Run causal models on outcomes that moves cannot plausibly change (e.g., birth length) to test whether observed 'effects' are actually selection artifacts. Eshaghnia shows that substituting birth length for adult earnings reproduces Chetty–Hendren–style exposure gradients, with stronger alignment the earlier the move—something a true causal neighborhood effect on earnings shouldn’t mimic on an inborn trait.
— If marquee neighborhood-effects results fail placebo checks, policymakers must revisit relocation and 'opportunity mapping' initiatives and demand stronger identification before scaling.
Arnold Kling
2025.09.10
85% relevant
Kling cites James Heckman and Sadegh Eshaghnia arguing Chetty/Hendren’s neighborhood effects are confounded by self‑selection and parental traits, aligning with Eshaghnia’s placebo‑outcome critique that exposure‑gradient results can be artifacts rather than causal neighborhood effects.
Alex Tabarrok
2025.09.06
100% relevant
Eshaghnia’s WSJ-linked placebo finding: children’s birth‑length ranks track destination–origin birth‑length differences, ≈0.044 stronger per year earlier parental move.
← Back to All Ideas