James Schneider  

Health Insurance as Collective Bargaining

Desert versus Identity... Who is the Italian prime minis...

McKinsey has studied the hospital networks that are being used by plans on the new public health exchanges. They divide networks into three groups that they call broad, narrow, and ultra-narrow. Broad networks contain more than 70 percent of the 20 largest hospitals in the relevant area, whereas ultra-narrow networks contain less than 30 percent. McKinsey found that over two-thirds of the networks are either narrow or ultra narrow. (Download the study here.)

McKinsey takes a benign view of the situation. They feel that these narrow networks are expanding consumer choice. Offering a narrow network is one of the few really effective means that health insurers have left to restrain costs. If insurers expected the exchanges to attract price-sensitive consumers, insurers would naturally offer networks that were narrower than their typical offerings. In the McKinsey study, policies with broad networks were 26 percent more expensive than equivalent policies with narrow networks. This is a pretty substantial price differential. McKinsey found that hospitals included in the narrow networks were comparable in quality to those being excluded.

Although narrow networks shouldn't be surprising to health policy wonks, they have shocked many patients. Recently Megan McArdle described the growing pushback against the the ultra-narrow networks. She thinks that they may not be "long for this world."

Regulatory overreach is a real possibility. The downside of narrow networks is instantly recognizable; however, the benefits are a little harder to explain. Many people assume that narrow networks save money primarily by excluding the expensive hospitals. However, insurers do more than shop for a good deal, they are actively bargaining on behalf of their customers. It's not just that "you get what you pay for," but "you get what you pay for" after insurers get the best deal possible. Essentially, an insurance company engages in collective bargaining based on the fact that they can steer a large number of patients to providers in their network. Insurers can steer more patients to a given provider by limiting the number of providers in the network.

Essential to the insurers' bargaining power is the option to exclude providers. Imagine a situation where there were two hospitals in town and the insurer was philosophically opposed to excluding either hospital. The insurer approaches the first hospital and says, "We are bargaining on behalf of our customers who represent 10 percent of the population in your area. We think this merits a substantial discount off your usual charges."

The hospital negotiator says: "Interesting. That many customers would really come in handy. But what could you do to make sure that your customers come to our hospital instead of the hospital across town?"

The insurer replies: "Nothing. Our customers need access to all the hospitals in the area. Steering patients to your hospital would be a great disservice to them."

The hospital replies: "If I'm going to get half your business regardless of my rates, why would I offer you any discount at all. Hmmm ... by law, your policies now have no lifetime limits. And with the caps on out-of-pocket costs, it might make sense for us to raise our rates. I wonder if we are charging heart patients enough for the aspirin that we give them."

The insurer reconsiders: "Okay, what if we exclude the other hospital in the neighborhood from the network. If we do this, you can be pretty certain that you'll get 90 percent of our business."

The hospital negotiator smiles, "For that we can offer a steep discount."

Networks can be maddening. Inadvertently using out-of-network providers can be financially draining. However, the more we insulate patients from the cost of medical care, the more important it is that insurers can effectively bargain with providers.

Comments and Sharing

COMMENTS (13 to date)
Shayne Cook writes:

Your hypothetical scenario of "collective bargaining" has several logical/practical flaws and is based on a couple of defective assumptions - as follows ...

1. No form of "exclusive rights" contract ever, ever, ever results in "steep discounts" for the final purchaser. Not ever. The sole purpose of exclusive (market) rights agreements is to allow pricing to be unaffected by competition.

2. The rational incentive for hospitals or other health care providers - especially the "small" ones - is to price their services at the same rate as the highest priced competitor. Your hypothetical scenario hospital negotiator reaction would only be plausible/rational if it were currently operating well below its current "production capacity" - either due to grossly high prices, or grossly inferior quality.

3. Under ACA, insurers are (ostensibly) limited to a fixed 20% "profit", over and above the actual health care reimbursement costs they bear. With a fixed percentage return - similar to that allowed real estate agents - the incentive for insurers is to elevate actual health care costs in order to elevate their fixed-percentage returns. Have you ever gone to a real estate agent to buy a home and have them argue that you buy a lower cost house than you can "afford"?

The biggest fallacy of the ACA is that it will "bend the cost curve" or reduce health care costs. That is not going to happen when health care providers AND insurers are incentivized to elevate health care costs, not constrain them. The only possible constraint to any service providers prices are those established by the highest-priced local competitor. Want proof? Notice that health insurance AND health care treatment costs vary widely and dramatically across the country. Given time - and relatively short time - all health care services and insurance costs will "gravitate" to those of the highest cost providers/insurers - and continue to rise from there. There is NO INCENTIVE anywhere in ACA to constrain or lead to to lower health care prices/costs.

Shayne Cook writes:

Follow-up ...

