Are Employer Health Insurance Mandates a Viable Policy Option

1. Introduction

Pay-or-play employer mandates – which require businesses to either provide health insurance directly or pay into a state fund to cover their workers – goes back more than thirty years. In 1974, Hawaii passed the “Prepaid Health Care Act” which requires nearly all employers to provide health coverage to their employees who work more than 20 hours per week for four consecutive weeks. In the thirty years since its passage, however, there have been only a handful of attempts by other states to pass similar legislation, perhaps out of concern about violating the Federal government’s Employee Retirement Income Security Act (“ERISA”), which superseded all state laws related to employee benefits.

The most noteworthy recent attempt at pay-or-play legislation was in California in 2003. In the days leading up to the recall election in 2003, then-Governor Gray Davis signed the “Health Insurance Act of 2003” (“HIA”, also known as Senate Bill 2 or SB-2). HIA required businesses with 200 or more employees to provide health insurance to all employees, as well as their dependents, starting on January 1, 2006. It also required businesses with 50 to 199 employees to provide insurance to all workers, but not their dependents, starting January 1, 2007. And if the state of California passed a tax credit, businesses with between 20 to 49 employees would also be required to meet the requirements for individual workers. HIA did more than simply require coverage, however. It mandated that firms pay the vast majority of premiums for a minimum-mandated-quality plan. In all cases, the firm was responsible for 80 percent of the premiums for the single/family plan (for firms with 200 or more employees) or 80 percent of the premiums for a single plan (for firms with 50 to 199 employees). The tax credit meant that firms with 20 to 49 employees effectively were responsible for 64 percent of premiums rather than 80 percent. This proposal, signed into law, was scheduled to start in 2006, but opponents of the mandate put a referendum on the ballot for the November 2004 election. Proposition 72, defeated by a narrow 51 percent to 49 percent margin, repealed HIA.

Despite its narrow defeat in California in 2004, pay-or-play employer mandates continue to be a relevant policy option in many states, and even some localities. According to the National Restaurant Association’s web site, thirteen pay-or-play mandates were proposed during 2005. Although most were defeated or tabled, many pieces of legislation are expected to reemerge in 2006. In Massachusetts, for example, the House recently passed by a 131-22 margin legislation which imposes a payroll tax on firms with 11 or more employees, and then allows firms to credit health care expenditures against the payroll tax. Although different from California’s legislation, this payroll tax (and the credit against it for health care expenditures) is very much an employer pay-or-play mandate. The Massachusetts Senate passed a competing bill that did not include such a payroll tax, but the compromise bill that goes to the Governor very well could. Other serious attempts in 2005 – that for now appear stalled – occurred in Washington and New York. In both cases, many of the actual legislative provisions appear to be motivated by California’s HIA. For example, “The Working New Yorkers Health Insurance Act,” introduced in February 2005, had language that was nearly identical to HIA, but with different cut-off thresholds by firm size.

Even more recently, the idea of employer mandates has trickled down from the state-level to the city-level. In November 2005, San Francisco Supervisor Tom Ammiano introduced legislation requiring San Francisco businesses with 20 or more employees to provide health coverage. Although voters in San Francisco overwhelmingly supported Proposition 72, only four of eleven members of the Board of Supervisors are currently supporting the measure, and Mayor Gavin Newsom has expressed reservations about the legislation.

Even in the wake of Proposition 72’s defeat in California, the prospect of pay-or-play mandates continues to be discussed. Health Access California, in a policy brief in April 2005, states “The closeness of the vote (for Proposition 72) means that the idea will continue to be advanced, as the need will only become more urgent.” Their first recommendation in expanding employer health insurance is to “Support new versions of SB 2/Proposition 72 to secure the employer-based coverage on which 19 million Californians rely, and to extend it to some or all of the 80% of the uninsured that are in working families.”[1]

The remainder of the paper is arranged as follows. Section 2 examines some of the claims that were made during the debate about California’s Proposition 72. Although many of the claims are technically correct, they are also quite misleading. Section 3 discusses findings based on my previous work. In Yelowitz (2003, 2004), I found that the cost to businesses would have been quite high in the short-term, and that a considerable portion of this increased cost would have been a shifting of financial responsibility from workers to firms. In total, I found that the increased cost to business would have been close to $12 billion (assuming no tax credit was enacted), and that the previously uninsured were responsible for roughly 30 percent of the cost.

Section 4 explores some of the economic effects from pay-or-play mandates. When faced with higher mandated costs – such as the costs that HIA would have entailed – firms face a variety of economic options. Some involve traditional labor market adjustments – reducing real wages, laying off workers or cutting back work hours, eliminating other fringe benefits, or substituting capital for labor. Others involve “non-traditional” adjustments that are largely driven by the nuances of the actual HIA legislation; for example, if a firm consolidated part-time workers into full-time workers, it might be able to get under one of the costly thresholds of 20-, 50- or 200-workers. In addition to these labor market adjustments, there is also the possibility that firms could relocate at least some of their operations outside of the state, or in the extreme, be forced to shut down because their profits become losses. Finally, firms may simply accept lower profits or be able to pass some of the costs onto consumers in the form of higher prices. Convincing empirical evidence exists on some, but not all, of these avenues of adjustment.

