The high proportion of people who get their health insurance through their jobs is one of the most distinctive features of the U.S. health care system. According to the Census Bureau, 56 percent of the population had employer-sponsored health insurance (ESHI) as of 2017. ESHI accounts for 83 percent of all of those with private insurance of any kind. People whose health insurance is tied to their jobs far outnumber the 38 percent of the population served by government insurance of all kinds.

What is more, most people on ESHI appear to be satisfied with the coverage they get. A survey by America’s Health Insurance Plans (AHIP), an insurance industry group, found that 71 percent of respondents were satisfied with their ESHI plans, compared with just 19 percent who were not satisfied. An independent survey by Gallup came up with similar results, finding 69 percent of people on employer-sponsored plans to be satisfied. A study by the Employee Benefit Research Institute found that 50 percent of workers were extremely or very satisfied with their own ESHI plans, with another 39 percent somewhat satisfied.

How should would-be reformers interpret these numbers? Clearly, one possible reaction is, “If it ain’t broke, don’t fix it.” The Affordable Care Act took that approach. Rather than trying to replace ESHI, it made it mandatory for employers with 50 or more workers.

Despite its popularity, though, serious health economists tell us that ESHI is “broke,” after all. No comprehensive reform can succeed unless it is phased out. This commentary examines three of ESHI’s biggest problems: job lock, which reduces labor mobility for ESHI beneficiaries; the fundamental inequity of the way the benefits of ESHI largely accrue to the highest -paid workers; and the increased fragmentation of health care finance inherent in a system administered by thousands of separate employers. We conclude with a plan for phasing out ESHI in a way that can fix these problems while minimizing the disruption for workers who are satisfied with their current coverage.

Job lock

The term job lock refers to the tendency of employer-sponsored health insurance to discourage people from changing jobs; from starting a business of their own; or from reducing their hours to care for family members or move gradually toward retirement. Job lock undermines labor market mobility, makes it harder to match workers to the most suitable jobs, and cuts labor productivity.

Anecdotal evidence of job lock abounds. Almost everyone seems to have a friend or relative who has taken a job that is otherwise unsuitable, or not left such a job, simply because it is the only way to get health coverage.

If you yourself do not know such a person,  I recommend a recent New York Times Op-Ed by former reporter Kurt Eichenwald. Eichenwald suffers from a severe form of epilepsy for which medication alone costs $50,000 a year. His op-ed vividly details 40 years of struggles to secure and keep health insurance: small employers who refused to hire him because he would send the company premium through the roof; frightening gaps in coverage when he had to appeal to his parents to cover costly ER visits; a humiliating incident in which he had to beg for an entry-level job far below his qualifications just to maintain coverage.

Eichenwald’s experience is by no means exceptional. In the AHIP survey cited above, 46 percent of respondents listed health benefits as an important factor in deciding to work for their current employer. That included 9 percent who said health coverage was the decisive factor in taking the job. An even greater number, 56 percent, reported that health insurance had an impact on their decision to stay in their current job.

There is a large academic literature on the extent of job lock, well summarized in a 2015 literature survey by Dean Baker, published by the AARP Public Policy Institute. Baker notes that there is wide agreement that people with ESHI are less likely to change jobs, become self-employed, retire early, or reduce hours of work. At the same time, there are many other factors that influence labor mobility. Still, Baker concludes that even when those complicating factors are accounted for, the preponderance of evidence shows that job lock is a reality.


A second criticism of ESHI is its inequity. Workers with high paid jobs get most of the benefits, while those with lower pay or part-time jobs get little help.

Differences in tax rates are one source of that inequity. Consider how differences in tax rates would affect the value of a typical ESHI policy. According to data from the Kaiser Family Foundation (KFF), such a policy has a total cost of just under $20,000 as of 2018, of which employers pay about $14,000 on average. Suppose you are a head of household earning $60,000 a year, putting you in a 25 percent federal tax bracket. In that case, having your employer pay $14,000 of your insurance premium, rather than getting that much extra in cash and paying the premium yourself, saves you $3,500 in taxes. If you are a top executive in a 40 percent tax bracket, the tax deductibility of the insurance is worth $5,600.

However, according to the Tax Policy Center, some 44 percent of Americans will pay no income tax at all in 2018. Sixty percent of the nonpayers work. Even if they get ESHI, it gives them no tax benefit at all. They would be no worse off if health benefits were not deductible and if employers added the cost of their insurance to their cash pay instead.

