The Bureau of Economic Analysis will release the second estimate of Q2 2026 GDP on Tuesday, 26 August. The advance estimate, published in late July, showed annualised real GDP growth of 1.5 percent — below the long-run trend rate and well below the Q1 2026 print of 2.3 percent. The second estimate incorporates additional source data that was not available when the advance figure was constructed. Understanding what that additional data typically revises, and what it does not, is the correct frame for reading Tuesday’s release.
What the second estimate adds
The BEA’s advance GDP estimate is constructed from preliminary source data covering approximately 70–75 percent of the income and expenditure components. The second estimate adds three material inputs: final data on trade in services (imports and exports of services are measured with a lag relative to goods trade, which is captured more completely in the advance figure); revised inventory data from the Census Bureau’s Monthly Wholesale Trade Survey; and revised data on certain federal government expenditure categories.
Of these three, inventories are historically the largest single driver of revision between the advance and second estimates. Inventory investment is volatile quarter-to-quarter and difficult to measure from early data; it is also a component that matters for the interpretation of underlying demand rather than just the headline number. A positive revision driven by inventory accumulation tells a different story than a positive revision driven by upward-revised consumer spending.
The BEA’s own research on revision patterns over the 1993–2023 period finds a mean absolute revision between advance and second estimates of approximately 0.5 percentage points (annualised), with a range that occasionally extends to ±1.0 percentage points in quarters with unusual inventory or trade in services dynamics. At 1.5 percent advance, a downward revision of 0.5 points would put Q2 at 1.0 percent — still positive, but the kind of number that combines with the 9-3 FOMC dissent to produce a more complex communications challenge for the Fed.
What to watch in the components
Three component signals are worth tracking beyond the headline annualised rate.
Personal consumption expenditures: PCE within the GDP release is a quarterly, real-terms measure. If the revision shows an upward move in real consumer spending — specifically services spending, which has been the more durable component in the current cycle — it complicates the case that the 1.5 percent headline reflects genuine demand weakness rather than inventory drag. Strong consumer spending in a 1.5 percent GDP quarter suggests the underlying economy is running warmer than the headline implies.
Gross private domestic investment: Equipment and intellectual property investment have been the leading indicators of corporate capacity decisions in the current AI build-out cycle. A downward revision here would suggest that the capital expenditure acceleration visible in Q1 carried less momentum into Q2 than the advance data implied.
Net exports: The advance estimate often shows larger swings in net exports than the subsequent estimates, because trade in goods data (which is relatively timely) dominates the advance figure while trade in services (which is revised in the second estimate) can move the net figure materially. If the services trade balance revised significantly, it can change the apparent contribution of the external sector to Q2 growth without touching domestic demand at all.
Why the Fed will not primarily use this number
GDP is a quarterly, backward-looking measure. It tells the Fed what happened in Q2 (April–June). The FOMC meets on September 16, eight weeks into Q3. The data relevant to the September decision is the trajectory of the economy in real time — current inflation, current labour market conditions, current financial conditions. GDP-2 is useful context, not live signal.
The PCE price index for July, releasing Wednesday 27 August, is a different category of data: it measures the price index the Fed explicitly targets, for the most recent completed month, less than four weeks before the FOMC meeting. If core PCE comes in at or above 2.7 percent year-on-year, the dissenter case for hiking at September 16 is straightforwardly supported by the most recent inflation data. If it prints at 2.5 percent or below, the hold case rests on data rather than on an uncertain projection.
Read Tuesday’s GDP-2 release for what it tells you about Q2 composition — the inventory versus demand breakdown, the services trade revision, the consumer spending revision. The September 16 FOMC decision will be set primarily by the number that releases the day after.