Seasonality and jobs, one more time | FT Alphaville

Seasonality and jobs, one more time

Our thinking on the US recovery has been that things were never quite as bad as we feared in the worst of times (the last two Augusts), and never quite as good as we’d hoped in the best of times (the last two winters).

Back in December, we first started writing about the possibility that distortions in seasonality adjustments — the result of the Census Bureau’s X12 algorithmic program having been thrown off by the severity of the downturn in the fourth quarter of 2008 — had exaggerated the economic upturns and downturns for the past couple of years.

That doesn’t mean the swings weren’t real, or that such events as the Japanese earthquake, commodity price increases, or the debt ceiling debate had no influence. They did.

But we suspected, based on some work done by the economics teams at Nomura and Goldman Sachs, that they weren’t quite as dramatic as the economic indicators had made them out to be.

(Which is wholly different from saying that the recovery at any point has been adequate given the scale of the unemployment problem. As usual the danger in writing a post like this one is that we’ll sound Pollyanna-ish. We’re really not, and find the ongoing fiscal policy constraints and messaging problems of the Fed to be outrageous.)

This morning’s jobs report inspired at least a couple of posts on the possibility that jobs growth in June was probably somewhat better than the headline numbers. And indeed the record warm winter was a factor unique to this year that might also be playing funny with seasonality issues.

We actually have no idea if seasonal adjustments are responsible. We worry especially that growth is slowing everywhere in both developed and emerging countries, and of course the seasonality distortions are themselves very likely to have diminished quite a bit by now. The possibility of big external shocks continues to loom, with Europe, China, and the US fiscal cliff only three of the most commonly cited examples.

But when we look at this updated chart from Nomura…

… we find the relationship between the seasonal shifts the past three years in these economic surprise readings to be a little too close to be all coincidence.

The chart certainly doesn’t account for all of the recent economic weakness and it applies to a number of indicators, not just the jobs report. But our guess — and please, we emphasise that it’s just a guess — is that that the underlying pace of jobs growth is somewhere between this month’s print of 80,000 and the six-month average of monthly job gains (150,000). We base this both on the seasonality issues and on the fact that certain other employment reports (initial jobless claims, the ADP private employment report, Challenger layoffs) indicate a slightly better situation.

To be sure, 80k-150k is a wide range, but our main point is that yet again, things probably aren’t quite as bad as the employment reports of second quarter would suggest, just as the economy wasn’t growing as quickly as the readings from the first quarter suggested.

Related links:
‘Tis the seasonality, hold the jolly – FT Alphaville
Winter seasonality, employment report preview edition – FT Alphaville
Wait! It looks like the job market is actually roaring – Clusterstock