November 30, 2011

November 30, 2011 | Backgrounder on Health Care, Budget and Spending

Government Price Controls for Health Care: A Deficit-Reduction Strategy to Avoid

Key Points

  1. It is imperative that Congress reduce the deficit—and it is equally important that Congress learn from the disastrous effects of previous price controls in health care.
  2. Price controls depend on government price-setting and market interference—and the magnitude of projected savings is often based on the false assumption that the health care sector will not respond to bizarre incentives created by a distorted market.
  3. Government price controls in health care have generally caused more problems than they solved, and imposing further top-down control over the prices of medical goods and services will adversely affect patients and doctors.
  4. Congress should embrace market competition, which includes moving health care entitlements from an open-ended, defined-benefit structure to a defined-contribution model.

Abstract: The Joint Select Committee on Deficit Reduction—the “Super Committee”—created under the Budget Control Act of 2011 has failed to recommend a strategy for reducing the federal deficit by $1.2 trillion over the next decade, leaving automatic cuts of 2 percent on the table for Medicare, other domestic programs, and national security and defense budgets. Rather than accept this as the status quo, the rest of Congress must now act to stop federal overspending by focusing on the main drivers of long-term runaway spending: health care entitlement programs. As they move forward, Members of Congress must bear in mind the importance of avoiding more harmful policies in the name of deficit reduction and instead ensure that reform results in the systemic changes necessary to success.

There are a variety of ways to achieve savings in health care entitlements, though not all are sound policy. One way—price controls—depends on government price-setting and market interference. Setting prices administratively, instead of allowing the market to determine prices, appears to be an appealing method of deficit reduction because, on paper, it offers savings. But the magnitude of projected savings is often based on the false assumption that the health care sector does not respond to bizarre incentives created by an artificially distorted market.

The Joint Select Committee on Deficit Reduction (the “Super Committee”) has failed in its responsibility to recommend reductions in the national deficit of at least $1.2 trillion, triggering across-the-board cuts in most areas of government spending, including Medicare. Despite the committee’s failure, Congress still has an obligation to the American people to reverse chronic overspending in Washington, and mindless across-the-board reductions will not suffice. In addressing the deficit, Members of Congress should learn from the disastrous effects of previous health care price controls and avoid more of the same. [1] Congress should embrace the principles of market competition, [2] which include moving health care entitlements from an open-ended, defined-benefit structure to a defined-contribution model.

Government Price Controls Are Not the Solution

The federal government sets prices for services paid for by Medicare, Medicaid, and the Veterans Administration (VA) employing a variety of mechanisms, with a variety of consequences. Medicare’s complex fee schedules overpay providers for some services, underpay them for others, and therefore do not reflect the value of medical goods and services accurately. As a result, providers who treat Medicare beneficiaries are encouraged by perverse financial incentives to offer inefficient, less-effective care. Seniors receive unnecessary tests and treatments, have less time with their physicians, and face mounting barriers to access to physicians’ services. Price controls for drug coverage have been just as damaging. Low payment for prescription drug coverage by Medicaid and the VA increases costs for other purchasers and bars access to effective treatments using restrictive formularies.

Across the board, price-setting in health care has failed to produce expected savings, meanwhile devastating the quality of, as well as access to, health care. The only way to undo the damage wrought by decades of government price controls is to address the true causes of the problems of America’s health care system with market-oriented reforms.

Price-Setting: The Effect on Medicare Patients

The Medicare payment system is a complex web of formula-driven rates and statutory updates that contribute to the program’s rising spending, and which do not reflect market conditions. The passage of the Patient Protection and Affordable Care Act (PPACA) only makes this problem worse, ratcheting down provider reimbursement and opening the door to even more cuts. Now that the Super Committee has failed to produce the required deficit reduction, more damaging across-the-board cuts to Medicare providers are looming. While there are several deficit reduction proposals before Congress that might garner a favorable score from the Congressional Budget Office (CBO), they would set in place harmful policies that are unlikely to actually yield tangible savings, meanwhile promising to threaten patient care.

