Death Panels vs. The Herfindahl-Hirschman Index

by: Paul Rosenberg

Sat Aug 22, 2009 at 12:30


Everyone's heard of "death panels" it seems.  But what the heck is the Herfindahl-Hirschman Index?  And what does it have to do with health care reform?

Well, before I explain it precisely, let me just start off by comparing to death panels.  First off, the Herfindahl-Hirschman Index exists.  Death panels do not.  Second, death panels go well with Carmina Burana.  The Herfindahl-Hirschman Index?  Not so much.  Maybe a Bach Fugue.  Or Music For 18 Musicians. Third, the Herfindahl-Hirschman Index actually does have something to do with health care reform.  Death panels do not.

Okay, I know the suspense is killing you.  So what is the Herfindahl-Hirschman Index, anyway? Simple: it's a measure of market concentration, which allows us quantify precisely how concentrated a market is.  You take the market share (percentage) of each participant in a given market, and square it.  You add up all the squares and that's the Herfindahl-Hirschman Index (or "HHI" for short, now that I'm in demystification mode.).  If there are 100 market participants, all with an equal 1% share in the market, the HHI is 100.  If there's just one company-a complete monopoly, the HHI is 10,000. If there's 10 companies each with an equal 10% share, the HHI is 1,000.  But the HHI allows us to compare any configuration, not just these easy-to-describe cases.  (I first learned about the HHI last December; I'll tell a little story about that at the end of this diary.)

The Federal Trade Commission and the Department of Justice Anti-Trust Division regard markets with an HHI of less than 1,000 as "competitive", those with HHI's of more than 1,000, but less than 1,800 as "concentrated," and those with an HHI of 1,800 or more as "highly concentrated."  A 2007 report by the AMA, "Competition in health insurance: A comprehensive study of U.S. markets (pdf)," covering 44 states and 313 Metropolitan Statistical Areas (MSAs) found that virtually all the combined PPO & HMO markets in country were highly concentrated.  (As were the PPO and HMO markets considered separately.) Here's a chart I generated of the state markets:

And here are the MSAs:

The extreme concentration shown in these charts is a clear and compelling reason why there's a need for a public option to ensure the existence of a genuine competitive market.

More background and information from the AMA survey on the flip.

Paul Rosenberg :: Death Panels vs. The Herfindahl-Hirschman Index
The report explains government use of the HHI as follows:

HHI analysis

The 1997 Federal Trade Commission (FTC)/DOJ Horizontal Merger Guidelines (1997 Merger Guidelines) define market concentration as measured by the HHI as follows:

  • Markets with an HHI less than 1,000 are "not concentrated." The DOJ and FTC will generally not restrict merger activity in markets where the post-merger HHI is less than 1,000.
  • Markets with an HHI between 1,000 and 1,800 are "concentrated." Under the 1997 Merger Guidelines, a merger in these markets that raises the HHI by more than 100 points may raise significant competitive concerns.
  • Markets with an HHI above 1,800 are "highly concentrated." Under the 1997 Merger Guidelines, a merger in these markets that raises the HHI by more than 50 points may raise significant competitive concerns, and mergers that raise the HHI more than 100 points are presumed to be anti-competitive.

Under the 1997 Merger Guidelines, barriers to market entry and other qualitative factors are also considered to determine whether it is likely that a merger will result in market power for the merged entity.

In summarizing their findings, the report first addressed concentration in metro areas, which is what actually describes the market as individuals, families and small businesses experience it directly:

A. Metropolitan areas
This edition of the study analyzed 313 MSAs. This compares with 292 metropolitan areas in the 2005 study, 84 in the 2003 study, 70 in the 2002 study, and 40 in the 2001 study.

In terms of market concentration (HHI), the study found the following:

  • In the combined HMO/PPO product market, 96 percent (299) of the MSAs are highly concentrated (HHI>1,800), applying the 1997 Merger Guidelines.
  • In the HMO product market, 99 percent (309) of the MSAs are highly concentrated (HHI>1,800), applying the 1997 Merger Guidelines.
  • In the PPO product market, 100 percent (313) of the MSAs are highly concentrated (HHI>1,800), applying the 1997 Merger Guidelines.

In addition to the HHI, the study also included an analysis of dominant individual insurers.  The main findings were summarized as follows:

B. Market share
In terms of market share of individual insurers, the study found the following for each product market:

HMO/PPO product market

  • In 96 percent (299) of the MSAs, at least one insurer has a combined HMO/PPO market share of 30 percent or greater.
  • In 64 percent (200) of the MSAs, at least one insurer has a combined HMO/PPO market share of 50 percent or greater.
  • In 24 percent (74) of the MSAs, at least one insurer has a combined HMO/PPO market share of 70 percent or greater.
  • In 5 percent (15) of the MSAs, at least one insurer has a combined HMO/PPO market share of 90 percent or greater.

