Friday, March 26, 2010

Still Can't Look at Polls

Does it matter what instapolls after health care passes say?

Mostly, no. 

Nate Silver surveys the polls taken over the week, and finds...well, really I don't care what he finds. 

Look, even if we're just talking about the fall elections, it doesn't matter what people think -- or, more to the point, what they'll say to pollsters -- on March 26.  It just doesn't.  No one knows what kind of issue health care will be in November.  As Ezra Klein says, the midterm elections will probably be about the economy, not about health care.  But you know what?  That "about" is a tricky little animal, there.  There's no guarantee that candidates will be talking about the economy.  But whether they do or not, the economy is going to be what drives things, because it always does.  If the economy is bad, people are going to think that Obama is a lousy president and that Democrats are losers, and therefore are going to respond negatively to questions about health care reform.  Whether that produces an election in which candidates are talking a lot about health care or not will be driven by internal polling and focus groups (and, to some extent, by bargaining and competition within party coalitions), but the bottom line for election purposes is that single issues, even something that seems as momentous as the health care act, don't really push a lot of votes. 

Nor can one poll the alternatives.  Silver supposes that at this point, it's better for Democrats to have passed the bill than to have failed to pass it, and I strongly agree with that -- failure would, indeed, likely have demoralized Democratic activists and perhaps even directly demoralized voters.  He also supposes that:

I think if you polled Democratic strategists and they were being honest, they'd probably concede that -- electorally-speaking -- Democrats would have been better off if they'd found a different direction last year, focusing perhaps on financial reform and then only turning to health care if their numbers warranted it.
I don't know what strategists might think -- they have a professional bias in favor of believing that elections are more manipulable than academic research has found -- but my reading of the elections literature, along with the way Congress and the media work, suggest to me that it would be foolish to choose legislative strategy in 2009 based on the 2010 elections.  People just don't care that much about issues.  And it's impossible to guess at the alternative scenarios. 

A couple other things to add.  First, as I've argued, thinking of health care reform as a choice for Obama and the Democrats gets it wrong.  Obama was able to win the nomination only by promising to make health care reform a priority.  He, and virtually every Democratic candidate for Congress, campaigned on health care reform.  Presidents don't take office, and Members of Congress don't take office, with blank slates; they take office constrained, and often severely constrained, by the promises they've made while running.  If Obama had abandoned health care reform, he would have broken promises and lost the support of his election coalition.  Presidents can do that sort of thing, but it imposes high costs.

The second thing, and what I've been harping on all week, is that it's far too early to know how health care plays out -- mainly because no one has experienced the perceived good and bad about the new law.  Medicare recipients have all sorts of vague fears about it now; over time, those will be replaced by tangible changes, including I guess rebate checks this year for those who are in the donut hole.  On the down side, if Advantage cuts do harm customers, they will be very aware of it.  And then there are the scare stories to come, in which Republicans blame any terrible health or insurance outcome, including raised premiums, on the new act.  The Democrats are betting that all of that works out, on balance, in their favor, but there's really no way to know right now -- and polling is not the way to get at it.  The instant reactions people have just aren't likely to predict how they'll feel about health care reform six months (or years) down the road, and they're even less likely to predict how those reactions will affect their voting behavior.  One thing I could guess, however, is that the benefits of financial regulation are highly unlikely to be rewarded by voters, since those benefits are largely intangible to most voters (yes, I know, a new financial panic would be plenty tangible, but no one ever rewards pols for avoiding a disaster before it starts).

So: the effects of health care reform on the 2010 elections are apt to be small, consisting mostly of avoiding disaster for the Democrats if the bill had failed; beyond that, it's too soon to know what small effects there will be, and polling is the wrong way to get at those effects.


  1. The first year of Obama's presidency was the ONLY time health care could get passed. The only possible exception to that would be the first year of his second term, if you won and with a good margin. Obama and the Dems made the right decision last year to tackle health care. The country needed it.

