Why expectations matter in Iowa
Posted January 28, 2020 11:32 a.m. EST
CNN — Vermont Sen. Bernie Sanders could be surging at just the right time. He's ahead in some of the highest quality polls coming out of Iowa with a week to go until the caucuses.
History suggests that a win in Iowa could allow Sanders to take the lead in the national polls -- but Sanders' surge could also turn out to be a bad thing for him if he underperforms expectations.
A look at past Iowa caucuses reveals what you probably already know: They really do matter. The winners of the caucuses get a 7-point median bounce comparing their national polls just before and just after (i.e. before New Hampshire votes). That's a huge bounce, especially if applied to a primary season as stable as this cycle's.
Visit CNN's Election Center for full coverage of the 2020 race
Right now in the average national poll, Sanders is down 5 points to former Vice President Joe Biden. With a win in Iowa, he would theoretically jump ahead or at least be in a race too close to call. And if history is any guide, Sanders would probably then jump ahead even further with a win in New Hampshire. Sanders, of course, is also leading in the New Hampshire polls right now.
Yet caucuses and primaries are also about expectations. None of the last three Republican nominees have won Iowa, and all of them won New Hampshire. They were able to do so in part because they weren't thought to be heavy favorites to win in Iowa.
To get more specific, once we control for the places candidates finish in Iowa, how much they under or outperform their Iowa polling makes a big difference. A candidate who outperforms their Iowa polling by a point in the final results gains about 0.7 points nationally. Meanwhile, a candidate who underperforms their Iowa polling by a point will lose about 0.7 points nationally.
A candidate who has the best of both worlds (i.e. winning and doing better than everyone thought they would) tends to jump nationally. Such was the case when John Kerry won 37% in the 2004 Democratic caucuses, even though he was polling in the mid 20s in the final polls. (Kerry was likely helped too by outperforming his national polling in the Iowa results.)
In the expectations game, Sanders was likely hurt by getting his strong polling now instead of right before the caucuses. The issue for him isn't so much that an Iowa win won't get treated as an important win. It almost certainly will, especially if he does even better in Iowa than his polling.
The problem for Sanders is what happens if he loses in Iowa. Pre-polling surge, he would have been in a decent position heading into New Hampshire. Today, I'm not nearly as sure he'll be able to survive nationally if he loses in Iowa by more than a few points.
Meanwhile, other candidates such as Biden, former South Bend, Indiana, Mayor Pete Buttigieg and Massachusetts Sen. Elizabeth Warren could be the beneficiaries of lower expectations than they were a few weeks ago.
The chance that any of them (especially Buttigieg and Warren, given their polling) see a boost nationally following an Iowa win is fairly significant.
Needless to say, a lot is at stake in these final days before Iowa votes.