In a FP diary on Wednesday, echoed by another one by Laura Clawson yesterday, David Nir made the argument that Hillary Clinton's falling favorable ratings are wholly unsurprising and in any case don't "matter nearly as much as you might think".
Without getting into the latter point, is it true that Hillary's falling favorables are nothing but the unremarkable, natural consequence of her transition from respected office-holder to campaigning presidential candidate? Is Laura right to suggest that Joe Biden would experience the very same drop in popularity if he were to wade in as well?
I tried to find out by looking for parallels, and digging up archived opinion poll data about Al Gore in 1999-2000, and Hillary Clinton herself, in 2007-2008. I feel there's not enough directly comparable data to warrant any all too confident statements -- but just enough to feel uneasy about the ones they made.
More (and charts!) below the swirl.
On Wednesday, David Nir wrote:
[T]hey're very focused on Clinton's favorability rating, which has indeed trended negative over this same timeframe. But it's also been headed downward for years--long before anyone ever heard about any email servers--and for a very simple reason: When you leave a high-profile, nonpartisan post like secretary of state to run for elective office, you can't possibly sustain the broad appeal you once did when you were globe-trotting to meet with foreign leaders. As soon as you're back in the muck of the campaign trail, you're going to get viewed through the polarized prism of American politics. For Clinton, it was a predictable development that many did indeed predict.
This contention is a bit problematic, because Hillary's favorables have not in fact "been headed downward for years," as
the link he helpfully provided illustrates. Yes, they did drop significantly once she left the "high-profile, nonpartisan post" of Secretary of State, from close to 60% to around 50% in late 2013. But then they evened out. For about a year, from September 2013 to September 2014, they remained remarkably stable, the trendline never deviating from the 50-51% mark; then, after a slight drop, they hovered around 47% for another half a year. Until this spring. Since then, they've dropped precipitously. They're down at 41% now, and her unfavorables climbed to over 50%. Here's what that looks like when you zoom into the timeline a bit:
In short, the more recent drop can't really be fitted easily into a narrative about how it's just because she isn't "globe-trotting to meet with foreign leaders" anymore. That leaves the second part of David's quote above, however, suggesting that the significant drop in her favorables is nothing but a natural result of what happens when "you're back in the muck of the campaign trail". This is the point Laura picked up and expanded, arguing that Joe Biden would befall the exact same fate:
Vice president is a more politicized role than secretary of State, but Biden benefits from this phenomenon, too. As long as he's not a candidate, he's the jovial sidekick we can all be fond of without thinking too hard about it [..]. Once he's back in "the muck of the campaign trail," his favorables will decline, too.
Al Gore was not exactly known as a "jovial sidekick", but he did already go down the path Laura sketches here: from Vice-President to presidential candidate. So did he suffer the curse of falling favorables as well? More relevantly, did he do so to the same extent as Hillary does now?
What about Hillary herself? As hard as it is to remember now, there was a time after she got back into politics, once she'd settled in as junior Senator for New York, when she was gaining praise even from Republican politicians as a hard-working, constructive colleague. For a brief while, it seemed like she was no longer the polarizing figure conservatives loved to hate. By 2005, not just Gallup but the Fox News polls, too, regularly had her over 50% favorable, and even dipping under 40% unfavorable. Secretary of State she wasn't, of course, and being a Democratic Senator is no "non-partisan" office by any stretch of the word. But still, she was out of the wind a bit -- even as her name recognition remained overwhelming. And that's a combination which describes where she was at directly prior to this campaign as well, and which did not apply to Kerry or Obama before they started campaigning, when their name recognition was low. So how did her return to national politics affect her favorability numbers the last time she started campaigning for President?
Those are the two examples I decided to explore.
First problem: finding comparable polling data. Some pollsters ask respondents to choose between favorable or unfavorable, and leave little room for an in-between choice. You can say you have "no opinion", or claim to have never heard of someone, but there's no neutral ground. Other pollsters have provided a more nuanced choice. The NBC polls asked respondents to choose between very positive, somewhat positive, neutral, somewhat negative, or very negative. That left a much larger share of the electorate expressing a neutral view. So did the CBS/NYT polls, which provided for a softer than usual 'out' by providing options for both "undecided" and "haven't heard enough ... yet to have an opinion".
