Summary

As we have documented in earlier research, 16 million US workers (about one of every 10 workers in the economy) have no health insurance. In this report, we examine the current sources of health coverage for those workers who do have some form of health insurance.

The most important source of health insurance for workers ages 18 to 64 are employer-based policies, which cover seven of every 10 workers (70.7 percent).

Medicaid is the second most important source of coverage for workers, insuring one of every 10 workers in the country (10.3 percent).

Individual policies purchased through the Affordable Care Act (ACA) cover just under five percent of workers (4.6 percent). Less-regulated private insurance policies issued outside of the ACA insure a smaller share of workers (3.7 percent).

Military-related health insurance through the Veterans Administration and other sources provide insurance for a smaller share of workers (2.9 percent).

But, as we document here, coverage arrangements vary substantially by workers’ individual demographics (such as race and ethnicity, gender, and age) and by the characteristics of the jobs they hold (public versus private sector, unionization, and pay levels).

In the bottom 40 percent of the wage distribution, 16.8 percent of workers lack any kind of health coverage; in the top 20 percent, the uninsured rate falls to 2.9 percent.

Employer-based policies cover a much smaller share of Hispanic (53.1 percent) and Black (65.9 percent) workers than they do white (77.1 percent) or Asian (74.0 percent) workers.

Medicaid is an important source of health insurance for women workers: 11.5 percent, compared to 9.2 percent for men. This is especially true for Hispanic (18.1 percent) and Black (17.5 percent) women who work.

Two looming changes to health insurance policy threaten the current coverage of millions of workers.

The first is the possible rescission of the ACA’s subsidies to states to support their large and successful “Medicaid expansion” programs, which we estimate here, if fully enacted, could lead to as many as 12.8 million workers losing their Medicaid coverage.

The second immediate threat is the scheduled termination at the end of this year of temporary COVID-19-related expanded subsidies for low- and middle-income households who purchase health insurance through the ACA exchanges. We estimate, conservatively, that ending these subsidies could result in more than 1.1 million workers losing their current coverage.

Acknowledgements

We are grateful to Dean Baker and Julie Cai for their very helpful feedback on this report and an earlier related report.

Introduction

In a recent report, we documented that 16 million workers in the United States between the ages of 18 and 64 had no private or public health insurance.¹ This figure includes 10 million full-time workers who held jobs throughout the entire year.² In this report, we examine how workers who do have coverage obtain the coverage they have.

Workers seeking health insurance for themselves and their families must navigate a complicated and often confusing system. Access to coverage depends on their employer, their household income, family size, age, veteran status, the state they live in, whether or not they are covered by a union contract, and other factors. As we documented in our earlier report, this system has consistently left 10 percent or more of the workforce without any form of health insurance.

The most important source of health insurance for workers 18 to 64 years old are employer-based policies, which cover seven of every 10 workers (70.7 percent, see Figure 1).³ The next most important source of coverage for workers is the Medicaid program, which insures one of every 10 workers in the country (10.3 percent). Individual policies purchased through the Affordable Care Act (ACA) cover just under five percent of workers (4.6 percent).⁴ Less-regulated private insurance policies issued outside of the ACA insure a smaller share of workers (3.7 percent). Military-related health insurance through the Veterans Administration and other sources provide coverage for a smaller share of workers (2.9 percent), who are veterans or members of their families.⁵ A small share of workers in the 18 to 64 year age range are covered by Medicare (1.6 percent), the health insurance program that focuses primarily on the population 65 and older. These younger workers generally qualify for Medicare because they have long-term disabilities.⁶ But despite this array of coverage options, the currently available sources of health insurance leave more than one in 10 workers without any form of health insurance (10.5 percent).

Figure 1

The coverage arrangements summarized for all workers in Figure 1, however, vary substantially by workers’ individual demographics (gender, race and ethnicity, age, and others) and by the characteristics of the jobs they hold (public versus private sector, union coverage, state, and others). In this report, we use data from the six most recent years of available data from the Annual Social and Economic Supplement (ASEC) of the Census Bureau’s monthly Current Population Survey (CPS)⁷ to document the ways that workers obtain health insurance coverage and how their coverage varies by their personal characteristics and the jobs they hold.

