Datasets
The stacked package ships five datasets — two real-world empirical panels, one individual-level microdata sample, and two tiny deterministic teaching examples. Every dataset is available in both R (.rda) and Stata (.dta) form.
| Dataset | Rows | Unit | What it teaches | Links |
|---|---|---|---|---|
medicaid |
714 | state × year | Headline staggered-adoption panel: ACA Medicaid expansion and adult uninsurance | rda · dta |
smoke |
10,615 | county × day | Staggered treatment on a date variable; daily PM2.5 around the first smoke plume | rda · dta |
acs_micro |
69,820 | person | Repeated cross-sections, survey weights, and propensity-score weighting | rda · dta |
signflip |
72 | unit × year | Deterministic sign flip: unweighted TWFE gets the sign wrong, Q-weights fix it | rda · dta |
unparallel |
120 | unit × year | Deterministic non-parallel trends: how a pre-trend shows up in the event study | rda · dta |
Note
The Stata package additionally ships pscore_stack.dta, a pre-built stacked dataset used by its propensity-score certification tests. It is not part of the R package.
Loading a dataset
data(medicaid, package = "stacked")stacked use medicaid, clear