"Free persons" would be those persons not counted as slaves. It may include non-Whites.
"Slave-holders" would include anyone who held one or more slaves to service.
You are not asked to accept anything. These are census figures. I did not create them or modify them. The census recorded what data it chose to record and that is all we have.
Probably you own a motor vehicle. Are your children motor vehicle owners?
Julia Dent Grant owned slaves. Did that make Ulysses S. Grant a slaveowner?
According to Allen C. Guelzo, Lincoln's Emancipation Proclamation, p. 23, "from 1850 until 1862, Lincoln and his wife, Mary Todd Lincoln, were embroiled in litigation in Kentucky over the settlement of the estate of Mary's father, litigation that netted the Lincolns a share in the proceeds of selling the Todd family slaves." When Mary Todd Lincoln's father died, she obtained property interest in the family slaves. Did this make Abraham Lincoln a slaveowner?
[r9etb] Or are we defining "slave-holders" as all members of a family that owned slaves?
There is insufficient data to make any valid determination of the number of family members in addition to each slave-holder. The census counted individual slave-holders. That is the data that is available.
[r9etb] You'll agree that it's an important distinction, as the relative percentages are tightly bound to the definitions used to produce the figures you cited.
I would disagree that the "slave-holder" figure necessarily refers to those who held actual title to the slaves. For example, Massachusetts Congressman Chaffee and his wife, Irene Sanford Emerson Chaffee held actual title to Dred Scott and his wife and two children. Whomever held Dred Scott and his family to service in Missouri may well have been counted as a slave-holder, as distinguished from the holder of the actual title.
[r9etb] This does not alter the fact, however, that in some sense the slave-holders' families were in some sense also slave-owners, who gained from the property held by the head of their family.
In much the same sense that your infant child would be considered a motor vehicle owner and a homeowner.
The census recorded no such statistic. I should like to remind that the issue to which I have dissented is the claim that "academically-accepted evidence" had been (past tense) presented to show that a claimed percentage of Southern families were slaveowners. There is no such census statistic. There is insufficient data to accurately derive such a statistic.
[r9etb] In this latter case, if we assume an average family size of four (a number I'm pulling out of thin air), then Virginia's 52,128 slave-holders expands to roughly 200,000 slave-owners, which places the percentage closer to 20% of the total population.
The average family size for the general population was approximately 5.6. This does not establish that a subset of the general population, comprised of privileged elite wealthy slaveholders had an average family size of 5.6. Their family size may well have been 4.0. There is no statistical data upon which to accurately determine this figure.
In simply multiplying by (arbitrary) family size, the resulting figure purports that infant children are slaveholders. If we were to eliminate minor children, the number would shrink markedly. It might well be that an adult child was given ownership of a slave upon reaching the age of majority. There is simply no data recorded to inform us of how many adult children of slaveholders were also slaveholders. (General Grant's father-in-law apparently gave him a slave.)
Multiplying the number of slave-holders by 4 yields 4 times the number of slave-holders. Multiplying the number of apples by 4 does not yield the number of oranges.
[r9etb] Only if we define "slave-holder" as being all members of a slave-owning family, do the percentages cited make sense.
No. Defining all members of a slave-owning family as slave-holders makes sense only to the extent that a one-month old infant girl should be counted as a slave-owner. Assuming you own your own home, it would not be sensible to count your one-month old daughter as a home-owner.
Virginia had 1,105,453 free persons. It had 52,128 slaveholders. And it had 4.7% slaveholders as a per cent of free persons. Divide 52,128 by 1,105,453 and you get 0.0471553 or 4.7%.
The census data table clearly states what the percentage statistic is. It represents the percentage of free persons (individuals) who were slaveholders.
The census recorded individual slaveholders. It recorded the total number of free persons. It did not record the total number of persons per slaveholding family. It did not record how many families contained more than one slaveholder.
[r9etb] So ... how are we to define slave-holder?
The census data clearly considers it to be an individual who holds slaves to service.
However, you do seem to be saying that the number of slave-holders is counted differently from the number of "free persons." Specifically, you appear to agree that one could be a member of a slave-holder's family and still not be counted as a "slave-holder."
It appears that the percentages you posted do not mean what they're purported to mean. Your example of motor vehicle ownership is a good one: I hold title to my vehicle, but my children also "own" it, in the sense that it's a "family possession." It's obviously "mine" within the context of my family, but it's also obviously "theirs," as seen from outside my family.
For the percentages to be a correct and factual statement of slave ownership, the proper comparison would be between "slave-holders" and "land-owners," or something else that is equivalent to the idea of holding individual title to property. Instead, your source apparently compared all free persons (including minors) to individual title-holders.
Bottom line: the good professor was lying with statistics.