Sources of Health Insurance Coverage Vary Widely by Workers’ Demographics

The pattern of health insurance coverage in Figure 1 holds for the workforce as a whole, but sources of coverage vary substantially by worker demographics. Table 1A summarizes how the types of coverage differ across workers’ gender, race or ethnicity, age, marital status, the presence of children in their household, formal educational attainment, citizenship status and nativity, veteran status, and whether or not they live in a rural area.

Race and Ethnicity

Figure 2 compares the types of coverage workers obtain by their race and ethnicity.

White workers are the least likely to lack coverage entirely (6.7 percent), primarily the result of having the highest coverage through employer-based insurance plans (77.1 percent). White workers are the least likely to use Medicaid, but about one in 14 white workers (7.5 percent)  are insured through the program. Smaller shares of white workers use non-ACA private plans (4.2 percent) and ACA marketplace policies (4.2 percent). About three percent of white workers have insurance through a military-related policy and just under two percent are covered by Medicare.

Figure 2

Almost one of every four Hispanic workers (23.2 percent) lack any form of health insurance – the highest rate of the groups in Figure 2. Hispanic workers are the least likely to have coverage through their employer (53.1 percent) and the most likely to be enrolled in Medicaid (15.8 percent). They make slightly more use than average of the ACA (5.6 percent versus 4.6 percent on average). But Hispanic workers are less likely than average to use private non-ACA policies (2.9 percent versus the 3.7 percent average), military-related policies (2.0 percent versus 2.9 percent), or Medicare (0.9 percent versus 1.6 percent).

Black workers are more likely than the average to lack insurance (11.5 percent versus a 10.5 percent on average). Most have insurance through their employer (65.9 percent) but they are still far less likely than white workers (77.1 percent) to have an employer-sponsored plan. Medicaid is the next most important source of insurance for Black workers (15.0 percent). Black workers also have the highest share of coverage from military sources (3.4 percent versus 3.2 percent for white workers). The share with ACA policies was very close to the overall average, but the share of Black workers with private non-ACA plans was the lowest for the four groups in Figure 2. Just under two percent of Black workers between 18 and 64 are covered by Medicare.

Asian workers are, by a small margin, the least likely to lack insurance (6.3 percent versus 6.7 percent for white workers), an outcome that reflects a high share with employer-based coverage (74.0 percent) and the highest rates of ACA coverage (5.8 percent) and non-ACA market coverage (5.0 percent). Insurance through Medicaid (9.8 percent) lies between Black and Hispanic rates (both about 15 percent) and white rates (7.5 percent). Asian workers are the group least reliant on military-related coverage (1.8 percent) and only one percent of Asian workers are insured through Medicare.

Educational Attainment

Differences in health insurance coverage are even larger by educational attainment (Figure 3).

One third (33.3 percent) of workers with less than a high school degree have no health insurance from any source, and only slightly more (37.0 percent) have insurance through an employer (their own or another family member’s). Among the groups analyzed here, workers with less than a high school degree rely most heavily on Medicaid (22.9 percent). They also are somewhat more likely than workers with more education to use subsidized ACA plans (3.9 percent) and Medicare (2.0 percent), but the differences are small and they are also less likely to use unsubsidized ACA plans (only 1.0 percent). Workers who have not completed high school are the least likely to pay for non-ACA private insurance (2.9 percent) or to have coverage through military-related plans (0.7 percent).

Figure 3

A large share of workers with only a high school degree also lack health insurance of any kind (15.7 percent). High school graduates are far less likely than average to have employer-based coverage (60.9 percent) and substantially more likely than average to use Medicaid (15.2 percent). One of every 20 high school graduates has policies through the ACA (5.1 percent, where 3.9 percentage points are subsidized ACA plans). Non-ACA plans cover 3.3 percent, military-related policies cover another 2.5 percent, and 2.0 percent use Medicare.

