Menu
Tax Notes logo

CRS OUTLINES POLICY CHANGES TO MODERATE EMPLOYMENT EFFECTS OF EMPLOYER MANDATE.

JUN. 15, 1994

94-497 S

DATED JUN. 15, 1994
DOCUMENT ATTRIBUTES
  • Authors
    Gravelle, Jane G.
  • Institutional Authors
    Congressional Research Service
  • Subject Area/Tax Topics
  • Index Terms
    health care, publicly funded, mandated
    health care, publicly funded, legislation
    employee benefit plans
  • Jurisdictions
  • Language
    English
  • Tax Analysts Document Number
    Doc 94-5759
  • Tax Analysts Electronic Citation
    94 TNT 118-89
Citations: 94-497 S

                       CRS REPORT FOR CONGRESS

 

 

             EMPLOYMENT EFFECTS OF HEALTH CARE MANDATES

 

 

                          Jane G. Gravelle

 

                Senior Specialist in Economic Policy

 

                    Office of Senior Specialists

 

 

                            June 15, 1994

 

 

SUMMARY

The possible adverse effects of employer mandates on jobs have been of concern during discussion of the health care proposals, leading to questions about the desirability of certain plans. Concerns about the effects of mandates on jobs have led to proposals for special subsidies for small businesses or to proposals to exclude very small businesses from a mandate. There are three different issues that will be discussed in this study: involuntary unemployment, a voluntary labor supply change, and potential misallocation of labor. Of course, employment effects are only one aspect to consider in evaluating health care proposals.

When discussing job loss concerns, per se, it is primarily involuntary unemployment that is at issue. This involuntary unemployment would consist of two kinds: short run transitional effects and effects on minimum wage workers. Voluntary labor supply responses, since they are based on individual choice, do not involve job losses. They may, however, reflect inefficiency in individual choices (although inefficiency can be associated with both an increase or a decrease in employment). Voluntary labor supply responses are also relevant to the budgetary effects of a proposal as they change the level of taxes and outlays. Misallocation of resources may arise from effects of subsidies in producing differential labor costs for firms. Universal coverage could, however, reduce existing labor market inefficiencies.

Evidence on labor market responses suggests that involuntary labor effects would likely be small, but could be larger without the overall cap. Special small business subsidies do not appear to have a large effect in reducing these estimated effects, perhaps because they are not targeted. Overall empirical evidence on labor supply suggests a small voluntary labor response.

Several potential policy changes might be considered to moderate the employment effects of a universal employer mandate as proposed by the Administration, or improve efficiency. If short-run transitional effects, or effects on minimum wage workers are of concern, offsetting macroeconomic policies or adjustments in the minimum wage are available. Special subsidies that encourage early retirement could be reduced or eliminated. Age-adjusted community rating would reduce involuntary unemployment of minimum wage workers, incentives for early retirement, and misallocation in the labor market (and the need for subsidies). The structure of the subsidies could be altered, by targeting them to individuals rather than firms, and/or by adjusting the current tax benefits. Small businesses might be exempted from the mandate, although these firms might be required to administer the plans. Individual mandates, although they introduce some complications, could be considered, although the firm might be used to administer the plans and withhold payments. A successful individual mandate would probably require some attention to how to structure subsidies to prevent an incentive to drop employer coverage. Mandates might be eliminated altogether and reform directed solely at some form of community rating. Finally, a mandate without community rating might be considered with the government providing subsidized insurance for high risk individuals.

                              CONTENTS

 

 

A DESCRIPTION OF THE ADMINISTRATION'S PLAN

 

 

JOB MARKET ISSUES

 

     INVOLUNTARY LABOR EFFECTS

 

     VOLUNTARY LABOR EFFECTS

 

     MISALLOCATION OF LABOR RESOURCES

 

 

THE FRAMEWORK FOR EXAMINING EMPLOYMENT EFFECTS

 

     VOLUNTARY VS INVOLUNTARY UNEMPLOYMENT

 

     PARTIAL VS GENERAL EQUILIBRIUM

 

     SHORT-RUN, NEAR-TERM AND LONG-RUN EQUILIBRIUM

 

     GROUPS MOST LIKELY TO BE AFFECTED

 

 

EMPIRICAL EVIDENCE AND RECENT STUDIES OF LABOR

 

  EFFECTS

 

     DIRECT EVIDENCE ON MANDATES

 

     LABOR DEMAND

 

     LABOR SUPPLY

 

     STUDIES OF THE ADMINISTRATION'S HEALTH PROPOSAL

 

          Lewin-VHI

 

          O'Neill and O'Neill

 

          Consad Research Corporation

 

          DRI-McGraw Hill

 

          Early Retirement

 

     SUMMARY

 

 

POLICY IMPLICATIONS

 

     OFFSETTING INVOLUNTARY UNEMPLOYMENT

 

     ELIMINATING THE EARLY RETIREMENT SUBSIDY

 

     AGE-ADJUSTED COMMUNITY RATING

 

     ALTERING THE SUBSIDIES

 

     EXEMPTING SMALL BUSINESSES FROM THE MANDATE

 

     INDIVIDUAL MANDATES

 

     ELIMINATING MANDATES

 

     MANDATES WITH GOVERNMENT AS INSURER FOR HIGH COST

 

       PLANS

 

 

APPENDIX: LABOR DEMAND AND SUPPLY

 

 

EMPLOYMENT EFFECTS OF HEALTH CARE MANDATES

The possible adverse effects of employer mandates or payroll taxes on jobs have been the focus of considerable concern during discussion of the health care proposals. Concerns about employment effects have led some to question the overall desirability of certain health care proposals. These jobs issues have also led to proposals for special subsidies for small businesses so that employer payments on behalf of their employees would not exceed certain percentages of salary, or to proposals to exclude very small businesses from a mandate.

Several recent studies have estimated that an employer mandate would cause job losses. These estimates have varied widely -- from a low of less than one-tenth of a percent of the labor force, to a high of over three percent of the labor force. The first number is smaller than the usual number of jobs created in less than a month, while the latter represents two years or more of job growth. These numbers reflect different assumptions with respect to cost, shifting of premiums to workers through wage reductions, presence of subsidies, and the behavioral response of workers. Some analysts have also suggested that the Administration's proposal would actually expand job growth by adding additional jobs to the labor intensive health care market.

Some of the studies of job effects have failed to distinguish between long run and short run effects, do not differentiate between voluntary and involuntary unemployment, and rely on arbitrary cost shifting assumptions that are not well grounded in economic concepts.

Other analyses of the employment effects of health care plans have been concerned with the structural effects on the labor market, including a possible misallocation of labor resources due in part to subsidies offered to minimize any potential adverse overall effect on employment.

The first section of the paper briefly describes the Administration's health care proposal, and mentions some of the elements in the plan that are subject to change (and that have been altered in various congressional proposals). The second section of this paper outlines the policy issues. The third section examines the framework for understanding the employment effects of a mandate. The next section discusses the empirical evidence on these effects, including evaluation of the studies that have been done, and the final section discusses the policy implications.

This analysis takes President Clinton's health care proposal as a point of departure (introduced as the Health Security Act, H.R. 3600/S. 1757). This plan has received the most attention, and the studies that have calculated jobs effects have focused largely on this proposal. There are several congressional proposals that have been made, some of which differ substantially in focus. Several congressional committees are also considering, or have considered, health care proposals. Various revisions discussed in the last part of this paper have, in fact, already been incorporated into some proposals. Because of the many different proposals that have been introduced and the frequent legislative developments, this study does not attempt to discuss and analyze proposals per se; rather, the focus is on the effects of basic features. 1

A DESCRIPTION OF THE ADMINISTRATION'S PLAN

The President's health care proposal has several features that could influence employment and work. Before discussing those features, however, it might be useful to explain the problems in the health care market that may lead to a need for Government involvement and that help explain why certain approaches, which may have implications for employment, have been taken.