Perverse Incentives Hurt Inpatient-Care Quality. Medicare pays hospitals according to the Prospective Payment System (PPS), which offers a flat fee per episode of care based on the patient’s assignment to one of the hundreds of “diagnosis-related groups” (DRGs). The PPS was intended to encourage more efficient care by paying providers the same amount regardless of which treatment is provided, since it allows providers to keep the savings when a patient’s care costs less than the reimbursement received. Even so, the system is flawed. Since hospitals are paid the same regardless of how long a patient stays in the hospital, paying a flat rate per episode of care creates financial penalties for every additional day that a patient remains in the hospital. As Heritage Foundation health policy analyst Ed Haislmaier writes, the PPS payment system has led hospital administrators to think in the following terms:

The more patients he treats, the more revenue his hospital receives. But those paying the bills, whether patients or insurance companies, want the hospital to treat patients quickly and effectively, and then discharge them to recuperate in a less costly setting. These conflicting incentives encourage the administrator to make his hospital a more efficient treatment center, with shorter patient stays and a high turnover rate.[3]

High patient turnover has contributed to a steady decline in length of hospital stays for those ages 65 and above. According to the Centers for Disease Control and Prevention, the average length of stay dropped from 10.7 days in 1980 to 8.7 days by 1985, and 5.5 days by 2006.[4] Data from the Medicare Payment Advisory Commission (MedPAC) shows that the length of stays between 1999 and 2009 for Medicare inpatients fell 12 percent, while the length of stays for non-Medicare inpatients did not change.[5] However, premature hospital discharges can also cause unnecessary hospital readmissions, which result in lower-quality care and increased spending. According to the Dartmouth Atlas of Health Care, of the more than nine million seniors admitted to hospitals annually, “almost one in five of these patients are readmitted within a month of discharge.”[6] The Centers for Medicare and Medicaid Services estimates that the cost of unplanned readmissions accounts for more than $17 billion of Medicare’s annual cost.[7]

Furthermore, the PPS does not encourage greater efficiency because it fails to give patients incentives to seek out the best value in health care, one of the reasons Medicare has been unable to reject expensive low-quality care through the decades. The possible savings from providers’ incentives are also limited, since they are only encouraged to offer care that is of equal or lesser value to the price set by Medicare, which in most cases does not reflect the true cost of the services provided.

Medicare hospital-payment reductions may also shift costs to the privately insured. Though it remains unclear to what extent cost-shifting occurs, studies have shown that it takes place. The impact of the hospital payment cuts from the Balanced Budget Act (BBA) of 1997 resulted in a shift to private payers. The degree of cost-shifting was small for most hospitals, but urban hospitals and those serving a greater share of privately insured transferred as much as 37 percent of the payment reductions from the BBA to private payers.[8] Hospitals also shift costs to other areas of care. According to the American Association of Retired Persons (AARP), “the shift of many services from inpatient to outpatient hospital settings is reflected in Medicare’s spending.”[9] In 1980, 67.4 percent of Medicare spending went toward inpatient hospital care; in 2010, this amount had fallen to just 26 percent.[10] In 2001, before the addition of prescription drug coverage contributed to Medicare’s overall cost, 39.1 percent of total Medicare expenditures were for hospital care.[11]

The Physician-Payment Mess. According to the CBO, when payment rates are set administratively, “annual updates to those prices may reflect statutory formulas or legislative purposes to budgetary and other pressures that may deviate from the changes in providers’ costs.”[12] Medicare’s physician-payment formula, which Heritage Foundation health policy expert Robert Moffit describes as “based on the methodology of social science rather than on market forces,” is a prime example.[13] Physician reimbursement is determined by the relative value of doctors’ services, calculated with a complex formula to take into account a variety of factors, and then adjusted to reflect geographical differences. The relative value units are then converted into a dollar amount.

Even if this system had initially borne some resemblance to the actual average costs that physicians incur when providing their services, since Medicare was enacted in 1965, Congress has struggled—and failed—to update the fee schedule in a way that controls cost growth without deviating dangerously far from market conditions. The current method is to use the sustainable growth rate (SGR), enacted as part of the BBA to tie annual updates to per capita growth of gross domestic product (GDP). If spending exceeds the expenditure target, the update to the fee schedule is reduced. In theory, this should keep growth in Medicare spending below the set target and prevent it from surpassing growth in the overall economy. In practice, no such thing has occurred. Instead, Congress has delayed the SGR update since 2003 because its annual reductions to physician reimbursement would threaten Medicare beneficiaries’ access to doctor services. If Congress allows the SGR update for physician payments to go into effect in 2012, the reduction of reimbursements will amount to 29.4 percent.

