On Fictitious Government Statistics (4.9% GDP Growth Edition)

Today’s GDP release showed third quarter growth at 4.9%. That number was such a howler that it promptly elicited the contempt it deserved. From Barry Ritholtz, in “GDP=4.9% (also, I have a bridge for sale in Brooklyn)“

This 4.9% number is one of the more “fanciful” government releases you will see in your lifetime, (outside of the state run media that exist only within totalitarian dictatorships).

Did this past quarter feel like the strongest growth quarter in 4 years?

Let’s begin with what we know about Q3 prices: They saw significant increases — yet the price index deflator was a 9 year record low of 0.9%. Rex Nutting observed: “Because of the way in the price index is constructed, it likely understates real-world inflation, and thus overstates real growth.”

Residential fixed investment, the GDP component that includes spending on housing, plunged by 19.7% in the third quarter (but Investments in structures increased 14.3%).

Profits for the 3rd quarter flipped negative, dropping 8.5%.

We have seen consumer spending falter, with the crucial opening salvo of the holiday weekend down 3.5%. That’s no surprise, given that second-quarter wages were revised lower by $44.8 billion. As a result, real disposable incomes fell 0.8% in the second quarter, instead of rising 0.6% as the Commerce Department had previously reported.

~~~

Question: How can Q3 GDP be 4.9% with corporate earnings, housing and retail sales so awful? Forget Goldilocks, this fairy tale sounds more like Cinderella . . .

The more anodyne Wall Street Journal’s Real Time Economics Blog also took issue, and pointed to other measures that showed far less robust growth:

According to the latest gross domestic product revision, the U.S. economy swelled at nearly a 5% clip last quarter, almost double the economy’s noninflationary limit.

Or did it?

Gross domestic income – a lesser-known gauge that the Fed has highlighted in the past as perhaps a better alternative — increased less than 2% last quarter, well below the economy’s potential. The first estimate of GDI is released with the second GDP estimate because it incorporates data that isn’t available earlier. (See line 11 on this chart.)

GDP counts economic activity based on expenditures, while GDI bases it on income. In theory, they should add up the same, though the often diverge — albeit not as much as they did last quarter.

Earlier this year when the Fed was trying to reconcile slower GDP growth with still-strong labor markets, it noted that GDI “might better capture the pace of activity.” GDI was running hotter than GDP at the time.

But the tables appear to have turned since early in the year. GDI has grown more slowly that GDP over the first three quarters of 2007 as a whole.

The main difference between the two gauges last quarter was corporate profits, which GDI includes and GDP excludes. Corporate profits from current production fell last quarter. GDI also doesn’t explicitly include net exports and inventories, as GDP does. GDI, in contrast, relies more heavily on employee compensation data.

But when there are differences, Fed officials may lean towards GDI, especially when it comes to signaling economic downturns. Fed economist Jeremy Nalewaik wrote in a March paper that GDI “has done a substantially better job recognizing the start of the last several recessions than has real-time GDP.”

Now even Ritholtz’s post acknowledges that a big contributor to the seemingly exaggerated GDP figures is the way inflation is treated, and GDI isn’t skewed by that. So we seem to have a culprit.

However, what is troubling is that this isn’t the only instance of dubious statistical releases by the government. Ritholtz, Michael Shedlock, Dean Baker, and Floyd Norris have remarked repeatedly on suspect birth-death adjustments to the Bureau of Labor Statistic’s monthly jobs report. Needless to say, the birth-death adjustments have led to increases in jobs allegedly created, and this pattern has persisted for months.

Similarly, Ritholtz, Wolfgang Munchau, and others have been critical of reliance on the consumer price index as the metric of inflation (it is one thing to say that the non-core elements, particularly food and energy, are volatile, but in this case, they’ve been unidirectional).

Ritholtz, Shedlock, and Nouriel Roubini haver also criticized previous GDP releases, and Shedlock in particular has written on the seldom-considered practice of using hedonic adjustments to GDP. They allow for the fact that computers are becoming more powerful at lower costs. In essence, the US grosses up the price of computers in its GDP reports to adjust for the fact that computer prices are dropping.

These adjustments have been going on since 1980 and the US is the only OECD country to use this approach. Shedlock obtained some data from the Bureau of Economic Advisers that indicated that hedonic index-related adjustments had added $2.257 trillion to 2005 GDP. That’s 22% of the total.

Why does this bother me? As Ritholtz suggested at the outset of his comment on the latest GCP release, this is banana republic behavior. I did a wee bit of work in Mexico a while ago, and then you couldn’t rely on a single government supplied statistic. it made any analysis an exercise in informed guesswork, and made it hard for businesses and individuals to plan. But it seems this nation is more interested in feel-good than in reality.

Update 11/30, 2:00 PM: Jim Hamiltion at Econbrowser provides an analysis of the GDP release. An increase in inventories apparently contributed 1% of the 4.9%, which is not a good sign, but he found the news that exports were another source of growth to be encouraging.

Print Friendly, PDF & Email

2 comments

  1. Independent Accountant

    I concluded the CPI was “fixed” about a decade ago. Remember the Boskin Commission in about 1997 which reduced the CPI by 1.3% annually? I guess the CPI UNDERSTATES inflation by about 2-3% per year. Therefore the GDP numbers are similarly overstated. The inflation statistics are a joke.

  2. Anonymous

    ” The main difference between the two gauges last quarter was corporate profits, which GDI includes and GDP excludes. Corporate profits from current production fell last quarter. GDI also doesn’t explicitly include net exports and inventories, as GDP does. GDI, in contrast, relies more heavily on employee compensation data. “

    Gee. I’m glad we got that straight.

    Not a particularly robust analysis from these sources.

Comments are closed.