Peter R. Orszag
Excerpted from Peter R. Orszag's keynote address from the October 2008 Kaufman Hall Financial Leadership Conference
My children used to refer to CBO as the Congressional Boring Office. They have now taken to referring to it as the Congressional Bailout Office.
We clearly are in for a period of potentially quite sustained and potentially quite significant economic difficulty that will manifest itself in a variety of ways, including in the budget deficit. In 2007, the budget deficit was $450 billion. In 2008, it likely will exceed $750 billion, and credible estimates suggest it could be in excess of $1 trillion. Many people have noted that in that context, it may be difficult to get other things done, whether healthcare reform or new initiatives. A major impediment is going to be simply time and energy and simply oxygen for other initiatives as we struggle with regulatory reform, a new stimulus bill, and a variety of other things.
I am very interested in health care and have been spending most of my time on it. But over the last two months, I would say under 10 percent of my time has been spent on health care as other activities have crowded out that focus. That's just one small reflection of the difficulties that will face a new economics team as they struggle throughout 2009 to contain what we are experiencing.
As we emerge from our current difficulties-and we will ultimately emerge from the current difficulties-it is crucial that we bend the path of healthcare spending, because if we don't and we roll forward 10, 15, or 20 years, we could find ourselves in the kind of situation in which we would lack the maneuvering room that we currently enjoy to address crises like the one that we currently face. We are lucky today that we can issue hundreds of billions-if not trillions-of dollars in additional treasury securities at relatively low interest rates to inject equity into financial institutions, purchase assets, or respond to our current difficulties however we want to. We will not always have that luxury. If we were to face the kind of crisis we face today without the ability to issue substantial amounts of additional treasury securities, however bad you think today's situation is, imagine how much worse it would be if it were to occur in 10, 15, or 20 years and we still hadn't dealt with the ongoing fiscal path that we are on.
View exhibit 1
Our fiscal path over the long term is determined primarily by Medicare and Medicaid. Most of the growth in our nation's long-term budget projections is occurring in Medicare and Medicaid. I want to plant that thought in your head, and then I'm going to take a little detour, because I don't want to spend the entire time with green eyeshades around me.
Lessons from Econ 101
The current crisis underscores an economics lesson: that we need to move beyond the Economics 101 approach to life in which we are all supercomputers that can costlessly process information and make decisions perfectly. That approach is insightful in some situations, but has often gone awry. In health care also, we need to be learning the same lesson because it's ignored to our detriment.
Let me give you a little example from economics. The Econ 101 approach to life would suggest that retirement saving depends on your income, the match rate your company offers, the tax rate for retirement saving, and what have you. It should not make any difference whether you check a box that you're in a 401(k) plan or not, because you're all hyper-rational people who can costlessly process this information, and the box checking shouldn't matter. In practice, we know it matters a lot. If employees have to affirmatively sign up for a plan, for example, the participation rate for low-income workers is about a quarter, but with automatic enrollment in the plan, the participation rate is about three-quarters. This shows that other forces-in this case inertia-matter a lot. The default setting matters a lot. There is nothing in the Econ 101 toolkit that can get you this kind of jump in participation rates. You can offer match rates of 200 percent, but you're not going to get low-income workers participating in this kind of way. Health care and medical science are ignoring these kinds of forces in human behavior.
My favorite example involves exercise. Take people who clean hotel rooms every day, and split them into two groups. Tell one group that vacuuming, changing the sheets, and all of that qualifies as physical exercise, and don't do anything with the other group. Then watch what happens. They don't clean the rooms any better or report any more physical exercise outside of work. It turns out that simply telling people what they do every day is physical exercise generates results that are similar to actual physical exercise. For example, their body mass index drops, their LDL cholesterol drops, and their systolic blood pressure drops. You have created the belief in their mind that they are exercising, and you have generated results that are similar to actual exercise.
In many settings, we need to be paying much more attention to social norms, defaults, and basically sociology and psychology and not assuming perfect rationality. To much too large a degree, we are not there yet in health care and medical science.
What's Driving the Rise in Healthcare Costs?
