Public Support for the Public Option

UPDATED at 5:45 PM to include Lake/HCAN poll

A major, though by no means the only, substantive point of debate regarding health care reform is whether the plan considered by Congress will include a "public option" -- a government-run insurance program that would compete with private plans. Barack Obama's plan on the campaign trail included a public option: "any American will have the opportunity to enroll in the new public plan or an approved private plan," it said.



Inclusion of a public option is a sine qua non for many progressives, who believe it would lower costs by increasing competition, or who may have an objection to the notion of requiring that people (since health care would have an individual mandate) purchase something through a private, for-profit entity. On the other hand, the public option has drawn the ire of conservatives and industry groups, who believe that it would gobble up profit margins from private industry and that it might have unfair competitive advantages. Both liberals and conservatives seem to acknowledge some possibility that a public option might gradually evolve into a version of a single-payer system; for liberals this is a big plus and for conservatives a big minus. The revised plan released by Max Baucus's Senate Finance Committee on Thursday did not include a public option, although the House's latest version does.

It is worth evaluating polling on the public option, which has begun to be widely cited in the blogosphere, particularly by liberals who believe most of the polling favors them. The balance of this post contains a summary of the five six polls that I am aware of on the public option, which produce widely disparate results and all of which require careful interpretation.

1. Kaiser Family Foundation Tracking Poll

Who They Are / What's Their Angle: A California based non-profit founded in 1948 by Henry J. Kaiser. The Foundation no longer has any association with Kaiser Permanente, which operates hospitals and insurance programs mostly in the South and the West. KFF released numerous materials on the candidates' positions on health care in advance of the 2008 election which generally took a neutral tone. KFF itself has not given money to political candidates, although its employees collectively donated $11,700 in 2008 and $9,550 in 2006 to Judith Feder, a public policy expert who is an adviser to KFF and ran unsuccessfully for Congress as a Democrat in Virginia's 10th Congressional District.

Specifications: 1,205 American adults, including cellphone and Spanish-language samples, conducted from June 1-June 8. The sample was split in half, however, for the two formulations of the public option question as expressed below, and so sample sizes for each one are closer to 600.

Question Wording and Results:

"Now I'm going to read you some different ways to increase the number of Americans covered by health insurance. As I read each one, please tell me whether you would favor it or oppose it [...]

"Creating a public health insurance option similar to Medicare to compete with private health insurance plans."

Favor: 68% (40% strongly favor)
Oppose: 28% (17% strongly oppose)

"Creating a public health insurance option to compete with private health insurance plans."

Favor: 65% (32% strongly favor)
Oppose: 29% (17% strongly oppose)

Discussion: One suspects that KFF is pro-reform, but they seem to have taken care to frame their materials in ways that avoid partisan scrutiny. Their question wording is fairly straightforward but does not include the phrase "government", which might provide more clarity to the respondent about exactly what the public option is. Note, however, that they phrase the question in two different ways: to half their sample they include the phrase "similar to Medicare" and to the other half they don't. Responses were about the same between the two question wordings, although the half that had the Medicare language included were notably more likely to strongly favor the public option. Including a sample of cellphones and Spanish-language interviewers are nice perks. Although the sample sizes are not huge, particularly since the sample was split into halves, KFF found nearly identical results in their April tracking poll.

Non-partisanship rating:
Question wording:
Sample size, sample selection and disclosure:
Overall informativeness:

2. Employee Benefits Research Institute (EBRI)

Who They Are / What's Their Angle: A Washington, D.C. - based nonprofit, which is focused -- as their name implies -- on research on employee benefit programs. I can find no evidence of lobbying activities or campaign contributions by ERBI. They are funded by a largely corporate set of donors such as American Express, Chevron, IBM, Shell Oil and Towers Perrin, although they also receive financing from noncorporate groups like AARP and Blue Cross Blue Shield. An issue brief that EBRI prepared on the public option was neutral to slightly skeptical about it. The poll was conducted in conjunction with Mathew Greenwald & Associates.

Specifications: 1,000 American adults aged 21 and over. Interviews conducted from May 8th through June 2nd.

