Jump to ContentJump to Main Navigation
Fiscal TherapyBalancing Today's Needs with Tomorrow's Obligations$

William G. Gale

Print publication date: 2019

Print ISBN-13: 9780190645410

Published to Oxford Scholarship Online: April 2019

DOI: 10.1093/oso/9780190645410.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 18 February 2020

Reforming Healthcare

Reforming Healthcare

(p.123) 7 Reforming Healthcare
Fiscal Therapy

William G. Gale

Oxford University Press

Abstract and Keywords

As Chapter 7 documents, healthcare already accounts for a quarter of federal spending and will grow significantly in the future. Almost half of Americans get health insurance through government programs. But Americans spend substantially more of their income on private and public healthcare than citizens of other advanced countries, without always getting good value. Policymakers should expand coverage—which improves people’s health, reduces financial hardship, and saves lives—by penalizing people who do not buy health insurance and incentivizing states to expand Medicaid. Policymakers should control costs—which would raise wages and reduce long-term debt— and improve healthcare quality by paying medical providers based on the quality of care and patient outcomes, not the services they perform. Medicare should use competitive bidding and get the same drug price discounts as other programs. Policymakers should reduce tax subsidies that encourage people to use more healthcare than they need.

Keywords:   healthcare, Medicare, Medicaid, Affordable Care Act, Obamacare, CHIP, fee-for-service, bundled payments, premium support, single-payer system

Healthcare reform must be a central part of the fiscal solution. Healthcare programs already account for a quarter of the federal budget, and they will grow significantly as the population ages and medical technology evolves. Almost half of Americans get their health insurance through public programs, including Medicare for the elderly and Medicaid for the some of the poor, disabled, and elderly. Other programs serve low- to moderate-income children, members of the military, and veterans; provide tax subsidies for healthcare insurance; and subsidize insurance purchases for low- and moderate-income families in the private market.

We like to think that we have the world’s best health system, and in some ways we do. Our doctors and hospitals use state-of-the-art technology, and we’re a world leader in pharmaceutical innovations and in treating heart conditions, hypertension, cancer, and other major diseases. But too many people lack health insurance, which impairs their health and causes economic hardship. And our costs are too high. Healthcare spending has been rising for several decades, and, counting both public and private outlays, we spend far more than any other country with little difference in health outcomes. Slower spending growth would free up resources for other uses, which would help the economy grow faster, boost wages, and raise living standards.

Notably, these are issues with the healthcare system, not just government programs. Our policies, therefore, must induce change in both private and government healthcare. Fortunately, precisely because it plays such a large role in health insurance, the federal government can induce private-sector changes by reforming its own programs.1

In 2010, President Obama and a Democratic-controlled Congress enacted the Affordable Care Act (ACA, also known as “Obamacare”). It required people to obtain adequate insurance coverage, created online marketplaces (“exchanges”) that centralized information about private insurance plans, stopped insurance companies from refusing coverage or raising premiums due to applicants’ preexisting conditions, heavily subsidized states that expanded their Medicaid (p.124) programs, slowed Medicare spending, and provided incentives for better medical care. The law has raised coverage, slowed healthcare spending growth, and improved the quality of care.

Despite these gains, President Trump and a Republican-controlled Congress undid parts of the ACA in 2017 and 2018. That’s the wrong direction for policy. Policymakers should reinstate the ACA provisions that were removed. To raise coverage further, they should convince the states that have not expanded Medicaid to do so, and they should create a “public option” for health insurance in the exchanges. To curtail costs, the government should pay medical providers based on the quality of care they provide and the patient outcomes they secure, not the services they provide. Medicare should expand its use of competitive bidding, and it should get the same drug price discounts as other government programs. Policymakers should limit tax subsidies for health insurance and adjust taxes on alcoholic beverages to better account for their social costs.

These reforms would raise insurance coverage, improve care, and make people healthier. They would take a big bite out of the fiscal shortfall, reducing the debt-to-GDP ratio by about 18 percentage points by 2050 (Appendix Table 7.1).

The two broad problems noted earlier—low access and high costs—help frame the discussion. Regarding access, most Americans believe the federal government should ensure that people have health insurance.2 About 91 percent of Americans had private or public health insurance in 2018.3 The remaining 9 percent were disproportionately young adults (aged 26–34), minorities, or those with low incomes. Among them, some don’t want to pay for coverage they think they don’t need, but most cite cost as a barrier. The failure of 17 states to expand Medicaid under the ACA explains why many poor people remain uninsured.4

About 175 million people received health insurance through employer-sponsored plans in 2018.5 Health insurance is the most popular fringe benefit among workers and it’s not subject to income tax, even though it’s a form of income that workers receive. That tax exemption reduces the cost of buying healthcare by a third for middle-income families and by half for high-income families.6 But it also encourages families to use more health services than they need and it costs the government lots of money—a whopping $341 billion in forgone revenues in 2017, almost as much as federal spending on Medicaid.7

About 24 million people bought private insurance directly in 2018.8 About 40 percent of them used the ACA’s federal- and state-run exchanges. The federal government subsidizes purchases in the exchanges through progressive refundable tax credits for individuals with household incomes between 100 and 400 percent of the poverty line, or up to $97,200 in income for a family of four in 2018.9 The other 60 percent bought insurance directly or went through an insurer or broker.10 Individual and small-group plans are typically more expensive (p.125) than employer-provided plans because the smaller group size means that overall costs are more variable and thus costlier to insure.

Medicare provides insurance for people aged 65 and over.11 In 2018, it covered about 60 million people and, after Social Security, was the largest federal program, costing $583 billion in net spending. Medicare accounts for 20 percent of all national healthcare spending. The program has four parts. Everyone who is at least 65 years old is eligible for Part A, often called traditional Medicare, which is free and covers care in hospitals, homes, nursing homes, or hospice centers. Part B, called supplementary medical insurance, is voluntary but requires participating individuals to pay premiums. Like private insurance, it covers physician services, outpatient care, lab tests and preventive visits; participants pay a monthly premium and face deductibles and copayments. Part D—jumping out of order for a moment—subsidizes prescription drugs for Medicare enrollees and is available only through private insurance plans. Part C—called Medicare Advantage—lets beneficiaries receive Medicare benefits (Parts A, B, and often D) by joining a private insurance plan. Medicare Advantage covers about one-third of all Medicare beneficiaries.12 To supplement Medicare coverage, some elderly buy “Medigap” policies, private insurance plans that typically cover deductibles and copayments that Medicare does not cover and sometimes cap lifetime out-of-pocket payments as well.

