Reforming insurance to support workers' rights to compensation


  • Martha T. McCluskey

    Corresponding author
    1. SUNY Buffalo Law School, State University of New York, J.S.D., LL.M. Columbia Law School, J.D. Yale Law School, O'Brian Hall North Campus, Buffalo, New York
    • O'Brian Hall North Campus, Buffalo, NY 14260-1100.
    Search for more papers by this author

  • Additional disclosure: The author worked as an attorney representing ratepayers (employers) for the Maine Public Advocate Office from 1989 to 1992 in some of the insurance rate proceedings discussed in this article.


The structure and regulation of the insurance system for financing workers' compensation affects the costs of workers' benefits. Using the example of Maine's insurance market restructuring in response to a crisis of the early 1990s, this commentary explores how changes in insurance regulation might better support the goals of workers' compensation. The commentary analyzes how insurance and its regulation should go beyond correct pricing of risks to questions of how to structure incentives for loss control to include workers' interests as well as the interests of employers and insurers. Am. J. Ind. Med. 55:545–559, 2012. © 2012 Wiley Periodicals, Inc.


After a century, workers' compensation's promise of reliable support for injured workers is widely accepted in theory but frequently evaded in practice. Linking workers' compensation to international human rights offers some potential for enhancing the moral ground for this promise in an economic and political climate in the United States generally not conducive to labor protections and social welfare programs (Compa and Hilgert, 2012). Yet in the prevailing political and economic framework, efforts to turn economic protections into fundamental rights are likely to backfire, hurting the intended beneficiaries of those rights. In the conventional wisdom, if we expand rights to compensation for injured workers, we are likely to raise costs for employers and insurers, with the result that workers could end up worse off due to fewer jobs, lower wages, or less secure funding for compensation systems.

While public policy should take seriously the question of complex consequences of expanded economic rights, this argument simplifies and distorts the analysis. All rights produce costs, but all costs also depend on a web of existing rights. The tough economic tradeoffs posed by expanding workers' rights are not natural and inevitable, but instead are created and sustained by regulatory frameworks and background legal rules likely to be hidden from view. To evaluate the possibilities for expanding economic protection for injured workers, it is important to examine the underlying legal and economic structures that often appear to make strong protections for workers too costly.

In workers' compensation, insurance regulation is a crucial background piece missing from the standard picture of a tough tradeoff between jobs and benefits. The conventional view portrays insurance largely as a passive messenger, charging employers prices that reflect the costs of a given system of workers' benefits plus the costs of administrating and financing those benefits. To the contrary, insurance plays an active role in creating the costs of benefits and in shaping the system's impact on workers, employers, and society in general.

Inadequate Insurance Structures Undermine Benefit Expansions

Workers' compensation developed in the early 20th century U.S., replacing workers' civil right to sue employers for tort damages with employer-provided insurance against work-related injuries and illnesses. Each state operates its own workers' compensation program, with benefits and claims procedures governed by individual state law without federal standards. The federal government also operates a few separate programs limited to specific groups of workers, such as federal employees and longshore workers [Sengupta et al., 2010, App. H]. All but one state requires employers to participate in workers' compensation programs, subject to specific rules about which employers and workers are covered, when injuries and illnesses count as work-related, and the terms and levels of benefits provided to workers. In general, workers who are disabled due to their employment can claim cash benefits, medical benefits, and rehabilitation services through a no-fault claims process. In practice, workers' compensation programs from the start left serious gaps in coverage and access, and have often provided meager benefits, particularly for long-term disabilities and occupational illnesses.

In 1972, a bipartisan commission established under the federal Occupational Safety and Health Act of 1970 unanimously found state workers' compensation systems were inadequate and inequitable, and as a result recommended comprehensive benefit expansions [National Commission on State Workmen's Compensation Laws, 1972]. Over the next decade, states responded with legislation and other legal changes improving benefits and coverage, but these reforms fell short of the Commission's recommendations [Burton, 1994]. By the late 1980s, hopes for further substantial gains in benefits had been dampened if not extinguished by steeply rising insurance prices charged to employers in many states [McCluskey, 1998].

State law determines the insurance structures that finance and deliver workers' compensation benefits. Through the 1960s, only seven states provided insurance exclusively through a state fund, without participation from private insurance carriers [Thomason et al., 2001]; that number declined to four by 2008 [Sengupta et al., 2010]. From 1933 to 1984, 11 to 12 states relied on a dual system where employers obtained insurance coverage either from private insurance carriers or from a state fund [Thomason et al., 2001]. Beginning in the mid-1980s, however, more states adopted a public or quasi-public insurance fund, leading to a total of 21 states now relying on this combination of public and private insurance sources [Sengupta et al., 2010].

Throughout most of the 20th century, therefore, the majority of states depended solely on the private insurance market to supply workers' compensation coverage. Private insurance continues to be the largest benefit payment source, accounting for 53.2% of benefits in 2008 [Sengupta et al., 2010]. Almost all states also permit large employers to self-insure (subject to solvency requirements), and some states have extended this option by allowing some employer groups to self-insure. Self-insurance typically involves re-insurance arrangements with private carriers and may be managed by outside firms specializing in third-party administration.

For much of the 20th century, employers purchasing private insurance typically paid rates established through an “administered pricing” system, rather than through market competition. Insurance rates were subject to prior approval by state regulators, based on estimates of benefit costs and related expenses submitted by a private rating organization on behalf of private insurers [Thomason et al., 2001]. Insurers are permitted to act collectively to compile comprehensive data on benefit costs and to standardize reporting and risk classifications. In the longstanding system for rate regulation, all insurers in each state generally charged the same rates for the same risk classifications, so that price competition was largely limited to dividends returned to policyholders after the policy period ended [Thomason et al., 2001]. Risk classifications are designed to reflect the differences in benefit costs (loss risks) among different industry groups. Further, individual employers (of sufficiently large size and duration) could be subject to “experience rating,” whereby the general rate for their industry was modified to reflect the particular past benefit costs of that employer.

Because workers' compensation coverage is generally mandatory, some source of insurance must be designed to take all comers, a role typically served by public insurance funds in states where such funds are either an optional source or the exclusive source of benefit financing. States relying on the private insurance markets through the mid 1980s needed a special “residual market” mechanism to cover employers who were rejected by private “voluntary market” insurers for reasons such as small size, high risk relative to others in the industry, or lack of experience [Thomason et al., 2001] (factors likely to also make self-insurance inappropriate). Typically these employers were relegated to an “assigned risk” pool operated collectively by private insurers, subject to a rate surcharge to reflect the theoretically higher risk and expenses associated with this business and with any net losses spread among the insurers operating in the state. If the number of employers in this “involuntary” pool remained relatively small, the risk of loss generally would be minimal to any one insurer, especially if insurers could expect generous profits in the voluntary market.

