View
213
Download
0
Category
Preview:
Citation preview
Mandating Insurance Offers for Low-Wage Workers: An Evaluation of
Labor Market Effects
Amy WolaverBucknell University
Timothy McBrideUniversity of Missouri–St. Louis
Barbara WolfeUniversity of Wisconsin–Madison
Abstract Employing a simultaneous model of part-time status, health insuranceoffers, and wages, we examine the impacts on employment and health insurance cov-erage of nondiscrimination rules in the tax code governing employer-sponsored healthinsurance. Using 1988 and 1993 Employee Benefits Supplements to the Current Pop-ulation Surveys and variations in health insurance premiums and minimum wages, wefind that health insurance coverage among low-wage primary earners is increased byat most 31 percent by the policy, at a cost of an estimated 0.8–5.4-percentage-pointdecrease in full-time employment for low-wage workers.
Since the failure of the Clinton health care reform initiative, comprehen-sive health insurance coverage in the United States has fallen off the polit-ical map, to be replaced by piecemeal measures to extend or improve thegenerosity of coverage for various subgroups of the population. Despitethese efforts and a booming economy in the 1990s, a significant propor-tion of the population, 14 percent in 2000 (Mills 2001), remained with-out health insurance coverage for the entire year. Lack of insurance is aproblem because the uninsured disproportionately go without neededmedical care or delay care. In 2002, 28 percent of uninsured adults did notreceive needed medical care, compared to only 6 percent of insured per-sons. The uninsured are also more likely than the insured to postpone care(41 percent compared to 20 percent) and more likely to skip filling a pre-scription (26 percent compared to 11 percent) (Kaiser Family Foundation2003).
Journal of Health Politics, Policy and Law, Vol. 28, No. 5, October 2003. Copyright © 2003 by DukeUniversity Press.
With the exception of the boom economic years, 1999 and 2000, thenumber of uninsured in the United States had been steadily growing and,in particular, low-skilled workers lost ground relative to high-skilledworkers. Between 1979 and 1993, employer-provided health insuranceamong workers with the lowest wages (the bottom fifth) fell thirteen per-centage points, whereas among those in the highest wage group (the topfifth) coverage fell only three percentage points. This trend continued intothe mid-1990s: between 1987 and 1996, employer-provided coverage oflow-wage workers (those earning less than $7 per hour) fell from 54 to 42percent, whereas coverage of high-wage workers (those earning more than$15 per hour) increased from 87 to 90 percent (O’Brien and Feder 1999).Early anecdotal reports of low-income families improperly losing Medic-aid coverage in the wake of the welfare reforms and joining the ranks ofthe uninsured (Pear 1999) have since been confirmed with data from fivestates (Ellwood 1999), making it clear that the problem has grown worse.
The majority of the uninsured are in families of the working poor.Among the uninsured in 1999, 71 percent were in families with at leastone full-time worker, and 65 percent were in families with incomes lessthan 200 percent of the federal poverty level (Kaiser Family Foundation2001). Lack of coverage among children is mostly found in families withworking parents. In 2000, 7.3 million of the 9.2 million parents of unin-sured children worked, 5.4 million of them full time (Hoffman and Pohl2002). In general, these are families that earn too much to qualify forMedicaid but do not have access to employer-sponsored health insuranceor cannot afford the premiums.
In this article, we argue that the nondiscrimination rules governing thetreatment of employer-sponsored health insurance have had the unin-tended side effect of harming low-wage workers, leading to underem-ployment and little, if any, increase in the probability of coverage. Of par-ticular concern is the nondiscrimination rule included in the tax code in1978 to extend self-insured health insurance coverage to low-wage work-ers. Since passage of the nondiscrimination laws in 1978, employers withself-insured plans have been obliged to offer fringe benefits to essentiallyall full-time workers (not just highly compensated employees) in orderto receive the full benefits of the favorable tax treatment. Thus pensions,health insurance, and a variety of other benefits—if offered—must beoffered across the board to all full-time employees. The intended impactof this policy was to increase the access of low-wage workers to these ben-efits, including employer-provided health insurance, with its tax and risk-pooling advantages over the individual health insurance market. We argue
884 Journal of Health Politics, Policy and Law
here that this policy, meant to benefit low-wage workers, may havedecreased full-time employment for the group that the law was intendedto help. The details of the rules and a copy of the text are found in Appen-dix A.
In the next section we describe the provisions of the tax code govern-ing employer insurance and use economic theory to generate hypothesesabout how this policy might lead to negative employment effects for low-wage workers. Then we use data from the Current Population Survey(CPS) and other sources to test these hypotheses, using a model of part-time employment status, health insurance offers, and wages. We commentupon the implications of these findings for tax policy and on initiativesto encourage increased health insurance coverage.
How Can Current Tax PolicyDisproportionately Harm Low-WageWorkers?
Our current health insurance system is heavily weighted toward theemployer provision of health insurance for two reasons. First, the grouphealth insurance market has several advantages over the individual mar-ket, including economies of scale in administrative costs and the sharingof risk.1 The second reason is the tax exemption for fringe benefits thathas been in place since World War II. Workers pay for the insurance inpretax dollars, increasing their demand for compensation in this formrather than in the form of wages.
Nondiscrimination rules amended the tax code in Section 105(h) of Title26 in 1978. If the nondiscrimination rules are violated at firms that self-insure, the differential value of the employer’s premium contribution toemployees that are highly compensated relative to non–highly compen-sated employees is added to taxable income. In Section 89 of the 1986 TaxReform Act (Pub. Law 99–514) an abortive attempt was made to strengthenthe nondiscrimination rules, applying them to all plans and setting new,complex standards for compliance. The new rules were intended to go intoeffect on 1 January 1989, but before the end of 1989 Congress retroac-tively repealed the Section 89 rules in the face of protests from small busi-nesses and others, so that they were never enforced (Pub. Law 101–140).Congress nevertheless favored retaining some nondiscrimination rules
Wolaver et al. � Mandating Insurance Offers for Low-Wage Workers 885
1. Economies of scale in insurance purchase exist because of fixed administrative costs thatmust be covered regardless of group size. Large groups also avoid an adverse selection premium.
because, first, tax expenditures for employer-provided health insuranceshould not disproportionately favor highly compensated workers and, sec-ond, because the rules were felt “to promote(s) the broad health care cov-erage of rank-and-file employees” (Rostenkowski 1989).
Expanding health insurance coverage is a laudable goal, but anothereffect is possible. The nondiscrimination clause indirectly raises the effec-tive minimum wage for firms that offer a self-insured health insurance pol-icy. Economic theory regarding mandated benefits predicts that, to theextent that workers value health insurance, they will be willing to acceptlower wages in exchange for the benefit. If health insurance does not affectproductivity, firms should care only about the total compensation paid toworkers, not the mix of wages and health insurance.2 If wages can be low-ered enough to cover the full cost of the benefit, then firms’ demand forworkers will not change as a result (Summers 1989; Gruber and Krueger1991; Gruber 1999).3
However, as Figure 1 shows, minimum wage laws interfere with thisadjustment by preventing wages from falling below the legal floor. The leftgraph shows a hypothetical labor market for highly skilled workers. Inresponse to the increased costs of labor from providing health insurance,firms decrease their demand, shifting the demand curve from D0 to DHI bythe amount of the per-worker cost of insurance. A change in labor supplyis also likely, making the total effect on employment ambiguous. If work-ers value receiving the health insurance, they will increase their supply oflabor; in the case shown, from S0 to SHI, workers value the insurance dol-lar for dollar and employment levels do not change. Wages adjust down-ward, from W*0 to W*HI, exactly covering the cost of the insurance to theemployer. The second panel shows the same effects for a group of low-skilled workers, but in this case, the minimum wage law prevents wagesfrom adjusting fully, and the quantity of full-time, low-skilled workers sup-plied, ESHI, exceeds the quantity demanded, EDHI. Clearly, as minimumwages are raised or as health insurance premiums rise (causing the demandcurve to shift even further), this constraint becomes more binding andlarger numbers of full-time workers face the risk that employers will reduce
886 Journal of Health Politics, Policy and Law
2. The form of compensation will affect payroll taxes, about which, at first blush, firms maybe concerned. However, Brittain 1971 shows that payroll taxes are passed onto the worker inthe form of reduced wages, therefore negating their effect on employment. They may also resultin reduced employment for low-wage workers, but since they are a variable cost and not a fixedcost per worker, we do not expect to see employment effects for payroll taxes as large as thosefor health insurance.
3. It could also be argued that health insurance could increase firms’ demand for labor, ifhealthier workers are more productive workers, decreasing the likelihood of adverse employ-ment effects.
demand for their services. Health insurance is a benefit even more likely tocause employment effects than pensions, because health insurance is afixed cost per worker whereas pension contributions are likely to be a vari-able cost tied to wages and earnings. As a result of the fixed cost, employer-provided health insurance (EHI) is a higher fraction of total compensationfor low-wage workers than for high-wage workers, meaning that the healthinsurance provision is more likely to hit the minimum wage constraint thanthe pension provision. Even if these workers would be willing to take jobswith wages that were below the legal minimum but that also offered healthinsurance, federal and state laws would prohibit this arrangement. The pol-icy may therefore raise the labor costs of certain workers above the mar-ket value of their labor.
The nondiscrimination clause acts, then, as a partial employer mandate.Employers are not compelled to offer any health insurance to any work-ers, but offers cannot be extended selectively to only the highly skilledworkers. If firms employ some workers whose productivity is less than thehourly cost of insurance plus the minimum wage and other workers whoseproductivity is above this level, they face a problem in complying with thelaw. Firms may want to cover very productive workers to attract higher-quality workers than their competitors or to reduce turnover costs, but ifthey offer health insurance to those workers, they must also cover theirfull-time, low-skilled employees. Firms would be forced to pay these low-skilled workers more than they are worth to the firm.
Wolaver et al. � Mandating Insurance Offers for Low-Wage Workers 887
Wages
W0
WHI
S0 SHIWages
W0
WHI
S0
SHI
Minimumwage
E*0 = E*HIHigh-skilled,Full-time Workers
EDHI <ESHILow-skilled,Full-time Workers
Figure 1 Labor Market Changes with Employer Provision of HealthInsurance
Firms have several ways to avoid this situation. First, they could chooseto not sponsor health insurance entirely or to lower the generosity of cov-erage for all employees. In this case, the nondiscrimination clause mayactually decrease total health insurance coverage. Second, firms coulddecide not to self-insure; they would therefore not be subject to the clause.This option is costly for firms, since they would then become subject tostate benefit mandates and state premium taxes. Third, firms could cheat:they could continue to offer health insurance to only their highly skilledemployees but run the risk of discovery and the loss of the favorable taxtreatment. Fourth, firms could alter waiting periods and contribution ratesto minimize the coverage of their low-wage workers. Fifth, firms couldreduce the number of low-wage workers but increase the hours of theremaining workers to spread the costs of the health insurance over morehours. Finally, since the rule applies solely to the coverage of full-timeemployees, firms may hire low-skilled workers under alternative workarrangements, legally freeing them from offering these workers the samelevel of health insurance benefits that they provide to highly compensatedemployees while retaining the favorable tax treatment. These arrange-ments might include part-time or part-year work, hiring temporary work-ers, or contracting/outsourcing tasks formerly done by direct employeesof the firm. Although any response that results in a loss of coverage is ofconcern to us, the employment effect is the primary area of concern. Low-skilled workers would not only remain largely uninsured but some wouldalso lose full-time employment.
Not all non–highly compensated employees under the legal standardare likely to fall into the category of “at risk for employment effects.”Workers who fall into this category have low productivity relative to theminimum wage and to their health insurance premium costs, as clearlyshown in Figure 1, and are a subset of the non–highly compensated employ-ees as defined in the tax code. Other forces and policies undoubtedly alsoaffect the provision of health insurance by firms. For instance, there aremarket incentives to offer broad coverage within the firm, such as shiftinghealth risk from older, highly compensated employees to younger employ-ees with lower compensation. This effect would tend to lower the proba-bility of finding employment effects as a consequence of the nondiscrim-ination rules.
