The Cato Review of Business & Government
Marvin H. Kosters is director of economic policy studies at
the American Enterprise Institute.
Income security provided by the government may not extend form the cradle to the grave, but calls for he government to apportion more resources to specific uses along the way have become commonplace. Child care for working people, health plan coverage linked to jobs, long-term care for the elderly, and family leave for workers to care for newborns, ailing children, or parents are among the initiatives now on the horizon. With the federal budget deficit already embarrassingly large and the public apparently unwilling to accept higher taxes, proposals for mandated benefits would have the government supply the mandate, while generally shifting to employers-and ultimately to workers and consumers-the responsibility to finance those benefits.
The push to enact new benefit programs has been emanating from Congress and represents a real departure form the policies pursued during the Reagan administration. Regulatory relief, for example, was a major element of the Reagan economic program. The goal was to reduce the costs imposed on business and increase reliance on private markets. Tax relief was a second major element of the Reagan economic program-leaving more income in the hands of those who earned it. And a third element was reducing the growth of federal spending, often by targeting spending more closely on those with low incomes. Legislative proposals to mandate new benefits would reverse, at least partially, each of these policy initiatives.
The proximate reason for the spate of new congressional initiatives was the shift to a Democratic majority in the Senate after the 1986 elections. Among other significant changes, the chair of the Labor and Human Resources Committee passed form Senator Orrin G. Hatch to Senator Edward M. Kennedy. In addition the federal budget deficit, and pressures to contain or reduce it, have left little scope for major new direct spending programs. Mandated benefit programs have opened a new avenue for responding to constituencies seeking to advance their interests through governmental policy. Constituent services can be expanded without explicitly raising spending or taxes to pay for them.
Federal benefit mandates that have already been enacted include requirements for continued health plan coverage of former workers and their family members, and advance notice (or severance pay) requirements for major layoffs and plant closings. Proposals for increasing the minimum wage and for guaranteeing reemployment for workers taking family leave failed because filibuster threats could not be overcome. These proposals, and others such as health plan requirements, are almost certain to be on the agenda of the next Congress.
While politically appealing, mandated benefit programs raise some important economic issues. For example, are the benefits worth the cost either for the individual worker or for society as a whole? How are individual choices affected? Who ends up paying the costs? And how seriously is job creation discouraged by mandated benefits that raise employment costs? This latter question is crucial to relatively low-wage workers for whom the increase in employment costs would loom large -the group the proposals are ostensibly designed to benefit.
Mandated Benefits and Mandated Costs
The many diverse proposals for mandated benefits have in common a promise of some particular benefit to workers: improved wages, job security, pension plans, leave policies, health care, and so forth. Of course, if only benefits were involved, there would be no need for a mandate. Each mandated benefit implies a mandated cost, however. We can distinguish among the various proposals by looking at the mechanisms used to finance the costs and the opportunities these mechanisms provide for shifting the incidence of the costs.
Some of the oldest government-mandated programs are financed by an explicit tax on labor. Social Security, the best known and by far the largest program, is financed by taxes levied directly on employers and employees. Others, such as unemployment insurance, workers' compensation, and pension fund insurance, are financed by taxes levied on employers. In each case the actual incidence of the tax is likely to rest primarily on workers because the total cost of hiring additional workers includes the taxes paid to finance their benefits. For these programs collecting tax revenues and dispensing benefits are both administered by a government agency.
More recent proposals would mandate benefits that are financed and administered privately. For example, parental-leave proposals create a benefit to which workers are entitled and regulations that define exactly what that entitlement is. Financing and administering the program would be the obligation of the employer. For certain other benefits, such as health-care coverage, the government contributes indirectly to the financing by exempting the costs from the income tax-at least that portion paid by the employer. The primary obligation to provide the benefit, however, remains the empolyer's.
