BLOG POSTS ARE PRELIMINARY MATERIALS CIRCULATED TO STIMULATE DISCUSSION AND CRITICAL COMMENT. THE VIEWS EXPRESSED ARE THOSE OF THE INDIVIDUAL AUTHORS. WHILE BLOG POSTS BENEFIT FROM ACTIVE UHERO DISCUSSION, THEY HAVE NOT UNDERGONE FORMAL ACADEMIC PEER REVIEW.
By Byron Gangnes
A little less than a year ago, I published a post explaining why the US economy was clearly NOT in a recession, despite contrary claims and the pain experienced by some families. Fast forward a year, and things look very different. Now there are signs that we may be flirting with a recession. What am I watching most closely?
What’s a recession?
While the economy is clearly slowing, to be considered a recession, it must actually contract.
Here’s a stylized view of the business cycle à la Econ 101. An economic expansion ends when a slowing economy begins to shrink in size. The high point before a recession begins is the business cycle peak. A recession bottoms out at the trough before the next expansion begins. Expansions typically last much longer than recessions.
Recessions matter because they entail lower employment, higher unemployment, reduced incomes, lower corporate profits, and other adverse economic effects on households and the companies they work for.
To know whether and when we have entered a recession, we need to know whether the expansion has peaked and the economy has begun to contract. And the economy must have made a significant and sustained downturn, because we don’t want to call every brief downward wiggle in the economy a recession. How do we decide when that has happened?
A common, simple rule of thumb for determining a recession is:
- two consecutive quarters of negative real GDP growth.
The National Bureau of Economic Research (NBER)’s Business Cycle Dating Committee, the semi-official arbiter of US business cycle turning points, has a more sophisticated definition. A recession,
- “…involves a significant decline in economic activity that is spread across the economy and lasts more than a few months.”
The NBER Committee looks at a range of monthly indicators of inflation-adjusted (real) economic activity to assess this, including. “…real personal income less transfers, nonfarm payroll employment, employment as measured by the household survey, real personal consumption expenditures, wholesale-retail sales adjusted for price changes, and industrial production.”
The NBER has tended to give more weight to some of these indicators than to others: “There is no fixed rule about what measures contribute information to the process or how they are weighted in our decisions. In recent decades, the two measures we have put the most weight on are real personal income less transfers and nonfarm payroll employment.”
They then judge whether these factors as a group demonstrate a decline in activity of sufficient “depth, diffusion, and duration” to confirm that a recession is underway.
A very clear (well, extreme) example of a recession was the “Great Recession” associated with the Global Financial Crisis of 2008-2009. You can see what that looked like in my blog post from last September cited above.
That was then, this is now
A year ago, nearly all the indicators tracked by the NBER were rising, with little sign of slowing, let alone decline. In part, this reflected continuing robust consumer spending that was driving overall growth.
But since the start of this year, US economic growth has slowed: Across several NBER measures, activity has been roughly unchanged since March. Here is a chart with these economic indicators extending through June- August 2025. (The release dates for different data series vary considerably.) While there has been variation across the six indicators, most have been essentially flat since early Spring. I have highlighted with arrows the two indicators that the NBER says they weigh most heavily.
NBER Recession Indicators and Conference Board Coincident Index, Seasonally Adjusted
In some cases, the indicators were already weak in 2024 (payroll jobs and personal income). One bounced up at the start of the year but has since stalled (industrial production).[i] Consumer spending has posted one of the weakest first-halves on record, after steady growth in 2024. Retail sales look strong, but much of that is catch-up from a truly dismal several years, and it too has flattened recently.
The Conference Board publishes an average of four of these indicators, termed a “Coincident Economic Index,” since it is intended to show the position of the economy at the current time. That Index has been flat since March.
I noted above that the NBER’s definition of recession also requires that it be “spread across the economy….” In other words, you don’t get a recession if, say, real estate alone is contracting. At present, we do see a pattern of slower payroll job growth across much of the economy, but no evidence of broad contraction.
Payroll Jobs, Annualized Percent Change, Past Four Months and Preceding 12 Months
The verdict?
To me, these data suggest a US economy that is at or near a cyclical peak and could easily tip over into recession. Whether that happens will depend in part on the causes of this year’s weakness and whether those factors continue to drag on the economy. Tariff hikes, federal layoffs, cuts (or threatened cuts) to large projects that are in part federally funded (such as renewable energy projects), and generalized policy and legal uncertainty may be playing important roles.
Could we already be in a recession? Perhaps. In June and July, initial estimates of monthly job growth turned out to be too high once more complete data became available, and a second revision to the June figure showed the first (small) net job loss since December 2020. If the early revisions are still too high, the job count may have already begun a sustained contraction.
Tuesday, we get the results of the annual benchmark of labor market series to population data in the Quarterly Census of Employment and Wages (QCEW). This major once-a-year revision will provide new estimates of labor market data through the first quarter of 2025. While the direction of revision is not known yet, a significant downward revision could suggest that the job count has already been contracting for some time.
But remember: You would need to see a sustained decline in more than just that one indicator before you would say the economy is in recession. For this reason, business cycle turning points are often not called for some time after they have begun.
No actual forecasts here
It is important to note that the “historical” data reviewed here do not provide a forecast of recession likelihood. In fact, most of the models that do exist do not (yet) suggest that a recession is imminent. Some heuristics that have often predicted recessions, such as the Conference Board’s Leading Economic Index, have been predicting a recession for some time; others, like the yield curve, are no longer signaling a coming recession. Statistical models such as Chauvet and Piger’s Smoothed Recession Probabilities indicate a very low likelihood of a near-term recession.
Will the effects of the weak NBER factors get worse or improve? Tariffs, for one, have yet to fully feed through to prices; as they do, this will weigh more heavily on consumer spending. Since that represents two-thirds of the economy, a drop in consumption would almost certainly cause a recession. Much of the rest may remain unsettled for some time. If policy uncertainty recedes, and with some fiscal stimulus coming to both households and firms, conditions could improve and a recession be avoided.
The most troubling trend?
One indicator that is particularly worrisome is employment measured by the household survey. The number of people reporting that they are employed has declined by 575,000 since April, even after an uptick in August. In part, this reflects an end to growth in the labor force itself. If sustained, a flat or declining labor force will also constrain the growth of companies’ payrolls, which have slowed to near zero in recent months.
The US Labor Force, Employment and Unemployment
The lack of labor force growth this year is almost certainly due to actual and “self” deportations since the start of the Trump Administration. According to the Pew Research Center, the number of immigrants in the US declined by more than one million people in the first half of the year, and the immigrant labor force by 750,000. Assuming unchanged federal policy, there is a high likelihood that this will worsen going forward. In the short term, a declining foreign-born labor force may produce a recession—even if labor supply tightness limits the resulting rise in unemployment.
But this also represents a looming threat to trend US growth, which depends in part on an expanding labor force. US demographics were already set to become much less favorable than in the past because of lower fertility rates and people aging out of the workforce. Fewer immigrant workers will only make things worse, with negative implications for economic growth, government finances, and the country’s standard of living.
Thanks to Steven Bond-Smith, Carl Bonham, and Peter Fuleky for helpful comments.
[i] The apparent upward jump in employment from the household survey in January in fact reflects a BLS population control adjustment to better match population change since the 2020 census. This creates an artificial discontinuity in the data.