JOHN WILLIAMS' SHADOW GOVERNMENT STATISTICS


Issue Number 11


September 7, 2005

_____


Katrina Will Be Blamed for Pre-Existing Inflationary Recession


Storm-Enhanced Data Swings: Third-Quarter GDP Flat, September CPI Plus 1.0%, September Payrolls Down 300,000, September Unemployment Up 0.4%


Pre-Storm Numbers Showed Intensifying Contraction and Surging Prices

During the last twenty years, policies of Bush I, Clinton, Bush II and the Federal Reserve never caused a recession, that is, of course, if you believe the popular political spin. Saddam Hussein invaded Kuwait in August 1990 and triggered the 1990 to 1991 downturn, yet the official recession was timed from July 1990. The terrorist attacks of 9/11 were responsible for the 2001 contraction, but that recession began officially in March 2001. Likewise, Hurricane Katrina of late-August will be found to have caused the 2005 to 2007 recession, even though it likely will be timed officially from July.

As with the prior economic downturns, the 2005 contraction showed signs of serious economic slowing six to nine months in advance of its probable official start, let alone the politically popular excuse. In each of the above cases, however, a shock to the system intensified sinking economic activity and, in the current circumstance, exacerbated an inflation crisis. For anyone last week who saw sporadic gasoline lines and recalled the gas lines that began with the October 1973 Arab oil embargo, it is worth noting that the ensuing 1973 to 1975 recession was unusually long and deep.

The tragedy and catastrophe resulting from Katrina continue to unfold, with New Orleans and much of the Louisiana, Mississippi and Alabama Gulf Coast thoroughly devastated. The costs in lives and dollars most certainly will exceed those of the prior historic U.S. catastrophe benchmarks: the Great Galveston Hurricane in terms of lives lost and the Great San Francisco Earthquake in terms of insured dollar costs. The 1900 hurricane in Galveston may have taken 12,000 lives, while current guesstimates of Katrina's toll are in the 10,000 to 20,000 range. Nearly a century ago, the earthquake and ensuing fire in the City by the Bay cost insurers roughly $30 billion, restated in today's construction dollars. The Crescent City and Gulf Coast's insured losses likely will top $40 billion, with total costs, including those to be picked up by Uncle Sam, likely to top $100 billion.

Immediate economic impact from Katrina will be significant. Hurricane damage, historically, has tended to boost GDP, since losses are not subtracted from growth, but reconstruction business is added to it. Hurricanes of recent memory, however, have tended to hit resort and retirement areas. Katrina's impact is different, shutting down significant regional economic production and port facilities, including portions of the area's energy-related industry. Ripple effects throughout the national economy already are seen in gasoline and fuel supplies and prices. Further, for the first time in memory, there is still underway a protracted, total evacuation of a major U.S. city, with the accompanying displacement of its population.

Impact on economic data will be twofold: first in actual changes to economic activity, second in disruptions to economic reporting. As in the wake of 9/11, there are few economic reports that will not be dominated by the storms effects for at least the next three months. The effects also will be seen in seasonal factor distortions and year-to-year comparisons for similar-period reporting of 2006.

The U.S. statistical bureaus face a reporting nightmare in the months ahead. Door-to-door surveying, telephone surveying and company reporting from the storm-damaged area will not be possible for a month or two, perhaps longer. Many businesses no longer exist. That means that employment and unemployment data, in particular, will have to be guesstimated, and those guesses mean that the Bureau of Labor Statistics can come up with any numbers it desires.

Politically, it makes sense to take the biggest hits possible in the data, now, while all can be blamed on Katrina. Any overestimation on the downside will allow for stronger upside reporting in the future. When queried as to how the employment data in the storm region were going to be handled in the month ahead, a spokesman for the BLS would not go beyond "we will be working on it" but added that it would be safe to say the storm's impact would be "significant." Here is how some of the near-term data impact shapes up:

GDP - The Bureau of Economic Analysis estimates that Louisiana and Mississippi, respectively, produced 1.2% and 0.6% of 2004 GDP. September, representing one-third of the third-quarter period, will be close to a full loss for parts of Louisiana, Mississippi and perhaps Alabama, on top of damage done to national travel and consumer spending by high fuel and gasoline prices and disrupted supplies, and by restraints to consumption from people sitting at home watching TV for a week instead of shopping at the malls. These factors should knock at least two percent growth off the annualized, inflation-adjusted third-quarter GDP growth, on top of what otherwise would have been reported. With most data already slowing from the recession, reported third-quarter GDP growth is a good bet to be close to flat. Fourth-quarter GDP and beyond will suffer lingering damage, but massive reconstruction projects will provide some offset.

