Ron Cooke: American GDP: Can We Trust The BEA Data?


18 Jan 2008
View all related to economics | economy | Money
View all related to Ron Cooke


In another life (circa 1962), I was an auditor for AT&T. Nothing spectacular. Mostly cash and property reviews. Then some business process analysis. It was my good fortune to have two older gentlemen as partners. They graciously decided to teach this green college kid how to be a good auditor. It was a great learning experience. One of the tricks they taught me was called the “reasonable test”. If the data under audit was within the parameters of like data from other audits, then it was reasonable to assume there were no problems of procedure or management. If, on the other hand, the data did not seem to make sense versus circumstantial criteria, then it would be reasonable to assume further audit investigation was warranted. This technique of measuring the quality of information has become a cornerstone of my work ever since.

In early November, 2007, the Commerce Department’s Bureau of Economic Analysis (BEA) announced the United States had achieved a third quarter Gross Domestic Product (GDP) of 3.9 percent. That number was later updated to 4.9 percent. Those numbers set off my “reasonable test” alarm. How, I wondered, with an accelerating rate of inflation and declining economic activity, could the United States turn in such a stellar performance?

The BEA’s report flunked the reasonable test.


The BEA reported American GDP in billions of Current Dollars (the money we actually spent for goods and services) for Q3 2006 and Q3 2007. It also reported this same data adjusted for inflation using “chained” 2000 dollars. As of December 20, 2007, the quarterly data, using seasonally adjusted annual rates for the National Domestic accounts, yields the Current-Dollar and “Real” Gross Domestic Product data shown in the following Table. It shows that annual GDP growth in current dollars grew from 4.53% in Q1 2007 to 5.30% in Q3 2007. Using inflation adjusted chained 2000 dollars, economic growth grew from 1.55% in Q1 2007 to 2.84% in Q3. Not bad.

But wait. Does this imply an inflation differential of only 2.46% for Q3? And do we really believe the inflation differential actually declined from 2.98% in Q1 to 2.46% in Q3? Didn’t the value of the dollar decline over these three quarters?

GDP in billions of current dollars

% Change from year ago quarter

GDP in billions of chained 2000 dollars

% Change from year ago quarter

Inflation Differential



















The BEA’s Price Index for Gross Domestic Purchases (which measured prices paid by U. S. residents) increased by just 1.8% in Q3. By contrast, the Labor Department’s Bureau of Labor Statistics (BLS) CPI-U inflation index was 2.36% for this same period. Which number is a better measure of inflation? Can we trust either number?

And to further compound the confusion, the BEA has reported a current dollar gain of 6.0% for Q3. BUT this is against average GDP for all of 2006, rather than a comparison of Q3 2006 vs. Q3 2007.

Collecting the copious amounts of data used to compute GDP has to be a tedious and sometimes frustrating job. Unfortunately, sophisticated analysis and hard work does not guarantee credible results. The BEA’s conclusions appear to be a bit optimistic.

Simple Net GDP Calculation

Pundits frequently ignore current dollar GDP (the total production of goods and services priced as though they were purchased with current dollars). Instead they use a number that has been adjusted downward called “Real” GDP that deducts the rate of inflation and makes other adjustments to current dollar GDP in an attempt to compare GDP from one period, with the GDP for a subsequent period, using dollars of a constant value .

I dislike the term “Real” GDP. There is nothing sacred about using inflation adjusted dollars as a measure of economic performance. Current dollar GDP is just as “real” as any other measure of value and provides a useful way to compare multiple sets of data from period to period. We should remember. Consumers can not spend inflation adjusted dollars to purchase goods and services. They can only pay their bills with the money that is actually in their pocket – current dollars. So .. if we want to adjust current dollar GDP for inflation, then let us do just that … and call it “Net” GDP. In other words, Net GDP is the percentage increase (or decrease) in current dollar GDP for a specified period vs. the current dollar GDP of a like prior period, less the rate of inflation from the prior period to the specified period. In the following example, seasonally adjusted current dollar GDP increased from $13,266.9 billion in Q3 2006, to 13,970.5 billion in Q3 2007 – an increase of 5.30%. The BLS reported a seasonally adjusted price index increase of 2.36% for these same two periods. If we subtract the BLS CPI from BEA current dollar GDP, that gives us a net increase in GDP of 2.94% from Q3 2006 to Q3 2007, - far less than the GDP gain of 4.9% reported by the BEA.

