Bookmark and ShareShare
Friday, February 26, 2010

Polls Make Their Appearance at the Healthcare Summit

Good news: the voice of the American people is being interjected into the debate on healthcare.  More and more elected officials and pundits quote survey data on this important topic. This includes a number of references to polls at Thursday's healthcare summit at the Blair House in Washington. The not so good news: some of the references to polls have puzzling or unknown origins.

The McConnell References
Republican Senate minority leader Mitch McConnell of Kentucky had obviously been well prepared by his staff at the healthcare summit. He made two separate, albeit nearly identical, references to polling. Both included a reference to our Feb. 23 USA Today/Gallup poll on reconciliation. Which I’m happy to say he quoted accurately. However, he also included difficult-to-verify references to aggregates of polling on healthcare legislation.

Here’s the first of McConnell’s statements:

One thing I think we need to be acutely aware of, ladies and gentlemen, we are here representing the American people. And Harry mentioned several polls. I think it is not irrelevant that the American people, if you average out all of the polls, are opposed to this bill by 55-37. And we know from a USA Today/Gallup poll out this morning, they're opposed to using the reconciliation device, the short-circuit approach that Lamar referred to, that would end up with only bipartisan opposition by 52-39.

And here’s the second:

But we know from the polling that's been done in this country how the American people feel about this 2,700-page bill. We know how they feel about it. This is not a close call. If you average all of the polls in America, we know that the American people oppose this proposal by -- on an average of 55 to 37 percent.

They have also been asked -- and we keep reading in the newspaper that where we're headed next is to the reconciliation approach. Well, Gallup also asked that question. It explained to the American people what it meant so they understood what this word that we use around Washington actually means. And in the Gallup poll, the American people were opposed using that, 52 to 39 [percent].

As a point of interest, New York Times columnist David Brooks included a similar reference in his Feb. 22 column:

As the year went on, health care reform grew more unpopular. If you average the last 10 polls, 38 percent of voters support the reform plans and 53 percent oppose.

Sen. McConnell references “all of the polls” and “all of the polls in America.” Brooks references “the last 10 polls.”

The first issue is the indefinite sourcing. Neither McConnell nor Brooks mentions where their data came from.

I’m pretty sure that Sen. McConnell didn’t stay up late at night searching the web for all recent polls on healthcare. We assume McConnell has a diligent staff.  And that staff may, indeed, have found all polls ever conducted in America on healthcare legislation. But we don’t know that. The reader has no way of ascertaining to which polls McConnell's statements refer.

I’m not sure if Brooks has a staff  --certainly not one the size of the Senate minority leader’s at any rate.  Brooks may do his own research.  But again, the reader of his column is left wondering which “last 10 polls” he is referring to.

There are other issues involved with this type of attempt to average across polls. We are dealing here with polls asking about a complex policy issue. This can be distinquished from polls measuring presidential approval or election ballot polling. Analyzing polling on policy issues demands that the researcher be very cognizant of the exact question wording involved in each question being considered. In the current case of healthcare legislation, pollsters sometimes use widely differing descriptions of the legislation they are asking respondents to rate. Pollsters also have significantly different ways of framing their questions about healthcare legislation. Attempting to compress all of these results into a simple average loses a lot of information derviable from patterns of responses to specific ways in which the questions are asked. 
The best approach is to actually examine poll questions one by one -- taking into account how each question is phrased, and how the differences in wording relate to the responses obtained.

Now, this is not to say that the analyst shouldn’t attempt some type of overarching conclusion after examining a variety of data on a policy topic. In this instance, as I have indicated again and again, a careful examination of polls on a case-by-case basis does in fact underscore the conclusion that Americans tend to be more opposed than in favor of the type of plan Obama and the Democrats have put forth -- across a wide variety of ways of asking about it.

Another sensitive issue with blind aggregation of all available poll data on a topic concerns underlying differences in the quality of the data. Polling experts who examine polls one by one usually make decisions on how much attention to pay to any individual poll. These decisions are based on methodology, mode of interviewing and other technical details. Simply throwing all available polls on a topic into a giant blender and spewing out some overall average is often ill-advised. I can’t imagine that a medical researcher would simply treat any and all studies on the efficacy of Vitamin D in preventing heart attacks exactly the same. The researcher would be very careful about the types of studies that he or she took seriously and which ones were discarded.

