The NY Times article refers to the 2010 Census hiring as a $2.3 billion "injection into the economy". Mark Zandi at Moody's says:
“It’s a form of stimulus. It’s like infrastructure spending, or W.P.A. in the Depression. It effectively does the same thing. It’s not on the same scale, but it is large enough, and it will make a difference.”I disagree. The $2.3 billion figure is misleading; it is simply earned income (about 800,000 jobs X $25/hour X 20 hours/week X 6 weeks), rather than direct spending. A worker that earns temporary income from a six-week job is more apt to save the income rather than spend it - that's why rebates don't work. Not much stimulus there.
And don't forget the other side of the Census hiring story: the inevitable laying off of workers following the peak Census month.
The chart illustrates annual growth in government employment according to the establishment survey. Like clockwork, the Census survey grows government jobs to a peak 2.5%-4.5% annual pace in April or May of the census year. This means a peak impact of 800k-900k new jobs created in April or May 2010 (I use the average annual growth rate of 3.8% and a historical monthly average to forecast the payroll until April). All else equal, the unemployment rate will drop.
But consider this. It is likely that a very large share of the monthly 800,000 hires will simply be unemployed when the Census culminates - hence, the unemployment rate will rise in the early summer. And it's very possible that the increase in the unemployment rate will be larger than its decrease. If any discouraged workers - those who have not searched for employment recently and are counted as "not in the labor force" - are hired and do not secure new employment after the six-week job is over, then they will join the ranks of "unemployed".
We'll see; but the economic impact of the 2010 Census, to me, is questionable.