In this analysis we assume that changes in other policies and marketforces over time and across states are similar for our two defined groupsof workers, but we test this assumption through a variety of robustnesschecks. We identify policy effects by using changes in the minimum wage
888 Journal of Health Politics, Policy and Law
and health insurance premiums over time and across states. If the nondis-crimination rules have the impacts we anticipate, we expect to see the fol-lowing differences between high- and low-wage workers:
� Low-wage workers should be less likely to work at firms that offerany health insurance coverage;
� At firms where coverage is offered to some employees, low-wageworkers should be more likely to be ineligible for those offers;
� Low-wage workers should be more likely to be part time and ineli-gible for the health insurance offers at firms where coverage is offeredto some employees; and
� The difference between low- and high-wage workers in the rates ofjoint part-time status and ineligibility for firm offers should increasewith firm size, because rates of self-insurance increase with firm size.
We can estimate the magnitudes of the potential benefits and costs ofthis policy by comparing the rates of health insurance offers, coverage,and part-time employment between low- and high-wage workers. If thereare no differences between high- and low-wage workers, then the law, incombination with the minimum wage, is not binding and employmenteffects should be of little concern to policy makers.
Previous theoretical and empirical work on the effects of employer man-dates similar to this policy has found few or no employment effects. Thatwork, however, concentrated on unemployment as the outcome, rather thanon part-time or part-year work or other alternative work arrangements(Summers 1989; Gruber and Krueger 1991). Furthermore, the employmentof low-wage workers has not been examined separately from that of high-wage workers; aggregate analysis may hide the detrimental effects of thepolicy on low-wage workers. It is theoretically possible, depending on therelative value workers place on the health insurance, for benefits to increasethe employment of some workers at the same time as the employment ofothers is decreased, resulting in little or no measured aggregate effect onemployment.
The important question in evaluating any policy is whether the benefitsoutweigh the costs. The cost of this policy is fewer full-time, permanentjobs for low-wage workers; the benefit is increased health insurance cov-erage for other workers. A brief review of economy-wide trends is con-sistent with an assumption that costs outweigh benefits. Over the past twodecades, alternative work arrangements, from temporary services and con-tracted work to part-time work, have increased in frequency. At the sametime, health insurance coverage for the working poor has eroded, opening
Wolaver et al. � Mandating Insurance Offers for Low-Wage Workers 889
the door for other policy interventions. Studies such as Farber and Levy2000 found that lack of eligibility was responsible for the decline in cov-erage for part-time workers (old and new) during their period of analy-sis. Garrett, Nichols, and Greenman 2001 found that nearly 90 percentof workers in firms that offer coverage are eligible for coverage but thatpart-time workers and workers with short tenure (less than one year) arethe least likely to be eligible. These statistics are complemented by sur-vey data that find that 21.3 percent of employers cite “saving on wageand/or benefits” as one reason for hiring part-time workers (Houseman2001). And Blumberg and Nichols 2001: 39–40 states that
According to the 1997 Contingent Worker Supplement to the CPS, 3.7million (about 18 percent) of 20.3 million uninsured workers were inoffering firms but were not eligible for that coverage. An additional 6.4million workers with insurance coverage from another source were alsoineligible for their own employers’ coverage. Of all those ineligible fortheir ESI [employer-sponsored insurance] offers, 53 percent reportedthe reason as being that they don’t work enough hours per week orweeks per year. About 8 percent said that contract or temporary work-ers are not allowed in the plan. Twenty-seven percent said that they hadnot worked for the employer long enough to qualify, and 1 percent citeda pre-existing condition. About 11 percent cited other reasons. . . .While the lack of eligibility for workers with an offering employer rep-resents a minority of those workers who are uninsured, it is a statuswhich appears to be growing in importance over time. Between 1988and 1997, eligibility conditional on an employer sponsored insuranceoffer fell from 94.3 percent to 91.3 percent. But eligibility for peripheralworkers (those on the job less than a year and those in part-time jobs)fell more dramatically, from 79.8 percent to 69.6 percent, over the sameperiod. The largest percentage point decline (23) was for part-timeworkers in jobs for less than a year— their eligibility fell from 58.6 per-cent to 35.5 percent. Part-time workers in longer standing jobs saw theireligibility rates fall by 10 percentage points, to 67 percent in 1997, andeligibility for full-time workers in new jobs fell to 80 percent.
Methods
Our analysis estimates the probability of part-time work, health insuranceoffers, and wages simultaneously, since these aspects of a job will bedetermined together and affect each other. We create a categorical vari-
890 Journal of Health Politics, Policy and Law
able that equals 0 if the worker is full time with offer of health insurance,1 if full time without offer, 2 if part time with offer, and 3 if part time with-out offer. Although it is logical that full-time employment with an offer ofinsurance is the most desirable category and part time without an offer theleast, it is not as clear whether part-time employment with an offer dom-inates full time without. To allow for flexibility, we model this categori-cal variable using multinomial logit specification. (See Appendix B for adetailed description of the econometric methods and for selected regres-sion results.)
These equations are estimated simultaneously, for the full sample andfor the low- and high-wage samples separately, using generalized methodof moments. They are run in two stages, first with the probability that afirm offers health insurance to any worker and second with the probabil-ity that the worker is individually offered insurance. Because of the simul-taneous nature of the estimations, covariates have both direct and indi-rect effects on the outcomes. The simultaneity in combination with thenonlinearity of the multinomial logic makes direct interpretation of thecoefficient estimates impossible. Therefore, in addition to presenting coef-ficient estimates in Appendix Table B1, we construct predicted probabil-ities and change the values of one variable of interest at a time to infer thetotal direct and indirect marginal effects.
In order to control for the possibility that the worker works part timevoluntarily and to help identify the model, we include household charac-teristics, such as the presence of a child younger than age six, the femaledummy variable interacted with a child younger than age six, and house-hold size, in the insurance/part-time status equations. We also include theinsurance status of the spouse and an imputed premium to capture theexpected cost to the firm (see below) in these equations. In the wage equa-tion only, we include region indicators to control for labor market condi-tions; we do not expect preferences for part-time work or health insuranceto differ by region. We also use two definitions of part-time work in theanalyses: all part-time workers and those who indicate that they are parttime for economic reasons (our involuntary part-time measure). The esti-mates using the latter definition shift some part-time workers to full-timeemployees though they work part-time hours.
Data
The data are pooled from the 1988 and 1993 Employee Benefits Supple-ments to the Current Population Surveys, which ask questions about cur-
Wolaver et al. � Mandating Insurance Offers for Low-Wage Workers 891
rent employment and benefits. We restrict our analysis to men and singlewomen because we are most concerned with involuntary part-time employ-ment, and married women are traditionally more likely to desire fewerwork hours and flexible work arrangements. This restriction strengthensour conclusions regarding the employment effects of the policy. The pri-mary advantage of these data is the information on whether or not a firmoffers coverage, even if the individual was not eligible for the offer orturned down coverage. We concentrate on these years because the growthof health insurance premiums was substantial and the minimum wageincreased over this time period.4
The Current Population Survey does not contain information aboutactual health insurance premiums for the respondents. Instead, we use animputation based on regression results, using the 1987 National MedicalExpenditure Survey (NMES) and worker and job characteristics to pre-dict the cost of health insurance for each worker. These characteristicsinclude job hazards merged from the Dictionary of Occupational Titles,occupation, industry, firm size, marital status, household characteristics,education, age, job tenure, urban–rural status, and region. We estimate themodel using a maximum-likelihood exponential regression on the loggedpremium and construct the predicted log premium using the coefficientestimates and the characteristics of the CPS respondents. By using theexponential regression instead of a linear regression on the premium lev-els, we avoid the retransformation problem. Appendix Table B1 containsthe premium regression results from the 1987 NMES. We inflate the pre-mium using the medical care price index for individuals in the 1988 and1993 sample and then use the CPI-U to convert the values into real 1988dollars.
Our method of identifying the effects of the law is to divide workersinto two groups, those who are at risk for employment effects and thosewho are not, strictly on the basis of the theoretical model. We do this byusing the real minimum wage (in 1988 dollars) in the year and state andthe real hourly cost of health insurance for each worker. A low-wageworker is defined as a worker whose wages are less than the hourly costof health insurance plus the minimum wage. By this definition, 11.7 per-cent of primary earners are considered to be low-wage workers. It is pos-sible, however, because the minimum wage and health insurance premi-ums vary, that some workers classified as high wage might have lower
892 Journal of Health Politics, Policy and Law
4. From 1995 to 1999 health insurance premiums grew at a lower rate than cash compensa-tion. This has now reversed again as premium growth is outstripping the growth of cash com-pensation (see below for more detail on this).
hourly wages than some workers classified as low wage. Approximately9.6 percent of high-wage workers have lower hourly wages than the max-imum wage in the low-wage sample.
Because high- and low-wage workers vary in other characteristics,Table 1 provides descriptive statistics for selected variables used in theanalysis for the full sample and for the high- and low-wage samples sep-arately. We focus on employment effects, using results from multipleregression analysis controlling for many factors that affect the probabil-ity of health insurance offers, wages, and part-time work.
Results and Analysis: Did theNondiscrimination Clause Increase theCoverage of Low-Wage Workers?
The nondiscrimination clause is limited in its potential for expandinghealth insurance coverage because it applies only to businesses that offerself-insured health insurance plans. These firms, however, employ a sub-stantial portion of the workforce. A conservative estimate is that 37 per-cent of primary earners are offered a self-insured plan. Larger firms aremore likely to self-insure; estimates of the percentage of insured employ-ees in self-insured plans vary from 11.1 to 16.5 percent in firms with fewerthan one hundred employees to 57.8 to 62.6 percent in firms with morethan five hundred employees in 1987 and 1993 (Acs et al. 1996). On thebasis of these data and the distribution of high- and low-wage workers byfirm size in the CPS data, we estimate that about 27 percent of low-wageworkers work at firms that offer health insurance and self-insure, com-pared to 39 percent of high-wage workers.5 The difference in probabili-ties between low- and high-wage workers is a product of the lower prob-ability that low-wage workers work at large firms as well as the lowerprobability that they work at firms that offer any insurance.6
The nondiscrimination clause does not provide any means to increasethe take-up of insurance offers. Some workers— indeed, a growing per-centage— turn down insurance that is offered to them by their employer,
Wolaver et al. � Mandating Insurance Offers for Low-Wage Workers 893
5. Authors’ calculations, based on the distribution of firm size by wage group, firm offers bywage group, and the distribution of employees in self-insured plans by firm size from Acs et al.1996. Among workers who are offered health insurance, however, low- and high-wage workersare closer in the probability that the plan is self-insured—-40 and 42 percent for low- and high-wage workers, respectively.
6. Throughout this article, some empirical results will be cited that cannot be found in theaccompanying tables or graphs. These findings are based on research completed by the authorsand further explanations or details are available upon request.
at least in part because their costs increased owing to higher employeepremium shares and higher premiums. The average proportion of the pre-mium paid by the employer has declined: for single coverage, between1988 and 1996, the proportion paid by small firms (those with fewer thantwo hundred employees) fell from 88 to 67 percent, whereas in large firmsit fell from 87 to 78 percent (Gabel, Ginsburg, and Hunt 1997). Someworkers who turn down coverage are insured by other means, perhapsthrough a spouse, but many remain uninsured.
As Table 1 shows, low-wage workers are less likely to have potential cov-erage through a spouse and are also less likely to have any coverage. Ouranalysis shows that 20 percent of primary earners (men and single women)do not have employer-provided insurance themselves or access through aspouse; 11 percent of these workers therefore have no insurance coverageat all. Further examination of these workers (not in Table 1) shows that 25percent of the workers without their own insurance were not offered cov-erage through their employer; the rest turned down an offer. Only a smallgroup of ineligible workers is employed at firms that offer insurance bene-fits to some, but not all, of their employees. Figure 2 illustrates this point;93 percent (81.3/87.1 percent) of workers at firms that offer insurance to anyof their employees are eligible for the offered health insurance.
