Two-thirds of Americans support Medicare-for-all (#3 of 6)

Informative polls show two-thirds support for single-payer
By Kip Sullivan, J

In Part 2 of this six-part series, I reported on the results of two “citizen jury” experiments in which advocates for single-payer, managed competition, and high-deductible policies spoke to, and were questioned by, “juries” that were representative of America. In the case of the 1993 “jury” sponsored by the Jefferson Center, 71 percent voted for single-payer. In the case of the 1996 “jury,” 61 percent voted for single-payer when no specific information about its cost to individuals was presented, and 79 percent voted for a single-payer system that would have lowered premium and out-of-pocket costs by as much as taxes rose. Both juries rejected proposals relying on health insurance companies by huge majorities.

Many polls that ask about support for Medicare-for-all produce results that confirm the citizen jury findings. But others don’t. What explains that inconsistency?

The more they know about single-payer, the more they like it

In this paper (Part 3 in a six-part series) I will present data from polls that ask about single-payer, and then inquire why some polls show landslide majorities for single-payer and some do not. We will find a clear pattern: Polls that convey more information tend to report higher levels of support than polls that convey little information, and polls that convey accurate information tend to report more support than polls that convey inaccurate information.

Table 1 lists 14 poll questions taken from 11 polls conducted over the last two decades which used the phrase “single payer” and/or referred to an existing single-payer system (Medicare, for example). All 14 questions found majority support for single-payer.

Three of these polls (representing one question each) were limited to doctors. I have included these physician surveys to debunk the false impression (created primarily by the American Medical Association) that the average doctor is opposed to single-payer. The three polls shown in Table 1 indicate that support among doctors is about 60 percent.

Table 1 indicates that public support for single-payer ranges from a low of 50 percent to a high of 69 percent. I have divided the polls of the general public into those that found support levels at 60 percent or higher (eight questions) and those that found levels in the 50-to-58 percent range (three questions).

Table 1: Polls indicating majority support for single-payer

……………………………………………………………For single-payer……..Opposed to single-payer

General public: Polls in which support is 60 percent or higher

Harvard University/Harris (1988)(a)……………………61%…………..not asked
LA Times (1990)(b)…………………………………………….66%………….not asked
Wall Street Journal-NBC (1991)(c)……………………….69%…………….20%
Wash Post-ABC News (2003)(d)…………………………..62%………….not asked
Civil Society Institute (2004)(e)……………………………67%……………..27%
AP-Yahoo (2007)(f)……………………………………………..65%…………..not asked
Grove Insight (2009)(g)………………………………………64%……………..28%
Grove Insight (2009)(g)………………………………………60%……………..27%

General public: Polls in which support is below 60 percent

AP-Yahoo (2007)(f)……………………………………………..54%……………….44%
Kaiser Family Foundation (2009)(h)……………………..58%……………….38%
Kaiser Family Foundation (2009)(h)……………………..50%……………….44%


New Eng J Med (medical school faculty and students) (1999)
………………………………………………………………………….57%…………….not asked
Arch Int Med (doctors) (2004)………………………………64%…………….not asked
Minnesota Med (doctors) (2007)……………………………64%…………….not asked

(a) The question asked by the Harvard University/Harris poll was described in the Health Affairs article reporting the results as follows: “The majority of Americans (61 percent) state they would prefer the Canadian system of national health insurance where ‘the government pays most of the cost of health care for everyone out of taxes and the government sets all fees charged by hospitals and doctors….’” An analogous question posed to Canadians found that only 3 percent of Canadians said they would prefer the American system.

(b) The question asked by the Los Angeles Times poll was: “In the Canadian system of national health insurance, the government pays most of the cost of health care out of taxes and the government sets all fees charged by doctors and hospitals. Under the Canadian system – which costs the taxpayers less than the American system – people can choose their own doctors and hospitals. On balance, would you prefer the Canadian system or the system we have here in the United States?” Sixty-six percent chose the Canadian system and 25 percent chose the US system.

(c) The question asked by the Wall Street Journal-NBC poll was: “Do you favor or oppose the US having a universal government-paid health care system like they have in Canada?”