  2. evie is right. For years now you need 60 Senate votes to pass any bill Republicans don't like, especially anything big and progressive. The Democrats grabbed the chance when they briefly had 60 Senators.

    All the stars were aligned just right and the Democrats just barely passed a decades long priority. Any thought that was the wrong priority is incredibly short-sighted. It may not be ideal for the 2010 elections (hard to say), but it was a fleeting and rare opportunity to do something big.

    Also, the endorphin rush from this success may just give the Democrats the confidence to dump the filibuster at the beginning of the next term. That would help the Democrats more than a Senator or two.

  3. Jon,
    Isn't there a problem with this argument?

    Issues don't matter to voters. Check.
    Your second claim here, though, is that elected officials are constrained by their promises. If voters don't care about issues, then why would pols be constrained? Does the constraint come from the media, who must have small minds and a love of hobgoblins?

  4. Matt,

    (Most) issues don't matter to voters, directly, but they do matter to groups (interest groups, party factions), and groups can pay close attention to what pols are doing -- and they can influence voters, directly and indirectly. Does that fill in the gaps I left?

  5. Jon, I agree that we really can't know the effect, if any, of passing health care on the next election. That said, I'm still somewhat interested in how this is polling today, because it offers some lessons for officeholders. The polls suggest (sorry -- I peeked!) that Democrats got a modest bounce for passing the bill. Well, that's actually good for them to hear, isn't it? No, it doesn't tell us anything about November, but the fact that the public likes them a bit more for actually passing one of their main priorities, even if it's a controversial one, should encourage future legislative progress, one might think. Or it at least offers a rejoinder to the panicky let's-back-off-our-number-one-priority strategy that Dems were flirting with in the wake of Scott Brown's victory.

  6. Seth,

    Well, if one is an advocate trying to convince pols that they should do what you want...well, sure, then the instapolls (that come back in the correct direction) are useful. They're also useful if you want to know things such as how popular the president is right now! Mostly, I'm just saying that they don't add anything to our guesses about the eventual popularity, in November or down the road, of the policy.

  7. I have to disagree with this thesis. Sure, polls taken in March don't directly affect the outcome of elections in November, but there are countless indirect effects. Even in March, congressmen are still deciding whether to retire, and candidates are still deciding whether to jump into or drop out of races. Polls are one metric they use to help themselves decide. March is in the middle of primary season, and especially in Republican primaries, repeal is a big issue.

    If reform polls well, then Democrats will take heart, be more likely to run for re-election, and be willing to risk running for office, while anti-repeal Republicans will look wise and realistic. On the other hand, if reform polls badly, the field will be pushed in the opposite direction. Democratic candidates will retire or wait for a more favorable cycle to run, while pro-repeal Republicans will feel themselves fully vindicated.

    Sure, politicians are in some sense dumb for listening to polls in March. But it's hard to deny that they do. You can see the effects of this poll-consciousness even now. When reform, the economy, and Obama were polling badly in the fall and winter, Democrats spooked and retired from a bunch of congressional districts that they didn't want to face a tough re-election fight in. Even if the economy improves -- even if polling improves -- by November, these seats will be tough to hold, based to a significant degree on polls from 2009.

  8. Ace-K --

    Yes, yes, that's all correct. See my comment above -- polls are very useful if you want to know what people think now, and you are absolutely right that popularity now has some consequences. Again, I'm talking about using polls now as predictors of polls later (about health care reform, that is), and that's where I think the polls are pretty useless.

  9. Jonathan,

    You're certainly right that polls today are a poor predictor of polls in the future. But you did seem to say pretty unambiguously that polls today just plain don't matter:

    Look, even if we're just talking about the fall elections, it doesn't matter what people think -- or, more to the point, what they'll say to pollsters -- on March 26. It just doesn't.

    Polls in November matter a lot, but polls in March matter to some degree, because they induce politicians to take actions they can't undo e.g. retiring.