For much of the last century, state regulators overseeing the private workers' compensation insurance market generally deferred to private insurers' rating organizations on pricing and underlying cost estimates, approving rate requests in non-adversarial proceedings with little scrutiny or public participation. As benefit costs and insurance premium prices rose in the 1970s and 1980s, however, regulators and outside critics began to more closely question insurers' rate requests and particularly to consider the importance of investment income to predicted profitability [McCluskey, 2001]. But without easy access to rate increases, in a context of rising uncertainty about long-term costs due after a period of expanding benefits, private insurers were likely to protect their profits by trimming their more marginal “voluntary market” business. From the mid-1980s through the early 1990s, in many states insurers relegated increasing numbers of employers to the residual “assigned risk” pool [Thomason et al., 2001], threatening to unravel the insurance system in the majority of states where the private insurance industry operated the pool and assumed its net losses. As these pools grew, even with substantial profit opportunities in the “voluntary market,” insurers confronted an increasing and highly uncertain risk from the long-term losses from those pools, inducing them to further narrow their business to the most profitable employers. The exodus from the voluntary market helped pressure state regulators to grant a burst of insurance rate increases from the mid 1980s through the early 1990s [Burton, 1994], but the size of the residual market expanded to cover more than one-fourth of workers' compensation insurance premium volume nationwide by 1992 [Thomason et al., 2001]. In several states, the “assigned risk” pool grew to insure almost the entire market [Thomason et al., 2001]. Under these conditions, even massive voluntary market price hikes would likely be insufficient to protect insurers against severe residual market losses [McCluskey, 2001]. As a result, regardless of rate increases, a number of states faced a “death spiral” where a growing “residual” market pool de-stabilized the system and threatened to push private insurers out of the market altogether [McCluskey, 2001].

Faced with a crisis in insurance supply along with rising costs due to these expanding residual market pools, states without public funds generally succumbed to strong political pressure for restrictions on benefits and on benefit claims processes in the late 1980s and 1990s [McCluskey, 1998]. Business groups and news media targeted workers' compensation costs as a significant economic problem, as insurance rate hikes and higher residual market costs combined with general economic pressures to generate political concern about the effect on employers. Despite general failure to achieve the 1972 commission's recommended coverage and benefit goals [Whittington, 2004], by the 1990s even many lawmakers and experts sympathetic to injured workers concluded that benefits had become unsustainably generous. Further, insurers and employers often promoted the theory that rising costs and declining insurance supplies were the result of a culture of fraud and malingering among workers induced by benefit expansions [McCluskey, 2003]. To some extent, and in some circumstances, employers may be able to alleviate rising workers' compensation insurance costs by reducing workers' wages or other benefits or by raising consumer prices [Burton, 2009]. In practice, the state-based nature of the system has tended to direct political attention to the possibility that employers will respond to high insurance costs by moving the business to lower-costs states (or to lower-cost foreign countries), creating the conditions for states to compete in a benefit-cutting “race to the bottom” [Graetz and Mashaw, 1999]. Though states with public funds had protection from the insurance supply crisis, they generally shared the problem that expanding benefits, rising insurance costs, and general economic pressures (especially affecting high risk, high cost industries) had made insurance costs highly uncertain and politically problematic. Some state funds invoked political control to reduce or control employers' rates without effective cost control or attention to solvency, resulting in insurance fund deficits followed by extensive benefit cuts [Spieler, 1993; Thomason et al., 2001].

Reorienting Insurance From Benefit Pricing to Workers' Protection

These 20th century workers' compensation insurance systems were not well designed to sustain the promising push toward benefit expansions in the 1970s. But contrary to conventional wisdom, the resulting insurance problems from benefit expansions were not inevitable. Instead, movements for benefit expansions must confront the inadequacy not only of workers' benefits but also of the traditional understanding of workers' compensation insurance as a largely passive delivery system focused on the highly technical problem of correctly pricing benefits. Insurance systems have the capacity to either lead or impede the goal of workers' protection, because these systems actively shape and regulate the competing interests involved in workers' compensation.

The effects of insurance on the economic feasibility of a given benefit system go beyond the standard technical controversies about how to accurately measure, predict and price the risks of injury and illness. Insurance pricing issues such as insolvency protections and excessive profits by private insurers are important, but not sufficient, factors in protecting workers' access to adequate benefits. More fundamentally, insurance regulation should focus on how insurance affects incentives for injury reporting, claims processing, safety and reemployment, all of which can produce substantial variation in the costs of a given set of statutory benefit rights.

At the core of insurance is an incentive structure whereby an insurer tends to gain more the less it pays out in compensation claims and the less it pays out in services to claimants and employers. A parallel incentive structure operates at the level of employers, who may face some higher costs (especially through experience rating) to the extent their own employees claim benefits and who face generally higher costs as overall insurance rates reflect general benefit expansions. In the existing system these incentives are a powerful force working against injured workers' interests in obtaining, and improving compensation for injuries and illnesses.

In the conventional theory of workers' compensation, accurate pricing addresses this incentive problem, balancing the interests of workers, employers, and insurers by inducing all three groups to reduce claims through safety, rehabilitation and re-employment. But insurance prices that rise with rising claims costs instead tend to induce insurers, along with employers, to shift the costs of higher risks to workers—by discouraging, resisting, or penalizing compensation claims. These loss-shifting efforts can extend beyond suppression of legitimate individual benefit claims to the level of public policy decisions, where insurers have been leaders in actively shaping the design, enforcement, and reform of workers' compensation laws to control workers' access to benefits. Furthermore, insurance carriers and employers often have tried to protect their interests in reducing claims costs by promoting public perceptions of injured and ill workers as undeserving, diverting public attention from workplace health and safety problems and stigmatizing workers' compensation claimants and their advocates.

Accurate risk pricing therefore cannot be the primary solution to balancing the potentially competing interests of workers, employers, and insurers. A sustainable, equitable, and affordable system will be better achieved by changing direct control of insurance services and management to put workers' interests in the forefront. Workers must share in the power to determine insurance questions such as the production of information about risks and costs, the design of loss control efforts, financial management, and claims administration.

To a significant extent, many states alleviated the late 20th century crisis in insurance supply not only by restricting benefits and raising prices, but also by shifting control and design of insurance to better respond to employers' interests. In an era generally promoting privatization, it is notable that in the 1990s eight states established a new public or quasi-public source of insurance to compete with private insurers [Thomason et al., 2001]. These state funds typically operate as a guaranteed insurance source especially for covering the smaller businesses more likely to be excluded from the private market. In addition, rate regulation systems in many states were changed in the 1980s and 1990s to allow for varying degrees of increased competition among insurers, combining some easing of regulatory prior approval with some reduced power of private rating organizations to determine final rates charged after submission of loss cost data [Thomason et al., 2001]. Although this de-regulation focused competition on price, risking problematic incentives for claims suppression along with underfunding and market instability, this new regulatory approach also opened up possibilities for some employers to have more market choices and drew attention to insurance management in addition to benefit levels as a factor in insurance costs. Further, these state regulatory changes also tended to open up more opportunities for pricing based not just on benefit costs but directly on employers' loss control actions, for example by “schedule rating” programs that tied discounts to adoption of specific safety measures. Though these changes may sometimes have helped protect workers' interests—especially to the extent they created more stable and solvent financing and alleviated political pressure for further benefit cuts—these changes generally have continued to leave workers at the margins of insurance system governance.

