bacipair

bacipair estimates a treatment effect on a paired Before-After-Control-Impact (BACI) panel with staggered treatment timing -- paired difference-in-differences with cluster-robust standard errors, a simplex-weighted synthetic-control comparator, a proponent-level cluster bootstrap, and a Goodman-Bacon diagnostic -- and grades every estimator against a known-oracle simulator before it is trusted on a real register.

Related packages

• did estimates group-time average treatment effects in difference-in-differences designs with more than two periods and staggered treatment timing (Callaway and Sant'Anna, 2021).
• Synth builds a synthetic control unit from a weighted combination of comparison units for a comparative case study of one treated unit.
• bacondecomp decomposes a two-way fixed-effects difference-in-differences estimate under staggered timing into its 2x2 comparisons (Goodman-Bacon).

bacipair is built for paired BACI panels, where each impact site has its own control. It puts paired difference-in-differences and synthetic control side by side on the same panel, uses bacondecomp for its staggered-timing diagnostic, and checks each estimator against a simulator with a known effect before it is used on real data.

Installation

# install.packages("remotes")
# remotes::install_github("max578/bacipair")
install.packages("bacipair", repos = c('https://max578.r-universe.dev', 'https://cloud.r-project.org'))

Example

library(bacipair)

sim <- simulate_staggered_panel(seed = 20260520L, n_pairs = 80L)
fit <- synth_control_baci(sim$panel)

cat(sprintf("ATT_hat = %.3f pp  (oracle = %.3f pp)\n",
            fit$panel_att_hat, sim$oracle_panel_att))

See vignette("getting-started", package = "bacipair") for the estimand and estimator mathematics, the paired-DiD comparator, the cluster bootstrap, and the Goodman-Bacon staggered-DiD diagnostic.

Contributing

Bug reports and suggestions are welcome through the issue tracker.

Citation

If you use this package, please cite both the package and the source methodology paper:

Macintosh, A., et al. (2024). Australian human-induced native forest regeneration carbon offset projects have limited impact on changes in woody vegetation cover and carbon removals. Communications Earth & Environment, 5, Article 149. https://doi.org/10.1038/s43247-024-01313-x

Licence

MIT