewise my interests have skewed rather more than I anticipated when I started this blog toward the Antebellum. When it fell off my radar, I hadn’t read an ebook and didn’t own a Kindle. Now I do and the digital version comes with a very reasonable price.

Before I get into the findings themselves, Glatthaar’s method deserves some explanation. Using existing records, he developed a random sample of 600 soldiers. The sample took in infantry, cavalry, and artillery in proportion to their numbers in the army and represents officers and enlisted men similarly. It does not attempt to achieve the same balance with regard to the home states of the soldiers, though it does include men from all eleven Confederate states plus Kentucky and Maryland. The most in the sample hailed from Virginia (239), followed by North Carolina (96) and Georgia (86). Arkansas, Florida, Kentucky, and Texas tied for last at four each.

According to Glatthaar,

Just under half (46.7%) of all soldiers in Lee’s army were born in Virginia or North Carolina. With South Carolina and Georgia added, 3 of every 4 (75.1%) troops came from those Southern coastal states. One in every 13 (7.8%) was born in the North (a state that remained in the Union) or in a foreign country. Those numbers included young Private Bishop, the son of a fisherman, who originally hailed from New York and moved with his family to South Carolina.

Glatthaar, Joseph T. (2011-06-15). Soldiering in the Army of Northern Virginia: A Statistical Portrait of the Troops Who Served under Robert E. Lee (Civil War America) (p. 4). The University of North Carolina Press. Kindle Edition.

I cannot claim any special knowledge of statistics, but this sounds like about what one would expect for a fair, random sample from which we can confidently generalize about the ANV. Glatthaar also notes that 55.0% of the men resided in the Upper South, so one can’t claim he cherry picked a sample from the Cotton Kingdom’s black belts and then shockingly found them especially involved with slavery.

Right then, we’ve got a decent enough sample. What did Glatthaar find out about the men of the Confederacy’s preeminent army?

Soldiers were more likely to come from heavier slaveholding counties than the recruiting states as a whole. […] Their home counties on average had 16.6% more slaves to whites than the average of all the counties in those states.

Glatthaar, Joseph T. (2011-06-15). Soldiering in the Army of Northern Virginia: A Statistical Portrait of the Troops Who Served under Robert E. Lee (Civil War America) (pp. 5-6). The University of North Carolina Press. Kindle Edition.

One would expect more enslaved counties to show up more prominently in the rolls of an army defending slavery in a nation created for that purpose. A persistent person might argue that residence in a highly-enslaved county doesn’t necessarily make you more likely to favor slavery. The argument doesn’t make much sense considering the centrality of slavery in the South as a whole, let alone in its more enslaved than average counties where human property would have a more prominent and pervasive role still.

We could stop here and content ourselves with a data point in favor of an already well-supported position, but Glatthaar had more data still. Here we get into the real meat of things. It turns out that not only did men from unusually enslaved counties, by the standards of their own states, appear more frequently in Lee’s army. Men from slaveholding households did as well:

According to the 1860 census, 1 in every 20 (4.9%) adults owned slaves and 1 in every 4 (24.9%) households had slaves. In Lee’s army, more than 1 in every 8 (13.0%) soldiers owned slaves, and for those who lived with family members, approximately 3 in every 8 (37.2%) had slaves. Four of every 9 (44.4%) troops resided in a slaveholding household, some 78.0% greater than the South as a whole.

Glatthaar, Joseph T. (2011-06-15). Soldiering in the Army of Northern Virginia: A Statistical Portrait of the Troops Who Served under Robert E. Lee (Civil War America) (p. 9). The University of North Carolina Press. Kindle Edition.

Near to half of all men in Lee’s army lived in a slaveholding household. They grew up intimately acquainted with and materially benefiting from the stolen labor of black Americans. Furthermore, that number far exceeds the typical proportion of slaveholding families in the South.

Glatthaar doesn’t provide breakdowns by state for context, but I have them from my past work with the 1860 census. If recast as a state, Lee’s army would have had a greater percentage of enslaver households than any state of the Border or Upper South by a large margin. North Carolina, the most slaveholding among those states, topped out with 27.71% of households owning at least one person. It would even beat the Lower South’s average (37.01%), coming in between South Carolina’s 45.53% and Georgia’s 37.38%. This would make the State of Lee the South’s fourth most enslaving.

The Deep South in the 1860 census. (Click to enlarge.)

The Deep South in the 1860 census. (Click to enlarge.)

I’ve seen the complaint that Glatthaar went through a tremendous geneology project, pinning the slaveholding of fifth cousins, twice removed and essentially a strangers on some poor solider out of pure malice. Those who want to believe such things can, but Glatthaar used the United States census. It lacks any such remote information. The census takers organized their data by household. The parlance of the time called everyone who lived under the same roof or on the same property a family, even inmates at insane asylums and boarding houses where everyone understood no blood relation need exist.