There are several fundamental problems with health care and health care insurance. The purpose of insurance is to protect the individual from the financial burden of a serious illness. The first problem arises from imperfect information -- since insurance companies do not know enough to separate less healthy from more healthy individuals, they tend to set a fixed price for certain groups (e.g. ordered by age or sex). This problem leads to ADVERSE SELECTION -- the tendency for healthier individuals to leave the market. This effect raises the price of insurance further and tends to again drive individuals out of the market, a process that in theory could continue until no health care insurance is provided. Of course, some health insurance remains because of the desire of individuals to insure against risk even without actuarially fair insurance, but such insurance is likely to carry a high price, and many individuals are not covered.

Of course, even if insurance companies had perfect information, individuals who are most likely to get sick could not get insurance that would be worth much to them -- insurance would only be available for random hits that insurance companies could not predict. This type of market outcome may not be what society in general is looking for; rather the preference might be to spread the costs across all individuals even if the less health individuals have been identified.

Some compensating mechanisms have developed in the market -- many employers offer insurance to their employees as a group benefit, which can overcome the natural workings of the market. And, society has provided for coverage for the elderly and the poor through Government programs. Charity care is also provided by many hospitals. The existence of charity care and Government programs for the poor, however, further encourages individuals to go without private insurance. Moreover, some poor individuals may not be able to afford coverage, particularly if they are in poor health or in a category that would be subject to higher premiums.

Another problem with health insurance is the need to control costs, which have been rising rapidly. This cost containment problem is also a serious one and one that occurs in a market where, for a variety of reasons, normal price signals do not work well. It is relevant to labor market effects in the sense that success in containing costs will lead to smaller premiums.

There are several important aspects of the Administration's health care proposal that should be kept in mind in assessing the labor market effects.

First, the plan aims at UNIVERSAL COVERAGE, so that all individuals would have health insurance. Secondly, the plan provides for COMMUNITY RATING, so that any insurer would charge all individuals and firms in an area the same premium for a benefit package without regard to individual health differences (or other characteristics except for family size). These plans would be channeled through regional alliances. Large firms could elect to form corporate alliances or self insure; insurers could experience-rate coverage sold to a corporate alliance. Thirdly, the plan would involve an EMPLOYER MANDATE which would require all employers to offer health insurance and pay 80 percent of the cost for their workers. 2 The requirement that employers pay for part of insurance applies to each worker, so that each employer of a married working couple would pay for insurance, but teenage and certain other dependents would not be covered separately. 3 Not-employed individuals who lack a workplace-covered spouse are subject to an individual mandate.

The plan has several types of SUBSIDIES that are potentially relevant to employment effects. First, there is a subsidy to ensure that no employer in a regional alliance would have to pay more than 7.9 percent of wages in premiums. Second, certain smaller firms whose average wages are low enough would have even larger subsidies to pay for lower limits on premiums. Third, the plan would include individual subsidies for low income workers to help pay for their share of insurance and to low income nonworkers to pay for overall purchase. Finally, there is a special subsidy that provides that the government would pay for eighty percent of medical insurance costs by early retirees (age 55-64) regardless of income (except for very high income levels).

There are several measures to contain costs which will ultimately affect the magnitude of any mandate, and hence any employment consequences.

The plan also retains most of the current tax benefit for employer provided plans, which allows the value of health insurance premiums to be excluded from income for tax and social security purposes, and extends this benefit to individual purchasers.

As noted earlier, several committees of Congress are considering the plan and there are a number of other congressional proposals. One difference between the Administration plan and some alternative proposals or revisions is the employer mandate. Proposals have been made to exempt certain small firms from the employer mandate. Some alternative plans would have only individual mandates, leaving employers the option to offer but not necessarily contribute to plans. Still other plans do not have mandates, but provide more limited changes in the health insurance market such as restrictions on varying the premium due to health status of the individual. Different proposals also vary in the effects their plans would have across individuals of different incomes and age groups, because of the mix of premium subsidies offered.

Since most attention, and most studies, have been directed at the Administration's plan, this report largely focuses on that plan. The issues surrounding potential modifications or different approaches will be discussed in the final section.

JOB MARKET ISSUES

The issues of policy concern are not always spelled out clearly in the analysis of the labor market effects. There are three different issues that arise from the possible influence of the health care plan on labor. These include involuntary unemployment, a voluntary labor supply change, and effects on the misallocation of labor across industries.

INVOLUNTARY LABOR EFFECTS

Voluntary unemployment occurs when individual workers decide to leave the labor force because they would prefer leisure or unpaid activities to paid work. (Some temporary voluntary unemployment occurs if individuals decide to leave work to search for another job). Involuntary unemployment occurs when individuals are laid off from their jobs, or not hired, even though they would be willing to work at the same or lower wage. (Involuntary unemployment can also occur when workers are not able to work full time even though they would be willing to work full time.) The concern about job loss is really a concern about involuntary unemployment, since voluntary labor effects derive from the free choice of the employee.

In general, there are two circumstances in which a tax or mandate might create involuntary unemployment. The first arises only in a short run framework with business cycles. This short run is typically characterized by the possibility of involuntary unemployment because of wage rigidities. One of the reasons cited for the existence of business cycles and periods in which the unemployment rate rises temporarily is the likelihood of sticky wages. That is, there appear to be some difficulties with lowering money wages, perhaps because of explicit or implicit contracts with employees. (There may not be such difficulties in lowering real wages via the mechanism of an increase in inflation). As a result, when demand falls, rather than wages falling to reduce the costs of production, the number of employees falls, reducing the quantity of production.

This scenario is usually derived from a fall in aggregate demand. A similar phenomenon might occur with mandated employer- provided health benefits, if employers find it difficult to lower nominal wages or slow the rate of wage growth. In that case, employers' costs would go up. This effect might show up in reduced profits. Firms might, however, respond by trying to raise prices and that action might in turn lead to reduced sales and layoffs. (The extent of this effect depends on the extent to which a firm's direct competitors are also subject to increased costs, a likely possibility in many cases, and how substitutable the industry's product is for those other goods whose prices do not rise as much.) Health care mandates may be less likely to be subject to the effects of wage rigidities because employees are seeing a benefit in return. Moreover, because many of the firms affected are smaller ones, they might be more amenable to informal agreements to reduce salaries as compared with larger firms where explicit labor contracts exist.

The second circumstance when a mandate might produce involuntary unemployment is when there is an institutional constraint that prevents firms from hiring workers and paying them based on the value of their production. This institutional constraint is the minimum wage, and, other things equal, the employer mandate can create involuntary unemployment for workers at or somewhat above the minimum wage, in both the long and the short run.

Generalized short run involuntary unemployment is, essentially, a transitory issue. It is an issue to be concerned about, but any adverse effects could be softened by phasing in a plan or could be offset by other macroeconomic policies, such as an expansionary monetary policy.

The effect on minimum wage workers is of some concern. In setting a minimum wage, society is trading off unemployment among minimum wage workers for a higher level of income among those who are working. Depending on the shape of the demand curve, such a policy may be largely creating implicit redistribution policies (as wages of other employees and capital incomes fall to accommodate the higher wages paid to these workers) or largely creating lost output and income, as unemployment increases. While a full evaluation of the minimum wage is beyond the scope of this paper, it is perhaps worth noting that it might be questionable to reject an otherwise desirable health plan because of unemployment effects created by a minimum wage policy that could be changed. There may, however, also be some potential revisions in the health care plan that might ameliorate this problem.