Employment/Unemployment - As of July 2005, before Katrina, the Bureau of Labor Statistics was estimating the number of people gainfully employed in some of the now hurricane-stricken areas as follows: New Orleans-Metairie-Kenner at 606,000, Gulfport-Biloxi at 114,000 and Mobile at 174,000, for a total of 894,000 people. By state, the breakout is Louisiana 2.0 million, Mississippi 1.3 million and Alabama 2.1 million, for a total of 5.4 million. Those estimates are not seasonally adjusted, so they reflect actual counts of working individuals, at least in theory.

A guesstimate of 500,000 people losing their jobs from the storm seems within reason. If that estimate proves out, September payrolls would decline by 300,000 instead of a pre-Katrina consensus of a 200,000 gain, and the national unemployment rate would spike by 0.4% to 5.3%.

CPI - If gasoline prices average $3.00 nationally in September, that will represent an 18.1% increase from the August average of $2.54 reported by the Department of Energy. In turn, with gasoline weighted at 3.93% of the CPI, the recent gasoline price surge -- blamed largely on Katrina -- would add 0.7% to September CPI, bringing the monthly inflation gain in at roughly 1.0%, and push annual CPI inflation over 4.0%. This would be on top of a likely 0.5% August monthly CPI gain to be reported this month.

Trade Balance - The disruption to Louisiana's port facilities also will tend to reduce trade volume as well as trade paperwork volume that flows through to reporting of the monthly trade statistics. To the extent that both imports and exports get reduced proportionately in reporting, such would tend to reduce the reported trade deficit.

Following 9/11, a surge of insurance payments from non-U.S. reinsurers provided a brief respite from the ever-deepening trade shortfall, but the Department of Commerce has since instituted a new reporting methodology that distributes the reporting of those payments over time, making them invisible in the data.

To the extent that storm damage leads to increased oil imports and impaired exports of grain, the trade deficit will increase.

Other Affected Numbers - Near-term downside effects on economic reporting should be seen in September retail sales and industrial production (these do not get reported until October). The federal deficit also will take an added hit, as tax revenues slow from the damaged region and billions of dollars are spent by the federal government for relief and reconstruction.

In regular economic reporting, slowing economic growth and rising inflation made a comeback in last month's pre-Katrina numbers. Such is detailed in the respective sections for each series. Despite moving in the right direction, however, the reporting catch-up was far short of reality, leaving in play the likelihood of a new era of direct political manipulation of economic data. Reporting problems resulting from the Gulf Coast disaster could cloud the issue for several months, unless the current manipulators are political morons.

In this month's "Reporting Focus," an unsustainable increase in income variance dispersion is examined. Implications are for protracted economic difficulties in the United States. Also considered are the latest reports from the Census Bureau of declining inflation-adjusted household income.

In general, the broad economic outlook has not changed. The 2005 to 2007 inflationary recession continues to unfold, but its near-term development has just been intensified a bit by Katrina.

Recession, inflation and dollar problems are upon us and do not offer a positive environment for the financial markets. With minor changes, here is an updated recapitulation of recent points as to the broad picture of what lies ahead:

The Shadow Government Statistics' Early Warning System was activated in May and continues to signal the onset of a formal recession early in third-quarter 2005.

The system looks at historical growth patterns of key leading economic indicators in advance of major economic booms and busts and sets growth trigger points that generate warnings of major upturns or downturns when predetermined growth limits are breached. Since the beginning of 2005 a number of key indicators have been nearing or at their fail-safe points. In the last several months, three indicators moved below those levels, signaling an imminent recession. Once beyond their fail-safe points, these indicators have not sent out false alarms, either for an economic boom or bust.

While negative GDP growth likely will not surface in regular government reporting until 2006, Katrina now offers the possibility of a shallow, one-quarter dip in third-quarter 2005. The National Bureau of Economic Research (NBER) should time the downturn to mid-2005 and announce same sometime in early-to-mid 2006. The popular financial media will begin to debate whether there is a recession underway by late 2005, but it will be blamed on Katrina. Those Wall Street economists who act as shills for the market will keep up their "strong growth is just around the corner" hype regardless of any and all evidence to the contrary.