Let me now put my green eyeshade back on and return to the fiscal problems we face. I'll emphasize something that has not been emphasized enough: Health care contains the largest inefficiencies in our economy. The variation in the way health care is delivered in the United States across regions, across hospitals in a region, and across doctors in a hospital is not driven by evidence, is not driven by medical science. Too much of the health care delivered in the United States is of the ilk "That is how we do it here" or "This is the way I've always practiced." The evidence that we do have suggests that the more interventionist, higher approaches do not generate better health outcomes than the more efficient approaches.
Health care not only is the key to our fiscal future, but also contains a massive opportunity to improve efficiency, with credible estimates suggesting that as much as $700 billion per year, or 5 percent of GDP, of healthcare services delivered in the United States don't improve health outcomes. I can't come up with anything that is even remotely close to that in terms of other distortions or inefficiencies, whether the impact of high marginal tax rates, impediments to international trade, or what have you. So let's try to unpack that a little bit.
First, we need to be clear about what's driving rising costs. There are two different forces, Medicare and Medicaid, and most of the popular mass media gets that wrong. Costs are going up in Medicare and Medicaid. We will have more 65-year-olds and more 85-year-olds in the future, which will drive up costs. Because those 85-year-olds are going to cost more in the future than the 85-year-olds cost today, how much of the projected growth is coming from these two forces? Despite what the media says about the coming tsunami of the retirement of baby boomers, most of the growth comes not from demographics but rather from rising cost per beneficiary.
View exhibit 2
There are things we can do about cost-per-beneficiary growth. That brings me to the sufficiency point. The massive variation across parts of the United States cannot be explained based on the cost of building a hospital, the health status of the population, the salaries that doctors are paid, or the prices of inputs in different areas. When you throw all the explanatory variables that are available into trying to explain this, you can narrow the variation by perhaps a third. There remains a very substantial variation that we are not able to explain based on any of the factors that you would normally think could explain these sorts of differences in cost.
View exhibit 3
It's true that our nation's leading medical centers deliver the best health care in the world. But I don't think there is a single way of practicing medicine at our best medical centers. They seem to practice medicine in dramatically different ways, even for the same kinds of diagnoses. As an example, look at Medicare beneficiaries in the last six months of life at a few leading medical centers. At UCLA Medical Center and the Mayo Clinic, the available quality indicators suggest that the Mayo Clinic does better on quality. But look at the cost per beneficiary: UCLA Medical Center, $50,000 per year; Mayo Clinic, $26,000 per year. The best medical care in the world is costing twice as much at one institution as at another. And we have absolutely no idea what we are getting in exchange for that extra cost other than a lot of tests and procedures that don't seem to be improving health outcomes.
We can replicate this within a hospital. You see dramatic variations across doctors within a hospital. If you look at people reporting lower back pain, the rate of spinal MRI ordering for the same diagnosis and the same clinical notes varies dramatically, depending on who you see when you report the lower back pain. So depending on whether it's a bone radiologist or a nerve radiologist ordering the test for the same diagnosis, you have dramatically different ordering rates. Then within each of those groups, you have dramatic variation in the probability of ordering a spinal MRI. Again, we are practicing medicine in dramatically different ways, in ways that are not informed. How can you explain that based on objective medical evidence when it's the same diagnosis, even the same clinical notes basically, and the rates of spinal MRI ordering are dramatically different? At each level of our healthcare system, you see dramatic variations that are not backed by specific evidence that more interventionist approaches work better and that have dramatic differences in cost that are not correlated with health outcomes.
When you look at the correlation between cost and quality, it looks like the correlation goes the wrong way so that in general, the higher cost providers, the higher cost hospitals, the higher cost regions are not generating better health outcomes than the lower cost, more efficient providers.
Why is this happening, and how can we improve the situation? How can we either get more for the money that we're putting in or reduce the money that we're putting in for the same health outcomes? My theory is: Too much of the medical care delivered in the United States has not been examined in the clinical setting. How many times should you go back to see your doctor after surgery? Three or four times a month? Twice a month? We don't have the evidence that will then inform whether or not a given frequency of visits is worth it.
