No. 589: December Employment and Unemployment, Money Supply M3

COMMENTARY NUMBER 589
December Employment and Unemployment, Money Supply M3

January 10, 2014

__________

Jobs Loss or Jobs Gain, Either Is Possible Within the
Reporting-Confidence Interval Around December Payrolls

Revisions Show Headline Unemployment Changes Are Meaningless

December Unemployment: 6.7% (U.3), 13.1% (U.6), 23.3% (ShadowStats)

Year-to-Year Growth Slows in December M3
__________

PLEASE NOTE: The next regular Commentary is scheduled for Tuesday, January 14th, covering December retail sales and consumer liquidity, followed by Commentaries on the 15th (December PPI and a preview of the new series), 16th (December CPI, real retail sales and earnings) and the 17th (December housing starts and industrial production).

Best wishes to all — John Williams

OPENING COMMENTS AND EXECUTIVE SUMMARY

Early December Data Suggest Some Slowing Activity. Lingering distortions in economic reporting—from the government shutdown—should resolve themselves in the next month or two. Today’s (January 10th) Bureau of Labor Statistics (BLS) release on December 2013 labor conditions may have been an early example of reporting catch-up, with both payroll- and household-survey details turning increasingly consistent with a renewed downturn in broad economic activity.

The decline in the headline U.3 unemployment rate, from 7.0% to 6.7%, was not good news. The large drop in the number of unemployed mostly reflected people becoming “discouraged” and being statistically removed from the headline labor force, instead of finding jobs and returning to work. The increasing flow of discouraged workers through the broader U.6 measure, into the ShadowStats-Alternate Unemployment measure, boosted the ShadowStats unemployment rate to 23.3% from a revised 23.1%.

The household-survey (unemployment) also was subject to its annual benchmark revision for seasonally-adjusted data; the unadjusted data are not and were not revised. As discussed in the section on unemployment revisions, the BLS practice of using unique concurrent seasonals for reporting the current headline unemployment rate, but not publishing the consistent history that also is re-calculated each month, is tantamount to reporting fraud (see the Concurrent Seasonal Factor Distortions section for further detail). The revisions show that this reporting practice leads to destabilized headline month-to-month unemployment rate changes, becoming meaningless and misleading indicators.

In December payroll-employment reporting, the headline 74,000 monthly jobs gain was statistically insignificant, even allowing for the moderate upside revision to November’s data. A headline contraction was an easy possibility here, within the normal reporting-error range of the series. Separately, in the booming construction trade, where recent industry reporting appeared to have been bloated by the shutdown disruptions, employment turned down.

The employment and unemployment data usually are the first major economic series published in a month, and they often set the tone for data that follow. As will be discussed with the next Commentary No. 590, on January 14th, recent measures and indicators of consumer liquidity continue to show heavy structural-liquidity impairment (income, credit and confidence) for the consumer. Without a healthy consumer, there has been and can be no broad economic recovery.

Separately, the ShadowStats-Ongoing M3 Money Supply Estimate showed an unusual slowing of annual growth in December, as discussed in the Hyperinflation Watch section.

Revisions Show Monthly Change in Unemployment Usually Is a Meaningless Number. The BLS published annual revisions to the household-survey, today. The revised unemployment-rate detail is covered in this section. Other household survey detail is covered in the Reporting Detail section, and all graphs related to the household survey reflect the revisions in place, unless indicated otherwise. Aside from the concurrent-seasonal-factor issues, revisions to population estimates with January 2014 household data reporting will leave the January 2014 household numbers inconsistent with those of December 2013.

The headline monthly change in the U.3 unemployment rate is a meaningless indicator of economic activity. This is so only because of the deliberate misreporting of the series by BLS, as discussed below.

The accompanying graphs show annual revisions to the seasonally-adjusted headline unemployment rate going back five years, to January 2009, but the term “annual” is something of a misnomer. As published, the revisions are nonsensical and without meaning beyond the month of December 2013. Next month’s January data also will be revised back five years. December’s numbers then will be obsolete, but the BLS never will publish the January detail. As frequently discussed in these Commentaries on unemployment, the monthly, seasonally-adjusted household data simply are not comparable on a month-to-month basis. Each month’s number and prior history are estimated with concurrent seasonal-factor adjustments, unique to the month. The current month’s estimate is published on that basis, but the revised prior data are not.

A short-lived correction is seen only once per year, when the December data are published along with five-years of data recast for the unique December seasonals. For example, in November 2013, the headline seasonally-adjusted U.3 unemployment rate was 7.02%, based on November’s unique seasonals. The December 2013 reporting was for a headline unemployment rate of 6.68%, based on December’s seasonals, but November shifted to 6.98%, based on the December seasonals, not that big a change, with November still rounding out to 7.0%.

Consider, though, January and February 2012. As reported in February 2012, headline February U.3 was 8.27% (rounded to 8.3%), versus a non-comparable 8.26% (rounded to 8.3%) in January, with no month-to-month change in rounded headline unemployment. In the December 2012 “annual” revision, February and January 2012 revised respectively to 8.27% and 8.26%, unchanged at the second decimal points, with both still rounding to 8.3%. Come the December 2013 “annual” revision, February and January 2012 revised respectively to 8.32% and 8.20%, rounding respectively to 8.3% versus 8.2%, with February unemployment now rising by 0.1% versus January. (See second graph following, “Net Revision …”)

That revision was based fully on the December 2013 concurrent seasonal factors, where some elements of current seasonals have no relationship to the 2012 period. Yet, the new and consistent reporting back into 2012 likely would have had a different impact on the financial markets, it had been published, than was seen with the headline, inconsistent numbers of the time. Nothing changed in the reporting of the basic, underlying data. The not-seasonally-adjusted data do not get revised from their survey levels. The only thing that changed here was that the BLS recalculated the concurrent seasonal factors as of December 2013, and that changed the pattern of all historical unemployment reporting, back to January 2009.