From the standpoint of common experience, this downturn will be considered the second leg of a double-dip recession, not an independent contraction as will be claimed officially.

Most economic data already have started to soften, and the trend will accelerate sharply, with monthly contractions beginning for payroll employment and industrial production in the near future, even net of any short-term hits from Katrina. Political manipulation, if used, can keep the payroll data afloat for a while longer. Significant deterioration also will be seen in federal tax receipts (a widening budget deficit) and corporate profits, in addition to the sudden problems created by Katrina.

This outlook is predicated on economic activity that already has taken place and does not consider any risks from exogenous factors such as renewed terrorist activity in the United States or another major hurricane. As an aside, given the apparent inability of FEMA to respond on a timely basis to the New Orleans disaster, where authorities had plenty of warning and foreknowledge of the damage risks, one has to wonder how well the federal government would respond to a major terrorist attack that devastated a U.S. city.

Market perceptions of the downturn in business activity may be accelerated, due to Katrina. When expectations begin to anticipate weak data, expectations also will be lowered for inflation. Consensus forecasts generally still will tend to continue to be surprised on the downside for economic reports, and on the upside for inflation reports, for some time to come.

The roots of the current difficulties are structural in nature. A consumer starved of income growth and overburdened with debt cannot sustain the real (inflation-adjusted) growth in consumption needed to keep GDP growth in positive territory. The income weakness is a direct result of the loss of a significant manufacturing base to offshore locations and the ensuing explosive, perpetual growth of the U.S. trade deficit.

Exacerbating economic and financial woes will be unusually high inflation during this contraction. Inflation, fueled by high oil prices and weakness in the U.S. dollar, will not be brought under control simply by weakness in economic demand. Instead, persistently high prices only will serve to intensify the 2005 to 2007 recession, making it unusually long and protracted. Ongoing inflation woes and dollar problems will maintain upside pressure on long-term interest rates, inhibiting the traditional flattening of the yield curve expected with a recession.

An inflationary recession remains a nightmare for the financial markets. Particularly hard hit will be the U.S. dollar, with downside implications for both equity and bond prices.


THE BIG THREE MARKET MOVERS

(Each of these series is explored in the background article "A Primer On Government Economic Reports," available on the home page.)

The likely impact of Hurricane Katrina on upcoming GDP, employment/unemployment and CPI reports is discussed in this month's opening comments.

Employment/Unemployment -- August's moderate jobs gain and reduced unemployment were politically comfortable for the Bush administration. Seasonal factor distortions reversed for payroll employment, tending to understate the monthly jobs growth based on last year's seasonal patterns.

The popularly followed unemployment rate U-3 for August 2005 eased to 4.93% from July's 5.01%, seasonally adjusted, well within the published +/- 0.2% error margin. Unadjusted U-3 unemployment dropped to 4.9% from 5.2% in July, while the broader U-6 unemployment measure declined from July's 9.1% to 8.8%. August's seasonally-adjusted U-6 rate held even at 8.9%. Including the long-term "discouraged workers" defined away during the Clinton administration, total unemployment remained roughly 12%.

The household survey showed a seasonally-adjusted 373,000 increase in August employment after July's 438,000 gain.

Moving in softer tandem with household employment, the August payroll survey showed a seasonally-adjusted gain of 169,000 (213,000 net of revisions) +/- 108,000, after July's initial reporting of 207,000 new jobs was revised upward to 242,000. Annual growth in payrolls held at 1.7% in August.

The August gain included an upward bias of 132,000 jobs from the "net birth/death" adjustment to potential reporting entities, a net positive swing of 208,000 jobs after July's negative bias of 76,000 jobs. In September, the fudge factor eases back to the 40,000 to 50,000 range, unless the modeled guesstimate is manually reduced to account for Katrina damage.

August's employment/unemployment data were against a background of still-weak help-wanted advertising, renewed softening in employment as reported in the August manufacturing purchasing managers surveys, and a slight monthly deterioration in new claims for unemployment insurance (see the respective sections).