Workers with some college, but not a four-year degree, are generally better off than those with only a high school degree, but still face major challenges. Almost one in 10 workers without a four-year college degree (9.4 percent) have no health insurance from any source. At 70.4 percent, employer-based coverage is their most important source of coverage, almost identical to the 70.7 percent average for the workforce as a whole. Workers with some college are somewhat more likely than average to rely on Medicaid (11.2 percent, versus 10.3 percent for all workers). Workers with some college are also the most dependent on military-related policies (3.9 percent). Medicare covers 1.7 percent of the group.

About one in 20 workers with a four-year college degree lack any form of health insurance (5.1 percent). College graduates overwhelmingly use employer-based plans (80.8 percent). While graduates make substantially less use of Medicaid (5.2 percent), the program still insures two million workers with a four-year college degree.⁸ Private plans through the ACA (4.6 percent) and private plans outside of the ACA (4.3 percent) together cover about one in 12 college graduates, while military-related plans insure 2.9 percent. College graduates are the least likely to be enrolled in Medicare (1.2 percent).

Only a relatively small share of workers with advanced degrees (those with more than a four-year college degree) lack any form of health insurance (2.9 percent). This group makes the highest use of employer-based plans (86.7 percent) and lowest use of Medicaid (3.0 percent), ACA plans (3.3 percent), and Medicare (1.2 percent) with the remainder in non-ACA market (4.0 percent) and military-related (3.1 percent) plans.

Age 

Young workers are more likely than older workers to go without health insurance. In recent years, the uninsured rate for 18-to-25 year-olds has averaged 13.9 percent, compared to 11.0 percent for those 26 to 50, and 7.5 percent for those 51 to 64 (Figure 4). (We have chosen the first age cut-off because the ACA allows children to remain on their parents’ employer-based coverage until age 26.)

Figure 4

Younger workers are less likely than older workers to have employer-based coverage through their own employer or their parents’ or spouses employer (63.6 percent for 18-to-25 year-olds, versus over 70 percent for the two older age categories, and these figures include employer-based coverage through a parental plan). Younger workers are also more likely than older workers to be insured through Medicaid (13.9 percent for 18-to-25 year olds, compared to 10.8 percent for 26-to-50 year-olds and 6.9 percent for 51-to-64 year-olds).⁹ Young workers are the least likely to rely on Medicare (0.6 percent), likely because Medicare eligibility for those under 65 is generally linked to long-term disability.

Nativity and Citizenship

Where workers were born and whether or not they have US citizenship is another major influence on their health insurance coverage (Figure 5).

Figure 5

Workers born in the United States, who as of this writing have a constitutional guarantee of US citizenship, are more likely than foreign-born US citizens and foreign-born noncitizens to have some form of health insurance. About 8.2 percent of workers who are US citizens lack health insurance, compared to 9.7 percent of foreign-born citizens, and 31.3 percent of foreign-born noncitizens.

US-born workers are more likely (74.1 percent) than foreign-born citizens (67.7 percent) – and far more likely than foreign-born noncitizens (44.5 percent) – to have employer-based coverage.

Differences in coverage through Medicaid close some of this gap: 9.4 percent of the US-born, compared to 12.2 percent of foreign-born US citizens and 15.8 percent of foreign-born noncitizens. Insurance through the ACA is also somewhat more important for foreign-born workers: 4.1 percent for the US-born versus 7.0 percent for foreign-born US citizens and 6.0 percent for foreign-born noncitizens. Foreign-born US citizens (1.4 percent) and foreign-born noncitizens (0.7 percent) are less likely than US-born citizens (1.7 percent) to be covered by Medicare.

Marital Status and the Presence of Children in the Household

Workers who are married are much less likely to lack coverage (7.0 percent) than those who are not (14.5 percent).

In a mechanical sense, the most important reason for this difference is the large gap in employer-based coverage (Figure 6). About 77 percent of married workers have employer-based coverage, compared to only about 63 percent of unmarried workers. This likely reflects an important feature of the US health insurance system, which is that employer-based plans typically allow workers (usually at additional cost to the worker) to include married spouses and children on their employer-sponsored plan. In all of the data presented here, a worker’s coverage through an employer-based plan may be through their own employer or an employer’s plan where another family member, such as a spouse or parent, is employed.