Nor did Glatthaar cherry pick the wealthiest soldiers about, counting on the fact that wealth meant slaves in the antebellum South to make his point. Slaveholders, including the wealthy ones, do appear somewhat more prominently, but in measures of personal and family wealth the plurality of soldiers still could claim no more than $400 (35.8%). Another 5.9% came in below $800. By period terms, this made them poor. The middle class, between there and $4,000 accounted for another 22.8% of the ANV. The wealthy made up the remaining 35.4%. This creates a substantial gap in the middle, but the very wealthy would include large slaveholders who one would expect to have a stronger enthusiasm for the cause:

Approximately 92% of all soldiers’ households with a minimum total wealth of $ 4,000 possessed slaves. More than 1 in every 15 soldiers or his family (6.9%) achieved planter status— owning 20 or more slaves— and 1 in 11 soldiers (9.3%) resided in planter households. By contrast, 1 in 32 (3.2%) households in the South qualified as a planter. This was not, therefore, a rich man’s war and a poor man’s fight. Slaveholders, who also happened to be rich, served in disproportionately high numbers in Lee’s army. It was a rich, moderate, and poor man’s fight.

Glatthaar, Joseph T. (2011-06-15). Soldiering in the Army of Northern Virginia: A Statistical Portrait of the Troops Who Served under Robert E. Lee (Civil War America) (pp. 9-10). The University of North Carolina Press. Kindle Edition.

I know none of this can persuade those who have convinced themselves that the ordinary soldier had no interest in slavery. If the documentary record and bare census figures can’t do the job, then one more study never would. But for the rest of us, the numbers clearly show not just an increased interest in slavery for Lee’s army, in every way one would think to look, but one radically higher than coincidence or mere statistical noise could ever account for. They also, I must add, exceeded my own already generous expectations. I imagined thirty to forty percent more slaveholding households than the Southern norm, not nigh eighty.

 

Advertisements

4 comments on “The Army of Northern Virginia and Slavery: By the Numbers

  1. Andy Hall says:

    I’ve seen the complaint that Glatthaar went through a tremendous genealogy project, pinning the slaveholding of fifth cousins, twice removed and essentially a strangers on some poor solider out of pure malice. Those who want to believe such things can, but Glatthaar used the United States census. It lacks any such remote information. The census takers organized their data by household. The parlance of the time called everyone who lived under the same roof or on the same property a family, even inmates at insane asylums and boarding houses where everyone understood no blood relation need exist.

    ____

    Exactly so. The 1860 Census used the term “family,” while the modern Census used “household,” but they mean exactly the same thing. Those who assert that Glatthaar was pulling slaveholders from all over the family tree of the soldiers he researched either (1) have never done any actual research with the census rolls, to understand its limitations, or (2) are simply disingenuous partisans counting on the inexperience of their audience to bluff their way through.

  2. Jimmy Dick says:

    As a researcher with training in both quantitative and qualitative research methods, I have found that many research studies involving statistical data have flaws in them which derive from the way the study is constructed. Having looked over the information involving Glaathaar’s study from a research standpoint, he constructed a very solid and well planned study using a sophisticated methodology. One of the goals of quantitative research is to develop a generalizable result which can be expanded upon for additional research and more importantly, help in working with qualitative research.

    Glaathaar could do a study using every single soldier who ever fought in the ANV and the detractors would still complain that it wasn’t fair, that it was inaccurate, and that it was biased. They will never be happy with the results of any research unless that study told them what they want to hear. The fact that multiple research studies have shown results that do not fit into their belief structures just makes them insist that their beliefs are right and solidifies their self-delusions.

    It all comes down to one thing. Slavery was the cause of the Civil War. That’s it. Until that fact is accepted by the heritage crowd, they will continue to insist that their beliefs are correct. The reason for that is once anyone admits slavery was the cause of the Civil War, they are accepting facts instead of beliefs. That leads them to question other parts of the mythology and the house of cards begins to crumble.

    What this study does for me is to promote my own qualitative research work in examining why men fought in this war and why people had the viewpoints they held. I’m working on this is some detail here in Northeast Missouri and it is interesting how fluid people’s opinions were. They changed with the ebb and flow of the conflict over time. Recent research has strongly indicated this was common all across the nation on both sides of the struggle. More and more I am seeing a very strong connection with people indicating that “the way of life” they were accustomed to played an important role in their perceptions. Those perceptions could and did change and that I think may have been a integral factor in why the war actually started.

    The perceptions of people were challenged with secession and the move to armed conflict. They began to adjust them as they grappled with the very real possibility of war over slavery. That led to an examination of conscience and reality for many people. The leaders of the Confederacy saw their initial support based on emotional responses slipping away and had to take action to maintain that support. As we know from history, that turned out to be a really bad mistake.

    • Thank you, Jimmy. Glatthaar’s work looked rigorous to me, but I’m lucky when I can do an average right twice in a row.

      I meant to say that last week, but just now realized that I’d forgotten. It’s a poor way to thank you for sharing your knowledge, but thanks all the same.

Your input is welcome

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s