No. 1, January 2007
Last year’s passage of a law in Massachusetts intended to substantially reduce the number of people without health insurance has given hope that such a program might be politically and fiscally feasible in other states and at the national level. Massachusetts has tried health-care reform before, without notable success, but this time may be different. Whatever the fate of the Bay State’s new program, what features of it are unique to Massachusetts? Which aspects can be and ought to be copied by other states, and which might require or benefit from modification?
In April 2006, Massachusetts governor Mitt Romney signed into law a landmark health insurance reform bill. The new plan has four key components:
- an insurance mandate on individuals backed up by financial penalties and low-income subsidies
- establishment of a statutory insurance mechanism called “the Connector” to select and approve insurance policies that will be available for purchase by small businesses with fifty or fewer workers and by other individuals and families without access to employer-sponsored coverage
- “affordable” premiums for policies offered through the Connector that are supposed to be significantly lower than current insurance costs in the Massachusetts market
- a mandate for employers to offer their workers the option of paying explicit insurance premiums with pretax dollars, using Section 125 plans
Both the political factors that led to the passage of the Massachusetts law and the resulting program are complicated. The new plan does not make a clean sweep. Rather, it is designed to fit into a complex set of longstanding rules, regulations, arrangements, and special deals. The Massachusetts fix highlights one of the downsides of any incremental reform: unavoidably mind-numbing detail fully known only by the people who managed the old system and designed the new one.
Massachusetts had approximately $380 million in federal Medicaid waiver funds (targeted for state programs to help low-income people without health insurance) that would have gone back to the U.S. Treasury had no new program used them to provide insurance coverage to more people in the state. Given a total price tag for the new Massachusetts program of about $1 billion, being able to pay for 40 percent of it with already-existing federal funds was surely an incentive to proceed. Rigorous research shows that, faced with varying levels of federal funds to cover the cost of Medicaid programs, states do the logical thing: they are more likely to spend more when the matching funds are generous. Reducing the initial cost of
a more generous state program to state taxpayers surely got the ball rolling.
Furthermore, Massachusetts already had in place a program in which the state effectively “taxed” private insurance premiums and hospitals to help fund the costs of charity and bad-debt care. Other states do not have this kind of program and precommitted funding. Since extension of coverage would, in theory, reduce the need for hospital charity care for the uninsured, state funds could be diverted to cover the cost of new insurance subsidies to the formerly uninsured. While the initial fiscal picture is almost surely brighter than it will eventually become, there is little doubt that this confluence of free money and already-available funds made the heretofore difficult political issue of the cost of helping the uninsured much easier to handle in Massachusetts than it is likely to be in other states.
Finally, a relatively small share of Massachusetts’s population is uninsured when compared with other states. According to the Census Bureau, about 16 percent of Americans are uninsured; only 10 percent of Massachusetts residents are. Reducing the ranks of the uninsured even further might also take the problem off the political table.
Assessing the Massachusetts Program
Many of the features of the Massachusetts plan are praiseworthy and duplicate those in other programs proposed more than a decade ago. The plan recognizes, for example, the need for all individuals to carry health insurance. It also means-tests (or “income-conditions”) the premium subsidies for mandated coverage. Finally, it relies on private health insurance and markets.
But just as some of the features of the Massachusetts plan are familiar to those studying health care, so too are many of the objections. Libertarians, for example, protest on fundamental philosophical grounds. Why should most people in Massachusetts--who already have health insurance--subsidize coverage for those who do not? What business does the government have requiring a citizen to buy health insurance? The ultimate answer to this objection is grounded in morality and humanity: most citizens surely feel concern about the relievable suffering of their fellow human beings and correctly perceive that this suffering can be largely alleviated if people have insurance. The Massachusetts law can be viewed as an attempt to transform that sentiment into political action. But the insight that motivates the Massachusetts plan and others like it is that taxpayer contributions toward insurance which allows access to more than minimal care may be better, if not cheaper, than leaving people to seek inadequate charity care just because it is free, or paying for similarly minimal care out-of-pocket.
This argument should not be taken too far. Neither morality nor rationality would suggest that we want our neighbors to be wholly unburdened of financial risk or responsibility, or that they must have access to unlimited amounts of medical care--no matter how small the benefit relative to its cost. No one should be allowed to free-ride; there is a limit to altruism. But a minimum level of insurance coverage is necessary to ensure that everyone can obtain an acceptable level of access to care.