A law may appear benign when it simply requires benefits that is many cases would have been provided voluntarily. But government-mandated programs often displace voluntary arrangements that might have been part of private employment contracts. When the government mandates benefits in instances where the parties would not have agreed to them voluntarily, some parties are bearing an unwanted burden and will attempt to avid it. The analysis of how these burdens effect private behavior, and how private behavior in turn shifts the incidence of the burdens, is in essence an analysis of tax incidence-regardless of whether the tax is explicit or implicit. Benefits that are mandated as condition of employment can be regarded as a tax on employment, even if no government agency directly collects the tax and dispenses the benefits. Analysis of the incidence and indirect effects of such a tax is less straightforward, however, than in the case of a payroll tax that is withheld for all workers.
A key difference between government-financed and employer-financed benefit programs is that government financing may effectively guarantee the benefits with the full faith and credit of the United States. One consequence is that, from the point of view of both employers and employees, benefits and costs may appear only loosely coupled. The cost experienced by firms is determined by the level of taxes the government imposes. Individuals' benefits are determined by the benefit and eligibility regulations. There is no assurance that the government will link these factors for an individual or a firm, or that it will set these factors to keep its fund in balance. Unfunded liabilities may therefore be shifted to other workers, to taxpayers in general, or to taxpayers in future generations.
In employer-financed programs benefits and costs are more closely linked. Even in the case of benefits that are due far in the future (such as health care for retirees), liability for the associated costs wilt be recognized immediately and will affect the performance of markets for goods, capital, and especially labor. Cost-shifting will generally he limited to a firm's customers, stockholders, bondholders, and labor force.
A third type of financing deserves mention here: a direct tax credit to qualifying individuals. While not strictly a mandated benefit (since there is no mandate), this type of financing can be a policy alternative to financing through employers or through a dedicated government tax, as well as to direct government spending. Under one proposal, child-care benefits would be supported mainly by federal spending out of general revenues. A sizable component of the funds would initially be devoted to developing regulatory standards for providers and procedures to administer spending on child-care assistance. Under another proposal, the dependent child-care tax credit would he made refundable to low-Income families, and a new children's tax credit would be introduced for low-income families with children, whether or not the parents choose to work. If a child-care program were enacted, however, it might well include some employer mandates. This article will focus mainly on employer-financed programs, keeping in mind the alternatives of government financing through special funds, general revenues, or tax expenditures.
Proposals to mandate benefits are usually discussed piecemeal in the context of the needs of particular segments of the populace, the quality and availability of particular services, and the affordability of services for those who need them. For the citizenry as a whole, however, devoting more resources to any one use means devoting fewer resources to other uses. That is, enhancing the quality and amount of any particular service entails assigning higher priority, and devoting a larger share of the national output, to that use at the expense of other uses, including spending and investment choices that would otherwise be made by consumers. Benefit mandates cannot, in and of themselves, increase the national output.
The cost of benefit programs already in place accounts for a very significant and growing share of employees total compensation. In 1987 employer-paid costs for health and retirement plans, and social security tax payments by employers and employees, accounted for 23 percent of total compensation, up from 17.5 percent in 1973. In addition workers must often pay a share of the costs of health and retirement plans. Many workers also contribute to flexible spending accounts that enable them to pay health and child-care costs with pre-tax dollars. Mandating additional benefits would raise still further the share of total compensation devoted to benefits, further reducing the share available for workers to spend at their own discretion.
The fact that something must be given up to get more of something else is illustrated by recent trends in employee compensation. During the past 15 years, output and total real compensation per hour of work have increased by about percent. The cost of supplements paid by employers, however, has increased much more rapidly raising the employer-payments share of compensation from 12.8 percent to 16.4 percent. As consequence the wage and salary component c real hourly compensation has been quite stagnant. Employer-paid benefits have absorbed most of the small increase in output per hour of won since the early 1970s, leaving virtually no room an increase in wage and salary income. Devoting a still larger share of the national output to mandated benefits would widen the gap between increases in real output per hour of won and in the wage and salary portion of employee compensation.
Taxing by Regulatory Mandates
Use of law and regulation to produce public benefits as an alternative to direct, tax-financed spending is quite common. Perhaps the best known and quantitatively most important example is pollution abatement regulation. Environmental regulations require firms to pay for equipment to reduce pollution, with the increase in costs passed on to consumers. Workplace safety and health regulations are yet another example of a program financed through requirements placed on businesses.