The differences across wage groups and between part-time and full-time status are clear. Recall our hypothesis that employers would like toavoid offering health insurance coverage to low-wage workers because theminimum wage prevents the worker from paying for the coverage vialower wages. Figure 2 illustrates the percentage of workers at firms thatoffer insurance to any employees and the percentage of workers who havetheir own offer of health insurance, according to whether they work partor full time and their wage status. We expect that high-wage workers aremore likely to work at firms that offer health insurance and more likely tobe eligible for that offer. Because the nondiscrimination clause excludespart-time workers, we expect part-time workers to have fewer offers thanfull-time workers.
Over 90 percent of the employers of high-wage, full-time workers offerhealth insurance, and almost all workers are individually eligible for thatoffer (Figure 2). Low-wage workers, however, are less likely to work atfirms that offer any coverage, are less likely to be individually eligible forthat coverage, and are less likely to accept those offers. Only 83 percent(50.8/61.5 percent) of low-wage, full-time workers are eligible for theoffered firm coverage. There are several possible reasons for the differ-ences in the eligibility of low-wage, full-time workers relative to high-
894 Journal of Health Politics, Policy and Law
Tab
le 1
Var
iab
les
Mea
ns
and
Sta
nd
ard
Dev
iati
on
s (i
n P
aren
thes
es)
Wei
ghte
d H
igh-
Wag
e L
ow-W
age
Var
iabl
eFu
ll Sa
mpl
eSa
mpl
eSa
mpl
eSa
mpl
e
Low
-wag
e w
orke
r.1
17(.
32)
.117
(.32
)–
–Fi
rm o
ffer
s in
sura
nce
.87
(.34
).8
7(.
34)
.91
(.29
).5
6 (.
50)
Wor
ker
is e
ligib
le f
or o
ffer
.83
(.38
).8
2 (.
38)
.88
(.33
).4
2 (.
49)
Invo
lunt
arily
par
t-tim
eFi
rm o
ffer
s in
sura
nce
.020
(.0
61)
.020
(.14
).0
14(.
12)
.067
(.25
)Fi
rm d
oes
not o
ffer
insu
ranc
e.0
19(.
14)
.020
(.14
).0
09(.
09)
.099
(.30
)Fu
ll-tim
e,a
firm
off
ers
insu
ranc
e.8
5(.
36)
.85
(.36
).8
9(.
31)
.49
(.50
)In
volu
ntar
ily p
art-
time
Indi
vidu
al o
ffer
of
insu
ranc
e.0
13(.
11)
.012
(.11
).0
10(.
10)
.032
(.18
)N
o in
divi
dual
off
er o
f ins
uran
ce.0
27(.
16)
.027
(.16
).0
13(.
11)
.13
(.34
)Fu
ll-tim
e,a
indi
vidu
al o
ffer
.81
(.39
).8
1(.
39)
.87
(.34
).3
9(.
49)
Cov
ered
by
own
empl
oyer
.76
(.42
).7
6(.
43)
.82
(.38
).3
1(.
46)
Not
cov
ered
by
any
sour
ce.1
1(.
31)
.11
(.32
).0
71(.
26)
.40
(.49
)A
nnua
l pre
miu
m (
$)22
49(6
03)
2250
(604
)22
74(6
08)
2065
(523
)R
eal w
age
($)
10.9
0(6
.05)
10.9
3(6
.10)
11.8
2(5
.84)
3.90
(.75
)In
volu
ntar
ily p
art-
time
.040
(.19
).0
39(.
19)
.023
(.15
).1
7(.
37)
Part
-tim
e.1
0(.
30)
.10
(.30
).0
74(.
26)
.32
(.46
)A
ge (
year
s)37
.8(1
1.4)
37.5
(11.
4)38
.4(1
1.1)
33.1
(12.
5)Fe
mal
e.2
9(.
45)
.30
(.46
).2
6(.
44)
.50
(.50
)Fi
rm 1
,000
+ e
mpl
oyee
s.4
5(.
50)
.46
(.50
).4
8(.
50)
.27
(.44
)Fi
rm <
100
em
ploy
ees
.39
(.49
).3
9(.
49)
.37
(.48
).5
8(.
49)
Hou
seho
ld s
ize
2.50
(1.8
7)2.
49(1
.88)
2.48
(1.8
4)2.
59(2
.07)
Tab
le 1
Var
iab
les
Mea
ns
and
Sta
nd
ard
Dev
iati
on
s (i
n P
aren
thes
es)
(Co
nti
nu
ed)
Wei
ghte
d H
igh-
Wag
e L
ow-W
age
Var
iabl
eFu
ll Sa
mpl
eSa
mpl
eSa
mpl
eSa
mpl
e
Pres
ence
of
child
< 6
.18
(.38
).1
8(.
39)
.18
(.39
).1
5(.
36)
Fem
ale
�Pr
esen
ce o
f ch
ild <
6.0
25(.
15)
.024
(.15
).0
19(.
14)
.066
(.25
)H
igh
scho
ol e
duca
tion
.57
(.49
).5
7(.
49)
.56
(.50
).6
5(.
48)
Col
lege
edu
catio
n.3
1(.
46)
.31
(.46
).3
4(.
47)
.10
(.30
)N
onw
hite
.13
(.34
).1
4(.
35)
.13
(.41
).2
0(.
40)
Les
s th
an 1
yea
r jo
b te
nure
.1
6(.
37)
.17
(.37
).1
3(.
34)
.40
(.49
)U
nion
mem
bers
hip
.23
(.42
).2
3(.
42)
.25
(.43
).0
68(.
25)
N24
,039
24,0
3921
,234
2,80
5
Sour
ces:
198
8 an
d 19
93 E
mpl
oyee
Ben
efits
Sup
plem
ents
, Cur
rent
Pop
ulat
ion
Surv
eys,
Pri
mar
y E
arne
r Sam
ple
(U.S
. Cen
sus
Bur
eau
and
U.S
. Bur
eau
of L
abor
Stat
istic
s 19
88, 1
993)
.a F
ull-
time
+ v
olun
tary
par
t-tim
e w
orke
rs.
wage, full-time workers. Low-wage workers are more likely to work atsmaller firms, which are less likely to self-insure and therefore not sub-ject to the nondiscrimination law. Moreover, part-time status is not theonly exemption in the clause; firms were also exempt from includingworkers under the age of twenty-five, part-year workers, and workers withfewer than three years of job tenure. Each of these characteristics is alsoassociated with lower wages.
If we assume that no low-wage, full-time workers would have beenoffered coverage by an employer in the absence of the law, an upper-bound value for the benefit of the law can be estimated by the fraction ofthese workers who are offered and accept employer-provided health insur-ance coverage. From Table 1, 42 percent of low-wage workers are eligi-ble for an offer, but only 31 percent of low-wage workers are covered bytheir own employer. Assuming that none of these workers would have
Wolaver et al. � Mandating Insurance Offers for Low-Wage Workers 897
Figure 2 Health Insurance Offers, by Wage and Part-time StatusNote: Males and single females only.Sources: U.S. Census Bureau and U.S. Bureau of Labor Statistics 1988, 1993.
100 -
90 -
80 -
70 -
60 -
50 -
40 -
30 -
20 -
10 -
0 -
Perc
ent
Low-wage,Part-time
Low-wage,Full-time
High-wage,Part-time
High-wage,Full-time
Total
44.4
23.6
61.5
50.8
76.1
63
92.289.9
87.1
81.3
Firm Offers Health Insurance Individual Is Offered Health Insurance
been offered insurance or would have been covered without the nondis-crimination clause—a strong assumption that maximizes the estimatedbenefits of the policy—our results suggest that the benefits of the clauseare an increase of thirty-one percentage points in coverage among low-wage workers. These estimates are clearly overly optimistic; since highlycompensated employees tend to be older, firms may have previouslyextended health insurance offers to the younger workers to increase therisk pool, subsidizing the costs of the older, less healthy employees.
Did the Nondiscrimination Clause ReduceFull-Time Employment for Low-WageWorkers?
Recall that economic theory predicts that low-wage workers, for whomthe minimum wage limits the wage adjustment that can be made to covertheir health insurance costs, should have higher rates than high-wageworkers of part-time work with ineligibility for coverage, who should notbe subject to any employment effects because their employers can fullylower wages to adjust for the health insurance premium. If workers arenot willing to accept lower wages (that is, they do not value the healthinsurance as much as the benefit costs the employer), then the high-wageand the low-wage groups should both manifest increased rates of part-time work and ineligibility for coverage; these preferences would be takeninto account in the estimates presented. The impact of the clause wouldbe overestimated, however, if low-wage workers are less willing thanhigh-wage workers to give up wages, and the observable characteristicsused in the regression analysis do not adequately control for this fact.
As discussed above, part-time or part-year hiring is among the availableoptions if firms want to provide insurance to only part of their workforce.Some employees prefer part-time work arrangements because they allowflexible scheduling, but many workers would prefer full-time employ-ment; these are the workers that are potentially harmed by the nondis-crimination rules. We have taken into account some of these differencesby controlling for factors that are associated with voluntary part-timework, such as the presence of a small child in the household and house-hold size. In our sample, the number of workers desiring flexible sched-uling should be small, because we selected only primary earners (men andsingle women). In one set of regressions, we also limit our definition ofpart time to only those who indicated that they were part time due to eco-nomic reasons. Our last alternative to take into account individual pref-
898 Journal of Health Politics, Policy and Law
erences regarding part-time work is to assume that the same fraction ofhigh- and low-wage workers desires part-time jobs. An estimate of theimpact of the law on part-time employment is, therefore, the differencebetween the fraction of low-wage workers and the fraction of part-time,high-wage workers who are ineligible for health insurance coverage. Basedon the raw regression results (sample in Appendix Table B2), in Table 2Awe predict employment effects of the following magnitude: 11 percent ofat-risk low-wage workers (those at firms that offer insurance) on averagewill be part time and ineligible for an offer of health insurance, comparedto only 1 percent of high-wage workers, producing an estimate of theeffect of the nondiscrimination clause of 10 percent.
Although we are examining only heads of households, some of theselow-wage workers could still be part time by choice. Thus, we also com-pare, in Table 2B, rates of coverage for high- and low-wage workers bylimiting the definition of what is a part-time job. The differences are smaller,but still striking; 5.8 percent of at-risk low-wage workers compared to 0.4percent of high-wage workers are both involuntarily working part timeand ineligible for coverage, resulting in an estimate of 5.4 percent increasedpart-time employment for at-risk workers. These differences are statisti-cally significant.
Differences by Firm Size
As discussed above, at any firm size, low-wage workers should be morelikely than high-wage workers to be part time and ineligible for coverage.Our first robustness test uses the variations in self-insurance status by firmsize to examine whether the law has an effect. Tables 2A and 2B show ourregression-based predicted probabilities that a worker is both part timeand ineligible for a health insurance offer according to firm size and wagestatus. (Standard errors are estimated by bootstrapping.) For high-wageworkers, the risk of part-time employment should be invariant across firmsize; low-wage workers should be increasingly part time and ineligible forcoverage as firm size increases.