Figure 6

At the same time, unmarried workers are much more likely than married workers to have insurance through Medicaid (13.4 percent versus 7.5 percent). As with employer-based coverage, this disparity likely reflects legal differences in eligibility. Medicaid is means-tested and limited to low-income households, and –  importantly – income limits are adjusted for family size. If two low-income workers have identical incomes, one may qualify for Medicaid if there is also a child present in the household, while the other may be above the household-size-adjusted income cutoff if the worker has no other members in their household.

Households with children have similar likelihoods of being covered by employer-based plans (71.7 percent) as households without children (70.0 percent). But, households with children are much more likely to have insurance through Medicaid (12.8 percent) than households without children (8.3 percent).

Medicaid appears to be particularly important for working households raising children. As Table 1B shows, this is especially the case for working women. About 14.8 percent of working women with children and about 16.4 percent of unmarried working women obtain their insurance through Medicaid.

Veteran Status

Veterans are substantially less likely to lack health insurance coverage than non-veterans (a 5.6 percent uninsured rate for veterans versus a 10.8 percent rate for non-veterans, see Figure 7). While veterans and non-veterans have roughly similar rates for employer-based insurance (69.0 percent for veterans and 71.1 percent for non-veterans), veterans have a very large share of coverage through military-related plans (24.2 percent). A small share of non-veterans (1.4 percent) have military-related coverage through spouses or parents who are veterans.

Figure 7

Rural and Non-Rural

Rural versus non-rural¹⁰ residence appears to make relatively little difference in overall coverage rates or the composition of coverage (Figure 8). Workers living in rural areas are somewhat more likely to lack coverage (11.4 percent) than workers in non-rural areas (10.4 percent). The fairly small difference in employer-based coverage (68.5 percent in rural areas, compared to 71.1 percent in non-rural areas) is partially offset by a higher rate of utilization of Medicaid (11.3 percent versus 10.1 percent) and Medicare (2.1 percent versus 1.5 percent).

Figure 8

Gender

The biggest gap in coverage rates between genders is in the share of workers who lack health insurance (Figure 9). About one in eight men in the workforce (12.4 percent) lack health insurance, compared to about one in 12 women (8.5 percent). The main difference in coverage patterns is that women are somewhat more likely than men to have coverage through an employer or through Medicaid (2.3 percentage points higher in both cases, in part for reasons just discussed) and less likely than men to have a military-related plan (a 1.1 percentage point difference).

Figure 9

Tables 1B and 1C present the same data on coverage rates by worker demographics as in Table 1A (and Figures 1 through 8), separately for women (Table 1B) and men (Table 1C). This allows us to note any differences in coverage rates by gender, within the other categories we’ve seen so far. The differences between these two genders are fairly small. The main exceptions are for coverage through Medicaid and employer-based insurance, where women are more likely to be covered, and for military-related insurance, where women are less likely to be covered.

Coverage Also Varies Widely by Job Characteristics

So far, we have examined how the use of different kinds of health insurance varies by worker demographics. We can also look at how characteristics of the jobs that workers hold affect the kinds of health insurance they have. Table 2A shows the type of coverage workers hold along four dimensions: whether they work in the public or private sector; whether or not they are covered by a union; whether or not they are in a state that has expanded Medicaid coverage (where state laws can affect health insurance availability); and how well their job pays relative to other workers.

Public or Private Sector

Workers in public sector jobs are far less likely to lack health insurance (3.2 percent) than their counterparts in the private sector (11.6 percent) (Figure 10). This primarily reflects much higher rates of employer-based coverage in the public sector (83.7 percent, compared to 69.3 percent in the private sector) and, to a lesser degree, the higher share of public sector workers with military-related coverage (7.7 percent, versus 2.1 percent in the private sector).¹¹

Figure 10

Private sector workers make greater use of Medicaid (10.7 percent, compared to 5.3 percent for the public sector), ACA policies (5.0 percent versus 1.9 percent), and non-ACA private plans (4.0 percent versus 2.0 percent).