Section 5 offers suggestions for policymakers who may consider pay-or-play in the future. The goal is to take the key economic motivation for covering the uninsured – the fact that the uncompensated care they use creates an expensive burden on the insured passed on through higher premiums – and combine that with some of the findings on economic responses. An employer mandate for workers-only that consisted of a high deductible and covered catastrophic costs, along with significantly more statutory cost-sharing on the part of employees, could substantially reduce total costs as well as at least some of the negative labor market adjustments. Moreover, recent work in behavioral economics suggests a great deal of “inertia” in decision making; it is possible that by allowing individuals to “opt-out” of the health insurance mandate, such a mandate would respect freedom of choice – an idea that has been called “paternal libertarianism.” Given the evidence in other contexts (related to 401k participation), it is likely that such an “opt-out” employer mandate would still reduce the uninsured quite substantially.

2. Efforts to distort the cost impact of Proposition 72

One of the reasons I became involved in the HIA discussion in 2003, and stayed involved in 2004, was the fact that I felt the numbers being presented to the public were incredibly misleading. None of the statistics were wrong; rather they were often irrelevant to the real debate at hand. Although there were inaccuracies and misleading statements on both sides of the debate, I shall examine only the claims that my two studies shed the most light on.[2]

One of the most misleading themes during the campaign was the idea that HIA was a modest, incremental step. The logic behind this idea was often summarized with four principal arguments. First, not very many firms would be affected by the law. Second, most of the firms that are affected already offer health insurance. Third, offering firms pay close to 80 percent of premiums for individual and family plans. And fourth, the cost per newly-insured worker was relatively small. I shall discuss these in turn.

Misleading Claim #1:

“Very few firms are affected”

Recall that HIA created different firm-size thresholds at 20-, 50-, and 200-employees. Based on tabulations of publicly available data, the fact emerges that most firms are very small. It led a number of advocates for Proposition 72 to make statements such as:

“SB 2 is actually a moderate and reasonable step that would affect less than 5 percent of California employers.” – California Medical Association (2003)

“It’s modest because it doesn’t cover everyone” – State Senator Jackie Speier (D-Daly City), one of the leading authors of the law.[3]

It is true that the overwhelming majority of employers are below the cut-off thresholds imposed by HIA, but many employees are working at these larger firms. Since potential costs are related to covering employees, not employers, they are the appropriate metric to use.

Figure 1 illustrates this point, using data from the California Employment Development Department for 2003 (the latest available online). I show the fraction of firms (in blue) and employees (in red) who would be affected by pay-or-play mandates with different firm cut-off thresholds. For example, a mandate that targeted firms with 1000 or more employees would affect just 0.1% of all firms, yet it would affect more than 14% of workers. A mandate with a cut-off of 50 employees would affect less than 5% of all firms (confirming the quotation from above), yet it would affect nearly 61% of workers. Even going down to a cut-off of 10 workers would only cover one-fifth of employers, yet it the overwhelming majority of employees. The basic fact is that many small firms, who by definition have very few workers, create a distortion between counting the impact based on employees or employers.

Misleading Claim #2

“Large firms already offer health insurance”

One motivation for targeting larger firms rather than smaller firms is based on the notion of ability-to-pay. In general, we perceive that larger firms have both more employees and higher profits. Thus, they may more easily accommodate the per-employee cost of such a mandate. It also turns out that larger employers tend to offer more generous total compensation than smaller employers – both higher pay and better health insurance. Part of the reason for large firms offering better health insurance stems from the fact that larger groups are able to obtain more reasonable insurance rates than smaller groups, due to the lack of adverse selection and the fact that they can spread the fixed costs of running a health plan across more workers.

As a consequence, it is often mentioned that virtually all large firms already offer health insurance, which leads many to believe HIA would have been a trivial change. For example, Eric Schlosser, author of “Fast Food Nation” wrote in a Los Angeles Times piece entitled “Super-Sized Deception From Fast-Food Giants” in October 2004:

“In fact, the proposition would apply to fewer than one-tenth of the state’s restaurants and retail stores. Among employers with more than 200 full-time workers, 99% already provide health insurance; among those with 50 to 150 workers, 94% do.”[4]

Again, on the surface, this is correct. The Kaiser Family Foundation and Health Research and Educational Trust (“KFF/HRET”) conduct an annual survey of businesses in California, known as the California Employer Health Benefits Survey (“CEHBS”), and this data source is credible and widely cited. The CEHBS, in fact, does reveal that almost all large firms do offer health insurance to at least some employees. As I show in Figure 2, using the 2003 CEHBS, offering rates (the red bars) around 98% are common across all firm sizes, once firm size reaches 50 employees.