Medicaid provides health insurance for some low-income children, families, and individuals; the aged; and the disabled. People move on and off Medicaid as their economic status changes. In any given month in 2018, the program financed healthcare services for about 68 million people. Over the year, though, it reached about 100 million individuals, or roughly 30 percent of Americans.13 Medicaid is a crucial part of children’s healthcare across the country. Children make up almost half of all recipients, and Medicaid covers an astonishingly large share of young children—about 45 percent—nationwide. Medicaid is a federal-state program, with the federal government paying half the costs in the richest states, rising to 75 percent in the poorer states.14 In 2018, Medicaid accounted for 17 percent of national healthcare spending, including federal spending of about $380 billion, making it the third largest federal program.15

In 2018, the Children’s Health Insurance Program provided insurance to roughly 7 million children and pregnant women in families with incomes that are too high to qualify for Medicaid but too low to afford private plans.16 Eligibility and programs vary by state. Most states cover children who live in families with income up to 200 percent of the poverty line. Like Medicaid, CHIP is a federal-state program. States can use CHIP funds to expand their Medicaid programs for children, create their own CHIP program, or have a combination of the two.17

(p.126) The Veterans Health Administration (VHA) provides an integrated network of healthcare facilities for almost 9 million veterans, including medical centers, outpatient clinics, and nursing homes.18 The military provides care—through a network of facilities and a health insurance plan called Tricare—to more than 9 million people around the world, including active duty and retired service members, their families, and surviving family members.19

Having health insurance improves people’s health and financial status and reduces death for preventable reasons.20 Moreover, better health for any particular individual can lead to better health for others by reducing the spread of disease. Of course, personal choices and social factors affect health, too, suggesting that the social spending programs advocated in Chapter 9 could well help to improve health in addition to their other positive impacts.

Medicare substantially raises access to health insurance and reduces financial risks among the elderly.21 Medicaid makes people more likely to use health services, reduces the chances of having unpaid medical bills and catastrophic out-of-pocket medical spending, reduces rates of depression, improves self-reported ratings of good health, and reduces the incidence of low-birthweight infants and infant mortality.22 Children enrolled in Medicaid or CHIP are more likely to have annual check-ups and flu shots, and less likely to have unmet medical needs (such as reading glasses).23 They’re more likely to complete high school and college, and less likely to participate in risky behavior.24 From a pure budgetary perspective, a significant portion of Medicaid spending pays off in the long run via higher tax receipts and lower EITC payments.25

Turning to cost issues, healthcare spending averaged more than $11,000 per person in 2018 and is distributed very unevenly. The 50 percent of patients with the lowest spending accounted for just 3 percent of overall costs, while the top 5 percent accounted for half of all costs.26 Most of the high spending is driven by patients who face a significant medical procedure (like treatment for a heart attack), have a chronic condition (like hypertension or diabetes), or are in the last year of life.27 As a result, significant cost savings will come from reducing the incidence and severity of extremely expensive cases. Getting the 50 percent of the population with the lowest costs to reduce costs further won’t save much money.

Healthcare spending—both public and private—has grown dramatically over time, from about 5 percent of GDP in 1960, to about 9 percent by 1980, and more than 18 percent by 2018 (Figure 7.1). Over the past 25 years, spending per beneficiary in inflation-adjusted terms rose at about the same rate in the private sector as in Medicare. Medicaid per-beneficiary spending grew more slowly.28

Reforming Healthcare

Figure 7.1. Health spending over time (1960–2040). Projections of Medicaid include only the federal share. Major federal programs include gross Medicare, Medicaid, Children’s Health Insurance Program (CHIP), and exchange subsidies. Author’s calculations; Boards of Trustees (2018); Centers for Medicare and Medicaid Services (2018a, 2018b); Congressional Budget Office (2018b).

The long-term rise in healthcare spending stems from several factors that interact with each other in complex ways. Five items—technological (p.127) innovation; health insurance expansion; fee-for-service plans; trends in income, aging, and health; and administrative costs—are worth a closer look.29

Technological innovations are the single biggest factor in boosting healthcare spending.30 Many are expensive, raising overall healthcare spending (in part by making new procedures and services possible). They include imaging technologies, surgical procedures, cancer drugs, and surgical robots. Many have greatly improved the quality of healthcare.31

Health insurance is a big part of the story as well. It spread dramatically after World War II. Insurance reduces healthcare costs for consumers, encouraging them to use more healthcare services—perhaps more than they need. Out-of-pocket healthcare spending has stayed constant at about 2 percent of GDP over the last 60 years even though total healthcare spending rose substantially.32

(p.128) The form of insurance has also boosted spending. Insurance companies, Medicare, and Medicaid in the past used a fee-for-service approach, with healthcare providers receiving payments for the services they perform, not the quality of the outcome.33 That incentivizes doctors, hospitals, and other medical providers to provide too many services for which they are well compensated and too few that are poorly compensated. To put it bluntly, a significant amount of medical spending—as much as one-third in some studies—is unnecessary. For example, there is significant variation in healthcare treatments and expenses across the country without discernible differences in outcomes.34

Over the past 50 years, income, aging, and health trends have raised the demand for healthcare as well. Per-capita GDP has risen substantially, giving more Americans the resources to afford insurance and health treatments. The aging of the population has greatly boosted Medicare’s rolls, with the share of the population over age 65 rising from less than 10 percent in 1970 to 16 percent in 2018. The share of Americans over age 85 has also risen dramatically, and Medicare spends twice as much annually per enrollee aged 85 to 94 as it does for those aged 65 to 74. Similarly, the share of Americans with a chronic health condition—for example, obesity or type 2 diabetes—has risen substantially.35 Some 38 percent of adults were clinically obese in 2014, up from 15 percent in 1980,36 and 60 percent of adults had at least one chronic disease in 2014, up from 45 percent in 2005.37 The one-quarter of Americans with at least three chronic conditions accounts for two-thirds of all health spending.38

Administrative costs have risen dramatically.39 Doctors deal with numerous insurance companies, each of which may have different billing procedures and requirements. Patients who deal with their providers and insurers face numerous filing issues.

Collectively, the factors above help explain why US healthcare spending has grown so much and point to why American healthcare spending is so much higher than in other countries. Whereas healthcare spending in the United States grew from 11 percent of GDP in 1990 to 17 percent in 2016, it grew from 7 percent of GDP to 11 percent in the rest of the G7 countries (Figure 7.2).40 Public spending on healthcare as a share of the economy is about the same in the United States as in other advanced countries, but private spending here is much higher.41

Reforming Healthcare

Figure 7.2. Total and public healthcare spending, United States vs. G7, 1990–2016. G7 average excludes the United States. OECD (2017).

Differences in healthcare spending between the United States and other advanced countries have grown dramatically since 1980 and are plausibly related to a variety of factors: the push toward deregulation in the 1980s and the absence of healthcare price controls, the greater increase in income inequality in the United States, and lower social spending in the United States.42

(p.129) Many of the trends driving US healthcare spending—including technological innovation, the spread of health insurance, and trends in aging and health—also occurred in other advanced countries. The other G7 countries have some form of national health insurance and/or control costs directly. This blunts some of the impact of the common trends. For example, the incentive to use new, expensive technology may be stronger when consumers have fee-for-service plans, in which doctors are paid for the services they provide, and when fees are determined in the private market, rather than under national health insurance plans, where doctors are paid salaries or are paid fees that are subject to government control.