VOLUNTARY LABOR EFFECTS

A voluntary exit from the labor force presents a different set of concerns from an involuntary exit. Note first that there are two possible labor responses -- a decision to leave or remain in the work force (the participation decision) and a decision to change the numbers of hours worked. (A labor supply effect does not typically distinguish between these two effects). Thus a labor effect may be in the form of changes in hours rather than changes in jobs.

In addition, the direction of a voluntary labor effect is not clear. The magnitude of the labor supply elasticity depends on whether there are both income and substitution effects. That is, when a tax is imposed, the tax both discourages work (by changing the relative price of leisure and consumption, and making leisure more attractive), and encourages work (by reducing income which tends to reduce the consumption of all goods including leisure, and hence increasing work).

There are essentially two reasons that the Government might be concerned about a voluntary decision to change labor force participation or to change hours. First, if a Government program alters incentives through taxation or through mandates that are not the true equivalent of fringe benefits, 4 the plan is creating distortions in individual choices between consumption and leisure. It causes individuals to make less than optimal decisions. This efficiency loss must, of course, be traded off against any potential efficiency gains from the plan, but they are nevertheless issues which have policy concerns.

One must consider efficiency issues carefully. Note that such efficiency losses tend to be measured as only a small fraction of labor income and, therefore, cannot be measured by changes in labor input or "voluntary job losses." Efficiency losses can occur when employment rises (if the income effect dominates) and efficiency gains can also occur when employment falls. For example, to the extent that individuals choose early retirement because individual insurance simply becomes available, that change may be efficient because it is allowing individuals to make decisions which are not distorted by market failures due to adverse selection or unnecessary administrative overhead.

Another reason to be concerned about voluntary labor supply is that it can affect the Government budget. If the labor supply increases, income and tax revenues increase, while if the labor supply decreases these revenues will fall. Such behavioral effects are important in evaluating the budgetary viability of a proposal.

MISALLOCATION OF LABOR RESOURCES

Under current health insurance institutions, there is a distortion in the allocation of workers since health insurance coverage, for a variety of reasons, tends to be provided primarily by medium size and larger firms. The purchase of individual insurance is often not an attractive alternative because of the adverse selection problem mentioned earlier (the tendency of healthier individuals to opt out of insurance leaving the pool of insured individuals less healthy than average) which raises the price, because of administrative overhead, and because of a lack of tax benefits. As a result, individuals work for firms not entirely on the basis of their productivity, but partly because of their preferences for health insurance. There is an efficiency cost that accompanies this misallocation of labor.

A related efficiency issue is the issue of "job lock." Because health insurance under the current system is largely associated with employment, there is a tendency for the system to discourage labor mobility and lead to inefficiency. In general, the evidence on the magnitude of this issue is mixed, but greater labor mobility is generally thought to be a benefit of universal coverage. 5

While universal health coverage could address these causes of inefficiency, it has the potential for creating another, potentially more serious, distortion. Because of the subsidies that produce overall caps on insurance in the Administration's plan (7.9 percent overall for firms in regional alliances, with lower caps for smaller businesses), the cost of hiring a given worker could vary enormously across firms. Some firms will be eligible for subsidies and thus subject to the cap. Others, either because they are not in the regional alliances, or because their average salaries are high enough, would not be affected by the cap. These differences will affect the cost of hiring different types of workers. For example, suppose that the cost of health insurance for a firm not subject to a cap is $2000. Now consider a firm that is subject to the cap. If the worker has a salary of $100,000 the additional cost in health insurance allowed will be 7.9 percent, or $7,900. If the worker has a salary of $15,000, the cost is $1185. Clearly, there is an incentive for a firm subject to a cap to find lower-paid workers relatively cheaper and a firm not subject to the cap to find such workers more expensive. Similarly, the firm without caps finds higher wage workers much less expensive than the firm subject to a cap. This discrepancy is further magnified in the case of small firms (under 75 employees) who have even lower compensation caps.

There will be considerable pressure for sorting of workers to occur -- for aggregating lower income employees in firms subject to a cap and higher income employees in firms not subject to a cap. There might also be some tendencies for firms to contract out services to accomplish this sorting effect. These pressures produce inefficiency and productivity loss since they encourage firms to organize in ways not consistent with maximum productivity. It also is likely to increase the amount of subsidies provided by the government through caps, and increase the overall cost of the program. (The increase in subsidies is also, however, likely to ameliorate the effect on minimum wage workers, but its overall effect on labor supply is unclear since additional budgetary costs would need to be made up elsewhere). 6

The proposal could also affect the use of contingent and part- time workers in a variety of ways, due to the new rules regarding coverage. 7

THE FRAMEWORK FOR EXAMINING EMPLOYMENT EFFECTS

Before reviewing the empirical research and the studies that have examined job effects, it is important to develop a basic framework for considering employment effects. This framework can differ on several important dimensions. First, are the changes in employment voluntary or involuntary? Second, is the analysis "partial equilibrium" or "general equilibrium"? Next, what time frame is being considered? The latter two questions are closely related. Finally, some groups that are more likely to be affected by the mandate are identified.

In this discussion, we consider the effects of a tax on labor income (such as a payroll tax) because it is similar to a mandate. Nevertheless, it is important to recall that a mandate is different from a tax because (1) it is accompanied by direct benefits, and (2) it may not have marginal effects on hours of work. The consequences of these differences will be discussed subsequently.

VOLUNTARY VS INVOLUNTARY UNEMPLOYMENT

It is important to distinguish between measures of involuntary and voluntary labor supply in discussing empirical results because, as discussed above, the concerns about these effects are different.

PARTIAL VS GENERAL EQUILIBRIUM

A partial equilibrium analysis looks at the consequences for a change that is confined to a single firm or a small enough segment of the economy that, for practical purposes, the behavior of that segment will not influence wage rates and other prices in the economy. A general equilibrium analysis is appropriate for a tax that applies to the economy as a whole; in a general equilibrium analysis wage rates and prices in the economy will change in response to a tax. Indeed, holding the aggregate price level constant, wage rates and rates of return will fall in response to a tax.

Because of the broad reach of the mandate, a partial equilibrium approach is generally not appropriate. It is not appropriate, for example, to take the job effects at a typical firm when no other firm is affected and then multiply it by the total number of firms to get an estimate of an economy-wide measure. Nevertheless, some estimates of labor demand that have been used in discussions of health care mandates are taken from cross-section estimates that reflect a partial equilibrium model.

One important aspect of a partial equilibrium model is that employment is not affected by a labor SUPPLY response, because of the assumptions that wages are fixed. Rather, the effects depend on labor demand response, or how firms respond to higher wages. This response by firms depends, in turn, on how easily labor and capital can be substituted in production and how easily other products can be substituted by consumers. In a general equilibrium model, labor supply response -- how workers change their willingness to work in response to their wages -- is crucial, except in those cases where wages cannot change because they are restricted by law (as in the minimum wage). (See the appendix for a derivation of these employment response measures, which are also discussed subsequently).

SHORT-RUN, NEAR-TERM AND LONG-RUN EQUILIBRIUM

The consequences of any policy will depend on the time frame being considered. These time frames, at least in theory, are quite different for a partial equilibrium model than for a general equilibrium one.

In the case of the partial equilibrium model which examines the effect on a single industry too small to affect other prices, the wage rate and rate of return required for new capital are fixed. In the short run, or near term, the firm is also characterized by a fixed capital stock -- that is, it cannot alter its existing capital stock, only its labor supply. In that model, the factor that drives labor demand is the ability to substitute between labor and capital in the production process. In the long run, the firm can increase or decrease its capital stock as well, although adjustment may take a number of years. In that model, the labor demand response is determined by both the ability to substitute between labor and capital and the ability of customers to substitute other products for the firm's output.

