Why health care price controls always fail
House Republicans, you may have heard, are trying to "end Medicare as we know it." And well they should—Medicare as we know it is the nation's biggest fiscal disaster. For years members of Congress and the executive branch have been trying, and failing, to find ways to restrain the growth of government health spending on seniors. Medicare is a $500 billion program on track to become a $1 trillion program before hitting insolvency in 2024, even under the rosiest projections. The program looms as a threat not only to itself but to the budgetary health of the nation. It is the single largest driver of long-term federal debt.
Despite the potential campaign effectiveness of the political charge that Republicans want to gut Medicare, President Barack Obama has positioned himself as a willing butcher of his own party's sacred cow. "We have to tackle entitlements" to control the federal debt, the president said in June, and "Medicare has to bear a greater part of the burden." Over the summer, Obama signed a debt deal with Republicans that allowed for a 2 percent cut to Medicare spending should a bipartisan deficit committee fail to come up with savings. In September he endorsed $248 billion in Medicare cuts as part of his own debt reduction proposal.
The cuts Obama proposed were not part of a fundamental Medicare overhaul, but they were cuts all the same. "Despite what some in my own party have argued," he said, "I believe that we need to make some modest adjustments to programs like Medicare to ensure that they're still around for future generations." Obama claimed he was open to reforms that would bring down the cost of Medicare, "not by shifting those costs to seniors but rather by actually reducing those costs."
"Actually reducing" the cost of Medicare has long represented the biggest pot of gold at the end of the public policy rainbow. It is treasure that Obama has been promising to deliver since early in his presidency. "If we do nothing to slow these skyrocketing costs," he said in 2009, "we will eventually be spending more on Medicare and Medicaid than every other government program combined. Put simply, our health care problem is our deficit problem. Nothing else even comes close.…We know we must reform this system. The question is how."
So what innovative solution does Obama propose to begin fixing America's biggest fiscal problem? Simple: He would change the way providers are paid for Medicare's services. Pay less, spend less. Right? It is so obvious that one might wonder why it hasn't been tried before. The answer is that it has—many, many times.
It is often said that you can't put a price on health. But for decades that is exactly what the federal government has attempted. Since the birth of the entitlement, a parade of legislators and bureaucrats has been playing billion- and trillion-dollar games of Whac-A-Mole with Medicare, knocking down spending with an elaborately constructed set of technocratic payment schemes in one area only to see it rise back up in some other part of the system. Obama is merely proposing to try it one more time.
All-You-Can-Eat Health Care
When Medicare, the federally run health care financing system for Americans who are 65 or older, passed in 1965, supporters knew the program would be expensive. Its lack of cost controls was the price of passage. Wilbur Cohen, a top health bureaucrat dubbed "The Man Who Built Medicare" by Medical World News, admitted that "the sponsors of Medicare, including myself, had to concede in 1965 that there would be no real controls over hospitals and physicians. I was required to promise before the final vote in the executive session of the House Ways and Means Committee that the federal agency would exercise no control."
Indeed, that promise was explicitly built into the legislation, which declared that "nothing in this title shall be construed to authorize any Federal officer or employee to exercise any supervision or control over the practice of medicine or the manner in which medical services are provided…or to exercise any supervision or control over the administration or operation of any such [health-care] institution, agency, or person." In other words, no rationing, no death panels. As Richmond University political scientist Rick Mayes explained in a 2007 essay for the Journal of the History of Medicine, Medicare was inaugurated with "a reimbursement system that neither imposed limits nor required outside approval." As a result, "unrestricted cost reimbursement became the modus operandi for financing American medical care."
Even then, the program's supporters grossly underestimated how expensive the program would be. The House Ways and Means Committee projected that 95 percent of the elderly would enroll in the program's doctor insurance component during 1967, its first year of operation. That estimate proved accurate. But the committee also projected that total costs for the first year would run no more than $1.3 billion. Total spending in the first year instead ran a whopping $4.6 billion.
As the program continued, its true costs rapidly departed even further from initial expectations. The committee had projected that hospital spending would amount to just $3.1 billion in 1970. Instead it was $7.1 billion. Hospital spending in 1975, initially expected to be around $4.2 billion, was actually $15.6 billion. The estimates were off because they didn't account for the increase in demand spurred by the program's offer of essentially unlimited benefits.
