It must be something about the end of Summer, but as we noted in our employment report preview yesterday, economists have historically done a horrible job predicting the change in Non-Farm Payrolls for the month of August. Using our Economic Indicator Database, which is available to all Bespoke Institutional clients, we screened for every Non-Farm Payrolls report dating back to 1998 to see how the initial report came in relative to expectations. As shown in the table below, the initial print of the August report has come in weaker than expected in sixteen of the last twenty years, and the average spread between the actual first reported number and the consensus estimate was a miss of 37K. Based on the average, you could argue that this month’s 24K miss was a good print on a relative basis!