Next Release (October 7): September payrolls and unemployment reporting will be dominated by Katrina's damage, as judged by the BLS. The numbers will be guessed, since hard data will be hard to get, aside from some initial claims for unemployment filings. With the slowing economy and weak bias factors, September payrolls would have been soft to begin with. As discussed in the opening comments, our best estimate is that payrolls will decline by 300,000, with unemployment spiking 0.4% to a level of 5.3%. Underlying fundamentals continue to favor declining payrolls and rising unemployment, beyond Katrina's impact.

Gross Domestic Product (GDP) -- The "preliminary" estimate revision to annualized inflation-adjusted growth for second-quarter GDP took growth to 3.29% from initial reporting of 3.41%, and down from 3.80% in the first quarter. Annual growth was 3.59% versus 3.64% in the first quarter.

The first estimates of alternate second-quarter measures to the GDP, Gross National Product (GNP) and Gross Domestic Income (GDI), also were published. GNP, which is the broadest economic measure, includes GDP and the trade balance in factor income (interest and dividend payments). Annualized real growth in second-quarter GNP was 3.05%, down from 3.88% in the first quarter. GDI, which is the income-equivalent measure of the GDP's consumption side, showed growth of 4.42%, up from 3.53% in the first quarter, due to a narrowing statistical discrepancy with the GDP.

Next Release (September 29): The "final" estimate revision for second-quarter GDP should be little more than statistical noise. The impact of hurricane Katrina will show up first in the "advance" estimate of third-quarter GDP, which should come in close to no growth, as discussed in the opening comments and as will be reported at the end of October.

Consumer Price Index (CPI) -- The seasonally-adjusted July CPI-U increased by 0.5% (0.5% unadjusted) after June's unchanged reading. Although there was some catch-up in gasoline prices, the overall CPI still was 0.2% shy of fully reversing recent underreporting. Going against weak inflation last year, unadjusted July year-to-year inflation surged to 3.17% from June's 2.53%.

Using the CPI's original (pre-Clinton Era) methodological approach of a fixed basket of goods (vs. substitution of hamburgers for steak as estimated by geometric weighting) would leave year-to-year inflation at about 6.0% instead of the official 3.2%.

The "experimental" Chained Consumer Price Index (C-CPI-U), the fully substitution based CPI that presumably is the eventual replacement for current CPI reporting, showed annual inflation at 2.62% in July, up from 2.17% in June.

Next Release (September 15): Inflation reports over the next several months should be miserable. Due to unmistakable energy price increases, the BLS likely will report around 0.5% inflation in August and something in the vicinity of 1.0% for September, given Katrina's effects as discussed in the opening comments. With last year's seasonally-adjusted inflation gains of 0.1% and 0.2% respectively for August and September, reported 2005 monthly gains close to our estimates will spike annual inflation sharply, from the current 3.2% in July to 3.6% in August to 4.4% in September, the highest annual CPI inflation seen since July of 1991.

If the BLS does not report the energy inflation, cries of manipulation will abound. If properly reported, high inflation will be excused due to the distortions of Katrina. Unfortunately for federal deficit projections, however, cost-of-living adjustments for Social Security recipients are based on the average annual inflation in the CPI-W for the July to September period. This year's adjustment would be about 3.7%, based on our projections, the strongest increase in 14 or 15 years, and the strongest cost-of-living-increase since geometric weighting of the CPI was introduced to depress such adjustments. Therein lies the government's primary motive for understating inflation in the next two months, as was done last year.


OTHER TROUBLED KEY SERIES

To varying degrees, the following series have significant reporting problems. All series (including the more trouble free) will be addressed in a monthly "Reporting Focus," with Income Variance/Dispersion covered this month. In addition to the big three, other series that have been detailed are The Federal Deficit (an original background article and an update in the Alert of July 7, 2005), Consumer Confidence (November 2004 SGS), the Trade Balance (December 2004 SGS), Industrial Production (January 2005 SGS), Initial Claims for Unemployment Insurance (February 2005 SGS), Retail Sales (March 2005 SGS), Alternate Payroll Employment Measures (April 2005 SGS), Money Supply - Part I / SGS Early Warning System (May 2005 SGS), Money Supply - Part II (June 2005 SGS), Financial- and Trade-Weighted U.S. Dollar Indices (July 2005 SGS), Short-Term Credit Measures (August 2005 SGS).

Federal Deficit -- The official deficit for the fiscal year ended September 30, 2004 was $412.3 billion, up from $374.8 billion the year before. For the twelve months ended July 2005, the rolling deficit was $319.1 billion versus $446.1 billion in June 2004.

