Hoisted from Comments: David Brooks's Innumeracy

John Schmitt writes:

Brad DeLong's Semi-Daily Journal: Fire David Brooks. Fire David Brooks Today (Why Oh Why Can't We Have a Better Press Corps?): I was also curious about Brooks's claim that: "The typical male worker with some college but no degree has seen his income rise from $34,000 in 2000 to about $40,000 today." These numbers probably give most readers the impression that the real annual earnings of the typical man with some college but no degree rose about 18 percent between 2000 and 2005.

A quick check at the Census web page shows median annual earnings for men 25 and over with "some college no degree" to be $35,463 in 2000, rising to $39,150 in 2005. But, both figures are in nominal terms.

Adjusting the Census numbers above for inflation suggests that the earnings of the typical man with some college but no degree actually fell about 3 percent over the same period.

Does anyone have any idea about the source for the David Brooks number? Or any thoughts on why there is such a discrepancy between his numbers and those at the Census Bureau?

If Brooks got this was as wrong as he appears to have gotten it, he should have to run a correction in his column.

My sources:

• for the 2000 number, http://pubdb3.census.gov/macro/032001/perinc/new03_071.htm
• for the 2005 number, http://pubdb3.census.gov/macro/032006/perinc/new03