This was a new problem for America. "Prior to Medicare," explains John Goodman, president of the National Center for Policy Analysis and a frequent contributor to the health policy journal Health Affairs, "we maintained a system that took up a reasonable percentage of the national income," holding more or less steady at 5 percent of GDP. But after Medicare, he says, the country "began to have health inflation that has never quit."
Coincidence? Not at all. Medicare was a major contributor to the problem. For beneficiaries, it transformed the health care system into a generously subsidized, all-you-can-eat buffet. For providers, it offered a steady revenue stream that they used to rapidly build out expensive new services. In 2007 MIT economist Amy Finkelstein published a paper estimating that the introduction of Medicare accounted for a 23 percent increase in total hospital expenditures between 1965 and 1970, with an even larger effect in the subsequent five years.
Part of the problem was that the program served up its smorgasbord all at once. On July 1, 1966—Medicare's very first day of operation—19 million individuals were instantly eligible for its benefits. Not one of them had ever paid a dime to directly support the program, but they collected full benefits anyway. That situation was at odds with the way the program had been sold, which was not as an entitlement but as a government-managed savings mechanism. When President John F. Kennedy outlined his original vision of the program, he declared, "We're not asking for anybody to hand this out—we are asking for a chance for the people who will receive the benefit to earn their way." In reality, when the program began, it was pure handout.
Seniors got the medical benefits, but doctors got the money. The payment system offered doctors and hospitals essentially unrestricted payments. Providers invoiced their expenses, and the government paid. The system gave doctors and hospitals both license and incentive to spend— and spend and spend and spend. Which is exactly what they did.
By 1970 the Senate was circulating memos examining ways to tamp down runaway Medicare spending growth. A set of 1972 amendments to Social Security included some unsophisticated attempts to cut Medicare payments, but hospital inflation continued to spiral upward, as did public concern about the issue. Hospital costs were growing at roughly twice the general inflation rate; between 1974 and 1977, total government spending on health care doubled.
Control Payments, Control Spending?
Medicare didn't just drive spending upward; it also made health spending a government problem. With polls showing that the rapidly rising cost of health care was one of America's top three policy concerns, elected officials took notice. And President Jimmy Carter came to believe that controlling payments was the best way to exert influence on the system.
In April 1977, Carter introduced a proposal to put strict limits on reimbursements to hospitals, which were believed to be a key driver of health care inflation. After years in which spending grew by an average of 15 percent, Carter wanted to impose a ceiling of 9 percent. Because he was responding to concerns about escalating costs across the medical sector, his plan imposed a federal cap on spending growth for both public and private payers. Rather than a single-payer system, it was a variation known as "all-payer."
But Carter's plan faced intense opposition from the American Hospital Association (AHA). The AHA successfully killed cost-control legislation twice under Carter by convincing a group of conservative Democrats that hospitals could control costs voluntarily. But the AHA's members failed to hold up their end of the political bargain. As President Ronald Reagan took office, hospital inflation continued to balloon, growing 13 percent in 1980 and 18 percent in 1981. Voluntary restraint, it turned out, was no match for the temptation of nearly unlimited federal funds.
In 1982 Reagan responded with the Tax Equity and Fiscal Responsibility Act (TEFRA), which, among other things, would have imposed strict new limits on Medicare payments. The law ditched Carter's "all-payer" idea and addressed public programs only.
The idea was to encourage cost-effective treatment by paying providers on a per-patient rather than fee-for-service basis. But TEFRA was intended more as a negotiating tactic than an actual reform. As David G. Smith, currently a professor emeritus of political science at Swarthmore College, points out in his 1992 book Paying for Medicare: The Politics of Reform, TEFRA was "designed to help the hospitals perceive the desirability" of reform. It wasn't an overhaul; it was a threat.
Hospitals got the message and relented, dropping their opposition to a new payment system. In less than two months, Congress approved a system of standardized payments that paid a flat rate per case and allowed federal officials to determine hospital payment rates in advance. Reagan administration staffers convinced themselves the reform was market-based because standardized prices rewarded more efficient providers, but there was no mistaking the system's fundamental element: bureaucratic price setting.
The only question was how the bureaucrats would decide what to charge. A robust system would need to create enough diagnostic groups to contain every possible patient and every possible diagnosis. That meant creating categories, and lots of them.