A second factor—also related to the spread of private fee-for-service systems versus national health insurance—is the higher level of administrative costs in the United States. Fragmented private healthcare markets generate higher administrative costs than healthcare systems with a single payer. By one count, such costs are 8 percent of health expenditures in the United States compared to 1 to 5 percent in other high-income countries.43 Medicare, which is a single-payer system, has far lower administrative costs than the US private healthcare sector.44

(p.130) The third factor driving divergent healthcare spending trends is that the prices of almost everything health-related are higher in America than elsewhere. General physician salaries average over $218,000 in the United States compared to a range of $86,000 to $154,000 in other advanced countries.45 On average, retail brand-name drug prices are 10 to 15 percent higher in the United States than in Canada, France, and Germany and 50 percent higher than in the United Kingdom.46 The average cost of hospital stays is more than 60 percent higher in the United States compared to other OECD countries.47

In principle, spending more on healthcare need not be a problem. We have higher incomes than other countries, and people naturally demand more healthcare as their income rises.48 In practice, however, we have less access to health insurance than other advanced countries and we do not generate better health outcomes along a wide variety of dimensions ranging from infant mortality rates to obesity and life expectancy. A 2014 Commonwealth Fund study found that the United States is “last or near last on dimensions of access, efficiency, and equity” among 11 advanced countries.49

The analysis above points to several major approaches to policy reform—restructuring provider payments, reducing administrative costs, limiting prices of healthcare services and products, and improving access to health insurance. We return to these issues below.

How much healthcare spending will continue to rise will greatly affect the budget outlook. Even with the overall rise in healthcare spending of the last several decades, the nation experienced two periods when such spending either rose at a reduced rate or fell as a share of GDP. In the 1990s, the emergence of health maintenance organizations drove down costs, but a consumer backlash spurred the spread of less restrictive health plans and healthcare spending resumed its earlier trend. Another slowdown began around 2003 and has persisted more or less to date.50 Why it occurred and whether it will continue is unclear. The Great Recession of 2007 to 2009 helped limit spending because people consume less when they have less income.51

Some of the slowdown, however, is related to more permanent changes in how Americans provide and use healthcare. The share of insurance plans with high deductibles has risen, inducing patients to use less healthcare. Providers are replacing fee-for-service systems with those that pay on a per-patient basis or on outcome-based criteria. Healthcare is increasingly provided by nurse practitioners, physicians’ assistants, and other health professionals, rather than doctors, further reducing costs. Healthcare delivery organizations that can provide less expensive service—for example, patient centers and minute (p.131) clinics—are increasing. Digital recordkeeping is spreading quickly, with the potential to reduce paperwork and administrative costs.

All of these trends can help dampen future spending growth, but population aging will almost inexorably raise healthcare spending. Total healthcare spending is projected to rise from 18.3 percent of GDP in 2018 to 20.6 percent of GDP in 2026.52 About half of the increase will come in major government programs. The rise in spending is projected to continue beyond 2026. By 2050, federal spending on the major programs is slated to reach 8.5 percent of GDP, about 87 percent of the projected deficit in that year.53 About 40 percent of this increase is due simply to population aging, with the rest owing to higher incomes, new drugs, and technological advances.

Paying for healthcare spending will be an increasing challenge. General revenues finance Medicaid and, along with participant premiums, Medicare’s Part B, C, and D. But traditional Medicare benefits (Part A) are financed out of a trust fund. The trust fund receives the 2.9 percent Medicare payroll tax on all earnings, split between employers and employees and the additional 0.9 percent tax on earnings above $200,000 ($250,000 for couples). Part A benefits can be financed only from the trust fund balance or new payroll taxes. The trust fund balance is slated to hit zero by 2026, leaving only dedicated tax revenue (mainly payroll taxes) to finance benefits. As a result, either payroll taxes must be raised, the financing will have to change, or benefits will have to be cut by approximately 9 percent in 2026 and 15 percent in the long term.54 The changes proposed later will help extend the date at which the trust fund is exhausted, but in any case policymakers should simply cover any shortfall in Part A financing by using general revenues, which is what is assumed in the baseline projections in Chapter 3.

Moving forward, we need to build on the ACA. As the nation’s most comprehensive healthcare reform since Medicare and Medicaid in 1965, it took an essentially conservative approach to healthcare reform, aiming to disrupt the existing system as little as possible. To ensure that significant and effective healthcare insurance was available to all legal residents regardless of income or health status, the ACA tapped the private market to make it easier to buy insurance rather than adopting a single-payer plan that covers all individuals.55 The mandate and subsidy structure were modeled on a plan that Mitt Romney pushed through as governor of Massachusetts—a plan based on a proposal by the conservative Heritage Foundation.56 To slow healthcare spending growth and increase the value in healthcare, the ACA created new incentives for private insurers to compete, and it made changes to Medicare that increased the rewards to providing value rather than just services. The key to the plan was the triumvirate of (1) the (p.132) individual mandate to obtain adequate insurance or pay a penalty, (2) subsidies to help households pay for insurance, and (3) the requirement that insurance companies issue policies and set premiums without regard to pre-existing conditions. Together, these policies create a viable and functioning market with near-universal coverage. However, each provision is needed to ensure that the others work properly. Without the mandate that everyone must obtain coverage, for example, some healthy people would drop out of the market, making insurance more expensive for everyone else, and making it harder for people to buy policies. Likewise, without the provision blocking insurance companies from considering pre-existing conditions, premiums for some people would be so high that they could not afford to buy insurance.

To help implement these changes and facilitate comparisons across insurance policies, the ACA created marketplaces where insurance companies bid on standardized health insurance packages. The act also heavily subsidized states to expand Medicaid for the poor and children, added measures to control spending and reform provider payment systems, and raised several taxes to pay for the changes.

The ACA has produced positive results. It was projected to reduce deficits by $350 billion in its first decade and $3.5 trillion in its second.57 It raised health insurance coverage (Figure 7.3); after hovering between 12 and 15 percent for (p.133) many years, the percentage of Americans who are uninsured fell below 9 percent in 2016.58 The increase in coverage reduced income inequality and personal bankruptcies.59 The ACA also boosted the quality of healthcare in some ways, which, in turn, has reduced costs. For example, the changes in Medicare policy have led to lower hospital readmission rates and patient infection rates.60

Reforming Healthcare

Figure 7.3. Uninsured rate before and after enactment of Affordable Care Act (ACA). Centers for Medicare and Medicaid Services (2018a, 2018b).

Nevertheless, President Trump and Congress moved to undo aspects of the ACA. The 2017 tax bill effectively eliminated the individual mandate, which will raise the number of uninsured individuals by an estimated 8 to 9 million by 2027.61 The Trump administration also altered cost-sharing reduction subsidies that help low-income earners pay for healthcare, slashed the budget for advertising enrollment in exchange-based insurance plans, approved waivers that let states make it harder for their residents to receive Medicaid, and lowered barriers to buying bare-bones insurance plans that don’t comply with the ACA. The net result of all these changes is that healthier people will drop out of the exchanges, which in turn will raise the cost of healthcare insurance for the remaining, less healthy pool of applicants and thus cause fewer of them to buy insurance as well.62

Instead of repealing and replacing part or all of the ACA, we should recognize it as the most promising way to expand access, improve the quality of care, and control costs. Building on the ACA, we should emphasize two themes: (1) harness the power of competition and consumer choice to cut costs, and (2) use the public sector to ensure that coverage is as broad as possible and that private insurance markets function well.