Insurance systems can go much further to protect the interests and dignity of workers in both financing and controlling benefit costs, making meaningful protection from harm and meaningful rehabilitation the primary goal, in place of claims reduction. Of course, insurance changes need to be combined with other reforms to improve protection of workers' health, safety, and income against the risk of work injuries and illnesses. The background legal rules that help create incentives to weaken workers' compensation go beyond insurance regulation to include laws governing benefits and workers' control over workplaces, including government regulation of workplace health and safety and labor and employment law. Nonetheless, these other potential sources of legal protection are likely to be undermined by the political, economic and cultural power of a workers' compensation insurance system shaped with incentives to restrict compensation for injured workers.

In short, the effort to affirm workers' compensation as a fundamental moral principle raises questions about insurance that are too often treated as a technical matters best left to experts. The policy debate should not simply focus on the relative merits of public versus private insurance, or on how stringently to regulate insurance pricing. Instead, renewed attention to the goals of workers' compensation requires a new vision of the function of insurance, going beyond an actuarial model focused on predicting externally driven costs to a more explicitly regulatory model focused on affirmatively and sustainably reducing costs for workers.


The state of Maine's response to its crisis of high workers' compensation costs in the 1980s and early 1990s suggests some possibilities for re-imagining insurance. This article does not attempt to answer the question whether Maine's insurance changes have actually improved health and safety or injury compensation for workers, and indeed it presumes that Maine is not likely to be exempt from the general concerns about shortcomings of occupational health and safety and compensation for injured workers in the U.S. More narrowly, this essay examines how Maine's experience may challenge the view that controlling workers' compensation costs is primarily a matter of controlling benefit levels. Like most states, Maine enacted legislation restricting benefits in the 1980s and 1990s in response to rising insurance costs [McCluskey, 1998]. Nonetheless, that state's insurance restructuring was crucial to reducing employers' costs while maintaining a stable and solvent insurance supply.

The Standard Story of Maine's Crisis: Control Benefits, Not Insurance Pricing

In the conventional view of workers' compensation insurance, Maine became the “most egregious” violator of sound insurance principles when state regulators repeatedly denied or reduced private insurers' requested rate increases during the 1980s, despite rising benefit costs [Kerr, 1992]. This view explains Maine's actions as an example of “rate suppression” that unsuccessfully attempted to scapegoat insurance companies to avoid the tough policy tradeoffs between workers' benefits and employers' costs. In a 1992 op-ed essay for the New York Times, the chief executive of American International Group (a major Maine workers' compensation insurer) charged the state's insurance regulators with substituting politics for economics, using insurers to “artificially” cover the “true costs” of a system that he described as bloated by benefit fraud [Greenberg, 1992].

Consistent with this story of excessive price controls making workers' compensation insurance unprofitable, Maine's private insurers increasingly refused to insure the more costly employers. Because workers' compensation is mandatory, as employers were rejected from this regular “voluntary” insurance market, they had to obtain coverage from the “residual market” pool as the insurer of last resort. Maine, like most states at the time, relied on the National Council on Compensation Insurance (NCCI, comprised of private workers' compensation insurers) to operate its residual market pool, which spread its risk among the state's private insurers [Burton, 1994]. In Maine, as in many other states during the late 1980s and early 1990s, the projected and highly uncertain costs of the residual market pool threatened to outstrip revenue, despite typically higher premium rates. Without sufficient profit to offset these losses, private insurers threatened to withdraw from the state. In an attempt to sustain this private insurance market, in 1987 state law reforms restricted workers' benefits, partly protected insurers from the growing residual market liability, and adopted regulatory changes that led to a series of substantial rate increases [McCluskey, 2001]. Despite these reforms, insurers did not significantly increase their voluntary market coverage in Maine, and in 1992 private carriers took steps to withdraw from the state entirely.

The private insurers returned to the state after 1992 legislation further limited benefits, deregulated prices, and bailed out the residual market deficit to alleviate insurers' exposure to losses from past insurance coverage. The rate suppression story concludes by explaining that the impending insurance market collapse and the growing deficit created the political pressure needed to shift focus from controlling insurance prices to controlling excessive benefit claims costs.

The Alternative Lesson of Maine's Crisis: Control Insurance More Effectively

A more complete picture of Maine's crisis shows the opposite: that regulators did not give sufficient attention to controlling insurance costs as the route to a system that better protects workers and employers. First, the high insurance costs were a problem of insurance management and design, not simply price pressure from expanding benefits and benefit claims. Second, the collapsing insurance supply was a problem of the structure, not simply the pricing, of the residual market in Maine and in many other states. The crucial step in resolving Maine's insurance crisis for employers was not lifting price controls on insurers or cutting benefits to workers, but instead was the establishment of a new source of insurance structured to respond more directly and effectively to employers' interests in reducing workers' compensation costs. This raises the question whether further changes in insurance structures might create a system that better reduces costs for workers as well for employers, without sacrificing insurance affordability or solvency.

Insufficient cost management, not insufficient benefit pricing

The picture of benefit prices in Maine's crisis period of the 1980s through the early 1990s is more complicated if we look beyond the floundering commercial insurance market to the flourishing market for alternative insurance forms. As traditional commercial insurance coverage shrank to about 10% or less of Maine's workers' compensation market, employers turned not only to the insurer-run residual market but also to self-insurance. By 1991, various forms of self-insurance constituted over 40% of the state's total workers' compensation premium volume, up from 29% in 1989 [McCluskey, 2001, p. 115]. In effect, these alternative insurance structures were successfully competing with traditional commercial workers' compensation insurance (which also changed to offer more risk-sharing arrangements).

If employers could save money by self-insuring, covering the same statutory benefits to workers at lower cost, that suggests insurance prices could be inflated rather than suppressed. It is true that self-insurance systems can duplicate or exacerbate problems of insurance price suppression, if inadequate regulatory protections lead to underfunding. Recent insolvency problems have cast doubt on the viability of New York's group self-insurance system, for example [Task Force on Group Self-Insurance, 2010]. Furthermore, self-insurers' economic advantage over commercial insurers may come from their increased incentives to suppress workers' injury and illness, not from delivering similar benefits at lower costs. Nonetheless, Maine's self-insurance market so far appears to have remained largely stable for employers and insurers, raising the question that traditional insurance systems may have some disadvantages in delivering and funding benefits.