Question Wording and Results:

"Creating a new public health insurance plan that anyone can purchase."

Support: 83% (53% strongly support)
Oppose: 14% (9% strongly oppose)

Discussion: Information about this poll was a little bit harder to come by than it probably should be. For example, I had to look in a separate press release to find details about its sample size. Nor is it clear that the entire battery of questions was released in ERBI's summary brief. The selection of adults 21+, rather than 18+, is also unorthodox, and is a strange enough choice that I wonder about the other decisions ERBI made in constructing its sample. EBRI's poll was also the only one which did not specify that the public option would be designed to compete with private plans.

Non-partisanship rating:
Question wording:
Sample size, sample selection and disclosure:
Overall informativeness:

3. Consumers Union

Who They Are / What's Their Angle: Not a labor union; Consumers' Union is instead a Yonkers, NY-based non-profit group and the publisher of Consumer Reports. They take a somewhat unabashedly liberal view on health care reform and the poll was released in conjunction with Democratic Senator Chuck Schumer. The Consumers Union' generally spends several hundred thousand dollars on lobbying activities each year.

Specifications: 2,009 American adults aged 18 and over. Interviews conducted from April 2nd to April 6th, 2009.

Question Wording and Results:

"Congress is discussing several ideas to address healthcare reform. One proposal provides everyone, whether insured or uninsured, an additional choice: the option of a public health plan that people can count on to cover what they need at more affordable rates. This option would allow people with good insurance that they like to keep it. Those without good insurance can gain access to reliable healthcare, regardless of preexisting medical conditions, and obtain a consistent menu of benefits. This public plan would be paid for by enrollees. Those that cannot afford to pay the full premiums would be subsidized based on their income.

Please rate your level of support for this proposal."

Support: 66% (33% strongly support)
Oppose: 16% (8% strongly oppose)

Discussion: This is more or less an explicitly partisan poll, both in terms of the organization backing it and in its question wording, which is leading and highly favorable to the public option. The large sample size is nice, but Consumers Union' should have picked more neutral phrasing.

Non-partisanship rating:
Question wording:
Sample size, sample selection and disclosure:
Overall informativeness:

4. Rasmussen Reports

Who They Are / What's Their Angle: Regular readers of this website will be very familiar with Rasmussen Reports, a standalone polling firm that releases a prolific amount of polling data on elections and public policy issues. Past FiveThirtyEight.com analyses have generally found Rasmussen's electoral polling to be quite reliable. However, some observers have questioned its issue-based polling, which frequently tends to elicit responses that are more conservative than those found on other national surveys. Rasmussen Reports' founder, Scott Rasmussen, is a Republican, although neither he nor Rasmussen Reports have appear to have contributed to political candidates in recent years. Nor to my awareness does Rasmussen Reports conduct a significant amount of polling directly on behalf of political candidates.

Specifications: 1,000 American adults on June 12th and 13th. Assuming that procedures here were the same as for other Rasmussen polling, surveys were conducted via the IVR ("robocall") method and were weighted for partisan identification and other factors.

Question Wording and Results:

"Would it be a good idea to set up a government health insurance company to compete with private health insurance companies?"

Yes: 41%
No: 41%

Discussion: I am not particularly fond of this question wording. For one thing, unlike the other polls, it focuses on the action of setting up the "government health insurance company" rather than the choice of insurance plans this ultimately presents to the consumer. For another, it is not clear that a new program would have to be "set up" in order to provide for a public option (i.e. an existing program like Medicare could be expanded), nor that any such entity would properly be described as a "company". The poll seems designed to juxtapose the terms "government" and "company" in a way that might elicit a negative response. (Note that I actually like the inclusion of the term "government" in conjunction with, or perhaps instead of, the term "public". The problem is not with the term "government" itself but instead with the overall way that the question is phrased.)

Non-partisanship rating:
Question wording:
Sample size, sample selection and disclosure:
Overall informativeness:

5. NBC / Wall Street Journal

Who They Are / What's Their Angle: Presumably you are familiar with NBC News and the Wall Street Journal. The NBC/WSJ polls themselves are conducted by Hart/McInturff, a pairing of Democratic pollster Peter Hart and Republican pollster Bill McInturff. Likewise, the NBC/WSJ pairing itself is a collaboration between a somewhat left-leaning and somewhat right-leaning news organization. This is an excellent model to avoid partisanship, both in appearance and in practice.