For starters, policymakers should reinstate and enforce the penalty for the individual mandate—or impose a similar penalty, such as eliminating the use of the standard deduction and itemized deductions in the income tax, for those who do not have qualified coverage.63 That would bring healthier people back into the market, raise coverage, and reduce healthcare insurance premiums. The Internal Revenue Service should have the power to collect the penalty as it would any other tax, and the funds to do so.64

In addition, policymakers should appropriate funds for the cost-sharing reduction payments that help low-income families pay for healthcare insurance.65 This would, on net, improve the fiscal situation.

Policymakers also should make it easier for working families to get affordable healthcare insurance.66 People with access to “affordable” employer-sponsored health insurance aren’t eligible for federal tax credits to help them buy coverage in the ACA exchanges. But because of a drafting error in the 2010 legislation, the so-called family glitch, the law is too restrictive with respect to the definition of “affordable,” and many families who should get subsidies do not. By loosening up the law, policymakers can help millions of families pay for healthcare insurance.67

(p.134) As of August 2017, every county in America had at least one insurer participating in its exchange, though about 3 million people had only one available insurer.68 Policymakers should boost the supply of insurers on the exchanges; increased competition would give consumers more choices and hold down prices. To do that, they should take three steps:

  • First, they should provide a stronger financial backstop for insurers. ACA redistributes funds to protect insurers on the exchanges against unpredictable costs that arise from the requirement that they not reject customers.69 But extremely high-cost enrollees still introduce significant risk into the market. Initially, the ACA had a reinsurance program that provided a backstop to address this problem, but it expired in 2016, at which point premiums rose significantly.70 Reinstating this program could help stabilize insurer participation in the exchanges, leading to lower premiums for households.71

  • Second, policymakers should create a public option for health insurance on the exchanges, raising coverage and creating competition for private plans to reduce costs. It would be based on Medicare but updated to improve its cost-sharing structure and account for differences in the population it would serve.72 That would save the government significant amounts through lower spending on exchange subsidies and increased tax revenue.73

  • Third, all states should expand Medicaid, as they can under the ACA with substantial federal subsidies. As of this writing, 17 states haven’t done so, leaving millions of low-income households without health insurance and without access to federal subsidies to buy a policy on the exchanges. States that expanded Medicaid have boosted healthcare coverage dramatically within their borders and have experienced lower premiums on their exchanges. Having all states expand Medicaid would increase state spending by about $60 billion over a decade, increase Medicaid coverage for more than 4 million people, and improve the health and finances of those who gain coverage.74

These changes will raise coverage and increase consumer choice. But we can do much more to cut costs. The traditional fee-for-service approach wastes money, rewarding providers for the services they provide, not the medical outcomes of these services. Provider payment reform offers the potential to reduce costs and increase the quality of care. Medicare, which has long been a pioneer in developing new payment systems, has piloted over 20 new types of alternative payment models, as the ACA authorizes.

The most promising alternative payment model is “bundled care,” through which insurers define large “episodes of care”—for example, a hip replacement and 90 days of post-surgery care—and pay a set amount that’s divided up among the providers involved.75 Bundled payments encourage coordinated, (p.135) cost-effective medicine since the payment depends on the medical situation, not the treatment. In a government-sponsored experiment in the 1990s, hospitals receiving a bundled care payment for a certain heart surgery cut costs by 10 to 40 percent and reduced patient mortality.76 Private efforts to develop bundled care have also shown considerable success in reducing costs and raising the quality of care.77 The government should adopt a policy that bundles services during a patient’s hospital stay plus any post–acute care that’s provided within 90 days thereafter.

Expanding competitive bidding in Medicare will reduce government spending while preserving beneficiaries’ options to get Medicare at no additional cost. There’s already a significant amount of competitive bidding in government health programs. In the ACA exchanges, insurers announce the prices at which they will offer a standard benefit package. In Medicare Part D, insurers announce the price at which they will provide a drug benefit package that meets certain minimum standards. In Medicare Advantage, insurers submit bids to provide beneficiaries with a wide range of Medicare standard and supplemental benefits.

I propose that we implement “premium support,” which is a form of competitive bidding. Medicare and private insurers would announce how much they would accept to provide Medicare benefits (Parts A and B) to a beneficiary in average health in each region of the country. Those eligible for Medicare would then choose a plan. For each enrollee in a plan, the federal government would pay the insurance company the average bid in the region less a standard premium that beneficiaries owed.78 The standard premium would roughly equal one-quarter of the cost of Part B services, about what beneficiaries currently pay. Beneficiaries who choose a plan that costs more than the government contribution plus the standard premium would pay the difference. Those who choose a less expensive plan would receive a partial rebate. This plan would generate significant government savings because private insurers’ bids would generally be lower than the cost of traditional fee-for-service Medicare. Beneficiaries’ total payments for Medicare premiums and cost sharing would be lower than projected under current law.79

Curtailing drug prices would help many people afford the drugs they need and reduce overall healthcare costs.80 US drug prices are higher than those in similar advanced countries with well-developed drug markets. The United States spends much more than any other country on prescription drugs. Drug spending has risen rapidly relative to overall healthcare spending since the 1980s, and that’s expected to continue over the next decade.81

At the same time, we should preserve incentives for drug companies to develop new therapies, which often requires large upfront costs for research, (p.136) development, and testing. If the law doesn’t protect newly developed drugs to some extent, any company could produce a drug without paying the upfront costs, undercutting the profits of the company that put in the work and offering little incentive for it to invest in the first place.82 Patents appropriately give drug manufacturers some protection for new drugs;83 once the monopoly ends, generic drugs can enter the market and tend to reduce prices dramatically.84

We should change the way Medicare pays for drugs. The nation’s largest prescription drug buyer, it has no authority to negotiate which drugs it covers and at what prices.85 Medicaid must cover all approved drugs, but it receives at least a 23 percent rebate on drug prices and doesn’t face price increases that exceed inflation. The Department of Veterans Affairs (VA) receives at least a 24 percent rebate, can negotiate prices, and can reject certain drugs.86 Medicare should be able to negotiate drug prices and reject certain drugs with the same authority as the VA.87

The tax exclusion for employer-sponsored health insurance is expensive and regressive, encourages gold-plated health insurance plans, and raises healthcare spending. With these considerations in mind, the ACA created a “Cadillac tax” on high-cost, employer-sponsored plans that’s supposed to take effect in 2022. Taxing health insurance is the right approach, but a different tax would work better. Policymakers should replace the Cadillac tax by eliminating the income tax exclusion for the benefits from employer-sponsored insurance plans that exceed the cost of the median healthcare plan. Among its other benefits over the Cadillac tax, this policy would make healthcare costs more transparent to beneficiaries.88

The federal government imposes “sin” taxes on purchases of alcohol and cigarettes. By raising the price of these products, the taxes reduce smoking and alcohol consumption.89 That improves people’s health and reduces the costs that smoking and drinking impose on others, such as drunk-driving accidents. In light of the 2009 tax increase from $0.39 to $1.01 per pack of cigarettes, studies suggest that cigarette taxes are now high enough to offset the societal costs of smoking.90 But taxes on alcohol should be raised to incorporate the full social costs of drinking, including drunk-driving accidents, increased violence and domestic abuse, and higher medical costs.91 Moreover, alcohol is taxed inconsistently. Beer and wine are taxed by liquid volume, whereas distilled spirits are taxed based on alcohol content. Raising the excise tax on all alcoholic beverages to $16 per proof gallon (or to about 25 cents per ounce of alcohol) would standardize the rate for all alcohol, reduce alcohol consumption, improve health, reduce drunk-driving deaths, and raise revenues. That roughly translates to a tax increase of 50 cents per six-pack of beer or 60 cents per bottle of wine.92

(p.137) The changes I advocate would produce (near) universal coverage and lower prices in both the over-65 and under-65 markets. But healthcare reform will be a process, with a lot of experimentation, not a single event. There are many additional promising approaches.