The revision volatility, based on the unique December 2013 seasonals, versus the headline numbers for 2013, and the December 2012 revisions to earlier headline numbers, are shown in the accompanying graphs. In the second graph of “Net Revisions,” changes of 0.5% or more mean that the month-to-month change in the headline unemployment rate likely changed in revision at the first decimal point, from its last reporting. Where no one but the BLS regularly knows what the actual, consistent numbers are on a monthly basis, the headline reporting here is unstable enough to make the monthly-change data worthless as indicators of near-term economic activity, and often to mislead the financial markets.

The unreported shifts in monthly seasonal adjustments also would allow for a simple rigging of the unemployment rate, where changed seasonals could shift growth from one month to another. While there is no apparent political pattern in the just-published revisions, that does not mean a thing. The new data reflect only the concurrent seasonal factors in place as of December 2013. What the adjustments actually were before the 2012 election, or what they will be going into the 2014 election, no one but the Bureau of Labor Statistics knows, and the BLS will not release that information, because, as the BLS has expressed it, they do not want to “confuse” their data users.

Again, these revisions have nothing to do with changes to the underlying data, they simply follow the high level of monthly volatility from the concurrent seasonal adjustment recalculations that are calculated but not published each month. To restore an honest accounting to the public, the BLS either should publish its monthly numbers on a consistent basis, or go back to using fixed seasonal adjustments, where the reporting of monthly data remains consistent over time.

Headline Unemployment Data. Headline unemployment (U.3) fell to 6.68% in December, from 6.98% (previously 7.02%) in November and 7.20% (previously 7.28%) in October. The 0.30% +/- 0.23% (95% confidence interval) decline in headline December unemployment was statistically significant, in this once-per-year event, when the December and November numbers are consistently prepared. December’s unadjusted U.3 unemployment rate was 6.5%, versus 6.6% in November and 7.0% in October.

Again, though, the decline in headline U.3 was not good news. As discussed in Commentary No. 521 and Commentary No. 554, instead of reflecting those who are unemployed finding jobs, the lower headline U.3 rate reflects those who are unemployed being defined out of the government’s unemployment measurement by restrictive definitions. Those leaving the headline labor force usually end-up moving to the broader accounting of the U.6 and ShadowStats measures.

U.6 Unemployment Rate. The broadest unemployment rate published by the BLS, U.6 includes accounting for those marginally attached to the labor force (including short-term discouraged workers) and those who are employed part-time for economic reasons (i.e., they cannot find a full-time job).

A seasonally-adjusted increase in people working part-time for economic reasons and an increase in short-term (unadjusted) discouraged workers, offset the decline in headline U.3 unemployment, with the headline December 2013 U.6-unemployment at 13.1%, versus a revised 13.1% (previously 13.2%) in November and a revised 13.7% (previously 13.8%) in October. The unadjusted December U.6 rate rose to 13.0%, from 12.7% in November and versus 13.2% in October.

Discouraged Workers and the ShadowStats-Alternate Unemployment Measure. The count of short-term discouraged workers (never seasonally-adjusted) was 917,000 in December 2013, up from 762,000 in November and versus 815,000 in October, likely reflecting an increased rollover of short-term discouraged workers to the long-term discouraged category

The current, official discouraged-worker number reflected the flow of the unemployed—increasingly giving up looking for work—leaving the headline U.3 unemployment category and being rolled into the U.6 measure as short-term “discouraged workers,” net of those moving from short-term discouraged-worker status into the netherworld of long-term discouraged-worker status. It is the long-term discouraged-worker category that defines the ShadowStats-Alternate Unemployment Measure. There appears to be a relatively heavy, continuing rollover from the short-term to the long-term category.

In 1994, “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 a large number of long-term discouraged workers. The remaining short-term discouraged workers (those discouraged less than a year) were included in U.6.

Where the U.6 measure includes short-term discouraged workers (those who have not looked for work in the last four-weeks, but have looked in the last year) and those working part-time for economic reasons. After being discouraged for a year or more, the short-term discouraged workers become long-term discouraged workers and move to the ShadowStats-Alternate Unemployment measure. More-complete definitions—including discussion on the increasing divergence between the ShadowStats number and U.3 and U.6—are found near the end of the Reporting Detail section.

The first graph following reflects headline December 2013 U.3 unemployment at 6.7%, down from 7.0% in November and versus 7.2% in October; headline U.6 unemployment at 13.1% in December, 13.1% in November and 13.7% in October; and the headline ShadowStats unemployment measure at 23.3% in December, 23.1% (previously 23.2%) in November and 23.4% in October. October is the new series high (since 1994), previously at 23.5%.

Two graphs follow the traditional plot of the three mentioned unemployment rates. The first graph (the second graph following) is of the ShadowStats unemployment measure, with an inverted scale. The higher the unemployment rate, the weaker will be the economy, so the inverted plot tends to move in tandem with plots of most economic statistics, where a lower number means a weaker economy.

