The current economic downturn highlights the degree to which Americans under age 65 rely on a healthy labor market for almost all facets of economic security—particularly access to health care.

While the Great Recession officially ended in the summer of 2009, the labor market continued to deteriorate into 2010. The unemployment rate increased from 9.3 percent in 2009 to 9.6 percent in 2010, and long-term unemployment (the share of the unemployed without a job for 27 weeks or more) grew from 31.2 percent in 2009 to 43.3 percent in 2010.

Most Americans, particularly those under age 65, rely on health insurance offered through the workplace. Thus, given these unemployment trends, it comes as no surprise that the share of Americans under age 65 covered by employer-sponsored health insurance (or ESI) eroded for the tenth year in a row in 2010, falling from 59.4 percent in 2009 to 58.6 percent. However, the situation started deteriorating long before the Great Recession: The share of Americans under age 65 covered by ESI eroded every year from 2000 to 2010, decreasing by a total of 10.6 percentage points. As many as 28 million more people under age 65 would have had ESI in 2010 if the coverage rate had remained at the 2000 level.

No demographic or socioeconomic group has been spared from the erosion of job-based insurance from 2000–10. Both genders and people of all ages, races, and education levels have suffered declines in employer-based coverage. Workers across the wage distribution, in small and large firms alike, and even those working full time and in white-collar jobs have also lost coverage.

The decline in ESI coverage has been accompanied by an overall decline in health insurance coverage. The number of uninsured non-elderly Americans stood at 49.1 million in 2010—12.9 million higher than in 2000. Increasing public insurance coverage, particularly among children, is the only reason the uninsured rate did not rise one-for-one with losses in ESI. In addition, key components in the Patient Protection and Affordable Care Act (also known as health reform) took effect in 2010, shielding young adults from further coverage losses.

This briefing paper begins by documenting the decline in ESI coverage among the entire under-65 population. It then examines at length a smaller subset of this population, workers age 18 to 64—an increasing share of whom have lost ESI as well as other forms of health insurance. It next analyzes declining rates of ESI coverage for children—and argues that if not for public insurance, the overall coverage rate among children would have fallen. This briefing paper then explores the ESI situation across all 50 states and the District of Columbia before concluding with a discussion of what the Patient Protection and Affordable Care Act means for the trends discussed herein.

This report’s central findings include:

• In 2010, the share of non-elderly Americans with employer-sponsored health insurance declined for the tenth year in a row, from 59.4 percent in 2009 to 58.6 percent. The total decline from 2000–10 now stands at 10.6 percentage points. In 2010, 13.6 million fewer non-elderly Americans had ESI than in 2000.
• As many as 28 million more people under age 65 would have had ESI in 2010 if the ESI coverage rate had remained at its 2000 level.
• Workers age 18 to 64 experienced losses in job-based coverage, with ESI coverage declining 3.1 percentage points from the last business cycle peak in 2007 to 2010. Among strongly attached workers (those who worked at least 20 hours per week for at least half the year), service-sector workers had the lowest rates of coverage from their own job and experienced the largest declines from 2007 to 2010. Strongly attached workers in small firms are far less likely to have coverage than those in large firms.
• Children’s employer-sponsored insurance coverage (obtained through their parents) fell 11.9 percentage points from 2000–10, and the gap in ESI access for children by income widened substantially over this period.
• The decline in ESI coverage from 2000–10 was felt nationwide, with a statistically significant decrease in non-elderly coverage in all but two states. No state had an increase in the share of its under-65 population with ESI coverage over this period.
• The decline in ESI coverage has been accompanied by an overall decline in health insurance coverage. In 2010, 49.1 million people under age 65 were uninsured, up by roughly three-quarters of a million people since 2009. The number of uninsured non-elderly Americans is 12.9 million higher than it was in 2000.
• Workers age 18 to 64 were 30 percent more likely to be uninsured in 2010 than in 2000. Uninsured workers are disproportionately young, Hispanic, less educated, and have lower incomes. The gap between coverage of full-time and part-time workers grew substantially from 2000–10: The share of full-time workers who are uninsured increased 3.2 percentage points, compared with a rise of 9.3 percentage points for part-time workers.
• Public health insurance is responsible for keeping millions from becoming uninsured, as job-based coverage sharply declined from 2000–10. Public insurance covered 22.5 million more people under age 65 in 2010 than in 2000.
• Though the Patient Protection and Affordable Care Act, informally known as health reform, will substantially insure more Americans (especially as the 2014 insurance exchange provisions take effect), high unemployment will likely lead to further ESI losses in the next couple of years.
• Critical provisions in the Patient Protection and Affordable Care Act have helped offset the declines in ESI coverage by insuring young adults through their parents’ health insurance policies.

