Tag
machine_learning
4 findings
Economics (general)Agnostic Machine Learning Heterogeneity Detection in Field TrialsWhile overall average treatment effects were small and statistically insignificant (-1.4 to -2.7 percentage points, p > 0.10), generic ML inference uncovered significant treatment effect gaps between the top and bottom quintiles of 20.9 percentage points for social stigma (p = 0.009) and 12.6 percentage points for salient stigma outreach (p = 0.022).RCTLabour & employmentAlgorithmic Intermediation and Bias MitigationAlgorithmic recommendation of workers to employers improved recruit pool quality and increased overall hiring. Algorithmic writing assistance on job seekers' resumes boosted hiring probability by 8 percent. Initial hiring algorithms selected far fewer Black and Hispanic candidates, but re-engineered algorithms successfully increased demographic diversity while maintaining high hiring rates.RCTEconomics (general)Compositional Selection Shift Under Zero Average Treatment EffectsProfessional stigma framing shifted applicant composition significantly, increasing the share of rich applicants by 62% (p < 0.01) and currently working applicants by 28% relative to control, despite near-zero net application growth.RCTMethods & evidenceHonest Split-Sample ML Heterogeneity InferenceIdentified statistically significant treatment effect divergence between the top and bottom predicted individual treatment effect (ITE) quintiles for social stigma (p = 0.009 in Exp 1, p = 0.022 in Exp 2) and professional stigma (p = 0.015 in Exp 1), uncovering hidden heterogeneity where average treatment effects were near zero.RCT