Two-thirds of Americans support Medicare-for-all (#4 of 6)

Jacob Hacker’s ambiguous polls
By Kip Sullivan, JD

In Part 2 and Part 3 of this series I reviewed rigorous evidence from multiple sources supporting the statement that somewhere between 60 and 80 percent of Americans support a Medicare-for-all system. A reasonably conservative averaging of the more rigorously conducted research I reviewed – the citizen jury results and the results of polls that asked accurate and relatively informative questions – indicates two-thirds of Americans support a single-payer or Medicare-for-all system.

In this part and in Part 5, I will examine the basis for the claim by representatives of the “public option” movement that only a minority of Americans support single-payer and a majority are opposed. The basis for that claim consists primarily of several papers written by Jacob Hacker and “research” done for the Herndon Alliance by pollster Celinda Lake. Until about two years ago, Hacker wrote about health policy primarily for the academic community; since then he has published frequently in the lay media. Since its formation in 2005, the Herndon Alliance has sought to create “research” that could be used to persuade the public, especially legislators and political activists, that single-payer should be taken off the table and the “public option” should be put on the table. I review Hacker’s work in this paper and Celinda Lake’s in Part 5.

Expediency-driven health policy

It may sound sacrilegious to say this …, but the greatest lesson of the failure of the Clinton health plan is that reformers pay too much attention to policy and too little to politics. If real estate is about location, location, location, health reform is about politics, politics, politics.

Thus spake Jacob Hacker in a paper published in Health Affairs in 2008 entitled, “Putting politics first.” Hacker argues that anyone who wants to achieve universal health insurance must somehow separate “politics” from “policy” and give highest priority to politics. If Hacker had merely said that anyone who seeks to achieve universal health insurance should devote resources to building public pressure for it, his statement would be incontrovertible. It would be a truism. But Hacker’s “politics, politics, politics” statement went beyond the truism that “reformers” must build a movement for universal health insurance.

Hacker’s demand that we distinguish between politics and policy and give high value to one and low value to the other is nonsensical. It’s equivalent to saying that process is separate from and matters more than outcome, or that means are separate from and matter more than ends. To make such a distinction amounts to an endorsement of opportunism and expediency. We will see in the remainder of this article that in fact that’s where Hacker’s “put politics first” mantra leads him. It leads him to attribute to the public anti-single-payer, pro-insurance-industry attitudes based on polling data that are so abstract they offer no guidance at all. As the events of 2009 have demonstrated, the exaltation of expediency – dressed up as political science – produces neither good policy nor good political strategy.

Unlike the Herndon Alliance, which commissioned its own polling and focus group “research,” Hacker relied on existing polling data to support his conclusion that single-payer is not feasible while the “public option” is. Hacker cites different types of polls depending on whether he is addressing the general public or health policy experts. His 2006 article for Slate cited one set of polls. A 2007 paper that he co-authored with Mark Schlesinger (“Secret weapon: The ‘new’ Medicare as a route to health security,” Journal of Health Politics, Policy and Law 2007;32:247-291) cited another set of polls. Inexplicably, neither paper discussed the Jefferson Center jury results I discussed in Part 2 nor the polls showing large majorities for single-payer that I discussed in Part 3 of this series.

In the course of examining these two papers, I will review in detail seven polls that Hacker cites. This may get tedious, but it’s important that you see for yourself how nebulous Hacker’s “evidence” is. Once you behold Hacker’s “evidence” directly, you realize that Hacker’s belief that Americans oppose single-payer is based entirely on polling results that resemble a Rorschach ink blot. You can see in them what you want to see. Where you and I might discern a public ready to support single-payer, Hacker discovers hulking impediments to single-payer.

Polls Hacker cited in his Slate article

Hacker’s article for Slate bore the condescending title, “Better medicine: Fixing the left’s health care prescription.” The problem in need of “fixing,” according to Hacker, was “the left’s” support for single-payer. Hacker urged “the left” to support instead his proposal to “give employers the option of providing … coverage to their workers through a new public program modeled after Medicare” or through the insurance industry, a proposal that would, by 2009, be called “the public option” for short.

