JOHN WILLIAMS’ SHADOW GOVERNMENT STATISTICS

 FLASH UPDATE

May 8, 2009

 __________

 Better-Than-Expected April Jobs Report Had A Bad Odor to It

539,000 Jobs Loss was 605,000 Net of Revisions, 491,000 Net of CSFB

Birth-Death Bias Showed Unusual Jump in April

 __________

PLEASE NOTE: The next planned Flash Update will follow the release of the April retail sales report on Wednesday, May 13th, with a subsequent update following the April CPI report on Friday, May 15th.

– Best wishes to all, John Williams

CBS news radio this morning (May 8th) was headlining and hyping a likely improvement in the jobs picture, well before the April employment report was released. Where the White House formally received the employment detail after the markets closed on Thursday (and probably had a good sense of the number a week before), today’s reporting looked very much like an orchestrated event. News organizations usually are pretty conservative about touting market-moving reports in advance of a release.

Continuing a pattern seen in the last seven monthly payroll reports, today’s estimates included negative revisions to the previously-report February and March payroll changes (see the Reporting/Market Focus in the most recent SGS Newsletter No. 50, for further background on this indication of flawed reporting), but the Concurrent Seasonal Factor Bias (CSFB) reversed in April (see below). There also was an unusual surge in birth-death modeling bias. Separately, unusual seasonal adjustments were apparent in the unemployment report, which, unlike the payroll reporting, was exactly as bad as expected by consensus forecasts.   

As indicated in the May 4th Flash Update, "Something shy of the consensus jobs decline — a decline less severe than the month before — would meet both political and short-lived financial-market needs…" The consensus outlook had been for a 620,000 jobs loss, and it narrowed to a 600,000 jobs loss (per Briefing.com) following the sharp narrowing in the April jobs decline estimated by ADP, on Wednesday (May 6th). The ADP results generally have not been overly predictive of government reporting, and appear to be of suspect quality. Reinforced by the ADP report, however, the government’s better-than-expected 539,000 April payroll decline happily will be touted as "confirmation" of the Administration’s recent shift to a rosy scenario for the economy and of Wall Street’s sales pitch that the worst is behind us.

Unfortunately, later reporting should confirm that the worst is not behind us. As discussed below, even if today’s data were accurate, the news in the April report remained bleak, with steep ongoing monthly and annual deterioration. Despite the pabulum put out for public and market consumption by the Fed, Federal Reserve Chairman Bernanke should have a pretty good understanding by now of both the ongoing and deepening economic and systemic-solvency crises. Accordingly, the chances should be very close to nil for any near-term Fed tightening.   

Payroll Survey. The BLS reported a statistically-significant, seasonally-adjusted jobs loss of 539,000 (down 605,000 net of revisions) +/- 129,000 (95% confidence interval) for April 2009, following a revised 699,000 (previously 663,000) jobs loss in March and a revised 681,000 (previously 651,000) jobs loss in February.

From peak-to-current (the peak month was December 2007; the current month of April also is the short-lived trough of the current cycle), payroll employment has declined by a seasonally-adjusted 5,738,000 jobs, or 4.2%. Year-to-year contraction (unadjusted) in total nonfarm payrolls continued to deepen, down 3.82% in April, versus a revised 3.57% (was 3.56%) in March. The seasonally-adjusted series also continued contracting year-to-year, down by 3.81% in April versus a revised 3.53% (was 3.48%) in March.

The unadjusted annual decline in April payrolls was the worst since July 1958. At the current pace of deepening annual decline, by the June 2009 employment report (due for release in July), the annual percentage contraction in payrolls will be the most severe since the production shutdown following World War II.  

Concurrent Seasonal Factor Bias. The pattern of impossible biases being built into the headline monthly payroll employment reversed with April 2009 reporting, for the first time since last July (see the accompanying graph). Instead of the headline jobs loss of 539,000, consistent application of seasonal-adjustment factors — net of what I call the concurrent seasonal factor bias — would have shown a less-severe monthly jobs loss of about 491,000. Beyond this month’s reverse result, this pattern has generated an upside reporting bias in 10 of the last 12 months, with a rolling 12-month total upside headline-number bias of 1,255,000. A worksheet on this is available upon request. (See the most recent SGS Newsletter No. 50, for further background.)

Birth-Death/Bias Factor Adjustment. An element that helped to improve the April jobs report was the monthly upside bias factor (birth-death model). Never designed to handle the downside pressures from a recession, the model adds a fairly consistent upside bias to the payroll levels each year, averaging about 60,000 jobs per month, assuming the BLS adequately seasonally adjusts for same. The seasonally-adjusted numbers, however, likely were skewed by a large increase in the unadjusted April 2009 bias factor to 226,000 from 176,000 jobs in April 2008. With a relatively small benchmark revision in place for 2008, one would have expected the April bias adjustment to have narrowed, not widened, as shown in the current report. The upside bias factor for March 2009 was 114,000 jobs.

Household Survey.  The usually statistically-sounder household survey, which counts the number of people with jobs, as opposed to the payroll survey that counts the number of jobs (including multiple job holders), showed April employment rose by an unbelievable 120,000 versus a decline of 861,000 in March. At work here were poor quality seasonal adjustments.

The April 2009 seasonally-adjusted U.3 unemployment rate, however, still showed another statistically-significant increase, to 8.87% +/- 0.23%, from 8.54% in March.  Unadjusted U.3 eased to 8.6% in April from 9.0% in March.  The broader April U.6 unemployment rate rose to an adjusted 15.8% (fell to 15.4% unadjusted) from 15.6% (16.2% unadjusted) in March. The less than proportionate seasonally-adjusted increase in the U.6 measure, versus the U.3 measure, again reflected seasonal factor distortions. The adjusted U.3 unemployment rate likely should have been reported higher than 8.9% and should catch-up in the reporting of the next month or two.

