To be clear he said correctional control. That includes those on probation and parole.

tl;dr this claim is based on a quote by the author of "the new jim crow" and her broader definition of terms is technically accurate.

an older source for the same claim:

http://www.huffingtonpost.com/2011/10/12/michelle-alexander-more-black-men-in-prison-slaves-1850_n_1007368.html

Others have pointed out the flaws out this but Michelle Alexander is considered a reputable source so i doubt she's out and out lying so perhaps we've got this wrong.

so i took another look and realized that she's not actually talking about "in jail" what she's saying is

More Black men are in prison or jail, on probation or parole than were enslaved in 1850,

Thus we need to measure the paroled

http://www.pewtrusts.org/en/about/news-room/press-releases/0001/01/01/one-in-31-us-adults-are-behind-bars-on-parole-or-probation

so the population is 7.3 million (1-31) so back of the envelope math means that if 44% of that population is black the modified claim is technically accurate.

now from the pew article

Correctional control rates are highly concentrated by race and geography: 1 in 11 black adults (9.2 percent)

Using wikipedia (lazy version of looking at 2010 census) gives 38 929 319 as the black population so the "super lazy" incarcerated number is 3,581,497.348 (one person got 65% murdered by the census) and

3,581,487 under the thumb of the judiciary>3.2 million but we're not done yet

as the "super lazy" indicates this is problematic because it's talking about adults and presumably kids have less but also both numbers include women while alexander's statement has 0 women in it. so we need to cut the slave population by about half (# of female slaves) and by about 20% for women (based on same pew poll and assuming no difference between the ratio of male/female incarceration rates by races) and by another 25% for the fact pew data excluded all under 18 (census data) but then we need to add the number of black children/people under 18 under the state's thumb. Both assumptions are flawed but it doesn't matter because the margins of error don't overlap and thus # slaves will be less than the other number though it could be somewhat close.

I'm sure others like Alexander have gone the route of pure data and gotten a precise figure but i think this basic runthrough shows at the very least that the claim is very likely to be true and thus we should probably assume it is true unless someone can dig out some crazy data. A more interesting question is also what does this mean considering population growth (1850 population was 23 million today it's 315-320 million).

http://www.naacp.org/pages/criminal-justice-fact-sheet

2.3 million in prison. 1 million black.

1850 http://en.m.wikipedia.org/wiki/1850_United_States_Census

3.2 million slaves, all black.

Doesn't seem to equate.