The Selzer Score: Does Rick Perry Deserve a Spot on the Debate Stage?
We now know who’s in and who’s out of the Fox News debate and forum, based on the network's calculations drawn from an exaltation of polls defining the Republican field of candidates. Quibble with the details if you will, the upside for Fox News in using this method is that it is data-driven. It’s hard to argue with numbers, and the polls used by Fox show remarkable consistency.
Maybe the candidates assigned to the smaller forum—former Texas Governor Rick Perry, former Pennsylvania Senator Rick Santorum, Louisiana Governor Bobby Jindal, former Hewlett-Packard CEO Carly Fiorina, South Carolina Senator Lindsey Graham, and former Virginia Governor Jim Gilmore—will be sanguine. And there are grounds for arguments. The idea that a fraction of 1 percentage point could seal the fate of candidates polling near the 10th-place mark—as was the case for Perry or Santorum—will likely lead to second-guessing the method.
We created the Selzer Score to be a way of taking more data into account than just the proportion of first-choice votes each candidate attracts. In a field as mammoth as this one, 100 points is just not very many to go around. There’s an obvious absurdity in making a consequential decision when one candidate gets 3 percent of the vote and another 2 percent.
The index is simple math. For each candidate, we take the percentage of first-choice votes and double it. We then add the percentage who say the candidate is their second choice. Finally, we take the percentage who say they could see ever throwing support, even if the candidate is not their first or second choice. That “ever score” is halved, because it is less important than winning actual votes.
In our latest Bloomberg Politics national poll, the Selzer Score does not substantially recast the field. The candidates who do best in the first-choice vote do well when we calculate the Selzer Score. The exception is Ohio Governor John Kasich and Perry. The publicity surrounding Kasich’s entry into the race earned him a 4 percent showing in voters’ first-choice selection. On its own, that would be enough to get him into the main event at Fox News Thursday. However, he is relatively unknown, with just 26 percent saying they could see themselves ever supporting him. That’s in the territory of Graham (23 percent). A low “ever” score didn’t hurt Donald Trump—he, too, gets 26 percent for his “ever” score. But his dominating lead overshadows what may seem like limited potential.
Perry benefits from a high “ever” score at 49 percent. So, even though just 2 percent said he would be their first choice, he makes up for it for having more upside potential than almost all the other candidates (former Arkansas Governor Mike Huckabee is in that league with a 48 percent ever score).
Beyond looking at upside potential in addition to the main horserace question, the Selzer Score has one additional benefit: It spreads the field. In our current poll, the candidates get anywhere from 21 percent to 0 percent. Without Trump, the range is from 10 to zero. With the Selzer Score, the field ranges from 63.5 points for Trump to 7 points for new entrant Gilmore.
Instead of breaking between 2 and 3 percentage points, Fox News—had it opted to use the Selzer Score—might have chosen to reveal what could have been a clean break between 30 points and 26.5 points, using just the Bloomberg Politics poll findings. This is a case where a little distance would likely make the ultimate decision more palatable.