". . . ZAP! Krugman Truth Squad member David Hogberg caught Krugman in a humiliatingly simple math error, which surely the Times will be obliged to correct. Krugman wrote,

As analysts at the nonpartisan California Budget Project point out, real state spending per capita was only 10 percent higher in 2002-03 than it was in 1989-90 — that is, most of the spending growth was simply a matter of keeping up with the population and inflation.

But Hogberg went to the CBP's website to fact-check this claim (Lord know the Times won't bother to do it). Hogberg noted in his Cornfield Commentary blog that the CBP stated that

Per capita spending was $2,211 in 2002-03, as compared to $1,950 in 1989-90 (in 2002-03 dollars).

Now let's see ... I'm no Princeton professor — I'm a Yale drop-out, in fact — but I think Hogberg is right when he says that spending growth from $1,950 to $2,211 is 13.4 percent, not "only 10 percent." How can this be? Is it an innocent error? Is it rounding? Or maybe Krugman thinks that 10 percent sounds better with the word "only" in front of it than 13.4 percent does. . . ."

Thanks to an observation by CalPundit, we can see why Luskin is wrong. If you read the *entire* CBP document, you will see this passage: "When the state lowered the Vehicle License Fee (VLF), it 'backfilled' lost revenues to local governments (VLF revenues flow to local governments). The state spent $3.9 billion to reimburse local governments for VLF losses in 2002-03, but had no such costs in 1989-90, which was prior to the rate reduction."

This VLF decrease is essentially a tax cut, and the $3.9 billion of reimbursed money should definitely *not* be counted as an expenditure. So when we knock off the cost per capita of this reimbursement, we see that 2002-2003 per capita real spending was not $2211 but less: [$2211 - ((3.9 billion)/(Population of Calfornia during 2002-2003))].

A back-of-the-envelope calculation shows that, if CA's population is roughly 40 million (an overstatement) for 2002-2003, 3.9 billion/40 million = $97.5 per capita which was included in the $2211 of per capita spending for 2002-03 but should not be.

Therefore the more accurate figure to derive from this CBP document is (2211-97.5)/1950 = 1.08410 and the more accurate percentage difference for the CBP document is roughly 8.4%!

Therefore, when Krugman wrote 10% in his 8.1.03 NYT article, he was actually overstating and definitely *not understating* the percentage differential of real per capita spending between 1989-90 and 2002-03, relative to the 8.4% that the CBP document yielded.

(Note that 40 million is an overstatement of the actual population of California and implies a smaller decrease due to the VLF reimbursement and hence higher real per capita spending for 2002-2003. Even under this assumption, which is charitable to Luskin's desired conclusion, the per capita real spending differential is less than 10%).

Originally published by Bobby on the Unofficial Paul Krugman Site, 8.22.03