opsr_2step function

Heckman Two-Step Estimation

This is a utility function, used in opsr and should not be used directly. Tow-step estimation procedure to generate reasonable starting values.

opsr_2step(W, Xs, Z, Ys)

Arguments

• W: matrix with explanatory variables for selection process.
• Xs: list of matrices with expalanatory varialbes for outcome process for each regime.
• Z: vector with ordinal outcomes (in integer increasing fashion).
• Ys: list of vectors with continuous outcomes for each regime.

Returns

Named vector with starting values passed to opsr.fit.

Details

These estimates can be retrieved by specifying .get2step = TRUE in opsr.

Remark

Since the Heckman two-step estimator includes an estimate in the second step regression, the resulting OLS standard errors and heteroskedasticity-robust standard errors are incorrect \insertCite Greene:2002OPSR.

References

\insertRef Greene:2002OPSR

See Also

opsr.fit, opsr_prepare_coefs