We know from New Zealand data this week that labour market data is requiring a bit of careful interpretation at the moment. The United States is no exception.
In data released overnight, the US unemployment rate dropped another 0.4 percentage points to 9.0% in January. It is now down 0.8 percentage points over the last two months. This is, however, despite apparent sluggish jobs growth. Non-farm payrolls rose only 36k in January, well below average market expectations of +150k. Poor weather copped the blame for the poor result.
The drop in the unemployment rate over the last two months has been driven by a significant drop in the participation rate. That is essentially telling us that people have "dropped out" of the labour force, 162k of them in January alone.
It is not unusual in times of high unemployment for people to drop out of the labour force. People become disenfranchised with the state of the labour market and give up looking for work. Or is there something else going on here?
The lacklustre jobs growth in this survey of employers is at odds with surveys of households which are showing significantly stronger job growth. The Labor Department survey of households showed jobs growth of 589k in the month! That is significantly different from 39k!
So what should we believe? Or is the more important question: What should Ben Bernanke believe about what is happening in the labour market when he considers the desirability (or not) of continuing on with the current programme of asset purchases, a.k.a. QE2. That's a really big and important question. Remember, 50% of the reason QE2 was launched was due to persistently weak jobs growth and high unemployment (the other 50% was price stability).
Could it be the case that, given the structural nature of the recession and the subsequent recovery, that surveys of employers are a poor indicator of current labour market activity? That's because old employers have disappeared, and the surveys havent' yet caught up with new employers or a large number of people who may have become self-employed in the aftermath of the recession.
So perhaps we are due some stronger employment numbers as the payrolls data plays catch-up. It's interesting that recently, as each subsequent payrolls number gets released, the two preceding month's employment numbers get revised up. With the release of this report, December and November got revised up a collective 40k. When the December data was released in early January, November and October were revised up a collective 70k.
And perhaps this helps explain stronger-than-expected consumption at the end of last year?
I tend to focus on employment trends more than unemployment when assessing the labour market data. That's because of the inherent difficulties in measuring unemployment. Over time, any disparities between the two tend to come out in the wash.
Policymakers are right to continue to focus policy actions on weak payrolls job-growth. But there's something else going on here that will become obvious soon. Perhaps we will be talking about stimulus exit strategies a bit sooner than we thought. Fascinating.