Autor, Dorn, and Hanson (2013) argue that the “China shock” negatively affected the United States from 1990 to 2007. Borusyak et al.’s (2022) corrections to the research design reaffirm the impact of Chinese imports on the manufacturing share of employment, but negative effects on living standards—unemployment, labor force participation, wages, and household incomes—disappear. Subsequent findings reported by Autor et al. (2019; 2020; 2021) are similarly fragile. Furthermore, Autor et al.’s famous claim that Chinese imports caused “one-quarter” of all national job losses in manufacturing rested on the whimsical assumption that partial R² could be used to scale up their local results. The “China shock” narrative thereby illustrates how the “credibility revolution” failed to take the con out of econometrics.