Cato Online Forum

Curtailing Subsidies for Health Insurance

According to many observers, the United States has entered a period of slower than historical growth, otherwise known as secular stagnation.1 The evidence for this view is suggestive but hardly definitive; past periods of slow growth, such as the 1970s, have been followed by periods of faster growth. Similarly, while the reasons offered for secular stagnation are plausible, each is open to reasonable debate.

Since the causes of the alleged slowdown are still up for grabs, it might seem hard to offer policy advice. The right response could differ substantially depending on whether the slowdown, if real, reflects insufficient aggregate demand, reduced technological innovation, or some other cause.

Fortunately, policymakers can improve economic prosperity without determining the causes of the slowdown or even whether it is occurring; this is because the United States has numerous ill-advised policies that ought to be scaled back or eliminated, regardless of current or future growth rates.

A crucial misguided policy is subsidized health insurance for the non-poor. Current U.S. policy does this via Medicare, Obamacare, and the tax exemption for employer-paid health insurance. For brevity, I focus on Medicare and Obamacare, emphasizing three points. These programs are difficult to justify on efficiency grounds; these programs generate substantial inefficiencies in health care markets; and these programs are unaffordable under current parameters.

The Efficiency Arguments for Subsidizing Health Insurance

Economists make two efficiency arguments for subsidizing health insurance.

The first relies on asymmetric information and adverse selection. According to this view, individuals have better information about their own health status than insurers. So if insurers set premiums at the actuarially fair value, unhealthy consumers purchase disproportionately; and their larger than average health costs force insurers out of business. Insurers must therefore set higher premiums. This makes insurance unattractive to relatively healthy consumers, who suffer the loss of not having insurance.

A crucial problem with the adverse selection defense of subsidizing insurance is that the model does not imply any subsidy. Rather, it suggests that government mandate insurance. But since this would force low income households to pay substantial amounts for health insurance, mandated insurance programs include subsidies for low-income households. Those subsidies, however, achieve a distributional rather than an efficiency goal.

A second problem with the adverse selection model-when applied to health insurance-is the assumption that insurers know less about applicant health status than applicants. Insurers can learn substantial amounts about applicants’ health by requiring medical histories and medical exams as a condition of insurance (as life insurers do routinely). Indeed, insurance companies are probably better than individuals at assessing the implications of any given health characteristics for future health expenditure; thus any asymmetry favors insurers, not applicants.

This view is consistent with the standard fear that in a free market for health insurance, insurers would know “too much: about each applicant’s health status (e.g., via genetic tests) and then charge high premiums or refuse coverage for those with pre-existing conditions. Thus some people would face higher insurance costs than others-because they are more expensive to insure-but risk would still be pooled among those with similar average health costs. This outcome is unfortunate for those who face high premiums, but it is not inefficient.

The alternative efficiency argument for subsidizing health insurance is that behind a veil of ignorance, everyone would buy insurance against being born with bad health. Private insurance against such risk, however, might not exist. Government can therefore improve economic welfare by providing such insurance. This is a standard public goods-that is, efficiency-argument; it does not require a distributional motivation and applies even if everyone is (ex ante) identical. Under the assumptions of this model, and setting aside inefficiencies generated by subsidizing insurance, everyone is better off from such subsidies.

The veil-of-ignorance story is reasonable as far as it goes, but the inefficiencies generated by government provision of health insurance are potentially large. This suggests the subsidies should extend only to low-income households, who are the presumably the ones receiving the worst health shocks and most affected by the absence of a market of (pre-born) health insurance. This is exactly what most societies due regarding income insurance more generally; social safety nets protect against the worst-case outcomes but do not attempt to equalize incomes across the entire distribution.

Subsidizing Insurance Generates Inefficiencies

The crucial inefficiency from subsidized insurance is increased moral hazard: people with insurance purchase excessive health care because they do not pay the full cost. Moral hazard occurs with all insurance, but subsidizing insurance makes moral hazard worse. In particular, private insurance uses co-pays, deductibles, and coverage restrictions to limit moral hazard; Medicare and Obamacare make modest use of these features.

The excessive demand for health care that results from moral hazard means higher prices for health care. This translates into faster growing government expenditure, so governments limit reimbursements to health care provides or ration care directly. Providers reply with creative accounting when seeking reimbursements and engage in other non-productive behavior to raise revenues. The limits on reimbursements also discourage innovation and reduce the supply of talented doctors, nurses, and the like.

