Who Pays for Employer-Provided Coverage?

 VI. Key Issues: Financing and Delivery >> C.  Health Financing >> Private Health Revenues >> Health Insurance Premiums >> Who Pays for Employer-Provided Coverage? (last updated 2.28.17)

Overview

The conventional wisdom is that health insurance employer-paid premiums are ultimately fully borne by workers in the form of lower wages or fringe benefits that they otherwise would receive. In short, there is assumed a compensating wage differential for any health benefits offered. (Assumptions underpinning this dynamic are also referred to as hedonic theory, the consistency hypothesis, and the equalizing differences theory.) Thus, for purposes of estimating the impact of proposed federal legislation mandating health benefits, “the CBO makes the simplifying assumption that total compensation is fixed and that changes in the costs of health insurance translate immediately into offsetting changes in wages and other forms of compensation” (even though the actual period of adjustment may take several years); the Joint Committee on Taxation staff makes the same assumption when estimating the effects of proposals on revenue collections (p. 5).
Given the difficulties in estimating hedonic models, research in this area has spilled over into other areas of the labor market, such as mandated benefits and other employer-sponsored fringe benefits.
Several studies show that the wage offset is individualized, i.e., based on each worker’s characteristics rather than an average adjustment across all workers:

• Income Differentials. Wage effects vary directly with income level with the share of health benefits offset being highest for high wage workers.
• Workers with Family Coverage. Married workers with family coverage experience a higher wage offset than single workers.  Workers with family coverage attach about twice as much value per dollar of health coverage as single workers.
• Older Workers. Older workers pay more for their health benefits through lower wages than do younger workers.
• Working Wives. Employed wives who accept health benefits are paid a lower wage than employed wives who do not.
• Maternity Benefits. The wages of women of child-bearing age and their spouses roughly offset the average costs of providing those benefits.
• Obese Workers. Obese workers with employer-sponsored health insurance pay for their higher expected medical expenditures through lower cash wages.

Theory

• Rosen, Sherwin (1987). The Theory of Equalizing Differences. In Handbook of Labor Economics, Edited by O. Ashenfelter & R. Layard. Vol 1:641-692. Amstersdam: North Holland. Argues that in a competitive labor market (where total compensation received by a worker equals that employee’s marginal revenue product), employees fully pay for the health insurance provided by their employer through lower wages or a reduction in other benefits compared to what they could receive from an employer not offering health insurance. Thus there is a wage-benefits tradeoff, also known as a compensating wage differential.
• Summers, Lawrence (1989). Some Simple Economics of Mandated Benefits. The American Economic Review, 79(2: Papers and Proceedings of the Hundred and First Annual Meeting of the American Economic Association), 177-183.
• Pauly, Mark V. (1999) Health Benefits at Work: An Economic and Political Analysis of Employment-Based Health Insurance. University of Michigan Press. Pauly suggests that confusion about how employer-sponsored health insurance is financed has complicated the debate on public policy and health reform.
• O’Brien, Ellen. 2003. Employers’ benefits from workers’ health insurance. The Milbank Quarterly 81 (1): 5-43. This article does not measure the trade-off. However, it provides an extensive survey of literature demonstrating some of the potential benefits to employers of providing health benefits (e.g., productivity gains). If the value to employers of providing health benefits matched their cost, there would be no need to recover the cost of such benefits from employees.

Why The Wage-Benefits Trade-off Is Difficult to Empirically Measure 

As shown below, the literature is inconsistent about whether a wage-benefits trade-off exists or if so, how large it is. A variety of explanations have been offered to explain this:

