Credit Suisse’ head of US rates Carl Lantz previews today’s NFP report from the perspective of one data series that gets virtually zero mention in the media: The Data Collection Rate.
The most important payroll number today (that you never look at)
Each month the BLS tells us, buried in an oft-ignored table, what percentage of businesses surveyed returned a response in time for the first payroll print. Despite payrolls remaining an intensively revised number (part of the reason we usually advocate fading an overreaction in the market), this data collection rate has climbed steadily over the years. The attached chart shows the first-print data collection rate for October going back to 1981. Collections have risen from about 40% to above 70% over this period.
If today’s number is to have near-full credibility we should see the collection rate around 74% – the current trend level. It is very likely that this collection rate will be lower due to the technical difficulties in receiving incoming electronic surveys during a period (the government shutdown) when many BLS systems were turned off.
October collections have generally clung to the uptrend seen in the chart with the notable exception in 2001 where presumably disruptions related to the 9/11 attacks had an impact on collections. The ultimate revision to this number was +88k.
Another big drop occurred after huricane Katrina. Collections were well below trend in the autumn of 2005. The worst collection rate then was in November 2005 (presumably lagged effects of business disruptions) and the ultimate revision from the first payroll print was +120K.
While payrolls should be much “cleaner” than the u-rate as furloughed government workers will be counted as employed in the former but not the latter, we need to look at the data collection rate to see just how clean this dirty shirt actually is.
via Zero Hedge http://feedproxy.google.com/~r/zerohedge/feed/~3/kVKlhltDwdE/story01.htm Tyler Durden