HMO product market

  • In 98 percent (306) of the MSAs, at least one insurer has a HMO market share of 30 percent or greater.
  • In 64 percent (201) of the MSAs, at least one insurer has a HMO market share of 50 percent or greater.
  • In 37 percent (117) of the MSAs, at least one insurer has a HMO market share of 70 percent or greater.
  • In 16 percent (49) of the MSAs, at least one insurer has a HMO market share of 90 percent or greater.

PPO product market

  • In 97 percent (304) of the MSAs, at least one insurer has a PPO market share of 30 percent or greater.
  • In 76 percent (238) of the MSAs, at least one insurer has a PPO market share of 50 percent or greater.
  • In 36 percent (112) of the MSAs, at least one insurer has a PPO market share of 70 percent or greater.
  • In 9 percent (28) of the MSAs, at least one insurer has a PPO market share of 90 percent or greater.

In the tables below, I present the data from all 44 states covered, along with the most- and least-concentrated MSA in each state.



Those are truly astonishing levels of concentration--levels that most other industries could only dream of. In light of all the above, there should be no doubt whatsoever that the health insurance market in America is anything but a competitive free market.

Indeed, this was not the first time the AMA had conducted such a survey.  And matters are only getting worse, as they noted:

Over the five years since the AMA's first study, the country's largest health insurers have continued to pursue aggressive acquisition strategies. The largest insurer, WellPoint Inc. (formed from the merger of Anthem Inc. and WellPoint Health Networks), has acquired 11 health insurers since 2000. The second-largest health insurer, UnitedHealth Group (United) has also acquired 11 health insurers since 2000.

To put this in perspective, in 2000, the two largest health insurers, Aetna and United, had a total membership of 32 million lives. As a result of mergers and acquisitions since 2000, the top two insurers today, WellPoint and United, each have memberships, respectively, of 34 million and 33 million, totaling more than 67 million covered lives. Together, WellPoint and United control 36 percent of the national market for commercial health insurance. In 2004 and 2005, 28 mergers valued at a total of $53.8 billion were completed or announced, which exceeded the value of all the deals completed in the previous eight years. (Corporate Research Group, The Managed Care M&A Explosion, 2005).

Observers predict that large health insurers will continue to acquire their smaller competitors. WellPoint's new chief executive officer stated in February that mergers will be one of the key drivers of WellPoint's future growth. Further, in March, United announced its proposed acquisition of Sierra Health Services, the largest health plan in Nevada. The AMA has asked the U.S. Department of Justice (DOJ) to block the merger, because if the merger is approved United will control 56 percent of the Nevada marketplace (compared with its current 11 percent market share).

Indeed, they've been at war with the FTC and the DOJ for a while now.  In March 2005, former AMA President Donald J. Palmisano, M.D., J.D. wrote:

Health insurance markets out of whack

The American Medical Association (AMA) has a very different perspective on health insurance markets from that reflected in the Federal Trade Commission (FTC)/Department of Justice (DOJ) report, A Dose of Competition, and in the January 2005 Physician News Digest interview with Mark Botti, an attorney with the DOJ. Our perspective is based on objective data and the experience of practicing physicians.

The AMA is very concerned that federal regulators continue to turn a blind eye toward the reality that in much of the country health insurance markets are not competitive. Health insurers have amassed significant market power through mergers and acquisitions but have received minimal scrutiny. Mr. Botti himself can point to just one instance where the federal agencies have challenged a health insurer merger: the 1999 Aetna/Prudential merger.

In contrast, physicians have been placed under a far higher level of scrutiny than is warranted by their comparative economic strength in today's market. In April 2002, the FTC announced its intention to "find and bring" cases against physicians. Between April 2002 and December 2004, it brought 21 complaints against physician entities. Of these, 20 entities decided to settle rather than engage in a protracted and financially devastating legal battle with the FTC. This scrutiny is misplaced given that physicians are by far the least consolidated component of the health insurance industry.

