Railroads of the Raj — expanded estimators

Archive index · Regression tests · Results & settings (JSON) · Methods & code

Donaldson (2018), AER. https://doi.org/10.1257/aer.20101199

1 specification. Effects are in outcome standard deviations; intervals are 95%.

Treatment adoption

Treatment adoption timing

S06-P001: seeded sample of 9 of 9 eligible units. Purple untreated; yellow treated; white unobserved.

Log real agricultural income per acre

RRR-T4-C1-REAL-INCOME · Equation (16); main-text Table 4 column 1; published page 924 · Data & fit details

278 observations · 9 units · 9 clusters · 6 treated / 3 control units.

Estimate comparison

S06-P001: point estimates, targets and available 95% intervals
Estimator Estimate (SDs) SE 95% CI Averaging
Static TWFE -0.004 0.202 [-0.470, 0.463] Coefficient
Dynamic TWFE 0.048 0.299 [-0.641, 0.737] Treated cells
BJS / did2s 0.307 0.185 [-0.120, 0.733] Treated cells
Sun–Abraham 0.013 0.401 [-0.913, 0.939] Treated cells
CS-DID (equal event-time) -0.441 0.168 [-0.769, -0.112] Equal post-event times
CS-DID (treated cells) -0.362 0.195 [-0.743, 0.020] Treated cells

Event studies

Periods −10 to +10; pointwise 95% intervals where available.

Dynamic TWFE event study BJS / did2s event study Sun–Abraham event study CS-DID event study