Since lately it seems that the BLS' sole mandate is to smooth and otherwise seasonally adjust noisy, incomplete employment data, even as the one part of employment that should not be rising - part-time jobs - continues to creep ever higher both seasonally adjusted and otherwise, it has become virtually impossible to get an accurate, unadjusted perspective of how the US labor force is truly doing. And with this Friday's non-farm payroll report "the most important ever", as every other month of course, this time because it will seal the fate of the taper (or won't) depending if the data is bad enough, everyone is eager to get a sense of what is going on behind the scenes and how to position for Friday. Unfortunately this Wednesday's ADP private payrolls report will be quite inconclusive as well: not only has it become a laughable predictor of the immediately following NFP data, it does not even disclose what the unadjusted data actually is.
Which leaves us with a third, and final, approach to analyze jobless trends in the US: Gallup, which every week polls thousands of adults to get an unadjusted, snapshot picture of who has a job, who has a part-time job and who has no job. The latest weekly results are good and bad. Good for those who fear that the NFP print on Friday will be so bad Bernanke will have no choice to delay (or reduce) the taper; bad for the economy. Because at 8.5%, unemployment for the week ended September 1 is now near the highest levels it has been in one year, following a spike in mid-August that sent it all the way to 8.8%.
And while there are seasonal variations in the data set, it is notable that on a year over year basis, unemployment one year ago only peaked at 8.4%. It has risen in a gradual at first, and then sharp, fashion since then.
Whether the NFP number will show a comparable deterioration in its own data set is debatable and quite unlikely: thanks to seasonal smoothing, it is very probably that the goalseeked number the BLS has in mind for August is already being being "built." However, while that particular NFP print at best impact machines scanning headlines and responding in the stock market, it is unlikely that the Fed, which by now has certainly made up its mind if it will taper or not, will care much if the jobs print goes +/-100K at this point.
Which means that while the Gallup data is largely useless from a policy, or trading perspective, it is very useful in one key fashion: it shows that the US economy continues to deteriorate, and that increasingly more people have no jobs when stripping out the BLS' favorite labor force participation adjustment gimmick.
A breakdown of the Gallup polling methdology can be found here.