The state-panel models use fixed effects, year indicators, lagged mortality and population weights. State prohibition coefficients of .238, .307 and -.089 differ across periods. Policy endogeneity and omitted state-time factors limit causal conclusions; these regressions were not independently reproduced.
16
13
PDF15–18; Table7/PDF35
Full44-page source and figures/tables read; paraphrase checked with inferential limits. No independent model replication.
No access
Evidence
Paraphrase, not a direct quotation.
The state-panel models use fixed effects, year indicators, lagged mortality and population weights. State prohibition coefficients of .238, .307 and -.089 differ across periods. Policy endogeneity and omitted state-time factors limit causal conclusions; these regressions were not independently reproduced.
Location and context
Printed pp.13–16; PDF15–18; Table7/PDF35. PDF pagination includes two preliminary pages.
Annotation
Historical-source analysis, not current health or legal advice. Verified here means faithful to this source, not independent validation of its causal conclusions.
Verification checklist
Dills–Miron 2003 — State regressions are sensitive to period and specification