What Presidential Speeches Can and Can't Do

I very much liked Matt Yglesias's approach to reacting to the president's oil spill address.  Yglesias:

The most important thing to keep in mind about the sort of “major” presidential speech we saw last night is that they don’t matter. At all. They don’t move votes in Congress. They don’t move public opinion. The bully pulpit method of governance doesn’t work. And that’s about the best I can say about Obama’s speech—even if it had been much better, it wouldn’t have done much good.

Kevin Drum disagrees:

It's possible to take this political-science-inspired view of the presidency too far..This is tantamount to saying that presidents shouldn't bother communicating to the public at all. But does anyone really believe this? Even the political scientists whose research suggests that presidential speeches don't move the public opinion dial much? I doubt it. A single speech may not have much effect, but let's face it: a single anything doesn't have much effect. Last night's speech was part of a much broader communications strategy from the president, and that broader strategy does make a difference in the long run. Obama had a chance to move the dial a little bit, to shift the topic of elite conversation, and to send a clear signal about what he supports and what he doesn't. Those are useful things, and he should have done a better job with them. 

OK, who is right?


I'll refer back to three posts, all again from around last year's health care address to Congress.  Here's John Sides; here's Brendan Nyhan; and here's, er, me.  John concentrated on showing that the speech was unlikely to have much effect on Obama's approval ratings.  Nyhan argued that Obama was unlikely to change public opinion about the underlying issue.  And I mostly talked about other things presidents do that matter a whole lot more.  I think all three of us were proven correct by events.  However, it's important to note that our comments were in each case limited to debunking specific claims of the potential effects of presidential speeches.  I mean, in my case I was reacting to a claim that the speech would be "the biggest moment in [Obama's] presidency.  Nyhan was taking on claims that the speech would be "a game-changer" and that the president would be able to get "the American people..to put aside some of the ridiculous falsehoods that have been perpetrated these past few weeks."  That's the context of the kind of debunking we did in our posts.

It's too strong to jump from that to claims that presidential speeches don't matter at all.  There are at least three possible effects worth considering:

1. Priming, agenda setting, and framing effects.  The president can't really convince people to change their minds about things.  But he can get people talking about something new, and it's certainly possible that the president would be better off if people focus on the energy/climate bill instead of the continuing spill. 

2.  To the extent that partisan Democrats were unhappy with the White House's response to the spill, it's certainly plausible that they are open to adapting the president's point of view, and giving it in a high-profile format might make them more likely to listen.  Beyond that, beyond approve/disapprove, partisan Democrats may be open to learning whatever it is that the president wants to teach.  In general, I'd say that it is certainly possible for presidents to affect elite opinion, which can then filter down to everyone else, but it's unlikely that a president can do much of that in a single speech, or even in a series of speeches.  However, it's fair to note that there is some evidence (in my opinion, very limited, but still) that presidents can influence Congress by mobilizing public opinion.


3. Presidential speeches can matter because they can send signals to Congress and to other Washingtonians about what presidents want.  Remember, there are multiple audiences here, and some of those audiences are less interested in the poetry of the speech than they are about substance.

One other important function of presidential speeches: regardless of their effects, just the act of producing presidential speeches tends to have effects.  Once the presidential mouth is going to be engaged, conflicts within the administration, or between the administration and interest groups, may suddenly need to be resolved instead of being allowed to continue.  If I recall correctly, the West Wing once did a pretty good episode of how the State of the Union speech becomes a deadline that forces action on policy initiatives.  Of course, just because these issues get "resolved" at that stage doesn't mean that the result becomes actual public policy, but it does tend to lock the president into commitments that, without the speech, he could have avoided.

So.  On reflection, I think Yglesias is a bit too strong when he says that these speeches "don't matter.  At all."  It's more accurate to say that they only affect public opinion on the margins.  Sometimes, the margins matter!  And there's more to it than just how much public opinion can be changed.  But, yeah, I think for the most part that Yglesias is on safer ground in downplaying the speech than pundits and activists are when they hype these things.  The practical difference in a great speech and a pedestrian one just isn't that great.