(d) The Washington Post-ABC News poll asked: “Which would you prefer – (the current health insurance system in the United States, in which most people get their health insurance from private employers, but some people have no insurance); or (a universal health insurance program, in which everyone is covered under a program like Medicare that’s run by the government and financed by taxpayers?)” Thirty-three percent preferred the current system while 62 percent preferred the “universal system.”

(e) The Civil Society poll asked: “Other major nations, such as Canada and England, guarantee their citizens health insurance on the job, through government programs, or via a nonprofit source. Would it be a good or bad idea for the United States to adopt the same approach to providing health care to everyone?”

(f)The AP-Yahoo poll asked two questions. One asked respondents which of these two proposals they agreed with: (1) “The United States should adopt a universal health insurance program in which everyone is covered under a program like Medicare that is run by the government and financed by taxpayers” (65 percent chose this option); (2) “The United States should continue the current health insurance system in which most people get their health insurance from private employers, but some people have no insurance” (34 percent chose this option). The second question was: “Do you consider yourself a supporter of a single-payer health care system, that is a national health plan financed by taxpayers in which all Americans would get their insurance from a single government plan, or not?” (54 percent said they were supporters of single-payer and 44 percent said they were opposed).

(g) The Grove Insight poll asked two questions. One asked: “Federal leaders are considering expanding Medicare to all Americans, so that people have another option besides private health insurance or an HMO. Do you favor or oppose the creation of this type of public health plan option?” (64 percent said they favor this proposal). A very similar question was asked which differed from the first by including information on the financing mechanism: “There is proposed federal legislation that gives any American, regardless of age, the option of joining the Medicare program. Americans who choose this option would share the cost of the coverage with their employer through increased Medicare payroll deductions, instead of paying private health insurance premiums. Do you favor or oppose this legislation?” (60 percent favored it and 27 percent opposed it). Both questions, especially the second one, imply private insurers will continue to exist alongside a Medicare program open to all. But the questions are so similar to questions that clearly ask about Medicare-for-all systems that I decided to include them here.

(h) The Kaiser Family Foundation poll asked: “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.?” This was followed by eight proposals which, with the exception of the question about the “public option,” were asked in a random order (the “option” question was always asked at the end). Two of these questions asked about single-payer. The first read: “Having a national health plan in which all Americans would get their insurance through an expanded, universal form of Medicare-for all.” Fifty-eight percent said they favored this proposal while 38 percent said they opposed. The second read: “Having a national health plan – or single-payer plan – in which all Americans would get their insurance from a single government plan.” Only 50 percent favored this proposal while 44 percent opposed.

For sources see Table 2 below.

If we examine the questions posed by all the polls of the general public, one difference between the two sets of poll questions jumps out immediately: The questions that generated levels of support at 60 percent or higher mentioned one of three existing single-payer programs – the Canadian system, the British system, and the US Medicare program. (I have bolded the words referring to these systems in the poll questions, which are presented in the footnotes to Table 1.) In other words, those questions didn’t just rely on the phrase “single payer,” a phrase most people do not understand.

On the other hand, the three questions that prompted support in the 50-to-58-percent range used the phrase “single-payer” but did not refer to an existing single-payer system or program. The second AP-Yahoo question, for example, merely asked respondents if they considered themselves to be “single-payer supporters.” Fifty-four percent said yes to that question, which was substantially below the 65 percent who indicated in the same AP-Yahoo poll that they supported a system of universal coverage “like Medicare.” These two AP-Yahoo questions taken together suggest that merely using the term “single-payer” and not comparing it to Medicare will cut roughly 10 percentage points off the support level for single-payer.

It might be argued that the second AP-Yahoo question shown in Table 1 produced a relatively low single-payer support rate (54 percent) because it also mentioned the words “taxpayers” and “government.” But that argument doesn’t work. All but one of the other questions that produced support levels of 60-percent or higher also mentioned “government” and “taxes.” The difference is they also mentioned an existing single-payer system or program.

