The promoters of Instant Runoff Voting (IRV, also known as Ranked Choice Voting) hope it will eventually be used for the big statewide races in Minnesota. So far, they've beaten back court challenges, demonstrated that it can work in last year's Minneapolis election (although it wasn't too stiff a test) and will try to do the same in St. Paul on the next round there.

Looking at a poll out of Texas this morning, I thought that Texans would be fortunate to have IRV this year, but there's a catch.

Texas is one of those states that requires 50 percent to win. If no one reaches that mark in the first round, there's an automatic runoff between the top two finishers. Republican Gov. Rick Perry was supposed to be facing a tough reelection challenge this year when U.S. Sen. Kay Bailey Hutchison decided to give up her Senate seat to try to become governor. It hasn't worked out for Hutchison and the latest polling shows Perry with 45 percent to 21 for Hutchison and 19 percent for a Ron Paul disciple named Debra Medina. The last few polls have showed this pattern, with Perry hovering just below the 50 percent needed to avoid a runoff and Hutchison holding a statistically insignificant lead in the race for second place.

It's still quite possible that Perry will blow through the 50 percent mark on primary day (coming up soon, March 2, and Minnesota thinks it's tough thing to move the primary all the way up to August). If he does, my point is moot. If he falls short of 50, Texas will have to put on a whole second election, with very little suspense.

That certainly wouldn't be necessary if Texas had IRV, and I would say this case is a good selling point for IRV backers.

It's also possible that Perry will finish so close to the 50 percent mark that the runners-up would be entitled to a recount. This also seems a waste of time and money that could be saved under the IRV system. Presumably, if Perry got 49-51 percent on the first round, the instant runoff would push him well above 50.

On the other hand, some IRV likers suggest that IRV is also a good way to prevent long, expensive recounts (like the agony of Minnesota's last Senate race.) But IRV is just as likely to cause a recount as to prevent one. In the Texas case, if Hutchison and Medina finished just a few votes apart, as the polls suggest they might, the IRV rules would still necessitate a recount because, for instant runoff purposes, Texas would have to figure out which was the third-place finisher so that candidate's vote could be redistributed to the top two.

(Of course, the existing Texas system would also require a recount under those circumstances, to figure out which of the runners-up would be on the runoff ballot.)

On balance, Texas would benefit from IRV. And in the Minnesota '08 Senate race, we'll never know how it would have come out, but IRV very likely would have prevented a recount.