New polls: Trump’s lead just 4 percent in one poll, 21 percent in another–Aberration or ???–Analysis.
The Quinnipiac national poll, released Tuesday, December 22, showed Trump with 28 percent, just a 4 percent lead over Ted Cruz–his closest rival (and the leader two recent Iowa Polls). The significance of the Quinnipiac poll is that it was taken over the period of December 16th to 20th; hence it was the first national poll administered, at least in significant part, following Trump’s praising of Putin’s authoritarian regime. Just as it appeared that Trump’s campaign was finally showing signs of imploding, a second poll was released, this one by CNN. The two polls showed unusual variance, not only in terms of absolute numbers, but in direction of change–at least for Trump. The CNN poll was taken the 17th to 21st. It showed Trump at 39 percent, a full 21 percentage point lead over Cruz. Some variance among polls is to be expected early in the primary season. But not like this. Obviously pollsters are having a difficult time modeling the likely voter.
To understand what I mean about “modeling the likely voter,” consider a state poll, for New Hampshire released Thursday, December 24. In it, among those described by the pollster (American Research Group-ARG) as “Likely Republican Primary Voters,” Donald Trump, with 21 percent, maintains a dwindling lead over four other hopefuls who are bunched together in a 10 to 15 percent range. In second place is the rising (in N.H.) Marco Rubio, with 15 percent, followed by John Kasich with 13 percent, and Chris Christie came in at 12 percent. Kasich and Christie have been concentrating their campaign efforts in New Hampshire, and, at least in this poll, it seems to be paying off. Cruz, who was second in a CBS poll completed four days earlier, finished fifth with 10 percent. Bush, Carson, and Fiorina were further back with 5 to 7 percent. Now, New Hampshire has a system whereby voters do not have to be long-time registered Republicans or Democrats in order to vote in one of those party’s primary. The voter can remain undeclared right up to vote day. So, how do the pollsters treat undeclared voters? They can count their “vote” in the polls according to how “undeclareds” split in previous elections. If so, how many elections should they use? What of the conventional wisdom that undeclared voters for Trump, on the Republican side, and Sanders, on the Democratic side, are much more motivated this time around than is usual for “undeclareds”–hence the use of an historical basis of weighting them will understate those two candidates’ support. It’s no minor problem. For example, Trump, at 15 percent only ties the rising Kasich for second place behind Rubio (and barely ahead of Christie) among registered Republicans. But Trump leads his rivals by a significant amount, with 29 percent among those who are undeclared, yet say they are likely to vote. Sanders, on the Democratic side, polls only 37 percent among registered Democrats, but his numbers rise, significantly, to 54 percent among undeclared respondents. Another decision polling organizations have to make that can affect their results has to do with the relatively recent phenomenon of widespread cell phone use. They have to decide what percentage of their sample will be land lines, the rest, of course, being cell phones. Does it really matter whether a respondent uses a land line or a cell phone? It may well make a difference. In the ARG poll, Sanders draws five percent more among cell phone users, possibly reflecting his popularity among young voters. As we get further along the primary trail, with evidence from the early primary states to guide them, pollsters might make adjustments to their model of the likely voter, but for now, there are a lot more “seat-of-the-pants” guesstimates made–and different guesses no doubt plays a part in accounting for the variance in poll results.
Why the emphasis on how recent the polls were taken when they vary by only a day or two?
The background: Putin praised Trump on December 17, and Trump at first limited his response to a typical “Donaldism.” He immodestly said: “If Putin likes me, and he thinks I am a good, smart person which I hope he believes that I am, he’s right… I am a brilliant person, you know that.” But then he got beyond his own ego, for one of the few times in this campaign, and scared many Americans by dismissing Putin’s “disappearing” and jailing of dissidents and critical journalists, noting instead that it resulted in a stable Russia and showed that Putin was a strong leader. This earned Trump the contempt of every other Republican hopeful (excepting Ted Cruz, who has consistently refused to criticize Trump, for strategic advantage), as well as the past two Republican Presidential nominees.
Recall that in the past week several national polls, whose results favoring Trump have become the staple of Trump’s campaign speeches since their publication. Those polls had him increasing his poll numbers to between 34 and 41 percent–something Trump crowed about at every campaign rally. It was assumed that Trump’s strong man image, in the face of the Paris terrorist attacks, was responsible for a goodly part of his rise in the polls. The San Bernardino shootings, it was assumed, further helped Trump by pushing terrorism to the top of Republican voters’ concerns. With his praise of Putin, some have thought that Trump had finally gone too far for a Republican electorate. Therefore, these two new polls, reflecting a few days of polling following the Trump-Putin love-sonnet exchange, gives us the first glimpse of what many had hoped would be the implosion of Donald’s presidential campaign.