Proponents of the rate suppression theory of state workers' compensation crises have tended to implicitly acknowledge that employers' flight to self-insurance reveals some degree of rate inflation in commercial workers' compensation insurance. However, the rate suppression story explains this cost-savings as a problem of improper cross-subsidy that has unfairly penalized commercial insurance. One version of this story explains that adverse selection occurred when employers with particularly low costs turned to self-insurance to escape rates reflecting the broader pool's higher risks, so that insurers were stripped of the most profitable policies [Kramer and Briffault, 1991]. But if rates were inflated for a substantial volume of workers' compensation coverage, in a competitive market enterprising insurance companies would presumably seek to identify and seize the profit opportunities posed by those better-than-average risks (by cherry picking or by creating risk-sharing arrangements), rather than ceding those risks to self-insurance. If insurers nonetheless lost market share, whether due to inadequate risk information or other shortcomings, it could suggest not that government regulators or the insurance market failed, but that the regulated market worked well to select for more cost-effective systems.

A more revealing argument reconciles growing self-insurance with the rate suppression story by pegging the problem to inadequate residual market rates. Under the standard NCCI residual market system, traditional insurers faced a competitive disadvantage because of their obligation to cover the costs of growing residual market deficits. As rates increased to cover those losses, many employers moved to self-insurance to avoid that residual market “tax” [Kramer, 1991]. Indeed, a number of insurers paid large settlements in litigation charging that major insurance companies fraudulently shifted as much as one billion dollars of residual market costs to employers in several states during the 1980s and early 1990s, circumventing regulatory restrictions and hiding residual market charges in complex risk-sharing arrangements used for some voluntary market policies [Bradford, 1996, 1999]. Furthermore, in 2006, New York's Attorney General Eliot Spitzer obtained a $343 million settlement from litigation with American International Group (AIG) on charges that the insurer had underreported its voluntary market premium volume from 1985 through 1996 in order to escape residual market deficit assessments, which typically are based on market share [Ceniceros, 2009], thereby making that market appear less profitable than it actually was and inflating charges to other insurers and employers. In related ongoing racketeering litigation, AIG and other major insurers have accused each other of widespread and intentional underreporting of up to one billion dollars of residual market costs [Best Wire Services, 2010].

But aside from questions about the legality and accuracy of insurers' charges for residual market deficits, the rate suppression explanation fails to account for the fact that much of Maine's exodus into self-insurance came from businesses in the residual market. If the problem was that regulators kept residual market prices too low to cover costs, employers in that pool would be particularly unlikely to self-insure. In Maine, employers moving to self-insurance during that time were to a great extent subject to similar price constraints for coverage of the same benefits, with self-insured employers required to set aside funds at levels based on insurers' premium rates and risk classifications, in addition to satisfying comparable capital reserve requirements [McCluskey, 2001, p. 115]. Moreover, self-insurance typically involves substantial start up expenses and ongoing administrative costs, lacking the economies of scale and scope enjoyed by Maine's typical commercial insurers.

Nonetheless, during the time when Maine's residual market pool was plagued by dramatically rising deficits, substantial evidence showed that many employers in that pool could successfully cover their costs or even accumulate substantial surpluses by switching to self-insurance. For example, a number of group self-insured “pools” found their “rates” were sufficient to support “dividends” returned to their employer members, apparently without creating solvency problems [McCluskey, 2001, pp. 115, 120]. Moreover, contrary to the adverse selection theory, Maine's self-insured employers actually had higher than average risks during the crisis period. It is also true that an important spur to self-insurance during Maine's crisis period was a 1987 state law change shifting liability for residual market deficits from insurers to insured employers. Nonetheless, the surpluses accumulated by self-insured employers during the period of rising residual market deficits were not primarily attributable to avoiding residual market surcharges, most of which were not imposed until well after that time.

To explain the puzzle of how self-insurance could reduce the costs of the same benefit system for relatively risky employers during this crisis period, the director of Maine's self-insured employers' association offered evidence that self-insured individual businesses and business groups could operate effectively with much lower charges for general administration and capital (despite disadvantages in economies of scale). In addition, self-insured employers claimed to reap major savings in claims costs and litigation expenses due to better safety and loss control; better medical management; and a greater emphasis on returning injured workers to their jobs promptly [McCluskey, 2001, p. 120]. For example, an executive from a high-risk manufacturing company switching to self-insurance in 1990 testified in an insurance regulatory proceeding that their lost-time accidents decreased dramatically. When insured through the NCCI-run residual market, this company had no access to accurate records of their employees' claims. Upon self-insuring, they hired a risk management contractor who analyzed the company's accident experience and instituted a new cost control plan.

This is an example of the nationwide growth during the crisis period of “third party administrators” specializing in managing (but not underwriting) employers' workers' compensation insurance needs. These businesses often marketed their services to self-insured employers by highlighting their innovative loss control services and more efficient administration, and this marketing effort presumably helped many employers become more informed and attentive to loss control. Of course, that loss control tended to focus primarily on control of employers' losses, not necessarily on reducing workers' losses, since the services are hired by and accountable to employers rather than to workers. It is unclear how much the savings from loss control in Maine's self-insurance systems involved increased attention to safety and appropriate return-to-work rather than claims suppression. Instead, the more modest point of this example is to suggest that changes in control over insurance restructuring can dramatically change the costs of losses and of benefits, and that further changes in insurance control will be necessary to ensure that loss control protects workers as well as employers.

Inadequate residual market governance, not insufficient insurance supply

Maine's residual market's failure was a central cause of the cost disadvantages of traditional insurance in that state, but that failure was not simply a problem of inadequate rates. Along with over thirty states, Maine's residual market pool was part of a system operated by the NCCI, which structured those pools to separate responsibility for insuring from responsibility for servicing, which includes functions such as collecting premium and processing benefit claims. This separation posed the problem of “moral hazard,” where losses tend to increase when those who control the degree of loss are protected from the costs of that loss [Feldblum, 2010].

Exacerbating that problem, through the early 1990s, the NCCI's residual market system was governed by a board comprised of major insurance companies, typically including many of those companies most active in servicing the pool's policies [McCluskey, 2001, p. 101; Hoffman, 1991, p. 20]. This board assigned the pool's servicing functions to individual insurance companies without competitive bidding. Employers relegated to residual market coverage also had no choice about which insurer was assigned to manage their policies.

In addition, through the 1990s, the NCCI's pool compensated its “servicing carriers” by allowing them to retain a servicing fee typically fixed at 30% of the pool premium they collected, leaving the remaining 70% of premium dollars to cover claims costs. This fee level was higher than the amounts NCCI normally reported for comparable voluntary market servicing expenses during the same period, and more than double the reported expenses for comparable (and better) services to Maine's self-insured group programs [McCluskey, 2001, pp. 97–98].

This fee structure appears not only to have unreasonably drained the pool's reserves for claims, but also to have produced incentives for claims costs to increase. Without effective performance monitoring or competition, servicing carriers had incentives to profit by cutting expenses for services such as claims handling, safety instruction, careful accounting of losses and premium charges, or by fraudulently misreporting or withholding premium from the pool (according to charges in ongoing litigation).