We find that very few high-wage workers are both part time and ineli-gible for health insurance, regardless of employer size, whether we use allpart-time work or involuntary part-time work to define the dependent vari-able. Low-wage workers, in contrast, are more likely to be employed parttime (Table 2A) or to be involuntarily employed part time (Table 2B) andineligible for insurance in firms of all sizes. As firm size grows, the ratefor involuntary part-time workers grows from 4.3 to 7.2 percent, and for
Wolaver et al. � Mandating Insurance Offers for Low-Wage Workers 899
Tab
le 2
APr
edic
ted
Fra
ctio
n o
f Pr
imar
y Ea
rner
s W
ork
ing
Par
t-ti
me
and
Inel
igib
le f
or
Hea
lth
Insu
ran
ce C
ove
rag
e(S
tan
dar
d E
rro
r in
Par
enth
eses
): W
ork
ers
at F
irm
s Th
at O
ffer
Insu
ran
ce t
o a
t Le
ast
Som
e o
f Th
eir
Emp
loye
es
Low
-Wag
e W
orke
rs,
Firm
Siz
eH
igh-
Wag
e W
orke
rsL
ow-W
age
Wor
kers
Dif
fere
nce†
Hig
h-W
age
Coe
ffici
ent
Dif
fere
nce
(Nem
ploy
ees)
(a)
(b)
(b –
a)
(c)
(b –
c)
100
or f
ewer
.011
(.0
03)
.10
(.02
).0
9* (
.02)
.024
.076
100
–99
9.0
13 (
.003
).0
8 (.
02)
.07*
(.0
3).0
53.0
2710
00 o
r m
ore
.013
(.0
02)
.12
(.02
).1
1* (
.02)
.029
.091
Tota
l.0
12 (
.003
).1
1 (.
01)
.10*
(.0
1).0
27.0
83
Dif
fere
nce
in d
iffe
renc
eL
arge
st to
sm
alle
st fi
rms
.11
– .0
9 =
.02
(0.2
).0
91 –
.076
= .0
15
Not
e: C
ateg
ory
a an
d b
pred
ictio
ns b
ased
on
sam
ple-
spec
ific,
wei
ghte
d w
orke
r ch
arac
teri
stic
s an
d re
gres
sion
coe
ffici
ents
; cat
egor
y c
pred
ictio
ns u
sed
low
-w
age
wor
ker
wei
ghte
d ch
arac
teri
stic
s an
d hi
gh-w
age
wor
ker
regr
essi
on c
oeffi
cien
ts. P
rim
ary
earn
ers
are
all m
ales
and
sin
gle
fem
ales
, bet
wee
n th
e ag
es o
f 18
and
64. L
ow-w
age
wor
kers
are
defi
ned
as th
ose
wit
h w
ages
less
than
the
min
imum
wag
e pl
us a
n im
pute
d ho
urly
cos
t of
heal
th in
sura
nce.
Par
t-ti
me
is d
efine
das
all
part
-tim
e w
orke
rs. I
nvol
unta
ry p
art-
tim
eis
defi
ned
as p
art-
time
for
econ
omic
rea
sons
.So
urce
s: A
utho
rs’
calc
ulat
ions
fro
m r
egre
ssio
n re
sult
s us
ing
1988
and
199
3 E
mpl
oyee
Ben
efits
Sup
plem
ents
, Cur
rent
Pop
ulat
ion
Sur
veys
(U
.S. C
ensu
sB
urea
u an
d U
.S. B
urea
u of
Lab
or S
tatis
tics
1988
, 199
3).
†Boo
tstr
appe
d st
anda
rd e
rror
s.*S
tatis
tical
ly s
igni
fican
t at t
he .0
5 le
vel.
Tab
le 2
BPr
edic
ted
Fra
ctio
n o
f Pr
imar
y Ea
rner
s W
ork
ing
Par
t-ti
me
and
Inel
igib
le f
or
Hea
lth
Insu
ran
ce C
ove
rag
e(S
tan
dar
d E
rro
r in
Par
enth
eses
): W
ork
ers
Invo
lun
tari
ly P
art-
tim
e an
d In
elig
ible
fo
r C
ove
rag
e, b
y Fi
rm
Low
-Wag
e W
orke
rs,
Firm
Siz
eH
igh-
Wag
e W
orke
rsL
ow-W
age
Wor
kers
Dif
fere
nce†
Hig
h-W
age
Coe
ffici
ent
Dif
fere
nce
(Nem
ploy
ees)
(a)
(b)
(b –
a)
(c)
(b –
c)
100
or f
ewer
.003
6 (.
0018
).0
43 (
.013
).0
39*
(.01
3).0
08.0
3510
0–
999
.004
4 (.
0018
).0
72 (
.035
).0
68*
(.03
4).0
09.0
631,
000
or m
ore
.005
2 (.
0016
).0
72 (
.020
).0
67*
(.02
0).0
11.0
61
Tota
l.0
04
(.00
14)
.058
(.0
12)
.054
* (.
011)
.009
.049
Dif
fere
nce
in d
iffe
renc
e:L
arge
st to
sm
alle
st fi
rms
.067
– .0
39 =
.028
(.0
2).0
61 –
.035
= .0
26
Not
e: C
ateg
ory
a an
d b
pred
ictio
ns b
ased
on
sam
ple-
spec
ific,
wei
ghte
d w
orke
r ch
arac
teri
stic
s, a
nd r
egre
ssio
n co
effic
ient
s; c
ateg
ory
c pr
edic
tions
use
d lo
w-
wag
e w
orke
r w
eigh
ted
char
acte
rist
ics
and
high
-wag
e w
orke
r re
gres
sion
coe
ffici
ents
. Pri
mar
y ea
rner
s ar
e al
l mal
es a
nd s
ingl
e fe
mal
es, b
etw
een
the
ages
of
18an
d 64
. Low
-wag
e w
orke
rs a
re d
efine
d as
thos
e w
ith
wag
es le
ss th
an th
e m
inim
um w
age
plus
an
impu
ted
hour
ly c
ost o
f he
alth
insu
ranc
e. P
art-
tim
eis
defi
ned
as a
ll pa
rt-t
ime
wor
kers
. Inv
olun
tary
par
t-ti
me
is d
efine
d as
par
t-tim
e fo
r ec
onom
ic r
easo
ns.
Sour
ce: A
utho
rs’ c
alcu
latio
ns fr
om r
egre
ssio
n re
sults
usi
ng 1
988
and
1993
Em
ploy
ee B
enefi
ts S
uppl
emen
ts, C
urre
nt P
opul
atio
n Su
rvey
s (U
.S. C
ensu
s B
urea
uan
d U
.S. B
urea
u of
Lab
or S
tatis
tics
1988
, 199
3).
†Boo
tstr
appe
d st
anda
rd e
rror
s.*S
tatis
tical
ly s
igni
fican
t at t
he .0
5 le
vel.
those who are both part time it rises from 9 to 11 percent. It is striking thatthe rates of part-time work vary so little (by at most one-tenth of one per-cent) for high-wage workers as firm size varies; this finding is consistentwith our theoretical story that higher-income workers should not gener-ally be subject to employment effect because employers can adjust wagesto pay for the employees’ health insurance.
Robustness Tests of Worker PreferenceDifferences
Several other tests of the hypothesis are possible to account for other dif-ferences between high- and low-wage workers. First, we predict the frac-tion of low-wage workers who would be expected to be part time on thebasis of their observable characteristics and the coefficient results fromthe high-wage regressions (column c in Tables 2A and 2B). Intuitively,these estimates take into account differences in preferences between high-and low-wage workers captured by differences in their unobservable char-acteristics. In other words, the difference between column c and columna in Tables 2A and 2B is an alternative estimate of the fraction of low-wage workers who are involuntarily part time due to the policy. Underthese assumptions, the predicted effect of the law falls to between 4.9 and8.3 percent of workers (depending on which definition of part-time workis used as an outcome).
We can also identify a subgroup of the high-wage workers that is closerto low-wage workers than is the full sample of high-wage workers. Someworkers who in 1988 were defined as high wage would be reclassified aslow-wage workers under the conditions of 1993, when real health insur-ance premiums and the minimum wage were higher; we term this sub-group pseudo–low wage 1993 (column a in Table 3). The pseudo–lowwage 1993 workers are those whose wages in 1988 could be adjusteddownward to pay for health insurance, but who would be subject to employ-ment effects in 1993. This is a subgroup of high-wage workers who shouldbe more similar to low-wage workers. Similarly, some workers defined aslow wage in 1993 would be reclassified as high-wage workers under theconditions of 1988, when real health insurance premiums and the mini-mum wage were lower; we term this subgroup pseudo–high wage 1988(column b in Table 3). This is a subgroup of low-wage workers who shouldhave greater similarities with high-wage workers than with other low-wage workers. Finally, some workers would be defined as low wage ineither year (column c in Table 3).
902 Journal of Health Politics, Policy and Law
Tab
le 3
Co
mp
aris
on
s o
f a
Sub
set
of
Hig
h-W
age
Wo
rker
s to
Lo
w-W
age
Wo
rker
s
1988
Hig
h W
age,
1993
Low
Wag
e,
Wou
ld B
e W
ould
Be
All
Oth
er
1993
Low
Wag
e 19
88 H
igh
Wag
eD
iffe
renc
e L
ow-W
age
Wor
kers
Dif
fere
nce
(a)
(b)
(b –
a)
(c)
(c –
a)
Wor
kers
at fi
rms
that
off
er h
ealt
h in
sura
nce
Part
-tim
e6.
218
.36
30.2
(24.
2)(3
8.8)
(45.
9)In
volu
ntar
y pa
rt-t
ime
3.5
7.9
14.1
(18.
3)(2
7.0)
(34.
9)Pa
rt-t
ime
and
inel
igib
le3.
28.
35.
111
.68.
4(1
7.6)
(27.
7)(3
2.0)
Invo
lunt
ary
part
-tim
e an
d in
elig
ible
2.3
4.3
2.0
5.9
3.6
(14.
8)(2
0.3)
(23.
5)Pr
edic
tions
fro
m r
egre
ssio
n re
sults
(on
ly w
orke
rs a
t firm
s th
at o
ffer
insu
ranc
e)Pa
rt-t
ime
and
inel
igib
le1.
410
.59.
110
.89.
4In
volu
ntar
y pa
rt-t
ime
0.5
4.1
3.6
7.4
6.9
and
inel
igib
leN
umbe
r of
obs
erva
tions
31
444
611
28(t
hose
wit
h fir
m o
ffer
)
Not
es: C
ateg
ory
a w
orke
rs w
ould
cha
nge
cate
gory
fro
m h
igh-
wag
e to
low
-wag
e be
caus
e of
rea
l sta
te o
r fe
dera
l min
imum
wag
e in
crea
se o
r he
alth
insu
r-an
ce p
rem
ium
infla
tion
betw
een
1988
and
199
3. C
ateg
ory
b w
orke
rs w
ould
cha
nge
cate
gory
fro
m lo
w-w
age
to h
igh-
wag
e be
caus
e of
low
er r
eal s
tate
or
fed-
eral
min
imum
wag
e or
low
er h
ealt
h in
sura
nce
prem
ium
s in
the
earl
ier
year
. Cat
egor
y c
wor
kers
are
all
low
-wag
e w
orke
rs e
xcep
t for
thos
e in
cat
egor
y b.
The
pred
ictio
ns f
rom
the
regr
essi
ons
are
base
d on
the
actu
al w
orke
r ch
arac
teri
stic
s, in
clud
ing
the
sam
ple
year
they
are
dra
wn
from
, but
wor
kers
who
wou
ld s
witc
hw
age
cate
gori
es a
re g
iven
the
coef
ficie
nts
from
the
coun
terf
actu
al w
age
regr
essi
on. P
art-
time
is d
efine
d as
all
part
-tim
e w
orke
rs. I
nvol
unta
ry p
art-
time
is d
efine
das
par
t-tim
e fo
r ec
onom
ic r
easo
ns.
Sour
ces:
Aut
hors
’ ca
lcul
atio
ns f
rom
198
8 an
d 19
93 E
mpl
oyee
Ben
efits
Sup
plem
ents
, Cur
rent
Pop
ulat
ion
Surv
eys
(U.S
. Cen
sus
Bur
eau
and
U.S
. Bur
eau
ofL
abor
Sta
tistic
s 19
88, 1
993)
.
