Economists at Societe Generale had a quite decent overview of the key issues yesterday (click on charts to enlarge, courtesy of Societe Generale):
Employment is becoming a central question in the economic outlook for 2010. Production rebound is underway and should cement positive growth in the second half of the year. However sustaining this rebound into 2010 will require a pickup in demand. Will a jobs recovery materialize in time to support a smooth handoff from inventory-led growth to consumption-led growth?
Understandably, there is growing concern among investors about another jobless recovery. The last two recoveries saw anemic employment and wage growth during the early years of expansion. This time, a repeat could be more damaging because credit creation is unlikely to be strong enough to offset any persistent weakness in income. Therefore the next “jobless recovery” may be no recovery at all.
The stakes are high, and there are not straightforward answers. Standard economic models, i.e. those reflecting normal lags between growth and employment, suggest that payrolls should stabilize by year end and begin to climb in early 2010. That is our baseline view.
However, these models failed in the early 1990s and early 2000s and we must consider the risk that they fail again. The difficulty in correctly forecasting jobless recoveries is that they require sustained productivity gains and productivity is notoriously difficult to predict.
Nonetheless, we can evaluate the risks in the context of explanations for previous jobless recoveries. There are legitimate arguments on both sides.
Why another jobless recovery?
One of the reasons for jobless recoveries in the previous two cycles was an unusually high share of permanent layoffs relative to temporary layoffs. Permanent layoffs tend to be more structural than cyclical. Reabsorbing those workers back into the workforce requires creation of new industries and/or businesses and this process tends to take more time than recalling workers back to the same jobs. In the latest downturn, several sectors – including construction and financial services – have undergone large permanent layoffs that are unlikely to be reversed when the recession is over. In addition, recent reports suggest that a large number of workers are involuntarily working part-time. Therefore, the initial demand for labour may be filled by extending the workweek of those part time workers rather than creating new jobs.
In our own analysis, we have also found that divergences between output and employment can be explained by relative pricing power of businesses. When margins are getting squeezed due to inability to pass-through higher material costs, businesses typically try to offset the pressure by reducing labour and squeezing more productivity out of their workers.
This occurred in late 2007 and early 2008 when energy prices squeezed profits. It also explains to some extent the jobless recovery of 1992 and 2002 when pricing power was weakened by a flood of cheap goods out of Asia and the rest of emerging world.
This has mixed implications for the next cycle. Pricing power of businesses is very weak, but declines in material costs relative to year-ago levels could be easing layoff pressures.
Why not?
There are also several good arguments that can be made in support of a more normal jobs recovery than 1992 or 2002. First, there was no overinvestment or over-hiring in the preceding expansion, which means that there is less need for any payback effects in employment. In fact, job losses are already undershooting GDP (or output) which suggests growing pent-up demand for labour. The only scenario in which this pent-up demand fails to materialize is a surge in productivity.
To dig a bit further into this argument, we have compared employment and output trends by sector. We found that the undershoot in employment is particularly strong in the service sector, far more than in any previous recession. To put this in perspective, the service sector has lost 3.4 million jobs so far in this recession, or more than half of all jobs lost. In previous two downturns, the service sector accounted for only 20% of all job losses, or about 450K jobs on average in each recession.
This may be a good sign, because it is generally easier to drive productivity gains in the capital intensive manufacturing sector than in services. It is also more difficult to outsource or export service sector jobs, including those in retail, business services, healthcare and education. This suggests a more normal jobs recovery as the economy begins to expand.
So, you know now!
No comments:
Post a Comment