Healthcare economics are in a death spiral, not the one gracefully done on ice. Contrary to a lot of rhetoric, there are both free market as well as government-run alternatives. There is a working alternative covering some millions of people, and both Bush and Kerry, during the campaign, endorsed making that alternative, which is free market and multiple payor, to part of the population (e.g., small business). Broadening that alternative, which I'll identify in a bit, has gone nowhere.

What happened to the campaign promises about these trials? I didn't seem to hear any specific proposals in the State of the Union Address.

As some of you may know, the US employer-based insurance mechanism is a historical accident. It was introduced as a way that WWII employers could attract people at a time of controlled wages. Of course, at that time, medical care neither did a lot nor cost a lot.

Think a little about classical free market theory. Prices are set by interactions between the consumer and the provider. Here, I am ignoring elasticity of demand. See [1]. In the case of healthcare, the consumer is logically the person needing care and the person/organization providing care.

Introducing employer-paid care breaks the market model. Pricing now get sets between a buyer and a benefits manager. The buyer is the employer, to which healthcare cost is basically overhead [2], and the buyer wants to minimize cost. The seller is a benefits manager, usually an insurer [3]. The seller competes for business based on low cost, but, in most cases, does not directly provide the care.

Instead, the benefits manager contracts with actual healthcare providers to deliver services, normally at a negotiated discount. The bigger the benefits manager, the more leverage they had. For example, when I had my pacemaker installed, the hospital "list price" for the equipment, installation lab, etc., was $24,000. At the time, I was covered under the then 2nd largest benefits manager. Their discount was such that the total reimbursement to the hospital, including my copayment, was $1,800. In other words, $22,200 went into the ether.

Since the discounts are often negotiated on specific treatments, diseases, etc., the payments may or may not cover actual cost. The hospital still has to pay its expenses, so, if they lose money on some procedures for some provider, they have to get the money somewhere.

How Today's Con Game Works

Enter the shell game of today's healthcare: cost shifting. The hospital or other provider takes its loss, and spreads it over the benefits managers that don't have such good discount, and socks it at list price to people paying out of pocket. Further, there are other cases of cost shifting. Medicare/Medicaid negotiates the deepest discounts, and plays certain games that shifts even more risk to the provider [4].

There is a well-intended law known as EMTALA, which essentially requires emergency rooms to evaluate and stabilize patients without checking for the ability to pay. Unfortunately, EMTALA is an unfunded mandate. We have very real situations where an uninsured, unemployed individual gets hit, intentionally or not, in a drive-by shooting by a drug gang. Lifesaving care can easily hit the hundreds of thousands. Sometimes, the hospitals can get the patient enrolled in Medicaid, but in other cases, they eat the cost.

While lots of providers boast of giving "free care", free care is mostly free only because costs are shifted. Rare providers have endowments to subsidize costs, or municipal/county facilities may get tax support.

Another zap is that the employer-based system disincentivizes preventive and maintenance care. Employers constantly shop for cheaper coverage, and frequently change benefits managers, a phenomenon called "churn". Even though preventive care may lower long-term costs, the benefits managers know that if they pay for the flu shot today, the employee might be under a different plan at the time they would have gotten sick, so the savings go to a different benefits manager.

Benefits managers that are also insurers try to minimize their risk, by methods such as "cherry picking" the population least likely to get sick. Since every employer contract is different, the insurers do "experience-based underwriting" that looks at the often small sample in a company, rather than the incidence of a disease in the population as a whole. One small business offered to hire me, rather than have me as a consultant, to get me insurance. Their insurer took one look at my health history, and told them their premium would immediately double, and they would not be renewed, if they hired me, even with preexisting condition waiting periods. Large businesses tend to be statistically even enough that their employee experience tends to be like the general population -- and sufficiently large businesses go to ERISA.

So what is to be done?