Table 3 shows the comparisons between these three groups based onboth the raw data and the predictions based on the regression results.These results are consistent with the hypothesis that the nondiscrimina-tion clause increased part-time employment for low-wage workers. Therates of all part-time work for the subgroup of high-wage workers who areclosest to low-wage workers are 3.2 percent, compared to 8.3 percent forthe subgroup of low-wage workers who are closest to high-wage work-ers and 11.6 percent of low-wage workers would not change status, regard-less of the sample year. The comparisons using involuntary part-time workas the dependent variable follow the same pattern: 2.3 percent comparedto 4.3 percent and 5.9 percent. When the predictions from the regressionresults are used, the effects become starker: only 1.4 and 0.5 percent ofthe subgroup of high-wage workers are predicted to be part time or invol-untarily part time and ineligible respectively, compared to 10.5 and 4.1percent for the low-wage subgroup and 10.8 and 7.4 percent for the alwayslow-wage subset, respectively.
These figures imply that the policy led to an estimated increase in part-time employment of between 2.0 and 8.4 percent, somewhat lower in mag-nitude than other estimates, but still a significant proportion of low-wageworkers at risk of employment effects. Using the predicted percentagesfrom the regression results to measure the differences, the effects arelarger, ranging from 3.6 to 9.4 percent of low-wage workers affected bythe legislation, depending on the definition of part-time work used.
Robustness Tests Accounting for Other EconomicChanges: Year Effects
It is possible that the methods used here overestimate the effects of thelaw on employment. These effects could, in part, be attributable to changesin the economy or other policies between 1988 and 1993. We account forchanges in the minimum wage and health insurance premiums, but ratesof part-time work were certainly increasing over this period, perhaps formany other reasons. Mishel, Bernstein, and Schmitt 1997 calculates thatrates of total part-time work increased from 18.1 to 18.8 percent between1989 and 1993, roughly the time period of the sample; rates of involuntarypart-time work increased from 4.3 to 5.5 percent. Our predictions arebased on the regressions, which control for year but are run over the com-bined 1988 and 1993 samples; they may thus be capturing other time dif-ferences in addition to minimum wage changes and inflation in healthinsurance premiums. As a test, we compare how the fractions of low- and
904 Journal of Health Politics, Policy and Law
high-wage workers who are predicted in Table 2C to be both part-time andineligible for coverage change between 1988 and 1993, using the coeffi-cients on the year dummies, for both definitions of part-time work. Usingthe involuntary part-time work definition, the impact of the law actuallydampens over time, so our results for this definition do not appear to bedue to other economic factors. The opposite is true for all part-time work,however; we estimate that those results may be inflated by as much as 1.8percentage points.7
Difference-in-Difference Estimates by Firm Size
As in most ex post studies of policy changes, the data for a perfect testof the effects of the nondiscrimination clause are not available. The dataused here are less than ideal in two respects: (1) The time period does notpre- and postdate the inception of the nondiscrimination rules; instead werely on variation in minimum wage policy and premium costs over timeand place to identify the role of the nondiscrimination policy. (2) The lack
Wolaver et al. � Mandating Insurance Offers for Low-Wage Workers 905
7. Percentage of the 3.3 (9.4 minus 6.1) percentage point increase in the difference betweenlow- and high-wage worker rates of part-time work and ineligibility.
Table 2C Predicted Fraction of Primary Earners Working Part-time andIneligible for Health Insurance Coverage: Involuntarily Part-time andIneligible for Coverage, by Year
High-Wage Workers Low-Wage Workers DifferenceSample Year (a) (b) (b – a)
All part-time1988 .017 .078 .0611993 .012 .106 .094
Involuntary part-time1988 .0061 .066 .0601993 .0044 .058 .054
Note: Category a and b predictions based on sample-specific, weighted worker characteristicsand regression coefficients. Primary earners are all males and single females between the agesof 18 and 64. Low-wage workers are defined as those with wages less than the minimum wageplus an imputed hourly cost of health insurance. Part-time is defined as all part-time workers.Involuntary part-time is defined as part-time for economic reasons.
Source: Authors’ calculations from regression results using 1988 and 1993 Employee Bene-fits Supplements, Current Population Surveys (U.S. Census Bureau and U.S. Bureau of LaborStatistics 1988, 1993).
of exact information on self-insured status hampers our identification.Firm size is the best correlate of self-insured status available and offerssome help in identification.8 If other policies or market changes explainedthe difference in involuntary part-time work between low- and high-wageworkers under our definition, they would not only have to operate differ-ently for both low- and high-wage workers but would also have differenteffects by firm size, coincident with minimum wage changes and growthin health insurance premiums cross-sectionally and over time.
An estimate of the policy’s effects that could potentially account for theabove possibility is the difference-in-difference method. This methodtakes, first, the difference in the rates of part-time work and insurance inel-igibility by wage status and, then, the differences in those rates betweenthe smallest and largest firms. Under this method, we estimate that thelaw might have increased the rate of part-time work among potentially eli-gible low-wage workers by 1.5 to 2.8 percent depending on the definitionof part-time work used (see the difference-in-difference in the fraction ofworkers who are part time and ineligible for coverage in Tables 2A and2B). Another, simpler comparison is the difference between rates of part-time work and ineligibility among low-wage workers at the largest andsmallest firms. Using this as a measure, the impact of the law ranges from2 to 2.9 percentage points.
The difference-in-difference estimates are best viewed as underesti-mates or lower bounds of the true effect of the legislation. In fact, theseestimates cannot be cleanly interpreted because rates of self-insurance arenot perfectly correlated with firm size. If no small firms and all large firms(or close to it) are self-insured, then the difference-in-difference resultswould be the best estimate of the impact of the clause. However, since 11.1(in 1988) or 16.5 (in 1993) percent of workers in small firms should beaffected by the legislation, compared to 60.9 (in 1988) or 60.6 (in 1993)percent of workers in large firms, the identification of the effects using thismethod is weakened. The results indicate that the nondiscriminationclause may affect as few as 1.5 and as many as 10 percent of low-wageprimary earners at firms that offer health insurance.
906 Journal of Health Politics, Policy and Law
8. Using the 1993 Robert Wood Johnson Foundation Employer Health Insurance survey, theauthors investigated other possibilities, including presence of union employees, for-profit versusnonprofit status, age distribution of the firm, and industry. The variation in self-insurance rateswith these variables is far less than that by firm size, and for some of these variables, for instance,industry, rates of part-time work vary widely for reasons exogenous to self-insurance status. Dis-entangling these outside rates from the policy effects would require more information than iscurrently available.
Premium Imputation Effects
An implicit assumption made in the analysis is that the incidence of thehealth insurance benefits is based on an average premium. For self-insuredfirms, the premium imputation method will systematically overestimatethe health insurance costs for healthy and underestimate the costs forunhealthy workers. This misclassification is likely to bias the results awayfrom finding any effect of the clause for the following reasons. In the firstplace, sicker workers are more likely to work part time and to have higherhealth care costs and higher premiums. Since we do not observe healthstatus, we will tend to misclassify some sicker low-wage workers as highwage, because the premium we impute to them will be too low. Con-versely, healthy workers are more likely to be misclassified as low-wageworkers, because the imputed premium will overstate their true premiumcosts. In combination, these two measurement errors will bias our differ-enced rates of part-time work and ineligibility for health insurance to zero.Our results therefore underestimate the effects of the nondiscriminationclause.
Another effect of imputation may be to overestimate premium costs foryoung workers and underestimate premium costs for older workers; this,combined with the higher probability nationally that younger workers arepart time, may cause our results to be an overestimate of the effects of thelaw.9 Restricting our sample to primary earners only limits this bias; as acheck, part-time workers in our sample are, on average, about a yearyounger than full-time workers.10
Henry David Thoreau pointed out in an 11 November 1850 entry to hisjournal that “some circumstantial evidence is very strong, as when youfind a trout in the milk.” In every real, predicted, and hypothetical com-parison of low- and high-wage workers, we find some evidence that thelegislation has effects. Even though the evidence is circumstantial, webelieve that those effects are present. Although we cannot offer a precisepoint estimate of the impact of the law, we have presented some boundsfor the effects. Our lower-bound effect is very likely underestimated,whereas our upper-bound estimates may be either over- or underesti-mated, depending upon which uncontrolled effects dominate. The strict
Wolaver et al. � Mandating Insurance Offers for Low-Wage Workers 907
9. As of 2001, the proportion of employed persons working part time was 26.3 for those agedtwenty to twenty-four, 11.1 for those aged twenty-five to fifty-four, and 24.2 for workers agedfifty-five and above (U.S. Bureau of Labor Statistics 2000).
10. The authors wish to thank an anonymous referee for the difference-in-difference inter-pretation of the results and for calling attention to the implicit assumption of an average, ratherthan individual, incidence of the health insurance benefit in the estimates.
division of low- and high-wage workers according to economic theoryfully justifies our base estimates for the impact of the law, but given theabsence of pre- and postlegislation data, we have performed every otherfeasible comparison to test the robustness of our results. We present theseestimates not as definitive measures of the impact of the law, but simplyto establish the presence of effects, with a range of possible measure-ments. These are small economy-wide impacts, but as the following sec-tion suggests, they may produce large distributional effects. Ironically, thecost of the law is placed on the very workers that the law was intendedto aid by increasing their access to health insurance coverage.
The Effect on Inequality
Economists have been grappling with an explanation for the increase inearnings and income inequality that began in 1979, after several decadesof earnings convergence. These events cannot be attributed to any singlecause, but the growing cost of fringe benefits and the nondiscriminationrules may contribute. Figure 3 shows the growth and composition of theemployment cost index. The inflation in health insurance slowed in the mid-1990s, but in general the costs of nonwage compensation have outpaced thegrowth in wages. Recently, however, the United States has experiencedfaster growth in nonwage compensation costs, as it did in earlier periods. Ashealth insurance costs grow, firms will pay more attention to their offer ofinsurance coverage and to recovering the costs of their firm’s contributions.
Earnings Inequality Effects
If low-wage workers are now increasingly likely to work fewer hoursbecause of the nondiscrimination policy, their labor earnings are furtherdepressed. These low-wage workers compose roughly the bottom 12 per-cent of the wage distribution of primary earners. According to the esti-mates presented here, between 0.6 and 5.6 percent of all low earners(based on the percentage of low-wage workers at firms that offer anyinsurance) experience reduced hours of work because of the nondiscrim-ination clause. Depending on how many fewer hours they work, theannual earnings loss could range between $975 and $3,900, with an aver-age loss of $2,925.11 This loss represents 12.5–50 percent of the average
908 Journal of Health Politics, Policy and Law
11. The average hourly wage of these workers in 1988 dollars was $3.90. If the workerreceives five fewer hours per week for fifty weeks a year, the income loss is $975. If the worker
annual earnings of these workers at full-time, full-year hours, and for themthe cost is substantial, even though the employment effect is a small per-centage of the total workforce.
The mandate regarding health insurance coverage can be compared toanother mandate, the minimum wage, which is also intended to help low-skilled workers. In both cases, one can ask whether the mandate achievedits goal. Critics of the minimum wage have argued that it is not success-ful because (1) many minimum-wage workers are not low income butfrom families with incomes well above the poverty line and (2) the mini-mum wage reduces employment because it requires a rate of pay that maybe in excess of a worker’s contribution to the firm. The latter point isclearly directly relevant to the investigation in this article. The evidenceis mixed, although the current view is that the minimum wage has eithernone or a small negative effect on employment, including hours worked(Card and Krueger 2000, Neumark and Wascher 2000). Since the nondis-
Wolaver et al. � Mandating Insurance Offers for Low-Wage Workers 909
loses twenty hours a week, the income loss is $3,900. The average part-timer in the sample workstwenty-five hours a week, representing a loss of fifteen hours per week translated into real annualearnings of $2,925; for comparison, an average worker at forty hours a week for fifty weeks peryear earns $7,800.
