The greenfield gap: America's record FDI year was bought, not built

The Bureau of Economic Analysis released its preliminary count of new foreign direct investment in the United States on 10 June, and the headline is emphatic: foreign investors spent $232.2 billion to acquire, establish or expand US businesses in 2025, up 49.5% from a revised $155.3 billion in 2024. Planned total expenditures, including future outlays on projects initiated last year, reached $284.5 billion. After two years of tariff escalation explicitly designed to pull capital onshore, the investment response looks, at first glance, like vindication.

The composition tells a different story. Of the $232.2 billion, $218.4 billion (94%) went to acquisitions of existing US businesses. Expenditures to establish genuinely new businesses were $4.6 billion, and expansions of existing foreign-owned operations added $9.2 billion, putting total greenfield outlays at $13.8 billion, or under 6% of the total. The employment data make the same point more sharply: of the 213,100 jobs at newly invested businesses in 2025, 211,700 came attached to acquired enterprises — existing payrolls that changed owners. Planned employment from everything initiated last year, including establishments and expansions when fully built out, adds roughly 20,700 net new positions.

The detail beneath is still useful to practitioners. Japan was the largest ultimate investor at $50.5 billion, ahead of Germany ($26.7 billion) and Canada ($23.5 billion), with Europe supplying half of all new investment. Manufacturing took $121.8 billion — 52.5% of the total — led by publishing (which includes software, at $50.7 billion), chemicals ($45.4 billion) and plastics and rubber ($19.0 billion). California ($59.7 billion), Texas ($21.5 billion) and Pennsylvania ($20.9 billion) topped the state table. But the greenfield geography is entirely different: Louisiana ($3.0 billion), Arizona ($2.7 billion) and Texas ($1.9 billion) led on new-build investment, with Australian, South Korean and Japanese investors the largest builders. Winning an acquisition headquarters and winning a new plant are different competitions with different rosters.

The gap between this measured reality and the announcement economy is now enormous. Investment trackers count roughly $1.77 trillion in US manufacturing commitments announced since January 2025, yet BEA's measured first-year greenfield expenditure by foreign investors was $13.8 billion, with planned total greenfield outlays of $66.1 billion. The reconciliation is partly definitional: most announced megaprojects are by domestic firms and never touch the FDI ledger, pledges are spread over a decade while BEA counts actual first-year spending, and some announcements simply never materialise. But the order-of-magnitude difference is the point. Announcement totals measure intent and political momentum; balance-of-payments data measure money that moved. Practitioners who conflate them in board reports will eventually be corrected in public.

There is also a policy reading. If tariffs were meant to make foreign firms build American factories, 2025's response was instead to buy American companies — a faster, lower-risk way to get inside the tariff wall that transfers ownership without adding capacity. That is not worthless: acquired firms gain parent-company capital, technology and export channels, and research consistently shows foreign-owned firms pay more and invest more than domestic peers. But it reframes the job. For US economic developers, the bulk of FDI activity now arrives as a change-of-control event in an existing local employer — making aftercare, retention and reinvestment programs the operative toolkit, not the ribbon-cutting. For IPAs outside the US, the data confirm the American pull is real but concentrated and transactional, and that the global pool of genuinely mobile new-build projects remains far smaller than the announcement noise suggests.