THE STASH NOVEMBER 4, 2009
Is Stanford's John Taylor -- who, according to this measure is the 10th most influential economist in the world -- exhibit A for the corrosion that occurs when politics meets academic economics?
In a blog post last week titled "National Accounts Show Stimulus Did Not Fuel GDP Growth," Taylor writes:
Along with the news that real GDP growth improved from -0.7 percent in the second quarter to 3.5 percent in the fourth quarter, the Bureau of Economic Analysis (BEA) released detailed National Income and Product Account tables...These tables make it very clear that the $787 billion stimulus package had virtually nothing to do with the improvement. Of the 4.2 percent improvement, more than half (2.36 percentage points) was due to firms cutting inventories at a less rapid pace, which has nothing to do with the stimulus. (For the details look at BEA’s Table 2 which shows that the contribution of inventory investment increased from -1.42 to .94 which equals 2.36.)
What about the other components of GDP? In particular what about government spending, which was supposed to be a big part of this stimulus? Government spending was a negative factor, subtracting 0.9 percentage points from the change in GDP growth.
While nothing about Taylor's math is technically inaccurate, his interpretation strikes me as way off. Taylor is looking at a table which breaks down the sources of the percentage change in GDP growth and then taking the difference between the two:
Examined in this manner, you certainly do get that there was a 0.9 percentage point drop in the contribution of government spending to GDP growth. (Here is a good explanation of the difference between percent and percentage point.)
But being a "negative factor" for the change in GDP growth isn't the same thing as saying government spending "had virtually nothing to do" with GDP growth, which is of first-order importance. (Gov't spending did contribute to GDP growth -- it was 0.48 percentage points of the total 3.5% gain as the table above shows. It just contributed less to GDP growth than it did the previous quarter.)
I'm not arguing that government spending was a huge part of the Q3 GDP story, but as Noam first wrote last week, it certainly doesn't mean the opposite either:
Meanwhile, federal government spending increased at a 7.9 percent rate after increasing 11.4 percent in the third quarter, which means we're moving past the point where the stimulus is going increase GDP growth. (That's not the same as saying the stimulus is decreasing GDP. The stimulus is still inceasing the overall level of GDP. But obviously growth is the change from one quarter to the next, so a smaller absolute increase is actually a drag on the change in GDP growth.)
So, why is Taylor bending over backwards here? My pet belief about these things is that any academic economist who advises a political candidate loses much of his or her credibility. Taylor is -- along with many top university economists -- probably no exception here, most recently working for John McCain during the '08 elections. But it's also a bit of the chicken-and-egg dilemma, too: Is Taylor pushing this particular view of Q3 GDP because of his prior beliefs, or did becoming a cog in a political party's machine change his views? It's probably the first, but having to ask this question means that politics and academia don't mix well. There's certainly a strong incentive for academics to become liars about their own beliefs or ignore what they've learned when they choose to work for political candidates, which just seems seriously inefficient.
On the other hand, I suppose you could argue that it's a lot easier to ignore Taylor now, whereas hardcore partisans like Krugman and Casey Mulligan (both of whom I don't believe have ever officially advised a candidate) can still get mileage out of their relatively pure academic credentials.
P.S.: According to the link at the top of this post, Krugman is the 8th most influential economist in the world and Mulligan is ranked 309, which isn't too shabby either.
P.P.S.: One simple way to assess the relative impact of government spending on the final GDP number is to compare it with the impact of other areas at the same categorical level. These other areas are personal consumption expenditures, gross private domestic investment, and net exports. By the table above, the average absolute value for all four categories in the second quarter was 1.68 percentage points and 1.15 in the third quarter. So, gov't spending's impact on GDP was 80% of the average in Q2 and 42% of the Q3 average.