Union Status

Only 3.0 percent of workers covered by a union contract lack some form of health coverage. The uninsured rate for non-union workers, at 9.8 percent, is more than three times higher (Figure 11). The main driver of this difference is the substantially higher rate of employer-based coverage for union workers (89.0 percent) relative to non-union workers (75.2 percent). Union workers also make less use of Medicaid (5.6 percent versus 8.6 percent for non-union), ACA policies (1.9 percent versus 2.6 percent for non-union), and non-ACA policies (1.4 percent versus 3.1 percent).

Figure 11

Medicaid Expansion States

Workers who live (and generally work) in states that participate in the ACA expansion have a much lower uninsured rate (8.6 percent) than those in states that have declined to participate (15.7 percent, see Figure 12). Medicaid coverage for workers in expansion states is significantly higher (12.0 percent) than in non-expansion states (5.7 percent). Employer-based coverage is also higher in expansion states – 72.0 percent versus 67.3 percent in states that have not participated in the ACA-related expansion.

Figure 12

Wage Level

Health insurance coverage varies systematically with workers’ wages (Figure 13). In the bottom 40 percent of the wage distribution, 16.8 percent of workers lack any kind of health coverage; in the top 20 percent, the uninsured rate falls to 2.9 percent. The biggest reason for this difference in coverage rates is the much higher likelihood of having employer-based coverage in higher-paying jobs (87.7 percent at the top, more than 30-percentage-points higher than for the bottom 40 percent). This gap is only partially offset by higher shares of coverage through Medicaid and the ACA for lower-wage workers.

Figure 13

Tables 2B and 2C present the same data on coverage rates by job characteristics as in Table 2A separately for women (Table 2B) and men (Table 2C). As with the earlier demographic analysis, men in the workforce are, on average and across job types, consistently more likely to be uninsured. As was the case with worker demographics, the biggest differences in coverage types by gender are limited almost entirely to coverage by employer-based plans and Medicaid, where women are in almost every case more likely to be insured than men. Men are more likely to have military-related insurance, but both the difference in take-up rates and the number of workers eligible for veteran-related policies is small.

Some Policy Considerations

The complicated maze that is the US health insurance system leaves one in 10 US workers without insurance. Employer-based health insurance is the central pillar of the current system and covers 70 percent of the workforce, but employer-based coverage systematically fails low- and middle-wage workers, workers of color, younger workers, and immigrant workers. The remaining private and public forms of insurance jointly fall far short of filling the gap.

But the coverage data do not tell the whole story. Two looming potential changes to the health insurance system threaten to greatly increase the share of workers without coverage.

The first threat is a legislative proposal that would reduce or eliminate the long-standing ACA subsidy to states that raise the income ceiling for Medicaid eligibility. Under the ACA, the federal government pays nearly all of the state costs involved in increasing the upper limit for Medicaid eligibility from 100 percent of the federal poverty level to 138 percent of the federal poverty line. According to estimates by KFF and the Center on Budget and Policy Priorities, this subsidy has allowed more than 20 million people in households just above the poverty line to obtain coverage through Medicaid.¹² A separate KFF study found that 64 percent of those covered by the Medicaid expansion were working either full-time (44 percent) or part-time (20 percent). Taken together, these figures suggest that elimination of the Medicaid expansion could lead to as many as 12.8 million workers losing their Medicaid coverage (64 percent of the 20 million total covered through Medicaid expansion).

The second immediate threat to coverage is the scheduled termination of the  COVID-19-related expansion of subsidies to low- and middle-income households who purchase health insurance through the ACA exchanges.¹³ These subsidies triggered a large increase in ACA enrollment; after holding steady between 2015 and 2020 in the range of 11 to 12 million total enrollees per year, ACA marketplace coverage more than doubled between 2020 and 2025 to the current total of over 24 million enrollees.¹⁴

The temporary subsidies driving this doubling in ACA coverage, however, are set to expire at the end of 2025. A recent analysis by the Urban Institute predicts that if these subsidies are not renewed, “four million people could become uninsured.”¹⁵ The Urban Institute estimate  refers to all beneficiaries of the enhanced subsidies, not just the workers we analyze here. But these numbers allow us to make a rough, if conservative, estimate of the impact on workers alone. The Urban Institute projects that 4 million people (workers, non-workers, and other family members) could lose their coverage. This figure of 4 million people is out of a total number of enrollees in 2024 (the final year of the Urban Institute analysis) of 21.4 million, or about 19 percent of the total. If we assume a proportional impact on the subgroup of subsidized ACA participants who were workers in 2024, which was 5.8 million,¹⁶ this would suggest that more than 1.1 million workers could lose their coverage if the enhanced subsidies expire at the end of this year. This is likely a conservative estimate of coverage losses for workers because the COVID-19 subsidies provided relief to families with higher incomes than in the past — meaning that low- and middle-wage workers who previously made too much to qualify for the standard subsidies became eligible, and would be the most affected by the rescission of these subsidies.