Hacker grudgingly acknowledged single-payer’s advantages, but then claimed single-payer advocates were “biting off too much.”

Americans like Medicare, and yes, Medicare is easy to explain. But that doesn’t mean most people are ready to say everyone should be covered by Medicare. Many of us remain stubbornly attached to employment-based health insurance, and proposing to abolish it entirely is likely to stir up fear as well as gratitude.

He hyperlinked the words “stubbornly attached” to an article in Mother Jones written by the Century Foundation. (In the fullness of time, the Century Foundation became a passionate advocate for the “public option.”) The Century Foundation article reviewed several polls on American attitudes about “universal coverage.” Amazingly, one of them was the 2003 Washington Post/ABC News poll showing 62 percent support for a Medicare-for-all system that I discussed in Part 3. Does Hacker read the documents he cites as evidence for his own claims?

Before we examine the Century Foundation’s article, I want call your attention to three features of Hacker’s argument.

First, he practices “put politics first.” He says that even though single-payer is a good proposal, it should be rejected entirely. It would be one thing to counsel single-payer advocates against trying to get a full-blown single-payer system enacted in a single session of Congress and to plan instead for a multi-year campaign (which is fact what the single-payer movement has been doing for two decades). But Hacker is not doing that. He is urging progressives to reject single-payer completely.

Note second that Hacker urges us to accept whatever polls say as the final arbiter of what is politically feasible. Hacker has no interest in a very obvious question: If everyone who supports universal coverage threw their weight behind the campaign for single-payer, how much higher could public support for single-payer be raised?

Third, Hacker can’t bring himself to say how many Americans are “stubbornly attached” to employer-based health insurance. He can only bring himself to say “many.” If Hacker is going to rest his entire argument that the “left” should abandon single-payer on the premise that “we” are “attached” to the current system, why is he so vague about what proportion of the populace he is talking about?

I urge readers to examine the Century Foundation article for yourself. Focus on the “What we know” section (it’s only a page long), which is where the poll results are discussed. It will become obvious quickly that this article provides no basis at all for Hacker’s claim that Americans are “stubbornly attached” to the current system. At most, only three paragraphs have any relevance to that claim, and these paragraphs produce results that are at best ambiguous and at worst (from Hacker’s point of view) supportive of single-payer.

Consider the two excerpts from the Century Foundation summary I quote below. The first asserts the public wants to replace “the current employer-based system” (yes, the very same “employment-based” system to which Hacker says “many of us remain stubbornly attached”). The second excerpt, which appears barely a half-page later, asserts just the opposite.

[Excerpt 1]
The public wants the government to play a leading role in providing health care for all. For example, in an October, 2003 Washington Post/ABC poll, by almost a two-to-one margin (62 percent to 33 percent), Americans said that they preferred a universal system that would provide coverage to everyone under a government program, as opposed to the current employer-based system.

[Excerpt 2]
The public generally wants to build on, rather than eliminate, the current employer-based private health insurance system. In a January, 2000 Kaiser poll, they preferred building on the current system to switching to a system of individual responsibility (54 percent to 39 percent) and in a November 2003 Kaiser poll, they preferred keeping the current system to replacing it with a government-run system (57 percent to 38 percent). (emphasis added)

How does one make any sense of these conflicting statements? How does Hacker find in these statements proof that Americans (a) like the current employer-based system, and (b) like it so much they would oppose a single-payer system? In these excerpts, the author of this summary, Ruy Teixeira, gives us not only two contradictory statements to sort out (the public does and does not want to replace the “current employer-based system”), but we’re supposed to understand what “a system of individual responsibility” and “a government-run system” means.

If we track down the polls these excerpts refer to, we discover that we have already encountered these polls, or polls like them, in Part 3 of this series.

I discussed in Part 3 of this series the 2003 Washington Post/ABC poll that Teixeira cites in the first excerpt. That poll found 62 percent support for a single-payer system, described in that poll as “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.” So how does Teixeira account for the difference between the 62 percent support for single-payer he reports in the first excerpt and the 38 percent level of support one or both of two Kaiser polls (Teixeira isn’t clear which) reported for a “government-run system” in the second excerpt? He doesn’t say.

