The Geography of Value Capture

In the AI Story

The economic specifics matter. The Lagos developer pays for AI tools in dollars — a currency whose acquisition cost, relative to her local income, is substantially higher than for a developer in a dollar-denominated economy. She hosts on cloud infrastructure priced for global markets but disproportionate relative to her revenue potential. She distributes through app stores charging commission rates calibrated to developed-world expectations. She processes payments through systems imposing currency conversion costs and compliance requirements. At each step, rent flows outward to geographic locations far from her workspace. She captures the residual.

The San Francisco developer faces none of these disadvantages. She operates within the institutional ecosystem the value chain was designed to serve. She pays in the currency she earns, accesses venture capital through proximate networks, distributes to a culturally familiar market, operates under robust IP protection, and benefits from complementary infrastructure that amplifies AI productivity rather than taxing it. Same tools, same nominal capability, dramatically different returns — determined not by talent but by institutional proximity.

The development-economics analog is instructive. Throughout the twentieth century, the primary pathway to closing the gap between rich and poor nations was industrialization — building domestic manufacturing capacity that captured value locally rather than exporting raw materials for processing abroad. The countries that successfully industrialized — South Korea, Taiwan, China — moved from the bottom of the global distribution toward the middle and top through institutional investment: industrial policy, protected infant industries, educational infrastructure, financial systems funding expansion.

The AI-era analog would be institutional investment in the infrastructure that determines local value capture: educational systems preparing workers for AI-complementary roles, financial systems funding AI-augmented enterprises, digital infrastructure maximizing tool productivity, regulatory frameworks requiring some share of value capture domestically, and platform alternatives reducing dependence on the concentrated infrastructure of wealthy nations. Without this investment, AI produces a digital periphery — economies that participate in the AI value chain but capture only a fraction of the value they help create.

Origin

The concept extends Milanovic's citizenship premium analysis by identifying the specific mechanisms through which institutional differences translate into value-capture asymmetries in the AI economy. It also connects to a longer tradition in dependency theory and world-systems analysis, though Milanovic's framework differs in treating the core-periphery structure as modifiable through institutional investment rather than as structurally determined by the world economy's architecture.

Key Ideas

Capability is not capture. Democratizing access to tools is not the same as democratizing the economic returns from using them. Most of the AI discourse conflates the two.

Every layer extracts rent. Cloud, model API, app store, payment processor, advertising platform — each layer of the AI value chain is an opportunity for geographic rent extraction.

The residual is what's left. Peripheral builders capture what remains after every layer of infrastructure has taken its share. The residual can be positive — she is better off with the tools than without — and still represent a minor fraction of the value she created.

AI amplifies institutional asymmetry. The multiplicative relationship between institutional quality and AI productivity means nominal capability convergence coexists with absolute outcome divergence.

Institutional investment is the path. Closing the gap requires the kind of sustained institutional construction that enabled industrial-era escapes from peripheral status — not faster tools, but better institutions.

Debates & Critiques

Some techno-optimists argue that AI will enable distributed development patterns that bypass the traditional core-periphery structure, with knowledge workers in low-cost jurisdictions serving global markets directly. The empirical evidence so far suggests that the direct-service pattern benefits individual workers without shifting the aggregate distributional structure, because the platforms that mediate the direct service capture the majority of the value. Individual winners in the periphery do not aggregate into peripheral value capture.