There are two basic ways to go. One is single-payor government-operated care. The Canadian and British systems look like this at first glance, but there are lots of complexities under the hood. For example, while the individual specialty boards in the US set the number of spaces in residency programs, sometimes causing oversupplies or undersupplies of specialists, Canada sets those spaces globally. Canada also emphasizes primary care over specialty training. Other nations do subsidize medical education, so new physicians are not under the usual crushing student loan debt contracted by most US students.

The other is mostly being called a consumer-driven health plan (CDHP). Today's best example is the Federal Employees Health Plan (FEHP). In the latter, negotiators (now government employees, but they don't need to be) get bids from benefits providers. In general, the bids must be based on population-wide underwriting, not experience-based, and there are enough Federal employees that this works. The negotiators establish a number of options of different plans with different benefits and out-of-pocket costs, so it isn't a single-payer system. With lower costs because there is a large population over which expenses are spread, the benefits managers want to keep their clientele, so they tend to be much more reasonable about preventive care and pre-existing disease management.

Bush and Kerry both made campaign statements about it being reasonable for certain small businesses, and possibly individuals, to buy into FEHP. No action.

The Clinton plan was somewhat like this, but with important differences. Under that plan, it was illegal to get private care out-of-pocket, which many healthcare economists consider a good quality control check. Hillary also so mishandled the introduction that it was dead on arrival in Congress, having offended just about every organization with influence on the process.

To achieve universal coverage, which does have measurable economic benefits to the population as a whole [5], what appears reasonable is a CDHP, to which, with very limited exceptions [6], everyone participates at the minimal-care level, with the option of paying more for better care. That minimum level might be out-of-pocket, or tax revenue -- and don't think that cost shifting doesn't cause higher taxes anyway.

For the unemployable, as in Medicaid, you don't need the Federal management, but instead provide CDHP vouchers for the agreed level of care. Social security, TRICARE, etc., all can work the same way -- a much simpler Federal involvement dealing with vouchers rather than individual case handling. Obviously, there needs to be auditing of the negotiators and the benefits providers, probably catastrophic event reinsurance, and there may need to be a mandate or tax revenue contribution per head.

There would be a lot of details to work out, but expanding the FEHP would be an excellent trial with a lot of mechanics already in place.

1. Elasticity of demand is an economic term that, in the most extreme case, not irrelevant here, says that the buyer will pay any price for something of infinite value. In this case, it literally may be that the price is for continued life. The theory breaks down, however, if the provider sets the price so high that no buyer has the money. Again, lots of complexity.
2. Some employers actually consider the health of their workforce as an investment, but that isn't usually part of the calculation.
3. Under a law known as ERISA, most large companies do not truly buy insurance, but pay the medical costs themselves. This can be transparent to the employee, as employers commonly hire insurance companies, with all the claims processing machinery established, to administer the financial aspects of payment. Using ERISA also exempts employers from some regulation that applies to actual insurers.
4. The basic model of Medicare reimbursement is not the sum of actual goods and services during a hospitalization, but something called a Diagnosis Related Group (DRG). Congestive heart failure, for example, is one DRG. Hospitals get a fixed reimbursement for each hospitalization for a person in that DRG. There's all sorts of gaming to minimize cost for DRGs, where the reimbursement doesn't cover the costs for the most severe cases. Again, I'm simplifying.
5. The argument that "I'm young, healthy, and don't need medical care" doesn't really work out statistically, because the young still get into major traumas (very expensive), need preventive care, and WILL need care as they get older.

   Further, not treating vulnerable populations will incur later costs, from birth defects to the spread of such things as multidrug resistant tuberculosis. Yes, many people in these populations don't go to the free clinics and such, but the reasons aren't always irresponsibility. For example, free clinics often have limited hours and may not be accessible to people dependent on public transportation.

   The underclass, then, goes and uses ERs, which have to see them under EMTALA, because those are available 24/7 and won't turn them away. It is false economy, however, to have situations where diabetics can't afford insulin, so they only show up at the ER in coma.