Beyond these negatives, the taxes required to fund health insurance subsidies generate their own economic distortions; and since these subsidies are large, the tax distortions are substantial.

The Solvency of Medicare and Obamacare

Regardless of how one balances the benefits and costs of subsidizing health insurance, expenditure programs that will bankrupt the country are ill-advised. Yet long-term projections from the Congressional Budget Office suggest that Medicare and Obamacare will do just that. These projections show expenditure on the major health subsidies growing at least 1% faster per year than GDP into the indefinite future. This growth, along with that of other entitlements, implies a debt-to-GDP ratio of roughly 225 percent by 2089 (and ever-expanding growth thereafter).

In recent years, growth in expenditure for the major federal health programs has slowed, suggesting the long-term fiscal imbalance might be not be so bad. Slower health cost growth is desirable, but this suggests only a limited basis for optimism.2 After all, slowdowns have occurred before but reversed later. And in any event, CBO projections that incorporate slower cost growth still show a steeply increasing debt to GDP ratios.3

What to Do?

The U.S. has two paths to avoid the fiscal crisis implied by current health insurance subsidies.

The bigger government approach is to impose more restrictive price and quantity controls in Medicare and Obamacare. This might reduce spending growth, but it will exacerbate the distortions, creative billing, and related inefficiencies caused by these controls.

The smaller government approach is to scale back Medicare and Obamacare.

The most aggressive “reform” would eliminate any federal role in subsidizing health insurance, leaving such policies to the states. Since the federal tax burden would fall dramatically, states could more easily raise taxes to finance these programs. Many states adopted Social Security programs before the federal government, and states routinely adopt more generous policies than mandated by federal law (e.g., higher minimum wages). Thus an extreme race-to-the-bottom, in which states offer little or no health insurance, is unlikely. But since states would fear in-migration caused by overly generous programs, spending would be restrained relative to current federal policy, plausibly yielding a better balance.

A less aggressive “reform” is higher deductibles and co-pays in Medicare and Obamacare. This would increase price sensitivity, slowing cost growth and moderating purchases of medical care. This approach, however, does little to address the growing demographic imbalance, which is one main reason for growth in Medicare.

Other policy changes could nudge the health care system in a good direction, regardless of changes to Medicare or Obamacare: expanded high-skill immigration to increase the supply of doctors and nurses; less restrictive licensing requirements, toward the same end; an FDA that better balances safety against costs and delay; and elimination of the favorable tax treatment for employer-paid health insurance.


The desire to provide universal health insurance is not well-justified on economic efficiency grounds, and the attempt to do so creates huge inefficiencies. Plus, current programs are unaffordable. If not cut back substantially, these programs will do serious harm to the health care system and bankrupt the economy. That is not a recipe for economic growth.


1 See, for example, Cowen (2013), Gordon (2014), Krugman (2013), Lindsey (2013), or Summers (2014).
2 See Chandra, Holmes, and Skinner (2013).
3 See Congressional Budget Office (2014).


Chandra, Amitabh, Jonathan Holmes, and Jonathan Skinner (2013), “Is This Time Different? The Slowdown in Healthcare Spending,” NBER Working Paper no. 19700.

Congressional Budget Office (2014), The 2014 Long-Term Budget Outlook, Washington, DC.

Cowen, Tyler (2013), Average is Over: Powering America Beyond the age of the Great Stagnation, Penguin.

Gordon, Robert J. (2014), “The Demise of U.S. Economic Growth: Restatement, Rebuttal, and Reflections,” NBER Working Paper no. 19895.

Krugman, Paul (2013), “Secular Stagnation, Coalmines, Bubbles, and Larry Summers,”

Lindsey, Brink (2013), “Why Growth is Getting Harder,” Cato Institute Policy Analysis No. 737.

Summers, Lawrence H. (2014), “U.S. Economic Prospects: Secular Stagnation, Hysteresis, and the Zero Lower Bound,” Business Economics, 49(2), 65-73.

The opinions expressed here are solely those of the author and do not necessarily reflect the views of the Cato Institute. This essay was prepared as part of a special Cato online forum on reviving economic growth.

Jeffrey Miron is the director of undergraduate studies in the Department of Economics at Harvard University and the director of economic studies at the Cato Institute.