visit ontolawgy™ LLC’s main site


Take a swim in the risk pool: It’s not as risky as it sounds

Until I can actually see the latest health reform legislation without painstakingly redoing the work that our Senators and their staff have already done for us, on our dime—valiantly and tirelessly, I might add, but on our dime nonetheless—let’s take a look “risk pools.”

So what is a “risk pool” and why would you want to swim in it? I know, it sounds risky, but chances are you’ve got prune hands from being in one so long already, so don’t be too scared.

A “risk pool”, very generally, is a group of people paying premiums into an insurance plan and who are entitled to some benefits from the plan. Generally, the larger the pool, the larger the amount of total money available to an insurance plan to pay for health care for the people who pay premiums, and the less each person has to pay into it.

You may also see the term “high risk pool” being bandied about. It is a “risk pool”, but for people who are at a high risk to spend a lot of money on health care. (Some sources equate the terms “risk pool” and “high risk pool”, but I am not doing that here.)

To use a very loose analogy (and perhaps taking too many liberties with the word “pool”), a regular risk pool is like an Olympic-sized swimming pool located in southern Florida, where it doesn’t take much energy to keep it nice and comfortable (heating, filtering)—let’s call the energy costs “premiums”. So, everyone who swims in it chips in a little bit to keep the pool comfortable; that is, except for the uninsured, who hop the fence and do cannonballs into it, but we’ll get to them.

A “high risk pool” is like a hot tub: It is smaller, and it may take a similar amount of energy to run and heat as compared to the bigger pool, but it can hold fewer people, so fewer people can chip in. Thus, it’s more expensive per person.

Seems fair, right? You want a hotter pool with bubble jets and such, you have to pay more.
The conventional wisdom has been that high risk pools are fair to the general population because the high risk people pay higher premiums to help compensate for the amount of health care expenditures they incur. Sounds great if you’re in the regular risk pool; not so great if you’re in the high risk pool. The problem is that this (generally) is not equitable in the health care arena. Whereas some people might choose to pay more for hot water and bubble jets, no one chooses to be born with a congenital heart defect or Type I diabetes, and that’s where the “hot tub”/ “big pool” analogy breaks down—the big pool is just as good for people with heart defect and diabetes as for everyone else, so we’ll assume (for the purposes of this discussion) that everyone goes into the big pool.

I won’t get into the grey areas that include obesity or smoking (or other similar health hazards caused by genetics and free will in varying unmeasurable degrees) or other issues such as paying for health care for illegal aliens—they’re not really relevant at this fundamental a stage of the discussion.

Back to the pool. Think of a mandate to purchase health insurance like going to a big pool used by a bunch of different swim clubs. But imagine that everyone must swim at some point and that you need to join a swim club before you can jump in. Some clubs offer different services (fresh towels, swim lessons, sunscreen, babysitting, free use of kick boards), and some are pretty bare-bones, (e.g., they'll pull you out if you start drowning), but everyone pitches in to use the pool. Just as with health insurance, some plans offer annual physicals, nutritrional counseling, and discounted gym memberships, whereas others just pay for you to get your hand sewn back on after a horrific accident.
Some people might never plan to go into the pool, just like some people (e.g., Christian Scientists, devout Rastafarians) might not plan to use the health care system. But the risk pool system breaks down unless everyone belongs to a swim club/health insurance plan—someday they or someone they care about is almost certain to go for a swim, voluntarily or not. One reason health care insurance premiums are so high is that not everyone has insurance. This wouldn’t be a problem if the uninsured didn’t use the health care system, but in the U.S., the uninsured use the system all the time, and often don’t pay for the privilege, so the cost gets passed on to the people who have insurance. This is not the place to get into why or how much, but I’ll lay out some of the consequences below.

So what happens if you put everyone in the same pool? For the most part, peoples’ insurance premium costs go down. There are a few people who will pay more (typically, those who are currently uninsured or underinsured), but generally, because everyone is pitching in, the per-capita cost goes down. If you understand fractions, you’ll see why very quickly.

5/4 > 5/5 (or for the fractionally deficient: 1.25 > 1.0)

That is, if 5 people use health care, but only 4 pay for it, those who pay must pay 5/4ths of what their care actually cost to keep the system solvent. If 5 people use health care and all 5 pay for it, then everyone only needs to pay for 5/5ths. So, before you whine about being “forced” to buy “universal” coverage, ask yourself: “Would I rather pay $1.25 in insurance premiums or $1.00”? Assuming you’re a rational person, it would be the latter, right?