Gross federal debt as of the end of September 2004 (U.S. fiscal year-end 2004) was $7.379 trillion, up $596 billion from a year earlier. At month-end August 2005, gross federal debt was $7.927 trillion, up $576 billion from August 2004, which, in turn, was up by $561 billion from August 2003.

The federal deficit will be inflated quickly by Hurricane Katrina. Government finances will suffer tax revenue losses from the Gulf Coast on top of the creation of new government spending to rebuild the region. The outlook for deficit deterioration also will intensify as the recession sets in.

Initial Claims for Unemployment Insurance -- The annual change for new claims on a smoothed basis, for the 17 weeks ended August 27 narrowed slightly to 5.4% from a 5.5% drop as of July 30. The moving average has resumed its trend of deterioration and eventually will turn positive (a negative economic development), which would support signals of a recession onset. Reporting over the next several weeks will see first: seasonal factor distortion tied to the Labor Day holiday, and second: increasing impact as Katrina's victims begin to register their unemployment.

The volatility of the seasonally-adjusted weekly numbers is due partially to the seasonal-adjustment process. When the series is viewed in terms of the year-to-year change in the 17-week (three-month) moving average, however, such is a fair indicator of current economic activity.

Real Average Weekly Earnings -- July's seasonally-adjusted real average weekly earnings fell by 0.2% from June, after June's change was revised from a gain of 0.2% to 0.3%. July's reported real earnings fell 0.5% from the year before after June's 0.4% gain. Recent volatility in this series has come from general misreporting of the CPI.

Allowing for the biases built into the CPI-W series used to deflate the weekly earnings, annual change in this series continues to signal recession.

Retail Sales -- Spiked by gasoline prices and unusual seasonal-factor adjustments used to smooth reporting of auto sales, seasonally-adjusted July retail sales jumped by 1.8% (the same net of revisions) +/- 0.7%, following June's unrevised 1.7% gain. On a year-to-year basis, July retail sales were up 10.3% versus June's 9.6% gain.

Inflation-adjusted growth in retail sales below 1.8% (using the official CPI-U for deflation) signals recession, and annual growth again remained comfortably above that level in July reporting.

Next Release (September 14): August retail sales likely will come in below expectations, mirroring a combination of impaired economic activity and sales volume puffed by rising consumer prices. Katrina's impact will be seen largely in the September numbers, although minor August impact is possible.

Industrial Production -- Seasonally-adjusted July industrial production was reported up 0.1% (down 0.3% net of revisions) although the actual index remained unchanged at 119.4. June's original gain of 0.9% revised to 0.8%, as the unseasonable surge in utilities eased. July's year-to-year growth softened to 3.0% from 3.6% (was 3.9%) in June. Industrial production is at an inflection point, about to turn down as a result of the recession. That process will be exacerbated by the impact of Katrina.

Next Release (September 14): With the economy entering a contraction, overall industrial production will enter a period of protracted decline within the next several months. The report for August still looks like the most likely one to take the series into monthly decline, with September losses intensified by the effects of the hurricane.

New Orders for Durable Goods -- Seasonally-adjusted new orders for durable goods in July sank by 4.9% (down 5.0% net of revisions) after a revised 1.9% gain in June (previously 1.4%). July year-to-year growth slowed to 3.5% from June's revised 11.4% (was 11.7%). The widely followed nondefense capital goods orders plunged 7.3% month-to-month in July after June's 2.2% decline (originally down by 1.9%).

As has been usual in most benchmark revisions, historical growth patterns were revised downward on August 19. Government statistical agencies are loath to underestimate economic growth in the initial reporting.

Monthly volatility is high for this series. Durable goods orders used to be one of the better leading indicators of broad economic activity, when smoothed using a three-month moving average. After the semi-conductor industry stopped reporting new orders, the series' predictive ability fell apart.

Trade Balance -- The seasonally-adjusted June trade deficit in goods and services increased to a near-record $58.8 billion from a revised $55.4 billion (was $55.3 billion) in May. Shy of short-term hurricane disruptions discussed in the opening comments, the longer term outlook for the deficit remains one of continual deterioration.