Overall health insurance trends among the entire under-65 population

Looking at the entire under-65 population in the United States, employer-sponsored health insurance remains the predominant form of coverage; 58.6 percent of this population is covered by ESI. However, as shown in Figure A, coverage through work eroded each year from 2000 to 2010, declining by a total of 10.6 percentage points. This is not solely due to unfavorable economic conditions: ESI fell 5.8 percentage points even over the previous full business cycle, from the cycle’s peak in 2000 to its peak in 2007. Declines continued even after the recession ended in 2001 and the economy expanded. These losses in the share of the under-65 population with ESI coverage moderated considerably as the economy finally began adding jobs in 2003, but never reversed. The relatively small declines in coverage over the expansion increased as the recession took hold in 2008 and accelerated as the unemployment rate soared in 2009 and continued to stay high through 2010.

While for many Americans a loss of ESI means they effectively have no health insurance, the increase in uninsured Americans was not as steep as the fall in ESI (Figure B). In 2010, the share of those under age 65 who were uninsured rose to 18.4 percent. It increased by 1.8 percentage points from 2007 (when the recession began) to 2010, and by 3.7 percentage points from 2000 to 2010. In 2010, 49.1 million people under age 65 were uninsured, up by roughly three-quarters of a million people since 2009. The number of uninsured non-elderly Americans is 12.9 million higher than it was in 2000.

While the data do not track individuals over time to see what happens to specific people as they lose ESI, it is clear that overall coverage rates would have fallen further had there not been increases in public coverage, including Medicaid, the Children’s Health Insurance Program (CHIP), and Medicare. Public insurance covered 22.5 million more people under age 65 in 2010 than in 2000. As shown in Figure B, public coverage increased 3.3 percentage points from 2007 to 2010 and 7.3 percentage points from 2000 to 2010, partially offsetting losses in ESI. Non-group or direct purchase insurance coverage remained relatively flat over the entire period, failing to compensate for the losses in ESI (not shown).

Although declines in the share of the under-65 population covered by ESI slowed due to economic growth in the mid-2000s, the losses accelerated in 2008 with the recession. The steepest declines since the previous peak in 2007 occurred in 2009—a sharp drop of 2.9 percentage points—but the decline continued in 2010 as the unemployment rate continued to rise. ESI coverage declined 4.8 percentage points between 2007 and 2010 (Table 1). This resulted in a total loss of 10.6 percentage points from 2000 to 2010. About 1.2 million fewer people had ESI in 2010 than in 2009; 13.6 million fewer had this coverage in 2010 than in 2000. It is important to note that these figures fail to show the true extent of the erosion because they ignore population growth from 2000–10. As many as 28 million more people under age 65 would have had ESI in 2010 if the coverage rate had remained at the 2000 level.

Coverage losses occurred across all age groups, but young adults (age 18–24) consistently have the lowest rates of coverage. Less than half of this group receives health insurance through the workplace—though the Patient Protection and Affordable Care Act stemmed these losses in 2010 (for more on this, see the Health reform and coverage in 2010 section of this briefing paper).

The greatest declines in ESI were among children (age 0–17), with an 11.9 percentage-point drop from 2000 to 2010. Prime-working-age adults (age 25–54) were close behind, with an 11.6 percentage-point loss in ESI coverage. Coverage declined for males and females alike and across racial and ethnic classifications. As shown in Figure C, racial and ethnic disparities in coverage persisted over time, with non-Hispanic whites in 2010 experiencing rates of ESI coverage 71 percent higher than those of Hispanics and 48 percent higher than those of blacks. ESI coverage among the native born is 33 percent higher than among the foreign born, though the native born experienced larger losses since the recession began and over 2000–10 as a whole.