Think of a filing cabinet with hundreds of drawers, each labeled according to a particular category of medical diagnosis. These are Medicare's diagnosis-related groups (DRGs), and every hospital patient is assigned to one. After that, prices are set based on the average cost of everything in the drawer. Anything from drawer number 707 —major male pelvic procedures—gets one price. Anything from drawer number 385—inflammatory bowel disorders—gets another price. What if a treatment ends up costing far more or less than the assigned price? The theory is that it doesn't matter, since everything averages out in the end. Everything is covered; everything has a code and a category.
Or at least it's supposed to. When the system was launched in 1983, there were 500 DRG codes. But like so many bureaucratic systems, this one grew larger and more complex over time. By 2010 there were almost 750 DRG codes. No matter how narrowly and meticulously Medicare's bureaucrats organized their files, they were always forced into an implicit admission of defeat in their quest to create an all-encompassing system. Each revision of the DRGs left a junk drawer—labeled "ungroupable"—at the very end.
Leave the Hospital, Go to the Doctor
Did the Reagan administration's price controls restrain Medicare spending? Within hospitals, yes. But the system had an unintended consequence: Hospitals, paid according to the diagnostic drawers their patients fit in rather than according to their own rates, were given an incentive to cycle through patients much faster. As a result, hospital stay lengths dropped dramatically, and so did inpatient hospital spending. A RAND Corporation study published in 1992 found that the new payment system reduced the length of hospital stays by an average of 24 percent.
Quicker discharges often meant sicker discharges. Where do sick patients who have been discharged from the hospital go? To doctor's offices, outpatient hospital services, or, in the case of seniors, nursing homes, none of which was covered under the new payment scheme. That's where the spending went too. Starting in 1983, outpatient hospital spending rose at three times the rate of inpatient spending. And according to a 2003 World Health Organization review of U.S. health policy, the savings generated by shorter hospital stays were offset by increased spending on nursing home and outpatient care. Throughout the 1980s, Medicare spending on physician payments grew an average of 13.7 percent—7 percent faster than all other services. By 1990 total Medicare payments to physicians had blown up to two and a half times what they were a decade earlier.
So by the mid-1980s, policy makers were hunting for yet another payment reform. This time physician payments would be the target. Whack one mole, another one pops up.
In 1985 Congress commissioned a review of the DRG system. But rather than stick to a retrospective, the authors of the report expanded its scope, using it as an opportunity to push for expanded physician price controls. The theory was that DRGs, focusing only on hospital payments, hadn't gone far enough. According to Smith's Medicare history, the authors of the review believed "the most viable approach…was to aim at increasingly global control of physician payments." A program passed on a promise to avoid control of the medical system was now looking to control its entire payment structure.
Who was behind this push for global control? Joseph Antos, an American Enterprise Institute health policy scholar who served as a senior economist in the Reagan administration —a role he describes as the "designated health and everything else person"—attended many senior staff meetings during which the new system was drawn up. "This was a group that consisted of high-level people, often political appointees from all sorts of different agencies," Antos tells me. "I don't think many of them were health policy experts. I know none of them were economists."
The system they were planning would become known as the resource-based relative value scale, or RBRVS. It attempted to divide physician's services into roughly equal work units and make payments accordingly. The assorted high-level officials had a naive confidence in their ability to accurately align the amount of work that went into a procedure with the amount of payment a physician received. "They knew that there was a problem paying physicians," Antos says. "They thought they knew what the problem was. This was going to be a new system that was going to rationalize the old system."
Antos, the only economist in the group, wasn't so sure. And so he began to ask questions: "How does the government know what the relative values should be? How is this related to any market-clearing process that anybody's ever known?" One idea was to set prices by committee. Antos pointed out that "asking committees of doctors to guess how much work is involved in something is the same thing as just setting prices."
In an October 2010 essay for The American, Antos described the initial plan as being "based on academic theory with its roots in the Soviet Union." Just as the Soviets made all economic decisions—how many tanks to build, how many jackets to sew, how much food to produce—through central planning, the RBRVS system is an effort to centrally plan medical prices. But as in the Soviet Union, those prices are not informed by market-based signals, which are generated by the interaction of supply, demand, and willingness to pay. In particular, the RBRVS system ignores how much value a patient receives from a service.
