2016 Elections

Trump Will Still Lose. Here's How.

A New Hampshire poll shows Rubio or Cruz would beat him in a head-to-head matchup.

All the way.

Photographer: Scott Eisen/Getty Images

Some of us keep explaining why Donald Trump’s poll results so far don’t make him a likely Republican nominee, yet others keep saying they do. So here we go again: 

True, turnout for primaries is better than in caucuses, but it’s still not high. New Hampshire has higher turnout figures for its primary than most states. Yet in 2008, the total primary tally (about 527,000) was still lower than the general election vote (about 708,000). Trump is reportedly relying on first-time voters and others who don't regularly vote in Republican primaries. We don't know yet if they'll show up, but habitual voters are disproportionately the ones who usually take part in primaries.

Next issue: Early polling leads, defined as surveys taken before the first voting event, the Iowa caucuses. Princeton’s polling prognosticator, Sam Wang, dismisses recent arguments (see Nate Cohn and Harry Enten) that polls, especially national polls, have limited predictive value at this point. He contends that those candidates who are far ahead by now usually win the nomination. 

That is true, but it means less than he believes. In some presidential cycles, the contest is basically over by now -- as it is on the Democratic side this year. When a candidate locks up leads in every indicator at this point, from endorsements to money to organization strength to polling, then, yes, her polling lead is likely to hold up. But when there’s no obvious front-runner and the party is slow to decide, early polling leads tend to be ephemeral -- as Howard Dean's was in the 2004 cycle and Rudolph Giuliani's was in 2008. Conclusion? It’s party support, not polling, that predicts winners.

Ezra Klein has an excellent column saying Trump won’t win, but Klein is unsure how he will be stopped:  

It's the Underpants Gnomes theory of Trump's loss. Step 1: Trump leads the polls for month after month. Step 2: ??? Step 3: He loses! Even if you think that's likely, it sounds a bit ridiculous when you say it aloud.

To fill in that "???" I’ll review a few possibilities

Perhaps something will still rapidly deflate his support. Maybe it will be losing in Iowa. Trump’s message is that he’s a winner; if he falls short, does that make him a loser? 

Or perhaps he’ll just fade a bit. Trump has dominated the information most Republicans have heard about the contest, as this "Trumpometer" shows. When the race gets to their states, other candidates will be running ads and gaining notice. It’s suggestive that Trump is running about 10 percentage points worse in Iowa and New Hampshire, the two states with full-on campaigns right now, than he is nationally. If Trump gets half of all news coverage (compared with the two-thirds he has been getting nationally), the Trumpometer predicts he falls to about 25 percent of the vote.

What if he keeps his vote share as high as his polling numbers? He still might not win the nomination. He has moved up from about a quarter of the vote in August to a bit over a third in polls right now, but that’s not enough to win once the field winnows down to two candidates. This will happen quickly: If Trump wins close to 40 percent of the vote in early contests, the pressure on losers to drop out will be enormous. 

So, even if we’re just looking at polling, Trump still needs to gain considerable new support. Of course, that’s true for the other candidates as well. But they all have the opportunity to introduce themselves to voters who don’t know them yet. And they all seem better positioned to gain when similar candidates drop out than Trump does, because there is no one like him in the race at all.

This isn't just a hunch. Polling supports it, including a new PPP survey in New Hampshire released on Wednesday. It has Trump leading by 15 percentage points, but also finds he would lose to either Marco Rubio or Ted Cruz in head-to-head matchups. That won't happen in New Hampshire, but it could soon after.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

    To contact the author of this story:
    Jonathan Bernstein at jbernstein62@bloomberg.net

    To contact the editor responsible for this story:
    Katy Roberts at kroberts29@bloomberg.net

    Before it's here, it's on the Bloomberg Terminal.