Public Insurance and Private Markets

Introduction:

What Role Should the Government Play in Insuring Private Market Risks?

Jeffrey R. Brown

The economic and financial-market events of 2008–2009 should remove any lingering doubt about whether the U.S. government’s role as an insurer of private market activity has consequences for taxpayers and economic efficiency. The government has made explicit its previously implicit guarantee of the government-sponsored enterprises Fannie Mae and Freddie Mac. The Federal Deposit Insurance Corporation (FDIC) has (at least temporarily) increased the insured limits on bank deposits. The U.S. Treasury has agreed to insure money market mutual funds. Government actions that would have been politically inconceivable as late as spring 2008 (such as the federal government taking equity positions in private firms on a massive scale, or using taxpayer funds to bail out troubled mortgage buyers and sellers) are now a reality.

Whether or not these interventions achieve their short-term goal of halting the general decline of the U.S. economy, the consequences of decisions about public versus private risk bearing are now very clear and tangible. The events of 2008–2009 should inform our analysis and policy debates about the appropriate role of government in insuring private market risk for decades to come. While recent interventions have been unprecedented in size, the role of the government (and thus the taxpayer) in insuring financial risks generated in the private market is not a new one. Indeed, even prior to these events, the U.S. government could fairly be called the largest insurance company in the history of the world. After all, the federal government operates the world’s largest pension and disability insurance system (Social Security) and the world’s largest health care insurance programs (Medicare and Medicaid), and it provides explicit or implicit insurance against a broad range of risks, such as pension funding shortfalls, crop losses, and property damage from natural disasters, to name a few. This is on top of the general risk exposure that taxpayers already share through the income tax system.

The government’s role as a risk bearer is typically justified using the theory of market failures. Indeed, the study of market failures is as old as the study of markets themselves. Even Adam Smith, typically viewed as the father of modern economic thought, recognized that markets can occasionally fail to achieve optimal social outcomes. Over the years, the class of market failures, and our understanding of them, has grown. In the twenty-first century, virtually any first-year PhD student in economics worth his or her salt knows how to write a simple model showing that in the presence of market failures, whether they arise from imperfect competition, externalities, asymmetric information, or other sources, an omniscient and benevolent government may improve social welfare by mandating, regulating, or incentivizing optimal behavior.

Of course, governments—even democratically elected ones—are not consistently omniscient or benevolent. Government policies, as well as private markets, can and sometimes do fail to achieve socially desirable outcomes. Well-intentioned policies often have unintended consequences that mitigate their efficacy or solve one set of problems at the expense of creating new ones. In the words of Milton Friedman, The government solution to a problem is usually as bad as the problem.

All too often, our policy debates reduce to ideological grandstanding. Many free-market advocates tend to ignore the possibility of real market failures. Many critics of markets believe that market failures are ubiquitous and that only the government has the power to set the world right again. A talented economic theorist can construct formal models justifying either position. To truly understand whether government intervention in markets—in this case, insurance markets—is harmful or beneficial on net, it is necessary to go beyond general ideological statements and assess the facts on the ground. For better or for worse, there are numerous U.S. government interventions in insurance markets available for study. This book looks at six of these interventions in an attempt to assess whether they are truly needed and how they might be improved.

An Overview of the Volume

The goal of this book is to bring economic theory and evidence to bear on existing U.S. government policy in six areas of public and private risk bearing, and to suggest policy options to improve the efficiency of public policy in these areas. These studies generally recognize that private markets are capable of failing to function efficiently. However, they are also forceful and expansive in their criticism of the poorly designed government programs in place.

In chapter 1, George Pennacchi examines the structure of the Federal Deposit Insurance Corporation. He points to the positive aspects of deposit insurance—namely, the protection of small, unsophisticated investors and the enhancement of overall financial-market stability through the prevention of bank runs. However, he also discusses the important economic distortions caused by politically induced limits on the FDIC’s ability to price deposit insurance accurately—based on the underlying risks. He finds that the FDIC’s attempt to set deposit insurance premiums to target deposit insurance fund reserves leads to excessive systematic risk taking by banks. He also discusses the limited progress toward combating the Too Big to Fail (TBTF) incentive for bailing out large banks.