While the ACA has undeniably made significant improvements in the scope of coverage and the quality of care, the US health insurance system remains a bewildering challenge for people trying to access health care. Unfortunately, the main lines of the current policy debate are focused almost exclusively on measures that would undermine the progress made under the ACA.

Despite the narrow focus of the current policy debate, there are alternatives to the current US health insurance system. A full discussion of the alternatives is beyond the scope of this report, but we can sketch some possible ways forward that would greatly expand workers’ coverage rates and provide continuity of their coverage as they move through the labor market — while maintaining and likely improving the quality of care received.

Possibly the simplest step in the direction of universal quality coverage would be to allow employers and individuals of all ages to “buy in” to the existing Medicare program, currently restricted to those 65 and older.¹⁷ This approach would build incrementally on three institutions of the existing health insurance system: Medicare, employer-provided health insurance, and the ACA Marketplace. An employer seeking to provide health insurance to its workforce could offer coverage through Medicare at a cost that is likely substantially below comparable private coverage. An individual without employer-based or other coverage could use existing subsidies to enroll in Medicare through the ACA Marketplace, where Medicare would compete on price and quality with private insurers that are also offering coverage.

Another option would be to gradually lower the eligibility age for Medicare, a program that already offers almost universal coverage to the population 65 and older.¹⁸ A plan to reduce the Medicare eligibility age by one year every year (or along some other schedule) could reduce coverage gaps for older workers and, if continued for a sufficiently long period, build a parallel public health insurance system available to the entire workforce, independent of their employer.

Beyond these more intentionally incremental approaches, the US could move more quickly to adopt a “single-payer” system. In single-payer systems, the government finances (almost) all health care costs for all residents, but the care is typically delivered by a mix of private and public entities.¹⁹

In the US context, proposals for a single-payer system are often referred to as “Medicare For All.” A US single-payer system would replace the current “multi-payer” system where individuals, employers, and the government pay for health insurance provided through a mix of private or public entities–with all the complications seen in the data presented earlier. In the best version of a single-payer system, all US residents would be covered for all medically necessary services, including doctor, hospital, preventive, long-term care, mental health, reproductive health care, dental, vision, prescription drug, and medical supply costs.

Canada’s Medicare program provides a useful, nearby point of reference for something close to a single-payer system. Their system is a decentralized, universal, publicly-funded single-payer health system.²⁰ All Canadian citizens are eligible for free, basic medical services. Services such as prescriptions, dental care, and private hospital rooms are often covered by private insurance. A common feature, in practice, of single-payer systems is that they often have a small private insurance market that offers private coverage for some services not included in the single-payer system.²¹

The US spends 16.7 percent of its GDP on health care, which is 5.7 percentage points higher than the average of comparable OECD countries. In 2023, US health spending per capita was $13,432, approaching double the $7,393 average per capita spending in this same set of comparable economies.²² A recent, careful comparison by The Commonwealth Fund of health care expenditures in the United States and OECD countries concluded: “More than half of excess US health spending was associated with factors likely reflected in higher prices, including more spending on: administrative costs of insurance (~15% of the excess), administrative costs borne by providers (~15%), prescription drugs (~10%), wages for physicians (~10%) and registered nurses (~5%), and medical machinery and equipment (less than 5%).”²³

The mix of private and public financing and provision of health care in the United States appears to be linked to higher rather than lower prices in the United States, relative to the systems in other countries where the private sector plays a much smaller role. These large differences in spending suggest that the US could, like our peer countries, feasibly provide high-quality universal coverage at a savings of multiple percentage points of GDP. In the short term, these savings could support a transition to a new system while addressing any longer term concerns about the scope and quality of care. In short, given how much we already spend, we could have a universal system with high quality care for all.²⁴