Specifications: 1,008 American adults on June 12th-15th, including a cellphone sample.

Question Wording and Results:

"In any health care proposal, how important do you feel it is to give people a choice of both a public plan administered by the federal government and a private plan for their health insurance––extremely important, quite important, not that important, or not at all important?

Extremely Important: 41%
Quite Important: 35% (76% Extremely or Quite Important)
Not That Important: 12%
Not At All Important: 8% (20% Not That or Not At All Important)

Discussion: I have no problem with the formulation of the question; in fact, I particularly like the wording "a public plan administered by the federal government" which makes clear that the public plan is in fact government-run. But I have a big problem with the choice of answers. "Importance" is a notoriously vague concept in public opinion polling and may be separate and distinct from asking someone whether or not they support a particular policy. How might someone respond to this question, for instance, if they had particularly strong feelings against a public option? Would they say that it was "not at all important", or would they say that it was "extremely important"? Conversely, how would someone respond if they had a weak preference for a public option, but didn't consider it an especially important component of health care reform? The 1,008 random adults that NBC/WSJ surveyed are going to interpret these dilemmas in a variety of different ways. In addition, the particular category of "quite important" is somewhat ambiguous and probably falls somewhere in between a favorable response and a neutral one.

Non-partisanship rating:
Question wording:
Sample size, sample selection and disclosure:
Overall informativeness:

6. Lake Research / HCAN

Who They Are / What's Their Angle: Lake Research Partners is a Democratic polling firm. Their poll was conducted on behalf of Health Care for America Now! (HCAN), an advocacy group that wants comprehensive health care reform and strongly favors the public option.

Specifications: Sample of 800 likely voters from January 8-13th, 2009.

Question Wording and Results:
"Which of the following three approaches to health care reform do you prefer: one, everyone getting health insurance through private health insurance plans; two, everyone getting health insurance through a public health insurance plan; or three, everyone having a choice of private health insurance or a public health insurance plan?"

73% choice of public or private
15% private only
9% public only

Discussion: Celinda Lake is an excellent pollster, but she is a Democratic pollster and this is a Democratic poll. I don't hate the question wording, but it really emphasizes the option part of the public option and somewhat de-emphasizes the public part; in this sense, it is sort of the alter ego to the Rasmussen poll. As in some of these other polls, it also may not be immediately obvious to the respondent that "public" means administered by the government. A couple of additional points of critique: the use of a likely voter model (as opposed to all adults or registered voters) is a bit unusual this far out of an election cycle, particularly when it regards how the public feels about a particular policy rather than how they want their elected officials to vote on it. And the poll is now a bit outdated, having been conducted in January.

Non-partisanship rating:
Question wording:
Sample size, sample selection and disclosure:
Overall informativeness:

Summary: The only poll I have a particularly high degree of confidence in is the Kaiser Family Foundation poll, which finds that between 65 and 68 percent of the public support a public option depending on how the question is phrased. The only thing I would change about their poll is to specify, as NBC/WSJ does, that the public plan would be administered by the government.

The other polls have one or more characteristics that give me pause about them. The question wording in the Consumers Union' poll is push-y and explicitly partisan; the question wording in the Rasmussen and Lake/HCAN polls is strange and probably implicitly partisan. The NBC/WSJ poll is otherwise terrific, but very difficult to interpret because they ask people about the importance of a public option, and not necessarily their support for one. I might be more comfortable with the ERBI poll if I learned more about it, but the comparative lack of disclosure coupled with the unusual choice to exclude adults 18-20 from the sample and a result that appears to be a mild outlier gives me some concerns about it.

Overall, polling points toward the public option being at least mildly popular and indeed perhaps quite popular. But more polling is required on this question, particularly by the news organizations and other unaffiliated groups like Pew and Gallup, and more care should be taken to frame both questions and answers in a neutral and informative way.