May 9, 2016
It’s still relatively early, but we can stop feeling guilty about paying attention to Hillary/Trump polls. With the nominees locked in, and John Kasich dispatched, we don’t have to listen to him talk about polling better against Hillary than Trump does. Bernie will still mention his relative standing, but a complete media blackout limits the audience.
Polls were very predictive about Trump’s strength against his fellow Republican contestants. Is there any reason to suspect they’re suddenly false when measuring against Clinton? The candidate who consistently led the primary polling race, consistently trails his fall opponent.
Trump is trailing. It’s by more than the margin of error. No pollster except Rasmussen has found him ahead nationally in the past three months. Clinton has led 90% of national polls over the previous six months. It’s more than a trend at this point. Still, the average gap is high single digits.
George H.W. Bush trailed Michael Dukakis by 17 points after the 1988 Democratic Convention. He won by 8. Gerald Ford made up 30 points on Jimmy Carter in a minute and a half in 1976 (unfortunately for Ford, he trailed by 33 and couldn’t find those final few stragglers.) Bill Clinton trailed both Bush 41 and Ross Perot at this stage in 1992. Ronald Reagan went from being about even to winning by 9 points in the final two weeks in 1980.
Trump is trailing, but not by a historically impossible margin. Though his large deficit with women, unprecedented gap with Hispanics, and now customary for Republicans lack of support among African Americans makes it easy for Clinton to lead him, this also provides Trump with plenty of places to improve.
If he only trailed by 10 points among women, or could get a full third of Hispanic voters, or could reach double digits with African Americans, he’d be damn close. I’m guessing the picture of him eating a taco bowl as he tweeted “I love Hispanics!” isn’t the move to immediately close the gap, but it’s theoretically possible.
Our job isn’t to predict whether Trump will modify his pitch, or determine what effect that would have on which group of voters. Instead, we want to figure out how we would notice improvement (or backsliding). What would make us think this is a 50/50 race, and what would allow Hillary to prep for her return to the White House.
Keep your eyes on the following:
Obama Approval Rating
At the moment, he’s back above water, +2.6 in the Real Clear Politics average, and scoring as high as +10 with Gallup. Obviously a positive number is better for Hillary than Trump. If Trump needs to get voters who approve of the president to vote for him, he’s got his work cut out for him.
You can figure that a percentage of those approving voters might fail to turn out, while some of those who strongly disapprove are motivated to pick Trump, but he’s far more likely to win if Obama’s approval number is upside down by late October. At the moment, the average approval number is 48.9%, and Hillary’s average head-to-head poll support is 47.3%.
One thing you can safely ignore is the right track/wrong track question. While a bad number is normally trouble for an incumbent party, that doesn’t seem to matter right now. The current average is -38 (26.9 right direction, 64.9 wrong.) If Obama has a positive rating and Hillary is ahead with those sort of numbers, it indicates the public is either holding congress equally accountable, or has completely given up assigning blame to the White House.
Percentage of Undecided Voters
Trump is struggling to hold the normal Republican advantage in places like Utah and Georgia. Just this week a WSB-TV poll showed him with a mere 1 point advantage on Clinton. That doesn’t sound good, and it’s a step back from the 9 point edge another pollster gave him in February.
But almost 20% of voters were undecided. A 42/41 result shows voters aren’t sold on Trump yet, but haven’t taken the leap to voting blue. If you see Trump even where Republicans should lead, and trailing where they should compete, it’s more manageable if he merely needs to move undecideds to his column instead of convince Hillary supporters to choose him.
We are seeing clear signs Trump will need to defend traditionally red states at least as much as he will get to contest blue ones. North Carolina, which Democrats have won once (narrowly in 2008) since 1976, is having some Trump indigestion so far. The Georgia poll is an irritant, Trump has plenty of time to bring voters home. North Carolina is another story.
Hillary Over 50% Nationally
Right now, she’s under 50% more often than not. Surveys trying harder to make voters pick someone can push her to or over the line. When we see more undecideds, she’s stuck in the mid 40s. It’s far more realistic to expect Trump to win over voters who are struggling to admit a choice than those who are comfortable declaring their support for Hillary.
Bernie has seen the pattern prove out during the primaries. In states where Hillary regularly crossed the 50% barrier, he struggled to catch up. Most of his victories were in states where she stayed below the line, even if she led him in polls earlier in the year.
Trump Over 45% Nationally
At the moment, he’s in the 40 to 41 percent range. When he’s competitive in surveys, Hillary is in the mid/low 40s. When he’s not, she’s at 50. Trump’s number doesn’t move much. Each of the last 11 polls show him between 38 and 43, with 40 and 41 by far the most common.
Even when he was doing better in these comparisons several months ago, Trump struggled to exceed 45%, only getting there in occasional outliers, only when virtually all voters made a choice. If Trump starts clearing 45%, even in polls with a decent amount of undecided voters, he’s made definite progress winning over some of his skeptics.
While presidents are picked on a state-by-state basis, Trump isn’t going to win a two-way race if he’s below 45%. Absent a major event, the last several elections have seen limited movement in the final few weeks. If he can’t get within range by the end of September, it won’t happen unless news intervenes.
Use Sanders Numbers to See if For/Against
Sometimes one of the candidates is actually popular. Donald Trump is popular in West Virginia. He leads Hillary 57/30 in a PPP survey. He leads Bernie 56/35. That means this is more pro-Trump than anti-Hillary. Don’t get me wrong, she’s not popular there. 30% in a two-way race is dismal. But Trump only loses 1 point of support against Sanders.
North Carolina is another story.The Tar Heel State does not like the Donald. He trails Clinton 49/40 in the most recent Civitas survey. Bernie leads by 18 (55/37.) That’s an indicator that neither nominee is well thought of, but Trump is especially disliked.
Meanwhile, Ohio has Trump losing by 3 to Hillary and 4 to Sanders. No particular distaste for anyone. This is an absolute must for Trump in November. Hillary is ahead by 3.5 in the RCP average, but that’s not exactly a safe lead. If you see Trump pulling ahead, or Hillary getting near double digits, it’s a sign something is changing.
We’ll continue to check in regularly and update as we see things change. The overall impression that Hillary is ahead, but not entirely safe yet is in line with the data. Should conservatives run an Independent Republican as a third party, or if Gary Johnson gains traction as a Libertarian option, we’ll revisit some of the above.