When the insurance giant American International Group was threatened with collapse in late 2008, its credit default swap business and other international operations were cited as the heart of its troubles. But the largest consequence of AIG's uncontrolled failure on consumers' pocketbooks could have come from the domino-like collapse of its businesses writing insurance on boats, cars, homes, lives, and just about everything else. If these businesses fell apart as a result of AIG's overall collapse, the argument went, the contagion could have brought a collapse of everything from retirement savings plans to auto insurance claims payments from companies unconnected to AIG. (In theory, the operations were firewalled from AIG's other operations, but the extremely slow rate at which they've found buyers indicates that many had significant exposure to the company's other woes.)
The source of the spreading trouble would have been an obscure, dated, and potentially dangerous system intended to make sure that insurance claims are paid even if an insurer becomes insolvent. This system, called state guarantee funds, has almost miraculously remained untested by the present financial crisis, but it poses a major worry for anyone looking to prevent future financial crises, or even a continuation of the current one.
Each of the 50 states runs its own guarantee fund. Any insurance company wanting to write insurance policies in a state must join its guarantee fund. The fund, usually operated by an industry-selected board technically independent of the government, serves as a receiver for insurers that become unable to pay their likely claims. The money to bail out these insolvent insurers comes from forced asset sales but mostly from assessments on the still-solvent insurers in a state. When there are no insolvencies, participation in guarantee associations costs next to nothing. When assessments come due, companies must immediately pay in proportion to their market share. In most states, they can pass on the assessments to policyholders but actually getting the money can take months or years. Although the amounts of individual claims covered by the guarantee funds differ (most are limited to either $300,000 or $500,000), there's no national cap on any insurers' total liability. (Since annuities-private, self-funded pensions-are generally sold as part of life insurance policies, the system also secures vehicles that look a lot more like investments than insurance.) Guarantee funds are obscure even to industry insiders: Most large insurers have policies against even mentioning them in their sales pitches, and several states forbid insurance agents from talking about them unless asked specifically.
The guarantee fund system has two important differences from the similar ones that back bank deposits (the Federal Deposit Insurance Corporation) and investments (the Securities Investor Protection Corporation). First, guarantee funds charge no annual premiums, and thus every penny they spend has to be rapidly raised. Second, unlike the FDIC, guarantee funds have no enforcement authority. Any effort to prevent insolvencies is carried out by state regulators who turn things over to guarantee funds only if they fail to prevent the collapse of a company.
As a result, the system carries an intrinsic danger: Medium-sized insurers-like Vermont Mutual or the Farm Bureau-affiliated insurers in most states-may have large market shares in one state but small or nonexistent operations elsewhere. Paying market-share based assessments for the collapse of a national rival could put them out of business almost immediately. Thus, a large assessment, due immediately, could cause further collapses and start a vicious cycle and bring down even more companies. Even if a regulator, aware of the risk, lets a company "fake it until it makes it" or extends government-backed credit, the private insurance-rating firms like Standard and Poor's and A.M. Best have no incentive to play along, and lenders would likely pull the credit lines that insurers need. The risk that this could happen is not negligible. Currently, one large insurer, the Hartford, has accepted TARP funds to stay afloat.
So far, however, the system has survived. In the past two decades, only two companies ranking in the nation's 100 largest-Florida's Poe National (which collapsed in 2006) and California's Executive Life (which went under in 1991)-have tested the system. Both states saw small and medium-sized insurers flee in the wake of these assessments. In Virginia, likewise, the unexpected collapse of Shenandoah Life Insurance (well-rated by private rating agencies) earlier this year put several small carriers on the ropes.
Two options for reducing the risk of the guarantee fund system have presented themselves.
First, guarantee associations could be pre-funded in a manner similar to the FDIC. New York State has, since the mid-1980s, pre-funded its guarantee associations. The existence of an already-in-place fund greatly decreases the possibility of cascading insolvencies by reducing the need for, and amounts of, sudden assessments. Pre-funding, though, has drawbacks. First, states might raid the funds. In 2006-as several times before-the New York state legislature tried to use the guarantee funds to bail out an ailing workers' compensation fund. Second, the greater costs of pre-funding will almost certainly end up in consumers' insurance bills. And consumers don't want to pay higher rates.
A national guarantee fund with essentially the same structure as the existing state funds is another option. Legislation now pending before Congress proposes one as part of the creation of a federal insurance regulator. The current version of this legislation, the National Insurance Consumer Protection Act, would start a national guarantee fund while still requiring national insurance companies to continue to participate in every state's guarantee fund. Although this would reduce the ability of one company's collapse to have a massive negative impact on large companies or the overall economy, a system of dual guarantee funds could well prove even more destabilizing to small and medium-sized companies than the current system in that it would essentially double their exposure to other companies' insolvencies.
Although no perfect solution exists, the best way to reduce the risks of the guarantee fund system probably lies with trying both ideas: a pre-funded agency for annuity insurance as part of a broader reform of retirement savings and a national fund for everything else.
Pre-funded annuity insurance will almost certainly become a political necessity if the country becomes serious about replacing Social Security. Given the amount of money involved in the collapse of any sizeable annuity provider, the assessments would prove just too much a shock to the insurance system. Such a system would probably have to figure out a way to distinguish between the "insurance" and "investment" components of these annuities and work to provide a partial safety net rather than an absolute "your money is safe" guarantee.
For the automobile, homeowners, and other insurance policies, the national guarantee fund option before Congress would significantly reduce risk. Although the collapse of any enormous insurer would still rattle the system, cascading insolvencies would be very unlikely. The system wouldn't be totally safe, particularly if insurers still had to participate in state guarantee funds, but its enormous assessment base would make it more stable and mitigate a lurking threat to the health of the economy.
Eli Lehrer is a senior fellow at the Competitive Enterprise Institute and director of its Center for Risk, Regulation, and Markets.