Apples-to-aardvarks comparisons also reduce support for single-payer

The two questions in Table 1 posed by the 2009 Kaiser poll (see question 13, page 8), which showed 58 and 50 percent support for single-payer, reveal another factor that seems to influence poll results – a factor I’ll call the “line-up effect.” The Kaiser poll asked about single-payer as well as a half-dozen other proposals without indicating what effect each proposal would have on costs, the number of uninsured, and freedom to choose one’s doctor, to name just a few of the variables most people would be interested in. By contrast, the polls listed in the 60-percent-or-higher category did not present single-payer in a line-up with other proposals; they simply asked whether respondents would support a single-payer system, or they contrasted single-payer with the current system. The “line-up effect” generated by the Kaiser polls would be minimized or eliminated in a citizen jury experiment because the jury would have plenty of time to inquire about the relative effectiveness of the competing proposals. Respondents to polls don’t have that luxury.

The 2009 Kaiser poll began with this announcement:

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.

Notice the phrase, “different ways to increase the number of Americans covered by health insurance.” It implies the “different ways” have all been shown by research to work, and perhaps to reach roughly similar results.

This question was then followed by a description of eight proposals, including “expanding Medicare to people between the ages of 55 to 64,” “offering tax credits to help people buy private health insurance,” and “requiring all Americans to have health insurance.”

This “line up” method of asking about support for single-payer is by no means fatal, but it does appear to reduce the pro-single-payer response rate by somewhere in the range of 5 to 10 percentage points. The Kaiser question that produced 58 percent support asked about “having a national health plan in which all Americans would get their insurance through an expanded, universal form of Medicare-for all.” Because this question did not mention taxes and government, you might think more than 58 percent of Americans would have said they favored this proposal. After all, when other polls that do not put single-payer in a line-up but do compare single-payer to Medicare and do mention “government” and “taxes” (see the upper half of Table 1), more than 60 percent indicate their support. The fact that only 58 percent of Americans responded favorably to this question from Kaiser – a question that does mention Medicare but mentions neither “taxes” nor “government” – begs for an explanation. It is reasonable to hypothesize that the explanation is the “line up” context in which the question was asked.

The second Kaiser question listed in Table 1, the one that produced only 50 percent support, contained a double whammy. Like the first Kaiser question, it used the line-up method; unlike the first question, it failed to compare single-payer with Medicare or another single-payer system. This suggests that the cumulative effect of the line-up method plus failure to compare single-payer to Medicare can diminish support for single-payer by about 15 percent.

Perhaps an analogy will help. Imagine if you were asked to indicate whether you “favored or opposed” six “ways to lose weight,” and the “ways” (“ways” is the noun Kaiser uses) ranged from the truly effective (for example, exercising for half an hour a day) to the barely effective (for example, weight loss pills or drinking more water). Imagine furthermore that the pollster gave you no information at all on the effectiveness of the various “ways” nor on their side effects. It seems likely that many respondents could be lulled into thinking all the “ways” are roughly equivalent in effectiveness and that respondents would, therefore, give less support to the effective methods of weight loss in response to this type of “line up” question than they would if they were simply asked, “Do you support exercise as a means of weight loss?”

Let me offer one more example of the use of the line-up method in a poll about health care reform, this one the July 2009 poll by Time Magazine. Time posed questions about seven different proposals that began with the phrase, “Would you favor or oppose a health care bill that…?” The implication of the phrase “a health care bill” is that members of Congress and experts in general think all of the proposals the respondent is about to hear will ameliorate the health care crisis to some degree, perhaps to the same degree. The single-payer question read:

Would you favor or oppose a health care bill that creates a national single-payer plan similar to Medicare for all, in which the government would provide health care insurance to all Americans?

Forty-nine percent favored single-payer, 46 percent opposed it. Like all the poll questions shown in Table 1 that showed support for single-payer in the 60-to-70-percent range, the Time question mentioned Medicare and “government.” (Oddly, unlike the high-scoring poll questions in Table 1, the Time question didn’t mention “taxes.”) You might think, then, that the Time poll would have produced a level of support for single-payer in the sixties. The fact that it produced only a 49 percent “favor” rating suggests, again, that something about the “line up” format reduces support for single-payer by about 10 percentage points.

To sum up this section: Polls that ask reasonably informative questions about single-payer show that somewhere between 60 and 70 percent of Americans support single-payer. This level of support can be reduced into the 50-to-60 percent range by two methods: Asking about “single-payer” without comparing single-payer to Medicare or the systems of Canada or the UK; and inserting the question about single-payer in a list of a half-dozen other proposals without warning respondents that the non-single-payer proposals, especially incremental proposals like tax credits, will have effects that are quite different from the single-payer proposal.

