by Calculated Risk on 3/28/2010 09:20:00 AM

I read this from Bloomberg this morning: Payrolls Probably Increased in March

Employers in the U.S. probably added jobs in March for the second time in more than two years, setting the stage for a broadening of the expansion, economists said before a report this week.

Payrolls probably rose by 190,000, the most in three years, after declining 36,000 in February, according to the median forecast of 62 economists surveyed by Bloomberg News before the Labor Department’s April 2 report.

Aargh.

As I noted earlier this month in Employment: March Madness, a headline number of 200,000 net payroll jobs might be considered weak!

The March report will be distorted by two factors: 1) any bounce back from the snow storms in February, and 2) the decennial Census hiring that picked up sharply in March.

These are real payroll jobs, but the Census hiring is temporary - and the Census jobs that are added in March, April and May will all be lost over the following 6+ months.

What we are interested in is the underlying trend of payroll job growth. To find that number we need to adjust for the Census jobs (although they are reported NSA), and we also need to adjust for the February snow storms. Later this year we will need to add the Census jobs back to find the trend.

The important point is 190,000 is probably a weak number for March - and probably not "setting the stage for a broadening of the expansion" - although we need to see the details.