A second factor adding to the inequity of ESHI is that low-wage workers, by and large, are not even offered the option of health benefits. The following chart, provided by the Social Security Administration, shows that only about a third of workers in the lowest fifth of the wage distribution are offered health benefits and that fewer than 20 percent accept those offers. In contrast, more than 80 percent of those in the top fifth of the wage distribution are offered health benefits and accept them.

Robert Kaestner and Darren Lubotsky, economists at the University of Illinois, Chicago, provide an estimate of the overall inequality of ESHI based on the combined effects of differences in tax rates and differences in offer and acceptance rates. As shown in the next chart, taken from their study, workers in the bottom fifth of the family income distribution get annual benefits of less than $500 from ESHI, while those in the top fifth get benefits averaging $4,500. What is more, the value of health benefits to well-paid workers grew substantially over the period shown in the chart, while the value for the lowest paid workers decreased slightly.


A third problem to which ESHI contributes is fragmentation of health care finance. As a Policy Brief from Brandeis University put it,

The US health care delivery system is expensive, fragmented, highly decentralized, and poorly organized . . . made up of a fragmented network of public and private financing, health care delivery, and quality assurance structures. There is no single national entity or set of policies guiding the health care system. States divide their responsibilities among multiple agencies, and providers who practice in the same community and care for the same patients often work independently from one another. The US health system is the most expensive system in the world and yet health outcomes and quality are no better and often worse than in most developed nations.

Fragmentation is a problem for small employers, who have little bargaining power in purchasing group policies from insurers, but also for larger employers. Many larger employers try to save on health benefit costs by self-insuring. According to Collective Health, a company that advises employers on their ESHI programs, 79 percent of companies with 200 or more employees self-insured as of 2017, up from 60 percent in 1999.

The problem is, companies that self-insure don’t always do a good job of it. Ali Diab, CEO and founder of Collective Health, puts it this way in a post on the company’s blog:

Contrasted with other areas of concentrated employer spending, like payroll, or sales and marketing, the distinct absence of a modern, technology-driven approach to managing this effort [ESHI] stands out. To understand why employers depend on these complex arrangements, we need look no further than the antiquated, error-prone administration systems that power the “back office” of today’s healthcare industry.

Recently three corporate giants, Amazon, JPMorgan Chase and Berkshire Hathaway, have joined forces to form a combined health care enterprise to manage benefits for their hundreds of thousands of employees. The stock prices of traditional insurers fell when news of the consortium was announced, but many analysts are skeptical of how well it will work. Leemore Dafny, a professor at Harvard Business School, told reporters from the New York Times, “Just because you know an industry is underperforming and you have a lot of money doesn’t mean you have a successful strategy.” Dafny said she was excited to see such serious players take on the problems of ESHI, but noted there were numerous examples of outsiders trying, and failing, to succeed in the health care system.

When it comes down to hard bargaining, health care providers, including big insurers, hospitals, and drug companies, are less fragmented than employers. Furthermore, health care is what they know best. For employers, whose main expertise lies in manufacturing, customer service, finance or other areas, health care is only a sideline. Given the structure of the system, providers will always come out ahead, driving up costs for workers and their families, who are the ultimate health care consumers.

Making the transition

As we noted at the outset, ESHI, despite its problems, is popular among its beneficiaries. Almost sixty percent of workers get some form of employee health benefits, and half to three-quarters of workers who receive ESHI are satisfied with what they get. Would-be reformers should take heed. A great many people would rise up in protest against any attempt to phase out ESHI without putting something better in its place.

Still, popular or not, the transition away from ESHI is already underway. The following chart from a 2018 KFF Employer Benefits Survey shows that the percentage of firms offering health benefits has been declining for years, most rapidly for small firms.

At the same time, deductibles for ESHI plans have been steadily increasing, as shown in this chart from the same source:

Furthermore, as this third KFF chart shows, workers have been required to pay ever-higher premiums for ESHI coverage when they do get it:

The fact that ESHI is already on the decline makes it all the more urgent to find a replacement. The alternative that I favor is universal catastrophic coverage (UCC).

