The third release of Q2 GDP data contained significant positive revisions. These revisions have a positive on our forecast for the rest of this year. New monthly data, as always, were a mix of good and bad. This month the balance was about even. Overall, our revised forecast beyond the near term is little changed from a month ago.
NIPA Data
The 3rd estimate of real output for 2026Q2 increased growth of real output from 1.5% to 2.2%. This makes two quarters in a row with a 2nd estimate that was virtually unchanged from the advance release, followed by large positive revisions in the final release for the quarter.
This time, however, the normal month-to-month changes were compounded by annual benchmark revisions. This means that 3rd estimates for Q2 growth rates are calculated from different Q1 values than those used for the 2nd estimate numbers.
This said, the pattern of growth in the new data is essentially the same as a month ago – strong spending by consumers and by business on equipment and intellectual property; weak or negative growth for spending on construction and by government. The estimates for growth in consumption (raised from 3.4% to 3.8%) and in residential construction (raised from 1.3% to 2.8%) are partly due to downward revision to their Q1 value.
Overall, an economy that is a little stronger across-the-board than we thought a month ago in the pre-revision data.
Recent Monthly Data
New data as we move from summer to fall have been ho-hum. Nothing worth three-cheers, but also nothing really worrisome. And no clear trends.
Household sector data are are illustrative. Real consumption was strong in August, but real income was weak, the latter due more to higher prices than to lower earnings. Consumer sentiment was lower in both August and September (and also in preliminary October data). Home sales (both new and existing) were disappointing in August, as were housing starts.
On the business side, total industrial production was flat after four positive months. Manufacturing IP declined, after six months of growth. ISM indexes for both manufacturing and services remained in the mid 50s (that is, solidly in expansion territory), as did their sub-indexes for new orders and production. Their employment sub-indexes were both in the low 50s. For services employment this marked a switch back to expansion after three months below 50.
Finally. the labor market report for September was mostly negative, but not disturbingly so. The establishment survey had total job growth of just 29 thousand. This included -17 thousand for government, putting the private sector at +46 thousand – weak, but rwithin sight of the 60-70 thousand averages over recent months. Employment in both manufacturing and construction was modestly higher; information, finance, and business services all saw job reduction.
In the household survey the good news was a second straight large increase in the labor force (485 thousand following 683 thousand in August). The downside is that this is only enough to bring the labor force within 200 thousand of its January level. The unemployment rate increased by just 0.04% in September, but that was enough to move the reported rate from 4.1% to 4.2%.
To summarize, nothing in the new/revised data to change our view that the core of the economy – private sector spending – is heading into the home stretch of 2026 at a steady gallop.
Baseline Forecast
As shown in the chart, our revised forecast has growth at 3.8% in the just completed third quarter. This is 0.3% above our September outlook, but essentially in line with the upward revisions to first half data. After that, through the end of 2027, growth averages 2.5%, which is 0.1% higher than in our previous forecast.
In the labor market our revised forecast incorporates the September data, which was weaker than implied by our previous forecast for the third quarter. After that, through the end of next year, in line with our stronger forecast for output, we have employment growth that averages 79 thousand per month. This is 11 thousand higher than last month. It is also somewhat above our estimate of long-run labor force growth, implying a decrease in unemployment. Our forecast has the unemployment rate reaching 3.7% toward the end of 2027.
Discussion
In summary, we see an economy with a foundation that can produce growth averaging about 2.5%. Our model is producing a forecast that is consistent with this outcome. But subjectively we think this is optimistic. Not in the sense that it requires a lot of things to go right, but rather that it rests on not much going really wrong.
Some of the underpinnings of our forecast model illustrate this. Consider, for example the two economic variables that dominate day-to-day economic news – oil prices and Federal Reserve actions concerning their federal funds rate target. Both are variables in our model, and both are “exogenous” which means we simple set their paths rather than having them be determined by statistical equations that are estimated based on data for other variables.
Starting with the oil price. The variable we use is the price of West Texas intermediate crude oil. Our current path has this price averaging $90 per barrel in the current quarter and then declining during the first half of next year to $70 where it stabilizes. This is not wildly optimistic – prior to the Iran conflict our assumption was for oil in the $60s range. But it is easy to spell out scenarios with much higher prices, particularly near-term, but also in the long-run. In addition, as we are now seeing with diesel prices, crude oil is not at all the entire story for fuel prices.
Regarding interest rates. Our current assumption is that the Fed, having raised rates last month, will skip any action at their meeting later this month (just before the election), but do a second 0.25% increase in December. After that we have the funds rate stable. Baring unforeseen shocks, we think that would be the best policy. But we doubt the Fed will end up doing nothing. We just don’t know what. And given their recent record this is not a source of confidence.
And in 23 days, there is the midterm election.