Checks on this self-dealing residual market system appear to have been minimal. The NCCI was structured to serve as an advocacy organization representing the interests of its insurance company members, a role seemingly in conflict with effective investigation and monitoring of individual insurers' poor performance. Although the NCCI conducted limited internal audits of servicing performance in the pool (and though these audits showed pervasive problems in Maine's pool servicing), the NCCI appears to have focused on seeking rate increases and benefit cuts rather than on insurer accountability as the main solution to deficit problems. Employers insured through the residual market (disproportionately small or new businesses as well as high-risk industries) tended to focus their frustration on restraining workers' benefits or seeking voluntary market coverage (or self-insurance) rather than on probing the technical and opaque operation of the residual market. State regulatory agencies appear to have routinely approved rates incorporating the NCCI's asserted fee levels without requiring specific supporting evidence of expenses or performance. Further, as long as the “residual” market remained a small part of each state's insurance coverage, the pool's costs and operation typically remained at the margins of regulatory scrutiny.

Similarly, as long as residual markets were small and voluntary market profits generous, individual insurers had little incentive to reduce the risk of residual market losses. These conditions gradually eroded as claims costs rose during the 1970s and 1980s, following efforts to increase benefit adequacy, and as state regulatory agencies began to move from rubber-stamping insurance rates toward substantive and adversarial ratesetting proceedings [McCluskey, 2001, pp. 72–94]. This increased—and increasingly uncertain—voluntary market risk exposure likely induced insurers not only to be more selective in their voluntary market underwriting but also, under some circumstances, to substitute the virtually risk-free up-front gains of residual market servicing for regular insurance underwriting. While this response profited some insurers in the short run, it set the stage for an eventual “death spiral” as projected shortfalls in growing residual markets accumulated and induced insurers to further reduce voluntary market share to avoid liability for those impending deficits, which the NCCI's pool system spread among insurers based on their portion of voluntary market business.

In Maine, the NCCI responded to the process of market collapse by successfully lobbying for legislative reforms in 1987 that shifted reinsurance responsibility for the residual market to employers. In exchange for this bailout, this legislation required insurers to increase voluntary market coverage to specified levels over a period of years or face assessment for a portion of the deficit liability. Though this legislation also included benefit cuts and rate increases, it failed to restore the state's voluntary market and provided only a temporary respite from the prospect of total market collapse. These law reforms failed not primarily because continued low insurance rates or high benefit costs made the voluntary market unprofitable on its own, but because even enormous profit opportunities would be insufficient to offset the risk of residual market liability for any one insurer contemplating increasing its voluntary market share [McCluskey, 2001, p. 110]. Even if insurers as a whole returned to that market en masse, the accumulating years of rising costs would pose substantial risk and uncertainty about its allocation. Further, that risk of residual market deficits was exacerbated by a regulatory deal giving one insurer, AIG, responsibility for servicing a large portion of the residual market business in exchange for special protection from any pool liability. Evidence of poor servicing (particularly by AIG) led to findings in ratesetting proceedings in 1990 and 1992 that insurers had substantial mismanaged the state's residual market through faulty premium collection, recordkeeping, and claims handling as well as inadequate investment practices; the state's insurance superintendent estimated that this mismanagement inflated residual market costs by 30% [Me. Bureau of Ins, 1990].

It was the possibility that insurers would finally face their first assessment for a portion of Maine's looming residual market deficits in 1992 that led insurance companies to then take steps toward withdrawing entirely from the state's market, forcing the legislature to enact more comprehensive changes that year, including further rate deregulation and major benefit restrictions. However, Maine's 1992 law reforms reflected agreement among otherwise divided political interests that residual market instability, not just benefit pricing, was at the heart of this insurance market collapse and its solution [Tri-Agency Report, 2010].

For that reason, along with several other states in the early 1990s, Maine withdrew from the NCCI pool system and solved the problem of insurance company flight by creating a new quasi-public insurance fund as the insurer of last resort beginning in 1993. That same year, responding to this rising competition and criticism, the NCCI began major changes in its residual market system, such as opening up residual market servicing to competitive bidding and rewarding servicing carriers for reducing fees and improving performance—with the result that pool servicing fees dropped while quality appeared to improve [Katten, 1995; NCCI, 1995]. Ultimately, deals struck through legislation and litigation divided financial responsibility for an estimated $220 million deficit in the remaining claims obligations of Maine's defunct pool among insurers (primarily servicing carriers), the state's guaranty fund, and employers (who bore the greatest share) [Fletcher, 1993; McCluskey, 2001, p. 113].

Restructuring the residual market

Easing Maine's insurance supply crisis involved confronting a perennial social insurance dilemma. Affordable guaranteed coverage requires spreading costs from higher to lower risks, but that risk-spreading can destabilize the insurance pool (or the insurance coverage) as resistance mounts from those forced to bear costs outweighing their own gains from the pool. Maine's reforms solved the immediate collapse of the voluntary insurance market by freeing its private insurers (prospectively) from responsibility for the costs of the insolvent residual market and instituting a new guaranteed source of insurance. Nonetheless, the basic insurance supply problem remained. How could the new pool act as an insurer of last resort without undergoing a new death spiral?

The mere fact of eliminating the insurance industry from residual market cost-spreading would simply have set the stage for lower-risk employers to avoid subsidizing the pool by escaping into the now-unencumbered voluntary market (aided by benefit cuts and deregulated rates). As the shrinking residual market pool's risk-spreading capacity diminished, it would either squeeze the remaining employers through ever-escalating prices or run up huge deficits, requiring further bailouts from employers or taxpayers and likely inducing more sacrifices from injured workers through additional benefit restrictions. Or, the public fund might attempt to maintain its market share by underpricing risks below sustainable costs, which (in a deregulated market) might destabilize the voluntary market by inducing a rash of competitive price-slashing, leading to mass insolvencies and further bailouts and benefit cuts, as in California's recent experience [Dixon et al., 2009].