• Brown C. (1980). Equalizing differences in the labor market. Quarterly Journal of Economics 1980 (February); 113-134. This article showed that the inconclusive empirical evidence of a trade-off is not due only to unmeasured dimensions of worker quality. Even when a statistical analysis carefully accounts for worker quality, the estimated coefficient is still often either positive or insignificant. He offers 5 explanations for why empirical studies may fail to detect a wage-benefits tradeoff.  (1) “Labor markets are simply not as competitive as the theory of equalizing differences assumes.” (2) “The marginal worker’s tastes may be different from those assumed in the a priori signing of the coefficients.” (3) “The job characteristics are not well measured.” (4) “Omitted variables – both individual characteristics that change over time and job characteristics – may be biasing the results.” (5) “Testing the hypothesis on a sample [of workers] in their early and mid-twenties is inappropriate.”
• Pauly, M. V., 2001. Making Sense of a Complex System, Empirical Studies of Employment-Based Health Insurance. International Journal of Health Care Finance and Economics 1, 333—339.
• Han, Seungjin (2009). The Myth of Equalizing Differences in the Labor Market. Unpublished manuscript, McMaster University (January 2009, Preliminary and Incomplete).  “Various reasons have been posited for the empirical difficulty in estimating equalizing differences in the labor market. Among many, they include unobserved heterogeneity (Hwang, Reed, and Habbard, 1992), omitted variables (Brown, 1980; Lucas, 1977), measurement error (Duncan and Holmlund, 1983), the labor market frictions (Boyd, Lankford, Loeb, and Wyckoff, 2003; Hwang, Mortensen, and Reed, 1998) and unappealing linear approximations (Ekeland, Heckman, and Nesheim, 2004).
• Devaro, Jed and Nan L. Maxwell (2014). The Elusive Wage-Benefits Trade-Off: The Case of Employer-Provided Health Insurance. International Journal of Industrial Organization 37(1) · October 2014. “Previous empirical research has failed to consistently identify a wage-insurance trade-off, even in the presence of detailed controls. We have proposed two new reasons for this failure. First, health insurance pricing depends on establishment size and firm size, which are omitted in the standard specifications and which must be included (interactively). Second, because benefits are typically determined at the firm level, establishments in multi-establishment firms are hindered in their ability to tailor their compensation offerings to appeal to the workers in their local labor markets.”

Literature Syntheses

• Currie, Janet and B.C. Madrian (1999).  Health, Health Insurance, and the Labor Market. In: Ashenfelter, O, Card, D (Eds), Handbook of labor economics, vol. 3C. Elsevier Science: Amsterdam. 1999. p. 3309-3416. Based on a review of 10 studies, the authors state “The empirical validity [of the trade-off] has been difficult to establish. The typical estimates [of β1] are either wrong-signed, insignificant, or both.”  They conclude that many of these studies did not have the appropriate data to estimate the magnitude of the compensating differential. Three studies reported a statistically significant positive  relationship between wages and health insurance costs (rather than an offset), two studies had mixed results and one found no statistically significant relationship.  However, the authors also explain in detail how data limitations–particularly the inability to account for differences across workers in productivity or the value that individuals attach to insurance–can explain why traditional methods are unable to detect the trade-off. Four studies found evidence of a trade-off, one showing that 83% of health insurance costs are offset by lower wages, and another showing 50-200% of health insurance costs were offset.
• Gruber, Jonathan (2000). “Health Insurance and the Labor Market,” in A.J. Culyer and J.P. Newhouse, eds., Handbook of Health Economics, vol. 1 (Amsterdam: North Holland, 2000), pp. 645–706. This provides an extensive summary of the evidence showing that employees as a group ultimately bear the costs of any payments an employer makes for health insurance. However, the author also finds that the estimated trade-off frequently has the wrong sign, with higher wages associated with higher fringe benefits.
• West, David L. (May, 2006).  Fair Share for Health Care: No Significant Employment Impact. Center for a Changing Workforce. To assess the employment impacts of the Fair Share for Health Care Act (FSHC), the Center for a Changing Workforce reviewed recent economic literature on employment and health insurance, and used several methods to develop a balanced estimate of possible positive and negative employment impacts.
• Madrian B.C. (2006). The US Health Care System and Labor Markets. Cambridge, MA: National Bureau of Economic Research; January 2006. NBER Working Paper No. 11980. This paper provides a broad and general overview of the relationship between the U.S. health care system and the labor market. The section on the wage-benefit trade-off only cites 1 broad literature review from 1999, arguing that the fundamental problem with early studies was a lack of appropriate data for estimating the magnitude of any such relationship. In contrast, 4 recent studies that took advantage of exogenous changes in health benefits costs found evidence of a wage-health insurance tradeoff.
• Kliff, Sarah (2015). Will cheaper health insurance really raise wages? The evidence is thin. Vox (11.2.15). “A close review of the evidence reveals something discomfiting: Economists have convincing arguments about why, in theory, the Cadillac tax should raise wages. But when it comes to actual data — real-life examples that show the wage-premium trade-off is happening — evidence is sparse” (the author claims to have read more than a dozen studies–though only 8 are cited–and interviewed a half-dozen economists). “Economists have strongly held views on this because there isn’t a good theory about why the trade-off wouldn’t happen,” says Darren Lubotsky, an economist at the University of Illinois Chicago whose research focuses on this issue. ‘But the hard part has been demonstrating the trade-off empirically. And it’s proven hard to get consistent estimates of something close to a dollar-for-dollar trade-off.'”

Studies Showing Negative Relationship Between Health Benefits and Wages

Studies are listed chronologically.

