TCS Daily

Middle Man Mess

By Arnold Kling - August 24, 2004 12:00 AM

"A relentless rise in the cost of employee health insurance has become a significant factor in the employment slump, as the labor market adds only a trickle of new jobs each month despite nearly three years of uninterrupted economic growth.

Government data, industry surveys and interviews with employers big and small indicate that many businesses remain reluctant to hire full-time employees because health insurance, which now costs the nation's employers an average of about $3,000 a year for each worker, has become one of the fastest-growing costs for companies."
-- New York Times

If employers bear the cost of health insurance, then I'm the Easter Bunny. It is fairy-tale economics to believe that "nice" employers give away health insurance, while "mean" employers withhold it. In reality, employers compensate their employees using a combination of cash and non-cash benefits. Workers bear the cost of health insurance.

The right way to think about health insurance is not as something that employers provide but as something that employers sell to their employees. Your employer is an intermediary between you and the insurance company. Step one, your employer company decides how much to pay for your labor. Step two, the company takes some of that pay and makes you buy health insurance with it.

What has happened to employee compensation recently? If you look at the data on total compensation per hour compiled by the Bureau of Labor Statistics, you find that the latest figures available show a 4.6 percent increase for the second quarter of this year over the comparable level a year ago. This is faster than inflation, and it might seem to show that workers are doing really well.

On the other hand, the conventional wisdom has it that wages are growing slowly, if at all. For example, labor economist Alan Krueger reports that "the increase in the wage rate was one percentage point less than the inflation rate." Thus, workers are doing poorly.

The discrepancy is between worker compensation and take-home pay. Non-cash benefits, primarily health insurance, are increasing relatively rapidly. Thus, when you factor out inflation, total compensation has gone up while take-home pay has gone down. The data prove that workers are paying more for health insurance.

The reality is that workers always pay for health insurance. The more money that a company pays for employee health insurance, the less it can afford to give the employee in wages. From the employer's perspective, health insurance is part of the cost of labor, and the laws of economics dictate that those costs will be borne by workers, regardless of whether the employer acts as an intermediary.

Opting Out

In theory, an employer can offer you a choice. The employer could offer you one salary if you choose employer-subsidized health insurance, and a higher salary if you opt out. For example, your salary might be $40,000 if you take health insurance, but $50,000 if you turn down health insurance.

If you opt out of health insurance and take the higher salary, you might choose to spend some of the extra money on buying health insurance directly. In that case, you are engaging in disintermediation, which is a fancy term for cutting out the middle man.

Health insurance disintermediation is widespread in today's economy. The New York Times article quoted above describes one type of disintermediation, which is the use of temporary workers who do not have health benefits. However, there are many other forms of health insurance disintermediation. When someone quits or is laid off and becomes a self-employed contractor, that is disintermediation. When a firm downsizes its work force and sub-contracts to a small firm that does not pay health insurance benefits, that is disintermediation.

In recent years, there has been a striking divergence between household and payroll measures of employment. According to households, there are millions of more jobs than there were when President Bush first took office, while there are fewer jobs listed on employer payrolls. I suspect that disintermediation accounts for some of this discrepancy, as workers effectively opt out of employer health insurance by becoming self-employed.

In theory, we could force everyone who works for a company, even as an independent contractor, to obtain health insurance through that company. However, you cannot force employers to pay more for labor than the value of what workers produce. If you forced every employer to offer generous health insurance benefits even to contractors and temporary employees, then the recipients' take-home pay would have to be reduced and/or hiring would have to be curtailed.

Disintermediation is a free-market response that is good for workers who otherwise would be unemployed or receive low take-home pay. Those workers who would rather not buy so much health insurance at the company store are able to obtain the employment and wages they prefer because of disintermediation.

Adverse Selection

What type of worker is likely to engage in disintermediation? If you would rather receive cash wages than health care coverage, that means that you place a relatively low value on health insurance. You might be young enough to obtain an inexpensive health insurance policy on your own, or you may be willing to gamble on going without health insurance.

On the other hand, if you are concerned about health risks, then you tend to place a high value on health insurance, and you will be willing to accept lower take-home wages in exchange for better health coverage. You will be looking for an employer that offers a lot of health insurance at the company store.

As health care becomes a larger and larger share of consumption, health insurance disintermediation will become a larger and larger factor in employment decisions. Low-risk workers will tend to opt out of the corporate labor market. Corporations will tend to hire from a pool of workers with relatively high health risks. This in turn will raise corporate health costs more, which will lead to even more disintermediation.

The phenomenon of low-risk employees opting out of the corporate health insurance market, leaving employers with a high-risk pool, is what economists call adverse selection. The employees who are willing to give up a lot of cash wages in exchange for health insurance represent an adverse risk population.

Joseph Stiglitz's work on adverse selection and insurance markets earned him a share of the Nobel Prize in 2001. One of his theories is that an insurance market may disappear entirely because of adverse selection. As more low-risk people opt out, the average cost of insuring high-risk people increases, to the point where high-risk people cannot afford insurance. Although we are a long way from that point in health insurance today, it is sobering to consider that disintermediation could be leading in that direction.

Choice, Pooling, and Insurance

The case for buying health insurance where you work has two main components. First, there is tax arbitrage. Unlike you as an individual, your employer can deduct your health insurance premiums from its taxable income. This tax treatment provides an incentive against disintermediation. However, this tax differential is an economic distortion, meaning that other things equal economic analysis says that we would be better off without it. There is no economic reason for health insurance to be taxed differently depending on whether I buy it for myself or through my employer as an intermediary.

The second reason to buy health insurance where you work is risk pooling. To the extent that an employee base represents a random selection of health risks, the insurance company is able to pool risks, which lowers costs.

Disintermediation may be undermining the advantage of risk pooling. Because of adverse selection, a corporate employee base may be less of a random sample of health risks and more of a high-risk sample.

The problem posed by Stiglitz is that free choice and insurance pooling may be incompatible. If multiple types of insurance coverage are available, then low-risk consumers will gravitate toward a different type of coverage than high-risk consumers. The high-risk consumers lose the benefit of being pooled with low-risk consumers.

The Consequences of Disintermediation

The consequences of health insurance disintermediation are generally favorable. It gives workers opportunities for jobs and take-home pay that otherwise would not be available. I would like to see complete disintermediation, so that health insurance is always bought directly by individuals, rather than through employers.

However, disintermediation raises two possible concerns for public policy. One concern is that some workers who opt out of the company store as their source of health insurance are doing away with health insurance altogether. A hard-core libertarian might not care, but others, myself included, would favor some form of mandatory catastrophic coverage.

A second concern is that disintermediation is reducing risk pooling, which may leave some high-risk consumers unable to pool risks with low-risk consumers. High-risk consumers might be protected by catastrophic re-insurance paid for by taxpayers.

If the government response to health insurance disintermediation were limited to a requirement for everyone to obtain high-deductible health insurance coverage and a provision for catastrophic re-insurance, that would be a relatively positive outcome. However, what is likely to happen instead is that health insurance disintermediation is going to lead to demagogic claims that we face a "crisis," and that this "crisis" justifies a much heavier hand of taxpayer funding and regulation as the "solution."


TCS Daily Archives