The inverted-scale ShadowStats unemployment measure also tends to move with the employment-to-population ratio, which is plotted in the third graph following. Discouraged workers are not counted in the headline labor force, which continues to shrink. The labor force containing all unemployed (including total discouraged workers) plus the employed, however, tends to be correlated with the population, so the employment-to-population ratio tends to be something of a surrogate indicator of broad unemployment, and it has a strong correlation with the ShadowStats unemployment measure.

These graphs reflect detail back to the 1994 redefinitions of the household survey. Before 1994, data consistent with today’s reporting are not available.

Headline Payroll Employment. December payroll data were published in the context of an upside revision of 38,000 to the November payroll level, with the adjusted October and September levels unrevised. Ongoing or catch-up surveying and reporting, offsetting the distortions from the effects of the government shutdown in October are unknown, but reporting-consistency issues continue from the ongoing concurrent-seasonal-adjustment debacle

The seasonally-adjusted, month-to-month headline payroll employment gain of 74,000 for December 2013 was well below financial-market expectations. Net of prior-period revisions, that monthly gain would have been 112,000. Where the standard 95% confidence interval around the headline monthly change in the payroll employment is +/- 129,000, the headline jobs increase was not statistically significant. Concurrent seasonal factor reporting issues, however, suggest that a much wider confidence interval could be justified.

The November headline jobs gain was revised to 241,000, with the October headline monthly jobs gain unrevised at 200,000, which became non-comparable and inconsistent with the prior September data, as of the December reporting.

The ongoing payroll-reporting issue here remains that the BLS publishes only two prior months of consistent data with concurrent-seasonally-adjusted payrolls. Accordingly, where the published October number no longer is consistent with September reporting, related month-to-month comparisons have no meaning, given the BLS adjustment and reporting policies discussed in Concurrent Seasonal Factors Distortions in the Reporting Detail section.

The preceding two graphs are updated for the headline payroll levels through December 2013. Even with the annual growth seen in the payroll series since mid-2010, the December 2013 level of employment is shy by 1.2-million jobs, or 0.9% in official reporting, from recovering its pre-recession high of January 2008. In perspective, the longer-term plot of employment levels shows the extreme duration of the non-recovery in employment, the worst such circumstance of the post-Great Depression era.

Not-seasonally-adjusted, year-to-year change is untouched by the concurrent seasonal adjustments, so the monthly comparisons of year-to-year change are reported on a consistent basis. For December 2013, the year-to-year percent gain in payrolls slowed to 1.62%, a seven-month low, from a revised 1.74% in November and an unrevised 1.69% in October (see graphs in the Reporting Detail section).

[For further details on December employment and unemployment, see the Reporting Detail Section.]

__________

HYPERINFLATION WATCH

Annual Growth in December M3 Money Supply Slowed to 3.3%, from 4.1% in November, with the Monetary Base Still Exploding. The ShadowStats-Ongoing-M3 Estimate for December 2013 year-to-year growth is 3.3%, down from a revised 4.1% (previously 4.0%) in November, as published on the Alternate Data tab of www.shadowstats.com. The December annual growth rate reading was the slowest pace seen since September 2012, down from the recent near-term peak in growth of 4.8%, seen in January 2013, at the onset of the expansion of QE3 into purchasing U.S. Treasury securities.

As discussed and graphed in the just-released Hyperinflation 2014—The End Game Begins (page 36), where annual M3 growth had been on the upswing into the expanded QE3, the continuing general pattern now of stagnant-to-falling annual growth, in an environment of continued rapid growth in the monetary base, remains a sign of likely mounting banking-system stresses.

Any revisions in the following numbers are due to benchmark revisions of underlying data by the Federal Reserve. The seasonally-adjusted, preliminary estimate of month-to-month change for December 2013 money supply M3 is for a likely gain of 0.2%, versus a revised gain of 0.3% in November. Estimated month-to-month M3 changes, however, remain less reliable than are the estimates of annual growth.

Initial Growth Estimates for November M1 and M2. For December 2013, early estimates of year-to-year and month-to-month changes follow for the narrower M1 and M2 measures (M2 includes M1, M3 includes M2). Full definitions of the measures are found in the Money Supply Special Report. M2 for December is estimated to show year-to-year growth of roughly 5.5%, versus a revised 6.1% (previously 6.0%), with month-to-month change estimated at roughly a 0.5% gain in December, versus an unrevised 0.2% contraction in November. The early estimate of M1 for December is for year-to-year growth of roughly 8.2%, versus an unrevised 8.4% gain in November, with a month-to-month December gain of 1.4%, versus a revised 0.9% contraction in November contraction.

Hyperinflation Outlook. With Hyperinflation 2014—The End Game Begins published on January 7th, a new Hyperinflation Summary for this section will be added in the near future.

__________

REPORTING DETAIL

EMPLOYMENT AND UNEMPLOYMENT (December 2013)

December Labor Data Began to Stabilize. Discussed in the Opening Comments, the December 2013 statistical release on labor conditions may have been an early example of some reporting catch-up from data gathering and reporting issues tied the government shutdown. Where the unusually violent data swings in October and November numbers may have warped the coincident seasonal factors, the nice thing about the Bureau of Labor Statistics (BLS) misuse of the concurrent seasonal factor process is that the monthly data do not have to be consistent or comparable with each other.

