The third estimate for GDP topped economist’s consensus estimates.
The BEA reports Real Gross Domestic Product (GDP) increased at an annual rate of 2.2 percent in the second quarter of 2026 (April, May, and June).
Real GDP was revised up 0.7 percentage point from the second estimate, primarily reflecting upward revisions to investment, consumer spending, and government spending
In the first quarter, real GDP increased 2.5 percent (revised).
Real Second-Quarter GDP and GDI
- Real GDP 1.5%
- Real Final Sales: 2.8%
- Real Final Private Domestic Sales: 4.6%
- Real Final Domestic Sales: 3.8%
- Real GDI: 2.6%
The difference between real GDP and Real Final Sales is Change In Private Inventories (CIPI) that nets to zero over time.
Thus real final sales are a better measure than the topline widely reported numbers.
The Fed focuses on Real Final Private Domestic Sales which was a whopping 4.6percent annualized.
Contributions to GDP
Contributions to GDP
- PCE Services: 1.57 PP
- PCE Goods: 094 PP
- Government: -0.01 PP
- Residential Investment: 0.01 PP
- Nonresidential Investment: 1.25 PP
- CIPI: -0.53 down
- Exports: 0.56 PP
- Imports -1.66 PP
Imports and Exports
Imports don’t subtract to GDP. The have no impact of GDP at all. The D in GDP explains why.
The BEA subtracts imports because its initial assumption is that all sales are domestic.
Otherwise the BEA would be trying to figure out things like “What percentage of this hammer from Home Depot is domestic?”
Reflections on the Economy
Based on private domestic sales, the economy is overheating. 4.6 percent final private domestic sales is a huge number.
The Fed is behind the curve. That is the bond market reaction today with long-term yields rising again despite slightly better than expected PCE price inflation readings by the BEA.
GDP Not Boosting Sentiment
On September 29, 2026 I noted US Consumer Confidence Plunges to the Lowest Level Since 2014
Views of the present situation and future expectations both declined dramatically.
On September 25, 2026, I noted Consumer Sentiment Drops in September to Just Above Record Lows
Republican attitudes decline the most, but from higher levels.
Partisan Sentiment Changes
- After particularly large declines in sentiment this month, Republican sentiment is now 20% lower than January 2026.
- Democrats are down 13% over the same period.
Consumers remain concerned over inflation no matter what economists and the GDP say about the economy.
Keep in mind that spending for the rich is discretionary. They can decide whether to jet off to Cannes, or maybe just go to Aspen this month.
The poors have to spend every dollar they have, just to do things like “eat”, “get to work” and “not be reduced to sleeping in a cardboard box”.
[…] September 30, 2026: BEA Revises Second-Quarter GDP 0.7 PP Higher to 2.2 Percent […]
Inflation will creep up some in the Q3 report, but much higher in Q4 when price increases of 5-10% at Walmart have taken their full effect
Why is a big real final domestic sales number a problem? Isnt that great news and mean the inflation part of nominal gdp is low?
Estimated GDP increase 2.2% annualized. I think that any reasonable estimate of inflation would exceed 2.2.5 annualized. Doesn’t this mean that the economy has become smaller? I am not trying to start an argument, I am trying to understand how this works.
This one goes to 11.
Dick Armey, of all people, pointed out that “you tell me who did the study, and I’ll tell you what results they got.”
When he’s right, he’s right.
The “G” in GDP is Gross, with a fairly accurate connotation of statistically crude or coarse. GDP is a total of a skewed distribution. We know that GDP is concentrated in large firms and most employment is in small firms. Proxies for skewness can be found on the income side. Unfortunately, looking at the SL Fred data, GDI (median or by income class) doesn’t seem to be published. The closest proxy I can come to is nominal “Income After Taxes by Quintile” but it’s pretty stale — the latest datum is over three years old. Even this is somewhat useful to get an idea of the hardship buried in the averages. Nominal income for the bottom quintile has been flat since 2021. If (big If) this has persisted until the present, these folks have experienced declining real income for the last five years. The trend (whatever it is) has probably worsened due to higher inflation since the Iran war started. Meanwhile the top quintile has been growing by leaps and bounds (the highest decile even faster).This may help explain the decline in consumer sentiment and the growing unhappiness of voters. A rise in real GDP of 0.7% doesn’t mean that much of the “P” or “I” is trickling down.
I think most War Department spending is included in GDP, and that spending has ramped up significantly
Over the last 25 years, the change in nominal PI is almost the same across deciles:
https://www.bea.gov/data/special-topics/distribution-of-personal-income. The K-shaped economy meme explains a lot of feelings, but less so most of the actual statistics.
If anything, the poor have done better nominally (in % terms) than the middle class
Thanks for the info. I was using nominal income after taxes by quintile, U.S. BLS statistics. Your link, to a study by the BEA and using Nowcast data, shows the growth in the bottom quintile significantly lower in 2024-25 than all other quintiles. This more recent time period is the one we should be most interested in for the political implications. In terms of the average growth in nominal income after taxes since 2020, St. Louis Fed says that the bottom quintile growth has been 3.4%/year vs. 3.7%/yr for the top quintile. The lowest quintile is skewed upward because of the pandemic-era stimulus payments in the 2020-2021 period. Over a long period of time, even a few tenths of a percent of sustained growth differential can add up. Probably a better comparison would be 2000-2019. We’ve had a K-shaped economy for the last two decades.
2.2/1.5 = ~1.47
47% is a pretty big error bar on a first estimate. What does this say about competence levels among whoever is running the show at the data collection bureaus? This is banana republic level.
Agreed, how do you trust either number when the estimate is that far off?
I know you exchange speed for accuracy on the first round but it was so far off that is was more misleading than useful.
Again, these estimates are out there to just give us an idea of what’s happening in a country with 350 million individuals. (Nobody on this blog is going to change their daily consumption significantly because they read this number)
And estimates change for a myriad of reasons, including timing of when the production/sales actually happened – known after the fact. The GDPNow estimate for 2026:Q3 went down today from 5.0% to 3.7% today. NOT because the third quarter was now 26% (1.3/5.0) worse than it was yesterday, but because some of the ‘hot’ production for 3rd quarter was now clocked for the 2nd quarter instead.
So instead of going from 1.5% growth in Q2 to 5.0% growth in Q3, it’s now estimated the growth rates are 2.2% in Q2 and 3.0% in Q3.
Nothing nefarious, just updates to data. Take a break on your ears, and try removing the tin foil hat.
”We’ve secretly replaced their data analysts with ferrets. Will they notice? Let’s watch…”
All the data is subject to revision no matter the scale or time period of the revision.
“The BEA revised GDP data going back to 2021 to reflect updated information.”
https://www.reuters.com/world/us/us-second-quarter-gdp-revised-higher-amid-robust-consumer-spending-2026-09-30/
“Comprehensive update recalculates quarterly statistics as far back as 1947”
“Comprehensive update recalculates annual statistics as far back as 1929”
https://www.bea.gov/news/blog/2026-08-17/annual-update-gdp-industry-and-state-stats-publicly-available-starting-sept-30
Considering the increase in the prices of domestically marketed and exported petroleum products, this can not be a surprise.
This is only the fourth increase from 2nd revision this large going back to the 1980s and including one in covid and one after the financial crisis. It smells.
Congrats. I thought there would be at least a couple of comments regarding the economics of this data and Mish’s take before the obligatory post here of “This is fake (because I know better)”.
Oh well