Does Competition Benefit Health Insurance Subscribers?
By UWE E. REINHARDT, New York Times, October 29, 2010
Uwe E. Reinhardt is an economics professor at Princeton. He has some financial interests in the health care field.
How does competition work in the health insurance market? Do more competitors in a market area — say, a state — invariably mean a better deal for the insured?
These questions are germane as regulators are busy implementing the Affordable Care Act, which imposes a series of stringent new regulations on the market for health insurance, particularly the market for small-group or individually purchased insurance.
After months of hard work, for example, the National Association of Insurance Commissioners on Oct. 21 submitted to Kathleen Sebelius, the secretary of health and human services, its recommendations for defining the minimum loss ratio that health insurers must meet as of Jan. 1.
As I noted in two earlier posts on the ratio, it expresses what portion of the premiums collected by insurers must be paid out in the form of “medical benefits.” For insurance sold in the small-group or individual insurance market, the portion must be at least 80 percent. For insurance sold in the large-group market it must be at least 85 percent.
One can debate whether, with properly conducted health insurance exchanges, this regulation of the minimum-loss ratio was actually necessary. I think not, because if both the benefits and the premium for them are clearly stated, I can comparison-shop. It’s not important for me to know how insurers allocate the premium.
But if it were to be imposed, it should have been phased in gradually, over, say, four years or so, as was done with the Medicare prospective payment system for hospitals introduced under the Reagan administration in the mid-1980s and the Medicare physician-fee schedule introduced by the George H.W. Bush administration in the early 1990s.
If major changes in the health system are imposed rapidly, they tend to generate untoward and sometimes unanticipated side effects that must then be addressed with ad hoc waivers or other adjustments, such as the esoteric “credibility adjustment” of the minimum-loss ratio, whose details most readers of this blog probably wish to be spared.
Now that the regulation is on the books, one wonders what impact it will have on the number of insurers competing in the market and what effect their numbers will have on the insured.
In a letter submitted to the insurance commissioners in May, the Council for Affordable Health Insurance reminded the commissioners that these states’ traditional “standards for individual-market loss ratios vary between 55 to 65 percent, depending on the type of plan.” The council, as well as the American Health Insurance Plans, argues that the effect of the regulation will be to drive many small insurers from the market and thus to reduce competition for health insurance.
I have no doubt that this prediction is on the mark.
The question is whether a reduction in the number of insurers in a given market area invariably is to the disadvantage of the insured. On this point one’s memories of Econ 101 can be quite misleading.
The narrative in Econ 101 centers on a producer of the legendary “widgets,” a standard, mass-produced, imaginary product whose characteristics satisfy all the conditions required for perfectly competitive product markets. Students are then shown convincingly that if only one producer offers widgets in the market — a monopoly — the price of widgets will be far above what it would be in a perfectly competitive market and the volume sold much below the competitive benchmark.
Things are a bit better for the buy side when there are a few producers-sellers — what we call an oligopoly — but widget prices will still be above the competitive benchmark and volume below it, especially when the oligopolistic sellers cooperative formally or informally.
Buyers will find the lowest conceivable widget price and can buy the largest number of them under perfect competition, in which many sellers and buyers interact and arbitrate, and market entry, as well as exit, is easy and cheap for the sellers.
One can see why, when this narrative is grafted onto the market for health insurance, the conclusion emerges that the more sellers in the insurance market, the cheaper health insurance and health care must be for the insured.
Evidently, President Obama and his Congressional allies profess to believe the narrative, as they advocate the public-insurance option or at least new nonprofit insurance cooperatives. Representatives of the insurance industry profess to believe the narrative as well. Perhaps most people actually do. In an editorial on Tuesday, The Wall Street Journal appears to deplore the likely shrinking of the number of insurers doing business in the United States (roughly 1,200), even though, as noted above, many small insurers pay out only 55 to 65 cents of every premium dollar they collect in medical benefits.
But is the facile Econ 101 narrative actually apt for the health insurance market of the real world?
That market does not trade in widgets. It trades in immensely complicated financial contracts that bundle into one deal a number of distinct product lines. Ideally, these product lines include:
1. A transfer of much or all of the financial risk of ill health from the insured to the insurer, facilitated by the statistical miracle of risk pooling;
2. Claims processing.
3. Management consulting services to help patients manage their health, including disease management.
4. Purchasing of health care on behalf of the insured, including:
a. bargaining on behalf of patients over theprices of health-care products and services
b. monitoring the quality of care delivered
c. monitoring the appropriateness of caredelivered.
Some widget that is! The premium paid by the insured can be thought of as a bundled payment for all of these distinct services.
So the question is whether production costs of each of these distinct product lines, and the prices that consumers are charged for them, will necessarily fall as the number of health insurers in a market rises. Do they?
