With so many candidates and so long to go until next November, we’re going to make plenty of bad predictions over the course of the 2016 campaign. But one of our very first predictions about 2016, one we made almost three years ago, has already proven true.
“If [Hillary] Clinton runs for president in 2016, one thing is almost certain,” I wrote back in December 2012, at a time when polls showed about 65 percent of Americans had a favorable view of Clinton, “she won’t be as popular as she is right now.” I added: “In an era of intense partisanship, there is a relatively low ceiling (and perhaps also a relatively high floor) on the favorability ratings that any politician can have in the most active stages of a presidential campaign.”
Clinton’s favorability rating has, in fact, fallen quite a lot, to an average of about 42 percent favorable and 48 percent unfavorable in recent polls.
Numbers like those, when combined with the “emailgate” scandal and Sen. Bernie Sanders’s position in the polls (he’s now running very close to Clinton in New Hampshire, although not in Iowa or nationally), have a lot of commentators saying Clinton’s campaign has had an unexpectedly rough start. “Hillary is probable, but no longer inevitable,” wrote David Horsey of the Los Angeles Times, assessing her chances to win the nomination.
Horsey is right to deal in probabilities rather than certainties. Personally, I give Clinton about an 85 percent chance of becoming the Democratic nominee. (The general election is a whole different story.) That’s a pinch higher than betting markets, which put her chances at 75 to 80 percent.
But those betting markets, unlike some pundits, haven’t changed their assessment of Clinton much. In the markets, her probability of winning the nomination is still close to its all-time high and has barely budged in the past few months, rarely falling much below 75 percent or rising much above 80 percent.
Emailgate? #feelthebern? Clinton’s declining favorables? The betting markets think everything that’s happened to Clinton so far in the campaign is pretty much par for the course. It’s not that these markets are clairvoyant; they presumably didn’t know there would be a scandal involving Clinton and her email server, for instance. But it was a pretty good bet that there would be some scandal involving Clinton. (It’s not as though there is an absence of them to pick from.) Likewise, while you might or might not have identified Sanders as the person to do it, it was a pretty good bet that some challenger to Clinton would be situated about where Sanders is in the polls. So events like these were “priced in” to her stock. Let’s look at each of them in a bit more depth.
The rise of Bernie Sanders
Being “inevitable” doesn’t mean you’ll sweep through all 50 states with no opposition. In the modern era (since 1972), the non-incumbent candidates who were similarly “inevitable” to Clinton, judging by the number of endorsements they had early on in the race, were Bob Dole in 1996, Al Gore in 2000, and George W. Bush in 2000. You can probably also add George H.W. Bush in 1988 to the “inevitable” list; he had a narrower endorsement lead but was the presumptive Republican nominee by virtue of being the sitting vice president.
Among these candidates, only Gore went undefeated in the primaries (and Bill Bradley came within a few percentage points of beating him in New Hampshire). In 1988, George H.W. Bush finished third in Iowa — behind Dole and Pat Robertson. In 1996, Dole lost New Hampshire to Pat Buchanan. George W. Bush lost badly to John McCain in New Hampshire in 2000.
But wait — wasn’t Clinton “inevitable” in 2008 too? Not to nearly the same extent. Her lead in the polls is considerably larger this time around, her edge in the endorsement race is much greater, and her opponents are weaker than Barack Obama and John Edwards were. If you set a lower threshold for “inevitability” and included Clinton’s 2008 campaign in your equation, you’d probably also need to include winning campaigns like Romney in 2012 and Mondale in 1984, in which case the front-runners would be six-for-seven — an 86 percent success rate, which is about where I’d put Clinton’s chances now.
In fact, Gore is the only non-incumbent in the modern era to have swept all 50 states. (Two incumbents, Gerald Ford in 1976 and Jimmy Carter in 1980, also lost some states.) More often, candidates similar to Clinton have lost Iowa or New Hampshire, along with a few other states, before consolidating their support and eventually winning fairly easily.
In Sanders, Clinton has drawn an opponent who is relatively well suited to New Hampshire and Iowa. The reason is simple: Sanders’s support comes mostly from white liberals, and the Democratic electorates in New Hampshire and Iowa have lots of white liberals. Furthermore, Iowa and New Hampshire are small states, which makes it easier for candidates who don’t have Clinton’s financial resources to compete there. But we’ve seen this movie before. Based on current polling averages, Sanders would almost exactly replicate Bradley’s performance in 2000, losing Iowa by double digits, giving Clinton a close call in New Hampshire, then losing badly once the calendar turned to more populous and diverse states.
Or Sanders could do better than that, winning New Hampshire and a few other states in New England, the Upper Midwest or Pacific Northwest, perhaps along with one or two surprises elsewhere. But that too would be consistent with the losses that “inevitable” candidates like Clinton have endured in the past.
Clinton’s email and “value over replacement scandal”
Aren’t Clinton’s polling numbers suffering as a result of the scandal surrounding her use of email while secretary of state? It’s certainly possible, but it’s hard to determine cause and effect. There hasn’t been the sudden change in ratings you see around historical scandals like Iran-Contra; instead, her favorability ratings have been declining at a steady but relatively slow pace for a long time.