What level of coverage--between none and total--is best to mandate or subsidize? The minimum level of coverage for each type of care should encourage use of high-benefit care (in other words, care that provides the best outcomes) that is highly responsive to insurance coverage. Even if we had that information, we must still decide where to draw the line in terms of how much care to provide and what degrees of improvement in health care are sufficient.
At present, there is not enough information even to outline the options. Health-care experts know fairly well how coverage affects care and somewhat less well how subsidies affect the demand for insurance, but good information linking coverage to health outcomes is sorely lacking, especially for the majority of the uninsured who are not poor. Trial and error is the most sensible way to proceed. Without frank acknowledgment of the uncertain nature of the task, political pressure and snap judgments will probably triumph by requiring some types of coverage and ruling out others, although budget constraints, if they are truly binding, should prevent the worst excesses.
Let the Consumer Decide
Faced with the combination of anxiety and abysmal ignorance about key elements of policy design, how should the levels of coverage be determined? The best strategy would cease trying to discover the unknown, and instead place as few restrictions or obligations as possible on the coverage a person must have. The simplest way to combine a hands-off approach with the desire to make a difference is to specify the obligation in terms of actuarial value. At a given risk level, a qualified policy would be any one with an actuarial value of $X or more. This value could be varied with risk (although doing that perfectly would itself be a daunting task), and it surely should be varied (inversely) with income.
Insurance should cover medical and surgical services--not spas, aromatherapy, and the like. But within broad limits, it would be left up to consumers--not professional advocates for any specific worthy service--to decide what they wanted their insurance to cover and how to arrange cost-sharing. People might choose high-deductible coverage, or they might prefer coverage with upper limits--that would be their choice. Of course, one could always spend more than the actuarial amount as long as the funding was from personal funds. This would surely be a feasible strategy for the middle-income-and-above “mandate,” but it could also serve as a useful template for the upper end of the subsidized coverage. Closer to the poverty line, insurance should cover everything essential, but lower-middle-income people should be given some choice. Adjustments to the rules could always be made later if many people were to choose foolish plans.
This permissive policy would have another dividend: because people would be choosing the policy they prefer most, more people would be eager to purchase the coverage. It is the fear that even a large subsidy or a difficult-to-enforce mandate would still leave some lower-income people without coverage that has provoked recent controversy in Massachusetts over what each household’s premium should be. The best way to induce people to accept subsidized coverage, or to ensure that a political mandate is truly effective, is to permit individuals to choose their form of coverage. Any restriction or obligation--no matter how well-meaning and even if it does not increase cost--will turn off some potential buyers. Getting decent coverage to almost everyone is a better initial goal than getting perfect coverage to fewer of the uninsured. Rather than focus on details of subsidized coverage--as Massachusetts seems to be doing because of its extensive list of coverage mandates--the initial focus should be on providing coverage of greater actual value than a given benchmark, leaving plan details to consumers, not special interests or public health experts.
There is no predetermined group of currently uninsured individuals to whom subsidies can be targeted or mandates aimed. Both uninsured and insured alike are spread across large firms, small firms, the self-employed, and the unemployed, and across a range of income levels. They have different levels of household income, health risks, and insurance preferences. The right principle is the opposite of the sort of targeting in which Massachusetts is engaging: subsidies should be neutral, of the same value for a household with a given set of characteristics regardless of how that household might obtain insurance coverage. The current variegated pattern of state and federal subsidies (explicit and tax-based) makes neutrality a challenge, but it is the right direction nonetheless.
A significant proportion of low-wage workers in Massachusetts and across the country already has private health insurance. More than three-quarters of households with incomes between 100 and 300 percent of the poverty level have private insurance. This means that any insurance subsidy offered to all low-income workers would be paying many of them to do what they are already doing: obtaining insurance. Low-wage workers are also much more likely to lack coverage if they work for small firms or are self-employed than if they work for large firms, generally because of the much lower administrative cost associated with furnishing insurance in the large-group setting. The problem is clear: subsidizing insurance for all lower-income (but not poor) people across the board will be costly, but limiting eligibility to some and not others will be politically charged and economically distorting.
Under the Massachusetts law, not all low-income households are eligible for the state’s insurance subsidies, and not all forms of health insurance qualify. In particular, subsidies will be directed only to people with incomes up to 300 percent of the poverty level who get their health insurance through the Connector, which is restricted to individual purchasers and employees in small firms. The low-income worker in a large firm gets no subsidy.