Contrary to these scenarios, Maine's reconfigured market has provided a stable, seemingly solvent insurance supply, with dramatically lower loss costs for employers and insurers (down 47% since 1993 [Tri-Agency Report, 2010]). Reported lost time injury rates, as well as compensation claims for disabling injuries, decreased substantially during the mid-1990s in Maine, though the degree to which this decrease accurately reflects reduced injuries is somewhat uncertain [Conway and Svenson 1998, Tables 6, 7]. Although private insurers quickly returned to the market, the new insurer of last resort, the Maine Employer's Mutual Insurance Fund (MEMIC) has held between 60% and 65% of the insured premium volume since 2002, maintaining an overwhelming lead since its 1993 inception. Originally capitalized through annual assessments from its employer policyholders (presumably a competitive disadvantage compared to private insurers), it accumulated this capital (set to be comparable to private insurers' surplus requirements) in half the expected time and then refunded the entire capital investment to employers by 2001. MEMIC has received consistently high financial ratings, and its apparent financial strength has allowed it not only to price its policies at or below those of private insurers, but also to return $110 million in dividends and capital repayments to its insured employers since 1998. In a similar example, Kentucky's quasi-public insurer and provider of last resort, established in 1994, has maintained a dominant market position along with excellent financial ratings and—in a reversal of the residual market bailout problems—recently had to fend off political attempts to control its large surplus accumulated even while regularly lowering employers' rates [Insurance, 2010]. Self-insurance has also maintained its position since Maine's crisis period, covering 44% of the premium volume in 2008 (ranging from 40% to 49% since 1998) due especially to the persistent market strength of self-insurance groups [Tri-Agency Report, 2010].

MEMIC has not only out-competed traditional insurers in Maine's insurance market, despite being required to cover the market's riskiest and costliest employers, but it has expanded (through a subsidiary) into the private workers' compensation market nationwide and into other lines of business insurance as well. By 2009, the MEMIC group was among the nation's 50 largest workers' compensation insurers [National Underwriter, 2009].

Though MEMIC is primarily designed to prioritize reducing costs for employers, rather than for workers, it offers some clues about how further, more worker-centered, insurance reforms could substantially reduce the costs of expanded benefits while solving the risk-spreading dilemma of guaranteed coverage. First, Maine's reconfigured market suggests that workers' compensation insurance often may be less costly for employers when structured to maximize direct ownership and control by stakeholders (rather than capital shareholders or political leaders) and when structured to allow for substantial specialization.

Second, this potential for cost-savings seems to stem from a more diffuse shift in insurance culture to focus on reducing risks rather than selecting, spreading, and pricing risks (or reforming benefits) as the route to financial strength. That risk reduction could potentially include improved safety and worker-oriented rehabilitation, though the extent to which that potential has been realized is unclear and would require further study.

Examining organizational structure, MEMIC is designed as a mutual insurance company owned and governed by its employer policyholders. Furthermore, MEMIC is structured to enhance management's accountability to policyholders, giving more substance to the mutual form than is typical among insurers. MEMIC's authorizing legislation requires its nine member board of directors to include six policyholders along with two board members appointed by the Governor, subject to review by the legislature's insurance committee, and this law also prohibits lobbyists and workers' compensation service providers from board membership. In the beginning, company governance included a system of nine industry-specific employer advisory boards, each with nine members elected by businesses in that industry; though eventually abandoned as unwieldy, this initial system of intense employer involvement created a pool of loyal and informed policyholders aware of and committed to the company's distinct vision [Gold and Gold, 2003].

Further, MEMIC's authorizing statute insulates it from the general insurance industry not only by giving it sole responsibility to act as insurer of last resort, but also by excluding MEMIC from the general insurance guarantee fund and by limiting MEMIC's provision of insurance lines (or other products) not tied to its workers' compensation business. MEMIC is a “state fund” distinct from traditional private insurers by being subject to political oversight through this authorizing law, its public board members, and its need for legislative approval of major structural changes (like the creation of its subsidiary for business outside of Maine). Nonetheless, unlike some traditional state funds, its authorizing statute protects policyholders from conflicting political interests by stating that MEMIC is not a state agency or instrumentality and by prohibiting the state, the state's General Fund, or any state agency or division both from supporting MEMIC and also from borrowing or appropriating MEMIC's funds.

Along with this quasi-public mutual structure, MEMIC's specialization may help protect policyholders' long-term interests. Maine's workers' compensation system remains its core mission, with MEMIC as the parent company even as it has grown to other states and lines. In contrast, despite similar origins as an employer-focused workers' compensation mutual, Liberty Mutual (also a significant provider in Maine) has restructured into a complex conglomerate where a holding company owns numerous subsidiary stock companies and other entities focused on many kinds of insurance in numerous states nationwide as well as in numerous foreign markets [Sclafane, 2000]. The narrower purpose and design of MEMIC and similar state funds may be particularly useful in reorienting workers' compensation insurance from a focus on spreading and predicting risk to give more attention to reducing risk. Without this focus, insurers' long-term performance in a particular workers' compensation market may be undermined by opportunities for short-term gains from expanding into new markets, moving in and out of states as benefit laws change, and using workers' compensation as a loss leader to attract other more profitable lines of business. Expertise in safety, reemployment, and claims handling may require extensive and initially costly specialized knowledge of particular industries, occupations, workplaces and social contexts, as well as the benefit systems of each state. Further, this expertise may depend not just on technical knowledge but on building trusting relationships with employers and workers and on otherwise cultivating local good will, qualities that may be enhanced by strong and concentrated long-term commitment from top management and directors to a distinct and stable market.

As a result, the seemingly detrimental duty to serve as a state's insurer of last resort can create a competitive advantage by creating a stable market niche. To some extent, high-risk employers and small businesses generally deemed less profitable by traditional insurers appear to have been an untapped asset presenting opportunities for innovative service and profitable loss control (though again, it is unclear the extent to which this loss control actually represents improved safety or simply improved claims control). This loss control expertise may then have proved valuable when applied to a broader market. At the same time, a major factor in the competitive advantages (for employers) of MEMIC and similar new state funds was that they took over “residual” markets that had grown sufficiently large to broadly spread the start-up costs of covering high-risk policyholders and to amass sufficient funds to support early investments in cost containment.

Second, MEMIC's success in reducing costs for employers stemmed from a cultural change among influential business leaders. Maine's extended crisis and insurers' egregious servicing performance in the dominant residual market sparked widespread criticism of insurance companies by employers, leaving them at least to some extent open to changing their ideas about how to reduce costs, despite their usual political alignment with insurers to restrict benefits. Examples of cost savings from better insurance management and from professed attention to safety and reemployment among the many residual employers becoming self-insured reinforced this criticism.

From the start, MEMIC proclaimed a mission of promoting innovative loss control distinct from traditional insurance [McCluskey, 2001, at 125]. Of the founding group of incorporators and initial board members, none had insurance industry experience or ties; instead, it was comprised of managers, owners, or health and safety specialists from a diverse mix of businesses, along with several factory workers and a legislative aide. This group declared its goal as creating “an anti-insurance insurance company” that would prioritize improved claims handling, safety, and rehabilitation [Gold and Gold, 2003, p. 39, p. 46]. As with other insurers, MEMIC's sometimes lofty rhetoric about safety and concern for workers has coincided to some extent with a practice of opposition to workers' rights, both in individual claims proceedings and in public policy arenas. Nonetheless, the substantive composition and culture of the company's founders and directors suggests some differences from other insurance companies in the extent to which it approached workers' compensation insurance as a means for ground-level workplace change (whether detrimental or beneficial to workers) rather than primarily a matter of actuarial and underwriting expertise.