Two major problems arise from the cost-control strategy behind the SGR. First, according to MedPAC, “the formula aggregates spending across all physicians and practitioners who furnish services to Medicare beneficiaries and, therefore, does not provide incentives at a more granular level…to control volume growth or improve care quality.”[14] The system arbitrarily rewards some physicians, and punishes others, in a pattern that is completely divorced from the value of care provided. Furthermore, tying updates to the activity of the overall economy is nonsensical, since the growth pattern of GDP has little bearing on supply or demand for physicians’ services.

Medicare spending has meanwhile continued to grow, while reimbursement cuts encourage provider behavior that benefits neither the patients enrolled in the program nor the taxpayers paying for it. Throughout the history of price controls, one fact has been apparent: When the price for a good or service is artificially lowered, its supply or quality,or both, falls as well. As provider payments fail to keep pace with growth in provider costs, doctors have had to find ways to make up for the loss, mainly by increasing the volume and intensity of their services. According to the CBO, between 1997 and 2007, the behavioral response of physicians to changes in payment rates offset 28 percent of the potential federal savings.[15] This means that Medicare reduced payments to physicians with the expectation that it would create savings, but because doctors responded to the new, adverse conditions by inducing demand, savings were lower than expected. Meanwhile, Medicare per-beneficiary spending on physicians’ services increased 39.4 percent as a direct result of increased quantity of services.[16]

Volume and use of services vary due to a number of factors, including disease prevalence, medical innovation, and demographics, but reduced physician payments have also played a role. At the same time, doctors’ responses to reduced payments have included decreasing the length of office visits in order to fit in more patients, which reduces the attention they can give each patient, thereby reducing the quality of their care.

Losing Sight of the Patient. Medicare reimbursements, rather than the needs of patients themselves, increasingly dictate provider behavior and the quantity and quality of the care they provide. The Medicare payment system has removed patients’ needs as the centerpiece of the health care system and replaced them with the priorities of bureaucracy. As John O’Shea, M.D., writes,

Administrative pricing systems pursue a simple-minded objective: cut costs. Costs, however, are only half of the value equation. In Medicare physician payment, the SGR mechanism has no link to the quality of the services provided and contains no incentives for physicians to provide, or for patients to demand, better quality of care.[17]

Price-setting has not led doctors or hospitals to provide more efficient care. On the contrary, by failing to financially discriminate against hospitals and physicians that offer lower-quality care at higher prices than their competitors, price-setting discourages effortsto offer better value.

Reduced access to physicians is another way that price controls hurt patients. Payment cuts are an indirect way for the Medicare bureaucracy to ration care by controlling the supply of physicians’ services. As reimbursement falls, more physicians become incapable of taking new Medicare patients or continuing to see the Medicare patients they have already accepted into their practice, which can lead to waiting lines for pivotal medical interventions.[18]

Price Controls in Medicaid and the VA

Medicaid and the VA also use administratively set pricing, which has similar adverse effects on access, cost, and quality of care as does price-setting in Medicare. Most notably, Medicaid and the VA employ price controls in the form of rebates from drug manufacturers.[19] These are often described as “negotiated” prices, though rebates are, in fact, determined and set into statute through legislation. The closest thing to negotiating in this process is the decision by manufacturers to accept the set price or forgo selling their drugs to the Medicaid and VA patient populations. This system has raised costs for other drug purchasers and restricted access to effective treatments.

Higher Costs for Everyone Else. Under Medicaid’s prescription drug coverage, manufacturers must pay the government a rebate that is determined by a complex formula; before the passage of the PPACA, manufacturers paid back either 15.1 percent of the average manufacturer price (AMP) for all brand-name drugs (11 percent for generics), or the difference between the initial Medicaid rate and the best available price, whichever was larger. Under the PPACA, the rebate rates will increase to 23.1 percent and 13 percent, respectively. Drug manufacturers must also pay an additional rebate, which makes up more than half of the costs recouped by Medicaid, if a drug’s price increases by more than general inflation from year to year. This rebate is the reason that Medicaid’s total rebates are significantly larger than those received by private purchasers in the commercial market or in Medicare Part D.

