Long-Term Trends in Employer-Based Coverage

Employer-based health insurance is the most important source of health coverage for the nonelderly, covering about 58% of this population in 2017. The workplace has long been a significant source of coverage for those in working families, although its importance has been declining for a number of years, particularly among those in lower and moderate-income households.  This brief presents data from the National Health Interview Survey to examine trends in the share of nonelderly people who receive and are offered coverage through a job. A description of our methods is available below.

We find that the share of the nonelderly with employer-sponsored health insurance fell by almost 9 percentage points between 1999 and 2017, with larger percentage reductions generally occurring for income groups below 400% of poverty than above. Overall, if coverage rates had stayed at the 1999 level (67.3%), almost 24 million additional people would have been covered by employer-sponsored plans in 2017. Although the long-term trend shows a decline in the rate of employer-sponsored coverage, the share of people with job-based coverage did rise modestly between 2013 and 2017. 

The share of people with employer-sponsored insurance has declined over the past 20 years

The share of the nonelderly covered by employer-based insurance fell from 67.3% in 1999 to 55.6% in 2011. Coverage in employer-based plans recovered somewhat, to 58.4% in 2017, as the economy and employment rates have improved. Nonetheless, if the coverage rate for employer-sponsored insurance was the same in 2017 as it was in 1999 (67.3%), almost 24 million (or 23.8 million) additional people would be covered through an employer plan in 2017. 

The total number of people covered by employer-based insurance is roughly the same in 1999 and 2017

The number of nonelderly people with employer-based insurance has fluctuated within a relatively narrow band over time, even as the nonelderly population grew by 31.5 million people. The number of people covered by employer-sponsored plans fell during the economic slowdown after 2008, and recovered after 2014, reaching about the same number as in 1999.  

The prevalence of employer-based insurance varies significantly with household income

Only 14.5% of nonelderly people with incomes below poverty were covered through a job in 2017, compared to 84.4% with incomes over 400% of the poverty level. 

Lower income workers are less likely to have employer-sponsored coverage

Even among full-time workers, nonelderly people with lower incomes were much less likely to be covered by employer-sponsored coverage in 2017 than those with higher incomes. 

The share of people with employer-based coverage fell from 1999 to 2017 in all income groups

The percentage reductions were smaller for people with incomes above 400% of poverty than for those in lower income groups. Although the coverage share increased between 2013 and 2017 only for those under poverty, the overall share of the nonelderly covered by employer-sponsored plans increased by 2 percentage points (from 56.3% to 58.4%). This occurred because the share of population with incomes over 400% of poverty, where the share of people with employer-based plans is highest, increased significantly over this period. About 7 million more nonelderly people were in employer-sponsored plans in 2017 as compared to 2013.

One potential reason for the drop in employer-sponsored coverage is the drop in percentage of workers offered coverage at their jobs

During the period from 1999 to 2017, the percentage of workers offered employer-sponsored coverage dropped in each income category, with the largest percentage drops among workers with incomes between poverty and 400 percent of poverty. 

The offer rates increased between 2013 and 2017 for workers with incomes below poverty, which may be due to the employer requirement to offer coverage under the Affordable Care Act, although offer rates were largely flat over this period in other income groups.


Although the long-term trend is clearly and significantly downward, the coverage rate for employer-sponsored coverage did rise modestly between 2013 and 2017. There are several factors that may help explain the recent trend. One is that the economy improved steadily from the deep downturn in 2008 and 2009, with both incomes and employment increasing, particularly over the last few years.  The number of people with full-time employment increased by over 9.5 million between 2013 and 2017. The share of the nonelderly with incomes above 400% of poverty, where employer-sponsored health insurance is most prevalent, also rose significantly during that period.

If coverage rates had stayed at the 1999 level (67.3%), almost 24 million additional people would have been covered by employer-sponsored plans in 2017 Click To Tweet

Another potential factor was the requirement under the Affordable Care Act (ACA) that people have health insurance or face a tax penalty (”individual mandate”). This may have encouraged some workers to take up coverage at work for themselves or their family members that they may have declined without the individual mandate in place. The ACA also required that employers with more than 50 employees offer coverage to full-time workers; this requirement may have increased the share of workers offered coverage at work, particularly among lower-wage employees. The tax penalty for not having insurance was essentially ended (set to $0) in the tax bill for years after 2018, which could affect coverage going forward. In a recent survey, almost a quarter of employers with 200 or more workers said that they thought fewer employees or dependents would participate in their plans due to the repeal of the individual mandate.

A complicating factor was the implementation of new coverage rules, premium tax credits and cost-sharing subsidies in the individual health insurance market beginning in 2014 as required by the ACA. These tax credits, which are available to lower income people not offered coverage at work, may have led some smaller employers to stop offering coverage (employers with 50 or more workers are required by the ACA to offer coverage to full-time workers) because their employees might have less expensive options in the individual market. Individual market premiums also were quite low in many places in 2014 and 2015, which may have convinced some small employers that the individual market was a more cost-effective option for their employees. More recently, however, individual market premiums have risen fairly rapidly and family incomes have improved, which has likely reduced the attractiveness of individual coverage as an alternative to employer-based plans. Overall, the number of people with individual coverage fell by about 4 million people between 2016 and 2018, the large majority of which were people not receiving premium tax credits. Enrollment in the ACA marketplace has continued to decline in 2019 (though information is not yet available on enrollment in the individual market outside the marketplace). Given the continuing economic improvement, it is quite possible that some of these people moved into employer-based plans.

Taking a longer-term view, the increases over the last few years brought the number of people covered by employer-sponsored health insurance back to about the same number of people as in 1999, despite nonelderly population growth of 31.5 million people over the period. Coverage in employer-based plans, while still the largest single source of coverage among the nonelderly, has become less important overall and particularly among people with incomes below 400% of poverty. Despite a recent, favorable period, there is no evidence to suggest that the share of nonelderly people covered by employer-based plans will ever approach past levels. 


The National Health Interview Survey (NHIS) is a national probability survey of American Households sponsored annually by the U.S. Census Bureau and the Center for Disease Control and Prevention (CDC).  Although NHIS was started in 1957, the survey is redesigned periodically and had a major redesign in 1997.  We used survey data back to 1999, which was the first year after the redesign that included all of the insurance categories we typically use (in 1999, the survey asks about coverage through Children’s Health Insurance Program and also includes the variable “NOTCOV”).  For more information on NHIS, please see http://www.cdc.gov/nchs/nhis.htm

The analysis focuses on people below the age of 65 covered by employer-sponsored insurance (ESI).  Some individuals have multiple types of coverage; individuals who are covered by Medicaid or Medicare Part B in addition to an employer plan are not included.  Individuals who are covered by Medicare Part A and an employer plan are assumed to have ESI as their primary coverage (less than 1% in 2017).  Full-time workers work at least 30 hours a week.

Households at more than four times the FPL earn different incomes depending on a family’s composition; for example, in 2017, a single nonelderly adult would earn about $51,000 and a family of four including two children would earn about $99,000.[i]

[i] NHIS uses the Census Bureau’s federal poverty thresholds, while HHS and this portion of the brief use federal poverty guidelines to determine eligibility for some public assistance programs.