That said, at the risk of demonstrating why lawyers shouldn’t be allowed to play with numbers, below is an illustration of what would happen if the uninsured were able to get insurance and start paying premiums (whether on their own or through a subsidy scheme) in Texas, which, as of 2008, had the highest proportion of uninsured people of any State (including DC) in the country.

The latest year for which data are available for all States is from 2004. Looking at the overall numbers shows how (sadly) dated the data are: Total national health expenditures in 2004 were “only” about $1.55 trillion (that is, $1,551,261,000,000.00) and Medicare expenditures were “only” about $303 billion (that is, $303,462,150,000.00). As of 2007, they were each roughly 40% higher.

The health expenditure data are mainly from the Kaiser Family Foundation’s excellent, crystal-clear, and many of the data come from the somewhat more daunting Centers for Medicare and Medicaid Services statistics page. I pulled the 2004 Medicare enrollment data from there directly.

Major assumptions: 1) The uninsured, once they obtain insurance, will use care at the same rate they do now (an arguable proposition, given that they may be using less acute care and more preventative care); 2) The inference of the number of uninsured people was taken from percentages in a 2007-2008 survey—the assumption is that the percentage of the population that is uninsured has not changed between 2004 and 2008. (Although this assumption likely does not hold true for Massachusetts because of its recently implemented universal coverage system); 3) a “medical loss ratio” of 85% (taken from § 102 of H.R. 3962.PCS), that is, 85% of premiums collected are used to pay for health care; 4) the "premiums" below may be a bit off, because for the sake of simplicity, deductibles and coinsurance/copayments were not considered, so this assumes that insurance covers 100% of all medical expenses, but for a person who satisfies their deductible and pays 15-20% coinsurance, the out-of-pocket expense would likely be similar; 5) even if people are paying money into different pots, if everyone is insured, it acts effectively as a unified risk pool on a per-capita basis—the currently uninsured don’t choose which insurance company to bill their expenses to, so the assumption is that the costs are currently evenly spread across all the insured; and 6) smaller risk pools are generally more expensive per capita, but without data on risk pool sizes (e.g., number of insurance companies in each State, and number of people each company insures and sizes of their respective risk pools), it’s not possible to work out the advantage of a Statewide pool vs. the current situation for those in high-risk pools, rather only 100% coverage vs. current overall coverage rates, thus, the 100% coverage rate insurance premium projection is based upon a single Statewide risk pool because no other data are available (to me, that is).

Note that the illustration is extremely simple in that it does not take into account possible benefits of 100% coverage such as improved economies of scale or reduction in acute care utilization, nor does it account for self-insurance (personal or organizational), possible increased overutilization, medical expenditure inflation, the potential overlap (double-counting) of non-elderly Medicare beneficiaries, or Medicaid (which is likely to be supplanted in large part by premium and cost-sharing subsidies in what I’ve been able to glean from the leading health care legislation). At a fundamental level, the illustration below roughly shows that 5/4 > 5/5, but it looks somewhat more impressive when you plug in the real numbers. That said, let’s proceed.

Using currently available data and the above assumptions (with the above caveats), here’s a comparison between Texas now, and what Texas would look like with everyone (everyone who’s not in Medicare, that is) in the same risk pool:

Now (estimated)
Per Capita Health Care Expenditures (“PCE”), 2004, including Medicare
Total health care expenditures (“HCE”) (millions, 2004)
Inferred population (millions, 2004, inferred from HCE / PCE)
Medicare enrollees (2004, actual numbers)
Medicare PCE (2004)
Total Medicare spending (millions, 2004, computed from Medicare PCE * # of Medicare enrollees)
Total HCE, less Medicare (millions, 2004)
Non-elderly population (millions, 2004, inferred from inferred population less Medicare enrollment)
PCE not including Medicare expenditures or enrollees
Percentage of non-elderly population uninsured (2007-08)
Estimated number of non-elderly uninsured (millions in 2004 using 2007-08 uninsured percentage from StateHealthFacts and inferred non-elderly population)
Insured (millions in 2004 using 2007-08 uninsured percentage)
Rough estimate of total cost of care for uninsured (millions, 2004)
PCE per insured person to cover care for uninsured
Total PCE for each insured person to cover uninsured (overall PCE + PCE per insured to pay for uninsured)
Estimated per-capita premium for each current non-elderly insured person @ 85% loss ratio to cover uninsured health care costs
Estimated total premiums paid @ 85% loss ratio (millions)
Estimated per capita uninsured care cost ratio (a.k.a., for every dollar an insured person spends on his or her own health care, this much is spent by an uninsured person or on such person’s behalf)
With 100% Coverage (everyone swims)

PCE assuming 100% coverage (would be identical to current PCE)
Estimated per-capita premium @ 85% loss ratio and 100% coverage
Estimated total premiums paid @ 85% loss ratio and 100% coverage (millions, 2004, computed as per-capital premium * total # of all non-elderly)
Estimated per-capita premium savings for current insured
Estimated per-capita premium increase for uninsured

As you can see, the total amount of insurance premiums would not change, but the distribution of the expense would.

