By KIP SULLIVAN, JD

Jacob Hacker, the godfather of the “public option,” and Donald Trump have much in common. They both think the solution to high health care costs is more competition within the insurance industry. They both acknowledge that the insurance industry is highly concentrated, and yet for reasons they don’t disclose they both think it’s possible for new insurance companies to break into such a highly concentrated industry. The only difference between their theories of competition is that Trump wants insurance companies to create insurance companies from scratch, while Hacker wants the government to create insurance companies from scratch.

In my last post , I criticized Hacker for not explaining how the “public option” (PO) will come into existence. All Hacker can say is the PO will be “like Medicare.” Hacker and other PO proponents don’t tell us how the PO will become “like Medicare.” We are simply supposed to believe the PO will leap into existence and, when it does, it will be big like Medicare and enjoy Medicare’s low overhead and payment rates. [1]

Trump’s “explanation” is just as empty. He simply asserts that insurance companies in one state will open shop in other states if the regulations in some states are reduced. [2] Here is how Trump “explained” his proposal during the second presidential debate  on October 19: “We have to get rid of the lines around the state, artificial lines, where we stop insurance companies from coming in and competing, because they want – and President Obama and whoever was working on it – they want to leave those lines, because that gives the insurance companies essentially monopolies….”

HMOs and the chicken-and-egg problem

You can see in the excerpt quoted above that Trump explicitly acknowledges how highly concentrated the insurance industry has become. He equates insurance companies with monopolies, which is not quite accurate. The most accurate characterization of US insurance markets would be “oligopoly” – control by a few companies. According to a report  by the Government Accountability Office required by the Affordable Care Act, “[E]nrollment [in 2013] was concentrated among the three largest insurers in most states. Specifically, in each of the three market segments [individual, small group, and large group], the three largest insurers had at least 80 percent of the total enrollment in at least 37 states.” (p. 4)

Hacker and PO proponents are just as explicit in acknowledging how concentrated the insurance industry has become. For example, in testimony before the House Education and Labor Committee in June 2009, Hacker stated, “In much of the country today, health insurance competition is remarkably limited. Most metropolitan areas have no more than a few dominant insurers in control of the market.” (p. 71) Hacker went on to state, “Even the largest insurers are hard-pressed to enter established markets.” (p. 72) And yet, like Trump, Hacker leaps from acknowledging the oligopoly problem to boasting the problem will disappear if only his vaguely defined solution is enacted into law. “A public health insurance plan would provide greater competition for insurers and providers and greater choice for Americans,” he told the House committee (p. 72). How would a PO crack insurance markets all over the country when even “the largest insurers” aren’t doing it? Hacker didn’t say. [3]

The problem for which neither Trump nor Hacker provides a solution is often described as the “chicken-and-egg problem” – a problem created by the insurance industry’s consolidation and aggravated by the industry’s near universal use of limited networks. We can thank HMOs and the managed care ideology which gave rise to HMOs for this problem. Consolidation was a problem prior to the 1980s, but the combination of unusually high consolidation and widespread use of limited choice did not make insurance markets virtually impregnable everywhere until HMOs began their rapid spread in the 1980s. Similarly, we can thank ACOs and the managed care ideology which gave rise to ACOs for aggravating this problem.

Beginning in the 1980s, HMOs, which as we all know pioneered limiting patient choice, seized market share from traditional insurers which at that time were not limiting patient choice. Limited choice (and the increased potency limited choice conferred upon other HMO tools such as utilization review) gave HMOs and insurers that mimicked HMOs leverage over doctors and hospitals – leverage insurers used to force providers to give up huge discounts.  This in turn allowed those insurers to keep premiums lower than those of traditional insurers, which brought them more market share, which gave them more power to extract discounts from providers, and around the vicious cycle went.

