With the Iowa Republican caucus only two weeks away and Donald Trump still leading the national polls and polls in New Hampshire by a huge margin and essentially tied with Ted Cruz in Iowa, the Republican Party establishment is understandably perturbed. An illogical and impulsive summer fling of the voters, which was supposed to fritter away to obscurity as the seriousness of the campaign began in earnest in fall, has instead turned out to be an obstinate obsession. The question inevitably is – will Trump fade away at all? And if so, how?
After having been made to eat their words throughout last year, pundits are no wonder, no longer very excited in making straight forward predictions on Trump’s demise. But at the same time, in the deepest corner of their hearts, almost none of them have accepted the prospect that Trump may win the Republican nomination. After all, how can a weird looking, sixty nine year old showman with zero policy experience and expertise and a long history of doing everything that an ideal presidential contender is not supposed to do, win the nomination of a party with which he shares little ideological proximity?
Out of this dilemma has emerged a few theories on how Trump will eventually falter, as if the outcome is rooted in inevitability and the only the means are subjects of conjecture. One of the more talked about among these theories is that most primary voters do not make up their mind about their voting choices till they are very close to actually casting the vote. As a result of this, most primary contests are highly fluid and the outcome may remain unknown till the last moment.
All this is fair and square. But why will Trump voters leave him in droves and start voting for some other candidates when they have been giving him more than double digit leads in national polls for the better period of the last six months? The answer, some of the experts beckon, lies in electability. Donald Trump is, they say, a much less electable leader compared to the likes of Marco Rubio, Jeb Bush, Chris Christie and even Ted Cruz in a general election against a democratic presidential candidate. And while a lot of GOP voters like what they see and hear about Donald Trump, they shall eventually come to the realization that voting for Trump shall hand over the election on a platter to the Democrats. As a result, they will be inclined to hold their noses when the Election Day comes and vote for a less palatable but a more electable candidate; given the depth of the Republican field this year, this should not be a gigantically challenging endeavour.
All of these sound good on paper. After all, this is how Howard Dean, the anti-war liberal governor from Vermont was rapidly brought down to earth on the election days in Iowa and New Hampshire after leading in those states for most of the preceding few months. And this is how favourite establishment candidates like Jon McCain and Mitt Romney won comfortably in 2008 and 2012 over fringe right candidates like Mike Huckabee and Rick Santorum respectively. As the conventional wisdom goes, primary electorate makes an optimization decision between electability and ideological purity and elect the most ideologically pure candidate who is electable. And as they start paying attention to the candidates, the primary voters will realize that Donald Trump is a suicide bomb in an expensive suit; and no sooner does this realization dawn on them, they will desert their candidate en masse and settle on a more sober choice.
A look at the numbers, though, suggests that the Republican voters do not perceive Donald Trump to have an electability problem. In fact, in almost every poll that has asked Republican voters their opinion on electability of various candidates, Donald Trump has led the field. Not only that, the percentage of GOP voters who perceive Trump as the strongest general election candidate has exceeded the percentage of voters who actually intend to vote for Trump.
In this election cycle, four pollsters namely CNN/ORC, Quinnipac University, Fox News and ABC/Washington Post have asked this question to the voters in a number of surveys in different forms. In case of Quinnipac surveys, the question is a more generic one, on the lines of “Do you think the following candidate has a good chance of winning the general election?” For CNN/ORC, the question is “Do you think the Republican Party has a better chance of winning the general election with Donald Trump as the nominee?” whereas for others, it is something like – Which of these following candidates has the best chance of winning the general election?
The results of these polls, carried out over the last six months, have been tabulated below:
As mentioned above, in Quinnipac surveys, the answer is to a broader question of whether Donald Trump would have a good chance of winning the general election
In each of the polls, more number of Republican voters considers Trump as the strongest general election candidate than voters who are actually willing to vote for him. This is particularly startling considering that most members of mainstream media actually consider Trump to be the least electable Republican candidate, at least among the so-called frontrunners.
Do primary voters always hold such delusional beliefs about the electability of candidates around this time of the election cycle? In order to find out, we also find out how the establishment and anti-establishment candidates did on the electability parameter in the closing months of 2007 and 2011. The results are tabulated below:
In fact, there is strong relationship between the percentage support of a candidate and the percentage of voters rating him/her as most electable. When plotted against each other in a scatter plot (excluding Trump’s number and including the numbers of only candidates only mentioned in the table above), the following graph emerges:
Thus Trump’s high support level is definitely related to his electability numbers, although it is difficult to identify which one is the cause and which is the effect. A more interesting observation, perhaps, is that for every so-called establishment backed candidate (Mitt Romney in 2011-12, Hillary Clinton, Rudy Guilani and Jon McCain in 2007-08), the electability percentage is higher than the percentage of voters supporting them, in almost all polls. The reverse is true for insurgency candidates (like Santorum and Gingrich in 2011-12, Obama and Huckabee in 2007-08). Infact, we can do a multiple regression using electability as the output variable, percentage support as one of the input variables and whether the candidate is establishment backed or an insurgent one as the second input variable (denoting 1 for establishment candidates and 0 for insurgent candidates). The results of the regression have been tabulated below:
As is clear from the table above, the R square value is very high at around 91%. Further, both the variables have very low p values (less than 0.01); thus, we can say with more than 99% confidence that both the support level and establishment backing are related to how many voters perceive the candidate as the most electable one.