Pennacchi considers several reform options that would combat the distortions created by the current FDIC system and help improve efficiency in the banking system. The first would centralize the clearing process for derivative contracts as a way of reducing counterparty risk. This reform would make it harder to assert the TBTF rationale for bailing out distressed banks, because a common justification for TBTF assistance (used in the case of Bear Stearns, AIG, Wachovia, and CitiGroup) is that counterparty risk in derivatives contracts serves to interconnect institutions, and the failure of one would trigger losses and potential failures at others. An advantage of a clearinghouse is that it allows parties to net their positions with their many counterparties, thus providing more expertise and cost efficiencies in setting margin requirements. The second reform would emphasize charging appropriate risk-adjusting premiums to banks by using more market information and possibly collateralization of deposits. The third reform would either abolish the Deposit Insurance Fund (DIF) or use market mechanisms such as swaps to transfer the risk of targeting DIF reserves to investors outside of the banking industry.

In chapter 2, Andrew Biggs and I analyze the Pension Benefit Guaranty Corporation (PBGC), an agency of the federal government that insures private-sector defined-benefit pension plans against the bankruptcy of the sponsoring employer. We find that the PBGC does not adequately risk-adjust premiums and so encourages risk taking by plan sponsors, that it fails to ensure adequate funding, that it does not provide timely and appropriate information to pension plan participants or financial markets, and that it is likely to impose substantial costs on future taxpayers. We also highlight how the PBGC’s 2008 decision to increase equity exposure is precisely the wrong medicine for curing the funding shortfalls facing the agency.

Looking to the future, we discuss both incremental and wholesale reforms aimed at increasing market discipline in the program. At the top of the list: Congress should authorize the PBGC to risk-adjust the premiums that it charges to plan sponsors. There is no good economic justification for charging the same premium per participant to a very healthy plan sponsor with an overfunded pension plan invested in a portfolio that immunizes the plan against changes to its funding status, and to a financially distressed firm invested in a risky portfolio of assets that could worsen its funding status just when the firm is likely to dump its pension liabilities on the PBGC. Other incremental reforms include allowing the PBGC to use private reinsurance or securitization to reduce some of its exposure and increase pricing transparency; requiring that pension promises be given seniority over other liabilities in bankruptcy proceedings; or further limiting PBGC coverage. We also discuss the more far-reaching reform possibility of eliminating the PBGC and replacing it with a government-mandated, but privately run, pension insurance system, though we acknowledge that doing so would not eliminate the large implicit debt overhang that already exists in the system.

In chapter 3, Dwight Jaffee and Thomas Russell discuss the market for insurance against losses arising from terrorist attacks. Explicit federal involvement in this market emerged in the aftermath of the September 11, 2001, attacks, when insurers realized that their insurance pricing models were seriously inadequate and when foreign reinsurers withdrew from this segment of the property and casualty insurance market. Thus, a purportedly temporary government insurance program was created in 2002 to provide the industry with a backstop against large losses. Perhaps unsurprisingly, this temporary program has already been extended twice. Because it is virtually impossible for a fully functional private reinsurance market to emerge when the government is giving away free reinsurance, it will always be possible for proponents of government intervention to justify further extensions by pointing to the lack of a private market alternative. The authors discuss the important resource allocation costs that occur when subsidies prevent individuals and firms from bearing the true economic costs of their decisions. Supporters of the federal program often argue that the resource allocation costs are limited in this context, because a policyholder is far removed from a terrorist’s decision about whether to attack the U.S. Such a view, however, ignores the myriad of decisions that can influence the likelihood that an attack succeeds (e.g., how much to spend on security) or the damages that result (e.g., whether to upgrade sprinkler systems). Like it or not, such decisions will be affected in aggregate by the marginal cost of investing in these activities relative to the cost of insuring against the loss.

Jaffee and Russell note that the current federal terrorism insurance program covers losses from conventional attacks, but excludes chemical, nuclear, biological, and radiological (CNBR) attacks. The authors argue that the magnitude of losses from a conventional attack could be absorbed by private markets, but that the CNBR losses are of a magnitude large enough to warrant an explicit government role. The authors recommend that, when the current legislation ends in 2014, responsibility for conventional acts of terrorism be returned to the private market, with no government subsidy, and that the government shift its focus to covering CNBR risk, although not simply by using the existing terrorism insurance model; the authors discuss a range of ways that the government can partner with private investors to help recapitalize private insurers in the aftermath of a large CNBR event. In the event that the government does not see fit to exit the market for conventional terrorism losses, the insurance should at least be provided at an appropriate, risk-adjusted cost to the insured businesses.

Kunreuther and Michel-Kerjan analyze the tremendous growth in financial losses arising from natural catastrophes in recent decades, which they attribute to a combination of increasing urbanization, increased concentration of economic value in risk-prone areas, and possible changes in weather patterns. Looking forward, they suggest that the growth in future losses is likely to continue to increase, posing important market and public policy changes.