Two things should be noted: 1) Trump’s core support has remained in the low twenty percent range. Polls reveal that this core support, relatively speaking, has been solid. More of those who say they will vote for Donald Trump also say that their mind is made up than for any other Republican hopeful. Yet Trump continues to have the highest unfavorables of any of the candidates. And when asked by Qunnipiac whether “Donald Trump has the right kind of experience to be President or not,” 67 percent responded “No.” Trump will have to rise above his core support level to get anywhere near the fifty percent needed for the nomination. 2) Despite this, and the outrageous statements from Trump that were expected to sink his campaign earlier, Trump, until this Quinnipiac poll, saw his numbers rise with every new poll. This despite his insults to women, his seeming to approve of the internment of Japanese-Americans in World War II, his suggesting a database on all American Muslims, and excluding all Muslims from entering the country. He can’t seem to stop insulting others any more than he can stop his constant boasting of being the best at whatever skill is mentioned. Even since praising Putin, a man referred to by other Republicans as a KGB gangster and thug, the garbage has continued to flow from Trump’s mouth, seeming uncontrollably. At a rally Monday, he dwelt on Hillary’s needing to use the bathroom during the Democratic debate, inspiring a CNN analyst to liken him to a 12 year old boy. More disgusting yet, referring to Hillary’s defeat for the Democratic nomination, Trump said that “Obama shlonged her.” In case any of you don’t know that term, it’s a widely used Yiddish word for the male member. Therefore it was no surprise that in this latest Qunnipiac poll, half of those questioned said that they would be embarrassed by a Trump presidency–and that was asked prior to his toilet and “shlong” comments! 57 percent in the CNN poll hold a negative view of Trump, despite his rise in support. It appears that most of those who don’t hold negative views of Trump, gravitate toward him, as opposed to any of the other candidates. Cruz’ increased support may have come from evangelistic conservatives who had earlier favored Dr. Carson.
Some of Trump’s earlier surge in the polls also can be attributed to desertions from his former closest competitor, Dr. Ben Carson. Similarly, as suggested above, a goodly number of Dr. Carson’s former supporters went to Cruz. In the Quinnipiac poll, which showed lower support for Trump than in the CBS poll, some of Trump’s recent surge– beyond his Core supporters–appear to have migrated to the Tea Party candidate, Ted Cruz. This was also suggested in recent Iowa polls, where Trump’s one-time lead has changed to a 10 percentage point deficit to Cruz. Hence, the anomaly of the CNN poll. Just as we thought we had Iowa figured out, on December 4, a Gravis poll conducted in Iowa showed Trump and Cruz in a tie for first place with 31 percent each, well ahead of the rest of the pack. Obviously voter choice is in a state of flux–not at all unusual a month ahead of the first primary.
What of the “moderates?” Clearly the moderate Republican vote is being split among Marco Rubio, Jeb Bush, and Chris Christie. In New Hampshire, you’d have to add Governor John Kasich. The latter seems to be putting all of his resources into the New Hampshire race. Jeb Bush, the early favorite, can’t seem to gain any traction despite having the largest campaign chest. His unfavorables rival those of Trump at 57 percent in the CNN poll, however Bush has far fewer core supporters than Trump. Bush’s numbers seem stuck in the 5 to 9 percent range. One can never dismiss anyone with the name recognition and money that Bush has, but if his numbers don’t show marked improvement by the time of the New Hampshire primary, his campaign may be over. Chris Christie moved up from 2 percent in the previous Quinnipiac poll about three weeks ago to 5 to 6 percent in their latest poll. Meanwhile, a CBS poll of New Hampshire Republicans released Sunday, December 20, showed Christie at 11 percent, and, as noted, the ARG poll had him at 12 percent, so there is some evidence of life in his campaign. John Kasich came in at 8 percent in the CBS poll, and at a surging 13 percent with ARG. Rubio has been feuding with Cruz, and, thus far, seemed stuck in the 10 to 14 percent range, while Cruz’ numbers increased in the CBS and Quinnipiac polls. First advantage, Cruz. But hold on, the new Gravis poll in NH, which showed Rubio well ahead of Cruz, 15 percent to 10 percent. For now we can simply note that the field of moderates will have to narrow to allow one of them to move into serious contention. Rubio seems the most likely to benefit if that happens, though Christie or Kasich with a surprise victory in New Hampshire could move center stage. Dr. Carson still polls 10 percent in national polls, mostly from voters identifying themselves as evangelical Christians.
Trump draws a considerable number of his supporters from the white, protestant, lower-educated demographic. He is also assumed to be popular with those who feel most alienated from society in general, and the establishment politician, in particular. Historically these groups, which probably overlap to some extent, show lower turnout numbers on election day. This cycle, they appear more enthused about their candidates, Trump, and to a lesser extent Cruz, than normal. Different assumptions by the pollsters about how many of this demographic will actually vote no doubt accounts for some of the variance in poll results.
Meanwhile, the media has focused on the controversial Trump. Meaning lots of free publicity, both good and bad. CNN’s poll showed that more respondents relied on post-debate analyses, rather than having seen the debates themselves. Nearly a quarter of those polled neither saw the debates or watched news reports about them. Further polling, which reflect the widespread dissemination of news of Trump’s recent controversial statements would have to confirm the downward direction of his poll numbers in Quinnipiac’s national poll before one can talk about Trump’s campaign’s implosion, though the newest New Hampshire poll lends support to that thesis. CNN’s national poll, on the other hand, shows no evidence of a Trump collapse. Quinnipiac’s poll suggests that it may have started, although, even then, as we have seen before, Trump may be just one terrorist atrocity away from a resurgence. Given the sizable number of respondents who neither watch the news or the debates, it likely will take several weeks for news to disseminate–and an equivalent amount of time before the polls reflect its effect on candidate choice. Given the number of strong negatives that Trump seems to have accumulated, some Republicans may sit out the general election should Donald Trump win their party’s nomination. In fairness, it should be noted that, on the Democratic side, Hillary also has high negatives, particularly when it comes to the trust issue. Campaign organization, always a critical dimension in elections, may play an even greater role this time in order to motivate reluctant supporters to show up at the polls. Insofar as the early primaries go, conventional wisdom has Trump with a relatively weak ground campaign. The effect of this should be more pronounced in Iowa, home of the first primary, with its caucus method of voting. Winning in most of the early primaries–defined as those held before the second week of March, known as Super Tuesday–may prove more symbolic than real due to the Republican rules this year that require most of them to apportion delegates on a basis proportional to the vote the candidates receive, as opposed to the winner-take-all nature of the later primaries.