Such incomplete targeting is presumably intended to hold down the cost of the subsidy program. Since there are many already-insured low-wage workers in large firms (even in Massachusetts), this policy seeks to achieve what is called “target efficiency,” and it has obvious fiscal appeal. Policymakers under budget constraints pursue target efficiency in order to direct as many designated program dollars as possible to currently uninsured individuals, thereby achieving the biggest coverage bang for our health-care bucks. The goal is to minimize government spending on lower-income individuals who would have already been insured without a new program or additional public spending. But refusing to offer equal subsidies to equally low-income workers who also have coverage is inequitable. The net cost of insurance will be reduced by the subsidy for those who can take advantage of it. Low-income workers in large firms will see their wages substantially reduced by the cost of their insurance (more than 10 percent of their income), but they will not have access to a state subsidy.
Some believe that lower-income workers in large groups are cross-subsidized by higher-income workers, but except for those constrained by minimum-wage laws, there is no strong support for this assertion. The similar notion that employers pay for the insurance they give their employees is largely an illusion: while the employer surely sends in the check that pays for health insurance, nearly the full cost of insurance purchased in the workplace setting ultimately falls on the workers themselves in the form of lower wages. Employers do not sacrifice profits to help their employees get insurance; rather, workers sacrifice a portion of their potential wages for it. A politically inconvenient corollary of this proposition is that laws compelling employers to bear the burden for coverage will almost surely produce the opposite result: payment for mandated coverage will come out of workers’ future raises.
A more serious concern is that the Massachusetts program’s failure to offer equal subsidies to all similar lower-wage workers could cause employers or employees to drop preexisting efficient group coverage in some settings in order to benefit from the Connector’s subsidies. Furthermore, larger firms might spin off their lower-wage workers to smaller entities (for example, contracting out the mailroom) in order to take advantage of the low-wage subsidy. There are reasons of fairness and efficiency to make lower-income subsidies neutrally available to all lower-income families regardless of how they might otherwise obtain insurance coverage.
Advocates of the Massachusetts plan argue that there are relatively few large firms in Massachusetts with enough low-income workers to make it worthwhile to dissolve the group in order to take advantage of the subsidy. Some of those advocates were the first to point out that President George W. Bush’s timid subsidy to the same set of low-income households would likely cause many groups to collapse as employers would take advantage of the subsidy to give themselves an excuse to stop arranging insurance. Perhaps Massachusetts is different from the rest of the country, but a more plausible explanation for the change in tune is the change in venue. The concern about the threat to groups is overblown, but the virtues of group insurance are also overestimated. The inevitable and potentially substantial long-term distortion of firm size and production arrangements is a greater threat. The only benefit of this distortion is in limiting the cost of the program, but a major drawback is in depriving the most responsible lower-income workers--those who look for and take jobs which provide decent health insurance--of the subsidy that they deserve as much as their less-circumspect peers. Although the cost of raising enough public money to fund all low-income households to the same extent and in the same way must be acknowledged, neutrality is the best policy for the long run.
Connecting the Dots
The most novel and hyped aspect of the Massachusetts program is the Orwellian-named Connector. This entity is meant to offer small groups and individuals four benefits thought to characterize large group insurance: lower administrative costs, a better array of low-cost plans, a more transparent choice of plans, and better risk pooling. The quid pro quo in Massachusetts is that the Connector must observe almost all of the state’s restrictive insurance laws. It is the only game in town for individuals and small groups seeking subsidized coverage.
While the Connector may be able to deliver some of the desired advantages, at least two of them are likely to be small. Lower administrative costs do characterize large groups, but those lower costs occur for reasons that have little to do with the Connector. A large group pays only one sales commission and only one bill, and it has only one human resources department asking questions. Large groups reduce neither the health-care costs of those insured nor the costs of processing claims, although they may reduce some other costs. The Connector, with thousands of individuals and small groups to serve, will have to sell thousands of individual policies; issue thousands of separate bills, reminders, and notices; and answer thousands of individual questions. There may be some reduction in administrative costs, because the subsidy will make for an easier sale, but we know that those costs drop nowhere near the level of large groups.
The other elusive feature of group insurance is its supposed advantage in risk pooling. Individual insurance is actually quite good at pooling the most serious risk: that a person who is healthy today will become chronically ill tomorrow, with high expected health expenditures for the rest of his life. The policy provision in individual insurance in all states that is relevant here is called “guaranteed renewability at class average premiums,” and it requires insurers who sell coverage to a person who is low-risk at the time to promise in the future not to raise premiums based on that person’s changed risk--so-called reunderwriting.
