I'll do the first one for now; the second one is quite complicated and I don't feel up to it this morning. Start with
this column by Krugman. He goes through the usual gyrations to prove that Gore actually won Florida in 2000. Then he talks about Ohio 2004:
That would certainly be improbable, and so
I went looking for information on Miami County, Ohio's 2004 results.
And in actuality the total turnout for the entire county was 72.23%.
Others, including fellow JREFer JamesB,
found more mistakes in Krugman's column, regarding Florida 2000. Krugman wrote:
James B replied, using Krugman's own paper to back him up.
Krugman wrote
an entire column the following Monday, trying to explain what his previous column had meant, while still claiming he was right. In newspaper terms, this is known as a rowback. He reiterates his earlier claim, but this time notes:
JamesB
noted that the definitive statement "would have given the election to Mr Gore" had now been watered down to "probable". He also notes that Krugman continues to claim that the recount that probably would have given the election to Gore "should have been". Of course, the fact that Mr Gore himself did not demand that particular recount, nor did the Supreme Court, or Florida election law.
Bizarrely, Krugman did not take that opportunity to correct the Miami County mistake, even though that was transparent. It was not
until the following Friday that he acknowledged that error:
Note as well that after much emailing back and forth, Calame himself was forced to write a further correction at the end of that column (Krugman himself petulantly refused to do so):