When it comes to inflation data, there are two parallel sources: the BLS, and ShadowStats' John Williams, who continues to plough through the underlying "data" using pre-pre-pre-revision protocols, and every month reveals a parallel universe in which something shocking is revealed: the truth. Here is his take on the October "weaker but really stronger than expected" jobs numbers. Here is what really happened.
Never Recovered, the Economy Remains in Terrible Shape. The large number of opening headlines in today’s (November 9th) missive reflects various stories, ranging from twisted unemployment data, to an election dominated by underlying economic reality, and to headline 2014 financial results on the federal government’s operations that should raise some troubling questions in the markets. The general outlook is unchanged
Twisted Unemployment Numbers. Headline October 2014 unemployment reporting, in particular, was skewed heavily by warped seasonal-adjustment factors that do account properly for last year’s government shutdown. When the U.S. government closed in October 2013, the shutdown encompassed the Bureau of Labor Statistics (BLS) base-period for determining the unemployment and employment detail in the household survey, as well as for determining employment in the payroll survey. The BLS was unable to determine fully the impact of the government shutdown on the monthly October labor data.
For last year’s October 2013 headline payroll-employment survey, the shutdown’s impact generally was guesstimated by the BLS, but it was not reflected in the headline reporting of the time. For the headline unemployment and employment detail, some government employees were counted among the unemployed as being on temporary lay-off, some were just counted as employed but absent from work. Where the headline employed dropped by 735,000, unemployed only rose by 17,000 in October 2013 (see Commentary No. 572 and Commentary No. 580 for the analysis of the time). Where all government employees should have returned to work by the November 2013 reporting, the headline employed rose by 818,000, the unemployed dropped by 365,000.
The BLS never attempted to correct its data, and heavy distortions to the regular seasonal-adjustment process were a virtual certainty. As revised in December 2013, only for the resetting of seasonal adjustment factors—not adjusting for poor-quality reporting—seasonally-adjusted household data now show employment dropping by 785,000 in October 2013 and rebounding by 958,000 in November 2013.
Distorted Numbers One Year Later. The seasonally-adjusted 683,000 jump just reported in headline October 2014 household-survey employment largely was an offsetting seasonally-adjusted artefact of those 2013 events. The October 2013 plunge in household employment, and the counter-adjusted jump in October 2014 employment, can be seen later in these Opening Comments in the Employment and Unemployment section, in the graphs of Civilian Employment Level and the Civilian Employment-to-Population Ratio.
Employment Should Fall Sharply and the Unemployment Rate Should Rise in November. Reversing the swings seen in 2013, November 2014 likely will move sharply in the opposite direction from October, as the reporting and adjusting imbalances catch up. Watch for an offsetting sharp headline decline in the November 2014 household-survey employment number, and an increase in the November headline unemployment rate.
Otherwise, the headline October 2014 payroll employment and unemployment reporting suffered their regular distortions and non-comparable month-to-month estimates, where the non-comparability results from BLS reporting policies tied to the concurrent seasonal-factor adjustment process (see Reporting Detail section), with general detail and misreporting summarized later in these Opening Comments.