• Eberts R, Stone J. (1985). Wages, fringe benefits, and working conditions: an analysis of compensating differentials. South Econ J. 1985;52:274-280. In this paper we argue that empirical tests of compensating differentials should take into account not only attributes of individual workers but also attributes of their employers. Even when considering labor markets that are competitive in the long run, the analysis should include information about employer attributes because labor markets do not adjust instantaneously. When examining collective bargaining and other noncompetitive markets, it is even more important to have information about employers. The consequences of ignoring firm-specific information are potentially serious. Omitting variables related to profitability and labor bargaining strength will tend to attenuate coefficients used to test for compensating differentials if these variables are positively related to other pecuniary benefits and favorable working conditions. When evaluated at the mean, an additional dollar of health benefits is associated with about an 83 cent reduction in salary.
• Gruber, Jonathan and Krueger, Alan B. (1991). “The Incidence of Mandated Employer-Provided Insurance: Lessons from Workers’ Compensation Insurance,” in David Bradford, ed., Tax policy and the economy. Cambridge, MA: MIT Press, 1991, pp. 111-44.  This study notes that many empirical studies fail to confirm Rosen’s theoretical claim, but explains the variety of factors that may explain this lack of empirical confirmation; for example, many studies cannot account for differences in worker productivity, so cross-sectional comparisons show that higher-paid (more productive) workers are more likely to have health benefits–i.e., no “offset” is observed. The empirical analysis in this paper shows that the increased costs of mandated workers compensation are largely shifted to wages with little adverse effect on employment. However, the shifting is incomplete due to workers at the minimum wage level whose wages cannot be adjusted downward in response to higher fringe benefits. Although extremely imprecise, the estimates suggest that every one percentage point increase in workers’ compensation rates is associated with an employment decline of .11%.
• Kaestner, Robert, and Anne Carroll. 1997. New estimates of the labor market effects of workers’ compensation insurance. Southern Economic Journal 63 (3): 635-51. An examination of the labor market consequences of workers’ compensation benefits reveals that the exclusion of the medical component of benefits may result in a serious bias of the estimates of the wage-benefit tradeoff. Using only the income component of benefits, estimates of the tradeoff are 100% larger than those that use both the income and medical components… a 1 percent increase in the ratio of workers’ compensation benefits to wages will reduce wages by .78 percent.
• Pauly, Mark and Brad Herring (1999). Pooling Health Insurance Risks. AEI Press, 1999. Authors use a difference-in-difference approach to show that for each year of tenure, the increase in wages is higher in firms that do not offer health benefits compared to firms that do; this effect is observed in both large and small firms.
• MaCurdy T, Rapoport D. (2003). Understanding health-insurance coverage of low-skilled worker in the US and in California: Trading wages for insurance. Sphere Institute: Burlingame, CA; 2003.  An overview of our empirical findings indicates that low-skilled workers valued access to employer health insurance at about 74 cents per hour in 1997, which means that in today’s dollars, the implicit price of this access is approximately 83 cents per hour in forgone wages. For a full-time worker, this translates into about $140 in salary per month.
• Royalty A.B. (2003). A Discrete Choice Approach to Estimating Workers’ Marginal Valuation of Fringe Benefits. Indianapolis: Indiana University–Purdue University; June 2003. This paper shows that the marginal valuation of health benefits is well below their actual costs.  Employees with families value health benefits at ~40 cents per dollar of benefits while those with single coverage value them at only 20 cents on the dollar. This implies that even if the wage-benefits trade-off exists, workers generally would be unwilling to accept a full dollar-for-dollar reduction in wages per dollar of health benefits.

• Miller R.D. (2004). Estimating the compensating differential for employer-provided health insurance. Int J Health Care Finance Econ. 2004;4(1):27-41. I find a compensating differential for health insurance equal to roughly 10 to 11 percent of wages. Some caution is advised here due to the fact that I was unable to control for other fringe benefits, the most important being paid vacation and sick leave.

