GDP says 1.5%; GDI says 2.2%. The 1.8% middle is a measurement hedge

The Department of Commerce complex at 4600 Silver Hill Road in Suitland, Maryland; the Bureau of Economic Analysis lists the same address for its offices. This undated General Services Administration photograph supplies institutional context; a building cannot show which national-account estimate is closer to the truth.[7][8]

Priced: second-quarter real GDP slowed to a 1.5% annualized rate from 2.1% in the first quarter. New: real gross domestic income accelerated to 2.2% from 1.2%, the equal-weighted average of GDP and GDI grew 1.8%, and real final sales to private domestic purchasers advanced 4.2%.[1] The gap does not reveal a second economy. It says the familiar slowdown headline is more certain than the underlying measurement—and broader private demand looks firmer than 1.5% on its own.

That is a narrower claim than “GDP is wrong.” The expenditure and income accounts are two imperfect routes to the same destination. Their disagreement is useful because it marks where source data, timing and estimation have not yet reconciled. With the Bureau of Economic Analysis preparing an unusually broad annual update on September 30, the spread is a live indicator of measurement uncertainty—not a confidence interval or a menu from which investors should select the more convenient growth rate.[2][3][4]

Evidence cut-off: September 12, 2026 at 23:38 UTC. Quarter-to-quarter growth rates are seasonally adjusted annual rates. “Priced” and “new” describe an analytical framing, not a claim about a specific asset's market price. Policy and portfolio implications are interpretation, not investment advice. BEA's combined September 30 third-estimate and annual-update release will supersede the current vintage.[1][4]

Image context: the cover is a real GSA photograph of the federal complex at the address BEA lists for its offices. It locates the institution responsible for the national accounts; it is not evidence for any growth, inflation or profit estimate.[7][8]

One economy, counted from opposite sides

GDP begins with final expenditure: personal consumption; gross private domestic investment, comprising fixed investment and the change in private inventories; government consumption and gross investment; exports; and, as a subtraction, imports. That route asks what buyers paid for domestically produced final goods and services.

GDI begins with the income created while producing those same goods and services. It assembles compensation of employees, taxes on production and imports less subsidies, net operating surplus, and consumption of fixed capital. That route asks who received the value created in production.[2][3]

In principle, the totals must be equal. A restaurant meal counted as consumption also becomes compensation, taxes, operating surplus and the economic depreciation of equipment on the income side. In practice, BEA cannot observe every expenditure and every income at the same instant from one complete ledger. The two estimates draw on different surveys, administrative records and reporting clocks. Corporate profits, bonuses and stock-option income are among the items that can be earned over time but recorded when paid. Sampling, coverage and timing differences create the residual BEA calls the statistical discrepancy.[2]

This is why “GDP versus GDI” is the wrong contest. Neither is a direct meter attached to the economy. GDP is featured because its quarterly source data are generally available sooner; GDI arrives later as more income-side information, including corporate profits, becomes available. The disagreement is not an accounting violation. It is the visible seam between two estimation systems.

The aggregate slowed; the private core did not

The second-quarter vintage tells three different but compatible stories.

First, real GDP decelerated. Its 1.5% annualized gain followed 2.1% in the first quarter, as lower government spending, slower investment and exports, and a larger increase in imports outweighed faster consumer spending. Imports subtract from GDP by construction because they are not domestic production.[1]

Second, the income side did not confirm the same loss of momentum. Real GDI rose 2.2%, up from 1.2% in the first quarter. Averaging the two measures produces 1.8% growth after 1.7%. GDP therefore says “slower quarter”; GDI says “faster quarter”; their middle says “little change.” The sensible takeaway is not to crown one winner but to lower confidence in the direction implied by either measure alone.[1][2]

Third, the domestic private-demand measure was stronger. Real final sales to private domestic purchasers—consumer spending plus private fixed investment—grew 4.2%. It strips out inventories, net exports and government, three components that can make headline GDP noisy without describing the same change in household and business final demand. It is not recession-proof: consumers and companies can still weaken, and a single quarter can be revised. But at this vintage, the part of spending most closely tied to private domestic demand was not behaving like a 1.5% economy.[1]

The profit data lean in the same direction without settling the issue. BEA's profits-from-current-production measure rose by $400.9 billion in the quarter, after a $74.4 billion increase in the first. That measure includes inventory-valuation and capital-consumption adjustments, so it is not S&P 500 earnings and should not be dropped into an equity multiple. It does show why the income account could look firmer while the expenditure headline slowed.[1]

The average is useful precisely because it is modest

BEA publishes an equal-weighted average of GDP and GDI as a supplemental measure. Reliability research cited in the national-accounts handbook finds that combining the two can damp some of their distinct measurement inconsistencies. The National Bureau of Economic Research also considers the average among the indicators used when dating business-cycle turning points.[3]

Calling 1.8% a measurement hedge is deliberate. An equal weight does not prove both estimates are equally accurate in every quarter, and it does not manufacture new observations. It simply refuses to put the entire analytical burden on one source system when the two disagree.

