A libertarian friend who sells health insurance asked me the following by email:

What are your views on the healthcare “Cadillac” tax?

I started to answer him by email but then realized that my response is probably of more general interest.

So here goes. This is Part I.

I’m sympathetic to what it’s trying to achieve, but I’m against it. Moreover, I’m against it for the same reason that I’m sympathetic to what it’s trying to achieve. I want to reduce distortions caused by the tax code. The Cadillac tax in some ways adds to the distortions.

First, some background that virtually all American health economists are familiar with, and this is from “Free and Healthy at Half the Cost,” Chapter 15 of my book The Joy of Freedom: An Economist’s Odyssey:

Our employer-provided health insurance system is the unintended consequence of government regulation and taxation. During World War II, the federal government had imposed general wage controls on the economy. Thus, an employer who wanted to pay employees higher wages than the controls allowed had to get the government’s permission. In a dynamic economy, especially one in which the government is financing wartime spending by printing money, employers often need to raise wages in order to keep more-productive employees. Thus, employers needed some legal way to pay employees more. The wage controls prevented employers from raising wages, but did not restrict what they could do with benefits. An obvious benefit to provide was health insurance, and that’s what many employers began to do.

Another advantage of giving employees health insurance was that, although the payments were deductible as expenses from the employer’s taxable income, they did not count as taxable income to employees. So, even after the wage controls of World War II ended, employers and employees found it mutually beneficial to keep health insurance as part of employee compensation. The Internal Revenue Service caught on and tried to tax the benefits as if they were taxable income. But, responding to employers’ and employees’ protests, Congress passed a law enshrining the tax-free status of health insurance.

The tax avoidance benefit of health insurance is higher at higher marginal tax rates. If the marginal tax rate is only 20 percent, so that you pay 20 cents in taxes for each additional dollar you earn, then an employee can avoid 20 cents of taxes if his or her employer pays one dollar less in salary or wages and one dollar more in tax-free health insurance. But if the marginal tax rate is 40 percent, shifting that same dollar helps avoid 40 cents of taxes. Throughout the 1950s, 1960s, and 1970s, with one major interruption for the Johnson-Kennedy tax cut of 1964, marginal tax rates rose, making it increasingly attractive for employers to shift compensation from taxable money payments to nontaxable health insurance. Tax rates rose for three main reasons. First, and most important, inflation put people in higher tax brackets because tax brackets were not adjusted for inflation. Second, throughout the post-Korean War period, more and more state governments introduced state income taxes, and most state governments increased income tax rates for a given income level. For example, in 1954, 31 states plus the District of Columbia taxed income from salaries and wages, and their typical tax rate on a middle-income family was about 2 percent. By 1981, 41 states plus the District of Columbia had a tax on income from wages and salaries, and the typical marginal tax rate for a middle-income family was about 4 percent. The third reason that marginal tax rates rose was that the federal government substantially raised the tax rate for Social Security and Medicare. In 1954, the Social Security tax rate was only 4 percent. By 1981, the payroll tax rate was up to 13.3 percent and is now 15.3 percent.

As the senior economist for health policy under Martin Feldstein, Chairman of the President Reagan’s Council of Economic Advisers, I worked on a proposal to dramatically reduce the distortion on the margin. How so? By making the employer’s contribution to the employee’s health insurance plan, above some annual amount, fully taxable to the employee. That way, when an employee is looking at an extra $1,000 in pay or an extra $1,000 in health insurance, he would want the latter only if the extra $1,000 spent on his health insurance is worth at least as much to him as the extra $1,000 in pay (minus tax) is worth to him. So, for example, if the employee’s 7.65% portion of the payroll tax plus the federal marginal tax rate plus the state marginal tax rate sum to 35%, he would want the extra health insurance only if he valued it at at least $650. Otherwise, he would take the pay increase and pocket $650.

I say that our proposal would have dramatically reduced the distortion on the margin. It would have left some distortion. Why? Because the extra annual expenditure on an employee’s health insurance over some base amount would not have been subject to the employer’s 7.65% portion of the payroll tax the way the extra pay would have been. So there would have still been a tax-induced incentive to spend too much on the employee’s health insurance.

That’s Part I. Later today or tomorrow, I’ll give Part II.