It is important to remember that physicians play a critical role as patient advocates in an environment where health insurers have limited accountability for decisions that impact patients. The physicians' role as patient advocate can be undermined when they have no leverage in negotiating contracts with health insurers. The AMA has found that in much of the country physicians face a true David and Goliath battle when dealing with health insurers

This article was published in the Physicians News Digest shortly after an earlier edition of the 2007 study was published.  The AMA is hardly the most progressive organization around.  But they have considerable clout and they make an excellent point here--one that is certainly not just in their own self-interest. And yet, how many here have ever even heard about the HHI?

In general, we know about this problem only anecdotally, when this study shows that there is overwhelming data available.  Why isn't that data part of the debate?

Could it be because including that data in the debate would simply end it?

How I Learned About The Herfindahl-Hirschman Index

I learned about the Herfindahl-Hirschman as a result of my coverage of air pollution issues around the ports of LA and Long Beach.  Over the past several years, the two ports, under great public pressure over a period of many years, have begun working on a joint plan to reduce air pollution.  It's called the "Clean Air Action Plan."  One of the most important and controversial components of this is their plans to reduce truck pollution.

I say "plans" because the two ports have passed separate plans which share a number of elements in common, while differing in a number of significant ways.  Even though the heart of the plans is a tremendous give-away subsidizing the mother of all "cash for clunkers" programs, the trucking industry is fighting it tooth and nail, and has had the support of a wide range of powerful allies, including the Federal Maritime Commission (FMC), which has not gone to court in a truck-related case since the early 1970s--long before the current law governing its authority was passed.  To say they know nothing about truck-related decisionmaking is an understatement--they "know" plenty of things that just aren't so.

One of things they "know" is that the combined plans of the two ports would create a dangerously concentrated, uncompetitive market that would drive up trucking costs.  Of course, their plans would drive up trucking costs, because the truckers are woefully underpaid, lacking any sort of bargaining power, which is one reason why the trucks they drive are the dirtiest trucks on the road. Changes in the plans would undoubtedly drive up costs--but not because they would decrease competition.  And it was in a deposition for the ports in fighting the FMC's court filing that I first learned of the HHI.

The FMC's chief economist, Roy Pearson, claimed that "key aspects" of the ports' to plans "will likely transform the drayage [port trucking] market from a perfectly competitive market to a severely constrained market."

Since we're talking about tens of thousands of truckers--most of them mis-classified as "independent owner operators", working for over a thousand different trucking companies, the claim is utterly ludicrous, once you learn about the HHI.

The Port of LA's plan was specifically intended to consolidate the trucking market--for the simple reason that under-capitalized firms could not be counted on to maintain new trucks, even if the initial purchases were very heavily subsidized.  A second, equally important reason is that firms must reach a certain size before they have enough business to ensure that drivers will not make a great number of trips carrying no load, thus polluting the air needlessly.  But reaching the size for a company to operate successfully would still leave room for hundreds of companies--and a market with hundreds of competitors simply cannot become concentrated in the way that the HHI measures, unless anti-trust regulations are purposely neglected, and a few large players are allowed to dominate.  Merely reducing the number of firms from ~1000 to ~200 would have no anti-competitive effect on pricing at all, according to the government's own guidelines, chief among which are the HHI and instructions for employing its guidance.

And that's how I learned about the HHI, through a corporate scam, backed by Bush Administration appointees, which tried to allege a dangerous lack of competition would result, where the HHI clearly indicated that that was utterly absurd.

The situation with health care plan markets is exactly the opposite--it's hyper-concentrated, very anti-competitive, and the big boys are fighting tooth and nail precisely to keep it that way.

In both cases, the HHI is a readily-understandable tool for separating truth from fiction.

Ain't math grand?


Tags: , (All Tags)
Print Friendly View Send As Email

Even Obama admits that universal coverage... (4.00 / 1)
...can ONLY come from a single-payer plan.

Or, with all the "public option" plans considered so far, we can have less than universal coverage that doesn't even start until 2013...

and that further subsidizes the for-profiteers (thru mandates) -- I mean these guys obviously need and deserve our help, like Goldman Sachs did, right? Percentage change since 2002 in average premiums paid to large US health-insurance companies: +87. Percentage change in the profits of the top ten insurance companies: +428.

Experience:
Portion of its membership that Washington State's subsidized health plan intends to lose this year: 1/3.
Average percentage by which it is raising its premiums to do so: 70.

Since "expropriate the expropriators" oddly doesn't get much support in a forum titled "OpenLeft," I'll just quote liberal icon Keynes: "Euthanize the rentiers."  


It's Not So Much A Question Of Support (4.00 / 1)
The emphasis here is as much on "open" as it is on "left."  And since we're rather focused on the legislative struggle here, we're definitely at the sweet spot where Keynes and Marx coincide for all practical purposes.