To raise access further, policymakers could let individuals buy into Medicare or Medicaid even if they are not eligible by current standards. To deepen consumers’ choices on the exchanges, policymakers could require that all non-employer insurance be sold on exchanges rather than through private brokers.

To reduce costs, policymakers could limit monopoly power among providers, hospitals, insurers, and drug makers; cut administrative costs; reform malpractice rules;93 and encourage greater use of non-physician providers (physician assistants, nurse practitioners, clinical nurse specialists, etc.). Policymakers could also consider “rate setting,” a policy under which the government directly limits how much hospitals and doctors can charge.94

To reduce drug prices, Medicare could stipulate that a generic drug that is approved in other advanced countries is by default approved in America and could prohibit manufacturers from charging higher prices here than in those other countries. Under “reference pricing,” Medicare could group drugs based on their therapeutic effects and set a price per group. This could simultaneously regulate prices of existing drugs while giving incentives to produce new ones.95

When drug makers charge what can only be described as obscenely high prices, due to their monopoly power, the federal government should step in. Drugs for hepatitis C, for example, can cost tens of thousands of dollars for a 12-week treatment, prompting insurers (including Medicaid) to restrict patients’ access to them. The government has the power to pay patent holders “reasonable and entire compensation” and then produce the drug itself or let others do so.96

Further provider payment reform could help reduce the cost of the expensive episodes that account for such a large share of healthcare costs. “Pay for performance” programs—or value-based payments—offer financial rewards to providers who achieve pre-established quality benchmarks. The ACA created Accountable Care Organizations (ACOs), groups of providers that are collectively responsible for the health and payment outcomes of a specific population, that have grown rapidly.97

Finally, we should encourage the states, as “laboratories of democracy,” to experiment with alternative systems as long as they do not reduce coverage or raise costs.

While policymakers have plenty of options to pursue, there are a few they should steadfastly avoid. For starters, they should not convert Medicaid to a block grant to the states. With a block grant, the federal government gives states a set amount of funds each year to address a problem (e.g., healthcare, hunger) (p.138) and, in exchange, states get more leeway in how to spend the money. If, however, a state’s costs for its Medicaid program exceeded its federal Medicaid funds—because, for instance, healthcare costs went up more than expected or a weak economy put more struggling people on Medicaid—that state would have to cover all of the additional costs itself or cut Medicaid services. Under this scenario, because states often lack the capacity to raise taxes or cut other popular programs, they would likely have to scale back Medicaid services or throw some low-income people off the rolls. Block grants have led to substantial reductions in welfare payments over the past 20 years, disproportionately affecting lower-income households (Chapter 9).

Nor should policymakers raise the age of Medicare eligibility. This would shift healthcare costs from Medicare to the individuals who would no longer be eligible, but it wouldn’t do anything to strengthen the healthcare system. It wouldn’t save Medicare much money either. The youngest elderly (e.g., 65- or 66-year-olds) are the least expensive people to cover in Medicare. Among the downsides of this policy, costs would rise for the seniors affected as well as for employers who provided coverage for employees who could no longer receive Medicare.

Policymakers do not need to provide new incentives for high-deductible (HD) healthcare plans or health saving accounts (HSAs), which give consumers a tax-preferred way to save money that they can then use, if needed, for their healthcare expenses.98 High-deductible plans have been on the rise in recent years. They have lower premiums but larger out-of-pocket costs. Proponents argue that high deductibles prompt patients to make wiser, more cost-effective choices when deciding on their care. In practice, however, consumers tend to “throw the baby out with the bathwater” by reducing all forms of care, including high-value care like preventative screenings.99 Meanwhile, HSAs cost the government a lot in forgone revenue, and they mostly benefit those with high incomes who least need the help. And the potential for HD/HSA plans to reduce healthcare costs more than they have already is limited, as so much of aggregate healthcare spending is dominated by a relatively small number of people with extreme health episodes whose costs far exceed even very high deductibles.

On a final note, a “single payer” system, in which the government finances healthcare for everyone, is the holy grail for many reformers, especially for Democrats. Many countries have a single-payer system, though they vary a bit. In Britain, the government finances and provides healthcare; in Germany, it finances private insurers; in Canada, it pays for national health insurance, while private providers supply the care. In the United States, Medicare is a single-payer system for the elderly.

(p.139) A single-payer system brings big potential advantages. Coverage is universal; costs are generally lower because the government can set rates for doctors, hospitals, and drugs; and administrative and advertising costs are minimized. I’m very sympathetic to those goals, and a single-payer system might be what we’d want if we were starting from scratch.

But we’re not. We’re starting from a system that accounts for a sixth of the economy, with entrenched government programs, insurance companies, and physicians, and 175 million people who get employer-sponsored insurance. The switch to a single-payer system would generate immense transition costs for the government and private sector, and people would face massive “sticker shock” as the program would greatly expand the size of government.

Three states have considered and rejected single-payer systems in recent years. Vermont, among the nation’s most liberal states, rejected it in 2014 in part due to its tax burden.100 California, where it would have cost more than the entire state budget at the time, rejected it in 2017.101 In Colorado, about 80 percent of voters rejected it in a referendum in 2016.102 Thus, in America, single-payer healthcare seems to be more of a slogan than a solution.

Healthcare reform is central to the fiscal outlook and to the lives and livelihoods of our citizens. We have the opportunity to make enormous gains along all of those dimensions if we embrace reform.




(p.147) Appendix Table 7.1. Healthcare Reform Policies


2050 Spending Change

2050 Revenue Change

2050 Debt Effect

Reinstate the individual mandate




Appropriate cost sharing reduction payments




Fix the family glitch



Create a public option




Expand Medicaid



Bundle Medicare payments



Install a premium support system



Allow Medicare drug negotiation



Change tax treatment of employer-sponsored health insurance




Increase alcohol tax







Notes: All numbers are referenced as a percentage of GDP. Total may not equal sum due to rounding. Installing the mandate, fixing the family glitch, costs associated with the public option, changing the tax treatment of ESI, and Medicaid expansion all grow with major healthcare costs. Revenue from the public option grows with GDP. Drug negotiation grows with Medicare drug costs. Premium support and bundled payments grow with Medicare costs. Alcohol proposal revenue falls as a share of GDP over time. Several policies mentioned in the text are budget neutral or do not yet have fully developed cost estimates, and thus do not appear in the table. They include reinsurance, reference pricing, reducing administrative costs, malpractice reform, and increasing the use of non-physician practitioners.