The intended beneficiaries of these programs have traditionally been broadly defined. Pollution abatement, for example, is intended to benefit the general public. Social insurance and workplace safety and health programs are intended to benefit the entire work force. These programs have traditionally provided benefits that "everyone might be presumed to want.
In contrast the new mandated benefit proposals, which would also amount to
new regulatory programs, are generally intended to provide benefits to much
narrower constituencies Family-leave and child-care programs are cases in
point. Moreover, benefits that some may not want- either at all or in the
form provided by government
would be mandated. And qualification for benefits would be conditional on decisions about work and family life that have hitherto been viewed as matters best left to private choices by workers and their families or by workers and their employers. Accordingly, some of the proposed mandated benefit programs are much more intrusive with
respect to personal and family choices than most traditional programs. They also place important economic benefits outside the realm of bargaining and other informal arrangements for setting pay.
Labor Market Effects
The targeted nature of mandated benefit proposals is part of their political appeal: politicians appear to be making promises to particular constituents at corporate expense. This is easy to do rhetorically but difficult to do in fact. Wealth redistribution cannot be effected through the intermediation of employers nearly as easily as it can through the U.S. Treasury. While a corporation may be obligated by law to finance certain benefits, it must turn to markets to obtain financing, and transactions in those markets remain largely voluntary. Real costs fall on real persons: stockholders, bondholders, consumers, and employees.
To the extent that mandated benefits are contingent on employment and other identifiable factors, so are the associated costs. For instance, workers in a cyclical industry benefit disproportionately from layoff notification requirements. From the company's point of view, the potential cost of notification looms largest in those same industries. For this reason the law inhibits job formation in exactly those industries where it was supposed to enhance job security.
We can expect this pattern to be quite general. Benefits that are contingent on employment will-like any tax on labor-be paid for largely by employees. Parental-leave benefits will handicap prospective parents in the labor market; generous health benefits will handicap those prone to illness; benefits of value only to women will reduce the wages women can command.
These tendencies should not be attributed to perversity on the part of corporate managements. Firms that are exposed to large labor costs will, in the absence of other adjustments, necessarily experience higher prices and lower rates of return. Regardless of what management thinks or does, anonymous consumer markets and capital markets will direct the burden of costs to those factors upon which they are contingent.
One likely political response to these effects is a proliferation of rules against "discrimination.' Of course, it is the government that attempted to provide benefits to one favored group at the expense of others. Markets simply react to that. Young women, old men, large families, part-time workers-whoever is given an entitlement-wi4l feel its burden as well. The difficulties that firms may encounter in administering mandated benefit programs can be illustrated by the requirement under the 1986 tax reform act that they demonstrate equitable distribution on the basis of the value of benefits workers actually choose instead of equality of access.
The effects of the implicit taxes on employment imposed through mandated benefit requirements can be expected to differ markedly among workers at different levels of earnings. Benefits that are mandated for all workers restrict individuals' ability to choose or to negotiate the combination of wages and benefits they prefer. Some might be forced to subsidize benefits for which other workers would be the main beneficiaries, while others might themselves be subsidized by their co-workers. While individual workers with relatively high earnings might prefer a different mix of wages and specific benefits, however, there would be ample scope for substitution between wages and benefits for high-wage workers as a whole. The mandated benefits are likely to be well within the range of what a high-wage worker would choose to purchase anyway. This is not so for low-wage workers.
Impact on Low-Wage Workers
For workers with low earnings, benefit mandates would have a severe impact on the portion of total compensation paid out in wages and salaries, and would likely result in a very different pattern of compensation than they would choose voluntarily. The problem is most severe for workers at the very low end of the earnings scale for whom the value of what they can produce is lower than the combined total of mandated benefits and wages in their compensation package. Squeezed between mandated benefits and the minimum wage, these workers would likely find their jobs in jeopardy. For workers who are able to retain their jobs or obtain new jobs, mandated benefits limit the flexibility of workers and employers to structure mutually beneficial arrangements, inducing employers to invest less in employee training, for example, than workers would prefer.
