This was to be the morning when we received the bill for last month’s government shutdown shenanigans. All those who had irresponsibly failed to agree upon a budget by the Oct. 1 start of the fiscal year were to learn what damage they had wrought. How many more Americans would be looking for work because of their inability to find common ground? And the verdict was…
204,000 new jobs in October! The number was dramatically above forecasts of 125K and above the 190K/month average of the preceding twelve months. This seems the equivalent of partying all night before a test, never cracking open the textbook, and then finding out you got an A on the exam. What gives? Three thoughts below.
1. Approach monthly job numbers with caution
The closely-watched report actually has a margin of error of around 100K jobs. As the economist Justin Wolfers reminds: “There’s a lot of noise…Don’t overinterpret every blip in the data.”
This is not just a technicality. Every release features revisions of past releases. This one was no different: The August employment number was revised up from 193K to 238K, while the September number went from 148K to 163K. So, on top of the 79K unexpected October jobs, it turns out there were another 60K jobs from before that we had not counted on.
2. The “but for” challenge
Numbers such as the jobs release are rarely dispositive in economics. Given the impossibility of running a controlled experiment (“Let’s try October again, this time without a shutdown!”) we are left to speculate about what would have happened “but for” the shock that occurred. To do that, we rely upon models that are often unreliable. There is always room for someone to say: “Just think how high the number would have been without the shutdown!”
The challenge is even more difficult when other policies change. One celebrated prediction of the shutdown’s impact called for 900K jobs lost and the unemployment rate jumping by 0.6 percentage points (it actually stayed unchanged at 7.3%). One mechanism by which such havoc would be wreaked was that interest rates would rise as markets were overcome by fear and uncertainty. And yet the benchmark Treasury 10-year interest rate was roughly 30 basis points lower throughout the shutdown than at its early September peak. Of course, there was a major policy move in the interim: the Federal Reserve postponed its “taper.” There was speculation that, in doing so, it was casting a wary eye on Washington developments. Perhaps so.
But how does one then disentangle the net effect of the shutdown? Should it be the combined effect of the shutdown and postponed taper? It seems unlikely that the Fed would have reneged on its earlier taper commitment had there been a new budget firmly in place for FY2014 and ample room under a new debt ceiling.
3. What did you expect?
With all of the above caveats, there is a deeper question about how markets and the public approach temporary measures.
In an unusual attempt to increase pressure on Republicans early in the shutdown, President Obama gave an interview with CNBC in which he urged markets to worry: “This time I think Wall Street should be concerned…When you have a situation in which a faction is willing to default on US obligations, then we are in trouble.” Presumably the intent was to invoke the wrath of the markets and terrify opponents as indices tumbled.
Markets refused to comply. The S&P 500 actually rose slightly over the first couple weeks of October. That index reflects a huge number of daily transactions, so it would be foolish to give a single reason for the way it moved. But market participants clearly expected that the standoff would be short-lived and that the United States would make good on its debts. Had someone taken the President up on his invitation to panic, they would have ended up selling low and then buying high when the crisis passed. That’s not what they’re there for.
These same questions of expectations and “but for” analyses have bedeviled discussions of stimulus. In a Keynesian approach, the citizenry is first surprised by the burst of government spending, then surprised again later when taxes go up to pay for that spending. In a “rational expectations” approach, citizens are not surprised, nor are they so responsive to policy maneuvers.
Returning to the question of jobs, Paul Krugman has a piece in today’s New York Times in which he laments the staggering cost of persistent high unemployment. In that, he’s clearly right. He then goes on to attribute the unemployment to a lack of greater government spending and the influence of “deficit scolds.” Others would differ sharply in their diagnosis. Today’s job numbers seem to warn against facile analyses and overconfidence that we can pair each act of government with a market response.