The case of the general equilibrium model is more complicated. One may think of three different time frames, although the first two may overlap somewhat -- the short run with wage rigidities that has already been discussed, the short run where wages can change but the capital stock is fixed (which we will call the near term), and the long run. (The following discussion abstracts from the minimum wage issue and would be applicable to an economy without institutional wage constraints.)

The near-term framework might be described as a period when wages are able to fall in response to the tax, but the capital stock is more or less fixed. In this case, there is no involuntary unemployment (except in those cases with institutional constraints such as the minimum wage). Employment may change because of a voluntary response to a change in wages, and both supply and demand factors will play a role. There is, of course, some overlap between the near term and the short run.

The long-run steady state occurs after full adjustment to the change has occurred, and the capital stock adjusts as well. In the long run, all capital is ultimately derived from labor earnings, and if employment changes an eventual change in the capital stock will occur and aggregate savings in the economy will change. Aggregate savings can change simply because the level of income changes and can also change if the induced effects on rates of return affect the savings rate. The full adjustment to the steady state will typically take many years, however. If there is only one sector, the effects will depend solely on the labor supply response.

The general equilibrium calculations would be more complex if derived from a multi-sector economy, in that they would likely include other effects. Nevertheless, the effects will still be driven heavily by the labor supply response (and the long run effect would be dependent only on that response if a uniform labor tax were imposed). For a tax imposed on a single sector, of course, the effects on labor would be influenced by other factors, but if the labor supply elasticity is small, the overall effects will be small. In general, for a partial tax, the effects on labor in THAT INDUSTRY would be smaller if there are no close substitutes in factors of production and if there is no easy substitution between the products produced by the industry and those produced by other industries as identified in the partial equilibrium calculation.

GROUPS MOST LIKELY TO BE AFFECTED

Any employment effects will occur primarily in those firms and industries that do not already have health plans for their employees. (Firms with plans may find their plans becoming more or less costly.) Firms without health plans tend to be smaller in size and to be in certain industries. According to the Congressional Budget Office, 94 percent of firms with over 25 employees have health insurance, while only 39 percent of firms with less than 25 employees have such insurance. 8

One group of workers that is likely to be affected by an employer mandate is the group at or slightly above the minimum wage. Since employers are constrained (by the minimum wage law) from lowering wages, this group of lower-income individuals may have their employment possibilities affected. The effects on this group are moderated by the presence of subsidies in the Administration plan that limit the cost of insurance for employees in qualifying firms. They are exacerbated, however, by the provision for community rating without adjusting for age. Since lower income workers tend to be younger, the averaging of premiums across all age groups will tend to raise their premiums above what they would be if insurers were free to adjust for age.

At the same time, there is an offsetting force that would cause lower income individuals to voluntarily increase their labor supply. That occurs because the universal availability of health care coverage, especially if there are premium caps, may make individuals who are currently on welfare programs more willing to take jobs. One of the implicit costs of leaving the welfare system is the loss of Medicaid benefits.

Other groups that might, according to evidence, be affected because they have larger than average labor supply response elasticities are secondary workers (second earners in a family, primarily wives). In addition, older workers may be more flexible in their work decisions, since retirement decisions may be influenced by health care costs. 9 These effects are less likely to derive from employer mandates, however, than from other aspects of the Administration proposal. The first is the special rule that provides for Government payment of 80 percent of the health care premium for eligible early retirees. The second is that the system is likely to make individually purchased insurance cheaper because of the reduction in adverse selection problems (i.e. the tendency of healthy individuals to opt out of the insurance market) and provision of community rating with no age adjustment.

For many higher income primary workers already covered by health care plans, and not likely to be flexible about hours of work or labor force participation, the proposal would be unlikely to have any effect.

Finally, recall that the proposal under discussion is not a tax, but rather a mandate. How this mandate corresponds with the theory surrounding the economic effects on the labor market of a tax will vary from one circumstance to another.

For secondary earners, the mandate may seem more in the nature of a tax, since such workers could be covered by their spouse's insurance if they left the work force. It would also be in the nature of a tax for purposes of changing hours for any worker if the individual's employer is eligible for subsidies. Only in that case would a change in the wage bill lead to a change in premium payments. (If there is no employer subsidy, then a change in hours would not affect premiums paid by the employer and thus there would be no marginal effects.) The mandate could also produce an implicit tax if the individual is eligible for an income based subsidy. As the subsidy is phased out as income rises due to increased hours of work, the economic effect is like that of a tax.

For the single worker or the only worker in a family, the nature of the mandate depends on whether individual purchase of insurance would be subsidized if the individual left the work force. For workers with adequate non-wage income, there is still an obligation to purchase individual insurance, which makes the mandate more like a fringe benefit and less like a tax. That is, the payment of premiums and the receipt of benefits is not dependent on work. Many of these workers, however, are likely to be of retirement age; in this case it is not the employer mandate per se but other features of the proposal that might cause the employer mandate to be seen as a tax.

For lower income individuals who would be heavily subsidized in the purchase of individual policies, the mandate is like a tax. At the same time, the general availability of universal coverage and the presence of significant employer and individual subsidies would encourage work effort for those currently on welfare.

EMPIRICAL EVIDENCE AND RECENT STUDIES OF LABOR EFFECTS

The previous section describes the theory and framework for evaluating job effects of health care mandates. What evidence might be brought to bear on the subject? The first subsection discusses the direct evidence on responses to already enacted mandates. Then evidence on labor demand is examined, particularly focusing on the effects on minimum wage workers, and also on the underlying forces that drive labor demand. Next, is the evidence on labor supply which plays a crucial role in affecting any voluntary labor response. Finally, the recent studies assessing the labor effects of the Administration's health care plan are discussed.

DIRECT EVIDENCE ON MANDATES

One type of evidence would be to examine the effects of other mandates (or tax financed plans) directly. Unfortunately, there is relatively little literature using this approach.

Gruber and Kreuger 10 in a study of employer-mandated workers compensation insurance found that most of the mandate was shifted back in wages (about 85 percent). If most of the mandated costs are passed back in lower wages, then the employment effects would be quite small. The authors include in their paper a review of previous literature that shows mixed results with respect to the incidence of payroll taxes, a related phenomenon.

A study by Gruber and Hanratty 11 found, contrary to expectations, that the introduction of National Health Insurance in Canada did not cause a significant fall in employment. In fact, there was actually a rise in employment at that time. The authors found some evidence that there was an increase in sectors with high initial private health insurance, and also some evidence that there were decreases in employment in Provinces that used general tax revenues rather than lump sum payments to finance the plan. (A lump sum payment does not vary with wages or hours worked. A flat insurance premium is a "lump sum" payment. The proposed U.S. health plan is a hybrid between lump sum payments and payroll taxes because of the cap on premiums.) The Canadian system, however, is quite different from the Administration's proposal and its effects, therefore, may not be relevant.

LABOR DEMAND

There has been a good deal of research examining the effect of minimum wage legislation and the implied elasticities of labor demand. An elasticity is the percentage change in quantity divided by the percentage change in price. Thus, a labor demand elasticity is the percentage change in labor demanded, divided by the percentage in the wage (inclusive of tax) paid by employers. It is presumed to be negative. (This measure may depend in turn on two other elasticities: the factor substitution elasticity and the price elasticity.) 12 Since the wage is largely fixed in the minimum wage case, the labor supply response plays little or no role.

On the whole, the literature in this area shows a very small employment effect, with the upper limit being an elasticity of -0.3. 13 (That is, a ten percent increase in wage would lead to a three percent decrease in employment). Of course, it is likely that such elasticities would be smaller in the short run than in the long run. These studies appear to be based on data that might appropriately generate short-run elasticities in some cases and long-run elasticities in others.

