In Ross Douthat’s latest column, he cited $94,900 as the median income for an American family of four. Critics jeered, with some accusing Ross of being out of touch. As Ross explains, however, this number reflects total compensation:

The median income figure the C.B.O. used (see Table 4-4 on p. 65 of the report) is $94,900 for a family of four, which (as a number of readers have noted) seems much higher than the usual estimates for median income in a four-person household. It turns out that I didn’t catch a crucial footnote in the C.B.O. document: “All income is assumed to be from compensation, which includes employment-based health insurance and the employer’s share of payroll taxes.” That is to say, the $94,900 in income includes the estimated value of the median family’s health care plan as well as their salary, which is not what most people think of when they hear the term “median income.” I apologize for the confusion.

A number of bloggers have been impressively petty in their reactions to Ross’s column, but that’s hardly a surprise.

What I find interesting is that Ross’s American family of four gives us some insight into how we think about American households. When we leave out employment-based health insurance and the employer’s share of payroll taxes — and oh my, imagine how much employers would love to do that — the median income for a family of four is $75,700. As most readers of The Agenda understand, this number translates into different amounts of purchasing power in different regions. This is considerably higher than the median household income as of 2009, which was just below $50,000. While the nominal median income for a family of four in New Jersey or Connecticut is considerably higher than $75,700, that same family would have considerably more disposable income in a low-tax, low-cost metropolitan area in, say, Texas.

As Scott Hodge observed in 2006, the 2002 Panel Study of Income Dynamics found that only 9.3 percent of households are composed of a wife, a husband, and two children, the kind of American family Ross may have had in mind. Moreover, these families are overrepresented in the top quintiles of the income distribution. When we talk about the rich, we’re often talking about fairly conventional two-earner families who’ve experienced a good year.

In a piece on the cap on earnings subject to the Social Security payroll tax, Charles Blahous observed the following:

It’s often said that only the very richest Americans – roughly 6% – would be affected if the [$106,800] cap were raised. This sounds very attractive to those in favor of taxing the other guy, especially if he’s “rich.”
In the real world, however, wages fluctuate. If the cap on taxable wages were raised, it would affect everyone with at least some years of earnings above the cap. This is about 20% of workers.
It’s very important to avoid confusing annual wages with lifetime wages here. Social Security benefits are calculated based on lifetime average wages, a figure that further excludes any income above the cap. If instead we only discuss tax measures in terms of annual wages, it will always (wrongly) appear as though raising taxes affects few people, whereas benefit reductions would affect many.

Note that Blahous is talking about the percentage of workers, not the percentage of households or the percentage of American families of four, for whom the share that has had earnings in excess of $100,000 in at least some years is considerably higher.