The decline in the headline U.3 unemployment rate (U.3) from 7.0% to 6.7% was not good news. The large drop in the number of unemployed mostly reflected people becoming “discouraged” and moving out of the headline labor force, instead of finding jobs and returning to work. The increasing flow of discouraged workers through the broader U.6 measure into the ShadowStats-Alternate Unemployment measure was reflected in an increase in the ShadowStats unemployment rate to 23.3% from 23.1%.

The household-survey also went through its annual benchmark revision, for seasonally-adjusted data only. The unadjusted data were not revised. As discussed and graphed in Opening Comments section on unemployment revisions, the BLS practice of using the concurrent seasonals for reporting only the current headline number, and not properly publishing consistent, revised history, is tantamount to reporting fraud. The revisions show this reporting practice leads to destabilized headline month-to-month unemployment-rate changes being meaningless and misleading.

On the payroll-employment front, the headline 74,000 monthly jobs gain in December was statistically insignificant. Of considerable interest will be the upcoming February 7th benchmark revision to the payroll data. The will change both the adjusted and unadjusted numbers and will be discussed further, in advance of the release.

PAYROLL SURVEY DETAIL. This morning’s (January 10th) December payroll data were published in the context of an upside revision of 38,000 to the November payroll level, with the adjusted October and September levels unrevised. Ongoing or catch-up surveying and reporting distortions from the effects of the government shutdown in October are unknown, but serious reporting-consistency issues continue, with the ongoing concurrent-seasonal-adjustment debacle

Those issues aside, the BLS reported a seasonally-adjusted, month-to-month headline payroll employment gain of 74,000 for December 2013, which was well below financial-market expectations that appear to have topped 200,000. Net of prior-period revisions, that monthly gain would have been 112,000. Where the standard 95% confidence interval around the headline monthly change in the payroll employment is +/- 129,000, the headline jobs increase was not statistically significant. Reporting issues, however, suggest that a much wider confidence interval could be justified. The current numbers continue to be so far out of balance as to be absolutely meaningless, here, due partially to concurrent-seasonal-factor distortions (discussed in the Concurrent Seasonal Factor Distortions section).

The November headline jobs gain was revised to 241,000 (previously 203,000), with the October headline monthly jobs gain unrevised at 200,000 (initially 204,000), which became non-comparable and inconsistent with the previous September data, as of the December reporting.

The ongoing reporting issue here remains that the BLS publishes only two prior months of consistent data with concurrent-seasonally-adjusted payrolls. Accordingly, where the published October number no longer is consistent with September reporting, related month-to-month comparisons have no meaning, given the BLS adjustment and reporting policies discussed in Concurrent Seasonal Factors Distortions (an issue related to the household-survey reporting of unemployed, as discussed in the Opening Comments section).

Using the latest concurrent seasonal-factor calculations from the BLS, ShadowStats is able to estimate that the consistent, actual revised (but not published) month-to-month gain for October versus September was 205,000, instead of the official 200,000. The month-to-month reporting discrepancies go in both directions and often are greater than the October difference of 5,000, with monthly differential magnitudes approaching 100,000 jobs, on occasion.

The BLS explains that it avoids publishing consistent, prior-period revisions so as not to “confuse” its data users. No one seems to mind if the published earlier numbers are wrong, particularly if unstable seasonal-adjustment patterns have shifted prior jobs growth into current reporting, without any indication of same in the published historical data.

2013 Benchmark Revision. [Further benchmarking detail will be covered in a Commentary prior to the February 7th release. The benchmark revision discussion here has not been revised from the Commentary covering the November labor data, but.] As discussed in Commentary No. 561, of September 26th, the announced benchmark revision to the 2013 payroll survey would be tantamount to fraud, if the entire historical series is not otherwise revamped for a major redefinition of nonfarm payrolls. As standardly reported, the March 2013 benchmarking lowered the payroll levels of that time by 124,000 jobs, instead of the 345,000 “increase” reported, which included 469,000 new workers who were classified and defined previously as not counted in nonfarm payrolls.

Indeed, as it has been configured, the payroll employment level in the benchmark month of March 2013 was found to have been overstated by 124,000 jobs, requiring a downside revision to the series in that month, with adjustments back to March 2012, and with adjustments forward in time through the reporting of January 2014 payrolls (to be released in February 2014). In the later months of the revision cycle, the downside revisions to monthly levels likely would have topped 200,000.

In a turnaround, the announced benchmark revision was restated so as to be to the upside by 345,000, thanks to the inclusion of 469,000 in employment that previously had not been counted as part of the nonfarm payroll survey. Aside from excluding agricultural employment, the payroll survey had excluded those on household payrolls. Now 469,000 of the household payrolls have been moved into the payroll survey, into the education and healthcare industries, and there is no indication that the BLS plans to restate prior history so as to have a consistent historical series.

