Employer-Sponsored Health Insurance Premium Cost Growth and Its Association With Earnings Inequality Among US Families

Key Points Question What is the association of increasing health care premium costs with earnings inequality and wage stagnation among US families with employer-sponsored health insurance? Findings In this economic evaluation of US families receiving employer-sponsored health insurance, the mean cumulative lost earnings from 1988 to 2019 associated with growth in health insurance premiums was $125 340 per family (in 2019 dollars) or nearly 5% of total earnings over the 32-year period. In all 32 years of the study, health care premiums as a percentage of compensation were significantly higher for non-Hispanic Black and Hispanic families than for non-Hispanic White families. Meaning This study suggests that increasing health insurance premium costs are likely associated with decreased earnings and increased income inequality, including by race and ethnicity, among US families receiving employer-sponsored health insurance and are meaningfully associated with wage stagnation.


Introduction
Health disparities and affordability of health care are 2 major challenges affecting US society.Half of US adults believe health care costs are difficult to afford, with differences in affordability reported by race and ethnicity. 1More than half of individuals in the US receive health insurance through an employer, 2 and the costs of employer-sponsored health care insurance (ESI) have increased faster than the overall economy for decades. 3The COVID-19 pandemic has reemphasized profound health inequities in the US, forcing the medical community to reevaluate how inequities are maintained today.
A potential important yet understudied association between health care affordability and health disparities in the US is indivduals' stagnant or even decreasing incomes.Since the 1980s, real wages have increased among the highest earners but have been flat for most workers, 3 leading to a widening earnings inequality. 4][8][9][10][11] Furthermore, most employers do not adjust the health care premiums charged to workers by employee earnings 12 ; thus, the displacement of wages owing to increasing health care premiums could be particularly problematic for lower-wage workers 13 and could be associated with earnings inequality. 14,15Little is known about the quantifiable associations with earnings inequality, including by race and ethnicity, of increasing health care premiums among US families with ESI.
We investigated 3 underlying questions about the increase in health insurance premiums.First, over 32 years beginning in 1988, how much of annual employee compensation do ESI premiums consume?Second, to what extent has the increase in ESI premiums exacerbated earnings inequality among workers receiving ESI, including by race and ethnicity and wage level?Third, over the 3 decades prior to the COVID-19 pandemic, how much have workers lost in wages due to increases in ESI premium costs?To address these questions, we combined several national data sources to analyze trends among workers with ESI family plans in employer-sponsored health care premiums, wages, and taxation, for 32 years from 1988 to 2019, overall and stratified by earnings bracket and race and ethnicity.

Study Design and Participants
In this economic evaluation, we used a serial cross-sectional analysis of national data to combine information on US family earnings (wages) among those with ESI, ESI premium costs, and payroll taxation from 1988 to 2019 for our overall sample and stratified by race and ethnicity and wage level.
We assessed this period because 1988 was the first year in which the US Census began including Asian as a racial category, and 2019 was the last full year of data prior to large disruptions in employment, wages, and health care coverage caused by the COVID-19 pandemic.We included data

Data Sources
Earnings from work (excluding income from other sources) among workers with ESI were obtained from the US Census Bureau's Annual Social and Economic Supplements of the Current Population Survey (CPS). 16For consistency with premium costs for a family plan, we assessed family earnings rather than household or individual earnings.The US Census defines a family unit as "a group of 2 people or more related by birth, marriage, or adoption and residing together," 17 which is similar to the requirement most employers use for eligibility for a family plan.We used median wages to depict a "typical" family with ESI because mean wages can be skewed by the highest earners.Race and ethnicity, assessed as sociocultural constructs, were self-reported by CPS participants and evaluated for the head of household in each family.
Employer-sponsored health insurance premium costs from 1998 to 2019 were derived from the Kaiser Employer Health Benefits Survey. 5][8][9][10][11] Therefore, we included both employer-paid and employee-paid premiums because together they represent the full costs of a health care benefit package that could partially displace wages.Employer-sponsored health care insurance spending is not subject to income or payroll taxes.Other studies have shown that the displacement of wages associated with ESI costs is mitigated slightly by ESI's tax deductibility 8,19 and that reductions in wages may lag a few years because of employee contracts and employer uncertainty about health care cost changes.Thus, we assumed that health care premium increases would displace wages, although not exactly dollar-for-dollar owing to payroll taxation. 8

