interFE: Interactive Fixed Effects Models
In gsynth: Generalized Synthetic Control Method

Description Usage Arguments Details Value Author(s) References See Also Examples

View source: R/interFE.R

Estimating interactive fixed effect models.

1 2	interFE(formula = NULL, data, Y, X, index, r = 0, force = "none", se = TRUE, nboots = 500, seed = NULL, tol = 1e-3, normalize = FALSE)

`formula`	an object of class "formula": a symbolic description of the model to be fitted.
`data`	a data frame (must be with a dichotomous treatment but balanced is not required).
`Y`	outcome.
`X`	time-varying covariates.
`index`	a two-element string vector specifying the unit (group) and time indicators. Must be of length 2.
`r`	an integer specifying the number of factors.
`force`	a string indicating whether unit or time fixed effects will be imposed. Must be one of the following, "none", "unit", "time", or "two-way". The default is "unit".
`se`	a logical flag indicating whether uncertainty estimates will be produced via bootstrapping.
`nboots`	an integer specifying the number of bootstrap runs. Ignored if `se = FALSE`.
`seed`	an integer that sets the seed in random number generation. Ignored if `se = FALSE` and `r` is specified.
`tol`	a numeric value that specifies tolerate level.
`normalize`	a logic flag indicating whether to scale outcome and covariates. Useful for accelerating computing speed when magnitude of data is large.The default is `normalize=FALSE`.

interFE estimates interactive fixed effect models proposed by Bai (2009).

`beta`	estimated coefficients.
`mu`	estimated grand mean.
`factor`	estimated factors.
`lambda`	estimated factor loadings.
`VNT`	a diagonal matrix that consists of the r eigenvalues.
`niter`	the number of iteration before convergence.
`alpha`	estimated unit fixed effect (if `force` is "unit" or "two-way").
`xi`	estimated time fixed effect (if `force` is "time" or "two-way").
`residuals`	residuals of the estimated interactive fixed effect model.
`sigma2`	mean squared error of the residuals.
`IC`	the information criterion.
`ValidX`	a logical flag specifying whether there are valid covariates.
`dat.Y`	a matrix storing data of the outcome variable.
`dat.X`	an array storing data of the independent variables.
`Y`	name of the outcome variable.
`X`	name of the time-varying control variables.
`index`	name of the unit and time indicators.
`est.table`	a table of the estimation results.
`est.boot`	a matrix storing results from bootstraps.

Yiqing Xu <yiqingxu@stanfprd.edu>, Stanford University

Licheng Liu <liulch@mit.edu>, M.I.T.

Jushan Bai. 2009. "Panel Data Models with Interactive Fixed Effects." Econometrica 77:1229–1279.

print.interFE and gsynth

library(gsynth)
data(gsynth)
d <- simdata[-(1:150),] # remove the treated units
out <- interFE(Y ~ X1 + X2, data = d, index=c("id","time"),
               r = 2, force = "two-way", nboots = 50)

Bootstraping
Call:
interFE.formula(formula = Y ~ X1 + X2, data = d, index = c("id", 
    "time"), r = 2, force = "two-way", nboots = 50)

Estimated Coefficients:
        Coef    S.E. CI.lower CI.upper p.value
X1     1.021 0.02806   0.9685    1.082       0
X2     3.052 0.03918   2.9788    3.130       0
_const 5.005 0.14020   4.7962    5.260       0