NY Post's Crudele falsely accuses Census of using "statistical tricks" to boost employment figures

››› ››› ERIC SCHROECK

In two New York Post columns, John Crudele falsely accused the Census Bureau of using "statistical tricks" by "repeatedly hiring and firing workers without any apparent reason" in order "to artificially boost the nation's employment figures." In fact, in making this claim, Crudele distorted the Bureau of Labor Statistics' (BLS) method for gathering jobs figures, and the Census Bureau has flatly denied that it is "repeatedly hiring and firing workers."

Crudele accuses Census of using "statistical tricks" to "artificially boost the nation's employment figures"

From Crudele's May 25 New York Post column:

You know the old saying: "Everyone loves a charade." Well, it seems that the Census Bureau may be playing games.

Last week, one of the millions of workers hired by Census 2010 to parade around the country counting Americans blew the whistle on some statistical tricks.

The worker, Naomi Cohn, told The Post that she was hired and fired a number of times by Census. Each time she was hired back, it seems, Census was able to report the creation of a new job to the Labor Department.

[...]

Each month Census gives Labor a figure on the number of workers it has hired. That figure goes into the closely followed monthly employment report Labor provides. For the past two months the hiring by Census has made up a good portion of the new jobs.

Labor doesn't check the Census hiring figure or whether the jobs are actually new or recycled. It considers a new job to have been created if someone is hired to work at least one hour a month.

One hour! A month! So, if a worker is terminated after only one hour and another is hired in her place, then a second new job can apparently be reported to Labor . (I've been unable to get Census to explain this to me.)

From Crudele's June 1 New York Post column:

A guy I'll call Mike has worked for Census 2010 several times in California over the past two years. The last time, he was trained at a facility that was an hour's drive from his home. He was paid for his commuting time at $17 an hour -- which is what he also got while working and training.

[...]

Mike says that after each stint with Census he, like everyone else, was given an official "termination" notice. And every time he was rehired Mike had to fill out a new employment application (more paperwork to be processed by paid workers).

A couple weeks ago I found out that Census was repeatedly hiring and firing workers without any apparent reason. I questioned if this was being done to artificially boost the nation's employment figures since the Labor Dept. considers it a new job created whenever someone is hired to work as little as one hour in a month.

Was Census churning jobs to make the economy look healthier than it really is?

Crudele distorts both Census' hiring practices and how BLS gathers jobs figures

Census director calls Crudele's May 25 column a "distortion of the U.S. Census Bureau's employment and reporting practices ... we do not hire, then fire, and then rehire anyone." In a May 25 letter, Census director Robert Groves flatly denied the claim in Crudele's May 25 column that Census is hiring and firing workers in order to "report the creation of a new job to the Labor Department." Groves wrote that "[w]e do not hire, then fire, and then rehire anyone. Any employee who is fired is fired for cause." Groves also called Crudele's column a "distortion of the U.S. Census Bureau's employment and reporting practices."

Census director explains that when temporary workers' "work is complete," they go "into an inactive status" and "may be reactivated if there is more work," but this "reactivation" is not counted as "a rehire." In his letter, Groves goes on to explain: "We train and hire temporary workers for various operations, most significantly Non-Response Follow-Up, to complete work assignments. When the work is complete, the temporary worker goes into an inactive status. They may be reactivated if there is more work to do, or for another subsequent operation. At no time do we count a reactivation from non-working status as a 'rehire.' "

FactCheck: Crudele wrong on "the way jobs figures are gathered. ... Anyone who had been hired, fired and re-hired would still be reported just once." In a May 28 post, FactCheck.org wrote that "Census denies that what Crudele described is actually happening. And we can report that even if it is happening, it would not have any effect on jobs figures reported by the BLS. Crudele misunderstands -- or at least, did not report accurately -- the way jobs figures are gathered." From FactCheck:

[T]he jobs figures that the Bureau of Labor Statistics reports each month are not based on any count of "new jobs" or new hires. Rather, the BLS gets payroll figures on how many persons are actually being paid at the time of its monthly survey. What reporters and economists commonly refer to as job growth or job losses reflect the net change from month to month, not any tally of those hired or fired (or retired or deceased, for that matter). Anyone who had been hired, fired and re-hired would still be reported just once, no matter how many times they may have been on or off the payroll before the reporting period.

Groves: Crudele's accusation about Census jobs reporting is "simply inaccurate." In his May 25 letter, Groves called Crudele's accusation that Census is artificially boosting employment figures "simply inaccurate," writing that "[i]t is simply not possible for the Census Bureau to engage in the manipulation of data to artificially inflate the employment report of the BLS in the manner alleged by" Crudele. Groves further explained:

The Census Bureau reports to the Department of Labor and on our public Web site the number of people paid for work during a given week. We do not report the number of jobs. The Census Bureau reports the total number of unduplicated temporary 2010 census workers that earned any pay during a specific weekly pay period. Temporary workers earning any pay during the week are counted only once. The Bureau of Labor Statistics (BLS) measures changes in employment levels -- not the actual level itself -- and looks only at the week that includes the 12th day of the month. It is simply not possible for the Census Bureau to engage in the manipulation of data to artificially inflate the employment report of the BLS in the manner alleged by this news column.

Stories/Interests
Census 2010
We've changed our commenting system to Disqus.
Instructions for signing up and claiming your comment history are located here.
Updated rules for commenting are here.