Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

From |
"Wooldridge, Jeffrey" <wooldri1@msu.edu> |

To |
<statalist@hsphsun2.harvard.edu> |

Subject |
st: RE: margins command after "control function" IV negative binomial |

Date |
Thu, 3 Mar 2011 06:51:01 -0500 |

Some quick comments about the CF control. You should only be obtaining one control function, from the regression reg Y1 X1 IV1 IV2 IV1_X1 IV2_X1, vce(cluster ID) predict y1_resid, resid If you believe the reduced form for Y1 is linear with an additive, independent error, then adding this to the model controls for endogeneity of any function of Y1 on the RHS of the Y2 equation. Of course, you might want to put the CF in the second model in a flexible way. One of the benefits of the CF approach is that it is a parsimonious way to allow for many nonlinear functions of Y1. I hope I do a better job of explaining this in 2e of my MIT Press book. See also my NBER and UCL/IFS lectures with Imbens. Some theorists will complain about your approach because if the underlying "structural" model for Y2 is NB then the estimable model with the CF (residual) added cannot be NB -- unless you make a restrictive distributional assumption. This does not bother me so much. You will have to compute the average marginal effects "by hand." This is not hard with an exponential function. But take the derivative with respect to Y1 to get (b1 + b2*X1)exp(.) And then average across all of your data. Use the bootstrap for a proper standard error. JW -----Original Message----- From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of rlhall@umich.edu Sent: Wednesday, March 02, 2011 5:52 PM To: statalist@hsphsun2.harvard.edu Subject: st: margins command after "control function" IV negative binomial I am estimating a negative binomial model with two endogenous regressors, the second of which is an interaction between the endogenous variable and an exogenous term. I am using what Wooldridge calls the control function approach. Cameron and Trivedi describe its implementation in Stata (2010, 607-609). The two endogenous variables, Y1, and Y1_X1 (Y1 interacted with an exogenous variable. Both are interval level variables.) The system is overidentified. I estimate the first stage equations. (IV=instrument; IV1_X1 = IV1*X1, IV2_X1- IV2*X1 ). reg Y1 X1 IV1 IV2 IV1_X1 IV2_X1, vce(cluster ID) predict y1_resid, resid reg Y1_X1 X1 IV1 IV2 IV1_X1 IV2_X1, vce(cluster ID) predict y1x1_resid, resid I then estimate the 2nd stage: nbreg Y2 X1 Y1 Y1_X1 y1_resid y1x1_resid, vce(cluster id) The problem comes in calculating the predicted counts at various levels of the key variables, e.g.,: margins, predict (n) atmeans at(X1=2 Y1=5 Y1_X1=10) This produces huge predicted counts, often several times the maximum predicted count for the model: predict nbreg_hat, n) Insofar as I can tell, the problem arises because I am setting the values of the endogenous variables but letting margins set the two residual terms at their means. (The endogenous variables and their respective first stage residuals are correlated at about .7). But if that?s the mistake, I don?t know at what values I should set the residual terms to get the correct predicted counts. (Setting the residual terms at the same values as the respective endogenous terms does not seem to produce sensible results either). Thanks in advance for any light you might shed on this problem... or guidance toward an altogether different approach. * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/ * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/

**References**:**st: margins command after "control function" IV negative binomial***From:*rlhall@umich.edu

- Prev by Date:
**Re: st: RE: Interpretation of regression outputs when variables are log transformed** - Next by Date:
**st: RE: Endogenous interaction terms in 2SLS** - Previous by thread:
**st: margins command after "control function" IV negative binomial** - Next by thread:
**st: margins command after "control function" IV negative binomial** - Index(es):