BEA Q3 2007 GDP Growth in Current Dollars

from Q3 2006


Increase of Q3 2007 GDP vs. Q3 2006 GDP

BLS CPI-U Q3 2006 vs. Q3 2007


Deduct Q3 2007 Rate of Inflation




If we take the BEA seasonally adjusted quarterly current dollar Gross Domestic Product percent change for Q3 2007, and compare it with this same data adjusted for chained 2000 dollars, the “inflation” differential is only 1.1 % even though the BEA price index was 1.8. In addition, note that while real world food and fuel prices have been going up, the inflation differential has been going down.

How is this possible?



GDP percent change based on current dollars

GDP percent change based on chained 2000 dollars

Inflation Differential













If you go to and click on the Cultural Economics tab, you will see my essay of the rate of inflation: “CPI: Sophisticated Economic Theory, Terrible Ethics”. To quote from that essay: “If we use the weighting and data points from the above factoids to calculate an alternative estimate of CPI (the Consumer Price Index), we get a very different picture of American inflation from Q3 2006 to Q3 2007. There is a dramatic increase in food and housing costs. …... Granted. Accuracy would require the acquisition and analysis of a lot more data than assembled for this effort. But the large discrepancy suggests something is wrong with either the survey methodology or the process of analysis. Whereas the BLS reported a CPI increase of 2.36% for this period, the actual rate of inflation was more like 4.02%.”

Ok. I like my economics simple, uncluttered, and straight. Assuming the credibility of the BEA current dollar estimates, let’s deduct my alternative CPI from the BEA data to estimate economic performance.

BEA Q3 2007 vs. Q3 2006 GDP in Current Dollars


TCE CPI-U Q3 2007 vs. Q3 2006

Dollar value inflation


“Net” increase in Q3 2007 GDP


Using this methodology, could one conclude America’s economy posted a modest performance in Q3 2007? And by the way:

which number reflects contemporaneous comments on the economy:

the 4.9% gain in GDP reported by the BEA, or the above estimate of 1.26%?

Last February, I projected America would fall into a recession by the end of 2008. To be honest, I believed our economy would turn negative in the fourth quarter of 2007. But the economy has remained remarkably buoyant. We appear to have avoided a recession in 2007 because:

  1. Government spending continues to inflate the economy;

  2. Higher food, fuel and commodity prices inflated GDP;

  3. Unemployment tends to lag GDP and the rate of inflation; and

  4. The world is awash in speculative cash.

Unemployment, which has increased dramatically in the financial services and home construction industries, will be followed by higher unemployment in all economic sectors during 2008. That is a given. Cash infusions from offshore resources have found their way into American financial instruments, investments, and markets. This has (temporarily) inflated economic activity. But these anomalies only serve to mask the America’s economic challenges.

When the BLS and BEA numbers for Q4 2007 are published, they should show a year over year increase in inflation and a decrease of GDP. By my methodology, inflation should exceed 5% and GDP should be neutral or negative.


GDP is one of the most closely watched economic statistics: It is used by the White House and Congress to prepare the Federal budget, by the Federal Reserve to formulate monetary policy, by Wall Street and the media as an indicator of economic activity, and by the business community to prepare forecasts of production, investment, and employment. Because of its extremely sensitive business and political ramifications, reported GDP (current or chained) needs to be accurate, unambiguous, and trustworthy.

And this brings up an interesting point. One of the issues in this election cycle is trust. Can we trust the information we receive from the Federal Government? Congress? The Administration? Federal agencies? Aside from outright falsification, and intentional or intrinsic bias, data and information can be rendered untrustworthy by establishing a misdirected premise for the methodology or by overly sophisticated manipulation.

Hopefully, we will elect a management team in November that has the ability to review our measurement objectives and the analytical processes used to achieve them. In other words:

In God We Trust. All others need an occasional audit.

Ronald R. Cooke
The Cultural Economist
Author: Detensive Nation