Educational attainment is correlated strongly with employer-based coverage; those with advanced degrees are nearly three times more likely to have ESI than those with less than a high school education.1 Just over half (50.6 percent) of those with only a high school degree have job-based coverage, compared with more than three-fourths (76.7 percent) of college graduates.

As with education, higher household incomes are strongly associated with an increased likelihood of having employer-sponsored coverage. In 2010, only 15.6 percent of those with household incomes in the bottom fifth had ESI, compared with 85.5 percent of those in the top fifth—a nearly five-and-a-half-fold difference in the likelihood of being insured through work. Each income group experienced losses from 2000–10; however, the declines were much greater for those at the bottom of the income scale. Those in the second-lowest fifth fared the worst following the onset of the recession, experiencing a decline of 9.3 percentage points from 2007–10, and a total decline of 18.3 percentage points from 2000 to 2010. While in percentage-point terms the bottom fifth fell less far (13.6 percentage points versus 18.3 percentage points), the rate of coverage for the bottom fifth fell 47 percent from 2000 to 2010, compared with 29 percent for the second-lowest fifth and 4 percent for the highest fifth.

Declining coverage for workers age 18 to 64

After having documented the decline in ESI coverage among the entire under-65 population, this briefing paper will now examine a smaller subset of this population: workers age 18 to 64. It first details the decline in ESI among these workers; it then examines the increase in the share of these workers without any form of health insurance, ESI or otherwise.

Employer-sponsored health insurance

Employer-sponsored health insurance coverage is higher among workers age 18 to 64 (68.5 percent) than among the under-65 population in general (58.6 percent). As shown in Table 2, workers’ declines in coverage have also been smaller—3.1 percentage points from 2007 to 2010 (compared with 4.8 percentage points for the entire under-65 population) and 7.3 percentage points from 2000–10 (compared with 10.6 percentage points). These trends are unsurprising given that ESI is secured via work; therefore, when workers lose employment, they often lose the health coverage that goes with it (unless they keep insurance as a retiree or a spouse, or never had it in the first place).

While declines in the share of workers covered briefly abated as the economy expanded from 2006 to 2007, coverage fell 4.2 percentage points from 2000 to 2007. The 3.1 percentage-point decline from 2007 to 2010 can be attributed partially to the start of the recession in December 2007 and partially to the overall trend of declining coverage.

Male workers have lower rates of coverage than female workers (66.9 percent versus 70.2 percent) and experienced larger declines from 2000–10. Similar to the overall under-65 population, large disparities exist in ESI coverage for workers by race and ethnicity. Nearly three-fourths of white non-Hispanic workers are covered, compared with less than half of Hispanic workers. Racial disparities in coverage widened since the last business cycle peak in 2007; the coverage rate for white non-Hispanic workers declined 2.7 percentage points from 2007 to 2010, in contrast to a 5.5 percentage-point decline for black workers. While black workers have higher rates of coverage than Hispanic workers, black workers’ rate of coverage fell faster from 2000–10 (8.6 percentage points versus 6.6 percentage points), due partly to the fact that Hispanic workers’ coverage increased from 2009 to 2010.

Native-born workers are 35 percent more likely to be covered than foreign-born workers, and a coverage gap of about 18 percentage points persisted over 2000–10.

Workers who are college graduates have far higher rates of employment-based coverage than workers with only a high school degree, at 80.0 percent and 61.8 percent, respectively. In addition, from 2000 to 2010, workers with only a high school degree experienced declines nearly twice as large as those of workers with a college degree (11.4 percentage points versus 5.9 percentage points).

Workers earning lower hourly wages are significantly less likely to have employer-sponsored health insurance than those earning higher wages; however, even those at the high end of the wage scale experienced declines in coverage over 2000–10.2 Only 41.0 percent of those in the lowest fifth, making at or below $9.38 an hour, have ESI, while 85.1 percent of those in the top fifth, with hourly earnings at or above $30.00, have coverage. Disparities in coverage by wage levels widened over 2000–10, with those in the top wage fifth being 74 percent more likely to be covered than those in the bottom fifth in 2000, but 108 percent more likely in 2010.

