opsr_2step

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

Estimates ordered probit switching regression models - a Heckman
type selection model with an ordinal selection and continuous outcomes.
Different model specifications are allowed for each treatment/regime. For
more details on the method, see Wang & Mokhtarian (2024) <doi:10.1016/j.tra.2024.104072>
or Chiburis & Lokshin (2007) <doi:10.1177/1536867X0700700202>.

Daniel Heimgartner

OPSR

Ordered Probit Switching Regression

Xinyi Wang

opsr_2step function

<dl><dt>W</dt>
<dd>matrix with explanatory variables for selection process.</dd>
<dt>Xs</dt>
<dd>list of matrices with expalanatory varialbes for outcome process for each regime.</dd>
<dt>Z</dt>
<dd>vector with ordinal outcomes (in integer increasing fashion).</dd>
<dt>Ys</dt>
<dd>list of vectors with continuous outcomes for each regime.</dd></dl>

Arguments

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 Greene:2002OPSR.

Remark

Heckman Two-Step Estimation — opsr_2step

<dl>

<dt>W</dt>
<dd>matrix with explanatory variables for selection process.</dd>


<dt>Xs</dt>
<dd>list of matrices with expalanatory varialbes for outcome process for each regime.</dd>


<dt>Z</dt>
<dd>vector with ordinal outcomes (in integer increasing fashion).</dd>


<dt>Ys</dt>
<dd>list of vectors with continuous outcomes for each regime.</dd>

</dl>

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 Greene:2002OPSR.

opsr_2step: Heckman Two-Step Estimation

Description

Usage

Value

Arguments

Remark

Details

References

See Also