On 06th August, in a crucial Republican debate held in Manchester, New Hampshire, just three days before the New Hampshire Primary, Marco Rubio, cruising in the campaign till now, committed a major unforced error. When confronted by Chris Christie with the accusation that he was nothing but an empty suit with no real accomplishments and just a few memorised talking points, Rubio panicked and repeated the same talking point over and over again, almost as if trying to validate Christie’s criticism. It was a painful, cringe-worthy moment, especially considering the importance of this debate and the fact that Rubio has perhaps been the most polished performer in the Republican debates so far.
The media was quick to latch on to the gaffe, accompanied by its customary hype-ups. A lot of commentators have since compared Rubio’s performance with the holy grails in the pantheon of debate debacles, including Dan Qualye’s “Senator, you are no Jack Kennedy” embarrassment in the 1988 Vice Presidential debate, James Stockdale’s “Who am I? Why am I here?” introduction in the 1992 Vice Presidential debate or Rick Perry’s “Oops” moment in one of the 2012 Republican Primary debates. If you were the read the weekend coverage of the Republican race, you would get the impression that Rubio’s candidacy has been dealt a debilitating blow and it is probably not going to recover from this.
A bit of context setting is in order and this is what this post shall strive to do. For all the media coverage and entertainment that the most talked about debate moments provided, one shall keep in mind that they hardly went on to significantly alter the course of a presidential race. Quayle suffered through a major embarrassment in the 1988 debate, but it had little lasting effect and he along with George HW Bush went on to win the race. Stockdale’s boss Ross Perot gave the best performance by a third party candidate since Teddy Roosevelt in 1992. And Rick Perry was already in the middle of a steady decline in poll numbers when the “oops” moment came. It hastened his fall but the die had already been cast.
There have been seven Republican presidential debates before this one. To see the kind of effect debate performances had on the poll number of candidates, we have looked at the difference in Huffington Post Pollster average national poll of candidates one day before the debate happened and seven days after the debate.
The average absolute change in a candidate’s national polls one week from the debates was a meagre 0.46%. The highest gain was by Donald Trump, of 2.40%, after the fourth Republican debate (held on 10th November, 2015 in Milwaukee, Wisconsin). The highest fall was also seen by Trump, of around 2.30% in the second Republican debate (held in the Ronald Reagan Library, Simi Valley, California on 16th September, 2015).
Is there any way one can predict the change in poll numbers after a debate happens? Fortunately, the pundits in media are quick to come up with their list of winners and losers moments after each and every debate. The website Ballotpedia publishes the results of its own internal survey of 70-100 Republican insiders (strategists, operatives and party activists from around the country) and asks them to select the candidate who fared best in each debate and the candidate who fared the worst in each debate.
In the next chart, we have plotted the debate performance of each candidate (measured by percentage of Ballotpedia survey takers from the Republican Party who considered the candidate as the winner in that debate minus the percentage of Ballotpedia survey takers from the Republican Party who considered that candidate as the loser in that debate) against the poll bump of each candidate seven days after the debate ended. The survey results are not available for the seventh Republican debate and for the second tier debates.
It is apparent from the chart above that the performances of the candidates as measured by the party insiders have almost no linear relationship with their poll bumps.
Ballotpedia also comes out with a survey question that asks every respondent if a particular candidate helped his or her chances in that debate. We tried to measure the debate performance of a candidate in a different way by subtracting the percentage of respondents who said the candidate hurt his/her chances during the debate from the percentage of respondents who said the candidate helped his/her chances. But when I plotted this against the poll bumps of the candidates, the result was hardly any more illuminating.
It may be apt to mention here that this above sample set does not contain the seventh Republican debate which was probably Exhibit ‘A’ in the case on why pundits should not be trusted to judge the winners and losers of debates. Trump, even when he was absent, was widely considered a winner of the debate while Cruz and Rubio were considered losers. But against all conventional wisdom, Trump went on to underperform his poll numbers in the Iowa election (held 5 days after the debate) where both Cruz and Rubio over-performed theirs.
If we were to look at some of the most unambiguous winners and losers of the Republican debates and the bounce in their poll numbers, the result would make little more sense; but even by this measurement the effects are underwhelming at best.
|Candidate||Debate||Performance *||Poll Bounce|
|Donald Trump||First Debate||-48%||1.00%|
|Donald Trump||Second Debate||-31%||-2.30%|
|Marco Rubio||Third Debate||48%||1.20%|
|Marco Rubio||Fourth Debate||56%||0.80%|
|Ben Carson||Fifth Debate||-30%||-0.80%|
|Ben Carson||Sixth Debate||-44%||-0.10%|
|Jeb Bush||Third Debate||-50%||-0.40%|
|John Kasich||Fourth Debate||-51%||0.10%|
|Carly Fiorina||Second Debate||49%||2.40%|
|Rand Paul||First Debate||-24%||-0.50%|
*According to the first metric
So will Rubio’s debate performance have no impact on his prospects in New Hampshire? No, we are not saying that. These three days since the debate form a very crucial period for every candidate and a significant percentage of voters are making up their minds now. So Rubio’s gaffe and the huge coverage it is receiving in the media will probably affect his chances. But the voters will also be grappling with a lot of other factors while taking their decisions. So we do not know how much the effect will be. And considering that we are now just one day from the election, we will probably never know.
But considering the track record of conventional wisdom in judging the winners and losers of each debate, we would rather wait until the election to find out what actually happened than take the media’s verdict at face value and declare Rubio’s candidacy to be over.