Simple extrapolation

The trend in employment growth since January 2009 (the worst month of job losses during the recent recession) has been remarkably consistent: on average, each month's job performance (in terms of net change in payroll employment) has been about 67,000 better than the month previous. Job gains in March put us right on the trend line for that period. Simple extrapolation of the data suggests a net gain of 232,000 jobs in April.

All of the net improvement in the jobs picture since January 2009 has come from private sector employment. Total government employment has fallen by about 6000 jobs per month on average: cutbacks at the state and local level have offset gains in federal employment. Thus the trend for private sector employment is similar. Extending the trend line suggests, again, +232,000 private sector jobs (implying no net gain in government employment in April).

I think the trend in employment has been driven by expansionary fiscal and monetary policy and the economy's natural recuperative powers. I don't see anything in April that would have caused a departure from that trend. Census hiring may pick up in April, which would imply higher total employment.

So I'm going to predict +232,000 private sector jobs in the report tomorrow, and let's say +250,000 jobs overall. Note however that the standard deviation around the trend is large (81,000 for payroll employment), meaning that my 95 percent confidence interval for payroll employment (ex-Census adjustment) is +68,500 to +395,500.

Naturally this puts me in optimistic territory relative to the experts. Calculated Risk says the consensus is +200,000 total payroll employment (other sources I've seen put the consensus in the +170,000 range), of which +28,000 is private sector. I don't know how CR or the conventional wisdom squares that meager private employment forecast versus strong overall forecast with the fact that government jobs have been virtually stagnant for the past year. CR says there should be 100,000 new Census jobs in April, but that will be offset by employment losses at the state and local level. Also keep in mind that the ADP report (which showed +32,000 private sector employment in April and is the focus of the post in CR) badly underforecast private sector growth last month. I suspect that survey does a much worse job forecasting the BLS numbers than advertised.