But of course, this diary is actually about something much more modest--albeit quite devastating for the powers that be--and that's the difference between the "free markets" of ideology, and the capitalist accumulation via any means necessary of reality.

Which is why I end my diary with that ancient Greek anti-authoritarian battle cry, "Ain't math grand?"

"You know what they say -- those of us who fail history... doomed to repeat it in summer school." -- Buffy The Vampire Slayer, Season 6, Episode 3


[ Parent ]
Check out that metro area with a HHI of 9100! (4.00 / 1)
I feel sorry for the BC/BS competitor there!

Well, There's No 10,000 HHI (0.00 / 0)
So the market must be doing something right!

Right?

"You know what they say -- those of us who fail history... doomed to repeat it in summer school." -- Buffy The Vampire Slayer, Season 6, Episode 3


[ Parent ]
Btw, totally OT, just an idea (4.00 / 1)
Progressives should put a graqh showing how many uninsured people died over the last decade on one panel, and a graph showing how many people died while their insurance companies denied them coverage on another, and call these "The Real Death Panels".

Hmm, how about that?  


Oh, and of course the panels should be made of teak! (0.00 / 0)
We progressives respect the vote of the people, right?
(even though I think American Redwood but be more politically correct. Dontcha all always say "buy american"?)

[ Parent ]
awesome post Paul, awesome (4.00 / 2)
I see the area marked (1) on both graphs is empty

guess what the (1) area represents folks?

competitive markets!

or am I reading this incorrectly?


That's Right! (4.00 / 2)
No competitive mess!

Health care industry profits don't just grow on trees!

"You know what they say -- those of us who fail history... doomed to repeat it in summer school." -- Buffy The Vampire Slayer, Season 6, Episode 3


[ Parent ]
so, every instance of a state market has an HHI of 1500 or more! (0.00 / 0)
a "free market"!

[ Parent ]
Yes, EVERY State. This Is What Happens "Accidentally On Purpose." (4.00 / 2)
Or, as policy wonk Britney Spears would say, "Oops! I did it again..."

"You know what they say -- those of us who fail history... doomed to repeat it in summer school." -- Buffy The Vampire Slayer, Season 6, Episode 3

[ Parent ]
Big markets, California (0.00 / 0)
Your list is only partial.

Going through the list of 313 in your link, 14 markets are concentrated, 299 are highly concentrated.  None are competitive or even that close to it. The 14 markets under 1800 include both NY City metro (NY-White Plains-Wayne) and LA metro (LA-Long Beach-Glendale) as well as 8 others in California, 2 big markets in Florida (Miami and Orlando), Akron and Colorado Springs.

Fwiw, San Diego-Carlsbad-San Marco at 1393 is the nation's most competitive market followed by Modesto, Riverside, metro NYC, Stockton, Miami, and Akron. Then comes San Jose, Orlando, Santa Ana-Anaheim, Colorado Springs, LA-Long Beach, and Oxnard-Thousand Oaks-Ventura.


[ Parent ]
Indeed (0.00 / 0)
I did notice some of those things as I was working with the data.  

Two thoughts occurred to me.  First, I wondered about a possible correlation between metro size and HHI.  Second, I wondered whether such a correlation might exist, but not always be as meaningful as it might seem.

Specifically, I wondered if the larger metro markets may not actually work in a less coherent, more fragmented manner, so that for practical purposes the consumers actually experience something that's more like an environment with a higher HHI.

Of course, if the correlation was very strong, it would be hard to imagine that hidden factors could overwhelm the trend.  But, if, for example, certain providers are tightly associated with certain types of employers, there could be population pockets that don't have as many potential providers who are actually out there competing for their business.

"You know what they say -- those of us who fail history... doomed to repeat it in summer school." -- Buffy The Vampire Slayer, Season 6, Episode 3


[ Parent ]
Great work, as always, pablo... (0.00 / 0)
"The extreme concentration shown in these charts is a clear and compelling reason why there's a need for a public option to ensure the existence of a genuine competitive market."

Isn't this also a pretty good index as to why the Health Insurance parasites are so adamantly opposed to competition from the (never on-the-table, according to Olympia Snow) "public option?"  


Oh For Sure! (4.00 / 1)
You don't get where they are without having a killer instinct for anything remotely threatening to your interests.

They make the Mafia look like pikers.