9 --
8 --
7 --
6 --
5 --
4 --
3 --
2 --
1 --
0 --
82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 2000
Year and Quarter
Perc
ent
Wages and Salaries
Benefits
Figure 3 Employment Cost Index for Civilian Workers, 1982–2001,Changes in Wages and Salaries and Benefit Costs, Twelve-MonthPercentage ChangeSource: U.S. Bureau of Labor Statistics 2001.
Year and Quarter
Perc
ent
crimination clause raises the effective minimum wage for full-time work-ers at self-insured firms, our estimates of its small negative employmenteffects are consistent with those of Neumark and Wascher 2000 for theminimum wage.
Additional Equity Effects
A second distributional problem lies in the tax exemption for the valueof the health insurance premiums. Workers with higher wages are likelyto fall into higher tax brackets. The tax treatment represents a sizable taxbreak for those workers who least need the help. Low-wage workers areless likely to receive any tax benefits, since they are less likely to be offeredand to accept coverage from an employer. The taxes of those low-wageworkers who are covered are reduced, but not by as much as a high-wageearner’s taxes on an identical policy.
An additional equity concern is that the nondiscrimination mandatedoes not cost the same across firms or workers in terms of geography,industry, or occupation because health insurance costs exhibit consider-able variation along these dimensions. These differences are compoundedby state variations in the minimum wage and create horizontal inequitiesin the incidence of the law. Workers in some states, industries, and occu-pations will be more likely to be part time due to the increased healthinsurance or minimum wage costs. Workers in poor health will also bemore likely to be subject to these employment effects.
All of these costs should be compared to the gains to those workers whoremain employed and receive the benefits from the increased health insur-ance coverage. In the next section we attempt to provide some insight intothe benefits in increased coverage of this legislation.
Discussion and Conclusion
Our tests of the effects of the nondiscrimination policy on employmentin this article have in every case found them to be negative, even thoughfirms have several other possible responses, such as noncompliance orshifting away from self-insurance. Comparisons of low-wage workers,defined as those whose estimated total productivity is lower than the min-imum wage plus the costs of providing health insurance, to workers whoare able to pay for health insurance through reduced wages always showhigher rates of part-time employment coupled with ineligibility for healthinsurance. Employment effects are greater for low-wage workers in large
910 Journal of Health Politics, Policy and Law
firms in which self-insurance is more common and the policy is morelikely to have an impact. These effects remain when many other charac-teristics of the workers and firms that may affect coverage and part-timework decisions are taken into account. Finally, comparisons of subsets ofhigh- and low-wage workers, who are more similar to each other in theircharacteristics than the broader groups of workers, also show the hypoth-esized effects.
This article illustrates the unforeseen effects which sometimes plaguepolicy interventions. Certainly, the reduction of full-time employmentopportunities for low-wage workers— the group the policy was expectedto help—was an unintended consequence. Although the fraction of theprimary-earner workforce that is part time as a consequence of the lawis low, probably less than half a percentage point, the set of workers at riskconsists of those who will suffer the most from underemployment. Themaximum estimated benefit of the law to low-wage workers is an increasein employer-provided health insurance of thirty-one percentage points.Casting the provision in the best possible light, this result implies that forevery 9.4 low-wage workers who gain insurance coverage, one low-wageworker, on average, loses full-time employment. Translating these effectsinto dollar terms, for every $19,411 in benefits under the law there is anexpected cost in lost earnings of $2,965.12 Using this average, the calcu-lated cost of the law is 15 percent of the benefits among affected work-ers. This calculation does not include any other employment effects notexamined in this article: loss of all employment, shifts to part-year work,and any costs of increased temporary work.
Possible Impact of the Nondiscrimination Clausein the Post–Welfare Reform Era
Welfare-to-work initiatives so prominent in welfare reform policies have,we believe, increased the population of workers vulnerable to the employ-ment effects of the nondiscrimination clause. How has this policy affectedthe probability of private health insurance coverage for former welfarerecipients now entering the workforce?
Wolaver et al. � Mandating Insurance Offers for Low-Wage Workers 911
12. The average earnings loss in dropping from full-time to part-time hours for low-wageworkers in this sample is $2,925; we assume that the average premium of $2,065 for low-wageworkers is the dollar value of the benefit of the law. The range of estimates for all part-time workis 0.8 to 5.6 percent of all low-wage workers affected, with a mean effect of 3.3 percent. There-fore, if the benefit of the law is an increase in coverage of thirty-one percentage points, for every31/3.3, or 9.4, low-wage workers, we expect benefits of 9.4 � $2,065 and a cost of $2,925.
We found that at most 31 percent of low-wage workers in our samplewould have gained coverage as a result of the nondiscrimination clause.However, recent welfare leavers may not be comparable to the sample wehave studied here for several reasons. First, they may have less job expe-rience and more dependent children. Second, health care costs haveincreased in real terms, increasing the proportion of workers who wouldbe in our low-wage group, and this may affect the estimates. The additionof welfare leavers into the low-wage labor market will also represent anincrease in labor supply, further depressing market wages for this groupas a whole, drawing more workers into the pool of those at risk for employ-ment effects and possibly changing the responses of firms.
What do recent data on private coverage in poor households tell usabout these questions? On average, across the United States about 37 per-cent of those with incomes below 200 percent of federal poverty levels arecovered by employer-sponsored insurance, whereas nearly 42 percent arecovered by private insurance, Medicare, or Civilian Health and MedicalProgram for Uniformed Services (CHAMPUS) (Holahan 2002). But pri-vate coverage rates vary among states, from less than 30 percent in NewYork and California to more than 45.2 percent in Wisconsin. Insurancegaps also vary, from about 50 percent to more than 67 percent of house-holds. Underlying this variation, Holahan 2002 suggests, are state-levelfactors such as industrial composition, extent of union membership, sizeof firms, and the human capital of workers. Shen and Zuckerman 2003found that states with low ESI rates have populations with fewer skills andless human capital. Because of these workforce characteristics, thesestates get fewer of the high human capital, high paying jobs that also haveESI and as a result have high uninsurance rates.13 It is plausible that wel-fare leavers will be disproportionately concentrated in the low ESI statesand therefore have more difficulties finding private health insurance cov-erage and perhaps will be more subject to employment effects.
Little is yet known about the long-term employment prospects of wel-fare leavers compared to other low-wage workers. Loprest 2002 finds thatearly welfare leavers (those who left prior to 1997) are more likely thanworking low-income mothers (those with incomes at or below 200 per-cent of the federal poverty line) to have full-time employment (69.4 com-pared to 64.2 percent) but less likely to have employer-sponsored healthinsurance (23.2 compared to 35.8 percent). This comparison is less thanideal because Loprest’s sample includes married women, which would
912 Journal of Health Politics, Policy and Law
13. It appears to us that states with high ESI tend to be states with very low unemployment.
tend to increase the rates of voluntary part-time work, and because earlywelfare leavers are more likely to be successful than women who exit therolls later. Moreover, Loprest’s data refer to the period prior to the reces-sion that began in 2001 and prior to the recent upswing in health insur-ance premium inflation, both of which will impact coverage and employ-ment rates of low-wage workers.
Our reading of this literature is that public insurance is the primarysource of health care coverage of former welfare recipients and especiallytheir children. However, many former welfare recipients who remain offcash assistance and are employed at least part year are covered by neitherpublic insurance nor their employer; Garrett and Hudman 2002 estimatesthat only about 20 percent of early welfare leavers had employer-basedcoverage in 1999. In addition, for many of these individuals, the nondis-crimination mandate may reduce employment opportunities.
Our concern is not primarily that there is an efficiency cost to thenondiscrimination clause, as shown by an increased probability of (invol-untary) part-time work coupled with ineligibility for health insuranceamong workers who are not productive enough to pay for their coveragethrough lowered wages. Rather, we are concerned with equity and the bestpolicy to increase health insurance coverage. If the efficiency losses wereconcentrated among high-wage workers, for instance, the price of thenondiscrimination clause would be more palatable. If we are correct inour estimates, the consequences of the policy we explore are being borneby vulnerable members of the population, including recent welfare leavers.We believe the evidence is convincing enough that alternative policies toincrease health insurance coverage among low-wage workers should becarefully examined.
Several other policy instruments could increase the coverage of low-wageworkers without the concomitant loss of full-time employment opportuni-ties. These include Medicaid and/or state Child Health Insurance Program(SCHIP) expansions, tax credits, and universal coverage, although each hasits own efficiency and equity concerns. Universal health care coveragewould eliminate the problems of adverse selection and provide more equityin coverage, but it is unlikely to become politically viable in the immediatefuture. Incremental reforms such as expansions of the public insurance sys-tem or tax credits are more viable policy options that should not have thenegative employment effects associated with the nondiscrimination clause.
Expansions in Medicaid and SCHIP (some states have included fami-lies in coverage as well as children) have an established record of increas-ing coverage among low-income populations. These programs avoid the
Wolaver et al. � Mandating Insurance Offers for Low-Wage Workers 913
adverse selection problems of the individual health insurance market(Pauly and Nichols 2002). Further advantages are modest copayments andexperience in provision of care among the low-income population. Insome states, however, access to care is problematic because providersreceive low reimbursements (Berman et al. 2002). Moreover, crowd-outmay reduce employer-provided coverage, possibly increasing both the rateof uninsurance and government budgets by providing public insurance tosome households that would have had private coverage in the absence ofthe policy (Cutler and Gruber 1996). If one of the goals of governmentintervention is to increase equity as well as health insurance coverage, thenthe only form of crowd-out that should be of concern is an increase in theuninsured rate.
Tax credits are perhaps the most politically attractive option in the current legislative environment. Credits might increase health care cov-erage without the negative employment outcomes of the nondiscriminationclause if the credits are refundable and generous enough to make a differ-ence.14 A study by Families USA (2001) finds that the tax credits proposedby the Bush administration, in which low-income uninsured individualswould receive tax credits of up to $1,000 and low-income uninsured fam-ilies would receive tax credits of up to $2,000, would have little impact onthe coverage of the low-income population. According to the report, inmany states, $1,000 plans were not available and those that were “gener-ally provided incomplete coverage, had high deductibles and required highcoinsurance or copayments.” A Kaiser Family Foundation study (2001)found that individuals with even relatively minor health problems had dif-ficulty in obtaining affordable coverage in the individual insurance market;any policy should take into account this factor if it is to be successful atincreasing coverage. Tax credits targeted to low-income families wouldmake the tax treatment of health insurance more equitable. But since pre-miums decrease with the size of the group, low-income workers withoutaccess to employer-provided plans would still be at a great disadvantagerelative to workers with generous fringe-benefit plans. The prospect ofcrowd-out is not eliminated with the introduction of tax credits for indi-vidual purchase of health insurance. Nevertheless, compared to the non-discrimination clause, such credits should not have negative employmentconsequences and, if made more generous, could lead to an increase in cov-erage of this low-wage population.
914 Journal of Health Politics, Policy and Law
14. In most plans, the proposed credits are roughly half the current annual premium for fam-ily coverage (Feder et al. 2001), which is still well below that provided on average in employer-provided plans.
Now that health insurance premium inflation is once again exceedingwage growth, as in the period from 1988 to 1994, the employment effectsof the nondiscrimination clause may grow. In combination with a weakeconomy, the effects of the nondiscrimination clause on employment andinsurance coverage of low-wage workers are likely to intensify. First, it isclear that low-wage workers are less likely to be employed by firms thatoffer health insurance. Even for those workers who have potential accessto employer-sponsored coverage, the nondiscrimination clause can onlyhave an impact if the plan is self-insured. The policy does nothing to makethose plans more affordable for the low-wage worker and creates anincentive for firms to limit hours for workers it wishes to exclude fromcoverage. Moreover, this article provides some insight into current trendsin the rates of the uninsured in the context of the recent welfare reformsand the recent economic downturn. It is not surprising that only a smallpercentage of former recipients of Medicaid, whether exiting the publicsafety net because they lost eligibility or not, has been able to find afford-able employer-based insurance coverage.
