Has Real Median Household Income Fallen Since 2000?

During the presidential campaign this year and throughout the current economic uncertainty, many in the media and political candidates have argued that real median household income has fallen since 2000 (often a reference point aimed at when Pres. Clinton left the White House).

That claim is made using the CPI-U measure of price levels combined with Census data from the Current Population Survey, which is probably the most popular microdata set among geeks like myself. But what nobody will ever tell you is that income is a nebulous concept…and I mean nebulous. No political candidate understands it. Nobody in the media understands it. But it is a truly difficult question to answer that has been debated for hundreds of years among economists and philosophers.

Just to give you an idea of the uncertainty over what truly is income: The income measure used by the Congressional Budget Office when it does its annual study of federal effective tax rates is different from the Census Bureau definition that is often cited. The Tax Policy Center definition is different from both of them. Adjusted gross income, the measured used by the Joint Committee on Taxation and the measured used in calculating your income tax every year, is narrower than virtually all other definitions of income. Personal income, used by BEA, is different than all of the others mentioned, and is different from national income as defined by BEA. Treaury’s measure of income is different (was changed in early 2000s for various reasons). To sum it up, nobody can agree on what is income. Haig and Simons agreed, but Fisher disagreed with them, as did the Supreme Court.

These disagreements over what is income come even before you get into the complicated question of real income and the problem of making comparisons over time when adjusting for inflation given the fact that the bundle of goods in a later period may have a higher dollar pricetag but the products are far superior in the current period relative to the previous period.

To give you an illustration of this problem of defining income, the table below show how the change in real incomes from 2000 to 2006 looks under 15 alternative definitions of income derived by the Census Bureau using CPS data. As you will see, some could say that real household income has actually risen by 3 percent and be technically correct, while others could say that it has fallen by 3 percent and be technically correct. Definition 1 (money income) is that most commonly used by those who argue that real incomes have fallen. And the typical response from others to that argument is that such a measure does not include many benefits provided to workers such as pensions and employer-provided health insurance.