This fact alone is insufficient to conclude that such a mandate will entail moderate cost, however. At least two other critical factors come into play. First, the numbers refer to offering health insurance to any employees. Around one-half of firms only offer coverage to full-time employees, and HIA’s definition of full-time – which was 23 hours per week – may differ markedly from a typical firm’s full-time definition.[5] Second, many employees do not take up coverage, even if their firm offers it to them. The reasons for not taking it up vary. Some workers cannot afford their share of the premiums. One estimate from the 2002 CEHBS reveals that 58 percent of employees who turned down coverage (and did not have health insurance coverage elsewhere) could not afford their share of the premiums.[6] Others workers are covered through a spouse’s employer plan, purchase a private plan directly (such as a Blue Cross/Blue Shield plan), or are covered by a government program. And others may simply decide to take the chance they will not get sick.

Regardless of the reason, we observe in Figure 2 (in the blue bars) that roughly 60 to 70 percent of workers are covered by their own firm’s plan.[7] An employer mandate therefore could create a substantial new cost burden for these firms. Part of the cost for these employees (and dependents) come from the previously uninsured, while another part comes from the crowding out of government health insurance, spousal coverage, and direct purchases. The crowding-out of other sources, such as a spouse’s employer health insurance coverage, could be viewed, in some sense, as redistribution (e.g., shifting the cost burden from the spouse’s smaller employer to the other spouse’s larger one). Nonetheless, from the larger firm’s economic decision-making perspective, its own cost, not the net societal costs, are what is important. The firm, in turn, would be likely to make a number of labor market adjustments, including cutting back on employment and wages.

Misleading Claim #3

“Most firms pay 80% of premiums”

The HIA was more than legislation to provide coverage to the uninsured. It also introduced a “premium sharing mandate” by requiring affected firms pay for 80% of the premiums for the worker (in the case of firms with less than 200 employees) or 80% of the premiums for a worker and dependents (in the case of firms with 200 or more employees). This “premium sharing” part of the HIA mandate was ignored by all of the studies on employer costs except Yelowitz (2003, 2004).

One possible motivation for ignoring this premium sharing mandate comes from data from the 2002 CEHBS. The data reveal that, on average, medium and large employers offering coverage already pay for 87% of premiums for single coverage (Kaiser Family Foundation, Chart 10), and, on average, large employers already pay for 79% of premiums for family coverage (Kaiser Family Foundation, Chart 11).[8] Figure 3 confirms these findings using the 2003 CEHBS – on average, affected firms paid between 81 and 87 percent of a single plan, and between 71 and 79 percent of a family plan.

Is it correct, then, to ignore the cost impact on currently insured workers? There are two compelling reasons why this is not correct. First, as is widely recognized, even though California employers nearly meet the premium sharing part of the mandate requirements on average, there is a great deal of dispersion with some employers paying more than 80 percent and others paying far less. KFF/HRET used the CEHBS to estimate that in 2002, 20 percent of small or medium employers and 21 percent of large employers did not cover 80 percent of the premium costs of a single plan, and approximately 50 percent of large employers did not cover 80 percent of the premiums for a family plan.

Using updated data for 2003 in Yelowitz (2004), I found that many currently covered employees in firms that already provide health insurance would have been affected by the HIA provisions. Nearly 5.0 million currently insured workers pay for some or all of the premiums of the plan (Yelowitz, 2004, Appendix Table 2, lines 3-4). The employee’s share of premiums for a family plan for the median employee is 25 percent; under HIA, this would fall to 20 percent. This shifting of payment responsibility represents nearly a 7 percent increase in costs to the employer. At least one-quarter of all covered employees pay at least 44 percent of the family plan premiums. Those covered employees affected by the HIA mandate would have their contributions reduced by more than 50 percent; the employers would have to absorb these additional costs under HIA. As I show below, the costs for currently insured workers and dependents exceed $5.6 billion.

Second, HIA imposed not only a “premium sharing mandate” but also a “minimum benefits mandate.” Some firms pay for 80 percent of the costs of a health care plan with fewer benefits than would be required by the HIA legislation. Although the actual fee to “pay” rather than “play” from HIA was never established, in my work I assume that the generosity of the state’s plan (and the fee) is equivalent to the expense of the median health insurance plan (taken from the KFF/HRET data). The ambiguity in the law was never cleared up during the debate on Proposition 72, and many researchers have taken an approach like this. In Yelowitz (2004), I used the 2003 CEHBS to exactly replicate the findings in KFF/HRET (2004), in particular that the mean annual premium costs for single and family plans in California, respectively, were $3,102 and $8,504 in 2003. I also found that the median annual cost was modestly lower than the mean because of some outliers; the median cost of a single plan was $3,001 and the median cost of a family plan was $8,345. When the sample was restricted to firms with 50 or more employees, the median annual premium is $3,022 for a single plan, and $8,482 for a family plan, higher than the $3,001 and $8,345 figures reported here.

Figures 4a and 4b incorporate this “minimum benefits mandate” with the premium sharing mandate. Given the numbers above, an employer would be responsible for $2400.80 for a single plan and $6,676 for a family plan. By firm size, the figures break out the fraction of firms that are currently compliant in terms of expenditure, as well as the fraction that are within 20 percent of these expenditure thresholds, within 20 to 40 percent of the threshold, and more than 40 percent away. For example, a firm that is more than 40 percent away would have to raise its premium contribution by at least $2670 per employee if it were subject to the family mandate.