Statistical Analysis
Statistical analysis was conducted from February 2022 to July 2023.To estimate net premium costs, we accounted for the taxation on the additional wages earned if an employee hypothetically received the value of their health insurance as additional wages (and did not receive any health insurance).
We subtracted this additional, hypothetical tax burden from the actual premium costs to calculate the net premium costs.We applied the federal payroll tax rate in each year from 1988 to 2019 20 to the median annual earnings for our overall sample and to the median annual earnings for each subgroup (ie, Asian, Hispanic, non-Hispanic Black [hereafter, Black], and non-Hispanic White [hereafter, White] and the 20th, 40th, 60th, 80th, and 95th percentiles of earnings among families with ESI).The benefits of tax deductibility are greater for higher-paid workers who are taxed at a higher rate; thus, the net premium costs are higher for lower-paid workers after accounting for their tax deductibility.All analyses used the specific net health insurance premium costs for each subgroup.
We hereafter use the term compensation to describe the combined value of wages and health care premiums, while acknowledging that total compensation may include additional benefits, such as paid time off or retirement fund contributions.We recognized that wages and health care benefits are the major, but not only, components of a compensation package that employers offer.The annual dollar value of other employee benefits, including retirement fund contributions and paid time off, were not available for our study population for all years in the analysis.Other nationally representative analyses suggest that contributions to retirement funds have decreased during large portions of our study years, while paid time off as a percentage of employee compensation has remained relatively constant. 21,22We chose not to include out-of-pocket medical costs because these are distinct from employee compensation and because premiums are a fixed component of employee compensation.Given trends in increasing out-of-pocket medical spending, 23,24 this exclusion of out-of-pocket costs likely underestimates the negative association of higher health care costs generally with family economic well-being.
To calculate the annual compensation package for the median US family with ESI, we summed median family earnings plus net premium costs for a family plan, adding contributions from both the employee and the employer.This approach recognizes that wages and health care premiums paid by an employer reasonably represent a compensation package that an employer is willing to pay, with employee premiums further offsetting wages for the employee. 8e

Results
In   Over the course of these 32 years, the mean cumulative lost earnings associated with increasing health care premiums was $125 340 (95% CI, $120 155-$130 525), equal to 4.7% of total earnings for the median family receiving ESI during the study period (Figure 3).If health care premium costs for families with ESI had increased but remained at the same proportion of the 1988 compensation package (at 5.5%), then by 2019 the projected median US family earnings would have been $104 774 (95% CI, $102 262-$107 287), an $8774 (95% CI, $8354-$9195) increase over the observed median earnings of $96 000 (IQR, $56 000-$150 000).Disparities in lost wages by race and ethnicity were also evident, with the lost earnings associated with increasing premium costs representing 7.5% (95% CI, 7.3%-7.6%) of compensation for White families in 2019, whereas they represented 9.6%  premium growth was compensated through lower wages 10 (consistent with our analysis when accounting for tax deductibility of health care premiums).Another analysis concluded that increasing health care costs from 1999 to 2009 caused families to lose an estimated $5600 per year in wages by 2009. 8Faced with excessive premium growth, some employers are also responding by pivoting to greater employee cost sharing through high-deductible plans, reducing plan generosity, or by shifting employees to part-time work. 6,9,27Our finding that families with ESI may have lost up to $125 000 over 3 decades due to increasing premium costs is consistent with former studies [6][7][8][9][10] and provides important new evidence on long-term associations over a 30-year career.
Fifth, few ESI plans vary premium costs by the wage level of the employee, 12 making increasing health care premiums more burdensome for lower-wage workers.Our findings should be interpreted to apply to most ESI plans that do not vary premium costs by wage level.Plans that offer more affordable premiums to lower-wage workers could help improve earnings inequality amid increasing health care premiums.However, even among the few large employers that adjust premium costs by earnings level, the relative price of premiums is often not directly proportional to the earnings gap between the highest-paid and lowest-paid workers in the company, 28 meaning low-income workers still pay a greater percentage of their income to premiums.
Our findings that lower-wage workers contribute more to ESI premiums as a percentage of their compensation aligns with evidence that, across the entire US health care system, the burden of health care financing is greater for lower-income than higher-income households. 29Economists, such as Kaestner and Lubotsky, 30 have further noted that the tax deductibility of health care premiums has a modest adverse association with income inequality because the net premium costs (ie, pretax value) are effectively lower for employees in the highest tax brackets.They found that eliminating the tax deductibility of health care premiums reduces the ratio of the 90th to the 10th percentile after-tax income distribution by approximately 4%. 30 Others have commented that ESI taxation in practice could be regressive because health benefits account for a larger share of compensation for low-income employees who thus would pay a higher percentage of their income to taxes. 31her nations, such as Canada and England, have a health insurance system that is separate from employment.Expanding US public insurance programs, such as Medicaid and Medicare, has been promoted by some as a means to improve equity. 32However, even with increasing premium costs, recent surveys show high worker satisfaction with ESI coverage. 33Prior to the passage of the Patient Protection and Affordable Care Act, ESI was voluntarily provided to recruit and retain talented employees.The Patient Protection and Affordable Care Act now requires companies with 50 or more employees to provide basic health insurance coverage, 34 although companies still compete on the overall benefits and costs of such coverage.

