< Pollbearers


Friday, July 24, 2009

So we know that political consultants on both sides of the health care debate are cherrypicking words to craft the narrative. Are pollsters doing the same thing? This week, new polls declared that public support for President Obama on health care is dropping, and media coverage of those polls made it seem as if a health care overhaul was in serious trouble. So how should a media consumer interpret this polling data, especially when it comes to a complicated and messy policy issue like health care? We asked Nate Silver, who crunches numbers and analyzes polls at his site, FiveThirtyEight.com. Silver says pollsters, like political consultants, know that very small changes in wording sometimes elicit a 15- to 20-point difference in a response. Here’s an example. Silver says if a pollster confuses the words “government-provided insurance plan,” which is the public option, with “government-provided health services,” which tends to connote, quote, unquote, “socialized medicine,” that will affect the poll’s results. So which poll should we pay attention to, if any of them?

NATE SILVER: There is a great poll put out every other month, a tracking poll by the Kaiser Family Foundation, and they're not-for-profit. And they're very careful with their wording. In fact, a lot of what they'll do is actually phrase things in two different ways and see how it tests. For example, they have a question about the public option, where in one case they use the phrase “like Medicare,” which tends to be favorable to that. In the other case, they omit it. It turns out not to make a lot of difference, but two or three points. That’s honest polling, when you’re willing to like ask the same question both ways. That’s the one poll that I will look at, in part because they have been asking the same questions every other month, so February, April, June, so we can really kind of make an apples-to-apples comparison.

MIKE PESCA: Is it the case that the polls out now are confusing or worthless, or has the public really not changed its – from what you can tell – really not changed its general opinion about the health care proposals working their way through Congress?

NATE SILVER: Oh, it certainly seems to me like it’s lost somewhere five points, plus or minus a few. That seems clear enough. But, you know, you have different polls who have different partisan leans. Some of the Rasmussen polling that’s been out, for example, that says, you know, 51 percent opposed now, you know, that polling was pretty decent during the campaign, but ever since the campaign it’s been much more Republican-leaning than most polls on most issues, sometimes by a matter of 10 or 15 points, you know. And if something is consistently off in that one direction, then you tend to give it less weight in terms of any individual issue.

MIKE PESCA: You said something to the effect of people can be asked about health care and they don't understand the question. Now, it could be the case that a pollster asks, what do you think of the health care plan, and that person’s mind will immediately jump to, oh, rationing and long lines, I'm against it. That’s not necessarily illegitimate. If people are convinced that that’s what the health care plan means and they tell pollsters, I don't like it ‘cause it means rationing, it still could be a good poll, right?

NATE SILVER: Yeah, it could still be a good poll. What it’s measuring exactly is maybe not support for that policy but in terms of who’s winning the messaging war. I mean, what I would like to see is some kind of daily tracking poll, whether it’s Gallup or someone else, where you ask kind of an abstract question like, do you think health care reform is necessary, and then maybe ask, do you favor the bill that’s currently before Congress, and see if those numbers start to diverge.

MIKE PESCA: You started out as a baseball stats guru. There is a lot of raw data there. Then you reinvented yourself as a political guy. There’s a pretty good apparatus to measure baseball. We know what the definition of winning is in politics - lots of polls about each individual state. Is there a good enough apparatus to measure the goal of a policy - will it pass or won't it - good enough polling, good enough raw data there?

NATE SILVER: You know, one thing we're starting to do at 538 is, for example, on the Climate Bill, which was passed, barely, by the House and hasn't passed the Senate, we just ran a bunch of analysis of what actually motivated that vote in the House. And it actually matters how much carbon does your district emit, you know, what’s the poverty rate, the unemployment rate? What’s your partisan ideology? How much did Obama or McCain win by, and how much lobbying money did you take from both the nuclear and alternate energy industry, as well as, obviously, big coal? That’s the bill, though, where ideology takes a back seat to local considerations. With health care it’s not quite as much like that. There are some areas where more people are uninsured. Ironically, some of those areas are areas where you have some of these conservative blue dog Democrats. It tends to be more kind of rural poor areas, by and large.

MIKE PESCA: So climate change may be not blue or red, but inhalers and exhalers. NATE SILVER: Yeah, exactly.


MIKE PESCA: Now, did you look at age with the, with the current health care bill?

NATE SILVER: Well, that’s one thing, and there was a little interesting anecdote put out by an economist named Tyler Cowen on his blog this week, and he suggested that if the Democrats hadn't passed Medicare under Lyndon Johnson, then you'd have all these old people who are clamoring for health care. And you do see support for the bill tailing off with people who are 65 and up, and they have no reason to change the system. I mean, Medicare has done a great job. So in that way, it’s almost been a victim of its own success, I think. That’s a real worry. I mean, it’s usually a key part of the kind of tapped-in policy constituency who will actually watch the evening news, still, and will watch C-SPAN and will kind of write their congressmen, and they don't have so much incentive to be, to be engaged here. You know, maybe Obama should challenge the public and say, okay, if you don't believe in any kind of public health care, we'll take away Medicare and pay down the debt a whole bunch here.

[MIKE LAUGHS] But I don't think he’s going to be willing to do that.

MIKE PESCA: “Oh, you hate all public options, do you?”

NATE SILVER: Yeah, yeah.


MIKE PESCA: “How about this one?” Nate Silver, great as always. Thank you.

NATE SILVER: Thank you.

MIKE PESCA: Nate Silver is the founder of FiveThirtyEight.com. This is On the Media from NPR.