Although all firm sizes are presented, I will restrict attention to firms that would have been affected by the HIA requirements. Figure 4a shows the gaps based on the single worker requirements. What is clear is that with the exception of firms between 150 and 199 employees, the majority of offering firms would not have to raise their contributions for single plans.[9] Roughly two-thirds of firms in each of the categories were already compliant. The percentage of firms that exceed the $2400.80 threshold – illustrated in green in Figure 4a – remains fairly flat as one moves between firm sizes of 50 to 99 employees all the way to firms with 1000 or more employees. Moreover, the percentage of firms that will have a drastic increase in their costs is relatively small. Between 2 and 8 percent of firms would see their employer contributions rise by 40 percent or more due to the mandated single worker coverage.

Figure 4b uses a similar methodology and shows findings for family plans. What is immediately clear is that a far greater percentage of affected firms would have had to raise employer contributions due to the family mandate, even for employees that already receive health insurance. Between 42 and 56 percent of larger firms were spending at least $6,676 on contributions toward family plans. Moreover, many firms would have faced a drastic increase in costs for currently insured workers. Twenty percent of firms with 200 to 499 employees would have been forced to spend at least 40 percent more in contributions to meet the $6,676 threshold, while 6 percent of firms with 1000 or more employees would have seen a cost increase of this magnitude.

In summary, these figures show that the cost-sharing requirements of the single worker mandate were not very costly for most firms where employees were taking up the coverage, while the cost-sharing requirements of the family mandate were much more expensive.

Misleading Claim #4

“Low costs per newly insured workers”

The final part of the argument is that, in addition to very few firms being affected due to the firm-size requirement and offering patterns, uninsured workers are relatively cheap to cover. In large part, advocates rely on a study by Dube and Reich (2003). They state:

“The median California covered business (i.e. with 20 or more employees) will see an annual increase in costs of $1,343 per worker it newly insures.”

To arrive at this number, they use average health premium costs in the 2002 CEHBS published by KFF/HRET. The total premium was $2,845 for a single plan and $7,471 for a family plan from the 2002 survey. They also estimate a marginal cost per dependent of $2,085 using the aggregate number of dependents in family plans in the Current Population Survey (“CPS”). The employer is responsible for 80 percent of these costs, or $2,276 for a single plan, $5,976.80 for a family plan, and $1,668 per newly covered dependent. They then account for corporate income tax deductions, and adjust the values for the fact that “some of these workers who are not insured through their own employer are dependents of spouses. These individuals do not represent added costs to employers, but rather shifts in costs between employers.”

In Yelowitz (2004, p. 16), I offer a host of critiques about the method by which Dube and Reich arrive at this number. The most important critique is that their findings pertain to the median affected business, yet the distribution of costs across covered employers is clearly bimodal. HIA creates relatively low per-employee costs for uninsured workers with 20 to 199 employees and a distinctly higher set of costs – more than double – for uninsured workers at employers with 200 or more employees.

This idea can be made clear by examining Figure 5. This figure uses the sample weights in the 2003 CEHBS to examine the distribution of firms with 50 or more employees.[10] The figure clearly shows the distribution of firms is skewed toward smaller firms – the median firm has 126 employees, conditional on having at least 50. This firm – the “median affected business” – faces the single worker mandate, not the family mandate. To help illustrate why this matters for cost calculations, imagine instead that the median affected business just hovered on the border of 200 employees. If the median firm had 199 employees, the estimated cost would entail the single worker mandate, while if the firm had 200 employees, the estimated cost would entail the family mandate. Clearly these will give very different values, even using the same premium data in Dube and Reich. By not separately presenting cost estimates by the firm groupings established by HIA, the cost number that is typically cited ignores the impact of the family mandate.

The bottom line is that all four arguments have serious problems with them. Although many voters might be fooled into believing that the costs to employers would be low from a pay-or-play mandate, it turns out that each of the commonly presented statistics is largely irrelevant for figuring out the cost of the mandate.

3. Coverage and cost estimates

This section briefly summarizes findings from Yelowitz (2004), which at the time used the most up-to-date information on California’s health insurance market. The interested reader may obtain the full report (which is 90 pages) from the web site of Employment Policies Institute.[11] My work combined information from the March 2003 CPS to health insurance premium data from the CEHBS. It also utilized information from the County Business Patterns, in order to make more accurate predictions based on firm size (the CPS questionnaire had different thresholds than the HIA law).