Sources: Buettgens, Dubay, and Kenney (2016); Centers for Medicare and Medicaid Services (2018a, 2018b); Congressional Budget Office (2013, 2016, 2017a, 2017b, 2018a, 2018c); Dorn and Buettgens (2017); Hall (2018); Gagnon and Wolfe (2015). (p.148)


(1.) Medicare reforms, such as changing the way providers are compensated, tend to make their way to the private sector relatively quickly (Clemens and Gottlieb 2013).

(3.) Centers for Medicare and Medicaid Services (2018b). Slightly over one-quarter of adults in the United States were “underinsured” as they had large out-of-pocket payments and deductibles relative to their income (Collins, Gunja, and Doty 2017).

(4.) Finegold et al. (2015). Among the uninsured in 2016, about 43 percent were eligible to receive Medicaid or discounted coverage in the marketplace. Approximately 10 percent fall into what is called the coverage gap: they would be eligible to receive Medicaid, but they live in a state that did not expand it under Obamacare and their income is too low to receive a federal subsidy on the exchanges. About 17 percent were offered a plan from their employer but did not accept it, and 11 percent had incomes too high to receive financial assistance for the marketplace. The remaining 20 percent of the uninsured population are undocumented immigrants who are not eligible for coverage under ACA (Garfield et al. 2016).

(5.) Centers for Medicare and Medicaid Services (2018b). Employer-provided health insurance is a bit of a historical accident. When the government instituted wage and price controls during World War II, firms offered health insurance as a way to compete for workers. Employer-sponsored group insurance offers an effective way to pool risks. Because workers are usually not hired based on health status, the set of employees at a firm often provides a stable and average risk profile (controlling for age and sex). This allows insurance companies to offer policies at lower premiums than otherwise.

(6.) For an individual facing a 12 percent federal income tax rate, a 5 percent state income tax, and payroll taxes, an extra $107.65 generates $75.35 in available resources after taxes, so the ability to buy health insurance with pre-tax dollars reduces the price by 30 percent (= 1 – (75.35/107.65)). For an individual earning above the Social Security limit, facing a 35 percent income tax rate, the 3.8 percent Medicare tax and a 7 percent state tax, the ability to pay with pre-tax dollars reduces the cost by about 50 percent.

(8.) This includes people whose employers do not offer health insurance or who are not eligible for their firm’s plan (or do not like the benefits they are offered), are self-employed, or do not work. It also includes so-called Medigap policies that households over 65 purchase to supplement their Medicare coverage.

(9.) In addition, individuals cannot be eligible for other types of health insurance coverage (including affordable coverage from their employer), must file federal income tax returns, and enroll in an insurance plan through a healthcare marketplace.

(10.) Hamel et al. (2016). This includes “Medigap” coverage.

(11.) Individuals under the age of 65 can also receive Medicare if they meet one of the following criteria: received Social Security disability benefits for 24 months, have Lou Gehrig’s disease, or have permanent kidney failure that requires dialysis or a transplant.

(13.) Center on Budget and Policy Priorities (2016). About 10 million elderly and disabled Americans receive both Medicare and Medicaid and are called “dual eligibles.” Medicare coverage provides hospital and physician care, while Medicaid fills in the gaps (such as long-term care and nursing homes). Over two-thirds of dual eligibles are fully dual eligible, meaning that they receive full Medicaid benefits in addition to Medicare. The remainder are partially dual eligible, in that they receive full benefits from Medicare and some cost sharing and premium assistance from Medicaid (Mitchell, Baumrucker, and Herz 2014). Dual eligible individuals tend to have more extensive medical needs than other beneficiaries, leading to disproportionately higher healthcare costs.

(14.) This matching amount is called the Federal Medical Assistance Percentage and is determined by per capita incomes in the state.

(18.) The Veterans Health Administration does not provide health insurance. Rather, it is a government-owned and -run source of care for veterans and active duty service members. This is different from Medicare or Medicaid, where the government reimburses private providers for care they provide to beneficiaries. Veterans pay no premiums or enrollment fees and out-of-pocket costs are low. Half of the VHA beneficiaries are enrolled in Medicare or Medicaid, and many others have supplemental private insurance (Rovner 2014; Congressional Budget Office 2014a).

(22.) Baicker et al. (2013, 2014). A series of papers published by the Oregon Health Study Group details the effects of the Oregon Medicaid expansion experiment. They find that Medicaid coverage did not impact the prevalence or diagnosis of high blood pressure or cholesterol in recipients but did increase the probability of their being diagnosed with diabetes and using diabetes medication. It also reduced their probability of having depression. Having Medicaid essentially eliminated the prevalence of catastrophic medical expenses (over 30 percent of income) for recipients. It also increased their use of preventative care and screening services. Medicaid does not have any impact on employment or earnings, but is associated with increases in SNAP receipts.

(24.) Cohodes et al. (2016) examine how Medicaid expansions in the 1980s and 1990s impact educational attainment. The authors find that a 10-percentage-point increase in a person’s Medicaid eligibility as a child decreases the possibility of being a high school dropout by 4 to 5.9 percent and increases the chances of getting a bachelor’s degree by 2.3 to 3 percent.

(25.) Brown, Kowalski, and Lurie (2015) analyze administrative tax data and find that the government recoups 56 cents for every dollar spent on a child’s Medicaid coverage by the time the child reaches 60. Children who received Medicaid tend to have lower EITC payments, and women have higher cumulative earnings by age 28.

(28.) Center for Medicare and Medicaid Services (2018a, 2018b); Congressional Budget Office (2018a, 2018c). Excess cost growth is the growth rate of healthcare spending per person (after demographic changes are removed) relative to the growth rate of potential GDP per capita. From 1990 to 2014, average annual excess cost growth was 1.2 percent for Medicare, 0.6 percent for Medicaid, and 1.3 percent for other sectors (including private insurance and out-of-pocket payments). For all sectors combined, average annual excess cost growth was 1.2 percent.

(29.) The rise in healthcare spending can be thought of as an example of what is called “Baumol’s cost disease.” This idea traces back to Princeton economists William Baumol and William Bowen, who studied the performing arts in the 1960s and noted that it took as many musicians as much time in the 20th century to perform a Beethoven string quartet as it would have in the 19th century. That is, there has been no productivity growth in this regard. Nevertheless, wages for musicians were higher in the 20th century than in the 19th century—that is, costs in the music industry have risen. Why is that? They argued that increases in productivity and hence wages in some sectors—like manufacturing—drive up wages in other sectors, since otherwise too many workers would leave the sectors with low productivity growth. Healthcare (as well as education and many other public services) exemplifies this trend. Healthcare depends heavily on labor, and the productivity of doctors and nurses has not changed very fast. It takes a similar amount of time to treat a patient now as it did decades ago. While Baumol’s cost disease implies that healthcare costs should be rising over time, it is still the case that reforms could reduce the growth of healthcare spending (Baumol 2013).

(31.) Economists Amitabh Chandra and Jonathan Skinner (2012) distinguish three types of technological innovations: changes that are cost-effective and help nearly everyone (e.g., antibiotics, washing of hands before surgery); changes that are cost-effective for some people but not for others (antidepressants, imaging technology); and changes that are technically feasible but may not be worth the costs or resources used (arthroscopic surgery for osteoarthritis of the knee). They estimate that the third type of spending has played a significant role in the rise of healthcare costs.

