June 13, 2016
It feels dirty to speculate how tragedy (or if you prefer, atrocity) will impact poll numbers. Fifty people are dead. A similar number are very injured. Yet this won’t actually stop the calculations. Instead it heightens them. Already, Donald Trump has shifted the topic of his speech for today from the Clintons to terror. Hillary Clinton has altered her schedule for the week, canceling a few events deemed inappropriate under current circumstances.
The Clinton team is quickly recalculating and recalibrating, while Trump is working out his own personal equation. It appears his recipe is way more of the same. Pundits were already questioning if his primary season approach would play in the general election. Now they will find out how a steroidal version does during a time of crisis.
Before shots rang out in Orlando, Trump was trailing Clinton by 4 points in the Real Clear Politics average. He’d moved into a tie in this index by May 23, and started slowly leaking support over the last couple of weeks. We don’t know if it was more due to extra scrutiny of Trump University itself, or his attack on Judge Curiel, or some combination thereof, but he peaked a few days before Memorial Day.
Clinton likely got a small extra boost from clinching the Democratic nomination, though most of the polling we have was completed or in progress before the AP announced she had the necessary delegates on the night of June 6. Even without that, the gap was growing again.
Rasmussen Reports surveyed from 5/17 to 5/18 and found Trump ahead by 5 points. This is the Rasmussen poll that was included in the RCP average at the time Trump was tied overall. By the next study point (5/23-5/24), Rasmussen had Clinton ahead by 1. They found the same a week later (5/31-6/1.) The most recent survey (6/6-6/7) has Clinton ahead by 4. This means Trump has given away 9 points to Clinton over the past three weekly reports.
Fox News shows a similar trend with fewer data points. From 5/14-5/17 they had Trump +3. More recently, a poll taken 6/5-6/8 shows Clinton +3, a switch of 6 points. Both Rasmussen and Fox News had better numbers for Trump than the average pollster when the mid-May surveys were taken. If he’s trailing with them now, he likely lost a good amount of ground.
Several other companies have polled multiple times over the past few months, but Rasmussen and Fox News were the only participants at the time of Trump’s recent peak and in the past several days, so we can’t make a direct comparison with the others. Now Orlando will further complicate the readings.
Trump unquestionably received a boost in GOP primary polling after both the Paris and San Bernardino attacks. But he’s not just trying to win over the GOP right now. Maximizing support within the party and among independent leaners is important. When Trump was at his best in polls a couple weeks ago, a big reason was his ability to grab a higher percentage of Republicans and GOP leaners than Clinton was among Democrats/Dem leaners.
There are more registered Democrats, so if they have equal conversion rates, she leads. But when it appeared Trump was consolidating his side, while Hillary was struggling to win over enough Berners, he was even in the poll averages and led in some. More recently, a few Republicans have had second thoughts about Trump, while a few Democrats have rallied around their presumptive nominee. This is hardly cast in stone, but the shift favored her.
If Trump uses the turmoil to consolidate his side, while Democrats flock to Clinton, the event will leave Hillary ahead, though not by a ton. Remember, while Trump did better against his foes after Paris and San Bernardino, Clinton picked up ground against Bernie. There was a Democratic debate the night after Paris, and Sanders was thrown off by the focus on national security.
For each person who is sure Trump will leverage fear and anxiety into preference for a strongman-style leader, there is another who believes voters will prefer Clinton’s measured approach. If you see Hillary regularly ahead by 8 to 10 points in the next week, voters are responding better to her Orlando response.
If the margin is between 4 and 7, it was likely a wash. That’s approximately where we stood before anything happened. If Trump is trailing by 3 points or less, he was able to use the attack as a way to flip the script again and regain momentum by saying extra Trumpy things. Should he find himself with an actual polling lead in another week, it indicates the country is tired of calm responses.
Which of those outcomes is more likely? We do have a few clues in the Trump/Clinton matchup numbers from November and December. In the immediate aftermath of Paris (the attacks were 11/13, polls taken 11/16-11/17 and 11/16-11/19), Trump recorded improved results against Clinton with both Fox News and PPP.
Fox News had him ahead by 5, tied for the best result from any poll from any pollster in the past year. PPP showed Clinton up 1, the best Trump has done in any PPP survey (they’ve measured 8 times in the past 10 months.) Not the biggest sample size, but it appears Trump scored well after the first tragic event.
We often think of Trump being Trump, but the proposed ban on allowing Muslims into the country didn’t happen until San Bernardino. After Paris, he confined most of his commentary to criticizing President Obama for being out of touch, not doing enough to combat ISIS, and failing to protect Americans.
While the Muslim ban went over well with a majority of GOP voters, and was an acceptable thought to a decent amount of non-Republicans, the San Bernardino aftermath did not help Trump against Clinton, at least not in a way we can find in the data. At the time of the attack on December 2, the two candidates were roughly level in the RCP average.
Ten to fourteen days later, when the polling average was comprised of surveys taken after San Bernardino, Clinton was ahead by 5 to 8 points. If Trump’s Muslim ban or general response to the first clear example of an ISIS-influenced attack on American soil helped him in polling, I can’t find it.
So far, Trump sounds like an elevated version of post-San Bernardino Trump, not the Paris version. If his numbers improve this time, it would indicate Americans are feeling differently now than they did in December.