Will AI make companies outsource more, or less?
The article explores how AI might affect corporate outsourcing decisions, arguing that AI could reduce outsourcing by making in-house work more efficient, but also enable more outsourcing by lowering coordination costs. The outcome likely depends on whether AI becomes a general-purpose technology that firms integrate internally or a specialized tool best used through external vendors.
Background
- The author, Noah Smith, is a prominent Substack writer (Noahpinion) known for sharp takes on economics, tech, and AI. He often synthesizes academic research for a general audience.
- This post responds to a debate sparked by economist Richard Baldwin, who argued that AI will dramatically increase services offshoring (like coding or accounting) to lower-cost countries, similar to what manufacturing saw in the 1990s/2000s.
- The piece weighs two opposing forces: AI as a "coordination cost reducer" (making it easier to manage remote workers abroad, favoring outsourcing/offshoring) vs. AI as a "productivity booster" (making domestic workers so efficient that there's less need to hire abroad, favoring onshoring or keeping work in-house).
- Key framing concept: The "Alchian-Allen effect" from trade theory — that trade disproportionately favors higher-quality (or here, more productive) workers or tasks, potentially concentrating work where productivity is highest rather than where wages are lowest.