Next Release (September 13): The July trade deficit should widen again, likely breaking the old record level of $59.0 billion set in November 2004. Paper-work flow interruptions from Katrina possibly could distort the next report, but most certainly will do so for the August report, with impact from actual economic effects of the hurricane surfacing in the September report.

Consumer Confidence -- August 2005 confidence measures were mixed, with the Conference Board's consumer confidence rising by 1.9%, while the University of Michigan's consumer sentiment plummeted by 7.7%. Year-to-year change deteriorated for both measures. On a three-month moving average basis, annual growth in confidence slowed from 3.7% in July to 2.7% in August, while the annual decline in sentiment widened from a contraction of 1.1% in July to a 3.1% August decline.

As lagging, not leading, indicators, these numbers continue to confirm the economy was slowing throughout fourth-quarter 2004, then stagnating and slowing again during the first three quarters of 2005. Katrina likely will take a toll on the September confidence readings.

Short-Term Credit Measures -- Consumer Credit has not been updated since the last newsletter. Commercial and Industrial Loans in July showed annual growth of 12.4%, up from 11.6% in June, while Commercial Paper Outstanding showed annual growth of 15.5% in August versus 14.5% in July. These growth rates are strong and likely will increase in the near term due to the impact of Katrina.

Producer Price Index (PPI) -- The seasonally-adjusted July finished goods PPI rose 1.0% (0.9% unadjusted) for the month, after June's unchanged reading. July's annual inflation rate jumped to 4.6% from June's 3.6%, and, come next month's report, it surely will top the 14-year high of 5.0% seen in the revised March numbers.

Next Release (September 13): Despite a large component of random volatility in monthly price variations, PPI inflation reporting over the next several months should soar in tandem with the CPI, topping already increasing market expectations.

Purchasing Managers Survey (Non-Manufacturing) -- Published by the Institute for Supply Management (ISM), there is nothing unusually wrong with this survey of the service industry, except it does not have much meaning. Unlike its older counterpart, the manufacturing survey, if service companies such as law firms, hospitals or fast-food restaurants have "increased orders," that does not necessarily mean that economic activity is increasing.

Such considered, the overall August index increased to 65.0 from 60.5 in July, running counter, again, to the manufacturing series. The index is a diffusion index, where a reading above 50.0 indicates a growing service economy, in theory. Both the employment and prices paid components, however, have some meaning.

The August employment component rose to 59.6 from 56.2 in July, suggesting strengthening employment in the service sector.

The prices paid component diffusion index is a general indicator of inflationary pressures. The August index softened to 67.1 from July's 70.3, still holding in strong inflation territory. On a three-month moving average basis, the annual change in August narrowed to a decline of 8.4% from July's 13.5% drop.


BETTER-QUALITY NUMBERS

The following numbers are generally good-quality leading indicators of economic activity and inflation that offer an alternative to the politically hyped numbers when the economy really is not so perfect. In some instances, using a three-month moving average improves the quality of the economic signal and is so noted in the text.


Economic Indicators

Purchasing Managers Survey (Manufacturing) - New Orders -- Reversing two months of gains, the August new orders index fell 6.9% to 56.4, down from July's 60.6. The measure breached its fail-safe point several months back, generating an SGS early warning indicator of pending recession.

The Commerce Department provides suspect seasonal factors for this series, and adjusted monthly numbers often can be misleading in the reporting of month-to-month change. This problem is overcome by using year-to-year change on a three-month moving average basis. On that basis, August's index was down 7.2% year-to-year versus a 10.6% contraction in July. The index gradually has notched lower from its peak annual growth of 36.6% in April of 2004.

Published by the Institute for Supply Management (ISM), the new orders component of the purchasing managers survey is a particularly valuable indicator of economic activity. The index is a diffusion index, where a reading above 50.0 indicates rising new orders. The overall August ISM index fell to 53.6 from July's 56.6. An index level of 50.0 divides a growing versus contracting manufacturing sector. Also of note, the August employment component eased, moving to 52.6 from July's 53.2 reading, suggesting weakening employment growth.

Help Wanted Advertising Index (HWA) -- The July index notched higher from 38 in June to 39 in July, three points off the low of the 2001 recession and current cycle. This series is best viewed on a year-to-year basis with a three-month moving average.

For July, that annual change in the three-month moving average was no change, following a 0.9% decline in June. HWA is one of those series that never recovered from the 2001 recession, having been at its cyclical nadir and multi-decade low of 36 as recently as November 2004. Accordingly, current movement still is bottom bouncing and remains worthy of close attention.