Further, this is an area that is not surveyed easily by the BLS on a monthly basis, so it becomes a new fudge-factor for re-jiggering the headline payroll numbers. As announced by the BLS:

“Each year, [payroll] employment estimates from the Current Employment Statistics (CES) survey are benchmarked to comprehensive counts of employment for the month of March. These counts are derived from State Unemployment Insurance (UI) tax records that nearly all employers are required to file. For National CES employment series, the annual benchmark revisions over the last 10 years have averaged plus or minus three-tenths of one percent of Total nonfarm employment. The preliminary estimate of the benchmark revision indicates an upward adjustment to March 2013 Total nonfarm employment of 345,000 (0.3 percent). This revision is impacted by a large non-economic code change [made by the BLS] in the Quarterly Census of Employment and Wages (QCEW) that moves approximately 469,000 in employment from Private households, which is out-of-scope for CES, to the Education and health care services industry, which is in scope. After accounting for this movement, the estimate of the revision to the over-the-year change in CES from March 2012 to March 2013 is a downward revision of 124,000.”

Trend Model. As described generally in Payroll Trends and expanded in detail available from our affiliate www.ExpliStats.com, the trend indication from the BLS’s concurrent seasonal-adjustment model is for a 145,000 monthly payroll gain in January 2014, based on December’s reporting.

The trend indication often misses actual reporting. The indication for December was for a 170,000 monthly gain, where consensus forecasts purportedly topped 200,000. Of course, the actual headline gain of 74,000 was much lower. Nonetheless, the trend number becomes the basis for the consensus outlook, more often than not.

Construction Payrolls. The graph of construction employment updates the one accompanying the coverage of November construction spending in Commentary No. 588. Headline December 2013 construction employment lost 16,000 jobs, versus a revised 19,000 (previously 17,000) jobs gain in November. Total December construction jobs were 24.5% shy of the pre-recession peak in April 2006.

Annual Change in Payrolls. Not-seasonally-adjusted, year-to-year change is untouched by the concurrent seasonal adjustments, so the monthly comparisons of year-to-year change are reported on a consistent basis. For December 2013, the year-to-year percent gain in payrolls slowed to 1.62%, a seven-month low, from a revised 1.74% (previously 1.70%) in November, and an unrevised 1.69% (initially 1.70%) in October.

The following graphs of year-to-year unadjusted payroll change reflect near-term detail as well as seventy-plus years of history. Graphs of the seasonally-adjusted payroll levels are found in the Opening Comments. Year-to-year change had shown a slowly rising trend in annual growth into 2011, which reflected protracted bottom-bouncing in the level of nonfarm payrolls. That pattern of annual growth flattened out in late-2011 and began a pattern of slowing-to-flat growth early in 2012, notching lower with the latest numbers.

With the bottom-bouncing patterns of recent years, current annual growth has recovered from the post-World War II record 5.06% decline seen in August 2009. That 5.06% decline remains the most severe annual contraction since the production shutdown at the end of World War II (a trough of a 7.59% annual contraction in September 1945). Disallowing the post-war shutdown as a normal business cycle, the August 2009 annual decline was the worst since the Great Depression.

Still, even with the annual growth in the series since mid-2010, the December 2013 level of employment is shy by 1.2-million jobs, or 0.9% in official reporting, from recovering its pre-recession high. In perspective, the longer-term graph of the employment level (see Opening Comments), shows the extreme duration of the non-recovery in payrolls, the worst such circumstance of the post-Great Depression era.

Concurrent Seasonal Factor Distortions. Reflected in the accompanying graph, seasonal-factor instabilities continued in the latest payroll reporting. Concurrent-seasonal-factor distortions likely were exacerbated in the October-to-December period, due to the effects of the government shutdown on the unadjusted data.

There are serious and deliberate reporting flaws with the government’s seasonally-adjusted, monthly reporting of employment and unemployment (see the discussion of headline-unemployment revisions in the Opening Comments). Each month, the BLS uses a concurrent-seasonal-adjustment process to adjust both the payroll and unemployment data for the latest seasonal patterns. As each series is calculated, the adjustment process also revises the history of each series, recasting prior reporting on a basis that is consistent with the new seasonal patterns of the headline numbers.

The BLS, however, uses the current estimate but does not publish the revised history, even though it calculates the consistent new data each month. As a result, headline reporting generally is neither consistent with nor comparable to earlier reporting, and month-to-month comparisons of these popular numbers usually are of no substance, other than for market hyping or political propaganda.

December Inconsistencies. For concurrent-seasonal-adjustment distortions to the household survey this month, and related revisions, see the Opening Comments section. Despite the one-month of revisions each year, all the monthly data will be inconsistent again with January 2014 reporting.

While the 74,000 jobs gain reported for December was consistent with the revised 241,000 jobs increase in November, on a concurrent-seasonally-adjusted basis, those increases were not consistent with the unrevised headline 200,000 jobs gain reported for October or with any earlier published data. The November number would have been consistent with a 205,000 jobs gain in October.

Unemployment Numbers Simply Are Not Comparable Month-to-Month. Except for the just-published, once-per-year December release of revisions to seasonally-adjusted data, the BLS publishes no other revised seasonally-adjusted data on a monthly basis for the household survey, even though those revisions are made and are available internally to the BLS for publication every month, as part of the concurrent-seasonal-factor process. Come the next round of inconsistent January 2014 reporting, irrespective of the headline change, that unemployment rate—in reality—could have be up, down or unchanged; there just is no way to know from existing BLS reporting.

As discussed frequently (see Commentary No. 473, Commentary No. 461, and Commentary No. 451, for example), the revisions to earlier data from the concurrent-seasonal-factor process can be significant. As a result, month-to-month changes in seasonally-adjusted unemployment rates are meaningless—not determinable under current BLS reporting policies—and use of monthly comparisons simply should be avoided. At this time, the BLS does not make usable, comparative data available to the public.