“But isn’t this unfair to those who can’t afford insurance”? Well, if insurance becomes roughly 28% cheaper, it probably will be somewhat more fair. Furthermore, the most recent version of meaningfully published legislation would offer some help to those people who truly can’t afford it. See §§ 341–347 of H.R. 3962.PCS.

The spreadsheet used for the calculations above is available upon request; I may post it or a link to it in the near future. It shows considerably more data, including some interesting projections on what would happen with a nationwide risk pool, but which I’ll summarize below:

1) Using the assumptions above, insured people every State (plus DC) would be better off with full coverage and a Statewide unified risk pool as compared to the current situation.

2) With a nationwide unified risk pool, people living in 22 States would likely be worse off financially in the short run as compared to a Statewide risk pool—28 plus DC would be better off.

3) Only the insured folks living in Utah and Arizona would be better off in their current situation as compared to a nationwide risk pool with full coverage, meaning that for 48 States and DC, a unified nationwide risk pool with full coverage would likely result in lower premiums than people pay now. 

© 2009 Alex M. Hendler. All Rights Reserved.


What Congress could learn from farting phones

With the Senate inching ever closer to passing health care reform legislation, unless you’re working on the Hill or have a few dozen hours to dig through THOMAS to piece together what the most recent version of the health reform legislation actually does, you’re out of luck if you actually have the gall to want to read it.

There’s a November 19 version of the available from the previous link (H.R. 3590.AS), but you need to look at the “Bill summary and status file” (specifically, “All Congressional Actions with Amendments”) to see what has happened to the multiple amendments that Thomas reports have been made to the bill. Then you’ll have to manually insert the adopted amendments into the Nov. 19 text yourself and see what you can figure out has been done. There’s no freely available unified text as of this writing. Some commercial services may have independently produced unified text, but I don’t subscribe to any of those—nor should I have to to find out what my Congress is doing.

Given that Congress actually uses computers now (at least, that’s the rumor), it is thoroughly shameful that we, for whom the Representatives and Senators work, cannot actually see what they are doing until well after they’ve done it—a major problem in a Republic where the Representatives and Senators are supposed to represent our interests. It is also highly doubtful that the Representatives or Senators can see what they are doing if this is the system they are actually using to keep track of things. Fortunately, there is software—much of it free—that could help.

It is no coincidence that we speak of computer “code” just as we speak of legal “code”. Each is a set of relatively arcane instructions to a computer (silicon or human, respectively) to perform or not perform a certain set of actions within a defined framework. Each is typically comprised of complexly interconnected elements, yet remains readable and (somewhat) intelligible to experts in their respective fields.

Typically, when one speaks of human-readable computer code, one is describing “source code”, that is, a human-readable equivalent of what a programmer intends for a computer to do. Once the code is “compiled” for a particular type of microchip, a computer can run it. For example, multiple developers can work on a program to make an iPhone simulate flatulence. The program is split into different sections, and the developers can revise these sections, tracking each revision along the way, debating changes, and, in some cases, letting anyone in the world track progress on the developing source code through a public website. Once the code is complete, it is compiled, and can be used by an iPhone to everyone’s juvenile delight.

Similarly, legislation is “source code” that must be “compiled” (either into the Statutes at Large or in the U.S. Code, or both) after its passage to be applied (“run”) by the legal system (judges, lawyers, police, etc.). However, it is put together in person (for the most part), on paper (in large part), then transcribed into a computer system that (at least to the general public) seemingly simulates paper in many ways. Changes are rarely integrated in real time, instead, they are often posted as monolithic “amendments” (if the amendments are posted at all) and not integrated into a “live” version of the text. Typically, with complex legislation, its full text is not available until after it has been passed by the House and the Senate and signed by the President—too late for mere mortal citizens to do anything about it (e.g., talk to their Representatives or Senators).

The analogy is not as useful when taking it much further, so let’s stop there for the moment and get to the point:

The same free software that people can use to collaboratively make their phones fart could be used to draft complex legislation and share it—in real time—with the general public.

There is a trove of free, (relatively) easy to use software that is designed to track the minutest of changes to extraordinarily complex documents and share those changes—including all the history back to the original text—with the entire world. Git, CVS, subversion, Mercurial, take your pick.