Nearly three-fourths of full-time workers have ESI, compared with less than half of part-time workers. Furthermore, part-time workers experienced a sharper decline in coverage since the start of the recession, with a decrease of 6.2 percentage points from 2007–10. Over 2000–10, their coverage fell 12.3 percentage points. The gap between full-time and part-time workers grew from 17 percentage points in 2000 to nearly 24 percentage points in 2010.

An important group of workers to examine more closely are those who are strongly attached to the private-sector labor force (i.e., those who work at least 20 hours per week and 26 weeks per year). Table 3 displays coverage for strongly attached workers who receive insurance through their own job (that is, not as a dependent receiving coverage via a spouse or parent) from 2000 to 2010 by selected job characteristics.3 Coverage for these workers fell 3.5 percentage points over the full business cycle from 2000–07. After a brief increase in the coverage rate in 2006 and 2007, coverage for these workers fell again for three years in a row, declining by 2.3 percentage points from 2007 to 2010. Only 53.1 percent of these steady workers receive health insurance from their employer, down 5.8 percentage points throughout 2000–10.

Strongly attached service-sector workers are insured through their own jobs at roughly half the rate of both white-collar and blue-collar workers and experienced the largest drop in coverage from 2007–10 (3.5 percentage points) and throughout 2000–10 (7.9 percentage points).

Workers in larger firms are more likely to receive health insurance from their own employer than workers in smaller firms. Only 26.3 percent of strongly attached workers in the smallest firms (with fewer than 10 employees) have ESI from their own job, compared with 43.6 percent in firms with 10 to 99 employees, and more than 60 percent in firms with 100 or more employees. Coverage losses from 2007–10 and over 2000–10 were greatest among workers in smaller firms (with 10–99 workers).

Low coverage rates among workers in small firms are due to many factors that make purchasing insurance much more expensive for small businesses than for larger firms, including an inability to offer attractive risk pools to potential insurers, high administrative and loading costs, and little competition in insurer markets (Gould and Hertel-Fernandez 2009). With the passage of the Patient Protection and Affordable Care Act (PPACA) in 2010, very small, low-wage firms could see considerable reductions in their premiums with the use of new tax credits. Using these data, it is not clear how these new tax credits, which came into effect in 2010, have affected coverage rates for workers at the smallest firms. However, firms with fewer than 10 workers did see smaller losses between 2009 and 2010 than firms with 10 to 99 workers (with a decline of 0.2 percentage point compared with a decrease of 1.6 percentage points). This trend may be related to the tax subsidies provided by PPACA, but there were many moving parts in the economy that also could account for those differences.

In 2014, all small firms will be able to purchase insurance though new insurance exchanges, which will make insurance costs more stable and predictable, even if one or more of their workers require(s) medical care or if their workforce size or composition changes from one year to the next.

Coverage rates in 2010 differ dramatically according to what sector of the economy strongly attached workers were employed in, ranging from 72.6 percent in the mining industry to 24.9 percent in the agriculture, forestry, fishing, and hunting industry. Overall, the highest rates of coverage are found in mining; manufacturing; and information; and the lowest in agriculture, forestry, fishing, and hunting; arts, entertainment, recreation, and accommodation; and other services (except public administration). However, all industries experienced declines from 2002 to 2010.4 The largest declines were in the sector with the lowest percent insured, agriculture, forestry, fishing, and hunting—with a 12.2 percentage-point decline in employer-sponsored insurance coverage. The smallest declines between 2002 and 2010 were in the finance, insurance, and real estate and leasing industry, with a drop of 0.4 percentage point.

In addition, previous research has shown that certain industries, such as public administration, mining, and manufacturing, are more likely to offer coverage to workers’ spouses or children, whereas arts and professional services fall short in this regard (Bivens, Gould, and Hertel-Fernandez 2009). In other words, the likelihood of receiving dependent coverage is higher among industries with higher rates of worker coverage.

Uninsured workers

Declines in ESI among workers tend to translate into growing numbers of workers without any type of health insurance. Uninsured workers are increasingly common in the U.S. economy; nearly one-fifth of the workforce is uninsured (Table 4). Workers age 18 to 64 were 30 percent more likely to be uninsured in 2010 than in 2000.