Payroll Growth Is Consistent Only One-Month Back, With Heavy Distortions Usual. With the payroll series, the level of payrolls is released for the headline month, and for the two prior months, on a consistent basis. That means that only the current headline month-to-month change and the change for the prior month are consistent and comparable. Unlike the household-survey circumstance, however, the BLS makes available the seasonal-adjustment models and data so that others can calculate the payroll revisions, and ShadowStats has done so for the accompanying graph. All these data were reset with the March 2012 benchmark revision, which was published in February 2013. The system (and the accompanying graph) will be reset, again, with the March 2013 benchmark revision to be published with the January 2014 data on February 7, 2014.

Distortions in the current post-benchmark environment are evident, even though the first data were based on the initial public reporting of the benchmark revision. The reason for this is that the BLS actually runs the benchmark revision internally, based on October numbers. With subsequent internal runs in November, December and January, the new accounting will encompass and be skewed by three months of revisions. In the current graph, the line for February reflects only one month subsequent of new seasonal-factor revisions, the March line reflects a second month and so on through December. Without distortions, the plotted lines would be flat and at zero.

Conceivably, the shifting and unstable seasonal adjustments could move 90,000 jobs or more, in either direction (based on last year’s full revisions, and approached by this year’s numbers) in earlier periods and insert them into the current period as new jobs, without there being any published evidence of that happening.

Note: Issues with the BLS’s concurrent-seasonal-factor adjustments and related inconsistencies in the monthly reporting of the historical time series are discussed and detailed further in the ShadowStats.com posting on May 2, 2012 of Unpublished Payroll Data.

As discussed in other writings (see for example Hyperinflation 2012 and to be updated in the second installment of Hyperinflation 2014), seasonal-factor estimation for most economic series has been distorted severely by the extreme depth and duration of the economic contraction. These distortions are exacerbated for payroll employment data based on the BLS’s monthly seasonal-factor re-estimations and lack of full reporting.

Birth-Death/Bias-Factor Adjustment. Despite the ongoing, general overstatement of monthly payroll employment, the BLS generally adds in upside monthly biases to the payroll employment numbers. The continual overstatement is evidenced usually by regular and massive, annual downward benchmark revisions (2011 and 2012, excepted). As discussed in the Benchmark Revision section above, the announced standard benchmarking confirmed an overstatement of currently-defined payroll levels and growth as of March 2013. Without new gimmicks added to the process (again, see the Benchmark Revision section), current reporting would be running at a payroll level roughly 200,000 jobs lower, based on that benchmark

The upside-bias process was created simply by adding in a monthly “bias factor,” so as to prevent the otherwise potential political embarrassment of the BLS understating monthly jobs growth. The “bias factor” process resulted from such an actual embarrassment, with the underestimation of jobs growth coming out of the 1983 recession. That process eventually was recast as the now infamous Birth-Death Model (BDM), which purportedly models the effects of new business creation versus existing business bankruptcies.

December 2013 Bias. The not-seasonally-adjusted December 2013 bias was a monthly add-factor of minus 15,000, versus minus 1,000 in December 2012 and a minus 15,000 add-factor in November 2013. The aggregate upside bias for the trailing twelve months was lowered to 624,000 in December, from 638,000 in November, or to a monthly average of 52,000 jobs created out of thin air, on top of some indeterminable amount of other jobs that are lost in the economy from business closings. Those losses simply are assumed away by the BLS as part of the BDM, as discussed below.

Problems with the Model. The aggregated upside annual reporting bias in the BDM reflects an ongoing assumption of a net positive jobs creation by new companies versus those going out business. Such becomes a self-fulfilling system, as the upside biases boost reporting for financial-market and political needs, with relatively good headline data, while often also setting up downside benchmark revisions for the next year, which traditionally are ignored by the media and the politicians. Where the BLS cannot measure meaningfully the impact of jobs loss and jobs creation from employers starting up or going out of business, on a timely basis (within at least five years, if ever), such information is estimated by the BLS along with the addition of a bias-factor generated by the BDM.

Positive assumptions—commonly built into government statistical reporting and modeling—tend to result in overstated official estimates of general economic growth. Along with happy guesstimates, there usually are underlying assumptions of perpetual economic growth in most models. Accordingly, the functioning and relevance of those models become impaired during periods of economic downturn, and the current downturn has been the most severe—in depth as well as duration—since the Great Depression.

Indeed, historically, the BDM biases have tended to overstate payroll employment levels—to understate employment declines—during recessions. There is a faulty underlying premise here that jobs created by start-up companies in this downturn have more than offset jobs lost by companies going out of business. So, if a company fails to report its payrolls because it has gone out of business (or has been devastated by a hurricane), the BLS assumes the firm still has its previously-reported employees and adjusts those numbers for the trend in the company’s industry.

Further, the presumed net additional “surplus” jobs created by start-up firms are added on to the payroll estimates each month as a special add-factor. These add-factors are set now to add an average of 52,000 jobs per month in the current year. The aggregate overstatement of monthly jobs likely exceeds 100,000 jobs per month.