Thanks at least in part to Antos's questions, 1986 came and went with no major overhaul of the physician reimbursement system. But Antos eventually left for a new post. And in 1988 researchers at Harvard University finalized a study that would bring a modified form of prospective payment to physicians. In December 1989, as part of an omnibus budget proposal, President George H.W. Bush signed the RBRVS system into law. It would take effect in January of 1992.
Antos, who eventually transitioned to a senior position at the Health Care Financing Administration (HCFA, now the Centers for Medicare & Medicaid Services within the Department of Health and Human Services), was put in charge of implementing the system—not in spite of his skepticism but because of it. "I had a long connection to it, so I understood it," he says. "And [HCFA Administrator Gail Wilensky] didn't mind that I was against it, because she was an economist and also agreed that it wasn't going to work."
The Socialist Calculation Problem
Why would an economist be so skeptical of the system? Even from a purely technocratic perspective, it is an enormous challenge. Antos warns of the "technical difficulty of creating a prospective payment system that wouldn't totally screw everything up."
But the problem goes deeper than that. Medicare's twin payment schemes are inevitably beset by what George Mason University economist Arnold Kling calls "the socialist calculation problem." The bureaucrats in charge of setting prices have to come up with a rational basis for the prices they set. They have to be justified, somehow, which is where the complex rate-setting formulas come into play. But without price signals, the result is almost always an arbitrary formula based on a limited, imperfect set of factors. When all is said and done, says Kling, "it's just a made-up formula. It has to be."
The other problem is that any payment system inevitably ends up being manipulated by savvy payees. "You price on the basis of one thing, but then people optimize their behavior to that thing," says Kling. In a sense this is the primary job of health care administrators: to understand payment systems and squeeze every possible dollar out of them.
In the wake of the two payment reforms, hospitals began to manipulate the system through "upcoding"—systematically shifting patients into higher-paying DRGs. Research by economists at Dartmouth University suggests that during the early 1990s, hospital administrators figured out ways to substantially increase the number of Medicare cases they billed to higher-paying DRGs. Payment games continue today. In October the Senate Finance Committee released a report accusing several large home health care companies of abusing Medicare's payment rules by pushing employees to perform extra therapy visits, thereby qualifying for Medicare bonus payments, even when those visits weren't strictly necessary. But for many health care providers, that's the business. Hospital administrators "are people whose job it is to game the system," Kling says. "They know every little detail of the rules."
Playing by the rules, and getting the most out of them, becomes the focus. Over time, the rules cease to guide the game and instead become the purpose of the game. Activities that are coded and paid for become the activities that providers do the most. The system encourages covered procedures, such as surgeries and child delivery, while discouraging doctors from spending time in nonpaid activities such as emailing patients or monitoring health data collected electronically at home by the patient. The provision of care bends to fit the shape, however quirky, of the payment rules.
Which may explain why controlling physician payments failed to restrain the growth of Medicare spending. As Antos expected, the system did not work. The RBRVS system took effect in 1992. By 1997 Congress had the mole mallet out once again.
The Unsustainable Growth Rate
The RBRVS system included a mechanism, known as the volume performance standard, aimed at preventing doctors from gaming the standardized payments by increasing the number of cases they processed. If total physician spending increased, fees went down. If total physician spending decreased, the fees went up. The problem was that the formula was based on historical trends in volume, which had been rising for years. But in the mid 1990s, that trend unexpectedly slowed, leading to substantial increases in Medicare's physician fees.
Under President Bill Clinton, a Republican Congress tried out a new payment mechanism intended to control volume as part of the 1997 Balanced Budget Act. It tied payments to the size of the economy in the hope of controlling inflationary spending by keeping costs per patient from rising faster than GDP. If total spending on physician reimbursements stayed under the spending target, fees would go up. If reimbursements exceeded the target, fees would go down.
For a few years, payments to providers rose as planned, and spending stayed within budgetary targets. But like previous payment reforms, the Clinton-era "sustainable growth rate" (SGR) formula put pressure on one part of the system, shifting costs elsewhere. Douglas Holtz-Eakin, a former director of the Congressional Budget Office, argues that the formula has two major flaws: First, it targets only one component of Medicare—physician spending—rather than the program as a whole. Second, despite its goal, it does not really control volume. "Congress failed to understand that physicians respond to incentives," he says. Lower reimbursements inevitably result in more services being performed. "Cut rates, they will respond." Rather quickly, the system started to break down. As the journal Health Affairs dryly noted in a recent primer on the issue, "The expectation that this payment system would control spending was not realized."