Figure 6 shows the coverage effects of HIA, based on the individual’s existing source of coverage. Overall, the March 2003 CPS reveals that 6.4 million individuals in California’s population of 35.2 million were uninsured during the 2002 calendar year, or 18.2 percent of the population. Assuming that the HIA provisions affected firms with 50 or more workers, then 1.6 million of the uninsured, or slightly more than one-quarter of the uninsured, would gain health coverage due to the mandate. Although an employer mandate provides far from universal coverage, it would substantially improve California’s national standing. In the calendar year 2002, California ranked 46^th out of the 50 states and DC in terms of the percent uninsured. Reducing the uninsured by 1.6 million would move California’s ranking to 27^th. As discussed in Yelowitz (2003), however, there are other ways of making improvements of this magnitude that are far less disruptive to the labor market. For example, nearly all uninsured children in California are eligible for Medi-Cal (or some other state-sponsored program); an aggressive expansion in outreach efforts to enroll these children could reduce the uninsured by the same magnitude, while securing additional federal matching funds.

As can be seen by the figure, the mandate also would have covered a substantial number of individuals who already had employer coverage or government coverage. For at least some of these individuals, this coverage resulted in lower premiums for the worker and higher premiums for the firm. The coverage would have led to crowd-out of some government health insurance sources, in particular Medi-Cal. This would have resulted in a savings of tax dollars to the state government, but also a potential “exporting” of dollars to the federal government.

Figure 7 shows the cost effects of HIA, for the population as a whole, broken out by existing source of coverage. The technical assumptions used in arriving at these cost estimates are discussed in the Appendices of Yelowitz (2004, p. 50-65). In total, the mandate would have entailed a cost of close to $12 billion (using 2003 premium data). Of this amount, approximately 30 percent – or $3.7 billion – would have been spent covering the uninsured. By far the largest slice of Figure 7 – 47 percent, or $5.6 billion – would have been spent on individuals who already had insurance coverage. This emphasizes some of the key points made by Figures 4a and 4b; in particular, the mandate was costly to firms that already provided health insurance. One final figure of interest is the fact that employers would have spent $1.6 billion on covering those with government health insurance, primarily Medi-Cal recipients. Roughly half of this expenditure would represent savings to the state of California, while the other half would have likely represented savings to the federal government.

Overall, the HIA mandate does significantly reduce the number of uninsured, but at a very high price. After incorporating the costs to other groups that are already insured, the effective cost is more than $7,200 per newly insured individual. The steep rise in health premiums over the last several years means that a comparable number today would be even higher. At the national level, premiums rose 13.9 percent in 2003, 11.2 percent in 2004, and 9.2 percent in 2005. The total cost of a single plan in 2005 is $4,024, while the cost of a family plan is $10,880 – around 25 percent higher than the estimates used in Yelowitz (2004). Thus, the effective cost per newly covered individual would likely be around $9000.

4. Economic impacts of pay-or-play mandates

When faced with higher mandated costs – such as the costs HIA would have entailed – firms face a variety of economic options. Some involve traditional labor market adjustments – reducing real wages, laying off workers or cutting back work hours, eliminating other fringe benefits, or substituting capital for labor. Others involve “non-traditional” adjustments that are largely driven by the nuances of the actual HIA legislation; for example, if a firm consolidated part-time workers into full-time workers, it might be able to get under one of the critical thresholds of 20-, 50- or 200-workers. In addition to these labor market adjustments, there is also the possibility that firms could relocate at least some of their operations outside of the state, or in the extreme, be forced to shut down because their profits become losses. Finally, firms may simply accept lower profits or be able to pass some of the costs onto consumers in the form of higher prices.

A number of credible, peer-reviewed studies suggest that whenever possible, profit-maximizing employers will react to pay-or-play by shifting costs onto employees in the form of lower wages. This is especially likely in the long-run, because firms can reduce real wage growth rather than nominal wage levels. In this case, the employee rather than the employer bears the cost of the mandate. In the case of the least skilled workers, however, wage shifting may simply not be an option. These employees are at risk of losing their jobs, either through labor force cuts or competition from more experienced workers attracted by the new benefits.

Theory and evidence on mandated benefits

Summers (1989) presents theoretical arguments for mandated benefits relative to public provision of a good. He notes that “if employers and employees can negotiate freely over the terms of the compensation package, they will reach a mutually efficient outcome.” Yet, Summers argues that there are potential market failures that could lead to the case for public provision or mandated benefits. These market failures include “merit goods”, irrational consumers, externalities, and adverse selection.

Figure 8 shows the typical supply and demand framework used to analyze a tax or mandated benefit. Ignoring the presence of the minimum wage, the efficient labor market equilibrium occurs at the intersection of the worker’s labor supply (S⁰) and employer’s labor demand curves (D⁰). This gives the equilibrium quantity of labor (L⁰) and equilibrium wage rate (w⁰). Assuming that the labor market is competitive and there are no market failures or government distortions, the employment level and wage rate are economically efficient.

Under a typical tax imposed on employers (the demanders of labor), the demand curve shifts down to D¹, and the new labor market equilibrium is (L¹,w¹). This is shown in Figure 9.

The tax imposes economic inefficiency, known as deadweight loss, represented by the yellow triangle. Government intervention lowers the employment level and wage rate.