HOUSEHOLD SURVEY DETAILS. Where the BLS admitted mistakes in the preparation and reporting of the October 2013 household survey data, and where those numbers otherwise were heavily distorted by the effects of the October government shutdown (see Commentary No. 572), the November 2013 reporting is warped not only by a reversal of the October distortions, but also by related, extreme concurrent-seasonal-factor distortions. The December reporting was put in place today, along with revisions to the seasonally-adjusted household data, back to January 2009. For this month only, the historical monthly estimates reflect consistent concurrent seasonal factor adjustment. That circumstance, however, will be overthrown by the reporting for January 2014.

Headline Household Employment. The household survey counts the number of people with jobs, as opposed to the payroll survey that counts the number of jobs (including multiple job holders more than once). On that basis, December 2013 employment rose by 143,000, following a revised 958,000 (previously 818,000) gain in November, a revised 785,000 (previously 735,000) drop in October, and a revised 91,000 (previously 133,000) gain in September. The differences are fully a reflection of the shifting volatility created by the monthly concurrent seasonal factors. The unadjusted data do not change. The October and November data were heavily distorted by the impact of the government shutdown.

Headline Unemployment Rates. The revisions to the headline data are discussed and plotted in the Opening Comments section. Headline unemployment (U.3) fell to 6.68% in December, from 6.98% (previously 7.02%) in November and 7.20% (previously 7.28%) in October.

The official 95% confidence interval is +/- 0.23 percentage-point around the monthly headline change in the U.3 number. Where that usually is meaningless in the context of the comparative month-to-month reporting-inconsistencies created by the concurrent seasonal factors, it suggests a statistically significant decline, for once, in the headline December 2013 U.3 number.

On an unadjusted basis, the unemployment rates are not revised and are consistent in reporting methodology. December’s unadjusted U.3 unemployment rate was 6.5%, versus 6.6% in November and 7.0% in October.

Discouraged Workers. The count of short-term discouraged workers (never seasonally-adjusted) was 917,000 in December 2013, up from 762,000 in November and versus 815,000 in October, likely reflecting an increased rollover of short-term discouraged workers into long-term discouraged workers.

ShadowStats-Alternate Unemployment Rate. Adding back into the total unemployed and labor force the ShadowStats estimate of the growing ranks of excluded, long-term discouraged workers, broad unemployment—more in line with common experience, as estimated by the ShadowStats-Alternate Unemployment Measure—rose to 23.3% in December, from a revised 23.1% (previously 23.2%) in November, and versus 23.4% (previously 23.5%) in October, the series high (back to 1994). The ShadowStats estimate reflects the increasing toll of unemployed leaving the headline labor force. Where the ShadowStats alternate estimate generally is built on top of the official U.6 reporting, it tends to follow its relative monthly movements and its annual revisions. Accordingly, the alternate measure often will suffer some of the same seasonal-adjustment woes that afflict the base series, including underlying annual revisions.

As seen in the usual graph of the various unemployment measures (in Opening Comments), there continues to be a noticeable divergence in the ShadowStats series versus U.6. The reason for this is that U.6, again, only includes discouraged workers who have been discouraged for less than a year. As the discouraged-worker status ages, those that go beyond one year fall off the government counting, even as new workers enter “discouraged” status.

With the continual rollover, the flow of headline workers continues into the short-term discouraged workers category (U.6), and from U.6 into long-term discouraged worker status (a ShadowStats measure). There was a lag in this happening as those having difficulty during the early months of the economic collapse, first moved into short-term discouraged status, and then, a year later into long–term discouraged status, hence the lack of earlier divergence between the series. The movement of the discouraged unemployed out of the headline labor force has been accelerating. While there is attrition in long-term discouraged numbers, there is no set cut off where the long-term discouraged workers cease to exist. See the Alternate Data tab for more detail.

Two further related graphs found in the Opening Comments section are of the ShadowStats-Alternate Unemployment Measure, with an inverted scale, the employment-to-population ratio, which has a high correlation with the inverted ShadowStats measure.

Great Depression Comparisons. As discussed in previous writings, an unemployment rate above 23% might raise questions in terms of a comparison with the purported peak unemployment in the Great Depression (1933) of 25%. Hard estimates of the ShadowStats series are difficult to generate on a regular monthly basis before 1994, given the reporting inconsistencies created by the BLS when it revamped unemployment reporting at that time. Nonetheless, as best estimated, the current ShadowStats level likely is about as bad as the peak actual unemployment seen in the 1973 to 1975 and in the double-dip recession of the early-1980s.

The Great Depression unemployment rate of 25% was estimated well after the fact, with 27% of those employed working on farms. Today, less that 2% of the employed work on farms. Accordingly, a better measure for comparison with the ShadowStats number would be the Great Depression peak in the nonfarm unemployment rate in 1933 of roughly 34% to 35%.

__________

WEEK AHEAD

Weaker-Economic and Stronger-Inflation Reporting Likely in the Months and Year Ahead. At the moment, markets generally are overly optimistic as to the economic outlook, based on data that likely were puffed-up in the process of going through the data-gathering and reporting distortions of the October shutdown to the federal government. Although expectations should soften anew, quickly, there remains the potential for unusual, irregular and eventually corrective reporting and revisions in the months ahead.

That circumstance, and underlying fundamentals that remain highly suggestive of deteriorating business activity, mean that weaker-than-consensus economic reporting should become the general trend.