During the Clinton Administration, "discouraged workers" — those who had given up looking for a job because there were no jobs to be had — were redefined so as to be counted only if they had been "discouraged" for less than a year. This time qualification defined away the bulk of the discouraged workers. Adding them back into the total unemployed, unemployment in line with common experience, as estimated by the SGS-Alternate Unemployment Measure, rose to about 20.0% in April versus 19.8% in March.

Employment Environment. Based on the better-quality underlying series discussed below (all are leading indicators to the jobs report) and updated for new data since Monday’s (May 4th) Flash Update, the April jobs loss should have exceeded 750,000. It still may end up reflecting such a loss in subsequent revisions.

New Claims for Unemployment Insurance:Despite the regular weekly volatility of the series, the ongoing rapid rise in initial claims for unemployment insurance has continued to reflect the severe deterioration in labor market conditions, where a rising growth trend in new claims is an economic negative. On a smoothed basis for the 17 weeks ended May 2nd, as reported yesterday (May 7th), year-to-year growth hit 76.8%, its highest level since the 1975 recession and still closing in on its historical peak growth rate of 78.8% in March 1975. The latest annual growth was up from 75.8% the prior week, up from 73.2% as of the 17 weeks ended April 4th, and up from 66.9% in the 17 weeks ended February 28th. A year ago (May 3, 2008) claims were increasing at an annual pace of 12.2%.

Help-Wanted Advertising (HWA):The March help-wanted advertising index (newspapers) collapsed to an historic low of 10 (lowest since the creation of the series in January 1951), from the prior record low of 12 that had held for three months through February. The Conference Board measure was down year-to-year by a record 47.4% in March, versus a 42.9% decline in February. 

Despite some of the historic weakness in the newspaper series being due to the loss of ads to the Internet, the HWA newspaper index remains a solid leading indicator to the broad economy and to the monthly employment report. It has continued to signal severe deepening in the recession and ongoing deterioration in labor-market conditions. The nascent online surveys are telling a similar story. The deepening annual fall-off in new online help-wanted advertising (Conference Board) has continued, down 38.6% in April, versus a 36.0% year-to-year decline in March. The Monster.com online survey estimated that online jobs offerings were down 29% year-to-year in March, versus a 26% decline in February.

Purchasing Managers Surveys:The April manufacturing employment index rose to 34.4 from 28.1, but it did not show an improving economy so much as it indicated one where deterioration had eased. Rapidly declining activity eventually tends to bottom-bounce, as business falls to a low-level plateau of activity. In the employment number, survey participants indicated employment conditions in April as follows: higher 7%, same 58%, lower 35%. Such contrasted with March’s: higher 8%, same 41%, lower 51%. The reported improvement came from a reduction in those reporting lower employment (shifting to same or higher), but the higher percentage also dropped. So, as would be typical with bottom-bouncing, part of the downside turned to unchanged, even the though the upside still deteriorated. Separately, seasonal adjustments distort the reporting, which best is viewed on a three-month moving-average basis in terms of assessing significant changes. On that basis, the April employment reading was 29.5 versus 28.0 in March. The March reading was the lowest since the series was started in January 1948.

Covering the real estate and banking industries, among others, the April services-sector employment component showed patterns similar to the manufacturing sector, rising to 37.0 from 32.3 in March. The services series only has a history back to 1997.

Week Ahead. Trade Deficit:With potential impact on the next revision to the first-quarter GDP estimate, the March trade deficit is due for release on Tuesday, May 12th. Some widening is likely, as per consensus, due simply to an upturn in oil prices. This increasingly volatile and distorted series, however, is due for significant deterioration in catch-up reporting. Results significantly worse than expected would suggest some downside revision pressures on the next GDP update, and vice versa.

Retail Sales:Due for release on Wednesday, May 13th, April retail sales should continue showing record or near-record year-to-year contraction. Reported month-to-month change, despite market hype of a rebounding economy, likely will not be statistically significant either as an increase or as a decline. 

Industrial Production:Due for release on Friday, May 15th, April industrial production also should continue showing record post-World War II year-to-year contraction. Consensus (per Briefing.com) is for a 0.6% monthly contraction. Reality likely is somewhat worse, despite recent bottom-bouncing in the purchasing managers survey.

PPI: Due for release on Thursday, May 14th, the April producer price index should reflect some upside pressure from rising oil prices, although much of that may be tempered by seasonal adjustments. The series is regularly volatile and was underestimated in March, suggesting the potential for some upside catch-up in April.

CPI:Due for release on Friday, May 15th, the April seasonally-adjusted consumer price index (CPI-U) is expected to be unchanged month-to-month, per Briefing.com. The 4.5% monthly average increase in April gasoline prices (per Department of Energy) largely will be muted by seasonal adjustments.

If the monthly CPI-U is unchanged, the pace of annual deflation should deepen a little. Annual inflation would increase or decrease in April 2009 reporting, dependent on the seasonally-adjusted monthly change, versus the 0.15% monthly increase seen in April 2008.  The difference in growth would directly add to or subtract from March’s annual inflation rate of negative 0.38%.

Nonetheless, the continued rise in oil prices — in response partially to near-term softness in the U.S. dollar — increasingly should begin to fuel a non-demand-driven increase in the pace of consumer inflation in the months ahead.

 __________