Import Replacement

In the AI Story

The process is gradual, messy, and impossible to plan from above. Tokyo grew from a provincial city importing American bicycles in the 1870s to an industrial powerhouse through a century of such substitutions. Manchester grew from a market town importing finished cloth to the textile capital of the industrial revolution through the same process. Every city that has achieved durable economic vitality has done so through thousands of small acts of local substitution, each adding a new node to the economy.

The digital economy of the 2010s operated in the opposite direction, producing import dependence rather than import replacement. A marketing manager needing an analytics dashboard imported Salesforce or HubSpot — standardized tools made elsewhere, designed for the median user, controlled by priorities other than her own. The gap between the tool's generic design and her specific need was economic leakage: value that could have been created locally was lost to the standardization required by the platform's business model. This was the structural condition AI tools have disrupted.

When AI collapses the imagination-to-artifact ratio, the marketing manager can build her own dashboard in days rather than importing one over months. This is import replacement in precise Jacobsian terms: she has replaced an imported product with a locally produced substitute better suited to her specific needs. She has not merely saved money. She has added a capability to her local economic ecosystem — the capability to modify the tool when her needs change, to understand what it does, to build the next tool when she needs one.

The teacher who builds educational software for her specific students, the architect who builds analysis tools suited to the scale of her practice, the nurse who builds patient-tracking tools for her ward — each is engaged in the same process, at the scale Jacobs identified as most consequential. The significance is never in the individual act. It is in the aggregate: when thousands of practitioners each replace an import with a locally created substitute, the economy diversifies at a rate no policy could achieve. This is the democratization of capability read through a framework that specifies what makes it economically significant.

But Jacobs would identify a structural problem immediately: the AI-enabled builder has replaced dependence on Salesforce with dependence on Anthropic. The new dependency is different in kind — broader, deeper, more foundational — and concentrated in a handful of companies funded by catastrophic money. Whether the builders can generate enough genuine independence to sustain diverse local economies is the question on which the framework's application turns.

Origin

Jacobs developed the theory across The Economy of Cities (1969) and Cities and the Wealth of Nations (1984), building on her earlier observations of urban vitality but shifting to explicitly economic analysis. The theory was developed against the backdrop of postwar development economics, which emphasized export-led growth and foreign investment as engines of development — the pattern Jacobs identified as catastrophic money and as the opposite of genuine development.

Key Ideas

Growth through substitution, not specialization. Regions grow by adding capabilities locally, not by deepening their role in global supply chains.

Compounding diversity. Each substitution creates conditions for further substitutions. The economy becomes denser and more various over time.

Local knowledge as competitive advantage. The local entrepreneur understands needs no external supplier can see. Her replacement is better fitted because she is the user.

Independence as the goal, not efficiency. The point is not to save money on imports but to reduce structural dependency and retain local capability.

AI as enabler and as new dependency. The tool enables import replacement of software products while creating a deeper dependency on its own infrastructure.

Debates & Critiques

Mainstream development economics has largely rejected Jacobs's framework in favor of export-led growth models, though the Asian developmental state cases — Japan, Korea, Taiwan — are more consistent with her account than with Ricardian theory. Critics note that import replacement can degenerate into protectionism and inefficiency if pursued as national policy. The distinction matters: Jacobs described an organic process emerging from local entrepreneurship, not a top-down industrial policy. The AI application raises a different question: whether replacing one imported product with another produced using imported infrastructure counts as substitution or merely as relocation of dependency. The answer likely depends on whether the builders can, over time, replace components of the AI infrastructure itself.