A Three-Step Approach to Better Care
We have a payment system that facilitates more interventionist approaches for those doctors, hospitals, and regions that for whatever reason have adopted a more interventionist approach, again not backed by specific evidence that it works better. We pay for it, thereby facilitating it, and we wind up with lots of stuff happening, such as spinal MRIs being ordered much more frequently in ways that drive up costs but don't improve health outcomes. If that theory is right, you'd expect to see a lot more variation where we are clueless about what should be done. And that's exactly what we do see. In other situations where it's clearer what should happen, we see less variation.
It's clear that we have a payment system, especially within Medicare, that leads to more care rather than better care. What do we do about it? I'll give you a possible three-step agenda. The first thing we need to do is dramatically expand the health IT backbone that will provide information so that we can examine the benefit of that fourth visit or the benefit of that spinal MRI. But by itself, that won't be sufficient. That's often held out as a panacea-that we just plop a health IT system down into a fragmented healthcare delivery system with distorted incentives and expect magic. Not going to happen.
We also have to ask how we can get toward universal health IT. I'll say something very unpopular: The current policy approach of trying to encourage health IT adoption by offering institutions small subsidies to adopt is unlikely to be effective, because it will induce only those institutions that are close to adopting anyway. If you offer a $5,000 or $10,000 tax credit or subsidy for a health IT system, an organization would put it in place only if it was at the margin anyway, and the extra $5,000 or $10,000 kicks it over. Or you can offer massive subsidies-say $200,000 for a health IT system. The problem with that is the budgetary cost becomes huge, and you're also buying out the base of a lot of institutions that would have adopted it anyway. If we were really serious about getting to universal health IT quickly, we might offer small subsidies for some period of time, as we did in the e-prescribing legislation recently. Then we would say after three or four years, you must have a health IT system that meets the following criteria, or you will no longer be reimbursed under Medicare and Medicaid. Guess what? We'd get there very fast.
The second step is to use the information that comes out of a health IT system and evaluate these questions like spinal MRIs. To do that we are going to need a much expanded comparative effectiveness effort, way beyond the small effort that currently exists within a small agency in the Department of Health and Human Services. Policy discussions are going on about a dramatically expanded comparative effectiveness effort, perhaps lodged in a new entity.
The final step involves payment reform, because unless we change the incentives for providers toward more efficient care, we are not going to get what we want. To put the point another way, as long as we have incentives for more care rather than better care, guess what we are going to get? More care.
To tie this three-pronged approach back to the comments about psychology and sociology, as we do all of that, we need to be paying a lot more attention not just to Econ 101 financial incentives, but also to the context, social norms, and inertia so that we can get as much bang for the buck from the healthcare system as possible. That will not occur if we maintain this simplistic Econ 101 approach to public policy that has been too dominant in many areas, including healthcare policy.
Peter R. Orszag, PhD, served as director of the Congressional Budget Office (CBO) from January 2007 until November 2008. He resigned that post upon being nominated by President-elect Barack Obama to be director of the Office of Management and Budget. Under Orszag's leadership, the CBO significantly expanded its focus on areas such as health care and climate change.
Before joining CBO, Orszag was the Joseph A. Pechman Senior Fellow and deputy director of economic studies at the Brookings Institution. There, he also served as director of The Hamilton Project; director of the Retirement Security Project; and codirector of the Tax Policy Center, a joint venture with the Urban Institute. In previous government service, Orszag served as special assistant to the president for economic policy and senior economic adviser at the National Economic Council during 1997 and 1998. Earlier, he served as a staff economist and then senior adviser and senior economist at the President's Council of Economic Advisers.
Orszag graduated summa cum laude in economics from Princeton University and obtained MSc and PhD degrees in economics from the London School of Economics, which he attended as a Marshall scholar. He has coauthored or coedited a number of books, including Protecting the Homeland 2006/7 (2006), Aging Gracefully: Ideas to Improve Retirement Security in America (2006), Saving Social Security: A Balanced Approach (2004), and American Economic Policy in the 1990s (2002). He is a member of the Institute of Medicine of the National Academies of Sciences.
Publication Date: Thursday, January 01, 2009