As Summers (1989, p. 180) notes, “mandated benefits do not give rise to deadweight losses as large as those that arise from government tax corrections.” The reason is that because the mandated benefit is potentially valuable to the employee, the labor supply curve shifts downward as well. The new employment level is given in Figure 10 by the intersection of the new labor demand curve, D¹, and the new labor supply curve, S¹.

This equilibrium represents a situation with lower employment than without any government interference, but higher employment than with tax-financed provision of a benefit. The new labor market allocation, (L²,w²), also has lower wages for workers than either tax-financed provision or no government intervention. The inefficiency from such a mandated benefit is given by the smaller yellow triangle. The allocation with mandated benefits could therefore involve substantial wage reductions for employees.

This straightforward framework ignores several important features, however. First, as Summers (1989, p. 180) points out, the “mandated benefits represent a tax at a rate equal to the difference between the employers cost of providing the benefit and the employee’s valuation of it, not a rate equal to the cost of the employer of providing the benefit.” One critical issue then becomes the employee’s valuation of the benefit. My calculations in Yelowitz (2004) found that more than 1.08 million current recipients of government health care in California would have had that insurance crowded-out by employer-provided health insurance. For these enrollees, the additional value of the benefits from the HIA mandate are surely quite small, so the economic inefficiency and employment losses look more like Figure 9 than Figure 10. The same is also probably true for the uninsured who are eligible, but not participating, in Medicaid. Brown, et al., (2002, p. 48) estimate that 1.12 million adults and children are eligible for Medi-Cal or Healthy Families but not participating.[12] It is also likely that some of the uninsured – especially younger, healthier adults – do not put a very large valuation on health insurance.

Second, as Summers (1989, p. 181) notes, if there is a binding minimum wage, then “wages cannot fall to offset employers’ cost of providing a mandated benefit, so it is likely to create unemployment.” In Yelowitz (2004), I found that there were 4.3 million Californian workers with wages below $9.31 per hour, including more than 680,000 workers in large firms who were either uninsured or on government insurance. It is likely that wages would not be able to fully adjust downward for such workers. More generally, when wages are rigid and do not move downward in response to the mandate (which is especially likely in the short-run), then the larger economic inefficiencies illustrated in Figure 9 become more likely.

Finally, the framework above shows that mandated benefits are still a government tax, even if they are not explicitly called a tax. Summers (1989, p. 182) cautions about the government’s use of mandated benefits. He says “There is no sense in which benefits become ‘free’ just because the government mandates that employers offer them to workers.” Reinhardt (1987, p. 124) notes that “the fiscal flows triggered by mandate would not flow directly through the public budgets does not detract from the measure’s status of a bona fide tax.”

Although there are a number of empirical studies that examine the impact of employer mandates for family leave and workers compensation, the study that provides the closest (albeit not that close) analog to the HIA mandate is Gruber (1994). Gruber (1994) studied several state and federal laws that mandated comprehensive childbirth benefits in health insurance policies, and therefore substantially raised the cost of insuring women of childbearing age. Between 1975 and 1978, some states passed laws that prohibited treating pregnancy differently from “comparable illnesses.” In October 1978, the Federal government passed the Pregnancy Discrimination Act, which prohibited any differential treatment of pregnancy in the employment relationship. Using the CPS, he finds shifting of the costs of the mandates from the employer to the employee in the form of lower wages on the order of 100 percent. In fact, some of his specifications suggest over-shifting of wages.

Thus, Gruber (1994) provides strong evidence that firms will lower wages (when possible) to “pay” for the mandate. Even those who have higher wages are affected by the mandate if their employer does not provide coverage.

Wage shifting and unemployment

The evidence from Gruber (1994) would suggest that ultimately employees will bear the cost of HIA, even if the statutory cost of HIA is formally 80 percent on the employer and 20 percent on the employee. Although the previous literature on mandated benefits does provide guidance on the possibility of wage shifting, such wage shifting is unlikely for those near the minimum wage. For workers that are close to California’s minimum wage of $6.75 per hour, it is simply not possible to shift the costs in the form of lower wages. In Yelowitz (2004), I found that wage shifting was constrained for approximately 1.4 million employees due to the current California minimum wage of $6.75 per hour (along with the costs of the HIA mandate). Operating under this constraint, employers are faced with a similar situation to an increase in the wage floor. They must accept lower profits, raise prices, change the capital/labor mix, or alter employment levels and skill levels to respond to the increased costs.

In studying the effect on increases in mandated wage levels, Neumark (1995) found that current employees were often displaced by higher skill individuals attracted by higher wages. Lang (1995) found wage hikes shift “employment towards teenagers and students… [T]he competition from [these] higher quality workers makes low-skill workers worse off.” Neumark and Wascher (2000) convincingly reevaluate Card and Krueger’s (1994) study of minimum wages in New Jersey, and using restaurant payroll data find an employment elasticity of -0.22.[13] In a prominent survey of labor economists, Fuchs, Krueger, and Poterba (1998) find that the mean estimate of the employment elasticity for teenagers is -0.21, and the median is -0.10. Yelowitz (2005) found a significant rise in the unemployment rate for less-educated workers and a significant decline in hours worked when the city of Santa Fe raised its minimum wage to $8.50 per hour, a 65 percent increase from New Mexico’s $5.15 minimum wage. As with Lang (1995), Yelowitz also found a shift in employment towards teenagers and students.

