Land, what a mess Florida is.

Narrow victory for gay rights in Dade: The fight to preserve a Miami-Dade County law that protects people from discrimination based on their sexual orientation has succeeded, a Herald analysis of uncounted precincts indicates. Leaders of No to Discrimination/SAVE Dade, the group leading the fight against the repeal effort, claimed victory Wednesday, with results available from all but a few precincts. The repeal supporters vowed to mount a petition drive to try again. Anti-repeal leaders took satisfaction in the win, although the margin of victory in Tuesday's referendum vote was less than they hoped -- and substantially lower than a Herald/NBC 6 poll last week predicted. [...] A poll conducted for The Herald and NBC 6 last week showed the repeal being defeated by about 20 percentage points -- a wider margin than the apparent 6 percentage point difference on Tuesday. Political analysts say polls on touchy subjects such as gay rights, immigrants and minority rights often miss the mark. ''On polls like this people often give the socially desireable answer and don't want to admit to somebody they don't know they're a homophobe,'' said Kevin Hill, a political science professor at Florida International University.

Of course the margin was lower. Any polling on anything having to do with sex or race is subject to a "social desirability effect"; people will tell the pollster what they think the pollster wants to hear. This is why, in other places, when polling indicates that nondiscrimination laws will pass, they fail; why, when minorities run for offices they have not previously held and polling indicates they will win easily, they win by small margins or lose outright. In anything having to do with sex or race, you can generally subract 5-10 percentage points off the top, and that will give you a better idea what to expect. I don't know why, after all this time, pollsters don't just admit that. (Mind, a full 15 points is a rather impressive amount of lying, even for this type of topic.)

(As a side note, I would tell the national groups interviewed in that article to stop fooling themselves. The failure of the Miami-Dade referendum will most certainly not serve as a deterrent to other anti-gay ballot efforts nationwide. Why on earth would it? Quite apart from anything else, Florida is demonstrably peculiar. In any event, all politics is local, as the man said. Miami-Dade's referendum would have nothing to do with a similar referendum in Orlando, or in Atlanta, or in Pittsburg, or anywhere else. It's utterly meaningless, and to take encouragement from it is unwarranted. But I digress.)

Apparently, Miami-Dade County doesn't have any law that prevents what amounts to referendum harrassment, so the conservatives can and will just keep getting signatures to shove this back onto each election ballot. Mind, both sides will be subject to some fatigue. It's difficult to get people to keep signing petitions, especially when the people putting the petitions out are known to be corrupt, and especially when the referenda keep failing. And it's difficult to keep trying to mount a defense, raising money and awareness. The question is, who will wear out first.

Miami-Dade anti-gay coalition calls election results fraudulent: One day after Miami-Dade County voters narrowly rejected an effort to repeal a county law that protects gay men and lesbians from discrimination, a leader of the coalition opposed to the law said the results were fraudulent. With nearly all votes counted, some 53 percent of voters supported protections for gays, while 47 percent opposed them. Eladio Jose Armesto, a spokesman for Take Back Miami-Dade, a collection of groups that includes the Christian Coalition, accused county elections officials of rigging the vote by not counting all of the votes to repeal. Armesto said that because his group’s exit polls showed that 67 percent of county voters chose to repeal the 1998 amendment to the county’s human rights ordinance, the county’s tally “should be considered highly suspect and with very little credibility.”

Unfortunately, because of the spectacular mess Florida made of this first major post-2000 election, he has a certain point. Not, to be sure, the one he's making -- it's highly unlikely that tthe vote would be off by 20 percent, really -- but it is virtually certain that the votes for and against have been somehow miscounted, or that people planning to vote on the referendum didn't get the chance or had their votes misapplied. The only question is whether or not it's a difference that made a difference.

In the meantime, in The Big Race:

Two days later, still no clear winner in gubernatorial primary: Two days after voters went to the polls and with 99 percent of precincts counted, Florida’s election meltdown continued to leave the Democratic gubernatorial hopefuls in limbo, with neither willing to declare victory or concede defeat. With only a handful of precincts left to be counted, political newcomer Bill McBride held onto a precarious 11,000-vote lead over Janet Reno. But most of the yet-to-be tallied votes were from South Florida, Reno’s stronghold.

Herald analysis shows McBride victory: ..... McBride's victory seemed assured late Wednesday based on a Herald analysis of outstanding precincts in Miami-Dade and Broward counties. The Herald examined the number of registered Democrats in the precincts that had yet to be counted. Using a turnout rate of 50 percent -- higher than the actual turnouts in either county -- and giving all the votes to Reno, McBride would still win by more than 2,000 votes. [...] Late Wednesday, Reno's campaign left open the possibility that it might file a court challenge in the wake of rampant voting irregularities in the two counties. Campaign strategists also began pointing to alleged problems in the counting of votes in Miami-Dade. For example, one precinct showed 900 percent turnout. Another showed only ballot cast out of 1,637 registered voters.

I'm assuming that given surrounding verbiage, the phrase should read "only one ballot cast..." The story also goes on to note that despite an executive order from Jeb Bush requiring polls to be kept open, some poll workers closed them on time anyway, meaning some people didn't get to vote. The problem is, even if you can prove this sort of irregularity, what can you do? People will be unable to prove that they were kept from voting due to that sort of thing. You can't invalidate the primary completely and hold another; even if it were a reasonable solution, there's really not enough time. Basically, for the second election in a row, many voters will feel that they have been screwed out of their vote by the state. It may or may not matter in November -- somehow, Jeb Bush doesn't seem terribly threatened either by Reno or McBride -- but people are building up one hell of a grudge in that state.

And let's face it: Florida had two years to clear up their problems, and they seem to only have come up with excitingly new and different ones. How likely is it that they'll manage to fix this lot in the two months they have until the gubernatorial elections? It's very probable that Florida will have three major elections in a row with major difficulties ... and it will be very interesting to see what happens after that.