In his latest column, David Brooks asserts that Americans “spend roughly $11,000 per student” on K‑12 public schooling. This is false. We actually spend about $14,000 per student.

Why is his figure so erroneously low? The answer may astonish and anger you.

The likely source for Brooks’ figure is the last row in column 3 of Table 191, in the Department of Education’s latest Digest of Education Statistics: “Current expenditure per pupil in average daily attendance” for the year 2008-09 (the most recent reported). The value was $11,231. The first and biggest problem with this number is that “current” spending is a subset of total spending—but Brooks didn’t tell you that, did he? He just said “we spend roughly…,” not “our current spending excluding the following items is…”

As it happens, “current” or “operating” expenditures exclude costs like construction and debt service. These, however, are real costs that have to be paid in real dollars (to the extent our dollars are “real” after yet more QE) by real taxpayers. What difference does this make? Just look one column to the left. “Total expenditures” per pupil for the same year were $13,015. Bit of a difference. In fact, on a national level, it’s a difference of over $90 billion annually (see the top two rows of column 9, Table 188).

Of course that was back in 2008-09, when U.S. dollars were worth more than they are now. Adjusting for inflation into today’s dollars ups the figure to about $14,000 / pupil. Moreover, we’re almost certainly spending more today than we were in 2008-09, despite the economic downturn, but for some reason the public school monopoly prefers to release only aged data—as though they were fine cheeses.

So the next time someone claims to tell you what U.S. public schools spend per pupil, ask if they’re just citing you the current operating expenses subtotal. If so, tell them you want your other $90 billion back, since they apparently don’t seem to be using it.