• Goldman D, Sood N, Leibowitz A. (2005).  Wage and benefit changes in response to rising health insurance costs. NBER Working Paper 11063. 2005. This paper examines employee compensation decisions during a three-year period when health insurance premiums were rising rapidly. The data come from a single large firm with a flexible benefits plan wherein employees explicitly choose how to allocate compensation between cash wages and other benefits. Under such an arrangement, higher health insurance premiums must induce changes in the composition of total compensation — either in lower after-tax wages or in decreased contributions to other benefits. The results suggest that about two-thirds of the premium increase is financed out of cash wages and the remaining one-thirds is financed by a reduction in benefits.
• Baicker, Katherine and Amitabh Chandra (2006). The Labor Market Effects of Rising Health Insurance Premiums, Journal of Labor Economics, 24(3), 609-634. We estimate the effect of rising health insurance premiums on wages,employment, and the distribution of part-time and full-time work using variation in medical malpractice payments driven by the recent “medical malpractice crisis.” We estimate that a 10% increase in health insurance premiums reduces the aggregate probability of being employed by 1.2 percentage points, reduces hours worked by 2.4%, and increases the likelihood that a worker is employed only part time by 1.9 percentage points. For workers covered by employer provided health insurance, this increase in premiums results in an offsetting decrease in wages of 2.3%.
• Gerakos, Joseph. (2008). Chief Executive Officers and the Pay-Pension Tradeoff (December 2008). The theory of equalizing differences predicts that workers trade pay for benefits, but empirical confirmation of such tradeoffs is rare. This study investigates the extent to which chief executive officers (CEOs) trade pay for pension benefits. For a sample of S&P 500 CEOs, I find that an additional dollar of pension benefits is associated with a 48 cent decrease in pay. Although the tradeoff estimate is significantly different from zero, it is also significantly less than the anticipated rate of dollar for dollar, especially for CEOs with relatively more power over their boards of directors.
• Marks, Mindy S. (2011). Minimum wages, employer-provided health insurance and the nondiscrimination law. Industrial Relations 2011;50(2); 241-262. This article exploits cross-state variation in minimum wages to investigate the impact of minimum wage changes on employer-provided health insurance. In contrast to the existing empirical literature, this article considers an environment where some firms are constrained by non-discrimination laws that govern the provision of health insurance. For these firms, minimum wage changes do not reduce the probability that workers will receive employer-provided health insurance. For firms not covered by the non-discrimination law, and free to tailor their fringe benefits, low-skilled workers experience a disproportionate reduction in the availability and generosity of health insurance after a minimum wage increase.
• Kolstad, Jonathan T. and Amanda E. Kowalski (2012). Mandate-Based Health Reform and the Labor Market: Evidence from the Massachusetts Reform. NBER Working Paper No. 17933. Issued in March 2012. We find that jobs with ESHI pay wages that are lower by an average of $6,058 annually, indicating that the compensating differential for ESHI is only slightly smaller in magnitude than the average cost of ESHI to employers. Because the newly-insured in Massachusetts valued ESHI, they were willing to accept lower wages, and the deadweight loss of mandate-based health reform was less than 5% of what it would have been if the government had instead provided health insurance by levying a tax on wages.
• Devaro, Jed and Nan L. Maxwell (2014). The Elusive Wage-Benefits Trade-Off: The Case of Employer-Provided Health Insurance. International Journal of Industrial Organization 37(1) · October 2014. This study shows that standard regressions estimated to measure a trade-off between wages and health insurance are misspecified by insufficiently accounting for establishment and firm size; an interactive, size-corrected specification is more likely to reveal a trade-off. Furthermore, because insurance decisions are typically made by firms, and wages set by establishments, the insurance constraint on establishments in multi-establishment firms weakens the trade-off. We use model generated data to show that both factors contribute to the failure in previous research to identify a trade-off, and data from a cross section of Northern Californian establishments to test for a tradeoff in multi-establishment and single-establishment firms.
• Qina, Paige and Michael Chernew. Compensating wage differentials and the impact of health insurance in the public sector on wages and hours. Journal of Health Economics 38, December 2014, Pages 77–87. We find modest reductions in wages are associated with having employer-sponsored health insurance (ESHI), although this effect is not precisely measured. The reduction in wages associated with having ESHI is larger among non-unionized workers. Further, we find little evidence that provision of health insurance increases hours worked.

Studies Showing Compensating Wage Differential May Be Individualized

Most studies do not examine whether the compensating wage differential varies by worker characteristics within the same firm. A number of studies that have examined this question have concluded that the wage offset appears to be individualized at least to some extent. The categories below are listed in approximate order of the number of workers having access to employer-provided health benefits.

Income Differential

• Swabish, Jonathan. (2004) Accounting for wages and benefits using the ECI. Monthly Labor Review, Vol. 127, No. 9 (September 2004), pp. 26-41. To date, no researcher has convincingly estimated a hedonic model although many have produced mixed results. The estimates in this article point more to correlations between wages and benefits than to tradeoffs….the estimates suggest compensating differentials for subgroups of the population, namely the (weakly) 10th and 90th percentiles but tell a stronger story about the positive correlation between wages and benefits and how they differ at various points in the wage distribution.
• Toder, E, Smith, KE. (2011)  Do low-income workers benefit from 401(k) plans? Center for Retirement Research at Boston College. [T]hese studies assume that employer contributions to 401(k)s do not affect the total compensation that each worker receives – that is, every worker “pays for” employer contributions in the form of lower wages. This brief challenges this assumption, testing whether employer contributions may actually increase total compensation for low-income workers, who may be more reluctant than high-income workers to accept wage reductions in exchange for retirement saving contributions…These results support the notion that the fringe benefit/wage tradeoff can vary for workers at different income levels. For high-income workers, additional 401(k) contributions are almost fully offset by lower wages. For low-income workers, additional contributions reduce wages only modestly – by just 11 cents to 29 cents per dollar – suggesting that employer contributions increase total compensation for lower-income workers.

