[So Cal Rocket]

Actually there was one group in Florida 2K who had their ballots invalidated, and obviously not by "mistake," since they had a spoilage rate Fifty Times higher than than others in their precincts: Black Republicans

Thanks for the link... I had a question in my mind when I went to that thread, expecting to find the answer there - but it wasn't - so I'll ask it here:

How can you tell if the ballot that was spoiled was from a black Republican vs. a black Democrat vs. a White Republican, etc? Don't we vote using secret ballots?

[Stultis]

Unfortunately, as you note, Glassman and Lott do not explain this, nor can I find where they have published a more detailed exposition of their analysis, except that there is a slightly longer version of the same op-ed at AEI.

Whereas the shorter version (as published in the Los Angeles Times) simply said, "African American Republicans who voted in Florida were in excess of 50 times more likely than the average African American to have had a ballot declared invalid because it was spoiled," the AEI version says, "fifty-four to sixty-six times".

It would appear from this that they did some kind of analysis of the media recount ballot marking database that gave them these upper and lower bounds. I'd guess they inferred race from precinct percentages, and inferred party affiliation from down ballot votes.