In this slowing U.S. economy, it’s important to understand why and how economists’ adjustments of actual data can produce misleading results. Worse, those adjusted results often are presented as absolutes, without discussion of the steps taken to calculate them.
So, here we go, using the GDP (Gross Domestic Product), the primary measure of the U.S. economy, as the example…
Step one: The data
Users of GDP data naturally want accurate, speedy reports. However, those two goals conflict. Accuracy takes time and speed requires data management shortcuts. Therefore, the government agency responsible, BEA (Bureau of Economic Analysis), releases three different versions of the report.
Note: The BEA “Quick Guide: GDP Releases,” offers explanations of the reporting
The first of the three reports, labeled “Advance Estimate,” contains the preliminary results about one month after quarter-end. About a month later, the “Second Estimate” is released. Then, about a month after that comes the final “Third Estimate.” (For 4th quarter 2022, the Advance Estimate was released January 26, 2023. The Second Estimate and Third Estimate are scheduled for February 23 and March 30.)
Can the second and third estimates be significantly different? Yes, as shown by the 3rd quarter 2022 estimates. Advance estimate = 2.6%. Second estimate = 2.9%. Third estimate = 3.2%. Therefore, know that the 4th quarter 2022 advance estimate of 2.9% could be revised.
Understanding by non-economists is not helped by casual wording that presumes familiarity with all the adjustments made. For example, here is the first paragraph in the January 26 report:
“Real gross domestic product (GDP) increased at an annual rate of 2.9 percent in the fourth quarter of 2022 (table 1), according to the ‘advance’ estimate released by the Bureau of Economic Analysis. In the third quarter, real GDP increased 3.2 percent.”
So, a quick read is the annual rate through the fourth quarter was 2.9%. And that final sentence says nothing about the 3.2% being an annualized number. Then, there is the graph of those numbers with the title, “Real GDP – Percent change from preceding quarter.”
Now you see the 2.9% is a quarterly number. “What?” you ask. “Did the U.S. economy really grow at an inflation-adjusted 2.9% rate in one quarter?” Nope. In fact, the entire year’s real growth rate was only 2.1%. The explanation is in the mini-font, bottom notation: “annual rates.” In other words, the calculated quarterly rate was only 0.7%.
Why balloon-up a quarterly rate? To attempt to put everything in annual terms. That’s why the BEA release’s main data tables show quarterly dollar amounts multiplied by 4 (annualized).
Step two: Calculating “real” (inflation-adjusted) data
This step should be well understood. The goal is to remove price changes from the GDP growth calculations. Actual (AKA, nominal or current-dollar) amounts include both “real” production increases and price increases. Removing the latter gets to the heart of GDP growth.
The data is included in the tables at the end of each BEA report, but most of the data are heavily adjusted. To get to the basic data, go to the last table: “Appendix Table B. Not Seasonally Adjusted Real Gross Domestic Product.” In that table, line 8 provides GDP in current dollars, as originally calculated before inflation adjustment, seasonal adjustment and annualization. And, finally, we get the unadulterated 4th quarter GDP: $6.67T.
Comparing to other numbers on line 8, we calculate a 3.6% gain over the 3rd quarter 2022, and a 7.6% gain over 4th quarter 2021. Now, go back up to line 1 for inflation-adjusted (2012 dollars) amounts (still without seasonal adjustment and annualization). The two gains are now reduced to 3.1% and 1.2%.
Step 3: Seasonal adjustment
Many components of the economy are seasonally variable. Perhaps surprisingly, GDP is also heavily influenced by the seasons. This graph shows both the actual and the real GDP quarterly growth rates over the five years preceding 2020. That same pattern is visible in every year outside of these five, although it is skewed during recessionary periods.
To understand each quarter’s relative growth, seasonal adjustments are applied, based on previous years’ shifts. Here, then, are the real growth rates from above, compared to the seasonally adjusted real growth rates.
Step 4: Annualization
Economists express a need – or, at least, a desire – to convert quarterly amounts and growth rates into annual ones. Multiplying amounts or compounding growth rates by four may allow comparison with true annual numbers. However, it is simply making a mountain out of a molehill. During unusual times, like now, annualization can produce misleading results.
The bottom line: Go beyond simplistic, one number economic reports
Employment, retail sales and homebuilding are among the other areas where seasonal adjustment and annualization are commonly used. The problem is they are too often relied upon, even when conditions are skewed.
Therefore, the only way to understand unusual periods is to start with the basic data and add in other observations. Applying thoughtful analysis is always superior to relying solely on some algorithm created in a different time period – particularly, when a trend change is afoot.
Financial System Shakeup Has Begun
SVB, The Banks, The Economy, And The Market
Will SoFi’s Lawsuit To End The Student Loan Freeze Alienate Potential Customers?