Although the federal government has programs in place to deal with other types of risks discussed in this book, it has no comprehensive program to insure against catastrophe ex ante. (One exception—flood insurance—is discussed in detail in chapter 5.) As such, rather than analyzing an existing program, the authors lay out a number of principles for how any future intervention into catastrophe insurance markets should be structured; they call specifically for premiums that reflect risk, the use of grants (rather than price subsidies) to deal with equity and affordability issues, taking steps to ensure sufficient insurance demand, and appropriate solvency regulation.

The authors acknowledge that individuals often fail to invest in effective mitigation activities. Rather than using this fact to justify the direct provision of insurance by the federal government, however, Kunreuther and Michel-Kerjan propose the creation of a new financial product. Specifically, they suggest moving from the standard one-year insurance contract for homeowners to long-term insurance contracts, which would provide financial incentives for individuals to undertake efficient mitigation efforts. They propose flood insurance as a natural candidate for implementing this idea. If a long-term (e.g., twenty-year) flood insurance policy were tied to a particular property, then the homeowner could take out a twenty-year home improvement loan, invest in risk mitigation, and achieve a net increase in cash flow for cost-effective mitigation efforts through the resulting reduction in the flood insurance premium.

In chapter 5, Mark Browne and Martin Halek discuss the National Flood Insurance Program (NFIP) in detail and explore a range of reform options. They trace the origins of the modern-day NFIP from the government’s approaches to flood control in the 1800s through the policies of today. They point out the flaws of the existing program, notably the mispricing of insurance, and show that many flood insurance policies are underpriced relative to the risk they insure, with some policies receiving a 60–65 percent subsidization rate. (Indeed, the extent of mispricing is even worse than these figures suggest, due to grandfathering of rates and other design flaws.) Given its inadequate pricing model, it is not surprising that the NFIP is underfunded on an annual basis.

The authors explore various routes by which the efficient transfer of flood risk could be accomplished through private-market alternatives, including the use of private reinsurance, alternative risk-transfer mechanisms (such as catastrophe options, industry loss warranties, or contingent capital agreements), and, most directly, private insurance provision.

In chapter 6, Vincent Smith and Barry Goodwin examine the federal government’s role in providing insurance against agricultural losses. The authors discuss how federal crop insurance programs have been driven by political considerations from the beginning, when President Roosevelt proposed the creation of multiple-peril programs in 1936. Over the past three decades, the programs have been expanded to offer subsidized coverage for more crops (and, recently, even some livestock); the result is a wider range of available insurance products, and an increase in subsidies—the taxpayer now provides approximately 60 percent of the actuarially fair premium. The authors explain that Congress’s primary rationale for expansion was to increase participation enough so that ad hoc disaster programs would be unnecessary. However, even though participation in federal crop insurance has increased to include over 80 percent of eligible crop acreage, Congress nonetheless included a disaster relief program in the 2008 farm bill while maintaining subsidy levels.

The authors find little evidence of a market failure that would justify government intervention in this market. Recognizing the political reality that federal intervention in this market is unlikely to go away, however, they offer a number of principles to guide how a government program should be designed and implemented, such as limiting moral hazard through controls on coverage levels and using a rational pricing mechanism that limits the distortions to production patterns. They also discuss some of the changes that would be required to bring the program into compliance with world trade agreements.

Lessons Learned

Taken as a whole, these six studies provide a number of both constructive and cautionary lessons for policymakers considering new government insurance programs or reforms to existing ones. While the details of each particular market matter, a number of general themes emerge from the analyses.

Government Intervention Must Be Justified. One of the themes that emerges involves the economic justification for government intervention in insurance markets. The mere existence of a market failure does not necessarily mean that the government is in a position to overcome this failure, even on theoretical grounds. For example, one common source of market failure in insurance markets is moral hazard, the well-known phenomenon in which individuals engage in riskier behavior when they know that they are insured against the downside loss. Having the government rather than the private sector provide insurance does nothing to overcome this source of inefficiency. Indeed, given the government’s apparent inability to properly price insurance, government provision may exacerbate this problem. Of course, even if the government is able to overcome the source of the market failure in theory, a discerning analyst must also question whether a government insurance program will be able to improve the situation in practice, given the political nature of the policy and regulatory process.