Under UCC, everyone would automatically be issued a policy similar to a conventional Medicare or Medicare Advantage policy, with no premium required, but with an income-based deductible. For families below a designated low-income threshold, for whom almost any medical expenses would be unaffordable, the deductible would be set at zero. For others, it would be set as a percentage of eligible income, that is, of the amount by which household income exceeds the low-income threshold. For everyone, regardless of income, deductibles would be waived on a package of basic primary and preventive services, such as prenatal care and childhood vaccinations.

For the sake of illustration, suppose the low-income threshold were set at the federal poverty level, currently about $25,000 for a family of four, and the deductible at 10 percent of eligible income. In that case, a four-member family with income of $25,000 or less would have first-dollar coverage with no deductible. A middle-class family with $75,000 of income would have a deductible of $5,000, comparable to the out-of-pocket costs of an ACA silver plan (but without the monthly premium). A wealthy family with an annual income of $1 million would have a deductible of $97,500, which they might decide to cover, in part, with some form of supplemental insurance. (See this discussion for more detailed discussion of UCC finances.)

Now compare how various families would fare under such a UCC plan compared to an ESHI policy that required a premium of $5,500 (roughly the current average) and a deductible of $1,000 (more than half of all ESHI policies currently have deductibles of that much or more). Under ESHI, the family’s maximum annual health care spending would be $6,500. That is the same as the deductible that would apply to a family with $90,000 of income under our illustrative version of UCC.

About 70 percent of U.S. families have incomes of $90,000 or less. Based on the Social Security Administration chart shown above, about 75 percent of families with incomes of $90,000 or more have ESHI. It follows that about 22 percent of all families in our illustration would have a UCC deductible higher than their $6,500 average maximum health care expense under ESHI. However, not all such families would actually be worse off under UCC. Instead, since UCC has a higher deductible than ESHI, but no premium, a family would be worse off under UCC only if their actual annual health care expenditures were over $6,500.

Data on the distribution of personal health care spending suggests that well over half of all four-member families have annual personal health care spending of less than $6,500 each year. Accordingly, it appears that no more than about 10 percent of all families would actually have to spend more under UCC than under ESHI. Those would be families that are now covered by ESHI, that have incomes of $90,000 or more, and that have health care spending higher than the median.

But even if such families are only 10 percent of the population, we don’t want to leave them entirely in the lurch. They ought to be offered something to ease the transition from ESHI to UCC. My suggestion is that their employers should be allowed to continue some scaled-down form of tax-deductible health benefits if they want to do so – for the sake of discussion, let’s call it ESHI-Lite. ESHI-Lite would be far less expensive than today’s version because it would only have to cover the personal health expenditures of employees up to the limit of their UCC deductibles.

I suggest also that ESHI-Lite should come with an opt-out clause for the employee. Suppose, for example, that your employer’s cost of ESHI-Lite coverage was $500 per year. Under the opt-out clause, you could waive ESHI-Lite coverage in exchange for a $500 cash bonus added to your annual salary – a bonus that, like today’s ESHI benefits, would continue to be tax deductible during a transition period.

Most workers with low and moderate salaries, if reasonably healthy, would probably take the opt out. If so, then after the introduction of UCC, ESHI would continue to exist mainly as a sort of “Cadillac” benefit for high-paid executives, whose seven-figure incomes would expose them to six-figure UCC deductibles. The tax deductibility of these ESHI-Lite policies (and of opt-out bonuses) could be phased out over a ten-year period. As the deductibility of ESHI-Lite diminished, more and more employees at all income levels would prefer to bargain for straight salary increases rather than health benefits. Over time, then, even the residual form of ESHI would fade away.


If employer-sponsored health insurance were a carefully crafted system established with some clear purpose in mind, we might think twice before recommending its abolition, but it is not. Instead, people who have tried to trace its origins, like Indiana University’s Aaron Carroll, portray ESHI as an accident of history. Job-linked health benefits first became widespread during World War II when American firms faced both a labor shortage and a wage freeze. Desperate to attract employees, the story goes, they started giving out benefits like health insurance instead of cash raises. The IRS boosted the popularity of ESHI by declaring such benefits to be nontaxable. When President Truman’s attempts to establish a national health care system failed after the war, ESHI became a central element of a complex health care system whose many disparate parts have never fit together well.

We can do better than that. It is time to stop putting up with the job lock, the inequities, and the fragmentation that characterize ESHI. Universal catastrophic coverage has the potential to be the workable and affordable replacement we are looking for. The time to start the transition is now.