Thursday, September 20, 2012

Late Changes and the Rally Effect

I'm a fan of Harry Enten, but I have to take exception to one thing he said today in his state-of-the-election summary:
But what about a game-changing event? Gaffes like Romney's 47% remarks have shown no ability to move the polls. Debates, as John Sides points out, have historically almost never made a difference. A foreign policy fiasco would almost certainly result in a rally around the leader effect, a la Carter in 1980, before the incumbent gets blamed. There isn't enough time for the "blame" part of the equation to occur before the election.
It's that last one that I need to talk about. It's not actually true that all major foreign policy events yield a positive rally effect, and certainly not true that all fiascoes yield a positive rally effect. This is one that regular readers will recognize because I've talked about it many times: it turns out that sometimes there can be no rally or even a negative rally. For example, there was no rally during the Libyan intervention last year. The biggest slump? The Iran-Contra affair in November 1986 produced a huge drop, with Ronald Reagan's approval ratings dipping 21 points (a dip that lasted for months, in that particular case).

So, yeah, something going wrong in the world could certainly hurt Barack Obama in the final weeks of the campaign. No guarantee it would; some events produce the rally that Enten is talking about even if they objectively appear to be failures.

Looking beyond the rally effect, bad news at home could surely hurt Obama this late as well (just as good news could help him). It's true that the later we get, the less likely some event like this will happen. And while I don't know if there's any specific research on this, I would guess that events at the end of a campaign would tend to get swallowed up by the campaign; the campaign cues everyone to switch to maximum partisanship, and so presumably voters will tend to see events through a partisan screen and therefore tend more than usual to simply reinforce their prior opinions.

But that doesn't mean that events couldn't change the election this late. If Obama really does have a solid lead this late (and I'm still being cautious about that), I agree that it's very unlikely that campaign events would undermine it. Events in the world, however, are another story. It's not especially likely that a major new event would happen, and there's no guarantee at all that such an event would hurt and not help Obama, but I'm not aware of any reason to say that an external event couldn't make a difference late in the game.


  1. Hopefully, this comes across from my email (I have too many emails).

    I think I phrased it incorrectly. What I meant to say was NOT that there would be a rally around the flag effect. Rather that a situation that would cause an incumbent to get blamed eventually would get a rally around first.

    Second, I wasn't thinking of Iran-Contra when I wrote fiasco. I think of that as a scandal. Yes, a scandal would hurt Obama. I meant a war or a terrorist attack (or something along those lines), and the response to that. I get that foreign policy is larger than that, but it was what I was thinking.

    And I'm a fan of yours and appreciate you keeping me honest as they might say :).

    1. Fair enough on Iran-Contra, but it's just the extreme example...also in Brody's database (1947-1986) with initial negative rallies are Soviet A bomb announced, Inchon, China over the Yalu, Gulf of Tonkin, the Pueblo, Tet, the Christmas bombing, Cambodia goes Communist, US shoots down Libyan fighters, and Beirut truck bomb (plus some summits and a couple of other obscure ones). Sometimes there's the pattern of positive initial reaction and then long-term damage, but sometimes the initial reaction is negative.

  2. One thing which may have changed the 1980 election from a fairly close Reagan victory to a Reagan landslide (at least in terms of percentage of the two party vote) was the fact that there was a lot of last-minute talk about the hostages being freed--followed by Iran saying it would not release them.

    Of course it was also terrible luck for Carter that Election Day was around the first anniversary of the hostage-taking and the media concentrated on it enormously.

    Carter would have lost anyway but the increased attention to the hostages issue probably cost the Democrats control of the Senate (a number of the GOP victories were extremely narrow).

    Admittedly, this was not strictly speaking a foreign "crisis" but the sudden dramatization of a long-standing foriegn policy failure.

    1. "Of course it was also terrible luck for Carter that Election Day was around the first anniversary of the hostage-taking and the media concentrated on it enormously."

      Well, that's not exactly luck, is it? The media focusing on it is a decision that someone made.


Note: Only a member of this blog may post a comment.

Who links to my website?