Complex Systems, Part I

October 20, 2011

Visitors to John Goodman’s Health Policy Blog quickly realize that it’s different. For one thing, we have a sense of humor. For another, we understand that socialism doesn’t work. But the single biggest difference between our point of view and others in health policy — both on the left and the right — is this: Everyone who posts at our blog recognizes that health care is a complex system.

Visitors to John Goodman’s Health Policy Blog quickly realize that it’s different. For one thing, we have a sense of humor. For another, we understand that socialism doesn’t work. But the single biggest difference between our point of view and others in health policy — both on the left and the right — is this: Everyone who posts at our blog recognizes that health care is a complex system.

The single greatest health policy failing of the Obama administration is that it does not recognize this central fact. Neither Barack Obama nor any of his health policy advisors seem to understand what a complex system is, let alone how to develop public policies to successfully deal with one.

In this Alert, I will identify some of the properties of complex systems. In the following Alert, I will spell out some of the public policy implications.


I Can’t Tell You Why


Health Care as a Complex System. Our health care system is an example of what social scientists call “complex systems.” These systems are so complicated that no one person can ever fully grasp everything that is going on. As individuals all we ever really see is a small slice of the system. That’s usually the part of it that we interact with.

The economy as a whole is an example of a complex system. Of the subparts of our economy, health care is by far the most complicated. It is many times more complex than a normal market, for example. The reason: in addition to garden variety economic forces, the medical marketplace is institutionalized, bureaucratized and enormously regulated. Doctors are heavily influenced by medical ethics and traditional ways of doing things. Almost everything they do is impinged upon by third-party payer bureaucracies. And almost everything they do is influenced by myriad regulations that are inconsistent, voluminous and complex, to say nothing of the ever present threat of tort law litigation.

In addition to all that, we have completely suppressed normal market processes in health care — in this country and all over the developed world. As a result, rarely does anyone in health care ever see a real price for anything. Employees never see a premium reflecting the real cost of their insurance. Patients never see a real price for their medical care. Doctors and hospitals are rarely paid real prices for the services they render. Instead, they are paid on the basis of “reimbursement formulas” and there can be a different formula for each payer. Enormous amounts of money change hands every day in the medical marketplace, but most of the conventional rules of economics do not directly apply.

Complex systems can never be accurately modeled. Since the days of Adam Smith, economists have been trying to understand some general features of markets by creating simplified models. These models are necessarily very crude attempts to replicate some basic forces in the system as a whole. No model tries to capture everything that is going on in the economy, however. That would be impossible.

Nonetheless, economists have been able to use highly simplified models to predict some general effects of parameter changes in ordinary markets. For example, we can say with some certainty that a substantial increase in the minimum wage will cause unemployment. Rent controls will cause housing shortages. Price supports in agriculture will cause crop surpluses.

There is no reliable model of the health care sector. Unfortunately, there is no model of the health care system that allows us to make anything like the predictions we can make in other markets.

In just a few years, ObamaCare will insure an additional 32 million people. In addition, most of the rest of us will have to convert to health plans that have more generous coverage than we now have. We know from past studies that when people become insured, they tend to consume twice as much health care. We know that when people have more insurance coverage they consume more care. But what happens when there is a system-wide increase in demand and no change in supply?

Will the excess demand drive thousands of people to hospital emergency rooms? Will clinics run by nurses start springing up everywhere to meet the demand that doctors cannot meet? In the face of a large increase in rationing by waiting, will everyone who can afford them turn to concierge doctors? As more doctors become concierge doctors, how will the system manage the even greater rationing problem faced by all those left behind? Will patients start going out of the country — seeking care in the international medical marketplace?

Unfortunately, there is no model that allows us to answer these questions with any confidence.

Why can’t we apply ordinary economic models to health care markets? One reason is that price doesn’t play the same role in health care as it does elsewhere in the economy. Although many would like to think that our system is very different from the national health insurance schemes of other countries, the truth is that Americans mainly pay for care the same way people all over the developed world pay for care. We pay with time, not money.

On the average, every time we spend a dollar at a physician’s office, only 10 cents comes out of our own pockets. As a result, the time price of care (the time it takes to get to and from the doctor’s office, waiting in the reception area, waiting in the examining room, etc.) tends to be greater — and probably much greater — than the money price of care, for most people.

In addition, there is increasing evidence that non-price obstacles to care are greater deterrents than price barriers.

In general, we have no reliable model to tell us who gets care and who doesn’t when the time price of care rises for everyone, as we expect to happen once ObamaCare gets fully phased in. Nor do we have any model that allows us to predict how care will be allocated when non-price deterrents are the principle form of rationing.

Complex systems have unintended consequences. An interesting characteristic of complex systems is that when you perturb them (by passing a law, for example) there are always unintended consequences. The less you know about the system the more unpredictable these consequences can be. In economic history, for example, there are numerous examples of governments that adopted policies in an attempt to improve things, but ended up making the situation worse. Consider two examples from health care:

In 1965, Congress passed Medicare in an attempt to increase access to health care for the elderly and improve their health status. They believed they could do so without any material impact on the rest of the health care system. Yet MIT professor Amy Finkelstein has discovered that Medicare had no effect on the health of the elderly; but the additional spending did set off a bout of health care inflation for all patients which never subsided.  

In 2003, Congress passed a Medicare drug benefit, largely out of concern that senior citizens couldn’t afford the coverage themselves. Since the new program (Medicare Part D) had no funding source, it created a $15.6 trillion unfunded liability for the federal government, looking indefinitely into the future — more than the unfunded liability in Social Security! Yet Andrew Rettenmaier discovered that almost all of the spending, about 93 percent, simply replaced spending the elderly were already doing. Only one in every 13 dollars represented new drug purchases. Interestingly, the help given to the small number of beneficiaries who needed it actually reduced Medicare’s overall spending, as drugs were substituted for more expensive doctor and hospital therapies. But this “profit” on the truly needy was overwhelmed by the cost of giving the benefit to those who didn’t need it and in the process creating an enormous obligation for current and future taxpayers.

Implications of unintended consequences. Why are unintended consequences so important? Because in trying to solve one problem we can create other problems. Also in trying to solve problems, we can end up making them worse. ObamaCare has three principal goals: control costs, raise quality and increase access to care. Yet there is no model which allows us to predict that any of the three objectives will be even partially achieved. In fact, readers of this blog know that we expect all three problems to get worse.

That last sentence was not a misprint. We can actually spend a trillion dollars, create 159 new regulatory agencies, force almost everyone into a  new health plan and — at the end of the day — end up with higher costs, lower quality and less access to care.