Workers with Family Coverage

• Sheiner Louise (1999). Health Care Costs, Wages, and Aging. Washington, DC: Federal Reserve Board of Governors; April 1999. This study provides evidence that workers who elect to take family coverage pay for their health benefits through lower wages.

Older Workers

• Sheiner Louise (1999). Health Care Costs, Wages, and Aging. Washington, DC: Federal Reserve Board of Governors; April 1999. This study provides evidence that older workers pay for their health benefits through lower wages.
• Jensen GA, Morrisey MA (2001). Endogenous fringe benefits, compensating wage and older workers. International Journal of Health Care Finance and Economics 2001;1(3/4, SeptemberDecember); 203-226. This paper uses 1994 and 1998 Health and Retirement Survey data to examine the wage-health insurance trade-off for older U.S. workers. Job and insurance choice are treated as endogenous in a two stage least squares framework. There is strong evidence supporting the treatment of nonwage benefits as endogenous. The preferred specification indicates an annual health insurance wage adjustment of $6,300. The magnitude of the trade-off is fragile, however.
• Adams, Scott (2007). Health Insurance Market Reform and Employee Compensation: The Case of Pure Community Rating in New York. Journal of Public Economics 91 (5): 1119-33. Pure community rating, which was enacted to improve access to health insurance in New York’s small group market in 1993, prevents carriers from charging different premiums based on the ages of a firm’s workers. If small firms were adjusting compensation packages prior to reform to offset higher health care costs of older workers, then community rating could lead to greater relative wages for older workers post reform and not necessarily induce adverse selection that results in changes in who is insured. I present evidence showing that relative wages of older workers in small firms increased in comparison with other states and with large firms within New York following reform.

Working Wives

• Olson, Craig A. (2002). Do Workers Accept Lower Wages in Exchange for Health Benefits? Journal of Labor Economics 20, no. 2 (April): S91-S114. This study finds that among wives eligible for health benefits coverage, those women who accept coverage earn a lower wage relative to women who decline the coverage. Finally, the dollar value of the estimated wage-health insurance trade-off is very close to the average expected health care costs paid for by health insurance.
• Zhu L (2005). Is there a trade-off between employer-provided health insurance and wages? Unpublished paper. Southern Methodist University: 2005. The results in this paper, based on Current Population Survey (CPS) 2004 data on full-time working wives, not only suggest the existence of a tradeoff, but also suggest the existence of an indirect effect of HI on wages. The results explain why previous researchers cannot establish the tradeoff. HI not only affects wages from the direct tradeoff point of view, but also affects wages by increasing the productivity through better health and higher morale of employees. These two effects work in the opposite direction and are closely related. Therefore, ignoring the indirect effect will result in biased estimation of the tradeoff.

Maternity Benefits

• Gruber, Jonathan (1994). The Incidence of Mandated Maternity Benefits. The American Economic Review 84, no. 3 (June): 622-641. This study examined the impact of a state mandate to cover maternity benefits finding that reductions in the wages of women of child-bearing age and their spouses roughly offset the average costs of providing those benefits. Results correspond to shifting of 50± 200% of the cost of the mandate onto wages (Table 8).

Obese Workers

• Bhattacharya, Jay and M. Kate Bundorf (2005). The Incidence of the Healthcare Costs of Obesity. NBER Working Paper No. 11303, May 2005. Authors  find that obese workers with employer-sponsored health insurance pay for their higher expected medical expenditures through lower cash wages. This conclusion is strengthened by their finding that these types of wage offsets do not exist for obese workers with insurance coverage through an alternative employer. Nor are there wage offsets for other types of fringe benefits whose cost to the employer is less likely to be affected by BMI.
• Bailey, James (2013). “Who Pays for Obesity? Evidence from Health Insurance Benefit Mandates.” Economics Letters 121, no. 2 (2013): 287-89.  In the late 1990s and early 2000s, many state governments began requiring health insurance plans to cover treatments for diabetes. Using difference-in-difference analysis of restricted geocode data from the 1979 National Longitudinal Survey of Youth to compare wages across states with and without diabetes mandates, I find that obese people pay for all of their own increased health costs in the form of lower wages, rather than passing them on to employers, insurers, and co-workers.

