Everyone's all wound up about tomorrow's employment report... as am I.
One could sagely stroke the chin and very correctly pronounce that given the statistical confidence intervals around these data -- their margins of error -- the monthly reports are given far too much weight. The closely watched net job growth figure, for example, has a margin of error of plus or minus 100k. The numbers also are subject to significant revisions in coming months.
Or one could hang up the wonkiness and face that we're entering the heart of the election season where each jobs report will be a very big deal.
Best to try to do both -- respect the limits of the data, try to figure out what the numbers are trying to tell us (which typically means taking averages over a few months), but don't ignore their elevated political salience right now.
The consensus expectation is for about 150,000 jobs overall, 160k in the private sector, meaning folks reasonably expect the state and local sector to keep shedding jobs (state/local cutbacks sucked three-tenths of a point off of real GDP growth last quarter). The unemployment rate is expected to stay where it is at 8.1%.
I'm a bit more pessimistic and wouldn't be surprised to see a payroll number south of 150k -- more like 130k. A lot of this depends on technical aspects of the jobs data, like the fact that unseasonably warm weather boosted earlier months this year at the expense of later months. If the "seasonals" settled down in May, which "may" be the case, we're more likely to hit a higher jobs number.
Either way, the main economic point to keep in mind is that all of this is pretty consistent with an economy that's slogging along at its trend growth, rate about 2% per year. That's not a bad thing and, in fact, it's worlds beyond the horrors of the Great Recession, when GDP was tanking at 9% per year and we were losing 750,000 jobs a month.
But it's not great either. After being down so far, you need some bounce-back months where you're well above trend for a while. And that is not something we've reliably experienced.
Now, to the politics. Simply put, my experience suggests -- and this is art, not science -- that any jobs number of 150k or above with unemployment staying put, at worst (i.e., not rising), will be considered quite favorable for the administration. Anything below 125k will likely not.
But just think about this for a minute. Given the margin of error, a 170k print on net job growth, for example, could actually be 70k jobs... or 270k!* If the unemployment rate ticks up to 8.2%, that's statistically indistinguishable from no change at all.
And think also about the policy implications behind this. Suppose it's a strong month. What, exactly, did the administration or Congress do last month that was such a great boost to jobs (Jay Carney did mention my work yesterday, so perhaps that's it)? And vice versa, of course.
For policy evaluation, you want to think about the nexus of jobs and policy in the longer term, which in this case asks, "what did the administration do to offset the massive contraction in labor demand, aka the Great Recession?" The answers to that are the Recovery Act, financial and auto rescues, unemployment insurance extensions, payroll tax cuts, and all the rest of that stuff. Those measures demonstrably pulled the recovery, tepid as it is, forward, saved and created millions of jobs, and hastened the turnaround in net jobs growth that began in February of 2010, and which has yielded about four million jobs so far.
You might also want to note that Republicans have generally tried to block all of the above, and since 2010, have successfully blocked efforts like the American Jobs Act to do more to help offset the residual drag from the downturn. In that regard, their fingerprints are the most prominent ones on the current slog.
So, tune in tomorrow and I'll be sure to be going all kinds of crazy regarding the May jobs numbers. I can't help it, and neither can anyone else around here. But the truth is you learn a lot less from such reports than the reaction to them would suggest.
* Technically speaking, there's a 90% chance that the true jobs number is between 70k and 270k.
This post originally appeared at Jared Bernstein's On The Economy blog.