Marco Rubio finished in third in Iowa — a “strong third” in which he outperformed his polls, but third nevertheless. And yet, his chances of winning the Republican nomination nearly doubled according to the bookmaker Betfair, from about 30 percent before the Iowa caucuses to 55 percent now.1 Meanwhile, Donald Trump, who finished ahead of Rubio although behind Ted Cruz, saw his chances halved from 50 percent to 25 percent.
Even if you acknowledge the importance of the “expectations game,” isn’t that a little excessive? Well, maybe.2 But I think Betfair is coming up with approximately the right answer.
Here’s why: Presidential nominations are a lot like the stock market. In the long run, they’re reasonably well governed by the fundamentals. In the short run, they can be crazy. Iowa represented the equivalent of a stock market correction, a sign that sanity might prevail after all.
In the stock market, the fundamentals consist of things like the profitability and growth of a company. In the nomination process, the most important fundamentals are what we call electability (can the candidate win in November?) and ideological fit (does the candidate hold positions in line with the consensus of her party?). A party would prefer to nominate a candidate who scores well in both categories.
Rubio fits the bill, perhaps uniquely among the remaining Republican candidates. His image with general election voters is not great, but it’s better than the other leading Republicans. He’s also quite conservative. That’s convenient, because Republican voters are quite conservative also. In fact, Rubio is almost exactly as conservative as the average GOP primary voter.
By contrast, Trump is problematic in both categories. It’s not always clear what Trump believes or where he would wind up as a general election candidate, but he hasn’t been particularly conservative for most of his career. His electability case isn’t good either; instead he has an extremely negative image among general election voters. If Rubio is a blue-chip stock, Trump is a risky mortgage-backed security.
And yet, Trump was leading in the polls for many months. We’ve spent a lot of time considering why, and I won’t rehash all of that discussion here. But one highly plausible answer is that his national polls partly reflected his overwhelming lead in media coverage, which allowed him to top the field despite having a narrower base of support than Rubio or Cruz.
Under this theory, Trump’s polls and his round-the-clock media coverage are self-reinforcing: Better polls lead to more coverage, which leads to better polls, and so on. In stock market terms, you might even call them a bubble. Back in the summer and early fall, it seemed likely that something would burst the bubble eventually: The media would grow tired of Trump or he’d do something to break his winning streak. At around the time of the terrorist attacks in Paris and San Bernardino, however, which brought even more attention to Trump, that became far less clear. Trump looked like he might ride the magic carpet all the way to the nomination in Cleveland.
Then Iowa intervened. Voters there researched their decision carefully and heard from all of the candidates, making the media playing field more level. And they decided they didn’t like Trump so much after all.
Maybe Iowa was just a fluke, and Trump will perform better in the next several states. Before Iowa, Trump had a big lead in New Hampshire, for instance, and in South Carolina. But Iowa was the first state to have voted, and the only test we’ve had so far of whether Trump’s support in the polls will turn into votes. Pretty much the whole case for Trump depends on the premise that it will; if the linkage is broken, it becomes futile to cite Trump’s polls in future states as evidence of his resilience.
Iowa might even prove to be Trump’s high-water mark. Rubio and Cruz are going to get a lot more coverage now, and Trump has lost his sheen of invincibility.
So although Iowa is just one data point, it was doubly important. If Trump’s campaign was a bubble, it might burst. If it wasn’t, Iowa nonetheless suggests that Trump might draw more like 25 percent of votes instead of the 35 percent or 40 percent support he receives in national polls. That happens to be an important range: A candidate getting a 35 percent or 40 percent plurality of the vote could easily win a majority of delegates under the GOP’s complex rules, but one winning 25 percent almost certainly couldn’t.
That doesn’t mean the betting markets have things exactly right; I think they’re too low on Cruz, for instance. But one data point can be awfully important when it’s the only data point you’ve got. New Hampshire will be the second, and you can bet it will be a market-mover too.
Hey, you should sign up for our weekly 2016 newsletter, “What I Thought About Over The Weekend.”
Listen to the latest episode of the FiveThirtyEight politics podcast.