The literature on labor demand elasticities in general (i.e. not confined to minimum wage workers) is quite complicated because it is not always clear what model is being considered. Those estimates that have been done, apparently based on the long-run, partial-equilibrium model, have showed relatively small elasticities, well below one, as well. 14

The short-run, general-equilibrium labor demand elasticity should depend on the ability of the firm to substitute capital and labor in production; in the short run, it is unlikely that capital and labor can be easily substituted, because such substitution usually requires a change in the technology of production. This theory suggests that the firm's response to wage increases is likely to be small in the short run for all workers. In the long run labor and capital would be more easily substitutable, but as we move towards the long run it is increasingly the labor supply rather than the labor demand function that is important. This view of the factor substitution elasticity is consistent with at least some empirical evidence which shows a very small effect from time series estimates that reflect short run effects and a larger effect from cross section studies that are more likely to reflect long run effects. 15

LABOR SUPPLY

Extensive research has been done on the labor supply elasticity. (The labor supply elasticity is the percentage change in labor supplied divided by the percentage change in the wage (net of tax) received by the employee). In the case of the labor supply of men, which can likely be taken as a proxy for primary workers, evidence suggests extremely small elasticities, which can be either negative or positive (substitution effects are always positive, income effects are negative, and thus the sign depends on the relative magnitude of the two). 16

It has been conventional wisdom that the labor supply elasticity for secondary workers (wives) has been large, perhaps larger than one, but more recent research suggests that it is smaller. 17 The labor supply response for wives is extremely difficult to estimate statistically because there are many women who do not participate in the labor market and there are no direct observations on the wages these women would have received. At this point, the magnitude of the response of secondary workers is very uncertain.

Finally, the evidence on the effect of wages on the retirement decision is mixed -- some evidence suggests that the effect is significant while other evidence indicates that retirement decisions are not very responsive to wage levels. 18

STUDIES OF THE ADMINISTRATION'S HEALTH PROPOSAL

Several studies have reported estimates of job losses arising from the President's health care proposal. A recent study by O'Neill and O'Neill has estimated job losses without subsidies to be 2.1 million, while the loss with subsidies would be 500,000 to 900,000. 19 Given the current labor force of 130 million, this ranges from less than one half of one percent with subsidies to close to two percent without subsidies. A study by Lewin-VHI Associates estimates (with subsidies) a loss of approximately 130,000 to 280,000, a much smaller number than that of O'Neill and O'Neill and amounting to one to two-tenths of one percent of the labor force. 20 DRI/McGraw-Hill estimated a job loss of about 300,000 for the President's health care plan, around two-tenths of one percent of the labor force. 21 A study by the Consad Research Corporation estimated a loss of 850,000 from the President's plan with subsidies, and a loss of 3.8 million without subsidies. 22 The Congressional Budget Office has suggested, without further elaboration, that the loss might be from one quarter of a percent to one percent of the labor force, an amount that apparently included a substantial component of retired workers. 23

The DRI-McGraw Hill study was presumably done on its macro- model, which is a large scale model originally developed for short term forecasting. Its framework is one which allows generalized involuntary unemployment. The other three studies -- Lewin-VHI Associates, O'Neill and O'Neill, and the Consad Research Corporation used essentially the same methodology. All three used a fairly straightforward calculation where employment effects are the product of a labor demand elasticity, the degree to which the tax/mandate is not passed back in wages, and the size of the premiums. That is, as a simple formula, the change in employment would be measured as:

Ed * (Premium/Wages) * (1-p) * Current Employment

where Ed is the labor demand elasticity and p is the share of the premium that is passed back in wages.

Note that this formula does not always bear a straightforward relationship to the theoretical elasticities developed in the appendix and none of the studies actually discuss the labor supply elasticity that should play an important role in determining voluntary employment effects. The premium divided by wages is, of course, the equivalent tax change. The level of employment is necessary to translate a percentage change into the number of jobs. But it is the remaining factors Ed times (1-p) that produce the total labor elasticity (percentage change in labor employed divided by the tax/mandate as a percent of wage). Since the employer mandate is like a "tax" that applies to a large segment of the work force, it is an appropriate to use a general equilibrium model. In the general equilibrium model, in the short run with a fixed capital stock and a single sector, the relationship (1-p) is reflecting the influence of the labor supply effect.

The differences in the various estimates are due to different assumptions about these three effects as well as to potential differences in the employment base. They also depend on whether involuntary or voluntary labor effects are being considered. None of the studies appear to consider a potential expansion in employment in sectors that are already providing health insurance and whose costs might fall as a result of the proposal. The effects appear on the whole to reflect a shorter term model (but not a cyclical one), rather than a long-run steady-state.

Lewin-VHI

First, the Lewin-VHI numbers assume that costs will be passed back to workers, and examine the effects when the minimum wage constrains this ability. (They discuss, but do not provide calculations of, the possibility of a labor supply response and a voluntary change in the labor supply). In this sense, their analysis is probably the most consistent with the theoretical analysis above, because in this case the labor supply response is probably not very relevant. In addition, there is a set of literature on the labor demand elasticity for minimum wage workers, which suggests a general magnitude.

Their study examines the proposal of the Clinton Administration with the caps on the payments as a percentage of payroll. They use two assumptions of demand elasticities -- 0.2 and 0.5. These elasticities are a bit high compared to the demand elasticities reported above, but the authors argue that those elasticities include individuals not affected by the minimum wage. Using their own estimates of premium increases and the minimum wage limits to passing them back in wages, they estimate, for the two elasticities, job losses of 155,000 and 350,000 respectively including government; and losses of 130,000 to 285,000 in the private sector for 1998. These amount to about one-tenth to three-tenths of employment.

O'Neill and O'Neill

This study performed a number of different types of calculations. First, however, consider the calculations that are restricted to the effects on minimum wage workers that are comparable to the Lewin-VHI figures. These authors used labor demand elasticities of 0.1, 0.3, and 0.5. They also used two different measures of premium costs, those done by the Administration and those by Foster-Higgins, an employee benefit consulting firm. On the whole their estimates are higher than the Lewin-VHI ones, even when adjusted to comparable elasticities.

Using the Administration's estimates of premiums, and interpolating, they obtain a 272,000 loss at a 0.2 elasticity and a 680,000 loss at a 0.5 elasticity. These differences may reflect differences in the estimates of the number of workers at or near the minimum wage and how much these wages can fall. They may also reflect differences in the year in which the estimates are made -- O'Neill and O'Neill calculate effects for 1993, when the minimum wage is higher in real terms, while Lewin-VHI look forward to 1998, while leaving the nominal minimum wage fixed (and thus the real minimum wage is lower). These calculations do illustrate how the minimum wage constraints will be automatically loosened if there are no increases for a few years.

They obtain somewhat higher estimates for the Foster-Higgins premium calculations. Again interpolating to calculate results for an elasticity of 0.2 and 0.5, they obtain losses of 350,000 to 875,000 jobs.

They also calculate the effects of varying the caps. Using the Administration premium numbers, and a 0.2 elasticity, they obtain job losses of 328,000 rather than 272,000 if the special subsidies for small businesses are eliminated. Using the Foster-Higgins numbers and a 0.2 elasticity, the loss would rise from 350,000 to 400,000 with the special small business subsidies. If there were no subsidies at all and thus no premium caps, the losses would increase to 648,000 (Administration) and 926,000 (Foster-Higgins). These calculations suggest that it is not so much the small business subsidies, but the overall subsidies (the 7.9 percent limits) that restrain the effects of the proposal on the minimum wages. (To obtain results for the 0.5 elasticity, multiply all of the numbers in this paragraph by 2.5; to obtain results for a 0.1 elasticity divide all of the numbers by 2).