"You know what they say -- those of us who fail history... doomed to repeat it in summer school." -- Buffy The Vampire Slayer, Season 6, Episode 3


[ Parent ]
Maggie Mahar again (with a hint of Robert Fuller's power) (4.00 / 1)
"Put simply, over the past 25 years, power in our health care system has shifted from the physician to the corporation." The power struggle goes on... And we learn from her that the insurance industry refers to being uninsured as "going naked." There's a putdown - maybe it's rankism?

[ Parent ]
That's true of every significant part of Murkin life (0.00 / 0)
wherein previously there was a professional/expert who has been replaced by a committee of bean-counting chickenshit, dumb-ass, sycophantic MBAs...

[ Parent ]
Is HHI sensitive enough? (0.00 / 0)
Paul, the absence of any 'competitive' markets according to the Herfindahl-Hirschman Index was shocking to me. But then it made me wonder if the HHI is in fact well-suited to assessing the market concentration of the insurance industry. As a hypothetical example, there could be a market of 6 firms, with shares of 30, 20, 15, 15, 12, and 8. Maybe that's not realistic, but let's just say it exists for the sake of argument. As a layman, I would probably put that state or metropolitan area high on my list of places to consider living, since I'd have six apparently competitive health-insurance choices.

But this market's HHI would be 1958, or Highly Concentrated. It doesn't seem to me that that's a good description of this imaginary market. Is it? I'm sure there are a lot of details about market competition and about what HHI specifically denotes to market analysts and anti-trust watchdogs that I am not aware of. I wonder what you think.


The Simple Answer (0.00 / 0)
is that this is what's been adopted by the government regulatory process, which has not been known for being especially strict over the past 30+ years.

Put that together with the fact that health insurance profits have been through the roof for the past half-dozen years, and it's really not hard to see that that level of concentration doesn't come close to being a true competitive marketplace.

"You know what they say -- those of us who fail history... doomed to repeat it in summer school." -- Buffy The Vampire Slayer, Season 6, Episode 3


[ Parent ]
National Health Insurance Exchange (4.00 / 1)
The extreme concentration shown in these charts is a clear and compelling reason why there's a need for a public option to ensure the existence of a genuine competitive market.

A strong public option would very much help, particularly in the areas with HHI above 5000, by adding one more insurer to each market.  It couldn't make any market "competitive", though.  And, of course, if the public option works as well as I hope, the index everywhere will climb back to 10,000!

However, for this specific issue it is more important that the Health Insurance Exchange be set up at a National level.  The larger the exchange, the greater the competition and the lower the HHI.  A true national Health Insurance Exchange open to all could actually make each region competitive (using the HHI definition) over time.


A Robust Public Option Wouldn't Be Like Another Private Enterprise (0.00 / 0)
That's precisely why they're trying to cripple it, if not do away with it entirely.

"You know what they say -- those of us who fail history... doomed to repeat it in summer school." -- Buffy The Vampire Slayer, Season 6, Episode 3

[ Parent ]
True (4.00 / 2)
But that would be a different point than "to ensure the existence of a genuine competitive market."

Heck, one might even claim "do away with it entirely" and "to ensure the existence of a genuine competitive market" where kind of opposites.  Not me, of course.  Being genuinely in favor of the public option and the Progressive Block, I've promised myself I'd act more like it here.

The real point, in my mind, is the same forces working against the public option are also working against a large health exchange.  We keep hearing about them being state based from various quarters, for example, which would do nothing to help these state HHI numbers.

By fighting for a strong public option, making that our line in the sand, we make the battle very simple and easy to understand.  While I disagree a strong public option is necessary for a good bill, I completely agree it will guarantee a good bill; a guarantee that cannot be easily evaluated without it.


[ Parent ]
Okay. But... (0.00 / 0)
While a regulated private insurance system could work in theory, the American business culture at this point is so antithetical as to make it impossible.  They would continually work to circumvent, subvert and undermine it.

That's why I don't ever expend any energy talking about it.  It's simply irrelevant to any realistic discussion.

"You know what they say -- those of us who fail history... doomed to repeat it in summer school." -- Buffy The Vampire Slayer, Season 6, Episode 3


[ Parent ]
Yes, but again.. (0.00 / 0)
A public option is very different from a competitive market. Describing it as "necessary for competition" is inaccurate at best and, I think, doesn't do a good job of selling the public option.

The important part is having at least one good option in every market. Competition is one way to (theoretically) achieve that, and a public option is another. Conflating them just confuses the matter and leaves proponents of the public option open to attack.


[ Parent ]
USER MENU

Open Left Campaigns

SEARCH

   

Advanced Search

QUICK HITS
STATE BLOGS
Powered by: SoapBlox