February 5, 2010 / 2:44 PM / 9 years ago

Labor Dept pains in the birth-death jobs model

WASHINGTON (Reuters) - The Labor Department on Friday blamed faulty estimates of how many companies were created or destroyed for its unusually large revision to payrolls that showed job losses were considerably steeper than first thought.

Once a year, the Labor Department compares its payroll data with unemployment insurance tax reports and releases a “benchmark revision” that adjusts for discrepancies.

Normally the difference is relatively modest. This time, the Labor Department revised the level of employment for December 2009 down by 1.39 million, bringing the total number of jobs lost since the start of the recession to 8.4 million.

The primary culprit behind that huge revision was the so-called “birth-death” model, a method the Labor Department uses to try to estimate how many jobs were gained or lost because of companies opening or closing in a given month.

While economists have long questioned the accuracy of the model, it had performed well up until the latest recession.

This time, it overstated job creation by 779,000 in the year that ended in March 2009. The Labor Department also revised its monthly jobs data for April through December 2009, and found the birth-death model had overstated job creation by another 405,000.

"The pronounced recession led to a breakdown in that stability from March 2008 to March 2009," the Labor Department said in an article posted on its website, detailing the massive revisions. The article is available here: www.bls.gov/web/cesbmart.htm

“These errors started to grow in the fourth quarter of 2008 and got significantly larger in the first quarter of 2009,” the Labor Department said.

Economists at the department are still researching ways to improve the model so that it can better estimate job creation and destruction, particularly in times of economic upheaval.

The Labor Department said it was investigating whether it would be helpful to run the model quarterly instead of once a year, and adding other variables.

Reporting by Emily Kaiser; Editing by Andrea Ricci

0 : 0
  • narrow-browser-and-phone
  • medium-browser-and-portrait-tablet
  • landscape-tablet
  • medium-wide-browser
  • wide-browser-and-larger
  • medium-browser-and-landscape-tablet
  • medium-wide-browser-and-larger
  • above-phone
  • portrait-tablet-and-above
  • above-portrait-tablet
  • landscape-tablet-and-above
  • landscape-tablet-and-medium-wide-browser
  • portrait-tablet-and-below
  • landscape-tablet-and-below