The Kaiser poll search engine (using the phrases “individual responsibility” and “government run”) and a Google search turned up only one of the two Kaiser polls Teixeira refers to in excerpt 2 above – the January 2000 poll. That poll, which Teixiera cited as evidence that Americans prefer “the current system” to a “system of individual responsibility,” reads as follows:

Which of the following, option one or option two, do you think would be the better way to guarantee health insurance coverage for Americans? Option One is, building on the current system in which employers contribute to their employees’ health insurance, which they get through their job, and the Government covers the cost of insurance for the poor and unemployed, or Option Two which is, switching to a system in which all individuals would buy their own health insurance but would receive a tax credit or subsidy to help them with the cost of the plan.

Fifty-four percent chose “the current system” versus 39 percent who chose what Teixeira called “a system of individual responsibility.”

In my last installment I discussed polls quite similar to the other (2003) Kaiser poll Teixeira cited (the one my search failed to turn up), a poll which, according to Teixeira, asked respondents to choose between “the current system” and “a government-run system.” The ominous phrase “government-run system” sounds very much like the frightening phrase “government-run health care system” conjured by the Gallup poll (discussed in Part 3). The 38-percent level of support Teixeira reports is within the range of Gallup poll results over the last decade – 32 to 41 percent – that I reported. This strengthens my hypothesis that the question Teixeira claims Kaiser asked in 2003 was very similar to the Gallup question. (It would help if people who urge readers to rely on polls for any reason would link readers to those polls or give more precise source information.)

The only other shred of information in the Century Foundation article that might give a “yes but” comfort was this excerpt, which again contained contradictory statements:

In a December 2003 Harvard School of Public Health/Robert Wood Johnson/ICR poll, 80 percent supported expanding Medicaid/SCHIP; 76 percent supported employers being required to offer a health plan; and 71 percent supported a tax credit plan. Trailing these options, but still garnering majority support, were a universal Medicare plan (55 percent) and an individual coverage mandate plan (54 percent). …. (Note: one of the only options that didn’t garner majority support … was a single or national health plan financed by tax payers that would provide insurance for all Americans [37 percent to 47 percent].)

Once again, Teixeira juxtaposes a poll showing majority support for single-payer (55 percent) with another poll showing 37 percent support, and offers no explanation for the difference. As you can see, the two single-payer questions Teixeira refers to appear to have been part of a line-up of another half-dozen questions or so, including questions about proposals that wouldn’t come close to achieving universal coverage and none of which would cut costs.

To sum up, the Century Foundation article Hacker linked his readers to for evidence of our “stubborn attachment” to the current system demonstrated nothing of the sort.

Polls Hacker relies on in his 2007 paper

In the paper he published in the Journal of Health Politics, Policy and Law in 2007 with Mark Schlesinger, Hacker argued for the “public option” and against single-payer. As he did in his Slate paper, Hacker argued that the “expectations” and “values” of the American people, not the insurance industry, constitute an intractable obstacle to single-payer. At the outset of this paper, in a section entitled, “Prevailing American values as barriers to universal health insurance,” Hacker sought to make two arguments: Americans value choice of health insurance company, and they are scared of their government. The data he relied on to make this case were even more abstract and ambiguous than the data he relied on in his Slate article. I’ll review the evidence he cites for his claim that Americans value choice among insurance companies first, and then examine the data he cites for his claim that Americans are afraid of a single-payer system.

Hacker’s argument that Americans value choice of health insurer (as opposed to provider) consisted almost entirely of these statements:

During the debate over health reform in the early 1990s, 81 percent of the public reported that it was important or essential for a proposal to give “people a choice of different types of health insurance plans” (Louis Harris and Associates in 1994). When asked whether “seniors should have the option of picking a private health plan approved by the Medicare program to provide their health benefits,” 82 percent of the public endorsed these choices (Zogby International 2003). Americans embrace choice of insurance not because they favor markets in health care per se but because they have so little trust in government, employers, or private insurance and want protection against problematic experiences (Blendon et al. 1998; Jacobs and Shapiro 1999).

