WASHINGTON (Reuters) - The U.S. government has initially underestimated job growth for almost every August over the last decade, a trend that could make it harder for the Federal Reserve to decide if an upcoming employment report will signal America is ready for a rate hike.
The Fed said in July it needs to see “some” further improvement in the labor market before raising rates. The Labor Department’s jobs report for August, which will be released on Friday, will be the last monthly employment report before the Fed’s Sept. 16-17 policy review.
But like many economic indicators, it is subject to revisions as employers turn in questionnaires late and government statisticians re-estimate seasonal swings in the data.
Between 2005 and 2014, August was the month with the lowest first estimate for job growth relative to revisions that were published in the employment reports for the subsequent two months, a Reuters analysis found.
Over that period, on average the government found 58,000 more jobs added to payrolls in August than initially figured. Job growth in August was revised lower only in 2005 and 2008.
For this Friday, the number that market participants and economists are looking for is 220,000, roughly in line with the pace of job creation over the last few months, according to a Reuters poll. Far fewer than that and bets on a September Fed liftoff will sink fast.
“The risk for a negative surprise in August seems to be much larger than the risk for a positive surprise,” said Harm Bandholz, an economist ay UniCredit in New York.
Despite the flaws, several Fed policymakers recently have signaled the report’s importance. Many analysts think it could be the deciding factor for whether the Fed hikes rates this month, although the U.S. central bank also has an eye on global financial market turmoil.
The Labor Department acknowledged its initial estimate for August has tended to be revised higher.
John Mullins, an economist at the department’s Bureau of Labor Statistics which produces the monthly employment report, said the revisions have been notable across a variety of industries but were most common in highly seasonal sectors.
Adjusting the data for seasonal swings is important because it can help give an idea about the economy’s underlying strength.
Job cuts at construction sites during the winter, for example, might signal the economy’s direction only if the layoffs are more than what normally happens when the ground freezes.
A key problem for government statisticians is that seasonal factors are constantly shifting. Mullins said in an email that adjusting for seasonal swings in the education sector “can be particularly troublesome in August due to moving seasonality as many schools change their start dates in the fall.”
He did not speculate on why revisions in August would be regularly higher.
August is not alone in this sense - at least since the last recession.
Every month but June and April have on average had upward revisions since 2010, the year the labor market recovery kicked into gear after the 2007-2009 recession.
During the recession years, initial job tallies on average were subsequently revised lower in almost every month.
It’s unclear why the business cycle could be affecting revisions. But August was among the few months to generally have upward revisions both during and after the recession. Between 2010 and 2014, the Labor Department found 76,600 more jobs that month.
Reporting by Jason Lange; Editing by Chizu Nomiyama