The IT Productivity Paradox: Insights from the NBER
Abstract
Despite enormous IT investments, productivity statistics have been remarkably flat. This paper examines possible explanations: measurement errors, time lags, redistribution of profits, and mismanagement. The paradox suggests that simply investing in technology is insufficient — organizational complementary investments are required to realize productivity gains.
★ Curator Summary
Brynjolfsson's IT Productivity Paradox from 1993 is playing out again with AI in contact centers. Organizations are investing billions in conversational AI, chatbots, and agent assist tools — and struggling to show measurable productivity gains. The four explanations Brynjolfsson identifies — measurement errors, time lags, redistribution, and mismanagement — map directly to what we see: organizations measuring containment rate (wrong metric), expecting instant ROI (time lag), shifting costs between departments (redistribution), and deploying AI without redesigning workflows (mismanagement).
Why It Matters
This paper arms WFM practitioners with a framework for explaining to leadership why AI investments aren't producing expected results. It's not that AI doesn't work — it's that productivity measurement is hard, benefits take time to materialize, and organizational change is required alongside technology deployment. The complementary investments lesson is directly applicable: AI without workflow redesign, skill development, and planning model updates will underperform.
Caveats
Written in 1993 about IT broadly, not AI specifically. Some argue the paradox was eventually resolved as measurement caught up. The current AI wave may be different from historical IT waves in important ways. Still, the structural lessons about technology adoption remain highly relevant.
Discussion (1 comment)
Testing discussion post.
Loading...