A reader writes:

I haven't read the NR article, and I'm not inclined to subscribe, but Conor's summary cited "nearly 100,000 claimants in the Pigford case," contrasted with a census figure of 18,500 black farmers. I work at a legal administration company, and I can tell you that the number of "claimants" in a class action and the number of approved claims are usually very different numbers.  We could get a blank, unsigned form in the mail, with no accompanying documentation, and we'll make high-res scans of everything and (if the return address is legible) that person becomes a claimant in our database.  That doesn't mean they get a check. Again, I don't know if this is pertinent to the controversy, but if it does indeed hinge on the use of the word "claimants" then it's a red herring.

Says another:

It's wrong to use the number of claimants as an indication of fraud, and I saw that as a clerk to a federal judge who doesn't like seeing this kind of thing propagated.  Class actions do dragnets to gather up all potential claimants so they can be sifted through and processed in one go.  You don't want people weeding themselves out and then realizing later they actually were entitled to something once the settlement fund has already been distributed.  So, you cast a wide net and get everyone to submit up front everything that could possibly fit, and then you weed it out in the next steps of the process.  Consequently, first-step advertisements are typically worded very broadly and include a lot of "if you don't put in your claim now, you'll lose it" language.  You've likely seen some of them on TV advertising for drug class actions, telling anyone who took a common painkiller and had a common disease or heart attack to submit a claim.

That process leads to invalid claims being submitted, and that's by design, not necessarily because of fraud.  Folks often genuinely believe they may be entitled to compensation, and anyway it's better to be safe than sorry, so they're encouraged to submit their claim and then the work of the case is to sift through and see who belongs and who doesn't.  That's the way the process is supposed to work - it only yields fraud if there's fraud, i.e., lawyers coaching their clients to lie or the administrators ignoring their duty to weed out baseless claims.  There's no evidence of that here - especially given that past filing windows have weeded out so many claimants - and you certainly can't deduce it from the two numbers that you mentioned.

And a third:

The Congressional Research Service looked at just this issue and has a reasonable explanation (PDF) for the gap: at p. 8. The Census was enumerating "farms" in 1987, not "farm operators." This goes a long way toward closing the gap. Add to this the number of people who tried to enter farming but could not because the discrimination underlying the settlement prevented them from getting loans, and you've closed even more of it. Finally, the 100,000 number is the number of claimants, not the number of actual award to successful claimants. Only 69% of the original claimants under the original consent decree actually received payments.

Ta-Nehisi, Adam Serwer, and Cynic analyze the anti-Pigford campaign.