Doubly robust DID estimator (stationary, repeated cross-sections)

Usage

drdid_stationary(
  y,
  d,
  post,
  fit.ps,
  fit.or,
  stabilized = TRUE,
  i.weights = NULL,
  boot = FALSE,
  nboot = NULL,
  boot_type = c("mammen", "normal", "bayes", "wild"),
  ps_min_treat = 0.005,
  inffunc = TRUE
)

Arguments

y: Outcome vector.
d: Treatment indicator (1 treated, 0 control).
post: Post-period indicator (1 post, 0 pre).
fit.ps: Matrix of generalized propensity score estimates (n x 4).
fit.or: Matrix of outcome regression fitted values (n x 4).
stabilized: Logical; use stabilized weights.
i.weights: Optional sampling weights vector (non-negative).
boot: Logical; if TRUE, use multiplier bootstrap for SE/CI.
nboot: Number of bootstrap draws.
boot_type: Multiplier bootstrap type: "mammen", "normal", "bayes", or "wild".
ps_min_treat: Minimum treated-group propensity used in stationary weights. The default (0.005) clips fitted treated propensities to [ps_min_treat, 1 - ps_min_treat] to avoid unstable inverse-probability weights. Set to 0 to disable clipping.
inffunc: Logical; include influence function in output.

Value

List with ATT estimates, uncertainty, and metadata.