Wolaver et al. � Mandating Insurance Offers for Low-Wage Workers 915
Appendix A: Details of the Nondiscrimination Clause
The clause only applies to employer-provided plans where the firm itselftakes on the risk of workers’ medical expenses, rather than purchasingpolicies from insurance companies. These firms may contract with an insurerfor administrative tasks such as claims processing. Self-insurance affordsfirms several advantages. These plans are exempt from state mandates andstate premium taxes that add to the cost of the insurance, ceteris paribus(Acs et al. 1996). In 1991, approximately 40 percent of employees who hademployer-provided insurance were in a self-insured plan (Sullivan et al.1992). The history of the nondiscrimination clause exemption for non-self-insured plans is somewhat convoluted. The Internal Revenue Code of 1986extended the nondiscrimination rules to all employer-provided healthinsurance; the provisions were scheduled to go into effect after 31 Decem-ber 1988, but public law 101-104 repealed this extension retroactively.
The law stipulates that the value of any differences in coverage betweenhighly compensated and other workers becomes taxable. In discussing therepeal of section 89 of the 1986 Internal Revenue Code, the House Con-gressional Record states that the justification for the revenues lost dueto the tax expenditure for employer-provided health insurance is throughincreased rates of insurance. On the other hand, “the Congress believedthat the cost to the Federal Government of tax-favored employer-providedaccident and health coverage is not justified if such coverage dispropor-tionately benefits highly compensated employees. In order to achieve thisobjective, nondiscrimination rules were enacted to permit the full exclu-sion from income of employer-provided health benefits only if the bene-fits are provided to required numbers of nonhighly compensated employ-ees” (Congressional Record, 101st Cong., 1st sess., vol. 135, pt. 155, 7November 1989). There are some exceptions that allow employers toexclude some full-time employees; for example, workers during a proba-tionary period and workers under the age of 25 may be exempted fromoffers. Plans negotiated by unions are also exempt from nondiscrimina-tion rules.
The following is the relevant passage from the 1978 legislation requiringthat employers offer insurance to all employees (reprinted from U.S. TaxCode, Title 26, Subtitle A, Chapter 1, Subchapter B, Part III, Section 105: (h);accessed on-line at www.fourmilab.ch/ustax/www/t26-A-1-B-III-105.html).
916 Journal of Health Politics, Policy and Law
(h) Amount paid to highly compensated individuals under a discrimina-tory self-insured medical expense reimbursement plan(1) In general
In the case of amounts paid to a highly compensated individualunder a self-insured medical reimbursement plan which doesnot satisfy the requirements of paragraph (2) for a plan year,subsection (b) shall not apply to such amounts to the extent theyconstitute an excess reimbursement of such highly compensatedindividual.
(2) Prohibition of discriminationA self-insured medical reimbursement plan satisfies the require-ments of this paragraph only if(A) the plan does not discriminate in favor of highly com-
pensated individuals as to eligibility to participate; and(B) the benefits provided under the plan do not discriminate
in favor of participants who are highly compensated indi-viduals.
(3) Nondiscriminatory eligibility classifications(A) In general
A self-insured medical reimbursement plan does not sat-isfy the requirements of subparagraph (A) of paragraph (2)unless such plan benefits(i) 70 percent or more of all employees, or 80 percent
or more of all the employees who are eligible tobenefit under the plan if 70 percent or more of allemployees are eligible to benefit under the plan; or
(ii) such employees as qualify under a classification setup by the employer and found by the Secretary notto be discriminatory in favor of highly compensatedindividuals.
(B) Exclusion of certain employeesFor purposes of subparagraph (A), there may be excludedfrom consideration(i) employees who have not completed 3 years of ser-
vice;(ii) employees who have not attained age 25;(iii) part-time or seasonal employees;(iv) employees not included in the plan who are included
in a unit of employees covered by an agreement
Wolaver et al. � Mandating Insurance Offers for Low-Wage Workers 917
between employee representatives and one or moreemployers which the Secretary finds to be a collec-tive bargaining agreement, if accident and healthbenefits were the subject of good faith bargainingbetween such employee representatives and suchemployer or employers; and
(v) employees who are nonresident aliens and whoreceive no earned income (within the meaning ofsection 911(d)(2)) from the employer which con-stitutes income from sources within the UnitedStates (within the meaning of section 861(a)(3)).
(4) Nondiscriminatory benefitsA self-insured medical reimbursement plan does not meet therequirements of subparagraph (B) of paragraph (2) unless allbenefits provided for participants who are highly compensatedindividuals are provided for all other participants.
(5) Highly compensated individual definedFor purposes of this subsection, the term “highly compensatedindividual” means an individual who is(A) one of the 5 highest paid officers,(B) a shareholder who owns (with the application of section
318) more than 10 percent in value of the stock of theemployer, or
(C) among the highest paid 25 percent of all employees (otherthan employees described in paragraph (3)(B) who arenot participants).
(6) Self-insured medical reimbursement planThe term “self-insured medical reimbursement plan” means aplan of an employer to reimburse employees for expensesreferred to in subsection (b) for which reimbursement is notprovided under a policy of accident and health insurance.
918 Journal of Health Politics, Policy and Law
Appendix B: Econometric Methods
Logged wages are modeled as a linear function of a vector of character-istics, Xw and the categories of the insurance/hours variable. The omit-ted category is Full-time with an offer of insurance. The probability thata worker falls into one of the four insurance categories is given in equa-tion 1 (Greene 1993):
[1] Pr(Category j) = exp(�j + �j,wagesLog(wages) + �j,kXj,k)
j=0to3 exp(�j +�j,wagesLog(wages) + �j,kXj,k)+�i,k
As described in the text, economic theory suggests that firms decidesimultaneously on the size of the total compensation package, the mix ofwages and benefits in this package, and whether to hire workers part timeor full time. Thus we model the wage equation 2 in the following form,
[2]Log(wages) = �0 + �pt,no offerPart time,no offer + �pt,offerPart time,offer + �pt,no offerFull time,offer + Aw'Xw +�w
where X is a vector of worker/job characteristics. We employ GeneralizedMethods of Moments (GMM). There are other alternatives to the GMMtechnique, such as full-information maximum likelihood, but we choseGMM because it is computationally easier to estimate and in cases whereerrors in specification are more worrisome, yields consistent estimateswhere a full-information, maximum likelihood model will not (Lechnerand Breitung 1996).
Appendix Table B1 shows the results from the regression used toimpute the health insurance premiums for individuals in the sample.Appendix Table B2 shows the raw coefficient results from the model,which are used to construct the predicted probabilities presented in thetext.
Wolaver et al. � Mandating Insurance Offers for Low-Wage Workers 919
920 Journal of Health Politics, Policy and Law
Appendix Table B1 Premium Imputation, Exponential RegressionResults
Coefficient Standard Error
Constant 7.24 .126 *Hazardsa .722 .344 *PUSa -.417 .402Mediuma -.163 .118Heavya -.387 .216 **Presence of familyb .415 .051 *State and local government .033 .059Industry 1 -.132 .312Industry 2 .082 .220Industry 3 .007 .138Industry 4 .165 .086Industry 5 .078 .092Industry 6 -.020 .091Industry 7 -.038 .100Industry 8 .065 .124Industry 9 -.093 .190Industry 10 .236 .230Industry 11 -.055 .076Union .195 .048 *Firm 1–25 employees -.0625 .125Firm 26–100 employees -.103 .070 ***West Central Region -.003 .113North Central Region .089 .097Southwest SC Region -.126 .107South East SC Region -.130 .112South Atlantic Region -.133 .095Pacific Region .078 .101Mountain Region .033 .12Mid-Atlantic Region .025 .098N = 3,045 Log likelihood = -3514.37 Model χ2(28) = 154.95, Pr > χ2 = 0
Source: Authors’ calculations from 1987 National Medical Expenditure Survey (U.S. Depart-ment of Health and Human Services 1987).
aDictionary of Occupational Titles data on average scores by occupation.bMore than one person in the household.*Statistically significant at the .05 level.**Statistically significant at the .10 level.***Statistically significant at the .15 level.
Ap
pen
dix
Tab
le B
2Se
lect
ed R
esu
lts
fro
m S
imu
ltan
eou
s R
egre
ssio
n M
od
el In
div
idu
al O
ffer
Invo
lunt
ary
Part
-tim
e,
Invo
lunt
ary
Part
-tim
e Fu
ll-tim
e,
No
Off
erw
/ Off
erN
o O
ffer
Log
Wag
e
Var
iabl
eβ
Std.
Err
orβ
Std.
Err
orβ
Std.
Err
orβ
Std.
Err
or
Hig
h-W
age
Wor
ker
Set
Con
stan
t-1
0.25
4.65
*-1
7.12
8.57
*-7
.89
2.28
*1.
640.
51*
Age
-15.
9715
.19
-35.
4218
.75
**-2
1.32
7.02
*1.
922.
63ag
e219
.14
17.1
540
.46
21.1
6**
25.3
97.
75*
-1.9
02.
81U
nion
-1.9
50.
71*
0.69
0.69
-1.4
50.
30*
-0.0
20.
23y1
992
-0.3
30.
580.
320.
45-0
.15
0.23
-0.0
40.
07H
sdeg
-0.3
10.
67-2
.10
1.09
**-0
.67
0.32
*0.
180.
14C
olld
eg-0
.88
1.29
-4.5
52.
40**
-1.2
40.
64**
0.33
0.13
*fr
m1k
1t0.
210.
390.
290.
510.
120.
160.
070.
06fr
m10
00p
0.29
0.27
-0.3
30.
320.
010.
120.
160.
11In
dfin
0.42
0.70
-0.8
90.
75-0
.72
0.20
*0.
040.
20In
dman
-0.5
20.
661.
030.
90-0
.38
0.23
-0.0
70.
06In
dtra
de1.
250.
761.
380.
900.
500.
30**
-0.0
30.
19In
dallo
t0.
330.
580.
920.
790.
260.
22-0
.11
0.06
**te
nlt1
yr2.
070.
76*
2.03
0.93
*2.
620.
38*
0.10
0.59
ten1
t40.
760.
521.
430.
64*
0.69
0.22
*-0
.04
0.09
tenu
re5t
93.
081.
31*
0.77
0.48
1.92
0.29
*-0
.03
0.06
wes
t0.
130.
09m
idw
est
0.04
0.18
neas
t0.
070.
17pr
iskf
am15
.47
9.30
**1.
228.
7711
.66
4.02
*sp
ouse
own
1.20
0.48
*1.
210.
71**
0.74
0.25
*
Ap
pen
dix
Tab
le B
2Se
lect
ed R
esu
lts
fro
m S
imu
ltan
eou
s R
egre
ssio
n M
od
el In
div
idu
al O
ffer
(C
on
tin
ued
)
Invo
lunt
ary
Part
-tim
e,
Invo
lunt
ary
Part
-tim
e Fu
ll-tim
e,
No
Off
erw
/ Off
erN
o O
ffer
Log
Wag
e
Var
iabl
eβ
Std.
Err
orβ
Std.
Err
orβ
Std.
Err
orβ
Std.
Err
or
log
wag
e1.
043.
147.
814.
25**
2.28
1.43
Hhs
ize
-0.1
60.
170.
250.
20-0
.10
0.07
kidl
t60.
250.
44-0
.15
0.52
-0.3
50.
20**
fkid
lt6-0
.73
0.71
1.61
1.17
0.01
0.38
wpt
niof
f-3
0.39
59.5
4w
ptio
ff1.
3613
.51
wft
niof
f1.
116.
48N
= 1
9,32
4O
bjec
tive*
N=
5.5
8
Low
-Wag
e W
orke
r Se
t
cons
tant
4.42
33.7
13.