The share of uninsured workers increased more during the full business cycle from 2000–07 (increasing 2.5 percentage points) than it did in the recession that began in December 2007 (increasing 2.0 percentage points from 2007–10). The total increase in the share of workers uninsured from 2000–10 was 4.5 percentage points, representing 6.7 million more uninsured workers.

Older workers (age 55–64) are more likely to have coverage than any other age group, with only 12.5 percent lacking insurance—while more than one-fourth of younger workers (age 18–34) are uninsured. Working men are more likely to be uninsured than working women (22.0 percent versus 16.7 percent). White non-Hispanic workers are far less likely to be uninsured than black and Hispanic workers, with 14.0 percent of whites lacking insurance, compared with 25.2 percent for blacks and 40.1 percent for Hispanics. Native- and foreign-born workers alike saw increases in uninsurance rates over 2000–10, though foreign-born workers are uninsured at more than twice the rate of native-born workers.

Nearly half of workers without a high school degree are uninsured, compared with about one-fourth of high school graduates and one-tenth of college graduates. While these levels illustrate clear inequities, the gap in coverage rates among workers of different education levels grew substantially over 2000–10. Workers without a high school degree or with only a high school education experienced larger increases in their ranks of uninsured (rising 8.5 and 8.0 percentage points, respectively) than did those with a college degree or post-college education (3.2 and 1.8 percentage points, respectively).

Nearly one-third of all workers in the lowest 40 percent of the wage distribution are uninsured, compared with less than one-eleventh of workers in the top 40 percent. About 4.0 million additional workers in the bottom 40 percent were uninsured in 2010 compared with 2000, while only 1.4 million additional workers in the top 40 percent were uninsured over the same period. At the extremes, 37.7 percent of workers in the lowest fifth were uninsured, while only 7.1 percent of workers in the highest fifth were. While public insurance is designed to cover children who lose access to employer-sponsored insurance, it is clear that workers generally do not enjoy the same safety net.

Part-time workers are more likely to be uninsured than full-time workers, and that gap in access to coverage has grown over time. Part-time workers were 28 percent more likely to be uninsured in 2000, but 58 percent more likely to be uninsured in 2010; over 2000–10, lack of insurance coverage grew 3.2 percentage points among full-time workers and 9.3 percentage points among part-time workers. The introduction of insurance exchanges and their accompanying subsidies as part of PPACA in 2014 should provide an alternative to workers, particularly part-time workers, who do not have health insurance through their job.

Table 5 examines the uninsured workforce side-by-side with the workforce as a whole in 2010. Uninsured workers are disproportionately young. Workers age 18–34 make up 36.0 percent of the total workforce yet nearly half of the uninsured workforce. In addition, working men are more likely to be uninsured than working women. Disparities among the working uninsured are stark by race and ethnicity. Whereas Hispanics make up only 14.6 percent of the total workforce, they represent 30.9 percent of the uninsured workforce. A similar trend is found by nativity: While the foreign born account for 15.8 percent of the overall workforce, they represent 28.7 percent of the uninsured workforce.

Insurance coverage among workers rises consistently with increased educational attainment. Workers with a high school education or less represent 36.7 percent of the workforce, yet they make up 60.1 percent of uninsured workers. Those with a college degree or higher represent nearly one-third of the workforce, yet less than one-seventh of those uninsured.

The starkest disparities occur at different points in the wage distribution. When the workforce is equally divided by wage into fifths (see Gould [2010] for methodology), it is clear that those at the bottom end of the distribution are far more likely to be uninsured than those at the top. Workers in the bottom two-fifths by definition represent 40 percent of the workforce, but represent a full two-thirds of the uninsured. By contrast, the top two-fifths, again 40 percent of workers, represent only 17.4 percent of the uninsured.

Declining ESI coverage for children

Second to young adults (age 18–24), children under 18 have the lowest rates of ESI coverage of the under-65 population, at 54.8 percent (Table 6). Coverage shares for children fell every year from 2000–10, resulting in an overall decline of 11.9 percentage points. Nearly 7.2 million fewer children had ESI in 2010 than in 2000, without even taking into account the growth of the under-18 population throughout this period. As many as 8.9 million more children would have had ESI in 2010 if the coverage rate had remained at the 2000 level.

