Synthetic Control Method Wrapper — synthetic_control

Wrapper function for synthetic control estimation using the augsynth package for cases where traditional matching is not feasible.

Usage

synthetic_control_method(
  data,
  outcome_var,
  unit_var,
  time_var,
  treatment_var,
  treatment_time,
  covariates = NULL,
  method = "augsynth",
  inference = TRUE,
  n_lags = 0
)

Arguments

data: A data.table containing panel data
outcome_var: Character. Outcome variable name
unit_var: Character. Unit identifier variable
time_var: Character. Time variable
treatment_var: Character. Treatment indicator
treatment_time: Integer. Time period when treatment begins
covariates: Character vector. Time-varying covariates. Default: NULL
method: Character. Augsynth method: "augsynth", "synth", "ridge". Default: "augsynth"
inference: Logical. Perform inference using conformal method? Default: TRUE
n_lags: Integer. Number of outcome lags to include. Default: 0

Value

A list containing:

synth_estimates: Synthetic control estimates
unit_weights: Synthetic control unit weights
balance_table: Pre-treatment balance
inference_results: Inference results if requested
plot_data: Data for synthetic control plots

Examples

if (FALSE) { # \dontrun{
synth_results <- synthetic_control_method(
  data = panel_data,
  outcome_var = "employment_rate",
  unit_var = "region_id",
  time_var = "year",
  treatment_time = 2015,
  method = "augsynth"
)
} # }