Useful framework. Our complementary finding: when AI accelerates an occupation's decline, workers don't just lose work — they get stuck. The mobility trap may bite hardest in high-AI-exposure declining occupations. (Song, Brand, Yang & Lachanski 2026, AEA P&P)

Show Post

Providing the first large-scale economic evidence on the frequency and purpose of workplace meetings, using administrative data from Norway, from David J. Deming, Katrine V. Løken, Alexander Willén, and Yaling Xu www.nber.org/papers/w35706

Show Post

Stress-testing the depreciation rate is where I would start. Hardware obsolescence cycles compress the amortization window to three years, raising the hurdle rate for replacing payroll. Testing convex adjustment costs on grid interconnects also breaks the smooth transition assumption.

Show Post

When executives discuss productivity on earnings calls, AI is increasingly top of mind. How bullish are U.S. firms on AI's expected impact? https://bit.ly/4h3dqti #EconSky

Show Post

A v cool paper providing a useful framework. ”In the central case, automation in the intelligence sector is completed after 6.9y, reaching its ceiling of 87%, at which point automation-driven unemployment reaches 8.4% (and average worker income fell by 5.8% relative to 2024“

Show Post