Published by the Conference Board, the HWA is a reliable leading indicator of employment activity.

Housing Starts -- Seasonally-adjusted July housing starts fell 0.2% from June, which had gained 0.2% (originally unchanged) in May. On a three-month moving average basis, July year-to-year change inched lower to 5.9% from June's 6.0%. Housing starts remains a series at the brink of generating a recession warning signal.

Money Supply -- Annual money supply growth still is generating a solid recession signal. Before inflation adjustment, M1, M2 and M3 monthly changes for August (four weeks) versus July were up 0.2%, 0.3% and 0.7%, respectively. Year-to-year rates of change in seasonally-adjusted August (four weeks) and July M1, M2 and M3 were 0.1%, 3.8% and 5.7%, and 1.2%, 3.8% and 5.4%.

Adjusted for CPI inflation, August's M1, M2 and M3 annual year-to-year rates of change were down 3.4%, up 0.2% and up 2.0%, respectively, versus down 1.9%, up 0.6% and up 2.2% in July. On a three-month moving average basis, these annual rates of change were down 1.9%, up 0.6% and up 2.2%, levels that still are well underwater using the old-style CPI.


Inflation Indicators

Purchasing Managers Survey (Manufacturing) - Prices Paid -- The August prices paid diffusion index exploded versus July, gaining 28.9%, with the index rising from 48.5 to 62.5, a serious inflation reading. On a three-month moving average basis, August's year-to-year change narrowed to a 32.6% decline versus July's 35.7% drop.

Published by the Institute for Supply Management (ISM), the prices paid component of the purchasing managers survey is a reliable leading indicator of inflation activity. The measure is a diffusion index, where a reading above 50.0 indicates rising inflation.

Oil Prices -- West Texas Intermediate Spot (St. Louis Fed) jumped to still another monthly record high average in August, at $64.97 per barrel. That was up 10.7% from July and a highly inflationary 44.5% above August of 2004. Spot prices spiked sharply at the end of the month in response to Katrina and remained high in the wake of release of some crude from the strategic petroleum reserves. Despite any near-term gyrations or relief sell-off, high oil prices will remain a major contributing factor to the inflation side of the inflationary recession.

Oil price changes permeate costs throughout the economy, ranging from transportation and energy costs, to material costs in the plastics, pharmaceutical, fertilizer, chemical industries, etc. Where effects of continued oil price volatility will affect CPI reporting, downside oil price movements tend to be picked up more quickly and fully than are upside movements. Even as currently understated, CPI and PPI inflation should be much stronger than commonly predicted for the next six-to-nine months, partially as a result of continued high oil prices.

U.S. Dollar -- The Shadow Government Statistics' Financial-Weighted U.S. Dollar Index is based on dollar exchange rates weighted for respective global currency trading volumes. August's monthly dollar average fell by 1.9% after July's 1.9% increase over June. Year-to-year change moved from a 0.9% gain in July to a 1.6% contraction in August.

Moving slightly less than this month's financial-weighted index, August's monthly average of the Federal Reserve's Major Currency Trade-Weighted U.S. Dollar Index declined by 1.8% after July's 1.0% increase. August's rate of annual contraction deepened to 2.8% from July's 0.7% decline.

The recent dollar rally may have run its course, with underlying fundamentals remaining extraordinarily negative for the greenback. Where serious shocks loom in U.S. economic and fiscal data on top of Katrina's negative impact, heavy selling pressure against the U.S. currency could intensify at any time, with little warning. New record lows for the dollar remain likely in the months ahead, despite any overt or covert supportive intervention by any central bank(s).

Generally, the weaker the dollar, the greater will be the ultimate inflation pressure and the eventual liquidity squeeze in the U.S. capital markets.


SEPTEMBER "REPORTING FOCUS" -- INCOME VARIANCE/DISPERSION
AS A PREDICTOR OF LONG-TERM ECONOMIC TRENDS
For the leading economic indicators analyzed monthly in SGS, lead times to broad economic activity range between three and nine months. There is one indicator, however, that signals broad trends in liquidity and economic activity three to eight years in advance: income variance/dispersion. Since it is updated only once per year in the Census Bureau's annual poverty report, it also will be updated in the SGS once a year, in September.

