"If a person interferes with an election unjustly, they have no right of privacy as to who they voted for," [Rep. Will] Hartnett said. "Historically, I don't believe they have been forced, but that doesn't bind me. I have the ability to compel illegal voters to divulge who they voted for."

You make an excellent point, but I suspect the courts will end up sorting that out. But even leaving that aside (as well as the question of how someone could interfere with an election "justly"), there's another obvious problem with "compel[ling] illegal voters to divulge who they voted for." How do you know whether they're telling the truth, given our voting system is set up to prevent tying any voter to his/her ballot after the fact?

Say you find some "illegal voters" who support Heflin. Presumably, they voted for him, but what's to stop them from telling Hartnett they voted for Vo, in order to give Hartnett an excuse to order a new election?

The reverse could also be true, of course. Any illegal Vo voters could claim they voted for Heflin in order to undermine his case. Sure, it's perjury, but so what? It could never be proven.

I don't think Hartnett has given his position much thought. Probably typical for a GOoPer, though.

Like Greg, I note that three of the four people quoted in this Chron story about subpoenas being sent to testify before the State House in the Heflin contest voted for the Republican. I think that ought to put a dent in the crazy idea that there was some kind of nefarious Democratic plot to steal that election.

I would think anyone with any knowledge of statistics wouldn't take that small sample quoted by a reporter to be indicative of much beyond the fact they were quotes from a small sample. :)

Leaving aside the hardcore partisans who obviously hope to win this however they might, I don't think most sensible people are talking about nefarious Democratic plots to steal elections. I'm certainly not.

But, if Heflin is determined to pursue this losing strategy to the bitter losing end, I think we can all agree that shining some sunshine on this registration/voting process of ours is at least a side benefit. Bettencourt's already admitted one person got a voter card but shouldn't have, so the sunshine has already done some good, no?

I would think anyone with any knowledge of statistics wouldn't take that small sample quoted by a reporter to be indicative of much beyond the fact they were quotes from a small sample. :)

That's right, Hubert won on a sample size of 42,000. Votes have been counted, recounted, voters subpoenaed, and still, El-Heflin lost. I mean seriously, what the hell is going on? How many times does Vo have to win this election?

I'm all for sunshine on the process: please direct your vision of enlightenment on Paul "not a citizen, well here's a voter card" Bettencourt and Beverly "here's a letter from the Secretary of State saying we suck" Kaufman. I hope this rays will have you voting against these folks for others that will do their jobs better and with a better since of Math.

When did the GOP quit believing in Math? I guess with W's budget, math is a logic our elected officials just don't like.

I would think anyone with any knowledge of statistics wouldn't take that small sample quoted by a reporter to be indicative of much beyond the fact they were quotes from a small sample. :)

Heflin needs the illegal votes to put Vo over the top to win his challenge. If there were 167 illegal votes, as alleged, then they'd have to have broken at least 100-67 toward Vo for Heflin to have "really" won.

If we choose that as our hypothesis, the odds of getting a pro-GOP sample of four were nearly 18%. That's large enough that it's plausible that the Vo side was just lucky with this small sample; but if Heflin is right, the odds of getting a pro-Vo sample of four were about 48%, and the odds of getting even a neutral sample were over 34%.

So Kevin's right that it's too small a sample to give a definitive result, but I think Kuffner's statement that it ought to put a "dent" in Heflin's theory is still justified - at least if these four are all being truthful.

Of course, as I've already commented, partisans on both sides have a strong incentive to lie, and virtually no chance of being caught, so perhaps we shouldn't take this testimony too seriously no matter how large the sample becomes.

As for whether this cloud will have a silver lining, as Kevin hopes, I suppose time will tell. But I fear the most likely reaction to a tiny handful of illegally registered voters in an unusually close race will be to erect barriers which will make it much harder for legitimate voters to get registered.

In other words, the legislature will probably overreact, with the result being more legal voters disenfranchised than illegal ones rightfully removed from the rolls. If so, I wouldn't consider that a good thing at all.

Of course, since the voters most likely to be affected are the sons and daughters of non-citizens, who still tend to vote Democratic, I suspect Kevin is less concerned about it than I am.