If you squint, you can perceive some acceleration in the downtrend in March, when details of the email scandal were first reported. But there were a lot of other things going on this spring too — notably, Clinton was in the midst of launching her campaign. And while Clinton has seen a decline in her polling in areas such as honesty and trustworthiness — consistent with the hypothesis that the scandal is affecting her numbers — she’s also seen a decline in categories like “cares about people like you,” which would speak to a broader set of challenges.
The reason I’m skeptical that the email scandal is the cause of Clinton’s problems is because it’s not as though she entered the race with a squeaky-clean reputation. Instead, given how long the Clintons have been in public life and their reputation for playing at the edges of the rules, it might take a fairly bad scandal to capture the public’s imagination and produce much “value over replacement scandal.” Search Google Trends for “Clinton scandal”1 and you’ll find a spikes of public interest around an actual or alleged scandal involving the Clintons once or twice per election cycle; the current one does not particularly stand out.
Favorability ratings and press coverage
The email scandal does yield lots of unflattering news articles about Clinton. But, as is the case for the public, the press isn’t lacking for real or imagined Clinton scandals to pick through (see the activities of the Clinton Foundation, Clinton’s paid speeches, Benghazi, etc.).
Even if there were no Clinton scandals, however, she’d probably still be receiving fairly negative press coverage. The campaign press more or less openly confesses to a certain type of bias: rooting for the story. Inevitability makes for a really boring story, especially when it involves a figure like Clinton who has been in public life for so long.
Instead, the media wants campaigns with lots of “game changers,” unexpected plot twists and photo finishes. If the story isn’t really there, the press can cobble one together by invoking fuzzy concepts like “momentum” and “expectations,” or by cherry-picking polls and other types of evidence. The lone recent poll to show Sanders ahead of Clinton in New Hampshire made banner headlines, for example, while the many other polls that have Clinton still leading, or which show Sanders’s surge slowing down in Iowa and nationally, have mostly been ignored.
As a result, the flow of news that Americans are getting about Clinton is quite negative. Indeed, the steady decline in her favorability ratings seems consistent with the drip, drip, drip of negative coverage, as opposed to the spikes upward and downward that one might expect if any one development was all that significant to voters.
There’s probably not a lot the Clinton campaign can do about the tone of her news coverage for now. The master narrative about her being “inevitable” — and the incentive the press has to frame a better story by portraying the Democratic race as “unexpectedly” competitive — is too powerful. Winning Iowa and New Hampshire might help Clinton, but even there, she’ll have to beat the media’s high expectations in addition to Sanders and her other opponents. This can be true to an absurd extent: When Walter Mondale beat Gary Hart 49 percent to 16 percent in the 1984 Iowa caucus, it was Hart who emerged with the favorable press coverage because he had done a bit better than polls and journalists had anticipated.
The good news from Clinton’s standpoint is that the master narrative will be reset if and when she eventually wins the nomination, Republicans choose their candidate and the general election begins. There are things that could work for Clinton (like that most journalists are politically liberal) or against her (Clinton has a fraught relationship with the media) in the tenor of her press coverage, but overall the coverage will probably have a more neutral effect in the general election than it does in the primaries. Barring a genuine “game changer” like a recession or Donald Trump winning the Republican nomination (don’t bet on it), the general election is liable to be close, and either Clinton losing the race (end of the Clinton dynasty!) or winning it (first female president!) would be a good story.
In fact, it’s fairly common for candidates, especially “inevitable” ones, to see their favorability ratings improve once the primary campaign is over. The chart below shows favorability ratings for Al Gore over the course of the 2000 campaign, for instance.2 Gore’s net favorability rating began at about +20 percentage points when he was a fairly popular vice president, but declined to a low of +7 by November 1999, when there were polls showing him “unexpectedly” falling behind Bradley in New Hampshire. But Gore’s favorability rating recovered back to +20 by November 2000.
George W. Bush’s favorability ratings, meanwhile, reached their low point of the 2000 campaign in February and March of that year, when McCain was doing “unexpectedly” well against him. I don’t have favorability ratings for George H.W. Bush in 1988 handy, but Michael Dukakis — a better story, since his nomination was more “unexpected” — led in head-to-head polls for much of the spring and summer before Bush overtook him. In 2012, Mitt Romney’s favorability ratings steadily worsened over the course of the nomination race as candidates like Rick Santorum ran “unexpectedly” well against him, but they recovered to about breakeven by Election Day.
Much of this logic applies to the Republican candidates too, by the way. The primaries are tough on a candidate’s favorability ratings: Either candidates are getting criticized by other members of their party (that mostly hasn’t happened for Clinton yet; if it does, her numbers could get even worse), or if the race isn’t so competitive, they’re getting a difficult time from the press. Be especially wary if Clinton wraps up her nomination before the GOP candidate does; that could temporarily inflate her head-to-head numbers against her prospective Republican opponents.
Clinton remains the heavy favorite to win the nomination
So, Nate, you’re telling us that everything that’s happened in the Democratic campaign so far is irrelevant? That’s not what I’m saying. Clinton would rather have stronger favorability ratings than weaker ones, no email scandal, and Sanders behind her in every poll rather than all but one of them. If Joe Biden actually enters the race, I’d lower Clinton’s chances a bit; then again, I’d raise them if Biden doesn’t.
Rather, it’s that the challenges Clinton’s campaign has faced — and the media narrative about an “unexpectedly” competitive race — is exactly what you’d expect relative to the historical experience of campaigns like hers. Being the heavy favorite to win the nomination does not mean you expect to go 162-0.