The Bureau of Labor Statistics reported Friday that the unemployment rate has fallen from 9.8% in November to 9.0% in January, as big a two-month drop as we’ve seen in the last 50 years (hooray!). But in the same report, BLS indicated that their seasonally adjusted estimate of the number of Americans employed on nonfarm payrolls increased in January by an anemic 36,000 (oh dear!). Reconciling the very contradictory claims is even harder than usual, but I’ll give it a try.
We can start with the fact that the two numbers come from two different surveys and are measuring different things. The unemployment rate comes from a survey of households, and counts someone as employed if they did any work at all as a paid employee or worked in their own business during the surveyed week, and also people who have a regular job but missed work due to temporary factors such as illness or bad weather. The nonfarm payrolls, on the other hand, come from surveys of establishments themselves. If bad weather caused someone to miss work for a two-week payroll period that included the 12th of the month, that person would not be counted as employed according to the establishment survey. Rebecca Wilder notes that Nomura economists accurately predicted prior to the BLS release that weather distortions would cause the reported nonfarm payroll gain to come in well below 56,000. The same Nomura report noted:
In one of the largest first reported declines on record, the BLS in its February 7, 1996 report calculated that non-farm payrolls FELL by 201,000 from the previous month. The outsized decline hit both manufacturing (-72,000) and services (141,000) but the construction industry registered a net job gain of 13,000. At the time, the BLS blamed the big winter storm for skewing the job loss and a month later reported that payrolls surged by 705,000 in February after a revised drop of 188,000 in January.
But even if we stick to just the household survey, there is still some serious reconciliation required. We can start with the fact that the household survey reports a value for the civilian noninstitutional population for January that was 185,000 lower than the value for December. Obviously that’s not what really happened, and here’s the explanation:
BLS introduces the annual population control adjustments into the CPS estimates beginning with the January data. The adjustment can either increase or decrease the population level, depending on whether the latest information indicates the population estimates have trended too high or low. Conceptually, the population control adjustments represent the cumulative over- or under-estimation of the population since the last decennial census point.
That makes the December-to-January comparison of any of the household survey magnitudes a bit problematic. But, as Rebecca also notes, Table C in the BLS report suggests that one might just subtract the population control effects from the published December-to-January change in employment to arrive at an implied net gain in employment of 589,000 jobs according to the household survey (hooray again!). Declining participation rates have been one factor in earlier improvements in the reported unemployment rate. But it looks reasonable to me to interpret the further improvements in January as indicating real progress.
Here’s my bottom line: if you had concluded (and I had) from recent sales data and manager surveys that we’re finally seeing the economy growing solidly, there’s nothing in the latest jobs report to persuade you otherwise.
Various rates inferred from household survey. Source: Calculated Risk,
Originally published at Econbrowser and reproduced here with permission.