The Atlas of Economic Complexity: supporting strategic economic planning and green industrial policy in Ministries of Finance

• Economic complexity measures an economy’s embedded knowledge, production capabilities, and patterns of specialization by comparing which economic activities tend to co-occur in different locations. While initially applied to export data, it is now being applied to the energy transition, technology, research, skills and workforce training, and scientific publications, among other things.
• Economic complexity correlates strongly with countries’ economic growth, indicating that the process of economic growth involves diversification into more—and more complex—industries. Climate action can thus offer an opportunity for green growth by supporting countries’ efforts to diversify into green industries via strategic industrial policy.
• Economic complexity analysis can help formulate green industrial policy around countries’ strengths and inform policymaking around economic growth and strategic economic planning more generally. Practical applications include identifying the industries in which a country is competitive, the nearby parts of green value chains, emerging sectors with growth potential, and whether (and why) a country has or has not been successfully diversifying in the past.
• Economic complexity analysis preserves the granularity associated with non-fungible, activity-specific, and hard-to-move capital assets and relevant know-how, which methods such as CGE modeling struggle to reflect. It also improves on older methods, such as input-output analysis, export analysis, and analysis of revealed comparative advantage.

One of the challenges of economic complexity analysis is that it often groups industries via industrial classification codes (e.g., the North American Industry Classification System [NAICS]), which are broader and less detailed than product markets. Analyses based on VAT data are being developed to overcome this concern. Moreover, the method is backward-looking in that it relies on historical data. Where the technology and market structure are changing, this may be a limitation. Bottom-up or “genotypic” approaches to measuring industry-relatedness are being developed to overcome this challenge. Using time-series data to discern how capabilities have been developing can also be helpful.

MoFs can access the Atlas through the online platform at atlas.cid.harvard.edu. It can be augmented by other datasets on firms, trade, and employment for more nuance. For in-house complexity analysis, datasets such as production networks derived from VAT data and information-matching workers to industries and occupations may be useful.