NBER study finds AI boosts developer productivity by 32.6%

A new working paper from the US National Bureau of Economic Research (NBER) has quantified the financial impact of artificial intelligence on developer output. Titled "The Macroeconomic Effect Of AI: Sizing The Software Engineering Channel," the study concludes that AI advancements between November 2022 and December 2025 have increased the market's expected present value of software engineering productivity by a permanent 32.6%.

To reach this figure, the academic researchers analyzed the stock market performance of enterprises operating outside the software and semiconductor sectors. The underlying theory is that companies heavily reliant on software developers stand to gain the most from AI-driven efficiency gains. Consequently, positive news regarding AI capabilities translates into higher projected profits, which are subsequently reflected in elevated share prices. By modeling these stock market reactions, the researchers isolated the specific productivity gains anticipated by financial markets.

This economic assessment comes as organizations grapple with the rising costs of deploying generative AI tools. While platforms like Anthropic Claude Code and OpenAI Codex have fundamentally altered modern software development workflows, they also introduce recurring monthly subscription fees and variable usage costs. The NBER paper provides a macroeconomic justification for these expenses, suggesting that the efficiency gains easily offset the licensing overhead.

For engineering leaders and CIOs, this 32.6% figure serves as a crucial baseline for strategic planning. Rather than relying on anecdotal evidence of speedups, decision-makers can use this data to justify investments in AI coding assistants. It provides a concrete framework for balancing payroll allocations against software licensing budgets, helping organizations determine whether paying for premium developer tools will yield a net positive return on investment.