What group insurance might reasonably do over and above this is not really the pooling of risk, but rather risk averaging—charging people at different risk levels the same premium regardless of their age or family composition. The political appeal of offering easy switching at uniform premiums to whatever plan best fits one’s existing health needs is high, but its cost in terms of adverse selection or regulation to prop up cross-subsidies can be enormous. There is very little real gain over guaranteed renewability. Even on its face, pooling in group insurance is not complete. Plans with different premiums for families of various sizes and shapes have proliferated in Massachusetts and everywhere else. While explicit employee premiums usually do not vary with age, the evidence that wages do is strong: to pay premiums, employers take more from the wages of jobs typically occupied by older workers. Massachusetts is one of a few states that regulated individual-market insurers’ ability to vary premiums with risk, but evidence suggests that such regulation has, at best, only small effects in helping higher-risk people, and probably causes more lower-risk workers to reject coverage as a bad deal.
The Connector might offer real value if it can secure better insurance plans for money spent and enable people to make better choices, the latter because more options mean more accurate matches and because the terms of choice are more appropriately bargained and constructed than in most instances of small-firm, employment-based insurance. Both of these goals are hypothetical today. Optimism would be in order were this situation not inherently political, or if alternative health insurance choices were to survive should the Connector eventually revert to its bureaucratic origins. This structure might be used to furnish the terrain for ideological battles about what is good insurance--take, for example, health savings accounts, which some love but others dislike, versus aggressive managed-care plans, with politicized choices likely to be based on clout, favors, or worse. Perhaps the Connector can serve as an honest broker or as a transparent marketplace, but it is a gamble on the untried and the insufficiently explained.
Massachusetts as a Model
The essential features of the Massachusetts plan are praiseworthy. We need serious health insurance subsidies to households with incomes above the Medicaid eligibility level but below median income. We also need some type of penalty for higher-income uninsured people taking a chance on the charity of strangers. The pro-regulatory politics that have historically characterized Massachusetts suggest that the vision of a transparent and dispassionate market, offering a wide variety of insurance coverage options chosen by people on the basis of preference, remains only a vision--in Massachusetts today and in would-be imitators tomorrow. Even if the plan does not collapse before implementation, it will offer lessons from the “laboratory of democracy” which may be of considerable value to other states. States considering a universal health-insurance mandate or major health-insurance coverage expansion should provide more flexibility in the types of plans permitted and greater neutrality in the distribution of subsidies. This will require cutting through the smokescreen of workplace-based insurance and slicing the Gordian knot of good coverage at affordable premiums.
Mark V. Pauly ([email protected]) is the Bendheim Professor at the Wharton School at the University of Pennsylvania and an adjunct scholar at AEI. Editorial assistant Evan Sparks worked with Mr. Pauly to edit and produce this Health Policy Outlook.
1. Thomas W. Grannemann and Mark V. Pauly, Controlling Medicaid Costs: Federalism, Competition, and Choice (Washington, DC: AEI Press, 1983).
2. Mark V. Pauly, Patricia Damon, Paul Feldstein, and John Hoff, “A Plan for ‘Responsible National Health Insurance,’” Health Affairs 10, no. 1 (1991): 5-25; and Stuart M. Butler and Edmund F. Haislmaier, “The Consumer Choice Health Security Act,” Heritage Foundation Issue Bulletin, no. 186 (December 23, 1993), available at www.heritage.org/Research/SocialSecurity/IB186.cfm (accessed November 7, 2006).
3. M. Kate Bundorf and Mark V. Pauly, “Is Health Insurance Affordable for the Uninsured?” Journal of Health Economics 25, no. 4 (2006): 650-73.
4. Len Nichols, “Of Once and Future Kings: The Past and Future of Employment-Based Health Insurance” (working paper, New America Foundation, Washington, DC, November 2005).
5. Jonathan Gruber, “Coverage and Cost Impacts of the President’s Health Insurance Tax Credit and Tax Deduction Proposals” (issue brief, Henry J. Kaiser Family Foundation, Menlo Park, CA, March 2004), available through www.kff.org/insurance/7049.cfm (accessed November 7, 2006); and Mitt Romney, “Health Care for Everyone? We’ve Found a Way,” Wall Street Journal, April 11, 2006.
6. Bradley Herring and Mark V. Pauly, “Incentive Compatible Guaranteed Renewable Individual Health Insurance,” Journal of Health Economics 25, no. 3 (2006): 395-417.
7. Vip Patel and Mark V. Pauly, “Guaranteed Renewability and the Problem of Risk Variation in Individual Health Insurance Markets,” Health Affairs web exclusive (August 28, 2002), available through http://content.healthaffairs.org/cgi/reprint/hlthaff.w2.280v1 (accessed November 7, 2006).