The authors also calculate some additional employment effects given that wages are not shifted back to the full extent of the legal ability to do so. These job changes would be voluntary changes, and they are relatively small. They range from (again, using the 0.2 elasticity) 150,000 for the Administration estimates and 172,000 for the Foster Higgins estimates (with the small business subsidies). With only the overall subsidies, and without the special small business subsidies, the numbers are 172,000 and 192,000 respectively.

It is difficult to know how meaningful this last set of estimates is, given that the method is not carefully grounded in theory and does not rely on empirical estimates of labor supply response. It is only their estimates of the minimum wage effect that is consistent with theory, with empirical evidence, and with measuring involuntary labor effects. But, these numbers reflect in general a view that most of the premiums will be passed back to wages, which is consistent with the evidence of low labor supply elasticities.

Consad Research Corporation

The Consad report considers only an elasticity of 0.2, but includes both an overall elasticity and an industry specific elasticity that averages to 0.2. The latter produces larger estimates, but it is not clear what the basis for varying elasticities is. Thus, we report here the numbers with the uniform elasticities.

The Consad report also uses several measures of premiums, but their preferred case is the Congressional Budget Office estimates, and those are reported here. They estimate, confining the numbers to the minimum wage effects, a job loss of 470,000, somewhat higher than that estimated by the previous study. The year is apparently 1993, the same as for the O'Neill and O'Neill report. Their estimate with no premium caps is much higher at 2.3 million. It is difficult to know why these numbers are so much higher than the O'Neill and O'Neill estimates.

They also do two other estimates. One is an upper-bound estimate for the labor effect (effectively assuming the wage net of the mandate is fixed), which yields a job loss of 824,000. This number is about 350,000 in excess of the minimum wage effects. The second is an assumption of partial adjustment, which is 682,000, which is about 200,000 in excess of the minimum wage effect. These numbers could rise to 2.9 million and 2.7 million respectively without employer subsidies. (The numbers that are highlighted in their report are those with a partial adjustment but variable elasticities, which produce an estimate of 850,000, which would be 3.8 million with no employer subsidies).

As in the case of the O'Neill and O'Neill study, it is impossible to determine what meaning these latter estimates have for employment since their estimates are not embedded in a model or derived from empirical research. It is only their estimates of 470,000 for the minimum wage that is consistent with theory, with empirical evidence, and with measuring involuntary labor effects. This study appears to produce the highest numbers of the three.

DHI-McGraw Hill

The DRI study was presumably simulated on the DRI macroeconomic model of the economy, a model that is based on aggregate demand models with business cycles. There is very little detail about this study, and it is difficult to know how to interpret it.

Early Retirement

Some studies have focused specifically on the potential effects of the Administration proposal on early retirement. There have been studies that relate retirement decisions to retiree health coverage, but these studies have found mixed effects. 24 The Administration has indicated an increase in retirees of between 350,000 to 600,000, and the CBO has suggested an estimate close to the upper end. The Administration has argued that this effect is mostly because of the universal availability of community rated insurance, along with income subsidies, and not primarily because of the retiree subsidies.

SUMMARY

There is relatively little evidence to bring to bear about the short-run effects of employer mandates due to sticky wages, but the studies do provide some insight. The upper limit estimates in the Consad report suggest the magnitude of effects if wages were completely rigid downward, but even these estimates do not appear to be very large. One would expect, therefore, somewhat modest short run effects with the subsidies and also because there is some likelihood of wage adjustment even in the short run. These effects could probably be further moderated by adopting an offsetting expansionary monetary policy.

The basic empirical evidence on direct mandates, and on supply and demand for labor presented in this section generally suggests a small labor effect from an employer mandate as proposed in the Administration proposal. On the whole, the evidence shows that both supply and demand for labor are relatively inelastic, which would tend to suggest small effects. The amount of involuntary job loss in the long run, restricted to those at or near the minimum wage, is relatively small. 25 Moreover, that effect could be offset by changes in minimum wage rules.

Outside of the effects on minimum wage workers, these studies, in general, represent fairly blunt instruments to measure effects (Lewin-VHI, of course, does not address effects beyond those on minimum wage workers). The connection of these estimates of voluntary withdrawal from the labor supply do not differentiate by type of worker (e.g. primary, secondary, near retirement, opting between work and welfare), do not have a clear grounding in theory, and do not distinguish between employment effects that reflect changes in participation versus changes in hours. It is important not to attach too much meaning to these estimates. Nevertheless, they do not appear to be very large.

The studies do suggest that the effects of the specialized small business caps are relatively small in the long run since they do not change the estimated job effects by a larger proportion.

The effects estimated for early retirees are subject to some uncertainty and represent, of course, voluntary labor responses.

POLICY IMPLICATIONS

In this section, several potential alterations to the Administration's plan are evaluated in terms of their effects on the job market issues. Some of these modifications have already been incorporated in certain plans. The revisions discussed include: adjusting other policies to counter involuntary labor effects, eliminating the explicit non-income based subsidy for early retirees, using age-adjusted community rating, providing exemptions from the employer mandate for small businesses, altering the nature of the subsidies (including revisions to the tax subsidy), converting the employer mandate into an individual mandate, eliminating mandates altogether, and having the government as insurer of last resort. There are, of course, many other issues that are not covered, or are covered incompletely, in this discussion, which largely focuses on employment effects. Various changes discussed could have effects on coverage, costs, and other factors that would need to be considered. Note also that these policies are discussed as departures from the Administration's original proposal; many have already been incorporated in certain congressional proposals.

OFFSETTING INVOLUNTARY UNEMPLOYMENT

To the extent that concern is focused on involuntary unemployment, there are ameliorative actions that might be taken to deal with these effects. For example, the subsidies for small businesses might be allowed, but might be phased out over time, to allow for short run wage stickiness that might result in unemployment. The plan might also be accompanied by some offsetting expansionary monetary policy that would help to accommodate the nominal wage increases that could cover health insurance. Finally, it might be possible to reduce, or slow, future increases in, the minimum wage itself on the grounds that low income workers are benefiting from health insurance.

ELIMINATING THE EARLY RETIREMENT SUBSIDY

One obvious adjustment that might be considered is to reduce or eliminate explicit non-income-related subsidies for early retirees. These subsidies make the cost of retiring smaller and thus make retirement more attractive. Eliminating them would reduce, to some extent, the voluntary decrease in employment due to early retirement. They would also reduce the cost of Government subsidies.

AGE-ADJUSTED COMMUNITY RATING

Another such revision that would reduce the loss in employment would be to introduce age adjusted community rating. On average, the cost of insurance for young individuals is smaller than the cost for older individuals since the latter have more health problems. Community rating without age adjustments tend to exacerbate the potential reduction in employment of both groups. For younger individuals who are more likely to have low earnings, the higher costs that employers face make it more likely that the proposal would cause firms to reduce the number of employees if the minimum wage prevents the costs from being passed back to individuals. In addition, the costs of insurance for lower income jobs in firms not subject to subsidized premium caps are higher under broad community rating than they would be under age-adjusted rating, making work less attractive than welfare as compared to a system with age adjustments. For older individuals considering retirement, the lower premiums available under community rating make retirement more feasible. Thus, community rating without age adjustment tends to magnify both the involuntary unemployment that might occur due to the constraints of the minimum wage and the voluntary withdrawal from the labor force of early retirees. Since adverse selection is not a problem arising from age adjustments per se, 26 there is no reason to see an age adjusted (as opposed to a non-age adjusted) community rating as producing any inefficiency in the choice of work or leisure.

Age adjustments are also likely to reduce the inefficiencies associated with firms reorganizing to respond to changes in the costs associated with different premiums depending on whether subsidies/caps apply (the sorting problem). Because both income and health costs rise with age, insurance premiums will fall less rapidly relative to income under an age-adjusted system. That is, the premium tends to be higher for lower income workers who are likely to be younger because it is not adjusted for the lower health costs associated with younger ages. Introducing an age adjustment would reduce the discrepancies between the premiums for firms with and without caps produced by the subsidies.