Obama Weighs In
Sen. McConnell wasn't the only one talking about polls at the summit on Thursday. President Obama made references to polls as well. Here's what Obama said late in the day Thursday afternoon:

I'm just going to make this remark, and then I'm going to call on Patty Murray -- I'm going to save the two lions of the House here for the end -- because there's been a lot of comments from every Republican about the polls and what they're hearing from their constituents. And, as I said, I hear from constituents in every one of your districts and every one of your states. And what's interesting is actually when you poll people about the individual elements in each of these bills, they're all for them. So you ask them, do you want to prohibit preexisting conditions? Yes, I'm for that. Do you want to make sure that everybody can get basic coverage that's affordable? Yes, I'm for that. Do you want to make sure that insurance companies can't take advantage of you and that you've got the ability, as Ron said, to fire an insurance company that's not doing a good job and hire one that is, but also, that you've got some basic consumer protections? Yes, we like that. So polls I think are important in taking a temperature of the public. If you polled people and asked them, is the system working right now and should we move forward with health reform, they'd also say yes to that.

It is true that surveys (see here and here and here) have shown that a majority of Americans favor a number of specific elements that would be included in a Democratic healthcare bill. As I’ve discussed a number of times, however, it's not clear that increasing the awareness of the public to the existence of these elements would cause a major shift in support for healthcare legislation.  The significant consideration for many Americans, based on what I see in the data, is the big picture. Americans are concerned about the cost of all of these elements put together and the role of government in managing a new comprehensive attack on healthcare problems. In this instance, it appears that the whole is more than the sum of its parts.


Anonymous said...
February 26, 2010 at 5:00 PM  

Question: is there any polling which breaks down the opposition to the health care bill by ideology? What I'm getting at is that some of the opposition seems to come from liberals who are upset that the bill "doesn't go far enough". The people at MoveOn, DailyKos, and Huffington Post who want a public option or single payer. What portion of those recorded as "opposed" fit this profile?

Anonymous said...
March 2, 2010 at 6:53 AM  

I agree with Anonymous (Feb 26) who was pointing out that some of the opposition to the current health care bill comes from the left -- even though the press commonly ignores this possibility.

Gallop would be doing a tremendous service to the country by conducting a poll that gave respondents three and only three options:
1. Scrap the bill and start over; 2) Pass the bill with the president's amendments; 3) Pass Medicare for anyone who prefers it, funded by the revenue generated from the elapseed Bush tax cuts.

It's hard to imagine a more timely and important public opinion poll.

Anonymous said...
March 2, 2010 at 9:42 PM  

I don't understand this statement in lew of the polls you have shown in it. The Newsweek polls asks a series of questions about what is in the actual bills which the people approve of most of it. Then re asks the question of if they now favor the bill and the numbers shift by almost 10% to have more favor than oppose it. The Kaiser poll shows that most americans both do not know what is in the bill and favor most of the things that are in it. Isn't that significant? Here is your statement "It is true that surveys (see here and here and here) have shown that a majority of Americans favor a number of specific elements that would be included in a Democratic healthcare bill. As I’ve discussed a number of times, however, it's not clear that increasing the awareness of the public to the existence of these elements would cause a major shift in support for healthcare legislation."

Anonymous said...
March 4, 2010 at 3:27 PM  

I think that basic part and parcel of each and every healthcare poll should require the respondent to state two things:

1] Does s/he currently have any healthcare coverage at all.

2] Whether your healthcare coverage subsidized wholly or partially by a third party.

Off the top of my head, I would expect most of those against a new healthcare bill already are covered to some degree or other. I also have the feeling that most of those covered are not paying the full, retail premiums; i.e., their employer, union, spouse or some other entity is paying or otherwise subsidizing their healthcare costs.

That would make a HUGE difference in an individual's opinion as to the necessity for reform--like a 180-degree shift!

Anonymous said...
March 12, 2010 at 12:27 PM  

I dont think anyone disagrees that we need health care reform. How to pay for it... that's the question. Maybe all of Gov. should have to take a pay cut to cover the people who does not have ins. Also put a cap on Hospital Charges and Doctor Charges, as well as Prescription Prices. To Make the hard working middle class pay for another hand out is crazy and will eventully break this country.

Post a Comment

Comments are moderated by Gallup and may not appear on this blog until they have been reviewed and deemed appropriate for posting.

Copyright © 2010 Gallup, Inc. All rights reserved. | Terms of Use | Privacy Statement