(33.) Medicaid can be offered as fee-for-service, managed care, or a combination of the two. Approximately three-quarters of the Medicaid population are enrolled in managed care plans, but fee-for-service incurs the majority of Medicaid costs.

(34.) Cutler (2018); Lallemand (2012). Evidence from the Dartmouth Atlas Project shows that areas of the country with more resources per capita (i.e., more hospital beds, more specialists and physicians) will have higher spending as well. A higher supply of resources leads to a higher provision of care, and thus higher costs, though not necessarily better outcomes. This pattern is especially prevalent in the Medicare system (Fisher et al. 2009; Fisher, Bynum, and Skinner 2009). For example, the variation in resources explains why areas such as Miami, Florida, have average Medicare spending per enrollee of over $16,000 compared to $6,700 in Rochester County, Minnesota, where the world-renowned Mayo (p.142) Clinic is located. And, it explains why an individual who moves from a low-treatment area to high-treatment area can face higher costs for the same procedures. Other studies point to differences such as patient health, consumer preferences, differences in the utilization of types of care, and differences in prices negotiated by different health plans (Cassidy 2014; Finkelstein, Gentzkow, and Williams 2016; Sheiner 2014). Health economist Atul Gawande (2009) provided a well-known example of wasteful spending by observing that in 2005 and 2006, McAllen County, Texas, spent almost double what neighboring El Paso County did on healthcare, even though the counties had similar health and demographic characteristics and the same available treatments and technology. He argued that utilization differences arose because the medical industry in McAllen County was highly fragmented and driven by profit rather than the well-being of patients. Doctors had incentives to recommended more treatments, leading to larger associated costs. They were able to act on this incentive because they were reimbursed for each service rather than on a per-patient basis or on quality of the medical outcomes.

(35.) A chronic health condition is a “physical or mental health condition that lasts more than one year and causes functional restrictions or requires ongoing monitoring or treatment” (Buttorff, Ruder, and Bauman 2017).

(38.) Buttorff, Ruder, and Bauman (2017). Almost half (46 percent) of Medicare spending goes to the 14 percent of enrollees with over five chronic conditions (Lochner et al. 2012). Obesity-related health conditions (e.g., hypertension, type 2 diabetes, and high cholesterol) raise costs by an average of $3,600 per year for each affected person (Cawley and Meyerhoefer 2012). Thorpe et al. (2004) estimated that a quarter of per capita healthcare spending growth between 1987 and 2001 was due to obesity. Hammond and Levine (2010) also note that beyond directly affecting healthcare costs, obesity can also affect the economy through labor market effects such as lost productivity at work, increased absenteeism, premature death, and increased disability payments.

(39.) As noted earlier, precise estimates of administrative costs in healthcare are hard to come by. The Centers for Medicare and Medicaid Services (2018a) reports a measure of administrative costs that rises from about 0.2 percent of GDP in 1960 to 1.4 percent of GDP in 2016.

(40.) These data are for cross-country comparisons and are not directly comparable to the National Health Expenditure numbers presented in this chapter.

(42.) Frakt (2018a, 2018b).

(43.) Papanicolas, Woskie, and Jha (2018). Other countries include Germany, the Netherlands, Switzerland, Canada, Australia, the United Kingdom, Sweden, Denmark, France, and Japan.

(45.) Papanicolas, Woskie, and Jha (2018). The other countries include Japan, Germany, the United Kingdom, France, Canada, Australia, the Netherlands, Sweden, Switzerland, and Denmark.

(50.) Center for Medicare and Medicaid Services (2018a).

(51.) Hartman et al. (2010); Martin et al. (2011); Martin et al. (2012). Sheiner (2015) shows that states with slower per capita income growth during the recession also experienced slower growth in real per capita health services.

(53.) After a couple decades, the projections of various government agencies—CBO, GAO, the Centers for Medicare and Medicaid Services (CMS)—begin to diverge due to differences (p.143) in assumptions about costs, population growth, and other factors. In Chapter 3, I use estimates from the Medicare Boards of Trustees (Boards of Trustees 2018). See Chapter 3 for more details.

(54.) Boards of Trustees (2018). The 3.8 percent net investment income tax is not dedicated to the Medicare trust fund. The supplementary insurance in Part B and the prescription drug benefit in Part D are primarily financed through the Supplemental Medical Insurance (SMI) Trust Fund, which is funded by general revenues and monthly premiums that beneficiaries pay. The SMI Trust Fund is not in immediate danger of running out of money because general revenues are added every year to cover expected costs. The government pays for a pre-determined amount of Medicare Advantage enrollees’ Part A and B benefits using funds from those programs’ dedicated trust funds. If the insurance costs exceed the government contribution, enrollees pay the difference in the form of a monthly premium. If the plan’s costs are lower, the insurer gives the funds back to enrollees through lower premiums or higher benefits.

(55.) “Significant and effective health insurance” means that all Americans have access to a quality insurance plan that provides a minimum set of benefits, as specified in the ACA, at an affordable rate. The opposite of this would be a plan with an extremely high deductible and very little coverage above that level. While the latter would technically be “health insurance,” its prohibitive cost and lack of benefits mean that enrollees would receive little real gain.

(57.) Congressional Budget Office (2015); Furman (2017). The analysis runs from 2016 to 2025.

(58.) Massachusetts saw similar changes when the state converted to a similar system (Kaiser Family Foundation 2012).

(63.) Congressional Budget Office (2017b). In 2017, if an individual chose not to purchase insurance, he or she was charged a monthly fee equal to the greater of 2.5 percent of household income (up to the total yearly premium for the average price of a Bronze plan) or $695 per adult and $347.50 per child in the household, up to $2,085. The fee was adjusted for inflation each year. Individuals were exempt if their income was below the taxable threshold, if they would have to pay over 8.13 percent of their income to purchase insurance, if they experienced certain hardships, if they were members of certain religions, or if they were undocumented immigrants (Healthcare.gov 2017).

(64.) Even when the mandate was in place, the government did not enforce it seriously, and about 40 percent of the people who should have paid a penalty did not (Congressional Budget Office 2014b).

(66.) Brooks (2014). The law states that if one member of a family receives insurance coverage through his or her employer that is deemed an “affordable” share (less than 9.66 percent) of his or her income, then the whole family is ineligible to receive subsidized care on the exchange, even if the price of the employer-provided insurance for the whole family is greater than the affordability threshold. For example, consider an employee who is offered an insurance plan that totals 9 percent of his income. The cost of that plan goes up to 14 percent when including the cost for his family of four. Even though the total cost of the insurance is unaffordable for the family, they would be ineligible for federal subsidies, since the determination of affordability is based only on the cost of insurance for the individual.

(69.) It appears as if risk adjustment is working as intended. In 2016, the program transferred about 11 percent of premium dollars toward plans with less healthy beneficiaries, underscoring its importance in maintaining marketplace stability.

(71.) For example, a reinsurance program could compensate insurance companies for 60 percent of the costs for enrollees who incur over $1 million in costs in a given year, financed by fees on insurers. Such a program would not have any budgetary effects since collections from insurance companies would be directly offset by payments to them, as evidenced by ACA’s initial reinsurance program.

