Writing Code vs. Shipping Code: Productivity Effects Across Generations of AI Coding Tools

2026-06-05 · source · source ↗ · wiki page

Abstract

How do the productivity effects of AI evolve across successive generations of tools, and to what extent do task-level gains ultimately translate into final output? We study these questions in the context of software development, using data on more than 100,000 GitHub developers combined with their AI usage telemetry. In a matched event study design, we find that autocomplete, interactive coding agents, and autonomous coding agents each significantly increase coding activity ("commits"), with respective cumulative effects of 40%, 140%, and 180%. These gains, however, attenuate sharply across the production hierarchy: the 180% cumulative effect falls to 50% for the number of projects, and to 30% for actual releases. This pattern is consistent with the weak-link hypothesis: the strong productivity gains from AI are attenuated by human bottlenecks in the production chain, with an estimated elasticity of substitution of 0.25 between AI and human effort, which indicates strong complementarities. We further confirm these results across four major app marketplaces, finding a moderate increase in the number of new apps but no increase in total usage. Large task-level AI productivity gains have therefore translated only partially into shipped and used software thus far.

Authors

Mert Demirer — Massachusetts Institute of Technology (MIT) - Sloan School of Management
Leon Musolff (Contact Author) — University of Pennsylvania - Business & Public Policy Department
Liyuan Yang — Massachusetts Institute of Technology (MIT) - Sloan School of Management

Keywords

generative AI, productivity, software development, multi-stage production

JEL Classification

D22, D24, L86

Versions

MIT Sloan Research Paper (forthcoming) — 96 pages, Posted: 01 Jun 2026
NBER Working Paper No. w35275 — 97 pages, Posted: 03 Jun 2026

Content fetched via

Browser automation (surf) due to Cloudflare anti-bot protection on SSRN. Page text captured 2026-06-05.