The ADP private sector job report came out Wednesday for the month of September. ADP showed a LOSS of 39,0000 private sector jobs in September--not the only month this year for which ADP has reported a loss of private sector jobs. ADP is the largest Private payroll processor in this country--probably the world. It has been putting out this monthly report on employment for years, and knows what it is doing. My friend Sylvia works for ADP in El Paso (although she would like to do it in Portland, which she prefers to El Paso). That should not be held against ADP. I am confident they did not know about Sylvia. I should have warned them. I digress.
I trust the ADP report more than I trust the government report due on Friday for one very good reason: The government report is a SUBJECTIVE number (not simply counting), which is seasonally adjusted using Labor Department formulas. It is also revised next month (as may be the ADP report). The Obama Administration, of course, has a massive incentive to fudge the report on Friday of jobs gained or lost for September, as well as the unemployment rate (also a subjective number). I am not sure the extent to which the ADP report is "adjusted", but I trust them more than I trust the government. Call me a cynic. Perpetually surprised "economists" and "analysts" had expected, by the way, that the ADP number would show a GAIN of 24,000 jobs in September.
Lost in all of this is that the numbers over this entire year on employment are BAD, and the ADP report--plus the weekly numbers on new unemployment claims--indicates that we are still not improving (no matter what the government reports on Friday). The unemployment rate, of course, remains at the same level as JANUARY. The weekly number for new unemployment claims remains at the same level as the average for all of DECEMBER. That average for December was 455,0000, and the present four week average is right around 460,000. A new weekly number will be released on Thursday, but will not change the fact that there has been NO IMPROVEMENT in the weekly number for new unemployment claims (a measure of layoffs) for more than NINE MONTHS. All of this is already set in stone, barring some miracle (or fudging). President Obama keeps saying that there has been an eight month streak of private sector job gains, but that is fundamentally not true. The ADP numbers show you why, There is little difference between a 24,000 job gain and a 39,000 job loss. Those are SMALL numbers in our economy, and that has been true of many of the monthly job "gains" for which Obama is claiming "credit". The "gain" has often been essentially zero, when you consider the inherent margin of error. These are not concrete numbers. In very few of the past eight months has there been a private sector job growth of as much as 100,000. Yet, we need a monthly AVERAGE of about 125,000 just to stay even with the number of people coming into the labor force. The employment picture cannot improve without MUCH better job growth.
Before this year, of course, there was no job growth at all. There were only massive job losses through most of 2009 (when Obama was already President). And we have made NO IMPROVEMENT in all of 2010, except for that tepid private sector "growth" in jobs not nearly sufficient to improve the job picture. This is total failure.
I did not do my usual article on last Thursday's weekly report on new unemployment claims, because it would merely have been the same article I have written ever since December. No improvement. Yes, the weekly reported number went down 12,000--to 453,000--but that merely returned the number to where it was the week previous to the increase to a reported number of 465,000. Note that the media headlines said that the number fell 16,000, but that was a LIE--comparing apples and oranges. The 465,0000 reported two weeks ago was REVISED upward to 469,0000. This has been a steady trend--revising the number up 3 or 4 thousand. If you compare apples and apples, you have to compare the 465,000 with the 453,0000 (the unrevised numbers reported the past two weeks). If the pattern continues, the 453,0000 will be revised on Thursday to 456,0000 or so. THAT is the only number that can be compared to the 469,0000--comparing revised number with revised number. But all of these details hardly matter, because there has been NO IMPROVEMENT over more than nine months.
Now I see these things more clearly than either the mainstream media--or even the whole media--and Rush Limbaugh. There was a spike UP in the weekly number of new unemployment claims in August. The four week average actually increased from 450,000-470,000 to 480,000 to almost 490,000. Since December, the range had been between 445,000 and 490,000 for the weekly number. In August, there was a one week spike to 504,0000. This might have indicated we were falling off a cliff, but the spike quickly dissipated--indicating it was not "real" (in the sense of changing the pattern since December).
What happened? There were a lot of government programs over the summer that ended around the end of August. The Census temporary hiring--although clearly not generally a part of the unemployment claims numbers--ended. That could have had some indirect effect on the August numbers. Then states were cutting back on employment--of teachers and others. The school year, of course, starts at the end of August, and state cutbacks seem to have keyed off the coming of Labor Day. Then Labor Day was later in September than usual. All of this may have messed up the "seasonal adjustments", or just have contributed to a misleading spike in new unemployment claims (after seasonal adjustment). There was a previous FALSE 429,000 number in early July, caused by GM not closing its plants for the usual two weeks at the beginning of July. As stated, these are NOT concrete numbers. They are adjusted, subjective numbers. The quick reversion of the weekly number into the same pattern that has existed since December indicates that nothing fundamental happened in August.These quick jumbps--even over a month--up and down indicate statistical glitches rather than a real "trend". A real trend is consistent over a greater time.
The consistnet TREND since December has been NO IMPROVEMENT. There has been a slight deterioration since the beginning of May, since the weekly number has not dropped below 450,000 since then (throwing out that false 429,0000 number). In other words, the low end of the range has seemed to drop to about 455,000 instead of about 445,0000. However, this is small enough as to not suggest much other than that we are NOT IMPROVING.
There you have it--a solid basis on which to evaluate both tomorrow's weekly new unemployment claims number (for last week), and the employment numbers which come out on Friday. Given the consistent trends, over moe than nine months, you should be suspicious of any dramatic changes. We are STUCK in the mud, and it is effectively impossible for any number on Friday to change that fact.