Studies With Mixed or Null Results

• Ehrenberg, R. G. & Schumann, P. L. (1981). Compensating wage differentials for mandatory overtime (NBER Working Paper Series No. 805) [Electronic version]. Our paper estimates the extent to which employees are compensated for an unfavorable job characteristic, being required to accept mandatory assignment of overtime, by receiving higher straight—time wages. Our estimating equations are derived from a model in which wage rates and the existence of mandatory assignment of overtime are jointly determined in the market by the interaction of employee and employer preferences. While on average, we do not observe the existence of a compensating wage differential for mandatory overtime, we do observe the existence of such differentials for unionized workers and workers with only a few years experience at a firm.
• Leibowitz, A. (1983). Fringe benefits in employee compensation. in: J.E. Triplett, ed., The measurement of labor cost (The University of Chicago Press, Chicago, IL) pp. 371±389. Positive but insignificant relationship between wages and HI (Table 8).
• Monheit, A.C. et al. (1985), The employed uninsured and the role of public policy. Inquiry 22: 348±364. Positive relationship between wages and HI (magnitude not reported) (Table 8).
• Miller, R.D. (1995). Estimating compensating differentials for employer-provided health insurance benefits. Unpublished paper (University of California at Santa Barbara), cited in Currie and Madrian (1999). Author finds mixed results of wage-health insurance trade-off (Table 8).
• Johnson, N, Provan, K G. (1995) The relationship between work/family benefits and earnings: A test of competing hypotheses. Journal of Socio-Economics, Vol. 24, Issue 4. Based on the results of this research, it is reasonable to conclude that compensating wage differential theory alone has limited value for explaining the impact of work/family benefits on earnings. Because work/family job benefits are becoming so important, particularly for women, the inability of the predominant theory to explain  earnings sheds doubt on its generalizability and suggests the need to develop new theory to help explain the  phenomenon. Only when our analysis shifted to increasingly narrow categories of benefits and recipients was compensating  wage  differential  theory  supported. Specifically,  significant  results  were  found  only  for  sick  leave benefits for professional women. What our results suggest is that work/family benefits may both negatively affect earnings, through compensating wage differentials arising from benefit costs, and positively affect earnings, through enhanced  productivity resulting from reductions in work/family conflict….employers may have little need or incentive to pass on these costs in the form of lower earnings if productivity is enhanced.
• Ryan, S. (1997). Employer-provided health insurance and compensating wage differentials: evidence from the survey of income and program participation. Unpublished paper (University of Missouri-Columbia), cited in Currie and Madrian (1999). Author finds mixed results of wage-health insurance trade-off (Table 8).
• Thurston, N.K. (1997). Labor market effects of Hawaii’s mandatory employer-provided health insurance. Industrial and Labor Relations Review 51: 117±135. Effect of HI coverage on wages depends on how changes in HI coverage are measured as well as on estimation technique (OLS versus median regression); effects range from negative and significant to positive and significant (Table 8).
• Levy H, Feldman R. (2001). Does the incidence of group health insurance fall on individual workers? International Journal of Health Care Financing and Economics 2001;1(3/4, SeptemberDecember); 227-247. We find no evidence of a significant wage offset at either the individual or group level and conclude that changes in health insurance status are not exogenous.
• Simon, Kosali Ilayperuma and Robert Kaestner (2004). Do minimum wages affect non-wage job attributes? evidence on fringe benefits. Industrial and Labor Relations Review, 58(1): 52–70, October 2004. We examine effects of state and federal variation in the minimum wages on groups likely to be affected by the minimum wage. These effects are compared to estimates found for groups unlikely to be affected by minimum wages. Our results indicate no discernible effect of the minimum wage on fringe benefit generosity for low-skilled workers. This conclusion is unchanged whether we use only state level variation or federal and state variation in minimum wages.
• Lehrer SF, Pereira NS (2007). Worker sorting, compensating differentials and health insurance: Evidence from displaced workers. Journal of Health Economics 2007;26; 1034-1056. We do not find any evidence of a health insurance compensating wage differential. We hypothesize that the inability to identify the impact of health insurance on wage levels is likely a consequence of omitting firm characteristics that arise from aggregate worker sorting rather than the heterogeneity of individual job-skill matching.