The regression also yields the following equation:
Electability percentage = 0.97*(Support Percentage)+0.12*(Candidate Positioning)
where 1 denotes an establishment candidate and 0 an insurgent one.
However, when we apply this equation to Trump’s support level (considering him an insurgent candidate), the electability numbers that come out consistently are less than his actual electability numbers.
The results show that the voters have, in the last two election cycles at least, bought the argument that the moderate, establishment candidates are more electable in a general election and if they have supported some fringe or ideologically more extreme candidate, a large number of them have done so in spite of knowing that he or she may not be as electable.
The support for Trump, on the other hand, is with the belief and maybe, because of the belief that he is the strongest candidate the Republican Party has to offer in general election. Thus, if the focus of the election shifts to electability, going by this data, more number of voters may theoretically jump onto the Trump bandwagon. Hence, going by this logic, the movement of poll numbers towards an establishment candidate at the later stages of a campaign may indeed be reversed in this election.
But why are the Republican voters taking such a contrarian position for Donald Trump? Trump may be called a moderate by some (by virtue of being extreme on both sides) but he is no body’s idea of an establishment candidate. The answer, honestly, is “I don’t know”. Like a number of mysteries of this election cycle (e.g., Why has the Republican establishment not fallen behind Rubio yet? Why has Jeb Bush’s poll number not shown any sign of recovering in spite of millions of dollars spent in advertising? Why have Mike Huckabee and Rick Santorum, two previous caucus owners struggled so badly in Iowa?), there are probably no easy answers to it. If one is willing to look beyond the obvious though, one may argue that the election of Trump may not actually doom the Republican ticket. Trump is currently running a campaign almost exclusively on the basis of the power of his celebrity and his ability to hog headlines. He does not relay have a great grass-root campaign to rely on. Further, a large number of his voters are white, less educated males and people who do not vote very frequently. If Trump manages to win the Republican Primary, it means he has somehow managed to make these voters come out in droves and vote for him. Trump’s winning the nomination, in other words, may be a solution to the ‘missing white voters’ problem’, a theory that implies that GOP can win the White House just by increasing the percentage turnout of white voters, rather than increasing the percentage of support among minorities.
I cannot say I am 100% convinced by this theory as these less educated white voters are mostly concentrated in solid red states, although some swing states like Iowa, New Hampshire and Ohio also have a large proportion of such voters. But I am also not very convinced by the doomsday scenario for Republicans predicted by certain experts. Politics in USA today is sharply polarised and it is difficult to imagine a large proportion of Republican voters switching over to the Democratic side, no matter who the candidate is, unless of course there is a third candidate in the fray. Further, Trump may lose heavily among minorities, but he does not need to win among them; anyway, it might be challenging for the Democrats to replicate the turnout of and support among minorities with Obama not on the ticket anymore. One may also argue that Trump’s favourability is very poor among the general electorate, but his favourability among Republicans was also abysmal at the beginning of his campaign and it has improved dramatically over the last six months. Further, Hillary Clinton is also not very popular right now (although admittedly much more popular than Trump), indicating the polarized times we live in.
So, if electability does not doom Trump, what may actually prevent a Trump nomination? Again, I don’t have any definite answer. I cannot even predict with any degree of certainty that Trump is going to lose. But here are some scenarios that may lead to Trump falling short:
– As mentioned above, a large proportion of Trump’s coalition consists of unregistered voters who may or may not turn up to vote. Further, Trump’s ground game, especially in Iowa, lags that of many of his rivals. Considering the inclement weather that may be expected on the day of the caucus and the tedious process of caucusing, a large number of Trump voters may not show up, resulting in Trump losing to Ted Cruz in Iowa (currently, the two are more or less tied there).
– Trump’s loss in Iowa, which looks very much possible at this moment, may lead to a cycle of negative news coverage and erosion of his poll numbers in New Hampshire. Cruz may gain a lot of momentum from the Iowa win, leading to a large bump in his poll numbers in New Hampshire. Also, Marco Rubio may finish a strong third in Iowa and this may result in consolidation of moderate votes behind him in New Hampshire. As a result, Trump may either perform poorer than expected in New Hampshire or even lose out to either Cruz or Rubio in that state.
– After New Hampshire, the conservative voters may unit behind Cruz while the moderate/establishment ones may fall behind Rubio, leaving Trump in the middle without any core group of supporters and no real organization. Trump may continue to perform well here and there, winning a few states and delegates along the way, but his probability of winning the nomination may rapidly diminish.
Of course, I am not a seer and all of these predictions may go wrong. But if you are to look at Trump’s possible demise, rather than looking at electability, a bett er idea would be to look at his relatively lower polling numbers in Iowa and New Hampshire, his high support level among first time/unregistered voters and the possible inability of his inefficient ground level organization to turnout such voters.