GitHub Data Reveals 'Digital Complexity' of Nations, Study Shows Software Activity Predicts GDP and Inequality

Breaking: Researchers Use GitHub Innovation Graph to Uncover Hidden Economic Signals

A groundbreaking study published in Research Policy reveals that software development patterns on GitHub can predict national economic outcomes—including GDP, inequality, and carbon emissions—more accurately than traditional trade data alone. The research, based on the GitHub Innovation Graph, introduces a new measure called Software Economic Complexity Index (Software ECI) that captures the digital footprint of countries.

“For the last fifteen years, economists have been measuring the complexity of national economies by looking at physical exports, patents, and research publications. But all these have a massive blind spot: software,” said Sándor Juhász, a research fellow at Corvinus University of Budapest and co-author of the paper. “Code doesn’t go through customs—it crosses borders through git pushes and cloud services. We call this the ‘digital dark matter’ of the economy.”

Background: The Invisible Economy

Traditional economic indicators rely on tangible goods—what countries export, what patents they file, and what research they publish. But software, which increasingly drives productivity and innovation, has been largely invisible. The GitHub Innovation Graph, which tracks developer activity by programming language and location (using IP addresses), offers a first-of-its-kind window into this hidden economic layer.

“We applied the Economic Complexity Index to software production data,” explained Jermain Kaminski, Assistant Professor at Maastricht University and another co-author. “The bottom line is that Software ECI significantly predicts GDP per capita, income inequality, and emissions—even after controlling for traditional complexity measures.”

How the Study Worked

The four researchers—Juhász, Kaminski, Johannes Wachs (Corvinus University and Complexity Science Hub Vienna), and César A. Hidalgo (Toulouse School of Economics)—used the Innovation Graph’s Q4 2025 data release. They analyzed the distribution of developers across programming languages in each country, then computed a Software Complexity Index similar to how economists measure product complexity from export data.

“Software as a source of economic complexity has been missing from the literature,” said Wachs. “Our work shows that digital production reveals the collective knowledge of a nation in ways that physical products cannot. Countries that produce a diverse array of complex software tend to have higher incomes and lower inequality.”

The findings held strong even when the team accounted for traditional economic variables such as export complexity, patent complexity, and research publication complexity. Software added new predictive power.

What This Means: Policy, Investment, and the Digital Divide

For policymakers, the implication is clear: fostering a vibrant software ecosystem is not just about tech jobs—it may directly influence a nation’s future economic growth and environmental performance. “Our measure helps reveal why some developing countries leapfrog stages of industrialization,” said Hidalgo. “They may not export many physical goods, but their open-source software output signals hidden productive capacity.”

For investors and international organizations, the Software ECI offers a new lens to assess national competitiveness. Countries with low software complexity risk falling behind in the digital economy, while those that invest in coding education and open-source infrastructure could see outsized returns.

“This is the first time we can measure software complexity at a global scale,” Kaminski noted. “It opens the door to tracking how AI, cloud computing, and the shift to services reshape national economies in real time.”

The research also raises questions about inequality: nations with high software complexity tend to have lower income disparity, suggesting that inclusive digital skills programs might reduce economic divides.

Full Study and Data Access

The paper is available in Research Policy and uses publicly accessible data from the GitHub Innovation Graph. The team plans to update the Software ECI as quarterly data flows, enabling real-time monitoring of digital economic change. “We hope other researchers build on this,” Juhász added. “Software production is now measurable, and that changes everything.”

For more details, see the background section or jump to what this means.