As with the under-65 population as a whole, there are stark disparities in ESI coverage for children. White non-Hispanic children have coverage rates nearly as high as the rate for overall workers (67.2 percent), and almost double the rate of Hispanic children (35.3 percent). Black children experienced the largest losses from 2007 to 2010 (7.8 percentage points) and from 2000 to 2010 (14.1 percentage points). Native-born children experienced greater losses than did foreign-born children over 2000–10, yet their coverage rates are still far higher (55.5 percent versus 37.3 percent).

Children’s coverage is highly correlated with the education of the family head. Less than half of children of high school educated parents have ESI, compared with about four-fifths of children with college educated parents. Only one in five children of a family head with less than a high school education has ESI. Similarly, access to ESI is closely tied to family income (Figure D). While children across the economic spectrum experienced losses in coverage over 2000–10, disparities have widened. The gap between the top fifth and bottom fifth grew 7.7 percentage points over the period, while the gap between the second and fourth fifths grew by 15.8 percentage points.

Publicly provided health insurance stemmed larger losses in overall coverage

While losses in ESI from 2000 to 2010 were greater among children than among non-elderly adults, the share of children without any coverage actually fell, as shown in Figure E. The uninsured rate for children fell 1.0 percentage point, while the share of uninsured non-elderly adults rose 5.4 percentage points from 2000 to 2010. Given that the share of children and non-elderly adults covered by privately purchased, or non-group, insurance coverage was relatively flat over this period (not shown), the differences in the overall coverage rates are due to differences in the incidence of public insurance for these groups.

The share of children with public coverage grew 13.7 percentage points from 2000 to 2010, compared with an increase of only 5.0 percentage points for the non-elderly adult population. While both increases lessened the impact of ESI losses on overall coverage rates, only the increase in public coverage for children was large enough to be fully offsetting. Children have greater access to public insurance through CHIP, but eligibility for public insurance for non-elderly adults is mostly limited to Medicaid or Medicare (e.g., for the disabled).

Some claim that the losses in ESI were actually driven by increases in public coverage eligibility or generosity, a phenomenon known as “crowd-out.” However, given the economic downturn that began in 2007, it is likely that an increasing number of children became eligible for public insurance rather than public coverage replacing private coverage. That ESI coverage rates for adults fell without the same counterbalancing rise in public coverage further reinforces this fact. Regardless of the cause, it is clear that if not for public insurance, the overall coverage rate among children would have fallen.

ESI across the states

The non-elderly population across the country relies on ESI as the primary form of coverage; however, the incidence of coverage varies widely from state to state. Table 7 compares ESI coverage rates for the entire under-65 population across states between 2000/01 and 2009/10.

New Hampshire has the highest rate of ESI coverage among the under-65 population, at 73.0 percent in 2009/10. It is followed by Connecticut (70.8 percent), Massachusetts (70.2 percent), Utah (69.3 percent), and Maryland (68.8 percent). In contrast, less than half of both New Mexico’s and Mississippi’s non-elderly population have ESI, at 48.6 percent and 48.4 percent, respectively.

Across the country, on average, ESI coverage for the under-65 population fell 9.4 percentage points from 2000/01 to 2009/10. Fifteen states experienced losses in excess of 10 percentage points over the period. The largest declines in coverage occurred in Indiana, South Carolina, Michigan, Georgia, Nevada, Delaware, and Mississippi, each with losses of at least 12 percentage points. Forty-eight states plus the District of Columbia had statistically significant losses in coverage rates for their under-65 population, while no state had a rise in the share of its under-65 population with ESI coverage over that period.5

In terms of ESI coverage for workers age 18 to 64, Massachusetts and Hawaii have the highest coverage rates, at 79.2 percent and 78.2 percent, respectively (Table 8). This is not surprising, as both states have mandates requiring employers provide at least minimal insurance coverage to their workers. The lowest rate of worker coverage is in New Mexico, at 59.9 percent, followed by Montana (60.0 percent) and Texas (61.0 percent). The largest declines in job-based coverage among workers occurred in Georgia, South Carolina, and Michigan, each with losses in excess of 10.0 percentage points, far above the national average decrease of 6.8 percentage points.