• Buchmueller, T, Dinardo, J and Valletta, R. (2009) Employer Health Benefits and Insurance Expansions: Hawaii’s Experience. FRBSF Economic Letter, June 29.2009. We find that Hawaii’s ESI mandate has substantially increased health insurance coverage in the state. Our evidence also suggests that employers’ primary response to the mandate was increased reliance on the exempt class of workers who are employed for fewer than 20 hours per week. We did not find reliable statistical evidence for corresponding reductions in wages or overall employment probabilities.
• Abraham, J, Lluis, S. (2009) Compensating Differentials and Worker Selection of Fringe Benefits: Evidence from the Medical Expenditure Panel Survey 1997-2006. Department of Economic, University of Waterloo. In this paper, we evaluate empirically the importance of worker selection into jobs with different combinations of fringe benefits and how this impacts mobility and wage outcomes… Overall these results are inconsistent with the theory of a wage-benefit trade off. Given the previous selection outcomes found for the sample of involuntary job changers, firms that offer other benefits (including health insurance) seem to be able to pay higher wages by attracting higher skilled and healthier workers. Based on our findings, one could speculate that if a tradeoff does exist, then it may be occurring through the allocation of other non-wage benefits (e.g., retirement plan, paid vacation, paid sick leave) rather than through the decision to offer/accept lower wages.
• Sood, Neeraj, Arkadipta Ghosh, and José J. Escarce. (2009) Employer-sponsored insurance, health care cost growth, and the economic performance of U.S. industries. Health Services Research 44 (5 Pt 1): 1449-64. The findings of our study indicate that “excess” growth in health care costs has adverse effects on economic outcomes in the United States, and that these effects are greater for industries where high percentages of workers have ESI. This does not prove or disprove the wage-benefits offset hypothesis, but is consistent with the tradeoff being smaller than 1:1. If the entire cost of health benefits were shifted to workers, then rising health costs would have no effect on total compensation, hence there should be no adverse effects on employment.
• Qu, Dan  (2010). Health Insurance, Pensions and Wages. University of South Carolina. Theses and Dissertations. Paper 392. The author finds mixed results. The first chapter uses 1996 and 2001 Survey of Income and Program Participation to examine the evidence for compensating wage differentials associated with EHI in the US. The results provide no evidence of a tradeoff between wages and EHI coverage. On the other hand, the results do suggest that employees who work in states with income taxes are more likely to receive EHI than those in states without income taxes. Fixed effects, first differencing and instrumental variable estimation are used to address the potential ability bias and endogeneity problems in wage models with EHI as an independent variable. While both fixed effects and first-differencing estimation provide evidence of a positive relationship between wages and EHI, validity tests cast doubt on fixed effects estimation. Instrumental variable estimates, however, provide no evidence of a tradeoff in either direction.
• Anand, Priyanka (2013). The Effect of Rising Health Insurance Costs on Compensation and Employment. Unpublished paper (Mathematica Policy Research, 3.25.13). “I find that establishments that offer health insurance reduce total compensation by $0.52 for each dollar increase in health insurance costs. Establishments primarily rely on increasing employee contributions when passing along the additional cost of health insurance to workers, while the effect on wages and non-health fringe benefits is approximately zero. Thus, this study demonstrates there is a trade-off even though it does not affect wages.”
• Clemens J, Cutler DM (2014). Who pays for public employee health costs? J Health Econ. 2014 Dec;38:65-76. “We find that roughly 15 percent of the cost of recent benefit growth was passed onto school district employees through reductions in wages and salaries... This estimated wage incidence is statistically indistinguishable from 0 but is statistically differentiable from −1.
• Anand, Priyanka (2016). Health Insurance Costs and Employee Compensation: Evidence from the National Compensation Survey. “I estimate the extent to which total compensation decreases with a rise in health insurance costs and decompose these changes in compensation into adjustments in wages, non-health fringe benefits, and employee contributions to health insurance premiums. The paper examines this relationship using the National Compensation Survey, a panel dataset on compensation and health insurance for a sample of establishments across the USA. The paper finds that total hourly compensation reduces by $0.52 for each dollar increase in health insurance costs. This reduction in total compensation is primarily in the form of higher employee premium contributions, and there is no evidence of a change in wages and non-health fringe benefits. These findings show that workers are absorbing at least part of the increase in health insurance costs through lower compensation and highlight the importance of examining total compensation, and not just wages, when examining the relationship between health insurance costs and employee compensation.”

