Clinton's Post-Convention Bounce: Will It Stay or Go?

Clinton's Post-Convention Bounce: Will It Stay or Go?
AP Photo/Andrew Harnik
Story Stream
recent articles

I’ve been holding off writing the post-Democratic convention state-of-the-race update for two reasons.  First, polls continue to pour in, as pollsters seek to measure the full effect of the two conventions.  Second, and perhaps more importantly, the news cycle is moving very, very fast right now; it’s hard to find a good time to stop and write something that won’t be out of date 15 minutes after publication.

I think the best way to approach this is to give the “normal” take on the state of affairs, then the “Trump-centric” take.  Which one you think is closer to reality probably depends a fair amount on your prior views of the race; my sense is that the truth is somewhere in the middle.

So the normal take on the race begins like this: Hillary Clinton’s pre-convention, post-nomination lead peaked at 6.8 percentage points on June 28, when she led 46.4 percent to 39.6 percent. It settled down to about 4.5 points, before dropping to around 2.5 points in the wake of FBI Director James Comey’s press conference about her emails. Donald Trump got a decent bounce of about three-to-four points out of the Republican convention, which culminated in him seizing a 0.9 percentage-point lead on July 28.  This may, however, understate the size of his bounce, as the Comey revelations may have begun the process of bringing low-hanging fruit on board.

But Clinton clearly received a nice bounce out of her convention as well. She now leads by 5.1 points in the RealClearPolitics Average, 46.5 percent to Trump’s 42 percent.  You may point to the Los Angeles Times poll as an outlier pulling her average down, but Reuters played a similar role during Trump’s convention.  That’s just how averages work.

Likewise, you may point to the Gallup poll, which received a large amount of press, showing that the Republican convention made voters less likely to vote GOP.  There are two problems here.  First, it may well be that more voters became less likely to vote for Trump on balance, but if these were already anti-Trump voters (or pro-Trump voters whose minds were not changed in the end), it wouldn’t make a difference in the outcome.

Second, it isn’t really clear how much these things predict anything.  To wit, the net difference in performance between the two conventions is 19 points in the Democrats’ favor, according to the Gallup poll.  The last time we saw a gap that large was 1988, when it was also 19 points in the Democrats’ favor.  Needless to say, that ended poorly for them, with George H.W. Bush winning the White House.

The bottom line is that Clinton is up in the polls now.  Will it last?  Second convention bounces tend to fade; it did for the Republicans in 2004 and 2008 (even before the Lehman Brothers collapse); it did for the Democrats in 1996.  Overall, since 1968 the in-party has tended to run behind its showing even two weeks out from the convention, but this is not a hard, fast rule.  In this regard, Clinton’s lead of about four-to-five points should probably make Democrats nervous.

But this is where we move into the Trump-centric view of the race. In a normal world, the news cycle would probably be stepping on Clinton’s bounce.  We’d be hearing about the weak GDP report, the news of payments to Iran, and Clinton’s less-than-nimble responses to Chris Wallace’s questions regarding her emails in her interview on “Fox News Sunday.”

Instead, Trump engaged in a fight with Khizr and Ghazala Khan, who had given a passionate denunciation of the GOP nominee in Philadelphia. Trump’s supporters complain that the Khans have political agendas the media is not exploring, or protest the supposed unfairness of requiring Trump to refuse to fight back. I’m reminded of an old rule from practicing law: When the other side puts an extremely sympathetic witness on the stand, unless you have absolutely damning evidence against them, don’t cross-examine.  Get them off of the stand and away from the jury as quickly as possible.

By ignoring this advice, Trump has turned the Khans – who, as parents of a Marine who died in Iraq, are awfully sympathetic witnesses – into mini-celebrities. A one-day story has become a one-week story.  This was followed by Trump refusing to endorse Speaker Paul Ryan and Sen. John McCain; a dustup over the fire marshal limiting space at a rally in Columbus, Ohio; an assertion by Trump that the election might be rigged; endorsements of Clinton by prominent Republicans; and even reports that Trump (possibly jokingly) went after a crying baby at the Columbus event.  There are now rumors that campaign manager Paul Manafort has basically called it a day, and that Rudy Giuliani and Newt Gingrich have been called in to stage something of an intervention with Trump – though those are just rumors, which the campaign denies.

In short, it does feel like the wheels are coming off of the Trump train.  I can’t help but remind myself that we’ve been here before, but we’re entering the heart of the campaign, when people are finally tuning in.  This time really is different, and Trump doesn’t seem to understand (or care) that if the campaign is about him, he has a very good chance of losing.

The FiveThirtyEight polls-plus model currently gives Clinton a 66 percent chance of winning. That may seem comforting to Trump opponents, until you recognize that models (not Nate Silver’s) gave Britain about a 90 percent chance of remaining in the EU, or gave the Golden State Warriors about a 95 percent chance of winning the NBA championship late in the finals.  Things with a 33 percent chance of happening occur all of the time.

With that said, the model does not include an express variable for a candidate who says the sorts of things that Trump says. It also does not include an express variable for a candidate with the sort of baggage that Clinton has.  And it certainly does not include a variable for the probability that Wikileaks will post the contents of Clinton’s hard drive, or Clinton foundation emails, to the Internet immediately before a debate, or a week before Election Day.

In the end, we do have to remember that it is only August, that the “fundamentals” of elections – the economy, presidential job approval, incumbency, and so forth – do not give Democrats much, if any, advantage, and that there are plenty of events that could derail either campaign.  While it certainly feels as though the Trump campaign is approaching a precipice, in reality that probably still lies a bit down the road. With that said, however, a precipice probably is what awaits him on his current path.

Sean Trende is senior elections analyst for RealClearPolitics. He is a co-author of the 2014 Almanac of American Politics and author of The Lost Majority. He can be reached at Follow him on Twitter @SeanTrende.

Show commentsHide Comments