I just stumbled across an article by John F. Wasik, originally appearing in/for "The Fiscal Times" - and linked to at finance.yahoo.com. It's a good read. The third and second to last paragraphs in the article pretty much sums up the real intent and purpose of the ACA ...

"What will a more "universal" plan look like? Ironically, it may take the shape of an expanded Medicare-like program with limited out of pocket costs, small premiums and no link to employment.

Such an idea was recently floated by Donald Berwick, who formerly headed the Centers for Medicare and Medicaid Services under President Obama. Berwick is also running for governor of Massachusetts, where he would like to see a state-run single-payer health system launched."

Curious. The "Massachusetts Model" of compelled, mandatory private health insurance - that has been argued as the "proven workable model" for ACA - is now all of a sudden up for replacement in Massachusetts (and Vermont) as unworkable - and in need of replacement with a (government) single-payer system, proposed by none other than the former head of Obama's head of the Centers for Medicare and Medicaid Services. Whodda thunk it?

It seems that Massachusetts (and/or Vermont) is/are destined to be the new "proven workable model" for the new nationalized, "single-payer" health care nationwide.

For all of the reasons and incentives contained in ACA that I mentioned in my first comment here - and more - it is glaringly apparent that ACA was explicitly and specifically designed to be so unworkable, so ineffective, so costly, that a nationalized, government-run (taxed), "single-payer" replacement system would appear preferable to the electorate. Or at least a simple majority of the electorate.

Hazel Meade writes:

Alternative interpretation:

Very narrow networks allow realtively healthy people to avoid subsidizing the sick, by effectively purchasing a policy that really sick people will not want.

It's the market finding a way around community rating. The relatively health will opt for an ultra-narrow network to minimize the price of insurance. The relatively sick will opt for a more-expensive plan with a wide network.
That allows the price differential between the healthy and the sick to adjust beyond the 3/1 ratio.

Steve Roth writes:

"McKinsey found that hospitals included in the narrow networks were comparable in quality to those being excluded."

That's certainly not true in particular markets.

In Seattle, only one insurer covers *any* of the city's important hospitals in-network:

Children's Hospital
Harborview (the regional trauma center)
Swedish except one outlying branch
Seattle Cancer Care Alliance

This works for insurers because of information asymmetry. Not nearly all information is known.

Insurance buyers just don't know what hospitals they might need when the worst case arises (or even what the important hospitals in town *are*). And determining which ones are in-network using the online tools is like playing a lengthy game of battleship -- which you have to play on multiple insurers' sites, while building a table to keep track of it all.

A vanishingly small percentage of buyers will do that due diligence.

Happily, the press sometimes does this legwork for us:


But again: how many buyers will know about this article, understand its importance in protecting themselves from catastrophic financial risk, or incorporate its findings intelligently into their buying decisions?

The Myth of the Rational Shopper.

Shayne Cook writes:

to Hazel Meade:

Your "Alternative interpretation" is based on sound, logical and rational thoughts - as far as it goes.

However, there is a very "telling" paragraph in the Seattle Times article, to which Steve Roth provided the link above, that explains why "low-price" and/or "small-network" insurers won't last long, even applying your interpretation ...

"If a patient needs a covered service, such as a heart transplant, but it’s not provided at in-network hospitals, insurers must cover it elsewhere."

As I illustrated above, ACA is designed to eventually gravitate both insurer and provider costs to the highest cost current providers. It's inevitable.

Steve Roth writes:

Shayne Cook:

Thanks for pointing out that line in the ST article. I read it several times and still have no idea what it means. Can you explain what you think it means?

Yes insurers have to cover some OON costs, but a far smaller percentage. In particular, many plans have unlimited out-of-pocket maxes for OON costs.

The means that insurers vastly limit their financial risk, while people's insurance doesn't protect them from financial catastrophe, which should be the primary purpose of coverage.

Very few people understand this. Insurers understand it very well. This is a recipe and formula for an incomplete market.

Hazel Meade writes:


That is only true if there are NO providers in the network that cover the service at all.
It doesn't count if there is a provider, but also a long waiting list, or if the provider is a long drive away.

If you go out of network to get a shorter waiting list or a more convenient drive, it is at your own expense.

Hazel Meade writes:

Also, I will add that single payer is never going to be politically palatable in the US, because, by definition, single-payer systems require banning private fee-for-service health care. Most advocates of single payer don't understand this. They think single payer means "free healthcare for all". They don't understand that it means making it illegal to pay out of pocket if you want to skip the waiting list.

Shayne Cook writes:

Steve Roth (and Hazel Meade):

What I think it means is precisely what is stated - "insurers must cover it elsewhere". How that will play out under all possible OON scenarios is yet to be determined. But the underlying law in ACA is that in order for an insurer to be considered ACA-qualified insurer, it must contain no dollar limits to coverage.

ACA hasn't been in place long enough, (nor is it completely in place now, given the ad hoc "relaxations" to business and insurance companies from the White House), to really determine what has happened or will happen in these sorts of instances.