Neither of the two polls and neither of the two papers Hacker cites support his conclusions. The papers deal exclusively with the backlash against managed care that occurred in the late 1990s. Those papers say nothing that could be construed as evidence that Americans “embrace choice of insurance” and have “little trust in government.” To give you some idea of how badly Hacker misinterpreted these papers, I have presented the abstract of the paper by Blendon et al. in the appendix to this paper (the Jacobs and Shapiro paper did not contain an abstract).

Now let’s look at the two polls Hacker cited to support his claim that Americans value choice of insurance company. The 1994 Harris poll posed this question:

As the Congress debates health care reform, they must consider several different goals. Please say for each of the following whether you think it is absolutely essential, very important, or not important …. Giving people a choice of different types of health insurance plans?

Thirty-six percent said “choice of… plans” was “absolutely essential” and 45 percent said it was “very important.” But does this poll demonstrate that Americans value choice of insurance company?

This poll was conducted during May 23 to 26, 1994, while the debate over the Clinton bill – a bill which would have pushed middle- and lower-income people into HMOs and other tightly managed health insurance companies – was still in full swing. The poll question deliberately asked respondents to think about the current debate in Congress and the “goals” that “Congress must consider.” The context in which this poll question was asked, and the opening statement to the question, must have induced all or most respondents to think they were being asked whether they would approve of Congress reducing their choice of insurance companies. It is not surprising they said no to this question. But saying no cannot be construed as “attachment” to the current system, and certainly not opposition to Medicare-for-all. Hacker’s claim to the contrary is equivalent to saying prisoners in a gulag are “stubbornly attached” to gulag food because they told a pollster they would object to being given less of it. (This question and the responses were emailed to me by the Roper Center for Public Opinion Research at the University of Connecticut.)

The other poll Hacker refers to – a Zogby poll – misled respondents. The poll, conducted June 18-21 2003, asked if “seniors should have the option of picking a private health plan approved by the Medicare program to provide their health benefits.” But the poll failed to ask respondents if they would feel the same way if they knew that allowing insurance companies to insure Medicare beneficiaries raises the cost of the entire Medicare program. This is a very well documented fact; every expert knows it to be true. Even Hacker and Schlesinger acknowledged it. How far support would have fallen had respondents been informed that their taxes would have to go up to give seniors the privilege of leaving the traditional Medicare program and enrolling with an insurance company? We don’t know. Zogby didn’t ask, possibly because the conservative Galen Institute was the sponsor of the poll.

Now for Hacker’s and Schlesinger’s claim that Americans are afraid of a government-financed single-payer. This claim relied primarily on two polls conducted over several decades: the “General Social Survey,” conducted by the University of Chicago, and the “National Election Studies” survey conducted by the University of Michigan.

Hacker and Schlesinger claimed the General Social Survey supported the following baffling statement:

“[W]hile approximately 80 percent of the public endorses some collective responsibility for health care finance, support for a completely collective role rarely garners majority support and, if so, then for only brief periods of time (see Figure 1)” (page 252).

What does “some collective responsibility” mean? How does it differ from “complete collective responsibility”? The latter seems to mean government pays for 100 percent of the national health care bill. But no country in the entire world does that. What does “health care finance” refer to? Universal coverage? Less-than-universal coverage? A single-payer system? The current multiple-payer system?

The figure Hacker and Schlesinger refer to as evidence for this baffling statement is a bar chart, based on the General Social Survey, showing bars for various years broken down by the proportion of the populace who support “collective,” “individual,” and “split responsibility for medical care.” The figure indicates that from 1975 through 2000 roughly 50 percent of Americans supported “collective responsibility,” 30 percent supported “split responsibility,” and 20 percent supported “individual responsibility.” How any reasonable person can conclude from these data that Americans oppose single-payer because they fear government and value choice of health insurance company is beyond me. If we really must ask whether such ambiguous data dictate that we abandon or support a Medicare-for-all system, it would seem more reasonable to interpret these data to say a majority of the public will support Medicare-for-all.

To enhance your impression of how flimsy this bar chart is, consider the actual question asked by the GSS survey:

In general, some people think that it is the responsibility of the government in Washington to see to it that people have help in paying for doctors and hospital bills. Others think that these matters are not the responsibility of the-federal government and that people should take care of these things themselves. Where would you place yourself on this scale [respondents were handed a card showing numbers running horizontally from 1 to 5], or haven’t you made up your mind on this?