To a certain extent, manufacturers accept the reduced prices set by Medicaid because of the program’s large market share. But they do not simply eat this loss. While Medicaid’s direct cost to taxpayers falls, growth in drug prices above what would have otherwise occurred increases costs across the board. According to the HHS Office of Inspector General, the wholesale acquisition cost, the AMP, and the unadjusted Medicaid payment rate for brand-name drugs (all of which more closely reflect the prices paid by private purchasers) increased by approximately 30 percent between 2005 and 2009. The increase in inflation as measured by the consumer price index (CPI) was 10 percent over the course of the same time frame. Meanwhile, Medicaid’s prices grew by just 2 percent, much slower than the rate of inflation.[20] As this trend continues, the impact of artificially low Medicaid payments on the prices that the rest of drug purchasers pay will become even more pronounced. Just a few years after the enactment of the rebate program, the CBO stated that,

Although the basic rebate has lowered Medicaid’s expenditures on outpatient prescription drugs, spending on prescription drugs by non-Medicaid patients may have increased as a result of the Medicaid rebate program. In particular, the best-price provision has increased the prices paid by some purchasers in the private sector. Since Medicaid constitutes between 10 percent and 15 percent of the market for outpatient prescription drugs, pharmaceutical manufacturers are much less willing to give large private purchasers steep discounts off the wholesale price when they also have to give Medicaid access to the same low price.[21]

Strict Formularies Limit Access to Effective Options. Since its enactment, Democrats have pushed to replace Medicare Part D’s competitive model with government-fixed pricing that mimics the system used by the VA. There are several reasons why the VA model would be unsuccessful for Medicare besides the fact that Medicare represents an enormous portion of the market compared to VA beneficiaries.A 2008 analysis by the Lewin Group shows that the VA’s low drug-pricing is not a result of achieving better value for the same drugs available to seniors under Part D. Instead, the VA employs a restrictive formulary to keep prices low by excluding high-cost, effective drugs. The Lewin analysis shows that, of the 281 top drugs offered by Part D formularies, only 183 (65 percent) were included in the VA formulary. By contrast, the Part D plan with the highest enrollment covered 99 percent of the top drugs, and the plan with the second-highest enrollment covered 100 percent.[22]

Veterans’ access to brand-name drugs was even worse: The VA formulary covered just 51 percent of the 116 brand-name drugs available under Part D plans. The Part D plan with the highest enrollment covered 98 percent; the Part D plan with the second-highest enrollment covered 100 percent. Earlier research by Frank R. Lichtenberg of Columbia University also showed that veterans have less access to new drugs than other insured Americans. Of the drugs approved by the Food and Drug Administration in the 1990s, Lichtenberg’s findings show that only 38 percent were included in the VA formulary. Of the drugs approved since 2000, only 19 percent were included.[23]

Restricting the availability of new and effective treatment has obvious consequences on the health and well-being of beneficiaries. According to Lichtenberg’s study, increased reliance on older drugs reduced the average age of death by slightly more than two months. Denying or delaying effective treatment clearly reduces quality of care, which might increase health spending by requiring patients to obtain more expensive medical treatment later due to the lack of certain drugs.

What Congress Should Do

The history of price controls in health care shows that they always fall short of expectations, while unleashing undesired side effects on patient care. To put programs like Medicare back on sound financial footing, Members of Congress should avoid enacting more failed policies and should:

  1. Reject new price controls.There are a variety of proposals for importing price controls to the health care system. President Barack Obama proposed applying Medicaid rebates to drug coverage for the low-income Medicare population who are eligible for both Medicare and Medicaid. These seniors are currently enrolled in Medicare Part D, which provides affordable prescription drug coverage through a menu of private options. Government intrusion into pricing is barred by a “non-interference” clause. Importing Medicaid’s price controls would risk increasing premiums for all Part D enrollees and could threaten their broad access to pharmaceuticals. Moreover, as Joseph Antos warns, the level of savings projected by the CBO would likely not materialize “because of the poorer deals negotiated by Part D plans, undercutting net savings to the government after factoring in the rebate revenue.”[24] Indeed, estimates of the financial impact of this policy show that cost increases could lead to a rise in monthly beneficiary premiums of 25 percent to 50 percent.[25]