6. For example, Christian Scientists have a First Amendment reason not to participate. Otherwise, it's much like the requirement for driver liability. If you think about it, the basic coverage, whether that comes out of tax revenue or not, is at least guaranteeing that EMTALA emergency care is covered, just as fire and police are covered out of tax revenue -- and you can still buy sprinklers and hire guards.

It's been said that he who would defend against everything defends against nothing. I'm constantly reminded of that when I see the Administration spending a great deal of money on protection against a relatively low threat, but with a defense that sounds dramatic: building a shield over the nation. The question here: is this a better use of a comparable amount of dollars than hardening the national electrical power grid against blackouts?

This is the first of several threads dealing with tradeoffs in investment in protecting national critical infrastructure, or protecting the nation against threats of legitimate concern, such as bioterrorism. There are no black and white answers to the tradeoffs involved, but there are questions if the Administration is prioritizing correctly.

I refer here to national ballistic missile defense (NBMD), and whether dollars spent here are better investments than patching vulnerabilities in national critical infrastructure. Let me note that I consider investment in, and deployment of, theater BMD (TBMD) quite wise. TBMD can be stabilizing, if used, for example, against North Korean missiles targeted on Japan, Iranian or Israeli missiles around the Middle East, etc.

If budgets were unlimited, I might be more enthusiastic about NBMD, but even if they were, there are arguments that it may be destabilizing with respect to the world nuclear balance. Nevertheless, the strategic balance is less my concern here than is budget. In this thread, I'll examine one critical infrastructure subsystem, electrical power. I'll introduce new threads on other subsystems.

Part of the vulnerability of the electrical grid is a result of unintended consequences of deregulation. In some respects, there are good technical effects of an economic environment that may make it cheaper for utilities to buy power, across an interconnection grid, than to generate it themselves. Buying it from a utility that generates using clean sources such as hydroelectric power certainly has advantages over emissions from coal burning, or fuel supply for petroleum-based plants.

One of the unintended consequences, however, is putting a major energy transfer load on older grids intended for more occasional use. A bit of technical architecture here: inside an electrical utility, it is absolutely sensible to use conventional alternating current (AC). AC allows power to be stepped down from high-power lines to the needs of the end user. AC signals, however, have some critical timing parameters. When interconnecting grids with different AC generators, a slight mismatch of these parameters can cause damaging surges in either or both sides of the interconnection. We now have the technology to interconnect with direct current (DC), and this is indeed the technology of new construction. DC is much simpler and safer to interconnect; the timing issues essentially are nonexistent.

If there are surges in AC interconnection, the problems are detected and isolated with the System Control And Data Acquisition (SCADA) network. SCADA networks are utterly critical resources and need maximum security.

A great many technical catastrophes have no single cause, but result from a series of failures. In the case of the Ohio Valley blackout of 2003, the first problems came from AC mismatches. Unfortunately, the SCADA network of one of the utiities had not been kept sufficiently secure, and was contaminated with a computer worm. That worm slowed SCADA long enough that surges propagated into multiple grids before they could be stopped.

Remember that we now have a deregulated environment, with shareholders wanting high stock price. When regulated utilities didn't receive value on market price, but on return on investment, they were incentivized to improve, and even over-improve, their physical plant.

There are industry estimates that fully modernizing the interconnect grids and securing the SCADA networks might cost $20-30 billion. We have, however, already seen regional blackouts caused by the lack of these upgrades, caused variously by pure accident or by the effects of generic malicious hacking, not aimed at the utilities. Consider the effect of a very few trained terrorists infiltrating power utilities and getting jobs as control operators -- and then initiating simultaneous blackouts in different parts of the country.

The current economic model for the deregulated utilities can't justify the upgrades. Is such an expenditure a valid use of Federal funds to secure part of the national infrastructure? The amount involved is roughly equivalent to the cost of NBMD deployment. Which is the better national investment?