Strengths and Limitations
Our investigation has several strengths.First, our study adds important new evidence that increasing premium costs may increase earnings inequality, including by race and ethnicity, among families receiving ESI.Second, to our knowledge, this is the first study to assess 30-year growth in the percentage of compensation associated with ESI premium costs and its association with wage stagnation among US families with ESI.Third, we combined data from large, nationally representative data sets on ESI costs and family earnings to focus on families with ESI and potential disparities.Fourth, the analysis incorporated the full costs of ESI premiums by including costs paid by employees and their employers.Fifth, our analyses accounted for relevant payroll taxation rates in each year for each subgroup's annual earnings.This approach acknowledges that employers do not pay taxes on premiums, and thus net premium costs are lower than would appear without consideration of taxation.
Our analysis also has some limitations.First, our study design cannot determine causality, and employers may not redirect all savings associated with slower premium growth toward increased employee pay.For example, employers may instead invest in capital, new business ventures, or an

JAMA Network Open | Health Policy
Employer-Sponsored Health Insurance Premium Cost Growth and Earnings Inequality expanded workforce.][8][9][10][11] Second, increasing premiums could be offset by reductions in other employer-sponsored benefits, such as vacation time or retirement fund contributions; however, data limitations prevented us from exploring these associations.Historical trends support potential decreasing employer contributions to retirement funds and unchanged amounts of paid time off as a percentage of employee compensation over large portions of the study period. 21,22Thus, our findings may underestimate the true economic costs to families of increasing ESI premiums if greater premium costs not only cut into wages but also degrade the generosity of other benefits because employers have a finite amount of money to contribute to combined employee pay and benefits.Whether premium growth has also exacerbated racial and ethnic wealth gaps in retirement savings is an area for future research.Third, these findings are applicable to US families with ESI family plans with fixed premium costs for all earnings levels and are not generalizable to US families with non-ESI plans nor to uninsured families.Fourth, some employees may be willing to receive lower wages if costlier premiums buy higher-quality health care, such as coverage of new medications and treatments.Although health care spending increased faster in the US than in other high-income countries during our 32-year study period, the US continues to have the highest maternal and infant mortality, the lowest life expectancy at birth, and the highest death rates for avoidable or treatable conditions among high-income countries, suggesting that increased health care spending does not guarantee better population-level health outcomes. 35Fifth, we did not incorporate state income tax rates in our net premium cost calculations.However, state income taxes are typically small compared with federal payroll taxes (the mean state income tax in 2022 was only 2.1%), 36 and a handful of states do not levy income taxes; thus, their inclusion would likely have little effect on our findings.Sixth, our analysis does not account for increasing personal deductibles and other out-of-pocket medical expenses; thus, our findings likely underestimate the full negative association of increasing health care costs with economic disparities by race and ethnicity and wage level.

Conclusions
This economic evaluation of US families with ESI suggests that increasing health care premiums were associated with decreased annual earnings and increased earnings inequality, including by race and ethnicity and wage level, and were also associated with reduced median cumulative family earnings by approximately $125 000 over 32 years.Our results depict the hidden costs of increasing health insurance premiums for the US worker: less opportunity for wage growth and a heavier burden of health insurance premiums on lower-paid workers and on Black and Hispanic workers.Our analysis suggests a need for future research and for a corresponding US health care policy to examine the role that increasing health care premiums play in stagnating employee wages and increasing income inequality.

Figure 2 .
Figure 2. Percentage of Compensation Associated With Health Care Premiums Among Families With Employer-Sponsored Health Insurance From 1988 to 2019, by Earnings Percentile35 30

Figure 3 .
Figure 3. Lost Wages Associated With Health Care Premium Increases Among Families With Employer-Sponsored Health Insurance From 1988 to 2019

JAMA Network Open | Health Policy Employer
on wages only among US families with ESI and did not include families with other health insurance, such as Medicare or Medicaid.All dollar amounts were adjusted for inflation and are reported in 2019 dollars.This investigation was deemed not human participants research by the Tufts University -Sponsored Health Insurance Premium Cost Growth and Earnings Inequality JAMA Network Open.2024;7(1):e2351644.doi:10.1001/jamanetworkopen.2023.51644(Reprinted) January 16, 2024 2/12 Downloaded from jamanetwork.comby guest on 01/27/2024 Health Sciences institutional review board because it used publicly available data.Our study adhered to the Consolidated Health Economic Evaluation Reporting Standards (CHEERS) reporting guideline.