When looking at ESI coverage for children, New Hampshire again leads the country, with a coverage rate of 74.1 percent (Table 9). Connecticut and Wisconsin follow, at 69.6 percent and 69.4 percent, respectively. At the other end of the spectrum, nine states and Washington, D.C., have ESI coverage rates among children of less than 50 percent. Mississippi has the lowest rate of coverage for children, at 41.3 percent, followed by New Mexico (43.3 percent) and Washington, D.C. (43.5 percent).

As is the case nationwide, losses in coverage across the states are greatest among children. Twenty-five states and Washington, D.C., experienced declines from 2000/01 to 2009/10 in excess of 10 percentage points. Indiana experienced the greatest losses in children’s ESI coverage (17.0 percentage points), followed by Illinois (15.6 percentage points) and Iowa (14.6 percentage points). Overall, 45 states and Washington, D.C., had statistically significant declines in children’s ESI coverage rates. No state had a statistically significant increase.

Health reform and coverage in 2010

In 2010, several elements of the Patient Protection and Affordable Care Act took effect. These provisions include reforms to the insurance market, tax credits to help small businesses provide insurance to their workers, and a stipulation allowing young adults up to age 26 to stay on or join their parents’ employer-sponsored health insurance policy. Last year, several analyses touted the success of health reform with regards to the young adult provision, noting that the uninsured rate for 18-to-24-year-olds fell between 2009 and 2010—and that this was the only age group with a statistically significant decline in its uninsurance rate (Collins, Garber, and Davis 2011; Sebelius 2011). While this evidence may not appear sufficient to prove that health reform is principally responsible for this decline, it is relatively easy to demonstrate that this is likely the case.

Figure F compares changes in the employment rates and the rate of employer-sponsored health insurance for various age groups between 2008 and 2009. Employment rates fell for each group, as did employer-sponsored health insurance coverage rates. This is not surprising given that most people find health insurance on the job; however, declines, though smaller, were also found in coverage as a dependent.

Figure G compares these same changes in employment rates and health insurance rates for various age groups between 2009 and 2010. Young adults (age 19–25) still did not fare well in the job market in this latter period.6 In fact, their employment rate actually fell further than that of any other age group. Given the close relationship between labor market outcomes and employer-sponsored insurance, we would expect declines in coverage for all groups. What we see instead is that employer-sponsored health insurance actually increased among young adults, while it fell for all other groups. It rose particularly dramatically among young adults who had ESI as a dependent.

If the simple relationship between employment and overall ESI in 2008–09 had held for 2009–10, there would be a 1.0 percentage-point drop in insurance rates for young adults in 2010.7 However, in 2010 overall employer-sponsored insurance rates actually rose by 0.4 percentage point. Given this difference, roughly 414,000 young adults may have obtained coverage in 2010 because of the health reform provision allowing them to remain or join their parents’ insurance policy. And, as Figure G demonstrates, this coverage increase came in the form of dependent coverage, the very type of coverage the provision should have affected.

While this provision appears to improve the low coverage rates for 19-to-25-year-olds, coverage for young adults through this avenue is dependent on parental coverage, which fell from 2000–10 and is more likely to be secured by those with higher incomes. Thus, in a struggling economy, fewer young adults will be able to secure coverage through their parents. Furthermore, young adults whose parents do not have the advantage of ESI (disproportionately non-whites and/or those with less education and/or lower incomes) will not be able to take advantage of this provision.

Conclusion

Employer-sponsored health insurance is increasingly failing American families. If the coverage rate had not fallen 10.6 percentage points as it did from 2000 to 2010, as many as 28 million more people under age 65 would have had ESI in 2010. Public insurance, primarily in the form of Medicaid and CHIP, has been working to counteract this trend. However, many Americans, particularly those of working age, are falling through the cracks.

In the future, elements of the Patient Protection and Affordable Care Act—particularly the provisions establishing health insurance exchanges and the accompanying subsidies, which will come into effect in 2014—will make it easier and more affordable for Americans to secure and maintain health insurance coverage. However, the continued weak labor market will likely lead to further losses in employer-sponsored insurance coverage before major relief from health reform is realized.

