Chris Rupkey, chief financial economist at Bank of Tokyo-Mitsubishi UFJ Ltd. in New York, said he was “stunned” by today’s U.S. employment report.
He wasn’t the only one.
Not a single economist among 85 surveyed by Bloomberg News correctly forecast the 18,000 increase in payrolls in June reported by the Labor Department. Estimates ranged from a low of 60,000 to a high of 175,000. The median was 105,000 -- almost six times the actual number.
“When I’m wrong, I usually go down in flames and this time my forecast crashed and burned,” Joel Naroff, president of Naroff Economic Advisors Inc. in Holland, Pennsylvania, said in a note to clients. Naroff, who wasn’t among the economists surveyed by Bloomberg, declined to give his forecast.
How did economists get it so wrong?
Rupkey pointed to yesterday’s report from ADP Employer Services, which showed that companies in the U.S. added 157,000 workers in June, more than twice as many as forecast by economists.
“People revised up their numbers basically 24 hours ago on a report that apparently had no merit when it came to forecasting the actual number,” Rupkey said.
The ADP report is based on data from about 340,000 companies with more than 21 million workers on payrolls. The Labor Department surveys 140,000 businesses and government agencies.
“There’s no specific explanation of the divergence,” said Joel Prakken, chairman of Macroeconomic Advisers LLC in St. Louis, which produces the report with ADP. “The data is collected differently. They’re different samples.”
Ian Shepherdson, chief U.S. economist at High Frequency Economics Ltd. in Valhalla, New York, was among those who raised their forecasts for today’s payrolls figure after the ADP report. His forecast was for a gain of 175,000.
Shepherdson also pointed to the way the Labor Department adjusts the data to account for seasonal variations. He said the “unusually aggressive” seasonal adjustment factors “may explain at least some of the payroll weakness.”
The Labor Department, which houses the Bureau of Labor Statistics, adjusts the employment figures each month to account for things like teachers falling off school payrolls in June and workers finding temporary employment with retailers during the December holiday season.
“There was a big adjustment this month,” Labor Department Chief Economist Betsey Stevenson said on a conference call with reporters. “It’s an art and a science doing seasonal adjustments and it’s really hard to predict.”
It’s not unusual for payroll figures to fall outside of the range of economists’ forecasts. The same thing happened last month, as well as in October, November and December of last year.
Some economists who came closer to forecasting the government figures said they didn’t see anything out of the ordinary in the data.
“There wasn’t any quirk that we knew about,” said Robert Brusca, president of Fact & Opinion Economics in New York, whose forecast for a 60,000 gain in payrolls was the most accurate. “People were digging everywhere to predict better growth. If there were something technical we would’ve picked it up.”