Last week, Fox News announced its guidelines for the first debate among presidential contenders endorsed by the Republican National Committee (RNC). The network plans to invite the top 10 candidates, with the ranking determined by an average of the five most recent national opinion polls before the August 6 event. This is similar to the approach it has taken in previous cycles.
Following historical precedent is often smart. In addition, using a hard-and-fast metric, like a candidate’s poll position, is better than subjective criteria to determine whether a candidate is “serious.”
However, Fox has adopted the wrong approach, and the RNC is wrong to endorse it. Several problems stand out:
* The “margin of error” in polling does not disappear when one averages polls together. For instance, five polls with 750 respondents apiece would still yield a margin of error of about 1.5 points. That may not seem like much, but it could be trouble early in the cycle. What if the candidate in 10th place is polling at 4 percent on average, while the 11th-place candidate is at 3.5 percent? Statistically speaking, there is no difference between the two, yet one would be included while the other would be left out.
* Polls have been misbehaving of late. They were wildly wrong in Britain and Israel, and they were wide of the mark in our 2014 midterms. Worse, there has been evidence of what Harry Enten of FiveThirtyEight calls “herding”: pollsters producing results that closely mimic one another, but not what is happening in the real world.
* Polls simply do not tell us very much so early in the cycle. Voters are hardly paying attention, which means their opinions can be arbitrary and easily changed. We saw this in both the 2008 and 2012 GOP nomination battles, where the primary debates rapidly moved public opinion. Why should pre-debate polls carry any weight?
* There is no meaningful separation between the candidates yet. The Real Clear Politics polling average has Jeb Bush in first place, with 15 percent, and John Kasich in 11th place, at 2 percent. A 13-point gap is insubstantial in the early days of a presidential campaign cycle. Just ask President Barack Obama. At one point in 2007, he trailed Hillary Clinton by 26 points in the RCP average.
* It is not the business of Fox News or the RNC to determine the range of acceptable choices for Republican voters. If this were a typical cycle, with maybe a half-dozen serious candidates, a threshold such as this would make sense. It is the only way to exclude obviously nonserious or fringe candidates. But this is not a typical cycle. If we used the current RCP polling averages, the proposed threshold would exclude John Kasich, Carly Fiorina, Bobby Jindal, and Lindsey Graham from the first debate. These are all serious candidates—two sitting governors, a sitting senator, and a former Fortune 500 CEO. Moreover, the RNC has talked a good game about how to grow the party. Does it make sense to exclude a woman, the son of immigrants from India, and the governor of a must-win purple state? Neither Fox News nor the RNC should take it upon itself to decide that such candidates are unworthy of consideration. That task is best left to the voters.
There is no doubt that the RNC faces a logistical challenge with these debates. It is simply not practical to include more than 10 candidates in a single session (and even 10 will be a stretch). However, excluding serious candidates based on statistically meaningless poll positions so early in the cycle is a terrible solution.
There has to be a better way. For instance, CNN intends to have two debates, one with “first tier” candidates, and another with “second tier” candidates pulling in at least 1 percent apiece. But this approach still creates an arbitrary and meaningless distinction between who participates in which debate.
Both Fox and CNN should hold more two (or even three) debates, with the candidates divided up by some random selection, including all candidates who meet some basic threshold like 1 percent in the polls or a minimum sum of money raised. It makes sense to apply more stringent criteria later in the cycle; however, there should be a maximally inclusive approach in the early days of the campaign, without discrimination between candidate “tiers.”
The GOP electorate would surely appreciate this. A recent Pew poll found that Republicans are more excited about this field than their choices in the previous two cycles. It is an easy bet that primary voters would eagerly watch multiple debates.
In fact, the RNC should insist on inclusion. The only way to produce the best candidate to defeat Hillary Clinton is to examine all the credible contenders carefully. This means they all should be included in the debates, even if this means two or three-tiered debates in the early going.