—The author thanks Jin Dai, Nicholas Finio, and Natalie Sabadish for their valuable research assistance. EPI is grateful to the Open Society Foundations, Atlantic Philanthropies, and Ford Foundation for providing support for the research and publication of this report.

Endnotes

1. The results under the education heading assign each child the education level of their family head, as children under 18 rarely complete their education by that time.

2. See Gould (2010) for a discussion of wage quintile analysis and balancing fifths.

3. In this section, to qualify as employer-sponsored health insurance coverage, workers must receive employer-sponsored health insurance through their own job, and employers must pay at least part of their insurance premiums.

4. Changes in industry classification make it impossible to compare 2010 with years earlier than 2002.

5. The second set of numbers in Table 7 displays the number of people with ESI in both sets of years, including the difference between the years. Declining coverage rates accompanied by increases in the number insured reflect the fact that the increases did not keep pace with population growth. Another comparison would be to compare the absolute level of people with coverage in the latter period to the level that would have occurred had the rate remained the same as in the earlier period (i.e., multiplying the population in the latter period by the coverage rate in the earlier period). The same fundamental point also applies to Table 8 and Table 9.

6. Elsewhere in this paper, “young adults” are defined as 18-to-24-year-olds. In this section on health reform, young adults refer to 19-to-25-year-olds to best capture those most directly affected by this particular provision.

7. To predict the ESI rate for 19-to-25-year-olds in 2010 in the absence of health reform, apply the ratio of the change in the employment rate to the change in the ESI rate over 2008–09, to the employment rate change in 2009–10.

References

Bivens, Josh, Elise Gould, and Alexander Hertel-Fernandez. 2009. The Health Care Free Ride: Reform Should Help Distribute Responsibility for Coverage More Evenly Between Industries. EPI Issue Brief #259. Washington, D.C.: EPI.

Collins, Sara R., Tracy Garber, and Karen Davis. 2011. “Number of Uninsured in United States Grows to 49.9 Million; Young Adults Already Benefitting from the Affordable Care Act.” The Commonwealth Fund Blog, September 13. http://www.commonwealthfund.org/Blog/2011/Sep/Number-of-Uninsured-in-United-States-Grows.aspx

Gould, Elise. 2010. “Employer-Sponsored Insurance Erosion Continues in 2008 and Is Expected to Worsen.” International Journal of Health Services, vol. 40, no. 4, pp. 743–776.

Gould, Elise, and Alexander Hertel-Fernandez. 2009. Senate Health Bill Scores Bill for Small Business: Bill Would Help Provide Affordable, Stable Health Coverage. EPI Issue Brief #268. Washington, D.C.: EPI.

Sebelius, Kathleen. 2011. “Affordable Care Act in Action: Fewer Uninsured Young Adults in America.” The White House Blog, September 13. http://www.whitehouse.gov/blog/2011/09/13/affordable-care-act-action-fewer-uninsured-young-adults-america

U.S. Census Bureau. Various years. Current Population Survey, Annual Social and Economic (ASEC) Supplement, conducted by the Bureau of the Census for the Bureau of Labor Statistics [machine-readable data file]. http://www.bls.census.gov/cps_ftp.html#cpsmarch

U.S. Census Bureau. 2009–2010a. Current Population Survey, conducted by the Bureau of the Census for the Bureau of Labor Statistics [machine-readable data file]. http://www.bls.census.gov/cps_ftp.html#cpsbasic

U.S. Census Bureau. 2009–2010b. Current Population Survey, Annual Social and Economic (ASEC) Supplement, conducted by the Bureau of the Census for the Bureau of Labor Statistics [machine-readable data file]. http://www.bls.census.gov/cps_ftp.html#cpsmarch

U.S. Census Bureau. 2010–2011a. Current Population Survey, conducted by the Bureau of the Census for the Bureau of Labor Statistics [machine-readable data file]. http://www.bls.census.gov/cps_ftp.html#cpsbasic

U.S. Census Bureau. 2010–2011b. Current Population Survey, Annual Social and Economic (ASEC) Supplement, conducted by the Bureau of the Census for the Bureau of Labor Statistics [machine-readable data file]. http://www.bls.census.gov/cps_ftp.html#cpsmarch