Table 1A

Table 1B

Table 1C

Table 2A

Table 2B

Table 2C

APPENDIX Data and Definitions

We use nationally representative data from Annual Social and Economic Supplement (ASEC), which is conducted annually by the Census Bureau as part of the monthly Current Population Survey (CPS).²⁵ The ASEC collects information from approximately 60,000 households on demographics, incomes, work status, and other characteristics, including detailed information on health insurance coverage. While the data are gathered primarily in March, most of the supplement’s questions focus on the household’s experience in the preceding calendar year. We focus here on the workers’ demographics, work status, and health insurance coverage throughout the calendar year preceding their March interview.²⁶ A household’s answers provided in March of 2024 (currently, the most recent survey), for example, refer to their employment situation and health insurance coverage throughout 2023.

The ASEC offers several methods to measure an individual’s health insurance coverage in the calendar year before their interview. Throughout this report, we follow the Census Bureau practice of referring to individuals as “uninsured” if they “did not have health insurance coverage for the entire calendar year” prior to their March interview. This produces a conservative estimate of the number of workers without health insurance because it excludes workers who had health insurance at some point in the year, but lacked coverage in one or more months of the year.

The ASEC data in the Appendix Table here update numbers that appeared in our earlier report. The new table reports coverage numbers as of March 2024, not the preceding calendar year. They also refer to coverage as of the survey month, not coverage throughout the entire previous calendar year.

We limit our sample to workers between the ages of 18 and 64. Workers 65 and older are overwhelmingly eligible for the federal Medicare program, which provides nearly universal coverage to the elderly, regardless of their income or employment status. Workers under 18 make up only a small portion of the workforce and their health insurance coverage generally comes through their parents (many of whom are in the workforce).²⁷ Our main interest here is to document the sources of the health insurance coverage that workers have and to explore how these sources vary by workers’ demographic characteristics and work situation.

With respect to demographics, the ASEC data allow us to identify workers by age, gender, race/ethnicity, marital status, presence or absence of a child in the household, and education level. The ASEC assigns two gender categories, male and female. We use detailed ASEC responses to define five mutually exclusive race and ethnicity categories: White, Black, Hispanic, Asian, and Other, where the Other category includes racial and ethnic groups with only small numbers of observations, even in the pooled multi-year data we use here. For simplicity, we limit marital status to two categories: currently married and not currently married (which includes single, divorced, and widowed). We distinguish between five mutually exclusive levels of formal education: less than high school, high school degree, some college but not a four-year degree, college graduate, and advanced degree.

To capture differences in employment status, we divide workers into three mutually exclusive categories that together cover all working-age adults who participated in work at some point in the calendar year: full-time, full-year workers; part-time or part-year workers; and workers who experienced at least one spell of unemployment of at least one week in duration. Full-time, full-year workers were employed in a job or jobs for at least 50 weeks in the year for at least 35 hours per week. Part-time or part-year workers were employed between one and 49 weeks in the year, or for fewer than 35 hours per week, or both. Our “unemployed” category includes survey respondents who had at least one spell of unemployment during the year, regardless of their full-time or part-time status when not unemployed.

Within each survey year, we also divide workers into five pay bands by first ordering all workers from the lowest to the highest hourly rates of pay, and then dividing them into five equally sized groups.²⁸ We define these wage quintiles across the entire wage distribution, grouping men and women together. This means that the wage quintiles we show below generally contain different numbers of men and women in each quintile. Importantly, women are overrepresented in lower quintiles and men are overrepresented in higher quintiles.

In order to increase the precision of the analysis for smaller demographic groups, we have pooled together the last six years of ASEC data. These surveys were fielded every March from 2019 through 2024, and refer to calendar years 2018 through 2023. Each year’s data are weighted to be representative of the population within each year. We include the March 2020 CPS, which was fielded just as the COVID-19 pandemic was declared. Excluding the March 2020 survey does not meaningfully alter any of our qualitative or quantitative findings.²⁹

Appendix Table 1