    The President also proposed subjecting prescription drug coverage offered through the Federal Employees Health Benefits Program (FEHBP) to government price-setting. Unlike the traditional defined-benefit model for employer-sponsored coverage, the FEHBP allows federal workers to select a health plan from a menu of options. This model, similar to that of Medicare Part D, has successfully contained costs and maintained access to high-quality coverage. The secret ingredient is competition among insurers and the ability of beneficiaries to switch plans if one stops working for them. The President’s approach would allow the director of the Office of Personnel Management, which oversees the FEHBP, to “contract directly for pharmacy benefit management services on behalf of all FEHB enrollees and dependents,” saving just $1.6 billion.[26] This strategy would require restricting federal employees’ access to drugs. Tampering with the means of the FEHBP’s success would create measly savings at the expense of quality coverage for federal workers and is another suggestion from the White House that Congress should avoid.[27]

    In the case of both Part D and the FEHBP, the only way the government can achieve costs below those that the market has already achieved is by doing something that private plans either cannot or will not do. Unlike government plans, private plans in competition with each other incur the pressures of consumer demand not to restrict access to certain drugs, and are thus unlikely to use this form of rationing as a means of cost-containment.
  2. Reverse existing price controls. Congress should reverse the price controls that are already embedded in Medicare, Medicaid, and the VA health plan. Since the PPACA reinforces many of the failed policies of the past, its repeal is the next vital step to undoing the damage wrought by government interference in the health care system. The PPACA increases the number of patients who depend on government-funded health care due in part to its expansion of Medicaid, which will increase the portion of patients for whom providers receive low, government-set reimbursements.The program’s low reimbursements already limit beneficiaries’ access to care and negatively impact physician behavior, and its expansion will further reduce quality of care.[28] The new law’s arbitrary cuts to Medicare threaten provider profitability and, if fully enacted, will introduce Medicaid’s access and quality challenges into Medicare. The PPACA also lays the groundwork for extending government price controls into private insurance through its rate-review provisions. There are plenty of other existing price controls, like the SGR, that need reform.
  3. Allow market forces to determine value in the health care system. Congress should pursue reform that allows market forces to achieve equilibrium between consumer demand for health care and the prices paid for it. Allowing the health care system to operate according to the principles of market competition would create real incentives for providers and insurers to offer higher-value services that respond to the needs and demands of patients. In Medicare and Medicaid, Congress should begin to transition toward a defined-contribution model, where beneficiaries are able to enroll in a private plan of their choice, using a government contribution to offset costs. Replacing the complex fee schedules and arbitrary price controls would benefit both beneficiaries and taxpayers.


Experience with government-imposed price controls in federal health care programs shows them to be an insufficient and unnecessary way to resolve the issues they aim to address, largely causing more problems than they solve. Imposing further top-down control over the prices of medical goods and services will not reduce deficit spending without adversely affecting doctors and patients. Instead, Congress should move Medicare and other federal health programs in the opposite direction—improving health care for Americans, and reducing the federal deficit at the same time.

Kathryn Nix is a Policy Analyst in the Center for Health Policy Studies at The Heritage Foundation.

About the Author

Kathryn Nix Policy Analyst
Center for Health Policy Studies

Show references in this report

[1]Edmund F. Haislmaier, “Why Global Budgets and Price Controls Will Not Curb Health Costs,” Heritage Foundation Backgrounder No. 929, March 8, 1993, at

[2]James C. Capretta, “The Case for Competition in Medicare,” Heritage Foundation Backgrounder No. 2605, September 12, 2011, at

[3]Haislmaier, “Why Global Budgets and Price Controls Will Not Curb Health Costs,” and Capretta, “The Case for Competition in Medicare.”

[4]National Center for Health Statistics, “Health, United States, 2010: With Special Feature on Death and Dying,” 2011, Table 99, at (November 14, 2011).

[5]Medicare Payment Advisory Commission, “A Data Book: Health Care Spending and the Medicare Program,” June 2011, Section 6, at (November 14, 2011).

[6]David C. Goodman, Elliott S. Fisher, and Chiang-Hua Chang, “After Hospitalization: A Dartmouth Atlas Report on Post-Acute Care for Medicare Beneficiaries,” Dartmouth Institute for Health Policy and Clinical Practice, September 28, 2011, at (November 14, 2011).

[7]Stephen F. Jencks, Mark V. Williams, and Eric A. Coleman, “Rehospitalizations Among Patients in the Medicare Fee-for-Service Program,” The New England Journal of Medicine, Vol. 360 (April 2, 2009), pp. 1418–1428, at (November 14, 2011). 

