solve_masc: solving for matching and synthetic control weights

In maxkllgg/masc: Matching and Synthetic Controls Estimator

Description

Solves separately for synthetic control weights and matching weights for a given matching estimator, using data up to the treatment period.

Usage

solve_masc(
  treated,
  donors,
  treated.covariates,
  donors.covariates,
  treatment,
  sc_est,
  match_est,
  tune.pars,
  ...
)

Arguments

donors	A TxN matrix of outcome paths for untreated units, each column being a control unit.
sc_est	A function which constructs weights associated with a synthetic control-type estimator. See sc_estimator for input and output if you'd prefer to substitute your own estimator.
treated:	A Tx1 matrix of outcomes for the treated unit.
treatment:	An integer. The period T' in which forecasting begins (either the true treatment period or the first period after a cross-validation fold).
tune_pars	A list containing one element: m: an integer identifying the candidate matching (nearest neighbor) estimator.

Value

A list containing the weights associated with the two estimators. Weights are ordered in the same manner as the columns in donors. The weights.sc and weights.match named components contain the vector of synthetic control and nearest neighbor weights respectively. The objval.sc component contains the objective value (pre-period fit) of the synthetic control.

References

Kellogg, M., M. Mogstad, G. Pouliot, and A. Torgovitsky. Combining Matching and Synthetic Control to Trade off Biases from Extrapolation and Interpolation. Working Paper, 2021

See Also

Other masc functions: cv_masc(), masc_by_phi(), masc(), sc_estimator()

maxkllgg/masc documentation built on Sept. 7, 2021, 8:44 a.m.