Aldermen Mon Mar 07 2011

Runoff Math: A Word Problem

By Ramsin Canon

Put yourself for a moment in the shoes of those men and women running the campaigns for the runoffs that will take place in April.

You ran a tough race, slogging through phone lists and knocking doors in frigid weather for three, six, nine months. Still, your candidate couldn't top 50%. You spent most of your money, but raising money won't be too much of a problem: you've got a list, and your candidate has proven their viability.

The question is: how do you get those X votes you need to top 50%?

Presumably, the people who voted for your opponent won't vote for you. So you need to try to get enough of the votes that went to the other candidate or candidates. If you're challenging an incumbent, maybe the answer seems easy: the people who voted for the candidates who couldn't cut know the incumbent and presumably don't like them, so just snatch up those votes.

Let's go with an easy set up. You've got Andrea, Beth, and Carl. Andrea is the incumbent, Beth and Carl challengers.

On Election Day, about 6,561 people vote. Of those, 3,046 vote for Beth, 2,224 vote for Andrea, and 1,291 vote for Carl. If you're Beth, your path to victory is clear: get enough of Carl's votes to get you to 50%+1.

This where variables enter into it and confuse the matter. The big ones: not everybody who voted in February is going to vote in April. And, the people who did not vote in February are only remotely likely to vote in April. Therefore, you can't just ignore the people who voted for you. You need to spend time and resources to make sure your voters re-vote. How much time? How much money? Since your opponent will probably lose some voters, too, do you need to spend time going after third-party voters at all? Maybe your opponent's voters were significantly less enthusiastic and simply won't come out a second time.

Suppose none of Carl's voters come out. They didn't like Andrea, but they hated Beth. Plague on both their houses! If Beth knew that with some certainly, she can just focus on getting her voters out, and know that mathematically she can't lose.

But that's a big risk. Given the likelihood that overall turnout will decline despite her best efforts, if Andrea can snatch up Carl's neglected-feeling voters and get her people out, she can catch up rather easily. Also, you don't really know how Carl's voters feel about your candidate. The candidate and his hard core volunteers may all loathe her, but the people who voted for him don't necessarily feel the same way--in fact, they very likely don't.

So, what do you do? Do you spend time and energy pursuing Carl's endorsement, and figuring out how to woo his supporters? Or do you throw them a bone but concentrate on your own voters?

I don't know the answer--I'm sure there are plenty of campaign people out there who could school us on the issue; but if they're working on a run-off campaign right now, they certainly aren't going to share their strategy publicly. But the example above, if you can't tell from the specificity, comes from an actual campaign in 2007 that we can use for its benefit of hindsight.

In that election, Rey Colon, Vilma Colom, and Miguel Sotomayor competed for the 35th Ward aldermanship. The numbers broke down respectively; Colon with the most, Colom in second, and Sotomayor in third.

For the runoff, turnout only dropped by 18 votes. At the time, however, they could not have known the drop off. In the 15th Ward, the votes decreased by 1,500. Had Colon merely tended to his own flock, it is possible (though in that circumstance, it was not likely) that Colom could have picked up Sotomayor's votes.

For our current example, let's look at the 25th Ward, where incumbent Danny Solis, the influential chair of the Zoning Committee, won 4,308 of the 8,804 votes cast, just shy of a runoff. Runner-up Temoc Morfin won 2,460 votes. Ambrosio "Ambi" Medrano Jr., son of a former Alderman, won 2,036 votes. If every voter re-votes, Morfin needs 1,848 of Medrano's votes. Solis, in contrast, would need only 95 of Medrano's votes.

Again, this is only if all the voters re-vote. Which is no certainty; and it is not easy to tell, as of right now, which side will lose more voters. Perhaps Solis' voters were less enthusiastic than Morfin's. Or vice versa--I haven't seen evidence either way. Suppose Solis loses 25% of his voters--about 1,000. That'd put him at 3,308. Suddenly Morfin's lag is just 900. If his voters are all come out again--as Colon's did--and half of Medrano's voters come out for him, he has made up the difference. Not likely, but certainly conceivable, particularly in a three-man race with a surviving incumbent.

Let's go back to Andrea, Beth, and Carl, but give them the 25th ward numbers: 4,308, 2,460, and 2,036. Let's say Andrea's cost-per-vote was $25 and Beth's was $15, and they both have $100,000. Let's spot them each 25% of their previous vote that will come out again no matter what: 830 and 525 respectively. How would you spend your next five weeks?

Show your work.