GDP Revised to 5.6% – Only Using Government Math

The Commerce Department released a lower than expected revision of 5.6% for fourth-quarter Gross Domestic Product (GDP) today, but that number still does not come close to meeting the true growth in GDP according to Godlman-Sach’s Jan Hatzius.

Holding on to the skewed hope that the economy is growing and recession is over, the administration cannot afford to see this bubble-esque stock market plummet; which it would if the administration were to reveal a GDP number minus government intervention.

However, as noted by Hatzius, when absent the government’s inflation-bound free-spending, the real GDP number is much closer to 2.2%. And despite US officials spending more than twice than the rest of the population, the reality still remains that stimulus has affected a very small portion of the population. Unfortunately, that small percentage still does not increase employment broadly enough to lower an already underreported unemployment number and with it, a sagging economy.

And that reality could mean an ugly 2011 for the US Dollar if expectations remain unattainably high for growth in the 2nd half of this year. Advice to investors and traders: Stay ahead of the game; don’t rely on government math.

How Strong Was GDP Growth Really in Q4?
by Jan Hatzius

On Friday, the Commerce Department will release its first estimate of fourth-quarter GDP as measured from the “income side” of the national accounts. This will provide a useful crosscheck on the higher-profile “expenditure side” estimate that real GDP grew 5.9% (annualized) late last year, which itself might be revised slightly. Although the two measures are conceptually equivalent—that is, any differences must be due to statistical errors in one or the other—recent work at the Federal Reserve Board suggests that the income side is often more accurate than the expenditure side.

We expect the income side number to fall short of the 5.9% expenditure side estimate. A weaker number might dampen expectations of a strong labor market rebound in 2010, which are partly based on the strength of the expenditure side GDP data in late 2009. Our own view remains that employment growth will be insufficient to push down the unemployment rate on a sustainable basis in 2010, although we expect Census hiring to result in strong headline job growth in coming months (including a 275,000 gain in March).

The second revision to the Commerce Department’s quarterly GDP report rarely generates much excitement, as the changes to headline growth and inflation estimates are typically small. However, tomorrow’s report may be of more interest than usual. The reason is that it will contain the Commerce Department’s first estimate of the “income side” measure of gross domestic product in the fourth quarter, which according to recent research by the Fed staff may be a more reliable indicator of economic activity than the conventional “expenditure side” measure.

First, some background. In principle, there are three different ways to estimate gross domestic product, i.e. the value of all goods and services produced in a given period:

1. The expenditure side measure, defined as the value of all domestic expenditures on final product (e.g. personal consumption) plus the change in inventories plus the change in the trade balance. Following the recent study by Fed staff economist Jeremy Nalewaik, we call this measure GDP (E). (See Jeremy Nalewaik, “Income and Product Side Estimates of US Output Growth,” Brookings Papers on Economic Activity, available at http://www.brookings.edu/economics/bpea.aspx.) GDP (E) is the Commerce Department’s “featured” estimate of US GDP.

2. The income side, defined as the sum of all incomes received in the economy, which we call GDP (I). Conceptually, GDP (I) must equal GDP (E) because one person’s income is always another person’s expenditure. However, in practice there are significant differences between the two, as both are subject to potential measurement errors. The Commerce Department publishes an estimate of GDP (I) in nominal terms in the first revision to the Q1, Q2, and Q3 GDP report, and in the second revision to the Q4 report. We can convert this into an estimate of real GDP (I) using the implicit GDP (E) deflator.

3. The production side, defined as the sum of all value added produced in the economy, which we might call GDP (P). Again, true GDP (P) must equal true GDP (E) and GDP (I). We include this for completeness only; there is no independent GDP (P) figure in the United States, although it is actually the primary measure in a number of other countries.

The reason why we care about both GDP (E) and GDP (I) is that both are “noisy signals” of the unobservable true growth rate of GDP. If so, statistical inference suggests that we should put some weight on both measures in estimating true growth. Indeed, the study by Nalewaik referenced above argues—quite persuasively in our view—that GDP (I) has been a better measure of true activity than GDP (E) over the past few decades. This suggests that one might want to put more weight on GDP (I) than on GDP (E). But even if one disagrees with Nalewaik’s findings, it makes sense to at least put some weight on both.

The argument for putting weight on GDP (I) seems particularly strong at the moment. The reason is that it has been consistently weaker than GDP (E) over the past few years and is therefore better able to explain why employment fell so sharply during the recession and has continued to decline even after the trough in the GDP (E) data in the second quarter of 2009. In other words, if we use GDP (I) to measure output, the departure of the relationship between the unemployment rate and output from the historical “Okun’s law” relationship is less serious than if we use GDP (E). (Our own view has been that the departure is not all that large even if we use the GDP (E) data, but the use of GDP (I) data further reduces the gap.)

So what can we expect from the fourth-quarter GDP (I) number? The continued weakness in personal income—which is published more quickly than the other components of GDP (I)—suggests that GDP (I) growth is likely to come in below the 5.9% GDP (E) number. Indeed, the Federal Reserve’s flow of funds report contains a rough estimate of the “statistical discrepancy” that can be used to generate a preliminary estimate of GDP (I). This shows real GDP (I) growing just 2.2% (annualized) in the fourth quarter, far below the 5.9% number currently on record for GDP (E). The Fed emphasizes that the flow of funds figure is only a rough estimate, and we would frankly be surprised if Friday’s number turned out to be quite this weak. Nevertheless, we do expect GDP (I) to look weaker than GDP (E).

The main implication of a weaker number concerns the labor market. The strong rebound in employment growth predicted by many economists is at least partly based on the acceleration in the GDP (E) data late last year. If this acceleration overstates the economic reality, then the case for strong employment growth would weaken as well. Our own view remains that the March jobs report will show a headline gain of around 275,000 jobs and the next few months will benefit from Census hiring, but employment growth will be insufficient to push down the unemployment rate on a sustainable basis in 2010.