That was not entirely the fault of SGR. Legislators in Congress did not let their own system work. The reform allowed for payment reductions in order to keep spending in line with the economy. But in the booming 1990s, when the law was passed, most policymakers never expected that payments would ever fall. As long as GDP grew and payments rose in response, the system worked mostly as planned. But in 2002 the formula called for a 5 percent reimbursement cut to keep payments in line with the flagging, post–tech bubble economy.
Congress allowed the cut to take effect, but doctors weren't happy. The grumbling was loud enough that when the formula called for another cut in 2003, Congress overrode it and voted to institute a small reimbursement hike. Since then that pattern has held: Each year the SGR has called for a reimbursement cut, and each year—sometimes multiple times a year—Congress, ever susceptible to outside influence, has instead voted to delay the cut though a temporary patch, now known widely as "the doc fix."
The lack of congressional fortitude has created additional headaches for doctors and patients. Although doctors took a pay cut only once, the temporary nature of the extensions still means that Medicare payments are riddled with uncertainty. Almost everyone—doctors and policy makers alike—assumes the cuts will never go through. But they don't know for sure. That makes doctors wary of relying too much on Medicare payments, which already lag behind the rates paid by private insurers. As a result, some are reducing the number of new Medicare patients they see, while others are dropping out of the program entirely. A 2010 survey by the Texas Medical Association found that 100 to 200 doctors in the state were giving up on Medicare each year. The SGR's unlikely but persistent threat of dramatic fee cuts has made it harder for seniors to obtain health care.
Doctors, led by the American Medical Association, have escaped those big cuts so far. But lobbying pressure to override the programmatic cuts has exacerbated the long-term problem. When Congress replaced a scheduled reduction in 2004 and 2005 with a 1.5 percent increase, it did not bother to change the long-term spending target. Consequently, the SGR called for even bigger cuts down the road to make up for the short-term hikes. As the overrides have mounted over the years, so have the cuts called for by the formula. There is now an enormous chasm between what physicians who accept Medicare are supposed to be paid under the SGR and what they are actually paid.
In December 2010, congressional leaders announced a tentative deal to pay for a one-year extension of the doc fix by trimming funding for the first year of ObamaCare's insurance subsidies. But time is running out. By the formula's reckoning, doctors face as much as a 29.5 percent cut at the beginning of 2012. Depending on how long Congress continues to delay the cuts, an even steeper reduction looms in the future—an estimated 40 percent if the charade continues until 2014.
Abolishing the SGR entirely, as many doctors would like, could cost up to $370 billion over a decade, according to the Congressional Budget Office (CBO). But the Obama administration, despite cutting more than $500 billion in Medicare payments to pay for the president's health care overhaul, and despite calling for another $248 billion in Medicare reductions as part of his debt reduction plan, has paid little attention to the problem. In February 2011, Health and Human Services Secretary Kathleen Sebelius told members of Congress that the administration thinks the doc fix is "very important to do." Early drafts of the health care overhaul included a doc fix. But in the end, Democrats chose to use the law's Medicare cuts to pay for expanded coverage rather than to stabilize physician payments. An administration budget proposal this year also called for the doc fix to be fully paid for. How? The administration won't say. Instead, it has offered up enough money to extend the doc fix for just two years—and then only by reducing the rate of Medicare spending growth over a full decade. This is rather like a lifelong two-pack-a-day smoker promising to quit next year, then spending the money he'll "save" on cigarettes over the course of the year at a bar that evening.
But the SGR puts Republicans in a tough spot too. CBO projections, based on current law, assume that the SGR's scheduled cuts will take effect. Few Republicans want to be seen as advocating what the CBO will count as hundreds of billions in additional Medicare spending. But neither do GOP legislators want to be seen as advocating a nearly 30 percent reduction in physician reimbursements, which will further reduce seniors' access to doctors.