While not evidence of its actual practice, MEMIC's aggressive marketing of its safety programs suggests its special attention to the relationship between workplace practices and insurance costs. Within a few months of its founding, it launched a half million dollar advertising campaign promoting workplace safety and branding the company's chief executive as “the safety guy.” The company brands itself using phrases such as MEMIC: Partners for Workplace Safety. The company's 2009 annual report web page features a series of case studies of the company's safety improvements in a variety of small businesses and non-profits, including a manufacturer, construction company, nursing home, and social service provider. These examples claim that safety services helped transform daily culture among both managers and workers, not just through specific training and equipment but by improving workplace communication and by building workplace trust. Managers touted the benefits of safety not only for saving insurance costs, but also for increasing worker retention and morale and for raising the competitiveness and quality of the business's product or services.

As Emily Spieler has incisively analyzed, incentives for innovative loss control often mean suppression of legitimate workers' claims, whether through fraud charges, delay, harassment, intimidation, or workplace penalties for reporting injuries and claiming benefits [Spieler, 1994]. It is unclear how much of MEMIC's professed attention to loss control actually works to benefit workers as well as employers, particularly since the company is not structured to directly protect workers' interests (reflecting insufficient political support for legislation that would have required labor representation on the board). Nonetheless, even if not the whole story or even an accurate representation of typical practices, such public relations efforts to promote enthusiasm for safety and workers' well-being “as a major social good” [Fletcher, 1995] may have significant indirect political and cultural impact by casting workers' rights in a positive light. In contrast, widespread negative public images (often promoted by insurers) blaming injured workers for high workers' compensation costs have helped support benefit cutbacks. The extent to which insurers' public relations efforts centered on the benefits of workplace safety can be effective in reducing injuries and the extent to which these can be the basis for coalition building in favor of workers' rights needs more exploration.

MEMIC claims to have supported its safety rhetoric through a number of substantive innovations, and research comparing the impact of such allegedly different insurance approaches could be an important step toward re-orienting insurance systems to better protect workers. MEMIC contends it made groundbreaking investments in loss control, starting with 5% of premium at its founding [Gold and Gold, 2003, p. 40]. In 2000, the company reported spending about 35% of operating expenses on safety and other loss control services, which it claimed was double the amount of most insurance firms [Gold and Gold, 2003, p. 89]. The company's founders explained that they rejected the easier route of subcontracting in favor of a building an in-house claims management department and underwriting database to avoid the accountability problems of the former residual market system. In its first year, the company began developing safety workshops tailored to each type of business, focusing especially on high-risk businesses like logging, where fatalities fell from up to five a year to hold at zero for at least several years [Gold and Gold, 2003, p. 90]. Training goals were enforced by a system of biannual workplace inspections by MEMIC staff. The company also designed and promoted individualized reemployment programs. Further, early MEMIC's innovations included substantial up-front price rewards (or penalties) and other incentives for individual employers based on their compliance with safety standards and participation in loss control programs (in addition to incentives based on actual loss experience) [Gold and Gold, 2003, pp. 50, 55, 60]. Note that these loss control efforts differ from the standard focus on risk-based pricing by targeting rewards more directly to safety rather than to claims reduction.

It is unclear the extent to which MEMIC's initiatives caused the state's subsequent drop in reported injury rates, and at least some of this drop may have been due to a concurrent OSHA initiative in the state as well as to the impact of changing economic conditions on the state's particularly high-cost industries. Other state insurance funds' claims to save money through improved safety have been found to be hollow; for example, in the early 1990s, Oregon's non-profit insurer, SAIF, was found to have used rhetoric of safety improvements to cover up widespread fraudulent suppression of workers' claims [Duin, 2005]. Without changes in the structure of these insurers, as well as in workplace regulation more broadly, such claims to safety programs and reducing injury rates remain unreliable.


Maine's reconfigured market should not be taken as evidence that competitive state funds are more likely to advance workers' rights, but instead should be viewed as a case study suggesting the need for further attention to the question of how insurance structures shape the long-term costs of benefit systems. New York and California provide examples of competitive state insurance funds that appear to have made less effort to distinguish their practices from traditional insurers, even though these funds similarly guarantee coverage while holding a dominant role in large markets. While such longstanding public funds have allowed many states to avoid the insurance market meltdowns that occurred in states relying on the private NCCI residual market system, these states also lacked the impetus presented by the private market crisis to scrutinize and change the role of workers' compensation insurance. In a study of state funds over the two decades before 1995, competitive state funds were associated with higher rates for employers than other insurance structures [Thomason et al., 2001; McCluskey, 2002].

That leaves the question of whether regulatory changes in insurance could actually bring cost savings for both workers and employers that have been claimed (though at best only partially realized) by some of the more recent quasi-public funds like MEMIC. Although further study of these examples is important, several themes emerge that suggest some principles for improved insurance regulation.

First, attention should focus on structural changes in the governance of state insurance funds and in insurance rating systems to increase accountability to the main stakeholders in the system (workers and employers), and to strengthen independence from the insurance industry and from other workers' compensation providers. Second, the regulatory framework should enhance the potential for public insurers to gain market control by cutting long-term costs to employers and workers. Third, regulatory protection against insolvency should focus not just on adequate pricing of predicted losses but instead on insurance fund governance, management and market structure. Finally, technical regulatory change must be accompanied by leadership in changing the cultural and ideological understanding of the role of insurance in workers' compensation.

Looking first at governance, the boards of directors of state insurance funds should be designed to ensure representation from a broader range of employers and especially to better represent workers, along with tighter restrictions on directors' and executives' ties to insurers and other workers' compensation vendors and service providers. The apparent success of Maine's state fund in responding to employers' interests in that state appears at least in part due to a governance structure that especially at the initial stages emphasized employers' interests and representation rather than ties to the traditional insurance industry. Other state funds could follow this lead by further shifting governance away from those with ties to the insurance industry toward workers and toward a wider range of employers. In 2010, the New York State Insurance Fund's Board of Commissioners included the president of the state's AFL-CIO, as well as the ex-officio membership of the state's labor commissioner; but of the six other members, two are members of very large corporate law firms presumably including major insurers as their clients; one is the president of a large insurance brokerage firm; and of the three business executives, one represents an international financial services firm. This composition may discourage strong leadership that might threaten the market power of competing insurers and seems poorly designed to foreground the interests of either workers or the interests of the small businesses particularly dependent on state funds. Policyholder and worker elections for some board positions may sometimes lead to greater accountability than political appointment, combined with other controls and oversight. Other possible avenues to enhanced employer and worker involvement and accountability in management could include refining the idea of policyholder and worker advisory boards tried by MEMIC.