Studies Showing Positive Relationship Between Health Benefits and Wages

• Monheit, A.C. et al. (1985), The employed uninsured and the role of public policy. Inquiry 22: 348±364. Positive relationship between wages and HI (magnitude not reported) (Table 8).
• Olson, C.A. (1992). The impact of permanent job loss on health insurance benefits. Unpublished paper (University of Wisconsin-Madison). “Displaced workers who lost health insurance have post- displacement wages 25% below those of displaced workers who were able find new jobs which also offer health insurance coverage  (Table 8). I find no evidence that workers who lost health insurance benefits received higher wages on their new jobs to compensate for the loss. In fact, a displaced worker that lost health benefits suffered a greater wage loss than a comparable worker who gained health benefits.”
• Gruber, J. and M. Hanratty (1995). The labor market effects of introducing national health insurance: evidence from Canada. Journal of Business and Economics Statistics 13: 163±174. NHI leads to a 1.4-4.2% increase in average weekly earnings; effects are bigger in industries with low initial private HI coverage rates  (Table 8). If the wage-benefit hypothesis were correct, wage growth should have been larger in industries with higher HI coverage rates since they were able to substitute public coverage for employer-provided coverage thereby freeing up more compensation resources to be paid in wages etc.
• Buchmueller, T.C. and M.K. Lettau (1997). Estimating the wage-health insurance tradeoff: more data problems? Unpublished paper (University of California at Irvine), cited in Currie and Madrian (1999). Author finds positive relationship between wages and provision of health insurance (Table 8).
• Simon, Kosali (2001).  Displaced Workers and Employer-Provided Health Insurance. International Journal of Health Care Finance and Economics. September 2001, Volume 1, Issue 3, pp 249-271. “Job changes that result from plant closings and mass layoffs provide an opportunity to see how workers respond to an employment shock that is arguably exogenous to individual productivity. Comparing compensation packages of displaced workers on their old and new jobs is a potentially promising method to infer a tradeoff between wages and non-wage benefits. Although displaced worker data overcomes many of the pitfalls to estimating wage/fringe tradeoffs by controlling for time-invariant unobserved productivity, time-varying unobservables could still bias estimates. In this analysis, I investigate the compensating wage differential for one particularly valuable benefit, employer-provided health insurance. I find that even after controlling for an extensive set of productivity factors, I obtain results indicating a wrong-signed tradeoff. Those who lose health insurance through the job change also lose wages relative to other displaced workers, while those who gain health insurance also gain in wages. Individuals expected to incur higher health care costs (older workers and workers who are likely to buy family coverage) do not experience steeper wage/health insurance tradeoffs as would be expected if employers were able to pass health care costs on to workers according to individual costs.”
• Oyer, Paul. (2008) Salary or Benefits? Work, Earnings and Other Aspects of the Employment Relation, 2008, 429-467. Positive relationship. Argues that firms offer benefits “packages that lower cost of effort and can be purchased efficiently.” (Table 1)
• Lubotsky, Darren and Craig A. Olson (2012). Premium Copayments and the Trade-Off between Wages and Employer-Provided Health Insurance. NBER, August 16, 2012 (now published in Journal of Health Economics, December 2015). “We find no evidence that teachers’ salaries respond to changes in insurance costs. Consistent with a higher willingness to pay for insurance, we find that premium contributions are higher in districts that employ a higher-tenured workforce. We find no evidence that school districts respond to higher health insurance costs by reducing the number of teachers.”
• Dizioli, A, Pinheiro, R. (2015) Health Insurance as a Productive Factor. “In this paper, we showed that health coverage has a positive impact on labor productivity, by reducing the number of sick days a worker needs to take…In equilibrium, firms offering health coverage are bigger and offer higher wages on average. These results are also corroborated by our empirical findings with the MEPS. According to our empirical results, increases in firm size and wage earned are positively related to the probability of a worker having health insurance coverage.”
• Aizawa, Naoki, Fang, Hanming (2015) Equilibrium Labor Market Search and Health Insurance Reform.  Federal Reserve Bank of Minneapolis Research Department Working Paper 727, July 2015. “Table 3 reports the summary statistics of the key variables in the 1996 SIPP data… The average 4-month wage for employed male workers with health insurance is about $9,050, higher than that for those without health insurance which is about $6,090; the average 4-month wage for employed female workers is about $7,370 for those with health insurance and $5,360 for those without health insurance… in Table 5 we provide the summary statistics for our firm side data based on RWJ-EHI 1997. In general, firms that tend to offer health insurance have larger size in employment and provide higher wage… firms that offer insurance on average pay more than firms that do not offer health insurance; and on average workers with health insurance earn more than those without health insurance.”

Worker Preference

Workers’ Desire Grows For Wage Increases Over Health Benefits. “More wages, less health insurance. In a recent survey, one in five people with employer-based coverage said they would opt for fewer health benefits if they could get a bump in their wages. That’s double the percentage who said they would make that choice in 2012. ‘I do these surveys all the time, and it’s rare where you see things change that quickly,’ said Paul Fronstin, director of the health research and education program at the Employee Benefit Research Institute, which conducted the survey of 1,500 workers with Greenwald & Associates… The growing willingness to trade health benefits for wages may be linked to some degree to the millennial generation’s growing share of the workforce, Fronstin said, referring to people born between roughly 1980 and 2000. ‘The younger you are, the less important health insurance is to you,’ Fronstin said.” (Kaiser Health News, 4.1.16)