Although THOMAS is an incredible resource, it still largely catalogues information as if it were written on vellum parchment with quills, bound together with catgut, horsehide glue, and a calfskin cover. It does give one some appreciation for the complexity of what Congress does, but as as legislation and the legislative process grow increasingly complex, THOMAS’s utility as a tool to extract meaningful information decreases dramatically. Hopefully THOMAS will git (or some equivalent) with the program soon.

© 2009 Alex M. Hendler. All Rights Reserved.


Time to bring out the big stick: Who’s afraid of McCarran-Ferguson?

OK, so what is “McCarran-Ferguson” and what does it have to do with health care reform or the law?

The McCarran-Ferguson Act*, codified in part at 15 U.S.C. § 1012, allows the U.S. Congress to supersede State regulation of insurance.

15 U.S.C. § 1012(a) specifically delegates to States the regulation of the “business of insurance”; 15 U.S.C. § 1012(b) allows Congress to supersede State regulation of insurance, so long as the Act of Congress “specifically relates to the business of insurance”. But doesn’t Congress already regulate health care insurance? To some extent, yes, but there is still a lot of room for improvement.

For the past six decades, Congress has generally spoken softly with McCarran-Ferguson, typically exercising its power under § 1012(a), i.e., to do nothing. Congress has swung the big stick of § 1012(b) for Medicare, Medicaid, and HIPAA, among other issues, but generally, Congress holds the big stick still and leaves insurance regulation to the States—to the seemingly ambivalent delight of health insurance companies (i.e., good for the bottom line, not so good for consumers or their health).

In a recent editorial, Michael Tanner of the Cato Institute deftly laid out several issues that he sees as limiting the potential effectiveness of current health reform efforts in Congress. Among other issues, Tanner noted that current regulatory systems “limit competition between insurers and providers by, for instance, prohibiting people from buying insurance across state lines”, which prohibition, he implied, could be contributing to increased health care insurance costs.

Kudos to Tanner for so publicly airing this ugly open secret. Tanner neglected to note, however, that the offending regulatory systems, are, by and large, State regulatory systems. His editorial implies that Congress may have the power to do something about this problem and it suggests the admirable goal of a “consumer-oriented free market” health insurance scheme. Tanner did not—understandably, for a variety of unstated reasons—propose a specific course of action for Congress to achieve this goal. At my peril, I’m going to get specific and try to do so. Back to McCarran-Ferguson.

A responsible and effective swing of § 1012(b)’s big stick would accomplish at least the following: 1) Prohibit States from regulating certain aspects of health insurance (e.g., pricing and interstate sales); and 2) Forbid health insurance companies from offering separate “group” and “individual” policies. In a future post, I’ll dive into why these two provisions are essential to meaningful reform (and perhaps how they could be implemented without causing people to storm Capitol Hill with pitchforks and torches).

UPDATE - 2009-10-15, 20:02: This article suggests that the McCarran-Ferguson Act can exempt insurance companies from federal antitrust regulation and refers to some movement to repeal the Act in its entirety. I do not know much the repeal movement, but what the Act does is slightly more complex than suggested in the article. Based upon my reading of the Act, Congress does not need to repeal the Act to make health insurers fully subject to antitrust law, although it must take out of States’ hands any competition-related aspects of health insurance industry regulation. Specifically, 15 USC § 1012(b) provides that federal antitrust laws do apply to insurance companies (since 1948 in any case), but only “to the extent that such business is not regulated by State Law”. Thus, I have amended point 1 above from “Require States to allow insurance to be sold across State lines” to read as it does now.

At the risk of horrendously mixing metaphors, even a liberty-focused organization such as Cato could rationalize this course of action: this use of § 1012(b) would be chemotherapy for the anti-competitive cancer of a fragmented health insurance market; CABG for the clogged arteries of health insurance that are strangling the heart of our economy; a machete cutting through the brambles that impede our progress as a 21st -century nation; Congress courageously doing what needs to be done, rather than what is politically expedient; or, in the alternative: Justice through economic rationality.

Comments are welcome below. In the next installment we’ll take a swim in the “risk pool”. (In the meantime, you may want review this rather lengthy post from another blogger—“Jay” does an excellent job of laying out the framework of many of the underlying issues).

*I would read this article for a history of the Act, not necessarily an accurate analysis of the Act or the court cases giving rise to it.

© 2009 Alex M. Hendler. All Rights Reserved.


Senate takes a major step in a long, hard slog on health reform

At this stage, I was hoping to be able to have analyzed some health reform legislation (namely, the “American’s Healthy Future Act of 2009”), but given the dozens (or hundreds, according to another source*) of amendments on which the Committee completed voting only hours ago (and me not having a direct link to the Committee), I haven’t gotten a chance to see the whole thing yet.

