Economic Complexity Index

In the AI Story

The index operationalizes Hidalgo's broader claim that productive knowledge is the fundamental determinant of economic development. Countries with high economic complexity grow faster over subsequent decades than countries with low complexity, even after controlling for income, education levels, governance quality, and every other standard economic variable. The predictive power is not accidental. The complexity of what you produce today reveals the depth of what you know, and that depth predicts the breadth of what you will be able to produce tomorrow.

The product space — the network visualization of proximity between products based on shared productive requirements — reveals that development is movement through a constrained topology. Countries move from products they currently make to nearby products sharing similar productive requirements. You cannot leap across the product space. The path is determined by what you already know how to do. This makes development path-dependent in ways that shape which countries can pursue which strategies.

AI creates new adjacencies in the product space. Products and capabilities that were previously distant have been brought closer by tool-mediated access to codifiable productive knowledge. A designer who had never touched backend code can, through Claude, build features end-to-end — traversing a gap that previously required years of specialized training. But these adjacencies are tool-dependent. They exist as long as the tool is accessible and collapse if access is disrupted. A country that reaches a new region of the product space by building underlying productive knowledge has made a permanent move. A country reaching the same region through tool-mediated access has made a conditional one.

The index makes a prediction that should unsettle triumphalists and reassure patient builders in equal measure: countries producing the most AI-augmented output in 2026 may not have the highest economic fitness in 2036. Output is cheap. Institutional capability is dear. The index measures what no productivity metric captures — the depth of what a country knows how to do, which determines the breadth of what it will be able to do next.

Origin

Hidalgo developed the index with Ricardo Hausmann at Harvard's Center for International Development, publishing the foundational methodology in PNAS in 2009 and later operationalizing it as the Atlas of Economic Complexity. The project grew from Hidalgo's conviction that standard development economics was asking the wrong question — what do countries have? — rather than the right one — what do countries know how to do? Mapping national exports according to their revealed comparative advantage and the ubiquity of the products they made produced a measure that captured productive knowledge indirectly but robustly.

Key Ideas

Complex products reveal deep knowledge. The sophistication of a country's output is a visible signal of the invisible productive knowledge embedded in its institutional fabric.

The index outperforms conventional predictors. Countries with high complexity grow faster than countries with low complexity, even after controlling for every other standard variable.

Adjacency constrains development. Countries move through the product space in small steps; the path is determined by accumulated capability, not by ambition alone.

AI creates tool-dependent adjacencies. New regions of the product space become accessible through AI-mediated knowledge, but the adjacency collapses if the tool becomes unavailable.

Fitness is being reshuffled. The knowledge distribution that determined national fitness is being restructured by a technology that makes codifiable knowledge universally accessible while leaving tacit capabilities untouched.

Debates & Critiques

The index has been criticized for being a backward-looking measure — it captures what countries have produced historically but may underweight emerging capabilities that have not yet translated into export sophistication. Hidalgo acknowledges the limitation and has responded by developing complementary measures of research complexity and technological capability. The deeper debate concerns whether the index's predictive power will survive the AI transition, which may decouple productive capability from the historical product mix in ways the index does not yet capture.