Age adjustment would also be likely to reduce the cost of subsidies provided to individuals as well as firms. Again, because income tends to rise with age, the age adjustments will, on average, reduce the premiums for those with lower incomes and increase them for those with higher incomes.

There may be concerns that an age adjustment is "upsetting the apple cart" since current employee plans do not typically differentiate among employees of different ages. Yet, age adjustments are effectively occurring under present rules since insurance companies will charge employers different premiums depending on the aggregate health characteristics of their employee. To the extent that different firms tend to have different aged workforces, age adjustment is already occurring. Businesses may also respond to this by altering their schedule of seniority raises that will offset the premiums differences, or they might offer plans with more or less comprehensive coverage to sort out employees between younger and older ones when individuals make contributions to the plans. In any case, it is not necessary to require that firms adjust premiums or wages explicitly by age, merely to allow insurance companies to age adjust.

One concern that might be raised about age adjustment is the possibility of age discrimination under an employer mandate, particularly for minimum wage workers where there is no opportunity to pass back the premiums in wages. This effect would be ameliorated if a set of individually targeted subsidies could be adopted, as discussed subsequently. Note also that for large firms that are self- insured or in corporate alliances, there is no community rating mandate in effect for insurers anyway, so that this effect will continue to occur in a somewhat random fashion. (The fact that firms cannot differentiate explicitly among their employees does not mean that they do not recognize the costs of hiring older workers). A second, political, concern is that introducing age adjustment might lead to pressure for other adjustments. 27 On the other hand, in another issue not related to employment affects, age-adjustment has the special quality that it is more consistent with ability to pay out of current income, since younger individuals would tend to have lower incomes and lower premiums.

At present, there is not any available study that indicates the magnitude of the effects of age adjustments on employment.

ALTERING THE SUBSIDIES

Another modification of the plan would be to alter the pattern of subsidies. Under the Administration plan, while individual subsidies are based on individual income levels, employer subsidies (or caps) are based on the average payroll. As a result, some individuals who are not lower-income individuals will receive effective subsidies simply because they happen to work in a firm with other lower-income individuals. Some other lower-income individuals will receive no subsidy because they work in a firm with higher income individuals. That is, the subsidies are not targeted very well.

An alternative approach would be to target employer subsidies directly on the basis of individual wages or incomes. There are two potential problems with this type of approach. First, as a matter of administration, it is easier to base subsidies on the total payroll of the firm. It would be more difficult to provide targeted subsidies to individual workers using total family income as a referent. Individually targeted subsidies could be based on individual wage levels however.

Of course, it would be possible to provide subsidies to individuals that take into account the portion paid on their behalf by employers, rather than giving the subsidy to the employer. This approach would, however, greatly exacerbate the minimum wage problem. In principle, an individual subsidy could hold harmless a minimum age worker from a health mandate driven employer's wage cut. However, the law does not allow the employer to offset his added cost by lowering the wage. To fix this problem, the employer could be mandated to administer an employment based health plan, but legally the mandate to buy coverage would apply to the individual.

A change which requires individuals to include some or all of individual payments in income for tax purposes, or which restricts some or all of the deductibility of insurance by employers could yield revenue that could be used in turn to finance subsidies more targeted and more generous to lower income individuals. Such an approach could involve some complexity, especially if the step were taken of including employer paid premiums in income (rather than restricting employer deductions). 28

Finally, the special subsidies for small businesses are in response to concerns about the reaction of these businesses to higher employee costs as firms adjust to the higher costs. The potential dislocation, however, primarily results from a sudden financial shock and should subside as firms discover that they are not greatly disadvantaged because other firms have similar costs and they have time to shift costs to workers or consumers. Accordingly, such special subsidies might be phased out over time.

EXEMPTING SMALL BUSINESSES FROM THE MANDATE

Small businesses are less likely to provide insurance now and may be more likely to hire minimum wage workers. Concern about the financial burden on these firms has led to proposals to exclude businesses below a certain size from the employer mandate. Excluded individuals, like current nonworkers, would still be required to buy insurance.

One problem is that, unless the structure of the subsidies changes, there might be pressures for these firms to drop coverage if the subsidy is greater under individual coverage. This effect would depend on what kind of employer vs individual subsidies are available, which would, in turn, depend on the overall wage level of the firm, the interaction of tax subsidies, and so forth. One option is to combine an individual mandate with a requirement that employers administer and offer insurance, but not require small employers to make contributions, or to have the required contribution share phase- down as the number of employees falls. Such an approach might make small businesses feel somewhat less pressured to provide payments while maintaining wages in the short run, and would decrease pressure on minimum wage workers in both the short and long run.

INDIVIDUAL MANDATES

On a broader scale, an individual mandate rather than an employer mandate would eliminate the employment considerations along with allowing a better targeting of low income subsidies. An individual mandate could simply require individuals to purchase insurance, and individuals covered by their employer's would, of course, have met the mandate. Others would have to purchase insurance individually.

There are, however, some disadvantages to individual mandates. If the individual mandate is to be accompanied by a pure subsidy to lower income individuals, many firms that presently provide insurance would have an incentive to drop such insurance, particularly if the tax treatment were equalized across employer and individual mandates. 29 Since there are some administrative advantages to employer purchases of insurance, such a decoupling of the employer/individual insurance arrangements might be undesirable. Such a system would also disrupt present arrangements. Again, the solution to this problem might be to require employers to offer and administer health plans although they would not be legally required to made contributions. They could simply withhold their employees premiums, much as they withhold income taxes and the employee share of payroll taxes currently.

ELIMINATING MANDATES

A more incremental change from present law would not involve mandates but would require community rating, perhaps limited to individuals and small group plans. (Such community rating could be adjusted for age, or for whatever general factors are deemed appropriate, as long as the individual's health were not taken into account). While this approach does not deal fully with adverse selection and would not achieve universal coverage, it may help deal with the worst of problems of availability of insurance and would not introduce employment effects. Some complications would still arise. A problem that would remain is that if an insurance company, by chance, tended to have a relatively unhealthy group of subscribers, it might be unable to cover costs and also compete with other firms. If left alone, this effect might lead to driving smaller firms out of the market and make the industry become more concentrated. It might still leave a lot of individuals and firms without coverage. (These effects on different firms would occur in any community rating plan, but the Administration's plan would provide for a set of corrections for this problem, through risk adjustments.)

MANDATES WITH GOVERNMENT AS INSURER FOR HIGH COST PLANS

A final approach would be to require coverage, but not introduce community rating, with the government acting as an insurer of last resort for individuals who face very high costs. By eschewing community rating, the need for risk adjustment or danger of driving some firms out of the market would be avoided. Such an approach might be combined with some other reforms, such as those designed to facilitate the combination of small firms into group purchasers. Individual or employer mandates could be used, but, as indicated earlier it would probably be necessary to arrange subsidies so that they do not encourage firms to drop coverage. This approach would differ from a straight individual mandate with subsidies, in that it would use direct provision of insurance rather than community rating to deal with the more serious health risks. The program could be partially financed by contributions from affected individuals, but would need to be subsidized because these individuals would reflect a high risk pool.

Effects on the labor market would depend on whether individual or employer mandates were used to collect the contributions to the plan and the financing mechanism used for the subsidies that would be necessary to insure these high risk individuals.

[The Appendix contains equations which are unsuitable for online reproduction and have been omitted.]