All of these studies suggest that in the absence of full wage shifting, there is a strong possibility of layoffs, job loss, and work-hours reduction as a result of HIA, especially for low-skill workers.

In Yelowitz (2004), I examined the possibility of job loss and used the Neumark and Wascher (2000) elasticity estimate of -0.22 under two different scenarios that are meant to represent the “short-term” and the “long-term.” The first scenario assumes no wage shifting for any worker, while the second assumes full wage shifting until the minimum wage and disemployment effects thereafter. That is, the second scenario shifts as much of HIA onto the worker as possible in the form of lower wages – as suggested by Gruber (1994) – and only to the extent that wages would have to be shifted below the California minimum wage of $6.75 would employment losses ensue.

Figure 11 shows the disemployment results found in Yelowitz (2004). By summing up the blue bars, we see that when wage shifting is possible, approximately 70,000 workers lose their jobs as a result of HIA. Nearly 25 percent of these workers already had employer-provided health insurance. More than 32,000 of these workers were uninsured meaning that in addition to not receiving health insurance, now they also lose their jobs. Around 11,500 workers with government insurance lose their job, meaning they continue to keep this insurance instead of being transferred to employer insurance. When wage shifting is not possible (as is likely in short-run), the results are even more dramatic. Summing up the red bars in Figure 11, we see that around 150,000 workers lose their jobs, with roughly equal numbers coming from the uninsured and covered by employers.

5. Are employer mandates viable? A policy proposal

Given the evidence above, the cost of pay-or-play insurance mandates is substantially more expensive than is commonly thought. Although it is not at all clear that these mandates are the best way to cover the uninsured, they continue to remain subjects of serious debate.

How then could such mandates be made more effective? That is, can legislation be designed that minimizes the negative labor market impacts, while at the same time reducing the costs of uncompensated care.

First, my work has found that a sizable part of the cost was due to the family coverage requirement. Future legislation should abandon such a goal. A non-trivial number of firms – even ones offering health insurance – had their costs substantially raised by this requirement. Moreover, having such a requirement increases the relative price of hiring a worker who has a family compared with a worker who does not. Given that the difference in costs between full-time workers based on single versus family coverage is in the range of $2.50 per hour, it is likely that there would be a detectable shift in large firms toward hiring single, childless workers. Rather than serving as a benefit for working mother and fathers, the mandate would work as a tax.

Second, there is strong evidence that costs would be shifted onto workers in the form of lower wages. A more honest representation of the law would reflect that reality. More importantly for the labor market, the disemployment effects presented in Figures 8 to 10 change when the statutory incidence of the mandate falls onto the worker rather than the firm in the presence of a binding minimum wage. That is, firms cannot wage-shift past $6.75 per hour, but if the costs were deducted from the employee’s gross pay, the wage-shifting property is preserved for the low wage worker. For example, a mandate that costs $1.00 an hour would reduce the minimum wage worker’s effective wage to $5.75 if it were imposed on workers, while the effective wage would remain $6.75 if it were imposed on firms. In the latter case, the likelihood of job loss is much greater than the former case.[14]

Third, although the actual “fee” to pay rather than play was never established, many thought the mandate’s minimum benefits would be quite generous. Yet, one of the principal motivations for the bill was reducing the costs of emergency room use and uncompensated care. The California Medical Association (2003) asserted that “5% of premium costs reflect absorbed costs of the uninsured.”[15] A high-deductible health insurance plan would cover the most extreme expenses of the uninsured, so it seems likely that individuals with such a plan could self-insure against more routine expenses. A casual inspection of www.ehealthinsurance.com reveals that a healthy 35-year-old male in Los Angeles could directly-purchase a high-deductible plan for roughly $70 per month. By mandating a plan like this, rather than the kind that was envisioned by the proponents of HIA, the infra-marginal costs to the currently insured would be minimized.[16]

Fourth, the mandate could take advantage of recent insights from the behavioral economics literature. Madrian and Shea (2001) find that 401(k) enrollment exhibits a great degree of “inertia.” That is, when workers were forced to fill out paperwork to “opt-in” the retirement plan, participation was quite low. When the company they studied switched to automatic enrollment – where the worker could easily “opt-out” of the retirement plan if he or she wanted to – participation rates skyrocketed. Sunstein and Thaler (forthcoming) term this idea – preserving freedom of choice, while changing the “default” behavior to a more desirable action – as “libertarian paternalism.” In the context of health insurance mandates, such a pay-or-play mandate could have a “default” option that the worker is signed up for the high-deductible plan, with an option of opting-out. Given the great deal of inertia associated with pensions, it is likely enrollment would remain high while preserving the worker’s sovereignty.

