Automatic Data Processing Inc., the payroll crunching firm better known simply as ADP, is updating its notorious monthly hiring indicator Thursday.
After a considerable amount of bad press over big misses compared with official Bureau of Labour Statistics (BLS) data, ADP brought in Moody's Analytics to overhaul its methodology. The ADP survey now includes 406,000 of the firm's clients, from 344,000, and 23 million employees from 21 million previously. ADP says its sample represents 20 per cent of all U.S. private sector employees.
ADP and Moody's say their new approach has a 96 per cent correlation with the Labor Department's monthly surveys.
Economists say they will wait and see.
"ADP seems to be proud of its 96 per cent correlation between change in ADP new numbers and change in private payrolls by the BLS," Matthieu Arseneau, an economist at National Bank Financial in Montreal, said Wednesday. "They forgot to mention that with its old methodology the correlation was not less than 95 per cent and discrepancies were sometimes significant."
Mr. Arseneau said the average miss of the old methodology was 61,000 and the new methodology would have produced an average miss of 54,000.
That's something, but probably not enough to convince investors to get ahead of the official data, which will be released Friday.
But while the number may not move markets, it will certainly induce plenty of anticipation. The Friday jobs report is the last before a presidential election on Nov. 6 that revolves almost entirely around the strength of the economy.
In September, the unemployment rate fell below 8 per cent for the first time since the first month of Barack Obama's presidency. Many on Wall Street called that a fluke.
We'll find out for sure on Friday with the October numbers. Until then, for better or worse, we'll have ADP to talk about.