 

FOOTNOTES

 

 

1 For a description of the Administration's proposal and of other proposals introduced as of the end of February, see Library of Congress, Congressional Research Service, Health Care Reform: President Clinton's Health Security Act, Report 93-1011 EPW, by Beth C. Fuchs and Mark Merlis, November 22, 1993, and Summary Comparison of Selected Health Care Reform Bills, Report 94-185 EPW, by the Health Section, Education and Public Welfare Division, March 2, 1994. For other discussions of the labor market effects of the Administration's health proposal see C. Eugene Steuerle, Economic Effects of Health Care Reform, Washington, D. C., AEI Press, 1994, U.S. Congress, Joint Committee on Taxation, Description and Analysis of the Employer Mandate and Related Provisions of H.R. 3600, Joint Committee Print, February 2, 1994; U.S. Congress, Congressional Budget Office, An Analysis of the Administration's Health Proposal, February 1994; and Douglas Elmendorf and Douglas Hamilton, Labor Market Effects of the Administration's Health Proposal, forthcoming National Tax Journal, September 1994. See also U.S. Congress, Office of Technology Assessment, An Inconsistent Picture.: A Compilation of Analyses of Economic Impacts of Competing Approaches to Health Care Reform By Experts and Stakeholders. OTA-H-540, Washington, D.C., U.S. Government Printing Office, June 1993.

2 Employees who work less than 40 hours a month are not included.

3 There is some cross subsidization of one-earner by two- earner families, since the employer pays the same for a two adult family. Two earner families will have two employer portions paid. See Health Care Reform: President Clinton's Health Security Act, CRS Report 93-1011, pp. 30-31 for more details.

4 A true equivalent would occur when the individual gives up wages for insurance benefits that he values equally. A true equivalent may not occur because the benefits may be of greater or lesser value than foregone wages or if the benefit would have occurred in the absence of work.

5 This issue is discussed in Congressional Budget Office, An Analysis of the Administration's Health Proposal, February 1994 analysis. Of course, it is also a possibility that preventing labor mobility could be beneficial because firms might be encouraged to invest more in generalized human capital of their employees if they expect their employees to be with the firm for a long time.

6 This job sorting issue is discussed in the studies that focus on labor market effects cited in footnote 1.

7 See The Health Security Act and the Demand for Contingent Workers, by Linda Levine, Congressional Research Service Report 94-26 E, January 3, 1994 for further discussion of these issues.

8 Congressional Budget Office, An Analysis of the Administration's Health Proposal (February 1994).

9 See Library of Congress, Congressional Research Service, Early Retiree Provisions in the Administration's Health Reform Proposal: An Economic Analysis, by Gail McCallion, Report 94-218 E, March 7, 1994.

10 Jonathan Gruber and Alan B. Kreuger. The Incidence of Mandated Employer-Provided Insurance: Lessons from Workers Compensation Insurance, in Tax Policy and the Economy, National Bureau of Economic Research, Cambridge: The MIT Press, 1991.

11 Jonathan Gruber and Maria Hanratty. The Labor Market Effects of Introducing National Health Insurance: Evidence from Canada, National Bureau of Economic Research Paper 4589, Cambridge, Massachusetts, December 1993.

12 The factor substitution elasticity is the percentage change in the ratio of factor inputs (i.e. capital to labor) divided by the percentage change in the ratio of factor prices (cost of capital to wage rate). The demand elasticity is the percentage change in quantity demanded by customers divided by the percentage change in price.)

13 A symposium on New Minimum Wage Research was published recently in the Industrial and Labor Relations Review, October 1992, Vol. 46, no. 1. Articles include: Symposium Introduction, by Ronald G. Ehrenberg, pp. 1-5; The Effect of the Minimum Wage on the Fast Food Industry, by Lawrence F. Katz and Alan B. Krueger, pp. 6-21; Using Regional Variation in Wages to Measure the Effects of the Federal Minimum Wage, by David Card, 22-37; Do Minimum Wages Reduce Employment? A Case Study of California, 1987-1909, by David Card, p. 38-54; Employment Effects of Minimum and Subminimum Wages: Panel Data on State Minimum Wage Laws, by David Neumark and William Wascher, pp. 55-81. Other recent papers include David Card and Alan B. Krueger, Minimum Wages and Employment, A Case Study of the Fast Food Industry in New Jersey and Pennsylvania, unpublished paper, March 1993 and Lowell J. Taylor and Taeil Kim, The Employment Effect in Retail Trade of California's 1988 Minimum Wage Increase, unpublished paper, March 1993.

14 See Daniel S. Hamermesh, The Demand for Labor in the Long Run, Handbook of Labor Economics, Ed. Orley Ashenfelter and Richard Layard, New York: North Holland, 1986 for a survey.

15 There is an extensive literature on this subject dating back to a paper by Robert Eisner, Tax Policy and Investment Behavior: Comment. American Economic Review 59 (June 1969): 379-388, which indicates that the ability to substitute in the short run is much less than in the long run. A more recent study suggesting a small short run response was done by Alan Auerbach and Kevin Hassett, Investment, Tax Policy, and the Tax Reform Act of 1986, In Do Taxes Matter: The Impact of the Tax Reform Act of 1986, Ed. Joel Slemrod, Cambridge, Massachusetts: The MIT Press, 1990, pp. 13-49.

16 See John Pencavel, Labor Supply of Men: a Survey, In Handbook of Labor Economics (1986).

17 Mark R. Killingsworth, James J. Heckman, Female Labor Supply: A Survey, Handbook of Labor Economics (1986).

18 Edward P. Lazear, Retirement from the Labor Force, Handbook of Labor Economics (1986).

19 June E. O'Neill and Dave M. O'Neill, The Employment and Distributional Effects of Mandated Benefits, American Enterprise Institute, Washington, D.C.: The AEI Press, 1994.

20 The Effects of the Health Security Act on Employee Wages and A Comparison of the Effects of the Health Security Act and the Individual Tax Credit Program on Households, Prepared for The Heritage Foundation by Lewin-VHI, Inc., March 9, 1994.

21 David Wyss, The Impact of the Clinton Health-Care Reform Plan, DRI/McGraw-Hill U.S. Review, March 1994, pp. 29-32.

22 Consad Research Corporation, Employment and Related Economic Effects of Health Care Reform, April 8, 1994. This study was prepared for the National Federation of Independent Business and Healthcare Equity Action League.

23 Congressional Budget Office, An Analysis of the Administration's Health Proposal, February 1994.

24 This issue and related literature are discussed in Congressional Budget Office (1994) and Congressional Research Service (Report 94-218 E, March 7, 1994).

25 In addition, these effects might be offset by increased hiring of minimum wage individuals by firms that had lower costs, effects that are apparently not accounted for in the studies. This effect, however, is likely to be small. The effects might also be moderated by sorting of minimum wage workers into those firms that are subject to caps, as discussed in the following section.

26 Adverse selection occurs with conditions that cannot be readily ascertained (e.g. individual health status) by the insurance plans. In addition, since each individual passes through the different ages if he survives, adjusting premiums by age does not prevent social sharing of risks.

27 One potential adjustment, for sex, would probably exacerbate employment effects of secondary earners, since premiums would be higher for younger females than for younger males.

28 The current tax benefit is discussed in more detail in Congressional Budget Office, The Tax Treatment of Employment-Based Health Insurance, March 1994.

29 Under current law individuals do not include in income the premiums paid on their behalf by employers, and this provision encourages employer plans to be used.

 

END OF FOOTNOTES
DOCUMENT ATTRIBUTES
  • Authors
    Gravelle, Jane G.
  • Institutional Authors
    Congressional Research Service
  • Subject Area/Tax Topics
  • Index Terms
    health care, publicly funded, mandated
    health care, publicly funded, legislation
    employee benefit plans
  • Jurisdictions
  • Language
    English
  • Tax Analysts Document Number
    Doc 94-5759
  • Tax Analysts Electronic Citation
    94 TNT 118-89
Copy RID