I decided to pick only pollsters that pressed for a clear answer, because polls with 20+ or 30+ percent undecided are of limited use. I picked the mid-date for when each poll was in the field. In addition, since every pollster has a 'house effect', I decided to only use pollsters for whom I could find data (on Pollingreport.com or the Huffington Post's Pollster site) for all three cycles. Unfortunately, that left me with just three lines of polling: Gallup, ABC/WaPo and CNN. Which provided plenty of data for the previous races but only a narrow selection of just 11 polls from this year. They still show a fairly clear trend, though, and that trend is backed up by the comprehensive HuffPo Pollster trendline.
Here's Al Gore, in the 1999-2000 cycle. Each dot indicates one poll. Ignore the single dot on the zero-line at the end of each of these charts, positioned on November 1 of the election year; those are there just to force Google Sheets to extend the chart frame to the full time range.
Here's how that translate as balance between favorable and unfavorable rating (favorable minus unfavorable):
Surprise #1: Al Gore remained pretty popular throughout. In these polls his favorables only started dropping below 50% in 2002, a couple of years after the election.
Observation #2: Al Gore did experience some initial slippage in his favorables. But it never took him much below 50%; and it took place very early, and soon evened out. In three polls between late May and mid-July 2009 (all by CNN) he had lower than 50% favorable ratings; but from late July on they'd stabilized, and in the subsequent 38 polls in this sample he only ever dropped below 50% twice again, much later in the campaign.
The Al Gore example is interesting because Laura argued specifically that as Vice President getting back into the muck of campaigning, Biden would inevitably see his favorables decline, too - much like Hillary did. Well, Al Gore's favorables did drop, but neither for as long, nor as low, as Hillary's are dropping now. (On a sidenote, it's a little startling to see Gore's favorability ratings at the time being described as "lackluster", when they were so much better than Hillary's now.)
Here's Hillary Clinton in the 2007-2008 cycle (click to enlarge):
She obviously struggled with worse ratings than Gore throughout. Even so, in the entire sample of 44 polls, there were only nine in which she was 'underwater', and only three in which her unfavorability rating exceeded her favorability rating by more than three points.
Moreover, most of the damage came very early on - as in: earlier than where we're at now. In Gallup polling, she dropped from 58% to 45% favorable between February and April 2007, but that was the lowest she ever went in that poll. CNN and ABC/WaPo similarly showed her hitting bottom in February-March 2007 -- before the campaign even started, really -- and then recovering somewhat again, to only ever reach a lower favorability or higher unfavorability rating twice more during the whole campaign.
This is what I've got for Hillary so far in this cycle (click to enlarge):
Very few data points, obviously. But what there is looks pretty similar to HufPo Pollster's chart of all favorability polling on Hillary this year.
This is why I'm a little worried, and find David's and Laura's dismissal of Hillary's dropping favorables as unsurprising and fairly meaningless a little all too glib. While both Gore in 1999 and Hillary in 2007 experienced some initial drop in their favorability rating, it had already turned it around by the time July came along. This time, Hillary's drop continues - with little sign, not in this chart with its scarce data but also not in the comprehensive Pollster chart, of bottoming out. Neither Gore's nor Clinton's favorability rating dropped as far as Hillary's rating seems now to be dropping. And of course, just to point out the obvious - neither Gore in 2000, nor Hillary in 2008, actually got elected.
Someone who did get elected, of course, was Barack Obama in 2008. At this point in time, during that campaign, he was enjoying a 20-point spread in his favor in his favorability ratings: around 50% favorable, 30% unfavorable. Incomparable, perhaps; he didn't have any of Hillary's baggage (though that might make you wonder about how smart it is to nominate someone with that much baggage). Same was true for Kerry in '04 and Bill Clinton in '92, who were largely unknown before the campaign. But the cases that seem to be more comparable don't make Hillary's current numbers look good either.
None of that proves by any measure that she's doomed, or that the downward trajectory of her ratings won't be reversed at some point in time. Her favorables have rebounded more than once before in her long time in politics, after all, so it wouldn't be the first time. Gore's 2000 numbers and Hillary's own 2008 numbers suggest that an initial drop is likely to stabilize and eventually, in the heat of the general election campaign, turn around. But a lot might depend on how low a point they will stabilize at, if the eventual rebound is to prove enough - because do you really want to bet the house on the Republican candidates continuing to do even worse? And with Hillary's current numbers, we're in uncharted territory.