The problem of mandated benefits for the lowest-wage workers is illustrated by proposals for mandatory health plan coverage. Among workers with no health insurance coverage, more than half are working at hourly wages below 125 percent of the federal minimum wage. One proposal of the 100th Congress would have raised the federal minimum wage by about 35 percent. Mandatory health plan coverage could raise the effective minimum legal level of compensation for full-time workers by an additional 25 percent, for a total increase in hourly compensation of about 60 percent.
Exempting pad-time workers, for whom compensation costs would be increased disproportionately, or exempting small firms, would help to mitigate the adverse effects of mandated health plan coverage on employment. But such exemptions would also leave more people without health plans. According to estimates reported by Dr. Deborah Chollet, senior associate at the Employee Benefit Research Institute, even if all those working at least IS hours per week and their dependents were covered by mandatory health plans-and no job losses resulted from setting such a high hurdle for compensation-only about two-thirds of the uninsured would be added to the rolls. Thus benefit mandates are an extremely inefficient way of transferring benefits to people who lack them, just as minimum wage requirements are an extremely inefficient way of raising the incomes of those at the low end of the pay scale.
Needs and Affordability
Those making a case for mandated benefit programs typically develop estimates of the numbers of people affected. The dimensions of the problem are documented in reports showing, for example, that more than half of mothers with small children are working, millions of people arc not covered by health plans, and growing millions of the elderly will need nursing-home care or other assistance. Data on demographic characteristics, income levels, and costs are widely publicized. Strong growth trends, both for numbers affected and for cost of services, help to convert problems of setting priorities and making choices into national crises. Judgments are usually made by experts that the people who need services are increasingly unable to afford them and, consequently, that the government needs to step in.
The question of affordability is frequently confused by the failure to distinguish clearly between regular and predictable needs and those that arise unexpectedly and rarely but are very costly. Needs that arise in these latter circumstances might only be affordable through insurance arrangements. But regular and predictable needs that are not affordable by particular groups can only be made available by providing income transfers. To illustrate this distinction consider regular payments for mortgage interest, maintenance, utilities, and taxes on a house, as differentiated from the loss that arises from complete destruction of the house by fire. Most people who own a home can afford the regular and predictable costs of home ownership, but cannot absorb such a major loss. They can, however, afford insurance that covers severe losses when the probability of occurrence is small and reasonably predictable. Thus, although most of the elderly cannot afford the cost of serious illness or prolonged nursing care, most of them can, as members of a group, afford insurance to cover such contingencies.
Some proposals for mandated benefits involve elements of insurance against risks, but many are intended to cover either regular and predictable expenses or contingent needs that may not warrant insurance coverage. Child-care benefits, for example, cover regular and predictable costs. In addition, of course, the need for such services is itself contingent on family decisions about childbearing and work. Other kinds of mandated benefits, such as family-leave programs, have apparently not been regarded as warranting insurance coverage by most members of the work force, judging by the fact that they have rarely been included in collective bargaining agreements or other voluntary arrangements between employers and workers.
Affordability of benefits can be improved in only two ways. One way is by developing insurance arrangements to spread risks. For most risks worth covering, insurance is available-or likely to become available-from private commercial providers. For risks that are beset by serious problems of adverse selection or moral hazard that cannot be handled through contract design, government policies might be helpful. In many instances, however, government policies tend to undermine the affordability and availability of insurance by setting minimum standards that preclude low-cost coverage. Requirements enacted by states, for example, regulating who qualifies for reimbursement under health insurance programs suggest that standards more often reflect the interests of provider groups, such as psychologists and chiropractors, than of health insurance consumers.
The second way affordability can be improved-at least for some people-is by shifting the cost of particular risks or by instituting other mechanisms for making transfers. This is frequently one of the effects of a policy and, even more frequently, a major goal of its proponents. Child-care and parental-leave requirements for employers, for example, are likely to impose costs on a much broader group than the direct beneficiaries, it is not at all clear, however, that the net effect of general benefit mandates would be to improve the economic circumstances of workers in low-income families, partly because of the depressing effect on wages and partly because the benefits would often be produced at higher costs than under informal private arrangements.
