Dobridge, Hooper and Slok in “Jobless Recovery III Seems Unlikely” Global Economic Perspectives (April 21, 2010) [not online]:
The sluggish performance of payroll employment and
jobless claims in recent months despite well-above-trend
growth in output has raised the specter of another jobless
… Various high-frequency indicators of labor demand, as well as labor market policies and historical trends within the labor market, show a mixed picture, with
some pointing in a positive direction and others in a negative direction. We judge the balance of evidence to fall modestly in favor of a strengthening of payroll employment growth in the months ahead. Under our central scenario, we see monthly payroll employment growth average near 200k per month for the balance of 2010. This outlook assumes 4% GDP growth, a rapid drop in productivity growth to an underlying trend rate of around 1.5%, and a moderate recovery in the hourly workweek. Key uncertainties include the assumed paths of labor productivity and hours worked. There are risks to these assumptions on both sides of the central scenario, but they appear to us to be balanced more to the downside than the upside.
This forecast, plus alternate scenarios, are presented in Chart 20.
Chart 20 from Dobridge et al., “Jobless Recovery III Seems Unlikely” Global Economic Perspectives (April 21, 2010). Note figures are for private nonfarm payroll employment.
I thought it of interest to see how this estimate fits into the econometric-based conditional forecasts presented in this post from last November.
Figure 1: Nonfarm payroll employment, SA, in thousands, (blue), dynamic forecast from error correction model (purple), dynamic forecast from first differences specification (light green), and from error correction model estimated over 1967Q1-09Q3 period (salmon). WSJ forecast for October 2010 (teal square). Shaded regions denote forecasting periods. Source: BLS via FREDII, WSJ October survey, and author’s calculations. See this Nov. ’09 post for details.
To see how things have played out since last November, I superimpose the latest nonfarm payroll employment series in dark blue. Note that the level has been shifted down, due to the benchmark revision associated with the January 2010 release.
Figure 2: Nonfarm payroll employment, Sept. ’09 release (blue), and Mar ’10 release (dark blue bold line), in thousands, seasonally adjusted; and dynamic forecast from error correction model (purple), dynamic forecast from first differences specification (light green), and from error correction model estimated over 1967Q1-09Q3 period (salmon); all forecasts based on October 2009 data. WSJ forecast for October 2010 (teal square). Shaded regions denote forecasting periods. Source: BLS via FREDII, WSJ October 2009 survey, and author’s calculations. See this Nov. ’09 post for details.
What I find interesting is that the pattern (although not the level) most closely matches the conditional forecast associated with the full sample error correction model (and not the one estimated over the last two jobless recoveries). The key conditioning variable is GDP. It turns out that GDP in 09Q4 was only 0.5% (log terms) higher than the October WSJ mean forecasted level I used to conduct the forecast. This suggests that differences in GDP are not central to the difference in forecasted growth in employment.
The Dobridge et al. forecast 180 thousand jobs created on net per month is quite close to my estimate of 192.5 thousand per month. One difference is that the DB forecast (3.8% q4/q4 in 2010) is substantially higher than the October 2009 WSJ mean forecast. (Update: There is a slight difference in measures; Dobridge et al. are forecasting private employment growth, while I forecasted total nonfarm payroll employment growth. If there is little change in government employment going forward, then the figures are comparable.)
Somewhat less optimistic implications can be drawn from the IMF’s recent assessment of output-employment linkages in Ch. 3 of the WEO, which indicates a 6 quarter lag between trough in output and resumption in employment growth, in the wake of recessions accompanied by financial crises (Figure 3.10).
Update, 11:45am Pacific Reader tj writes:
“At the low end of the wage scale, we have a minimum wage that jumped 3 times from 2007 – 2009. This is our first recovery in which hiring decisions are impacted by the new minimum wage.”
This is factually incorrect:
Figure 3: Minimum wage in dollars per hour (red) and 2009$ per hour, deflated by CPI-all. NBER recession dates shaded gray. Source: St. Louis Fed FREDII, NBER and author’s calculations.