There is an important counterweight. BEA says early GDI estimates have not shown statistically significant power to predict later GDP revisions. A higher GDI print is therefore not a reliable advance notice that GDP will be marked up. Over time the two tell a similar story; within a fresh quarter, the income side can be late, volatile and revised.[2]

The average is most valuable as a discipline. If GDP, GDI and private final demand all point in the same direction, conviction can rise. When they split, forecasts should carry wider error bars. That affects how investors read rate sensitivity, cyclicals and earnings risk even if no trading rule follows mechanically from the arithmetic.

Growth did not get a clean dovish interpretation

The strongest market counterargument is that the distinction may not matter. The Federal Reserve targets inflation and employment, not the GDP–GDI spread. The July minutes showed a 9–3 decision to maintain the 3.50%–3.75% target range; three members preferred an immediate 25-basis-point increase, and many participants assessed that tightening would likely be necessary if inflation did not decline.[6] A growth estimate that is merely “less soft” does not dictate the next rate move.

The same BEA release reinforces that boundary. The gross domestic purchases price index rose at a 5.8% annualized rate, while the PCE price index excluding food and energy rose 3.6%.[1] Those readings make a firmer income-side growth signal less comfortably bullish for duration: resilient demand can support earnings, but it can also preserve price pressure and a restrictive policy path.

For equities, the $400.9 billion profit increase is encouraging but not a free pass. National-account profits cover a different universe and use different adjustments from listed-company earnings. For bonds, 1.8% average growth is softer than 4.2% private final demand, but neither figure cancels the inflation data. For macro positioning, the honest conclusion is frustrating and useful: Q2 looked slower in total output, firmer in incomes and private demand, and still hot in prices.

September 30 is the revision event that matters

BEA's annual updates ordinarily replace early source data with more complete information, incorporate revised seasonal factors and can rewrite the path of prior quarters. The September 30 national and industry update revises the period from the first quarter of 2021 through the first quarter of 2026; the same-day Q2 third estimate is a current-quarter revision. State results also begin September 30, while county GDP and personal-income results follow on December 2.[4][5]

That breadth matters more than a cosmetic third decimal. A revision can alter the level from which growth is calculated, reassign activity among components, and change the apparent sequence of acceleration and slowdown. The combined September release will contain the regular third estimate for Q2, revised corporate profits and new industry detail alongside the annual update.[1][4][5]

Investors should separate two questions when it arrives. Did the estimate of Q2 growth change? And did the history leading into Q2 change? A stable latest-quarter rate sitting on a revised base can alter the economic story even when the headline appears unchanged.

Falsifier

The current interpretation—that the GDP headline overstates Q2's loss of private momentum—loses both stated supports if the September 30 release revises real GDI and real final sales to private domestic purchasers to rates no stronger than revised real GDP. GDI tests the income-side leg; private final demand tests the domestic-demand leg. This is an author-set falsifier, not a BEA forecast. A narrower GDP–GDI gap by itself would improve measurement agreement; it would not decide whether the convergence happened at a strong or weak rate.

Four dated checks

September 15–16, 2026: the FOMC meeting arrives before BEA's combined September release. Watch whether the committee treats demand resilience or price persistence as the binding constraint, while remembering that policymakers are still working with the current national-account vintage.[6]
September 30, 2026 at 8:30 a.m. EDT: compare revised GDP, GDI, their average and private final demand—not just the third-estimate GDP headline. Then inspect which history and components the annual update changed.[4][5]
October 29, 2026 at 8:30 a.m. EDT: the advance Q3 GDP estimate supplies the next expenditure-side read. GDI and corporate profits are not part of that first release, so the two-route comparison will again be incomplete.[5]
November 25, 2026 at 8:30 a.m. EST: the Q3 second estimate and corporate-profit release should restore the income-side comparison. Test whether GDP, GDI and private final demand finally point in the same direction.[5]

The practical edge is not choosing 2.2% because it is higher or 1.5% because it is familiar. It is recognizing that the gap signals measurement uncertainty, the 1.8% average hedges against it, and 4.2% identifies where current expenditure strength actually sits. Until the September release incorporates more complete data, the second quarter was neither a clean slowdown nor a clean reacceleration. It was an economy whose two scorekeepers agreed on the game and disagreed on its pace.

cronfeed.work