Two more examples of polls that convey too little information

To explore further the hypothesis that vagueness in poll questions diminishes support for single-payer, consider polls that are even vaguer than the polls in Table 1 that use “single payer” but offer almost no details about it. Let’s examine three polls that did not use the phrase “single payer” and offered no details about how the proposed “government” program would work.

In Part 2 of this series, I described a CBS poll conducted in June and August 2009 which asked:

Do you think the government would do a better or worse job than private insurance companies in providing medical coverage?

This question has the ring of a single-payer question, but it leaves numerous important questions unanswered, including whether the program in question would provide coverage to everyone and whether “provide” means cover people directly or give them subsidies so they can buy coverage from insurance companies.

We saw that when this question was asked in June 2009, 50 percent said “the government” would do a better job, but when this question was asked in late August 2009, only 36 percent said “the government” would do a better job. Does this CBS poll contradict the more precise polls listed in Table 1 that found two-thirds support for single-payer?

The answer is no. The CBS poll conveys so little information about how “the government” would do the “job” of “providing medical coverage” that it isn’t even clear if this question was meant to be about single-payer. In the context of the current debate, Americans are much more likely to think the question refers to the Democrats’ 2009 “reform” bills, which require Americans to buy health insurance from insurance companies, than to single-payer legislation. The sharp drop in support for “the government” in the CBS poll between June and August is evidence that the highly publicized town hall meetings held in August to discuss the Democrats’ bills influenced responses to the poll, which in turn indicates many respondents thought the question was about the Democrats’ legislation, not HR 676 (the single-payer bill introduced in the House of Representatives) or S 703 (the Senate single-payer bill).

We see a similar problem in the following question, contained in both a CBS/New York Times poll and a Harvard School of Public Health poll, conducted over several decades:

Do you favor or oppose national health insurance, which would be financed by tax money, paying for most forms of health care?

Like the phrase “government providing medical coverage” in the CBS poll, the phrase “national health insurance” in this poll could mean government financing of universal coverage through a single-payer system or through a multiple-payer system. If you look at Exhibit 1 on page 35 of this article from Health Affairs, you’ll see that between 1980 and 2000 the percent of respondents saying they favor “national health insurance” ranged between 46 and 66 percent. The vagueness of the phrase was unquestionably a significant reason why support fluctuated so much.

Another way to diminish support for single-payer: Convey inaccurate information

In addition to conveying vague information about single-payer there is, of course, another time-tested method of diminishing support for it, and that is to convey inaccurate information about it. This can be done explicitly and implicitly. It can be done explicitly by, for example, asserting in the question that single-payer systems raise taxes but do not lower premium payments and out-of-pocket costs. We have already seen one example of how reducing support for single-payer with inaccurate information can be done implicitly – by inserting the single-payer question into the middle of several other proposals, including incremental proposals such as tax credits for small employers, without warning respondents that the proposals have very different benefits and side effects.

Since 2001, the Gallup poll has been asking this explicitly misleading question (apparently each November):

Which of the following approaches for providing health care in the United States would you prefer: replacing the current health care system with a new government-run health care system, or maintaining the current system based mostly on private health insurance? (emphasis added)

“Government-run health care system” has garnered somewhere between 32 and 41 percent support since 2001 (while keeping the “current system” has attracted the support of 48 to 63 percent). But this poll is so biased it is irrelevant to the current debate. The problem here is the use of the phrase “health care” three times instead of “health insurance.”

The government does not “run health care” under single-payer systems (or any other system currently under debate in the US, for that matter). Under single-payer systems, clinics, hospitals, and makers of drugs and equipment that are privately owned today would remain in private hands. What the government will “run” in a Medicare-for-all system is health insurance, not health care. The latter phrasing conjures up nightmares of a gigantic government HMO in which the federal HMO owns all the clinics and hospitals and government bureaucrats decide whether you may have the surgery you and your doctor think you need or whether you must take Lipitor when your doctor prescribed Crestor.

