In April we noted that economist Mark Zandi's failure to accurately predict the monthly employment numbers had not stopped him from issuing predictions, nor had it stopped journalists from quoting him.
In February Politico quoted him as predicting "The January employment report will be on the soft side. I expect payroll employment to increase by just over 100K." In fact the payroll survey jumped by 243,000.
In April the Wall Street Journal quoted Mr. Zandi as an optimist in the context of a "consensus expectation" for 210,000 new jobs for March. In fact the payroll survey showed only a 120,000 increase.
This morning Politico reported:
MOODY'S MARK ZANDI emails: "All of the data that lead the BLS report suggest a much stronger gain in June than last month's 79,000 job gain. This includes ADP, the ISM surveys, UI claims, help-wanted advertising and Challenger layoff announcements. I think there are better than even odds of an upside surprise tomorrow. ...
FROM ZANDI's client notes : "I expect payroll employment to increase by 125,000 in June...
The actual June number announced this morning was 80,000.
As we said in April:
At a certain point, one starts to wonder why anyone pays any attention to these forecasters, or to the press that reports on them. Complex systems are hard to predict. Mr. Zandi doesn't seem particularly accurate at predicting the monthly payroll numbers. Neither do the professional journalists who do so for a living at places like Politico and the Wall Street Journal. Here's one prediction you can bank on: next month, when it is time for the Bureau of Labor Statistics to roll out its monthly numbers, there will be a preview article in some major news outlet that quotes Mark Zandi. Chances are, it won't be any more accurate than these two examples.
Make that three examples.
And I don't mean to pick on Mr. Zandi; I don't think he's that much worse than a lot of the other people who try to make these predictions.