Skip to main content

The context you need, when you need it

When news breaks, you need to understand what actually matters — and what to do about it. At Vox, our mission to help you make sense of the world has never been more vital. But we can’t do it on our own.

We rely on readers like you to fund our journalism. Will you support our work and become a Vox Member today?

Join now

After Brexit and Trump, it’s time to stop trusting our guts and start trusting the polls

There’s much to be said about Brexit. Start with Tim Lee’s rundown of what comes next, and Matt Yglesias’s look at the winners and losers.

But here’s a simple lesson from the GOP primaries and from Brexit: Polling is proving a much better guide to political outcomes right now than recent history or established norms.

We saw this during the Republican primaries. Donald Trump led poll after poll, but seasoned political observers simply refused to believe he would win. I was one of those skeptics, and I had no shortage of company. Even analysts who stick close to the numbers refused to believe the data. “Dear media, stop freaking out about Donald Trump’s polls,” demanded Nate Silver. For months, American political punditry was little more than an escalating series of rationalizations for why the guy dominating the polls would obviously, inevitably lose the primary.

And then, of course, he won.

Brexit followed much the same path. Even as the polls tightened, elite sentiment remained confident. On the eve of the vote, polls showed “Leave” even with “Remain,” but betting markets overwhelmingly assumed Britain would stay in. “I also walked by the European Commission in Brussels,” reported economist Tyler Cowen, “and saw not the slightest sign of panic or for that matter interest.”

And then, of course, Leave won.

On some level, we’re all good Bayesians: We’re skeptical of data that baldly contradicts how we know the world to work. But in politics right now, the world isn’t working the way we think it does. Polling is proving a much more reliable guide to political outcomes than the “does-this-seem-insane?” test most people use to guide their predictions.

The lesson here is not that the least likely, most disruptive outcomes are suddenly a safe bet. Brexit’s win doesn’t predict Trump’s victory, contrary to some of the chatter I’ve seen on Twitter. Rather, when evaluating the likeliest outcomes, look to trustworthy polls rather than your gut. Hillary Clinton’s 7-point lead over Donald Trump still makes her the favorite, but if those numbers flip, take it seriously. Pundits who stick to their priors even when the data tells them to abandon ship are not faring well this year.

More in Politics

Podcasts
The Supreme Court abortion pills case, explainedThe Supreme Court abortion pills case, explained
Podcast
Podcasts

How Louisiana brought mifepristone back to SCOTUS.

By Peter Balonon-Rosen and Sean Rameswaram
Politics
Trump’s China policy is nearly the exact opposite of what everyone expectedTrump’s China policy is nearly the exact opposite of what everyone expected
Politics

As Trump heads to China, attention and resources are being shifted from Asia to yet another war in the Middle East.

By Joshua Keating
Politics
Are far-right politics just the new normal?Are far-right politics just the new normal?
Politics

Liberals are preparing for a longer war with right-wing populists than they once expected.

By Zack Beauchamp
The Logoff
Flavored vapes doomed Trump’s FDA headFlavored vapes doomed Trump’s FDA head
The Logoff

Why Marty Makary is out at the FDA, briefly explained.

By Cameron Peters
Politics
Virginia Democrats’ irresponsible new plan to save their gerrymanderVirginia Democrats’ irresponsible new plan to save their gerrymander
Politics

Democrats just handed the Supreme Court’s Republicans a loaded weapon.

By Ian Millhiser
The Logoff
Can Trump lower gas prices?Can Trump lower gas prices?
The Logoff

What suspending the gas tax would mean for you, briefly explained.

By Cameron Peters