By Robert C. Fry Jr., Ph.D.
U.S. real gross domestic product grew rapidly in the second and third quarters of 2025. According to the Federal Reserve Bank of Atlanta’s GDPNow nowcast, it grew rapidly in the fourth quarter, as well. This happened even as employment growth slowed sharply. Nonfarm payrolls grew by just $93,000—less than $12,000 per month—from April to December. The unemployment rate rose during the year, but the increase was muted by the slowdown in labor force growth caused by President Donald Trump’s immigration policies. According to the Congressional Budget Office, net immigration fell from 2.27 million in 2024 to 410,000 in 2025.
If correct, the GDP data and employment data, taken together, imply a surge in productivity growth. Nonfarm output per hour rose at a 4.1% annual rate in the second quarter and a 4.9% rate in the third quarter, versus a long-term trend of about 1.5%. There are several possible explanations for the surge in productivity growth. Deregulation means there are fewer people doing nonproductive tasks. Artificial Intelligence may be replacing some workers and making the remaining workers more productive. Investment in equipment and intellectual property products (e.g., software) was very strong in first three quarters of 2025, perhaps boosted by immediate expensing of investment, which was made retroactive to the beginning of the year. And to some extent, productivity growth is just a necessity. If you can’t find workers, you do what you can to get your existing workers to work harder and smarter.
The other possibility is that either the employment data or the GDP data (or both) are wrong—or at least misleading. The strong growth in GDP in the second and third quarters (3.8% and 4.3% annual rates, respectively) came after a decline in the first quarter. That decline was due to a big increase in imports in anticipation of President Trump’s tariffs. Mathematically, imports reduce GDP but only if they replace domestic production. Otherwise, those imports add to the consumption and investment components of GDP, particularly the change in private inventories, which is considered investment in the national income accounts. I believe the decline in GDP in the first quarter and much of the apparent strength in subsequent quarters was caused by an understatement of inventory change in the first quarter. Putting things mildly, I don’t think the Bureau of Economic Analysis is particularly good at measuring inventories.
Governments aren’t good at measuring exports either. (That’s why the world, as a whole, runs at a huge trade deficit.) They have a strong incentive to measure imports well because they collect tariffs on imports, but they have no such incentive to measure exports well. Because of the inability to measure inventories and exports well (and a philosophical belief that more government spending is not a good thing), some economists prefer to focus on real final sales to private domestic purchasers rather than on GDP. Real final sales to private domestic purchasers removes the impact of inventories, net exports, and government spending from GDP, thus focusing on consumer spending and fixed investment. Real final sales to private domestic purchasers rose at a 1.9% annual rate in the first quarter, a 2.9% rate in the second quarter, and a 3.0% rate in the third quarter: still strong but much “smoother” than GDP.
But even if you remove the impact of inventories, net exports, and government spending, it can be difficult to reconcile the remaining components of GDP with other economic data. Housing starts and home sales, as well as investment in residential structures (where housing starts and home sales show up in GDP), have been declining for most of the last two years, although they’ve shown signs of a nascent upturn in recent months. While residential construction accounts for only 4% of GDP, the indirect impact of housing on the economy has historically been much larger than that. That’s because people who buy or sell or build houses generally buy things to put in those houses. The weakness in housing would normally pull down consumer spending by pulling down spending on carpeting, furniture and furnishings, appliances, paints, etc. But consumer spending has remained strong despite the weakness in housing. Housing, which has historically been a very reliable leading indicator, has called for two recessions in the last four years. The first never happened. The second doesn’t seem to be happening either.
Industrial production in U.S. manufacturing rose 0.2% in December, leaving it up 2.0% year-over-year but still well below the cyclical peak reached in September 2018 and the all-time high reached in December 2007. Recent growth in industrial production lies between the strong growth in GDP and the weak growth in employment. Industrial production indexes for many industries are based on hours worked, not on actual measures of physical production. Slow growth in employment is holding down measured growth in industrial production in these industries.
In the case of inflation, we must reconcile the numbers with what we thought would happen, as well as reconciling one data series with another. Although tariffs have definitely raised the prices of goods, they haven’t had the impact on overall inflation that many expected. That’s largely because services account for most of the weight in the price indexes we use, and tariffs have been imposed only on goods. The Consumer Price Index (CPI) rose 0.3% in December, but the November index was distorted by the government shutdown, so we should probably ignore the one-month change and focus on two other measures. The CPI was up 2.7% year-over-year in December, the same as in November. Over the three-month period from September to December, the CPI rose at a 2.1% annual rate. Excluding food and energy prices, the CPI was up 2.6% year-over-year in December, the same as in November but otherwise the lowest since March 2021. From September to December, it rose at just a 1.6% annual rate.
The small increase in the core CPI from September to December suggests inflation might be approaching the Federal Reserve’s 2% target, but the annualized three-month change fell to 1.7% in 2024 and earlier in 2025 before reaccelerating. So it’s too soon for the Fed to declare victory, especially after Producer Price Indexes rose much more than expected in November (latest data available), casting doubt on the veracity of the CPI data. And it’s likely that inflation will rise in the first half of 2026 due to tariffs (which haven’t been fully passed through yet), rate cuts (if you still think rate cuts are stimulative; I don’t), and the large tax refunds that will result from the One Big Beautiful Bill Act. Those refunds will boost growth, especially in the second quarter, but they will also boost inflation. After a strong second quarter, growth will slow to a more sustainable trend, unless continued strong productivity growth allows strong growth to continue or a shock (e.g., much higher oil prices or much lower stock prices) ends the expansion. █
Robert C. Fry Jr., Ph.D. is chief economist of Robert Fry Ecnomics LLC. He can be reached at 302-743-8553, robertfryeconomics@gmail.com, or www.linkedin.com/in/robertcfryjr.