[8]Vivian Y. Wu, “Hospital Cost Shifting Revisited: New Evidence from the Balanced Budget Act of 1997,” International Journal of Health Care Finance and Economics, Vol. 10, No. 1 (2010), pp. 61–83, at (November 14, 2011).

[9]AARP, “Medicare at 40: Past Accomplishments and Future Challenges,” July 2005, p. 3, at (November 14, 2011).

[10] 2011 Annual Report of the Boards of Trustees of the Federal Hospital Insurance and Federal Supplementary Medical Insurance Trust Funds, May 13, 2011, p. 9, at (November 14, 2011).

[11] Ibid.

[12]Congressional Budget Office, “Key Issues in Analyzing Major Health Insurance Proposals,” Chap. 5, December 2008, at (November 14, 2011).

[13]Robert E. Moffit, “Comparable Worth for Doctors: A Severe Case of Government Malpractice,” Heritage Foundation Backgrounder No. 865, September 23, 1991, at

[14]Medicare Payment Advisory Commission, “Report to the Congress: Medicare and the Health Care Delivery System,” June 2011, at (November 14, 2011).

[15]Congressional Budget Office, “Factors Underlying the Growth in Medicare’s Spending for Physicians’ Services,” Background Paper, June 2007, at (November 14, 2011).

[16] Ibid.

[17]John O’Shea, “A Predictable Mess: Medicare’s Physician Payment System Offers Lessons Against Drug Price Negotiation,” Heritage Foundation WebMemo No. 1330, January 25, 2007, at

[18]Bacchus Barua, Mark Rovere, and Brett J. Skinner, “Waiting Your Turn: Wait Times for Health Care in Canada 2010 Report,” Fraser Institute, December 6, 2010, at (November 14, 2011).

[19]Greg D’Angelo, “The VA Drug Pricing Model: What Senators Should Know,” Heritage Foundation WebMemo No. 1420, April 11, 2007, at, and Derek Hunter, “Government Controls on Access to Drugs: What Seniors Can Learn from Medicaid Drug Policies,” Heritage Foundation Backgrounder No. 1655, May 27, 2003, at

[20]Daniel R. Levinson, “Medicaid Brand-Name Drugs: Rising Prices are Offset by Manufacturer Rebates,” Department of Health and Human Services Report OEI-03-10-00260, August 2011, at (November 14, 2011).

[21]Congressional Budget Office, “How the Medicaid Rebate on Prescription Drugs Affects Pricing in the Pharmaceutical Industry,” January 1996, p. 11, at (November 14, 2011).

[22]The Lewin Group, “Comparison of VA National Formulary and Formularies of the Highest Enrollment Plans in Medicare Part D and the Federal Employee Health Benefit Program,” December 10, 2008, at (November 14, 2011).

[23]Frank R. Lichtenberg, “Older Drugs, Shorter Lives? An Examination of the Health Effects of the Veterans Health Administration Formulary,” Manhattan Institute Medical Progress Report No. 2, October 2005, at (November 14, 2011).

[24]Joseph Antos and Guy King, “Tampering with Part D Will Not Solve Our Debt Crisis,” American Enterprise Institute Health Studies Working Paper 2011-03, June 29, 2011, at (November 14, 2011). 

[25]The Lewin Group and Ingenix Consulting, “Financial Impacts on Medicare Beneficiaries if Larger Part D Rebates Are Required for Medicare/Medicaid Dual Eligibles,” June 25, 2009, p. 5, at (November 14, 2011).

[26]Office of Management and Budget, “Living Within Our Means and Investing in the Future: The President’s Plan for Economic Growth and Deficit Reduction,” September 2011, p. 43, at (November 14, 2011).

[27]Robert E. Moffit, “Bad Medicine for Federal Workers and Taxpayers: Killing FEHBP Competition,” Heritage Foundation WebMemo No. 3403, November 2, 2011, at

[28]Once the State Children’s Health Insurance Program (SCHIP—later renamed CHIP) went into effect, affected physicians reduced the duration of individual patient visits. Craig L. Garthwaite, “The Doctor Might See You Now: The Supply Side Effects of Public Health Insurance Expansions,” National Bureau of Economic Research Working Paper No. 17070, May 2011, at (November 14, 2011).