The government measures the size of the economy two ways — and for 15 straight quarters, the spending count has topped the income count

The Bureau of Economic Analysis publishes two figures for the size of the U.S. economy, built from independent data. Gross domestic product (GDP) adds up spending; gross domestic income (GDI) adds up earnings — wages, profits, interest, rents, taxes on production. Because one person's spending is another's income, the two are the same quantity measured from opposite sides, and in theory they are equal. The distance between them has a name: the statistical discrepancy. We pulled every quarter BEA has released — 318 of them, 1947 first quarter through 2026 second quarter — from the agency's series via the St. Louis Fed. In the second quarter of 2026, GDP ran to $32,486.1 billion at an annual rate; GDI, $32,245.1 billion. The spending side was larger by $240.9 billion — 0.742% of GDP — the value on BEA's own statistical-discrepancy line, computed from the agency's unrounded levels; subtracting the one-decimal figures shown here gives $241.0 billion, off by $0.1 billion from rounding. In inflation-adjusted terms the gap was $180.0 billion in chained 2017 dollars. The direction is the story. GDP has exceeded GDI for 15 consecutive quarters, from the fourth quarter of 2022 through the second quarter of 2026 — the last quarter before the run, GDI was the larger of the two. That is the longest stretch of the spending measure topping the income measure since a run that ended in 1997. Across the 15 quarters the gap averaged $279.0 billion, or 0.951% of GDP, and was widest in the third quarter of 2023 at $459.2 billion, 1.64% of output. The persistence shows up in growth, not just levels. Since the run began, real GDP has grown 9.69%; real GDI, 8.50% — a 1.19-point gap over roughly four years, with the income side growing more slowly over the 15-quarter period. BEA, aware the two rarely match, publishes both.

The following is analysis, not fact. The arithmetic is settled — GDP minus GDI is $240.9 billion, and it has carried the same sign for 15 quarters — but the meaning is not, and the honest reading is narrow. The two measures should describe one economy. When the income side comes in lower quarter after quarter, the pattern is consistent with the income tally being understated, the spending-and-output tally being overstated, or a combination of the two ^sspeculative. These data do not separate those cases. The discrepancy is a residual, not a diagnosis; it says the books do not close, not why. What is unusual here is duration, not size ^sspeculative. A single quarter's gap of three-quarters of a percent of GDP is ordinary — the country has seen wider in both directions many times. What the full record surfaces is that the sign has not flipped since 2022, and that over the run the income tally has grown more than a point less than the spending tally in real terms. Read only the headline GDP print and the expansion looks steady; read the income side beside it and the same years look a shade weaker ^sspeculative. Which of the two better tracks the economy is a genuine open question in the research, not one this table resolves — any claim that GDI is the truer signal here is a hypothesis, not a finding ^sspeculative.

The gap does not by itself establish manipulation or identify an error in either measure — it is a known feature of measuring one thing two ways from different source data, which is why BEA reports the discrepancy openly and averages the two. Magnitude, in particular, should not be oversold: 89 of the 318 quarters on record carried a larger gap than this run's average, and the single longest stretch of GDP above GDI was not now but the 62 quarters running from 1973 into 1988. Both figures are also heavily revised — the most recent quarters most of all — so annual revisions can narrow, extend, or unwind the current streak. The levels quoted here are the present vintage, retrieved September 1, 2026. The durable claim is the one the complete series supports: the sign has held for 15 quarters and the income side has grown more slowly than GDP — not that either number is the right one.

From the topline alone, the last three years read as an ordinary expansion ^sspeculative: nominal output up, quarter after quarter of positive growth (real output rose in all but one quarter, dipping once in early 2025). Set the income measure beside it and the same span looks slightly softer and, more to the point, internally unsettled ^sspeculative — two official tallies of the same economy that have leaned the same way, income below output, for 15 quarters straight. The reassuring headline and the quieter disagreement underneath it are the same economy, counted two ways that have diverged in the same direction.