There are three potential justifications for government intervention that arise across the multiple insurance markets analyzed in this book. The first is adverse selection, a type of asymmetric information in which the potential insurance purchaser knows more about his risk than the insurance seller, and in which it is possible that only the more risky individuals will purchase the insurance. In theory, adverse selection can cause a voluntary market to completely unravel (Akerlof 1970; Rothschild and Stiglitz 1976). To combat adverse selection, the government has one tool in its arsenal that no private insurer has—the power of compulsion. Specifically, the government can mandate that all individuals or firms participate in a particular market, thus removing self-selection from the equation. For example, the government can require all sponsors of defined-benefit pension plans to purchase insurance, or all homeowners to insure their residences against floods. As has been frequently noted in the economics literature, however, just because the government may have an important role in mandating insurance coverage, that does not mean that the government must be the one to provide the insurance. A government-mandated but privately run insurance program may be the most efficient answer.

A second rationale that frequently appears in discussions of government insurance programs is known as the Samaritan’s Dilemma (Buchanan 1975). The Samaritan’s Dilemma is quite simple: it arises when individuals can increase future benefits from an altruistic government (or altruistic individuals) by behaving in a manner that is socially inefficient. After a loss, be it one arising from a terrorist attack, a hurricane, or a widespread failure of pension plans, it is politically difficult for a democratically elected government to do nothing. There will be political pressure for the government to step in as the Good Samaritan to offer some form of aid (whether in the form of disaster relief, victim compensation funds, financial bailouts, or any number of other policies). If there is a widespread ex ante belief that such relief will be forthcoming after a bad event, then individuals will behave as if they are insured. This is essentially a form of moral hazard, but one due to a political failure (i.e., the inability to credibly commit to not intervene) rather than a market failure. In light of this, it is often suggested that having an explicit, well-defined, ex ante government role makes it easier for politicians to resist ex post relief. If the ex ante insurance program that provides appropriate incentives for risk mitigation is more efficient than ex post relief, then an explicit government program may actually reduce the taxpayers’ ultimate exposure.

Of course, using the Samaritan’s Dilemma to justify intervention presupposes that intervention will truly lead to a more efficient allocation of resources even after accounting for all behavioral responses of the affected population and the policymakers. Given that the ex ante government programs are often designed inefficiently themselves, and that politicians may still step in with ex post assistance even in the presence of a preexisting insurance program, the jury is still out on whether the evidence supports this justification for a government role. Even if it is determined that an ex ante program is desirable, one still must choose whether the government should directly provide the insurance, or whether a minimalist approach of ensuring that the private market operates effectively (such as by mandating participation and/or regulating the insurers) would be sufficient.

A third rationale that economists often use to justify a role for government is that the government has the unique ability to diversify risk across generations through fiscal policy. Some risks are large and systematic, meaning that they cannot be diversified away simply by insuring more individuals or firms. Furthermore, in some cases, effectively insuring a risk can influence the extent to which it becomes systematic: for example, one rationale for deposit insurance is to prevent financial distress at one bank from causing negative externalities on other participants in the financial system.

While the government has no advantage over the private sector in diversifying risk within a given generation, it is uniquely situated to contract across generations. As noted by Bohn (2006, 11), Welfare improvements are possible because a government’s power of taxation gives it a unique ability to make commitments on behalf of future generations. However, there are at least three reasons to observe caution when using the intergenerational risk-sharing argument to justify a government insurance program. First, while intergenerational risk sharing has the theoretical potential to spread risk more efficiently, it does not eliminate the risk. As the Congressional Budget Office (2004, 5) has noted: Risk is not reduced by the government’s power to print money, because financing credit losses by creating money substitutes an inflation tax for a pecuniary tax. In the end, someone must bear the consequences of unpredictable financial returns, and markets determine a price for assuming that risk. Second, as Bohn (2006, 11) points out, The government’s power to oblige future generations also creates potential for abuse. Risks might be shifted haphazardly onto future generations by governments catering to current voters. Third, as noted by Brown and Orszag (2006), the possible advantage of spreading risk more efficiently across generations must be weighed against the fact that fiscal policy may not spread risk efficiently within each generation. Markets use price signals to spread risk efficiently so that it is borne by those most willing to bear it. Fiscal policy spreads risk based on a host of political factors that may not bear any relation to the preferences of those chosen to bear the risk.

Other arguments in favor of government intervention in insurance markets are less convincing. For example, it is often argued that parameter uncertainty, particularly when there is the possibility of rare events, makes it difficult for the private market to function. In the immediate aftermath of the 9/11 attacks, insurers expressed concern about their inability to accurately model the risk distribution of terrorism losses. Especially at the time, no one knew whether 9/11 was a rare event unlikely to repeat itself in our lifetimes, or whether the world had entered a new era when large-scale attacks would be more common.

While it is true that parameter uncertainty makes insurance pricing more difficult, the government rarely has a comparative advantage in overcoming this uncertainty. Even in the unique case