Economist warns "missing intercept" traps distort AI wage-impact papers

paulnovosad · x · 2026-09-16

Economist Paul Novosad cautions that papers on negative wage/employment effects of AI must be read very carefully: the same result can be presented in two very different ways. In his view, a finding framed as declining wages for AI-exposed workers does not justify concluding those workers are worse off—the issue lies in how the intercept/baseline is handled. Arpit Rague amplified it as the latest in a continuing "missing intercept" series.

Related event: Economists Clash Over How to Read GenAI's Wage Effects(11 posts)→

Original post →

More from AGI Musings

AGI Musings channel →