I will discuss another example of a poll that delivers explicit misinformation in Part 5 when I discuss the “research” Celinda Lake did for the “option” movement.

The Bermuda Triangle

Finally, there is the occasional outlier poll that produces very low favorability ratings for single-payer about which I can only offer a plausible hypothesis. The August 7-8, 2009 Rasmussen Poll (not shown in Table 1) is an example. The poll asked:

Do you favor or oppose a single payer health care system where the federal government provides coverage for everyone?

We would expect this poll to produce “favor” responses below the 60-percent level because it offers so little information about what a single-payer is (it doesn’t mention Medicare or the Canadian or British systems, and offers no other details). But Rasmussen reported that only 32 percent supported single-payer while 57 percent opposed it. This question was not asked as part of a “line up,” so the line-up explanation doesn’t help us here. The two explanations that occur to me are sloppiness and deliberate manipulation of the process (for example, sampling a lot more conservatives than liberals). That possibility has occurred to others as well. Rasmussen’s non-electoral polls seem to show more support for conservative positions than other polls.


We have now reviewed three categories of polls that correspond roughly to support levels of 60 to 70 percent, 50 to 60 percent, and below 50 percent. Polls that produce greater-than-60-percent levels of support for single payer not only use the phrase “single-payer” but compare the concept to an existing single-payer program, typically Medicare. Polls showing 50 to 60 percent support inquire about “single payer” without comparing the concept to Medicare or to the single-payer systems of other countries or they pose the question about single-payer in a line-up context. Polls that seem to ask about single-payer and which show less than 50 percent support use phrasing that is so vague respondents cannot know whether the program being asked about is a single-payer and, if so, how it would work.

We saw in Part 2 of this series that two citizen juries conducted in the 1990s produced landslide votes for single-payer – votes equal to roughly 60 to 80 percent of all the participating “jurors.” These lengthy “jury” experiments are far more reliable than any poll could possibly be. And yet some polls confirm the “jury” experiments and some don’t. If we ask why, the answer is the polls that show support in at least the 60-to-70-percent range use the phrase “single payer” and give respondents concrete examples of single-payer programs.

If we couple the “jury” experiments with the polling data reviewed in this part, we see a pattern: The more people know about single-payer, the more likely they are to support it. We see this pattern when we compare the “jury” results with poll results, and we see it when we compare polls that show high levels of support for single-payer with those that don’t.

Stay tuned for Part 4: “Jacob Hacker’s ambiguous polls”

Table 2: Sources

Harvard/Harris poll: Robert J. Blendon et al., “Views on health care: Public opinion in three nations,” Health Affairs, Spring 1989;8(1)149-157.

Los Angeles Times poll: “Health Care in the United States,” Poll no. 212, Storrs, Conn.: Roper Center for Public Opinion Research, March 1990, cited in Robert J. Blendon et al., “Satisfaction with health systems in ten nations,” Health Affairs, Summer 1990;9(2): 185-192. Actual wording of the question is available at American Public Opinion Index, 1990, p. 649.

Wall Street Journal-NBC poll: Michael McQueen, “Voters, sick of the current health –care systems, want federal government to prescribe remedy,” Wall Street Journal, June 28, 1991, A4 (question available here).

New England Journal of Medicine poll: Steven R. Simon et al., “Views of managed care: A survey of students, residents, faculty, and deans of medical schools in the United States,” New England Journal of Medicine 1999; 340:928-936, 929.

Washington Post-ABC News Poll: Health Care, October 20, 2003.
Archives of Internal Medicine poll: Danny McCormick et al., “Single-payer national health insurance: Physicians’ views,” Archives of Internal Medicine 2004;164:300-304.

Civil Society Institute poll: Opinion Research Corporation, Americans and Health Care Reform: How Access and Affordability are Shaping Views, September 15, 2004.

Minnesota Medicine poll: Joel Albers et al, “Single-payer, health savings accounts, or managed care? Minnesota physicians’ perspectives,” Minnesota Medicine, February 2007:36-40.

AP-Yahoo poll: Knowledge Networks, (page 15).
Grove Insight poll: Grove Insight memo to Jamie Court, January 30, 2009.

Kaiser Family Foundation poll, 2009.