Furthermore, at this stage, a bill may not even be publicly available in legislative language anytime soon. The most recent information posted on the Committee’s website is a 223-page outline explaining the current law and what the Committee wanted the bill to do as of September 22.

The next step is getting the Office of the Legislative Counsel to draft a bill. (They may have already done so, but I cannot find any publicly available legislative language). If the House’s bill is any guide (the version of that bill as amended more than two months ago is still not available, but you can try cobble together your own version from the House Committee on Energy and Commerce’s health reform page), we may be looking at something that can be measured in kilopages and that no single person could fully understand.

Now, back to the Office of the Legislative Counsel to the U.S. Senate. This Office has an extremely daunting task before it. The Office regularly excels at quickly converting legislative intent into legislation that, if enacted, would modify the law in accordance with that intent. However, there is some risk that with a bill of this size and complexity, the outline may not correspond to legislators’ intent; alternatively, if legislation has actually been drafted, there is no guarantee that the Committee’s summary accurately reflects what the legislation actually does.

To be sure, legislation always includes some intentional “wiggle room” for a variety of reasons—e.g., political expediency, delegation to regulatory authorities, or (for the cynical among us) keeping judges and lawyers employed—but as the size and complexity of legislation increases, so does the likelihood of unintentional wiggle room, i.e., gaps, oversights, or human error. I have met with Senate staff to discuss complex legislation only to find them shocked at what the legislative language would actually do as compared to what they intended it to do. To be fair, it was an amendment to the notoriously inscrutable Internal Revenue Code, but look at what they’re up against this time:

As metioned in a previous post, the Finance Committee’s bill outline proposes to amend not only the Internal Revenue Code, but also to amend (or incorporate) the equally-inscrutable (albeit somewhat less famously so) following laws/systems:

  • CHIP
  • Medicare
  • Medicaid

Also remember that statutes are just the beginning of the story: This legislation would require the enactment or amendment of multiple regulations that are based upon or authorized by those systems, not to mention various official policies, bulletins, newsletters, advisories, and necessary modifications to State laws.

How do you plan to keep track of it all? Please comment below. I’ll explain my strategy in a future installment.

*One would think that the Washington Post has a copy editor on hand to review things before publishing them, but apparently not for that article (at least the version I read), e.g., “the fifth and final to contribute to the legislation” and “the committee’s work of churning through 564 amendment”. I understand that it was published at 2:36 a.m., but reporters and editors are paid to work through the night on a regular basis. What major crisis could have prevented a front-page article from being proofread and published at 2:37 (or even 2:38) instead of 2:36?

© 2009 Alex M. Hendler. All Rights Reserved.


An Introduction

My name is Alex M. Hendler. I am an attorney from the U.S. I own and run a small company called ontolawgy™ LLC. My company's goal is to help make the law more accessible to attorneys and non-attorneys alike, using an internet-based computer system called the ontolawgy™ platform.

The ontolawgy™ platform is a semantic legal analysis system. What is semantic legal analysis? In a nutshell, it is what an attorney does in his or her head to analyze a legal issue; the ontolawgy™platform simply allows the coherent electronic storage and retrieval of the key elements of that analytical process and dynamic, customizable presentation of the results. So, why the reference to "connect the dots™"? Semantic legal analysis helps make legal research almost as straightforward as drawing a connect-the-dots picture on a placemat in your neighborhood family dining establishment or in one of your childhood coloring books.

As for this site, I plan to post bi-weekly (i.e., once every two weeks) thoughts on semantic legal analysis, health care reform, and the relationship between the two.

To set the stage a bit, in the U.S., there is increasing movement towards revamping about 16% (and growing) of our country’s economy through what is commonly called “health care reform”. A comprehensive approach to health care reform will likely include changes to federal tax, employee benefits, and Medicare and Medicaid law (four of the more complex legal regimes in the U.S.), as well as the possible creation of a new legal regime to govern a “public option”, “regional co-ops”, “health insurance exchanges”, or some other variant designed to increase access to health care insurance in the U.S., and possible reforms to the Social Security system and the various laws governing health care insurance plans available to civilian, military, and law enforcement federal employees.

Because of its necessary legal complexity (at least in the U.S.), health care reform provides a unique opportunity to explore semantic legal analysis, which I will do in part here, and in part on a deployment of the ontolawgy™ platform that will address health care reform once some solution is enacted into law.

For more information about me or the ontolawgy™ platform, please visit or come back in a couple of weeks for the first substantive installment.

© 2009 Alex M. Hendler. All Rights Reserved.