The challenges facing America’s middle class are real enough. But the income data used to illustrate those challenges are often deeply flawed and highly misleading.

Take, for example, the claim — currently being peddled by Elizabeth Warren and others — that the bottom 90 percent of U.S. households have, on average, not seen any income growth at all since 1980. It is a claim based on statistics compiled by celebrity economists Thomas Piketty of the Paris School of Economics and Emmanuel Saez of UC-Berkeley. And it is complete rubbish.

As Cato Institute economist Alan Reynolds recently observed, the Piketty-Saez figures also show zero income growth for the bottom 90 percent since 1968 (!). “This is totally inconsistent with the data the Bureau of Economic Analysis uses to calculate GDP,” writes Reynolds. “For example, real personal consumption per person has tripled since 1968 and doubled since 1980, according to the BEA. Are all those shopping malls, big box stores, car dealers and restaurants catering to only the top 10%? The question answers itself.”

To understand why the Piketty-Saez data are so deceptive, we need to keep a few things in mind. First, the data reflect pre-tax income reported on tax returns. Therefore, they do not account for progressive tax policies or refundable tax credits, employer-provided health insurance, non-cash government transfers, non-taxable Social Security benefits, or other sources of income.

Moreover, Piketty and Saez analyze trends among “tax units” rather than households. That’s an important distinction, because many households contain multiple tax units. Just think of all the 20-somethings who share living costs with roommates but file separate tax returns, or the young workers who reside with their parents. As the Tax Policy Center has noted, “the number of tax units will tend to be larger than the number of families or households reported elsewhere” — meaning that tax-unit income will be significantly smaller than household income.

Something else to keep in mind: Since 1980, America has become an older country with a bigger share of retirees. We would not expect retirees to have much wage or salary income. Thus, any dataset that includes retirees but omits (1) the entire value of their Medicare/Medicaid benefits and (2) a large portion of their Social Security benefits, will inevitably make market incomes among the working-age population seem lower than they actually are.

Final point: The Piketty-Saez data are adjusted for inflation using a measure that, as Manhattan Institute scholar Scott Winship has explained, “fails to fully account for the ability of consumers to substitute between goods and services as their relative prices change.” A superior measure is the Personal Consumption Expenditures (PCE) deflator, which is preferred by the BEA, the Federal Reserve, and the Congressional Budget Office (CBO).

As I discussed in a recent IWF policy focus, CBO has compiled household-income data that capture a broader range of market income, that account for all taxes and transfers, that use the PCE deflator, and that are adjusted for variations in household size. The CBO figures indicate that, between 1979 and 2011, America’s real size-adjusted median household income grew by 47 percent, while the non-size-adjusted median household income grew by 40 percent. Among households in the middle quintile of earners, the average income increased by 35 percent. Among households in the bottom two quintiles, the gains were 37 percent (second quintile) and 48 percent (lowest quintile).

More on middle-class incomes in subsequent posts.