Above number 1 on the card is the label, “I strongly agree it is the responsibility of government to help” and above number 5 is the label, “I strongly agree people should take care of themselves.”

Similarly, Hacker and Schlesinger use data from the National Election Studies survey that is at best ambiguous and at worst (from Hacker’s point of view) favorable to single-payer to spin a picture of Americans so “deeply divided” about the role of government that single-payer isn’t possible. They claim that a single question from this survey supports the following conclusions:

“Americans have long been deeply divided about their preferred approach to expanding health insurance…. Americans … split evenly between those who favor administration of insurance benefits by the government and those who prefer subsidies for private insurers (table 2)” (page 255).

The table they refer to shows that over the last half century roughly 45 percent favor “government insurance” versus about 40 percent for “private insurance.” Here is the question:

Some people feel there should be a government insurance plan which would cover all medical and hospital expenses for everyone. Others feel that all medical expenses should be paid by individuals, and through private insurance plans like Blue Cross and some other company paid plans. Where would you place yourself on [a seven-point] scale…. ?

There was, of course, no other information to help respondents interpret the key phrases in this question including “government insurance plan.” Respondents had to rank themselves as a “1” if they were strongly in favor of a “government insurance plan” that paid all expenses for everyone, and 7 if they felt strongly in favor of “individuals and private insurance plans” paying some unspecified portion of expenses, or some number in between if they felt less than strongly about their opinion. Hacker and Schlesinger treated everyone who ranked themselves as a 4 as undecided, and then treated all the 1, 2, and 3 people as for “government” and all the 5, 6, and 7 people as for “private insurance.”


Even if we didn’t know that Hacker was an avid proponent of the “politics, politics, politics” mantra, and that this mantra amounts to little more than an excuse to make policy decisions based upon ambiguous and cherry-picked polling data, we might reach these conclusions simply by reading the two papers by Hacker I have reviewed here. In his 2006 article for Slate, and his 2007 paper for the Journal of Health Politics, Policy and Law, Hacker urged his readers to abandon single-payer based on poll results that were not merely cherry-picked (with one unintended exception he excluded polls that showed two-thirds support for single-payer), but, even after careful cherry-picking, were still unclear in their implications.

I am not saying polling data reveal that only a single-payer system attracts majority support. A fair reading of the polls (although not the citizen jury results) suggests that Americans would accept a variety of solutions to the health care crisis if they could be convinced that they would cover everyone and bring costs down.

I strongly disagree with Hacker, however, that the polling data demonstrates a majority wants to defend the current employer-based multiple-payer system and oppose a single-payer system. And I strongly disagree with the assumption that people who care about solving the health care crisis should examine polls first and then decide how to solve the health care crisis. If we must put our finger in the wind before we decide whether to support single-payer, then let us at least consult research that used rigorous methodology, e.g., the citizen juries, and polls that inform their respondents about actual proposals. Let us not consult polls that use vague phrases like “people should take care of these things themselves.”

Appendix: Abstract of one of two papers Hacker misrepresented

In his paper with Mark Schlesinger published in the Journal of Health Politics, Policy and Law in 2007, Hacker cited two papers for support of this sentence: “Americans embrace choice of insurance not because they favor markets in health care per se but because they have so little trust in government, employers, or private insurance and want protection against problematic experiences.” Neither paper discussed lack of trust in government or employers. Both papers were about public hostility to the insurance industry. Below I present the abstract of one of the two papers (there was no abstract for the second one).

This paper examines the depth and breadth of the public backlash against managed care and the reasons for it. We conclude that the backlash is real and influenced by at least two principal factors: (1) A significant proportion of Americans report problems with managed care plans; and (2) the public perceives threatening and dramatic events in managed care that have been experienced by just a few. In addition, public concern is driven by fear that regardless of how well their plans perform today, care might not be available or paid for when they are very sick. (Robert Blendon et al., “Understanding the managed care backlash,” Health Affairs 1998;17(4):80-94))

Stay tuned for Part 5: Celinda Lake’s “research” for the Herndon Alliance