The need to improve affordability for the poor is often the most persuasive argument used to make the case for imposing benefit requirements on employers. Yet the programs that are developed usually do not limit benefits to the poor. Either the requirements are made applicable to most of the work force or the benefits are phased out only after extending well into the middle class. As a consequence, these programs are very costly, and it becomes essential to scale back their costs. Ironically, the likely result is inadequate benefits for those whose low incomes and benefit needs played a major role in making the ease for benefit mandates in the first place.
Quality and Cost Escalation
When policy discussions turn to mandatory benefits like health insurance, where coverage must meet certain minimum standards, the "minimum" is typically defined in relation to current plans. Any distortions in present arrangements are thus carried over to mandated programs. Although employers have tried to intro duce incentives to discourage excessive usage, the health plans they provide still reflect the fact that their costs are paid in pre-tax dollars. These tax advantages lead workers to prefer relatively low deductible amounts and relatively little cost-sharing. In other words, workers choose health benefit plans that are strongly oriented toward reimbursement for regular and predictable expenses (which would otherwise have to be paid in after-tax dollars), rather than for extremely serious and costly health problems that are experienced only very infrequently. Thus the health benefits provided by most employers are more appropriately characterized as health plans than as health insurance.
Tax incentives have another effect on health plans: they produce a form of quality escalation of plan characteristics. Standards based on these plans would impose this (tax-induced) level of quality on everyone including low-income employees for whom the tax consequences may be far less important. Many of the beneficiaries of such mandated plans might be better off with plans designed primarily to provide insurance against serious risks.
The cost of services provided in response to government mandates is also likely to be higher at each quality level. Mandated benefit programs inevitably bring with them government regulation to specify in detail what is provided. Qualified providers are usually defined in terms of characteristics of facilities, credentials of care givers, and formal licensing of both. Informal arrangements and arrangements within extended families are frequently unable to meet regulatory standards, even when their quality is equivalent or superior to that of institutional providers. Standard setting also tends to limit diversity and to reflect the preferences of care givers instead of clients. Regulatory standards that define "qualified" plans can thus be expected to produce cost and quality escalation.
Federal policy initiatives have increasingly turned to mandated benefits
linked to employment to influence uses of the national output. Because the
benefits are financed by employers, and eventually paid by consumers, shareholders,
and workers themselves, their costs are not reflected in federal spending
and the budget. Costs are
obscured and often bone by a broader group than the beneficiaries.
The primary effect of requiring that more output-and a greater share of employee compensation-be devoted to particular benefits is that less is available for other uses. From the standpoint of employees, this reduces their discretion to choose the mix of benefits and wages that best satisfies their diverse needs and preferences. It also discourages job creation. Mandated benefits tied to employment, which are equivalent to a tax on employment, reduce job opportunities, particularly among those with the poorest earnings prospects who can least afford the benefits proposed.
|Direct government spending is preferable to benefit mandates and the regulation that comes with them.|
To increase the amount and quality of particular benefits to low-income people-whether day care, health care, or pension benefits- government policy should attempt to provide additional income support directly. This would best accommodate the preferences of intended beneficiaries. If the problem for low-income people is seen primarily as one of too little spending for particular uses instead of too little income, tax credits or in-kind subsidies (such as food stamps or rent vouchers) should be considered. Such an approach would leave room for choice while helping to control costs. Introducing new federal spending or tax benefit programs when the federal budget is deeply in deficit is undoubtedly awkward politically. Raising taxes to pay for these programs could also fuel concerns that bigger government, not smaller deficits, would result. Nevertheless, direct government spending is preferable to benefit mandates and the regulation that comes with them.
Benefit mandates restrict the scope for individual choices to reflect differences in circumstances. In particular, they limit opportunities to exercise one of the most basic choices: taking a job, gaining work experience, and thereby adding to the national output-instead of further subdividing its uses. Despite their understandable political appeal, mandated benefit programs are little more than costly efforts to escape the dilemma posed by setting priorities on the uses of our national output.
Subscribe to Regulation