Income dispersion, also known as variance, measures the distribution of household income from very low levels to extremely high levels. The more income is distributed in the extreme ranges, the higher is income dispersion; the more income is distributed in the middle income brackets, the lower is the income dispersion. Extremely high levels of income dispersion have tended to precede great economic and financial catastrophes. Those events, in turn, have tended to redistribute income towards the middle, in something of a self-correcting cycle.

Some dozen or so years ago, while offering one of his occasional gloom-and-doom talks, Alan Greenspan noted that, "A potentially significant factor in the current state of long-run concerns is that the distribution of family income has become more dispersed."

Triggering Greenspan's concern at the time was that variance in U.S. household income had retaken its pre-1987 stock crash high, a level not previously seen, except perhaps before the 1929 stock crash and the ensuing Great Depression. Last week, the Census Bureau released estimates of income dispersion for 2004 that were 30% higher than when the Fed Chairman's knees were knocking. Is this cause for concern? Absolutely! Does Mr. Greenspan have any idea of what really is happening to the economy and the underlying financial system? He does, although he rarely admits or indicates same.

The distribution of wealth and income provides the fundamental structure for all economies. Where income drives consumption, shifts in income distribution impact economic activity. When income distribution shifts away from extremes, a greater portion of the consuming public is in the middle-income bracket, consumption becomes more broad-based, and the economy booms.

Consider, for example, that someone earning $10 million per year likely will buy fewer automobiles than 100 people each earning $100,000 per year.

As income distribution shifts toward extremes, variance and dispersion rise, consumption become less broad-based and the economy turns down. A shift in income distribution leads to a structural shift in economic activity.

The current distribution of income in the United States has developed in something of a free-market system, with government modification in the form of tax, monetary, welfare, labor and antitrust policies. Without getting into the merits of various forms of modification, and without getting into the larger philosophical questions of wealth distribution, shifts in income variance can be shown to have significant long-term impact on general economic activity.

Our analysis shows a leading relationship between the level of income dispersion and annual growth in systemic liquidity, not adjusted for inflation. This is one reason why the Federal Reserve sometimes finds itself pushing on a string when it tries to stimulate economic activity through money supply growth. As shown in the accompanying graph of inflation-adjusted household income vs. income dispersion, income variance was at an all-time extreme as of 2004.






The measure used here is the median logarithmic deviation of income as published by the Census Bureau in its annual poverty report. There are a variety of income dispersion/variance measures, but they all show the same basic pattern: U.S. income variance is at an unprecedented level that portends a possible major break in the economic/financial system.

Also reported by the Census Bureau last week were estimates of inflation-adjusted median (meaning the middle measure) and average household income levels for 2004. Of some note, as shown in the second graph, both median and average household income, adjusted for inflation, have been shrinking since 1999/2000.

The combination of negative annual growth in household income and the accompanying spike in income dispersion, as shown in the first graph, is a particularly ominous pairing. Growth in household consumption cannot be sustained without income growth. If household consumption is down, the economy is going to grow at a slower pace than the number of households. The number of households grew by 1.0% in 2004 and by 0.6% in 2003, growth rates well below reported GDP growth.

The latest numbers from the Census Bureau show the system to be at the brink of instability and unable to sustain the GDP growth being reported by the Bureau of Economic Analysis.


UPCOMING "REPORTING FOCUS" FOR OCTOBER
-- SGS ALTERNATE CONSUMER PRICE INDEX
As announced in early June, Shadow Government Statistics is developing an alternate, monthly consumer price index, to be published starting fourth-quarter 2005. A summary of the methodology and advance results will be covered in next month's "Reporting Focus."

The SGS index numbers will be set -- not subject to revision -- and usable in calculations in the same manner as the official CPI. The SGS index calculations will be fully transparent and replicable, based on publicly available data, not on massaged surveying by the Bureau of Labor Statistics, or over-modeled and over-theorized price levels. A history going back to 1975 will be reconstructed, with a bridge to pre-1975 CPI reporting.

The full methodology will be published shortly following the next SGS newsletter, followed in November with the reconstructed historical data.

___

October's Shadow Government Statistics is scheduled for release on Wednesday, October 12, 2005. The monthly newsletter regularly is posted the Wednesday following the Friday release of the employment statistics. The posting of the October SGS on the Web site, as well as any interim alerts, will be advised immediately by e-mail.

_____