I will explore this question next:
Competition’s Shortcomings in Curtailing Health Care Costs
By UWE E. REINHARDT, New York Times, November 5, 2010
Uwe E. Reinhardt is an economics professor at Princeton. He has some financial interests in the health care field.
In last week’s post, I asked whether we can expect the health-care costs of individuals and families to fall as the number of health insurers competing in a given market area increases – as was so often posited by all sides during the recent health-care debate. (I posed that same question earlier this year in the policy journal Health Affairs.)
As I said last week, a health-insurance contract is unlike a widget (which, in standard Econ 101 discourse, is a simple, standard manufactured good whose price tends to fall to the lowest possible level as more producers enter to compete in the market for widgets). Rather, health insurance, in exchange for one premium, provides what is known as a tie-in sale of a set of services produced by the insurers, along with the care purchased on the insured’s behalf from doctors, hospitals and other providers.
Consider the production cost of the set of services produced by a health insurer, which I enumerated last week — marketing to enrollees, risk pooling, claims processing, disease management, information services and more.
There are apt to be significant economies of scale in the production of most of the services produced by insurers — and that means the unit cost of these services is likely to fall as scale increases, especially for services that rely heavily on electronic-information systems. Why else would small insurers, represented by the Council of Affordable Health Insurance, need 35 to 45 percent of premiums for marketing, administration and profits, while large insurers manage with a much smaller share, often less than 20 percent, devoting a correspondingly larger piece of the premium to paying for health care?
Then consider the cost of the care produced by hospitals, doctors and other providers and purchased for the insured by the insurer, at prices determined in negotiations between insurer and provider.
Common sense suggests that the total cost at which an insurer can procure a given set of health-care benefits from a given provider depends on the insurer’s and provider’s relative market power.
If an insurer would find it difficult to sell policies without having a particular provider in the insurer’s network, then that provider has significant market power over the insurer in negotiations over prices.
This condition prevails in areas where physicians join together in large groups or hospitals consolidate in large systems, as has happened in many markets in the last two decades. Studies have shown, for example, that consolidation in the hospital industry has raised the prices of hospital services significantly.
If a provider cannot afford to lose the patient revenue that an insurer could bring to that provider, the insurer will have more market power over the provider in negotiations over prices, other things being equal. This condition is more likely when an insurer has a large market share in any given area.
If more insurers enter a given market, it will be easier for providers to walk away from the table in negotiations with any one insurer over prices. Consequently, insurers in such markets will pay relatively higher prices. This price discrimination is perfectly legal and rampant in American health care.
An insurer, of course, might seek to reduce the total payments made to a provider not only through lower prices but also through controls of what and how much of a service is provided to patients. When an insurer accounts for a large portion of a hospital’s revenue, a provider who declined to contract under those conditions would lose significant revenue and is therefore likely to put up with the controls, however nettlesome.
It is safe to conclude, then, that the smaller the share that an insurer has in a particular provider’s total revenue, the more it will cost that insurer to procure from that provider the health-care benefits promised the insured. Thus the reduction in health-care costs some people expect from a larger number of insurers is not likely to come on this front, either. In fact, costs are more likely to rise.
This leaves the insurer’s profits as the only line item that might possibly be reduced through competition. These profits can be expressed either as a percentage markup over the insurer’s cost of those services it produces itself, in which case the markup will appear large, or as a fraction of the premium charged the insured, in which case the markup will appear to be much smaller.
From the perspective of the total health-care costs borne by the insured, the relevant profit measure is the fraction of the premium absorbed by the insurer’s profits. That fraction may well fall with an increase in the number of insurers competing, but the total relief for the insured is likely to be very modest.
Over all, however, the cost savings from economies of scale and the price discounts from the added market power enjoyed by larger insurers are apt to swamp, in most cases, any downward pressure on profit margins that might be had from a larger number of insurers.
Econometric research on the effect on prices of the allocation of market power between insurers and providers is challenging, because price can be affected by so many variables and the direction of causal flows (the influences that variables might have on one another) may be more complex than can be modeled with the data at hand. Even so, what research there is tends to support my arguments, such as a nationwide study published earlier this year by Asako Moriya, William Vogt and Martin Gaynor.
A very recent study by Glenn Melnick and associates, currently under peer review and not yet published, explores the effect of the allocation of market power among insurers and hospitals and finds, as theory would predict, that “health plan market concentration reduces hospital prices, while higher hospital market concentration results in higher prices.” The authors also find that “hospital concentration exceeds health plan concentration in most markets.”
In short, the widely held notion that more insurers in a market area will reduce the premiums paid by the insured is not supported by either economic theory or empirical research.
3