(72.) The proposal for a public option builds on that outlined by the Congressional Budget Office (2013).

(73.) Congressional Budget Office (2013). The presence of a public option may push more people onto the exchanges, including those who previously received insurance from their employers. If employers are less inclined to offer health insurance, taxable wages, and thus income tax revenue, will increase. The downward pressure on premiums from enhanced competition will reduce government expenditures on subsidies.

(74.) Dorn and Buettgens (2017). The authors estimate the potential federal budget effects through 2027 if all 19 states that did not expand Medicaid as of 2017 did so. The vast majority of cost increase is due to increased Medicaid caseloads. This spending is offset by reduced spending on marketplace subsidies (since some people would plausibly move off the exchanges into Medicaid) as well as a small amount of reduced spending on uncompensated care (federal expenditures for healthcare provided to those who are unable or unwilling to pay for care).

(75.) An episode of care is defined on the basis of a major condition or procedure and includes a number of pre-determined services related to this issue over a period of time. For more on bundled payments, see Romley and Ginsburg (2018).

(77.) Porter and Kaplan (2016). In the private sector, the Geisinger health system, an integrated health delivery organization headquartered in Pennsylvania (serving over 1 million patients annually), implemented a bundled payment method under its ProvenCare system. Initial results showed that this method increased the percentage of patients receiving all recommended processes of care from 59 percent to 100 percent and decreased the mean length of hospital stay for certain heart disease patients by 16 percent (Shih, Chen, and Nallamothu 2015). Another interesting development has been the creation of the Prometheus payment system. This payment method generates single, risk-adjusted payments given to providers, plus an additional post-care redistribution to providers based on patient outcomes. The current system has the potential to impact payment for almost 30 percent of the entire adult population, showing that it can be further expanded for an even wider group (Hussey, Ridgely, and Rosenthal 2011).

(78.) The federal payment to insurers would be adjusted for the health of beneficiaries.

(80.) A 2016 Kaiser Family Foundation poll found that among Americans who take a prescription medicine, a quarter have trouble affording the medicine they need (Kirzinger, Wu, and Brodie 2016).

(82.) While manufacturers frequently argue that the price of their drugs reflects the research and development costs involved to create them, evidence shows that this is not true. Rather, drugs tend to be priced based on what the market can bear (Kesselheim, Avorn, and Sarpatwari 2016).

(83.) Patent protections last 20 years beginning from the filing date and are issued by the US Patent and Trademark Office. The Food and Drug Administration (FDA) offers its own form of protection through offers of market exclusivity, which prevents drug competitors from entering the market for a certain period. Currently, small molecule drugs receive (p.145) 5 to 7 years of guaranteed protection, and biologic drugs receive 12 years of protection before generic versions can be sold. Patents and market exclusivity can occur at the same time but do not necessarily have to. Companies can apply to have patents extended to compensate for the time spent in regulatory review (up to 5 years) or clinical trials (up to 14 years), and they receive an additional 6 months of market exclusivity if they test the drugs on children.

(84.) Food and Drug Administration (2015). When two generic manufacturers enter the market, the price of a drug falls on average to 55 percent of the original branded price. It continues to decrease to one-third of the price with the entrance of 5 manufacturers and to 13 percent with 15 manufacturers.

(85.) For most classes of drugs, Medicare is required to cover at least two types. The six classes of drugs for which Medicare is required to cover all or substantially all drugs are immunosupressants (for prophylaxis of organ transplant rejection), antidepressants, antipsychotics, anticonvulsants (for seizures), antiretrovirals (for HIV/AIDS), and antineoplastic (cancer) classes (Centers for Medicare and Medicaid Services 2008). One study found that these drugs account for 17 to 33 percent of total outpatient drug spending in Part D and that these drugs are, on average, priced 10 percent higher than they would be without this requirement (Kipp and Ko 2008).

(86.) Specifically, under Medicaid, drug manufacturers are required to provide rebates on drugs equal to the lesser of 23.1 percent of the average price paid by other buyers or the lowest price paid by other buyers. States are also allowed to further negotiate with drug manufacturers to lower drug prices. The VA and the DoD require that drug manufacturers offer them rebates of 24 percent or the lowest price paid by other non-federal buyers. They can negotiate directly with drug manufacturers, either alone or together, and they can reject drugs from their formularies (Blumenthal and Squires 2016).

(87.) Gagnon and Wolfe (2015). The authors estimate Medicare Part D savings if the program used the same pricing strategy as the VA. They estimate that in 2010, the VA paid 46 percent of official manufacturer prices for drugs and Medicare Part D paid 83 percent (for a total of $36 billion to brand-name manufacturers). Using these data, the authors estimate a ratio of the Medicare to VA discounts to roughly estimate what Medicare would have saved if it had used VA pricing. They find the following: $36 billion—([46/83]*$36 billion) = $16 billion. Frakt, Pizer, and Feldman (2012) obtain very similar results. They estimate that retail VA drug costs are 40 percent of those paid by Medicare, and each beneficiary cost Medicare $1,275 in 2009. If Medicare received an additional 40 percent savings per beneficiary, it would have saved 40 percent*$1,275 = $510 per enrollee each year.

(88.) However, because limiting the deduction would make insurance more expensive, it could raise the number of uninsured individuals by an estimated 1 million (Congressional Budget Office 2016, Health Option 18).

(89.) A 10 percent increase in price reduces cigarette consumption by 3–7 percent and alcohol consumption by 5–11 percent (Chaloupka et al. 1993; Chaloupka et al. 2002; Chaloupka and Warner 2000; Congressional Budget Office 2012; Cook and Durrance 2011; International Agency for Research on Cancer 2011; Gallet and List 2003; Lowry 2014; Roodman 2015).

(90.) Furman (2016); Viscusi (1995). Cigarette consumption creates secondhand smoke, which is detrimental to those exposed to it. It also leads to higher public and private healthcare costs in the short term, since smoking is associated with a host of various diseases and cancers. However, since cigarettes also cause people to die earlier, baseline long-term health costs may be lower than what they would be if the smoking rate were lower.

(92.) Congressional Budget Office (2016). The 2017 tax act temporarily reduced excise taxes on alcohol through 2019.

(93.) To avoid malpractice lawsuits, healthcare providers sometimes use “defensive medicine,” ordering tests and other services that they don’t really think are necessary but can protect (p.146) them against a lawsuit. The standard way to address malpractice is to limit liabilities, but a better way might be to provide a “safe harbor” so that a provider who can show that he or she used best-practice protocol could not be sued (Orszag 2013).

(94.) Rate setting has been experimented with at the state level. Evidence from Maryland shows that state-based rate setting can slow price increases but not necessarily overall cost growth (National Conference of State Legislatures 2017). Under one proposal in California, a new state board would have the authority to regulate the prices of medical procedures that private health plans charge, using Medicare’s rates as a baseline. On the international front, Switzerland and Germany use rate setting in their health systems (Kliff 2018).

(98.) Kaiser Family Foundation (2016). For 2018, the HSA contribution limit was $3,450 for an individual and $6,900 for a family (Internal Revenue Service 2017).