The Citizenship Premium

In The You On AI Encyclopedia

The premium operates through mechanisms that are invisible to the individual experiencing them. A developer in Lagos with AI tools produces output that is, in capability terms, comparable to output from a developer in San Francisco. But she pays for those tools in a currency whose acquisition costs a larger share of her income, hosts her product on infrastructure priced for wealthier markets, distributes through platforms charging commission rates calibrated to developed-world expectations, and operates within legal and financial institutions whose frictions add to every transaction. At each step, institutional rent flows outward to the locations where the infrastructure is owned.

The popular intuition — that global tools will erode the citizenship premium — rests on a misreading of what institutional quality provides. Institutions are not merely neutral platforms on which capability operates. They are amplifiers of capability, and the amplification is multiplicative rather than additive. Faster connectivity means faster AI iteration; reliable power means more productive hours; robust IP protection means higher expected returns on creative work; proximity to capital means faster scaling when a product succeeds. The same tool produces different returns in different institutional environments, and the differential widens as the underlying productivity gain grows.

The pattern has historical precedent in the industrial era. The textile worker in England and the textile worker in Bengal used comparable physical machinery, but the English worker's output was captured through institutions — patent law, financial markets, imperial trade networks — that directed surplus toward England, while the Bengali worker's output was extracted through colonial arrangements that captured the majority of the value to the metropole. The machinery was the same; the institutional architecture determined the distributional outcome.

In the AI era, the institutional architecture is even more asymmetric than in the industrial era because the value chain is longer and more disaggregated. Cloud providers, model APIs, app stores, payment processors, advertising platforms — each layer is an opportunity for rent extraction, and most of the layers are owned in the same small set of jurisdictions. The geography of value capture reproduces the institutional core-periphery structure through a new technical substrate.

Origin

Milanovic developed the citizenship premium concept through his decomposition work on global inequality in the 1990s and 2000s, using household survey data to separate individual-characteristic variation from location-based variation. The finding that location dominated individual characteristics was genuinely surprising to economists trained to emphasize human capital variables, and it has become one of the most widely cited results in development economics.

The concept connects to a longer tradition in political economy — Adam Smith's observation that the wealth of nations depends on institutional arrangements, not natural endowments; Douglass North's work on institutional determinants of long-run growth — but Milanovic's specific empirical contribution was to quantify the relative weight of citizenship versus individual characteristics at the global scale.

Key Ideas

Location beats individual attributes. National institutional context explains more income variation than education, talent, or effort. The meritocratic frame that organizes professional identity in wealthy nations is, at the global scale, descriptively inaccurate.

The premium is institutional rent. What looks like personal success in the core is, in significant part, the capture of value generated within institutional infrastructure that the individual did not build and cannot easily transfer.

AI amplifies rather than erodes. The same tool produces different returns in different institutional environments. Multiplicative amplification widens absolute gaps even when nominal capability converges.

The core-periphery structure persists. AI value chains direct rent to the jurisdictions where infrastructure is owned, reproducing through new technical substrates the institutional asymmetries of earlier eras.

Individual democratization is not distributional democratization. Giving everyone access to the same tools is not the same as giving everyone an equal share of what the tools produce. The first is capability; the second requires institutional construction.