Recent corruption scandals underscore the problem that formal employer governance alone does not bring accountability sufficient to restrain self-dealing by employer board members or managers. For example, Beacon Mutual, a quasi-public fund that took over Rhode Island's residual market from the NCCI, was found to have given improper premium breaks to favored companies [Rousmaniere, 2007]. Further insights would come from examining the impact of California's 2008 legislation reforming governance of that state's insurance fund, after the fund was found to have paid more than $500 million in dubious fees to firms associated with board members, and to have been plagued by extensive general waste and claims mismanagement [Workers' Comp Executive, 2008]. Changes in that state fund's governance included expanding the board's size and providing board member training and compensation, along with increasing transparency by subjecting the board to open meeting and public information laws.

Another impediment to protecting stakeholders against competing insurers' interests is the relationship between state funds and rate advocacy organizations that include private insurers. Until 2008, New York's state fund's ability to compete against private insurers based on price was limited because it followed rates submitted by an insurer-controlled rating board (subject to regulatory approval). Recent reforms have shifted to the now-standard more competitive national ratesetting practice whereby the insurance industry rating board instead submits to regulators more limited “loss cost” data, allowing individual insurers to adjust these rates for their own profit and expenses, along with limited loss adjustments. Nonetheless, this system continues the potential for collusion and conflicts of interests by mixing data collection with rate advocacy. The goal of developing data to support the possibility of profits through superior and long-term loss control may conflict somewhat with the goal of gathering data to support profitable pricing of losses. The reforms of New York's system addressed some of these concerns by including four “public” members on the board along with four insurance industry members (one of whom represents the state insurance fund). In 2009, these public members included one leader of organized labor, one leader of organized business, and one representative each for the state's insurance department and workers' compensation agency [New York Compensation Insurance Rating Board Annual Report, 2009]. A question for further study is the effect of this changing membership; the technical expertise and resources for active and independent participation by workers and employers is likely to continue to be a challenge even when they have significant representation on governing boards. Independent public data collection agencies that can develop long-term expertise may be a better approach over the long run.

Second is the question whether a “public option” (in workers' compensation as in health insurance) should be structured to permit it to become the most cost-effective insurer possible, even if this means overshadowing the private insurance market. Private insurance companies often have tried to limit state fund competition through litigation and legislation, typically charging that these funds enjoy unfair advantages like tax breaks [Hays, 1999; Lopez, 2000]. By acting as insurers of last resort, some state funds may subtly or not be treated as the analogs of the longstanding NCCI residual market model, designed primarily to support the private market by freeing it from the costliest coverage and by allowing insurers greater flexibility to cycle in and out of the market. This view (echoed in the health reform debate) is problematic in assuming that the private market is inherently the most cost effective structure, and that guaranteed coverage necessarily means higher costs and not greater efficiency. Instead, state-created guaranteed insurance suppliers might under certain conditions turn restrictions on risk selection into specialized expertise in producing more cost-effective insurance for both employers and workers. Maine's experience with MEMIC perhaps best illustrates the potential for giving state funds both the independence and power to compete against traditional private insurers not simply by offering lower premium prices, but also by investing in specialized services tailored to the specific needs of small businesses in a particular state.

In one example suggesting the problems of limiting competition by state funds, an executive of New York's State Insurance Fund recently defended against insurer charges that it engaged in aggressive marketing by insisting that it did no paid advertising—except for anti-fraud campaigns designed not to take away business from other insurers but to encourage the public to participate in weeding out fraud [Rosenfeld, 2007]. In this way, New York's insurance market seems structured to discourage competition for promoting employer safety while encouraging collaboration in discouraging injured workers' from making claims.

On the other hand, allowing state funds to fully compete for market dominance does raise concerns about lack of competitive checks on poor service, or conversely, about destructive competition through predatory pricing. Although these are real concerns, hobbling state funds from gaining market share will not be an effective means of preventing those problems. Much of the workers' compensation market is not subject to broad competition, but is rather a “natural monopoly.” Small businesses in particular are likely to be dependent on guaranteed suppliers, and therefore most at risk in a system relying on competition from private insurers to keep insurance costs low and stable. Further, unlike simple price competition, gaining market share by convincing employers of the benefits of superior loss control and claims management services is likely to require long-term investment in producing and disseminating information to raise employer awareness and interest. Shifting competition from a short-term focus on low prices to create more nuanced interest in reshaping workplaces to achieve long-run savings requires attention to governance and structural incentives other than simply increasing the presence of private insurers.

These concerns relate to the question of solvency regulation, a third area of potential improvement in workers' compensation insurance structures. A conference of experts examining the recent solvency crisis in California's workers' compensation system revealed many overwhelming obstacles to ensuring that prices are sufficient to fund the long-term costs of often uncertain and changing benefit systems [Center for Insurance and Policy Research, 2010]. Efforts to improve and implement actuarial estimates of claims costs while allowing competitive pricing appear daunting, and likely to lead to political pressure to reduce uncertainty (and loss costs) by restricting benefits. Yet returning to the historic system of regulated noncompetitive pricing is unlikely to solve that uncertainty, other than to give insurers more opportunities to overprice, leading to further pressures for benefit restrictions. Instead, solvency regulation should be re-oriented to more heavily emphasize insurance structures and governance rather than pricing predictions, as appears to be the approach of some European regulatory systems [Grace and Klein, 2009]. Further, the recent problem of New York's group self-insurance insolvencies provides evidence that the apparent cost management advantages of this form of insurance supply could be better achieved through a larger public or quasi-public employer mutual structure. The goal should be to design insurance structures to induce competition on grounds other than short-term price savings and long-term cost-shifting to workers; and to design structures that use methods other than competition to preserve long-term accountability and stability.

Fourth, perhaps most important, is the question of how to change not only the formal rules but also the ideas and even spirit undergirding workers' compensation insurance. The most comprehensive incentive and oversight mechanisms can be manipulated and circumvented; cultural change is crucial for giving substance to regulatory reform. Nonetheless, seemingly technical details of insurance regulation can contribute to making insurers leaders in producing (or impeding) positive cultural change. In Maine, MEMIC's rhetorical efforts to build a public identity focused on workplace safety reflects the company's stated mission to vigorously challenge or even supplant the traditional insurance market in that state. New York's state fund's contrasting emphasis on fraud appears to be indirectly a product of more ambiguity about its institutional role, which could be clarified through regulatory changes. More fundamentally, national advocacy groups and coalitions between labor and business interests could play an important role in helping to produce and disseminate a story of workers' compensation emphasizing the potential for long-term cost-savings through increased protection of injured workers.

In conclusion, the system for financing workers' benefits plays a central role, though one that is often obscure and complex, in shaping both the adequacy of benefits and the risk of workplace injury and illness. Regulation of insurance must control against incentives for the insurers funding workers' compensation benefits to shift those benefit costs back to workers, whether through claims mishandling, claims suppression, insolvency, excessive employer costs leading to job loss, or legislative benefit restrictions. With much more comprehensive change than has been implemented so far, insurance systems might be designed to go beyond simply funding the existing level of benefits to help make expanded workers' compensation rights more affordable.