Stronger inflation reporting is likely. Upside pressure on oil-related prices should reflect intensifying impact from a weakening U.S. dollar in the currency markets, and from ongoing political instabilities in the Middle East. The dollar faces pummeling from continuing QE3, the ongoing U.S. fiscal-crisis debacle, a weakening U.S. economy and deteriorating U.S. political conditions (see Hyperinflation 2014—The End Game Begins). Particularly in tandem with a weakened dollar, reporting in the year ahead generally should reflect much higher-than-expected inflation.

A Note on Reporting Quality Issues and Systemic Reporting Biases. Significant reporting-quality problems remain with most major economic series. Headline reporting issues are tied largely to systemic distortions of seasonal adjustments. The data instabilities were induced by the still-ongoing economic turmoil of the last seven-to-eight years, which has been without precedent in the post-World War II era of modern economic reporting. These impaired reporting methodologies provide particularly unstable headline economic results, where concurrent seasonal adjustments are used (as with retail sales, durable goods orders, employment and unemployment data), and they have thrown into question the statistical-significance of the headline month-to-month reporting for many popular economic series.

PENDING RELEASES:

Retail Sales (December 2013). The December 2013 retail sales estimate is scheduled for release on Tuesday, January 14th, by the Census Bureau. With the consumer still in an extreme liquidity bind (consumer liquidity issues will be updated here, in what will be the accompanying Commentary No. 590), odds favor headline retail sales reporting coming in below-market expectations, which appear to be for moderate growth in the last month of the holiday-shopping season. An outright month-to-month contraction in seasonally-adjusted sales remains a fair possibility for this dominant retail-sales month, even before adjustment for consumer inflation.

Producer Price Index—PPI (December 2013). The December 2013 PPI is scheduled for release on Wednesday, January 15th, by the Bureau of Labor Statistics (BLS). This lame-duck series faces a complete overhaul and redefinition come the release of the January 2014 data in February 2014 (see Commentary No. 582). The new series will be reviewed here, in what will be the accompanying Commentary No. 591.

That said, depending on the oil contract followed, not-seasonally-adjusted monthly-average oil prices, were up by 2.8%-to-4.0% for the month of December, along with a 1.1% increase in average retail gasoline prices. There will be some added inflation gain from positive seasonal adjustments to energy prices. Accordingly, December PPI energy prices should prove to be the dominant upside contributor to the headline inflation.

Overall, a moderate increase is likely in the finished goods PPI, with likely upside contributions from food and “core” inflation adding to the higher energy costs.

Consumer Price Index—CPI (December 2013). The release by the Bureau of Labor Statistics (BLS) of the December 2013 CPI is scheduled for Thursday, January 16th. The headline CPI-U is a good bet to increase by 0.2%, or more, based on gasoline prices and dependent on upside contributions from food prices and core inflation.

Average gasoline prices rose month-to-month in December 2013 by 1.1-percentage points, on a not-seasonally-adjusted basis, per the Department of Energy. The BLS seasonal adjustments will give gas prices a boost. As last revised, an unadjusted 4.2% monthly decline in December 2012 gasoline prices was narrowed to a 1.9% contraction, with upside seasonal adjustments. Similar effects in the December 2013 number, by themselves, would add 0.2% to the headline CPI-U number. Any upside surprise here would come from food prices or core inflation.

Year-to-year, CPI-U inflation would increase or decrease in December 2013 reporting, dependent on the seasonally-adjusted monthly change, versus an adjusted and negligible 0.03% increase in the monthly inflation reported for December 2012. The adjusted change is used here, since that is how consensus expectations are expressed. To approximate the annual unadjusted inflation rate for December 2013, the difference in December’s headline monthly change (or forecast of same), versus the year-ago monthly change, should be added to or subtracted directly from the November 2013 annual inflation rate of 1.24%. For example, if the headline December 2013 CPI-U increased by 0.2%, the new year-to-year inflation would be in the 1.4%-to-1.5% range.

Index of Industrial Production (December 2013). The December 2013 index of industrial production is scheduled for release on Friday, January 17th, by the Federal Reserve Board. Net of the irregular volatility in utility output tied to seasonable or seasonable weather, moderate expectations for December production growth are a fair bet to be disappointed, as companies increasingly move to reduce excessive inventory levels. There also remains the potential for unusual reporting volatility and revisions tied to data disruptions resulting from the October shutdown of the federal government.

Residential Construction—Housing Starts (December 2013). Also on Friday, January 17th, the Census Bureau will publish its estimate of December 2013 housing starts. This series was distorted heavily by data-gathering and reporting issues that resulted from the government shutdown in October. Given last month’s seriously-flawed, coincident release of initial reporting for September, October and November 2013 monthly data, the December housing starts reporting could reflect catch-up reporting, significant prior-period revisions and/or further haphazard monthly detail. The markets appear to be looking for a large, downside correction, which is reasonable. Anything is possible in the December detail, but ultimately, the unbelievably strong data in the prior reporting should disappear in a sort-term correction or revisions in the next several moths.

Despite near-perpetual market expectations for strengthening activity in housing starts, reported month-to-month change likely will resume its regular pattern of statistical-insignificance, soon, with ongoing stagnation and renewed downturn seen in the aggregate series, as well as particularly for single-unit housing starts.

In the wake of a 75% collapse in aggregate activity from 2006 through 2008, and an ensuing five-year pattern of housing starts stagnation at historically low levels, little has changed. There remains no chance of a near-term, sustainable turnaround in the housing construction market, unless there is a fundamental upturn in consumer and banking-liquidity conditions. That has not happened and still does not appear to be in the offing.

__________