Here is how two HMO executives, Georgia Halvorson (formerly the CEO of HealthPartners in Minnesota and of Kaiser Permanente in California) and George Isham (formerly a VP at HealthPartners) described this vicious cycle in a book they co-authored entitled Epidemic of Care published in 2005. “[T]hose … health plans [referring to HMOs] use volume purchasing leverage to reduce the fees charged by providers. Those plans … negotiated deep discounts – 30 to 50 percent discounts in many cases. These discounts cut health plan costs to the point that most plans could significantly under-price the old pre-HMO traditional insurance products.” (p. 63) “[M]any plans used their negotiating power without mercy against both physicians and hospitals.” (p. 64) “As HMOs grew, their negotiating leverage increased, and their discounts grew with them.” (p. 65)

The outcome of this vicious cycle – oligopoly and near-universal use of networks – created a chicken-and-egg problem for any entrepreneur contemplating creating a new insurance company. In a 2012 report  on the Republicans’ “across state lines” proposal, Sabarina Corlette and three other scholars at Georgetown University described the problem using the chicken-and-egg analogy: “Out-of-state insurers thus face a chicken-and-egg dilemma: They must build a sufficient membership to negotiate competitive rates with providers, but, to garner that membership, they must show customers they have an adequate network of providers and charge a premium that is comparable to their competitors.” (p 11)

Corlette et al. examined laws passed in six states between 2006 and 2011 that either permitted out-of-state insurers to enter those states or authorized the study of such a law. Only two of the six states, Georgia and Wyoming, had implemented such laws as of 2012, but no out-of-state insurer entered or even expressed an interest in entering. The laws in the other four states also fizzled.

Corlette et al. interviewed representatives of insurance companies and state officials to determine why these laws failed. The explanation they got from everyone they interviewed was the chicken-and-egg problem. Under the sub-heading, “It’s the network, stupid!” Corlette et al. wrote: “Respondents universally reported the enormous difficulty that out-of-state insurers face in building a network of local providers, and insurers identified doing so as a significant barrier to market entry that far surpasses concerns about a state’s regulatory environment or benefit mandates. This difficulty is compounded in states like Maine, Washington, Wyoming and Georgia, which face provider shortages in rural areas.” (p. 11)

It’s the oligopoly, stupid!

I would have said, “It’s the oligopoly, stupid!” The people Corlette et al. spoke to were essentially saying insurers in one state just couldn’t imagine walking into another state and creating an insurance company large enough to solve the chicken-egg problem. That’s not to say that insurance companies never enter new markets. But when they do, it is almost always via acquisition of a smaller insurance company in the new market.

If the large insurance companies that control today’s markets do not think they can solve the chicken-and-egg problem, why do Hacker and Trump entertain that fantasy? [4] It’s reasonable to infer that Trump has never given ten seconds of thought to that question. But I suspect Hacker has. In the fall of 2009 Forbes published an interview    with Hacker in which Hacker was asked, “What do you think of the idea for [sic] nonprofit co-ops instead [of a PO]?” Hacker replied: “I think it’s pretty much a joke. I don’t believe that the cooperatives will be able to fulfill any of the vital goals. They’re not going to be easy to create, and they’re not going to be able to provide a strong counterweight to private plans.” Bingo! Not easy to create, not likely to grow big enough. As we all know now, Hacker’s prediction was accurate. Only six of the 23 Obamacare co-ops that opened for business in the fall of 2013 remain alive today, and one of those six has suspended operations. I know of no evidence indicating the remaining five are going to “hold the insurance industry accountable,” as PO and co-op advocates like to say.

So Hacker has obviously made a distinction in his own mind between the PO and the co-ops. But why does Hacker make that distinction? Only Hacker knows.

No explanation for Hacker’s double standard, one for the co-ops and another for his PO, makes much sense, but of the hypotheses I can think of the one that makes the most sense is that Hacker has never understood that in the post-HMO era the “Medicare-like” PO he has talked about since 2009 would not leap into existence virtually overnight as the original Medicare program did. Rather, the PO would have to establish a customer base market by market across the entire country. Hacker’s PO could conceivably be run by a single national entity, but because of state licensing laws, because of the unwritten law that every insurance company must establish its own limited network these days, and because Hacker’s current version of the PO is not pre-populated as his original version was, the PO would in practice have to administer dozens, maybe hundreds, of separate local insurance companies. Each of those insurance companies would have to hire its own sales staff, finance its own advertising, negotiate its own limited network, collect its own premiums, process its own claims, etc. In short, it is far more accurate to describe Hacker’s current, aspirationally defined version of the PO as “like the co-ops” or “like Humana” than “like Medicare.”