But I gravely doubt that an allegedly ACA-qualified insurer would survive a patient's court challenge if it tried to deny payment for a legitimate procedure delivered by an OON provider under any number of scenarios/circumstances.

The "If you go out of network to get a shorter waiting list or a more convenient drive, it is at your own expense. " argument is indefensible in court for any insurer that claims to be, or has operated under a claim to be, fully ACA-qualified.

As I said, it's too early to know if I am right on this - such cases haven't been encountered as yet and consequently haven't been argued in courts (to my knowledge). But they will be.

Hazel Meade writes:

Shayne, well, that remains to be seen.

My point is that, as written, the ACA does allow insurers to deny coverage for OON providers, as long as you COULD get the same treatment in-network.

Of course the COULD leaves a lot of wiggle room.
Is a long wait unreasonable? If so, how long?
Is a long drive unreasonable? If so, how long?

No doubt, someone will take them to court at some point, but in the meantime, the default state of affairs is going to be that insurers will be denying OON claims as long as there is some provider somehwere in the network that the patient could theoretically see.

Shayne Cook writes:

Hazel Meade:

You are correct that what I'm postulating hasn't been tested yet - "remains to be seen".

I'm not convinced that the ACA allows insurers to deny coverage for OON providers if the same (or comparable) treatments are available in-network. I suspect that if it came to a court case, the insurer would still be compelled to pay (not deny payment), but perhaps only to the extent of the charges an in-network provider would have charged. That is certainly plausible.

James Schneider writes:

@Shayne I don't really understand the first 2 points of your original post so I'm just going to address #3

The ACA essentially limits profit plus the administration costs of the company to 20% (in the individual and small group market). The latter will almost always be greater than the former. There is no fixed return. Let's say that the company was operating at the 80% minimum loss ratio. To make room for an additional $1 of profit, claims would have to go up $4. This might work in a monopoly situation, but in general a company is going to lose a lot of business when their rates increase relative to the competition.

The ACA also has requirements about covering emergency out of network benefits. But by and large, a company can still design a contract where they don't pay anything for typical out of network claims.

@Hazel Your alternative theory is pretty insightful.

Shayne Cook writes:

James Schneider:

On #3 ...
Keep in mind that what you are seeing transpire right now - negotiated prices between service providers and insurers within "networks" (irrespective of size) - is an exercise in insurers attempting to have some initial basis for actuarially determining their exposure under the new ACA. Those negotiations are not being conducted for the purpose of constraining or reducing costs to the insured.

Your statement, " To make room for an additional $1 of profit, claims would have to go up $4" is arithmetically correct. And that is precisely why insurers have exactly zero incentive to reduce total claims or prices of individual claims. Insurers have exactly the opposite incentive - in order to increase their administrative-cost/profit returns they are incentivized to increase claims payments.

Your next statement, "This might work in a monopoly situation, but in general a company is going to lose a lot of business when their rates increase relative to the competition", isn't quite right. The exclusive rights agreements nature of "networks" has some characteristics of monopoly, at least with respect to pricing freedom. Insurers may lose some of their clients with premium increases, but only if they raise their premiums substantially above, not equal to the competition. And then only if they don't justify the higher premiums with offsetting "benefits" such as superior quality, larger network of providers or any number of other perceived goodies.

Your statement in the third paragraph, "But by and large, a company can still design a contract where they don't pay anything for typical out of network claims", may be true. An insurer can certainly design such a contract. But I strongly suspect that such a contract would be ruled unenforceable in a court challenge, as I indicated above.

Such a contract would be in violation of major tenets of ACA. I'll readily admit I can't say for certain that will happen. I merely strongly suspect it will happen - if challenged in court. I don't see a jury allowing an insurer to deny a claim payment just because the service was rendered by an OON provider. At best/worst I would expect a jury to rule the insurer cover the claimant, at least to the degree/amount it would have paid had the service been rendered in-network. Besides which, the insurer suffers no loss. Any "loss" is suffered only by the in-network health care provider that didn't get paid to render the service.

On my first comment points #1 and #2 ...
The entire model of health care delivery "networks" is based on "exclusive rights agreements" - with the in-network health care providers having (ostensibly) exclusive rights to the clients of the insurer. Similar exclusive rights agreements are built into most every franchise you've ever seen, such as a MacDonald's franchise. When a franchisee purchases the franchise from the parent (MacDonalds, for example), the franchisee is granted exclusive rights to the geographical area. All that means is that the parent company will not sell additional franchises to others for operation within that geographical area - the original franchisee will not have to suffer competition (price wars) from another franchisee. The MacDonalds franchisee does however have to compete with other dissimilar franchises (Wendys, Taco Bell, etc., etc.). But they rarely compete on cost/price basis - that is a zero-sum game for all competitors. Instead, they compete on product differentiation.

Comments for this entry have been closed
Return to top