Medicare's resident technocrats have been somewhat more forthcoming with proposals. In October, the Medicare Payment Advisory Commission, a 17-member panel of experts that advises Congress about how to structure the program's reimbursements, voted to recommend a decade-long doc fix. Its proposal, which can't go into effect without congressional approval, would cut specialist rates by 5.9 percent annually for three years while freezing primary care reimbursements. But this plan pays for only $100 billion or so of the 10-year cost; the remaining $200 to $300 billion would come from cuts in payments to hospitals, drug makers, acute care facilities, and other providers. Provider groups immediately launched an aggressive opposition campaign; given Congress's historical reluctance to let doctors take a hit, it seems unlikely that this proposal will succeed where others have failed.
Holtz-Eakin, the former CBO director, argues that members of Congress need to recognize that the cost of recurrent SGR overrides is already built into the system. "We've already really committed to spending this money," he says. After accepting that reality, he says, Congress should move to a new system that puts the entire Medicare program on a budget and turns control of that budget over to individual seniors. "It's the first step that's killing every-one, though," he says. "No politician wants to be seen as adding $300 or $400 billion to the deficit."
The second step won't be easy either. In April, House Republicans voted for a budget plan authored by House Budget Committee Chairman Paul Ryan (R-Wis.) that would have transformed Medicare from an essentially unlimited program, committed to endless spending, into a premium support system that would allow seniors to buy regulated plans from private insurers. Democrats spent the following months accusing House Republicans of having voted to "end Medicare as we know it," a line that many promised to repeat all the way up to the 2012 election. But according to Antos, the American Enterprise Institute health policy expert, transforming the system is the only way to escape the flaws of SGR and other price controls. "If you leave the structure of Medicare alone," he says, "you cannot solve the problem."
ObamaCare and Medicare
Obama didn't transform the system so much as double down on its faults. Much like the introduction of Medicare, ObamaCare extends subsidized insurance coverage to millions of people, a move that is likely to spur additional demand.
It also starts a new round of price-control Whac-A-Mole via yet another mechanism designed to rein in Medicare spending. The Independent Payment Advisory Board (IPAB), an ostensibly independent 15-member panel, is charged with reviewing Medicare spending each year and making recommendations to Congress aimed at keeping costs below a threshold based on inflation in both the health care sector and the economy as a whole. Most of the board's recommendations are likely to address reimbursement rates. Those changes will take effect unless Congress votes them down and approves an alternative savings plan.
The board's boosters have great hopes that it will finally control the growth of Medicare. But the program's own number crunchers have less confidence. In May 2011, Medicare's actuary and its director of cost estimates warned that "it is doubtful that Medicare providers can take steps to keep their cost growth within the bounds imposed by these price limitations, year after year, indefinitely." Over time, they said, the new "price constraints would become unsustainable," noting that ObamaCare's payment updates likely would result in 15 percent of hospitals, home health agencies, and nursing facilities operating at negative margins by the end of the decade.
The current head of the CBO, Douglas Elmendorf, also seems wary. In the summer of 2009, when IPAB was first discussed, Elmendorf wrote a letter to Congress that said "in CBO's judgment, the probability is high that no savings would be realized…but there is also a chance that substantial savings might be realized"—a polite way of signaling minimal confidence in the board's ability to cut costs.
That suggestion was affirmed the following summer, just months after ObamaCare was enacted, when Elemendorf released the CBO's long-term budget outlook. In the alternative fiscal scenario, which is based not on current law but on the CBO's best guess as to how legislation will change and evolve, Elmendorf assumed ObamaCare would fail to achieve its intended Medicare savings. The CBO's mission does not include making policy recommendations. But it was hard to read the CBO's report as anything other than an implicit jab at the notion of relying on an independent advisory board to control costs.
And why shouldn't the CBO—or anyone else—be skeptical? Congress has not given up the game of Whac-A-Mole. It has merely assigned an independent committee to play for it. The board will face the same problems that price controllers have always faced: Without market signals, prices are inherently arbitrary. It's the socialist calculation problem all over again. There is no way for policy makers to avoid it—except, of course, by refusing to play the price-setting game at all.
At this point, that does not seem likely. Although it was passed based on an explicit promise not to control the practice of medicine, Medicare is now defined by its many payment processes, which exert substantial influence on how doctors and hospitals treat patients. "Medicare screws up the American medical system, period," says Holtz- Eakin. It does so not by telling doctors what to do but by deciding what to pay. He who controls the price, controls the system.
Peter Suderman is an associate editor at reason.