More insurance companies practicing managed care is not the solution

Let us pretend for a moment that Hacker’s and Trump’s dreams come true. Let us imagine that Congress has not only been persuaded to enact their proposals, but by some miracle new insurance companies have sprung up in all 50 states and these new companies have sold so many policies and accumulated so much capital they have guaranteed their survival indefinitely. So what? Will the addition of one or two new insurance companies to today’s markets matter much? To put it in terms of the three-firm concentration ratio the GAO used in the quote I presented above, should we really care if the next GAO report says, “[E]nrollment was concentrated among the four largest insurers in most states rather than the three largest insurers we reported in December 2014?”

Asking this question forces us (and one would hope Trump and Hacker) to ask what if anything insurance companies bring to the table. What is it insurance companies would do to lower costs that they haven’t already tried over the last half-century if they were (miraculously) exposed to more competition? I can’t think of what that would be. More bureaucrats arguing with doctors about whether patients should have access to statin A instead of statin B? More boxes for doctors to click on to game a pay-for-performance scheme? More guidelines inflicted on doctors when doctors don’t have enough time to consult every guideline on the books now for every patient?

Hacker definitely thinks insurance companies have much to contribute. He is an avowed fan of the insurance industry and managed care, but his explanations for his admiration are as vague as his explanation for how his PO will become “like Medicare.” For example, he stated in his October 28 New York Times op-ed  “[P]rivate plans are much better poised to develop integrated systems that closely monitor outcomes for a smaller circle of providers.” Paul Ellwood, Alain Enthoven and the captains of the insurance industry could not have said it better.

But evidence does not support Hacker’s praise of the insurance industry. The evidence simply doesn’t support the claim that insurance companies have some un-tried trick up their sleeves that will lower health care costs once reinvigorated competition comes to pass. Insurance companies can only help by getting out of the way.  That applies to Medicare as well in the age of MACRA and Medicare Advantage. Congress should return Medicare to a simple fee-for-service program that does not seek to control volume of medical services by meddling in the doctor-patient relationship but focuses instead on price. And if Congress really wants to achieve universal health insurance, it should extend the original MACRA-free, HMO-free Medicare to Americans of all ages.

Which reminds me, there is one significant difference between Trump’s and Hacker’s positions on health policy. Trump was at one time a supporter of single-payer. Hacker never was.

[1] I do not discuss here the “piggyback” version of the PO that some Democrats support. Under the piggyback version, an established public program such as Medicare would sell insurance to the public. This version of the PO presents different problems from those presented by the original PO proposed by Hacker and endorsed by congressional Democrats in the summer of 2009. The single biggest problem the piggyback version faces is inaccurate risk adjustment of the subsidies given to individuals who buy into the public program. Unlike the original version of the PO, finding and retaining a base of enrollees is not an immediate issue for the public program and, depending on how the inaccurate risk-adjustment issue is handled, may never be an issue.

[2] Just as it is unclear whether PO advocates are proposing a PO for all non-elderly Americans or just for shoppers on state exchanges, so it is unclear whether proponents of letting insurers sell across state lines are proposing to let them sell to all non-elderly Americans or only to individuals. Trump’s website says his proposal applies to “all Americans,” but a bill introduced in Congress by Republicans would have applied only to the individual market. Here is a “finding” proposed at the beginning of HR 2355  , a bill introduced by Republicans in Congress in 2005: “Congress finds the following: (1) The application of numerous and significant variations in State law impacts the ability of insurers to offer, and individuals to obtain, affordable individual health insurance coverage, thereby impeding commerce in individual health insurance coverage.” (p 2, emphasis added)

[3] Here is another example of a PO advocate enthusiastically admitting how concentrated the insurance industry is, then claiming a PO would fix that problem. The writer asserts, “A public plan offers the promise of being able to enter these markets currently controlled by monopoly or oligopoly for-profit insurers.” No evidence precedes or follows this claim.

[4] The oligopoly problem that Hacker and Trump must contend with afflicts the hospital sector as well, and to a lesser degree the physician sector. Consolidation among providers accelerated in the 1980s, thanks to the spread of managed care tactics within the insurance industry. As Halvorson and Isham put it in Epidemic of Care, “[T]he deep discounts [managed care insurers extracted from providers] strongly encouraged providers to consolidate into larger organizations to create their own negotiations leverage.” (p. 69)