1927
.11
5.22
15.1
21.
881.
41ag
e-2
0.77
29.3
43.
0214
.62
-11.
9814
.64
-2.6
55.
16ag
e224
.09
30.9
3-4
.34
17.6
712
.35
16.6
42.
855.
72un
ion
-1.3
61.
761.
500.
93-0
.84
1.23
-0.0
50.
28y1
992
-0.2
10.
840.
691.
18-0
.35
0.66
-0.0
20.
17hs
deg
-0.7
71.
21-0
.09
0.87
-0.0
80.
500.
030.
10co
lldeg
-1.2
53.
56-1
.12
3.55
-0.7
61.
52-0
.09
0.20
frm
1k1t
0.51
0.94
0.65
0.39
-0.6
20.
74-0
.14
0.27
frm
1000
p0.
640.
650.
060.
750.
170.
340.
020.
09in
dfin
-0.5
22.
82-0
.15
4.59
0.51
1.81
0.01
0.25
indm
an-3
.16
2.29
-0.2
20.
960.
010.
660.
070.
23in
dtra
de-0
.31
0.61
0.27
1.35
0.59
0.60
0.09
0.24
Ap
pen
dix
Tab
le B
2(C
on
tin
ued
)
Invo
lunt
ary
Part
-tim
e,
Invo
lunt
ary
Part
-tim
e Fu
ll-tim
e,
No
Off
erw
/ Off
erN
o O
ffer
Log
Wag
e
Var
iabl
eβ
Std.
Err
orβ
Std.
Err
orβ
Std.
Err
orβ
Std.
Err
or
Low
-Wag
e W
orke
r Se
t
inda
llot
-0.8
10.
680.
202.
650.
671.
020.
120.
20te
nlt1
yr1.
543.
940.
823.
472.
352.
980.
490.
96te
n1t4
0.25
1.93
-0.2
01.
690.
481.
380.
050.
07te
nure
5t9
1.88
4.66
-0.0
20.
721.
973.
610.
140.
33w
est
0.07
0.13
mid
wes
t0.
040.
10ne
ast
0.01
0.10
pris
kfam
8.72
55.3
7-1
0.69
56.8
09.
3325
.48
spin
s1.
081.
35-0
.20
0.99
0.10
0.68
log
wag
es-4
.15
31.9
0-3
.96
27.2
2-6
.84
16.0
8hh
size
-0.0
70.
390.
190.
32-0
.09
0.21
kidl
t60.
251.
190.
991.
250.
300.
78fk
idlt6
-0.3
82.
98-1
.29
3.84
-0.3
91.
29w
ptni
off
-0.2
22.
46w
ptio
ff-0
.07
2.97
wft
niof
f-2
.33
4.65
N=
1,5
74O
bjec
tive*
N=
2.2
9
Not
e: O
ther
var
iabl
es in
clud
ed b
ut n
ot s
how
n ar
e: r
egio
n, y
ear,
gend
er, c
lass
of
wor
ker,
race
, urb
anic
ity, o
ccup
atio
n an
d m
arita
l sta
tus.
Sour
ce: A
utho
rs’
calc
ulat
ions
fro
m 1
988
and
1993
Em
ploy
ee B
enefi
ts S
uppl
emen
ts, C
urre
nt P
opul
atio
n Su
rvey
s (U
.S. C
ensu
s B
urea
u an
d U
.S. B
urea
u of
Lab
or S
tatis
tics
1988
, 199
3).
*Sta
tistic
ally
sig
nific
ant a
t the
.05
leve
l.**
Stat
istic
ally
sig
nific
ant a
t the
.10
leve
l.
References
Acs, G., S. H. Long, S. Marquis, and P. Short. 1996. Self-Insured Employer HealthPlans: Prevalence, Profile, Provisions, and Premiums. Health Affairs 15(2):266–278.
Berman, S., J. Dolins, S. Tang, and B. Yudkowsky. 2002. Factors That Influence theWillingness of Private Primary Care Pediatricians to Accept More Medicaid Patients.Pediatrics 110(2):239–248.
Blumberg, L. J., and L. Nichols. 2001. Why Are So Many Americans Uninsured? AConceptual Framework, Summary of the Evidence, and Delineation of the Gapsin Our Knowledge. Paper prepared for a conference on the uninsured, Ann Arbor,MI, July.
Brittain, J. A. 1971. The Incidence of Social Security Payroll Taxes. American Eco-nomic Review 61(1):110–125.
Card, D., and A. B. Krueger. 2000. Minimum Wages and Employment: A Case Studyof the Fast-Food Industry in New Jersey and Pennsylvania: Reply. American Eco-nomic Review 90(5):1397–1420.
Cutler, D., and J. Gruber. 1996. Does Public Insurance Crowd Out Private Insurance?Quarterly Journal of Economics 111(2):391–430.
Ellwood, M. 1999. The Medicaid Eligibility Maze: Coverage Expands but EnrollmentProblems Persist: Findings from a Five State Study. Washington, DC: Kaiser Com-mission on Medicaid and the Uninsured.
Employee Benefits Research Institute (EBRI). 1999. Contingent Workers and Workersin Alternative Work Arrangements. Executive Summary. EBRI Issue Brief No. 207.Washington, DC: EBRI. Available on-line at www.ebri.org/ibex/ib207.htm.
Families USA. 2001. A 10-Foot Rope for a 40-Foot Hole: Tax Credits for the Unin-sured. Congress Daily, 13 September.
Farber, H. S., and H. Levy. 2000. Recent Trends in Employer-Sponsored Health Insur-ance Coverage: Are Bad Jobs Getting Worse? Journal of Health Economics 19:93–119.
Feder, J., L. Levitt, E. O’Brien, and D. Rowland. 2001. Covering the Low-Income Unin-sured: Assessing the Alternatives. Washington, DC: Kaiser Commission on Medicaidand the Uninsured. Available on-line at www.kff.org/content/2002/2258/2258.pdf.
Gabel, J., P. Ginsburg, and K. Hunt. 1997. Small Employers and Their Health Bene-fits, 1988–1996: An Awkward Adolescence. Health Affairs 16(5):103–110.
Garrett, B., and J. Hudman. 2002. Women Who Left Welfare: Health Care Coverage,Access and Use of Health Services. Washington, DC: Kaiser Commission on Medic-aid and the Uninsured. Available on-line at www.kff.org/content/2002/4050/4050.pdf(accessed 25 July 2002).
Garrett B., L. M. Nichols, and E. Greenman. 2001. Workers without Health Insurance:Who Are They and How Can Policy Reach Them? Urban Institute Report to theKellogg Foundation.
Greene, W. H. 1993. Econometric Analysis. New York: Macmillan.Gruber, J. 1999. Health Insurance and the Labor Market. In Handbook of Health Eco-
nomics, vol. 5, ed. A. J. Culyer and J. P. Newhouse. Amsterdam: Elsevier Science.
924 Journal of Health Politics, Policy and Law
Gruber, J., and A. Krueger. 1991. The Incidence of Mandated Employer-ProvidedInsurance: Lessons from Workers’ Compensation Insurance. In Tax Policy and theEconomy, vol. 5, ed. D. Bradford. Cambridge: NBER and MIT Press.
Hoffman, C., and M. B. Pohl. 2002. Health Insurance Coverage in America: 2000Data Update (February). Washington, DC: Kaiser Commission on Medicaid andthe Uninsured. Available on-line at www.kff.org/content/2002/4007/4007.pdf.
Holahan, J. 2002. Variations among States in Health Insurance Coverage and MedicalExpenditures: How Much Is Too Much? Washington, DC: Urban Institute.
Houseman, S. 2001. Why Employers Use Flexible Staffing Arrangements: Evidencefrom an Employer Survey. Industrial and Labor Relations Review 55(1):149–70.
Kaiser Family Foundation. 2001. The Uninsured: A Primer. Uninsured Basics (March).Washington, D.C. Kaiser Commission on Medicaid and the Uninsured. Accessed viathe Kaiser Family Foundation Web site at www.kff.or/content/2001/2228/2228/pdf.
———. 2003. The Uninsured and Their Access to Health Care. Fact sheet (January).Washington, DC. Kaiser Commission on Medicaid and the Uninsured. Accessedvia the Kaiser Family Foundation Web site at www.kff.org/content/2003/142004/142004.pdf.
Lechner, M., and J. Breitung. 1996. Some GMM Estimation Methods and Specifica-tion Tests for Nonlinear Models. In The Econometrics of Panel Data: A Handbookof the Theory with Applications: Advanced Studies in Theoretical and Applied Econo-metrics, vol. 33, ed. L. Matyas and P. Sevestre. Dordrecht: Kluwer.
Loprest, P. 2002. Families Who Left Welfare: Who Are They and How Are They Doing?Urban Institute Discussion Paper 99–02. Washington, DC: Urban Institute.
Mills, R. J. 2001. Health Insurance Coverage: 2000. In Current Population Reports,September, 60–215. Accessed via the Bureau of the Census Web site at www.cen-sus.gov/prod/2001pubs/p60–215.pdf.
Mishel, L. R., J. Bernstein, and J. Schmitt. 1997. The State of Working America:1996–1997. Armonk, NY: M. E. Sharpe.
Neumark D., and W. Wascher. 2000. Minimum Wages and Employment: A Case Studyof the Fast-Food Industry in New Jersey and Pennsylvania: Comment. AmericanEconomic Review 90:1362–1396.
O’Brien, E., and J. Feder. 1999. Employment-Based Health Insurance Coverage andIts Decline: The Growing Plight of Low-Wage Workers. Washington, DC: KaiserCommission on Medicaid and the Uninsured.
Pauly, M. V., and L. Nichols. 2002. The Nongroup Health Insurance Market: Short onFacts, Long on Opinions and Policy Disputes. Health Affairs Web Exclusive (23October) at www.healthaffairs.org/WebExclusives/Nongroup_Web_Excl_102302.htm.
Pear, R. 1999. Poor Workers Lose Medicaid Coverage Despite Eligibility. New YorkTimes, 12 April, A1.
Rostenkowski, D. 1989. Announces Introduction of, and a Public Hearing on, H.R.1864, a Bill to Simplify the Nondiscrimination Rules Applicable to Employee Ben-efit Plans under Section 89 of the Internal Revenue Code. Committee on Ways andMeans Press Release, No. 814, 14 April.
Shen, Y.-S., and S. Zuckerman. 2003. Why Is There State Variation in Employer-Sponsored Insurance? Health Affairs 22(1):241–250.
Wolaver et al. � Mandating Insurance Offers for Low-Wage Workers 925
Sullivan, C., M. Miller, R. Feldman, B. Dowd, and J. Gabel. 1992. Employer-sponsored Health Insurance in 1991, Witness to a Thousand Stories: A Look atInsurance Data. Health Affairs 11(4):172–191.
Summers, L. 1989. Some Simple Economics of Mandated Benefits. American Eco-nomic Review 79(May):177–183.
U.S. Bureau of the Census and U.S. Bureau of Labor Statistics. 1988. Employee Ben-efits Supplements, Current Population Survey. Washington, DC: U.S. Bureau ofLabor Statistics.
———.1993. Employee Benefits Supplements, Current Population Survey. Washing-ton, DC: U.S. Bureau of Labor Statistics.
U.S. Bureau of Labor Statistics. 2000. Health Insurance Coverage: 1999. ReportP60–211. Washington, DC: U.S. Department of Commerce, Economic StatisticsAdministration.
———. 2001. Employment Cost Index. Washington, DC: U. S. Bureau of Labor Sta-tistics.
U.S. Department of Health and Human Services. 1987. National Medical ExpendituresSurvey: Household Survey File. Washington, DC: Agency for Health Care Policyand Research.
926 Journal of Health Politics, Policy and Law
Copyright of Journal of Health Politics, Policy & Law is the property of Duke University Press and its content
may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express
written permission. However, users may print, download, or email articles for individual use.
Recommended