    I'd also suggest that polls in March do give us at least a hint about what the popularity of healthcare will look like in November. If HCR were polling at 20%, then I think it would be safe to say that even if all the contingencies that you mention -- closing the doughnut hole, cutting Medicare Advantage -- played out identically, that HCR would be less popular in November than it will be if we assume that it's currently polling around 45%. I think, in other words, that 45% is relevant as a "baseline" popularity for HCR. Depending on the economy and how the specific provisions take effect, that number will move, but in the near term it will hover around 45%, rather than 20 or 80 percent.

    I don't exactly have any evidence for this sense but heck, this is a blog comment -- I don't need evidence.

    I should probably also note that your blog has rapidly become my favorite political site.

  10. If reform polls well, then Democrats will take heart, be more likely to run for re-election, and be willing to risk running for office, while anti-repeal Republicans will look wise and realistic. On the other hand, if reform polls badly, the field will be pushed in the opposite direction. Democratic candidates will retire or wait for a more favorable cycle to run, while pro-repeal Republicans will feel themselves fully vindicated.

    Unfortunately, the current polls are highly ambiguous. Several have shown a modest bump for approval of the bill, but the results are far from uniform, and no poll yet has shown a majority of the public in favor of the bill. People looking at the polls right now are bound to draw different conclusions.

    What the Democrats have been hoping is that they'll have an easier time informing the public what's actually in the bill, now that it has passed. Citizens who are curious about the bill's contents can read several summaries online and in newspapers, which state plainly the provisions, and when they will kick in. People could do this before the bill was passed, but they were less inclined to do so, especially since the contents were subject to change.

    Furthermore, while right-wing propagandists will no doubt continue to spread misinformation about the bill, the wind has gone out of their sails somewhat. They can chirp about repeal all they want, but most of them realize, even if they don't say so outright, that repeal-the-bill is substantially less realistic than kill-the-bill once was. What they're gambling is that they'll encourage enough voters to punish the Democrats, regardless of the actual fate of the bill. This strategy may work, but they are on far riskier ground than they were before the bill was passed. They anticipated this situation, which is why they tried so hard to make sure the bill never passed.

  11. Ace-K hits the Jacobson-Kernell nail on the button. Events today matter for the fall elections via the strategic choices of politicians (who runs, who challenges, who retires or chooses not to challenge) and donors.


  12. You and always tell when one side is worried about the polls when their mouthpieces decide to write about them "not mattering".

    There is nothing "highly ambigiuos" about these results:

  13. Chris Bowers said a few months back that inasmuch as you want to argue the administration should have pursued different policies, it's pointless to do so unless you think that by following those policies THE ECONOMY WOULD BE BETTER THAN WHAT IT IS NOW.

  14. "mainly because no one has experienced the perceived good and bad about the new law."

    This is false. Thousands of small and medium healthcare related businesses and their employees have been suffering for almost a year simply by the extended debate on healthare. Many are going to go out of business or be forced to fire people as a direct result of this legislation. I had a cancellation of an order for a diagnostic apparatus just this morning and was given the "healtcare bill" as the reason for the cancellation...this heathcare provider does not believe they can survive as a direct result of the legislation. BTW, they had been able to deliver this medical service less expensively to the system than a hospital can...and soon, they won't be able to. Kudos to the powerful hospital lobby.

  15. "but no one ever rewards pols for avoiding a disaster before it starts"

    Really? It would seem that operating in a less constrained and challenging political climate is a pretty big reward.

  16. Smart post, but I think financial reform could in fact move voters a great deal -- if the Dems spanked bankers hard in some way. The lingering resentment over the bailout among voters all over the spectrum is enormous. If the Dems took Wall street on in a big way (unlikely, I'll admit), and the GOP resisted, all of a sudden the Dems seize the populist rage the GOP currently enjoys. Big votes there.

  17. I also think you're mistaken when you say "the benefits of financial reform are unlikely to be rewarded by voters."

    You underestimate the rage at the bailout.


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

Who links to my website?