A natural question arises with this proposal however: Why not simply impose an individual mandate to purchase health insurance, as states do with automobile insurance? First, by using an employer mandate, individuals would retain the advantage of buying insurance in the group market rather than individual market. Thus, they would enjoy some of the savings associated with reduced per-person administrative costs as well as the reduced amount of adverse selection from buying as a group. Second, the employer’s contribution – which should be much lower than that proposed in HIA – is not counted as taxable income, which is known as the “tax subsidy for employer provided health insurance.” Third, by imposing this mandate on employers and employees, rather than non-working individuals, we would ensure that the mandate is affecting those with at least some ability-to-pay. Consider the catastrophic plan proposed before, at $70 per month. For a full-time, full-year employee (e.g., 2000 hours of work per year), this cost would be around 42 cents per hour. Assuming that the statutory incidence were reversed from HIA’s requirements (e.g., employees pay for 80 percent of the costs), then the statutory hourly cost for employers would be around 8 cents per hour. Even applying an employment elasticity of -0.5 – more than twice as large as the Neumark and Wascher (2000) estimate for the restaurant industry – such a policy would likely reduce employment by less than one-half of one percent. Clearly such a proposal would be viewed as measured approach to the problem.

6. Concluding remarks

This paper has examined number of critical facts related to “pay-or-play” employer health insurance mandates. First, California’s HIA legislation from 2003 was more than a modest, measured step to cover the uninsured. Many of the facts that were brought into the Proposition 72 debate, which would lead voters to conclude that the cost was small, were often grossly misleading. Four facts in particular – that very few firms are affected, that large firms already offer health insurance, that firms already pay for 80 percent of premiums, and that the cost of the uninsured is modest – all fail to stand up to careful inspection. Second, my own work in Yelowitz (2003, 2004) carefully models the provisions of HIA using credible, publicly available data. In that work I found that the costs of HIA to employers were more than ten times what was often being quoted. As a consequence of this cost, there would have been a very strong likelihood of labor market adjustments – wage shifting and job loss in particular – as profit-maximizing employers tried to adjust. Third, I explore the viability of employer mandates more generally. I propose a scaled-down employer mandate for workers-only that exploits the most promising features of the group health insurance market while cognizant of the possible labor market consequences. My policy proposal explicitly recognizes a) the financial externality of uncompensated care, b) the moral hazard and economic inefficiency from excessive health insurance, c) the idea that mandate is not “free” to employees, d) that employment loss is a real possibility for individuals near the minimum wage, and e) that recent research in behavioral economics could make implementation of the mandate easier while preserving individual choice.

Acknowledgements: A number of people have given valuable input on this draft, as well as my other research on pay-or-play mandates. A special thanks to Debbie Reed, David Neumark, Ellen Hanak, Craig Garthwaite, Larry Yelowitz, and Beth Yelowitz for their valuable comments.

[2] For example, the San Francisco Chronicle noted that the “No on 72” campaign featured advertisements that represented an actress as a restaurant owner, and in a restaurant that employed 12 people and would therefore be exempt from HIA’s requirements. See http://www.sfgate.com/cgi-bin/article.cgi?file=/chronicle/archive/2004/10/04/BAG9S93CAE1.DTL .

[6] See KFF/HRET, October 5, 2003, “The Health Insurance Act of 2003 (SB2): Updated Findings from the 2002 California Employer Health Benefits Survey,” Chart 6.

[7] I obtain these percentages by multiplying the offering rate with the coverage rate (amongst firms that offer coverage). For example, in firms with 20 to 49 employees, 84% of firms offer coverage and 72% of employees are covered in firms that offer. Thus, I compute 60% of employees taking up coverage through their own employer.

[9] Note, however, that the offering firms still have employee take-up of far less than 100 percent. Thus, for workers that are uncovered, there is still a substantial cost.

[10] Dube and Reich examined firms with 20 or more employees, rather than 50 or more employees. By doing so, the median firm size would be smaller than what I compute here, simply reemphasizing my main point.

[12] Healthy Families is low cost insurance for children and teenagers in California. It provides health, dental and vision coverage to children who do not have insurance and do not qualify for free Medi-Cal.

[13] This elasticity means that a 10 percent increase in wages leads to a 2.2 percent reduction in full-time equivalent employment.

[14] Jonathan Gruber’s (2005) textbook, “Public Finance and Public Policy” explains this point nicely. On page 531 Gruber states “When there are barriers to reaching the competitive equilibrium (as in this minimum wage example), the side of the market on which the tax is levied can matter. There are a number of potential barriers, ranging from the minimum wage to workplace norms, that do not allow employers to explicitly cut workers’ wages. Such rigidities are often not present in output markets. For this reason, the party on whom the tax is levied may matter more in input than in output markets.”

[16] “Infra-marginal costs” refer to the costs to firms that already offer health insurance. Recall that between 2 and 8 percent of firms (depending on firm size) would see their employer contributions substantially raise due to the mandated single worker coverage under HIA. A high-deductible plan, as is proposed here (rather than the more comprehensive coverage under HIA) would lessen the cost impact for these firms. Moreover, high-deductible plans such as these, by forcing the individual to realize the full marginal cost of health care until the deductible is reached, tends to reduce moral hazard.