heckman estimation output

[ann panditharatne]

Hi all,
Im using the heckman method to impute wages for homemakers in my sample. I have 416 employed women and 422 homemakers in my sample.

Is there a way to get stata to output the wage and selection model results both in stata?I get the final results but not the first two sets.This is the code I use.

global lsmodel="shareexbymale ln_pc_hhexpense ln_pc_hhexpense_sqrd "
global dem1=" age age_square"
global education="second_education olevel advancedlevel degree"
global metgoals="goals_homemaker goals_both"
global perceptions=" whtwifesuddo_cadinal relwzchldrn_cadinal wrktorolemodel_cadinal "
global HHcharac="debt ownfamhouse tot_child_lessthan5 tot_child_5n11 tot_child_12n16 relwzcare motherliv_wzu hvdomestics adultwomentotalwzoutself shreofwhite"
global spatail="rural"
global overallhappy= "overallhappy"
global ln_wage= "ln_wage"

heckman ln_wage ///
$dem1 ///
$education ///
$spatail ///
if empld==1, select(age age_square second_education olevel advancedlevel degree rural)
eststo table1h
estout table1h, cells(b(star fmt(%9.4f)) se(par fmt(%9.4f))) stats(r2 N, fmt(%9.0f %9.0g)) ///
margin legend starlevels(* 0.10 ** 0.05 *** 0.01) prefoot("") postfoot("") ///
varwidth(16) modelwidth(12) delimiter("") r

esttab table1h using heckwage.csv, b(4) se(3) margin legend starlevels (* 0.10 ** 0.05 *** 0.01) prefoot("") postfoot("") ///
varwidth(16) modelwidth(12) delimiter("") r wide

predict heckwage1

Thanks
ASP

[Phil Bromiley]

You'll increase your chances of a useful answer by following the FAQ on asking questions - provide Stata code in code delimiters, readable Stat output, and sample data using dataex (or illustrate the problem using a Stata supplied data set). You make everything complicated by using a ton of macro, piles of options, and irrelevant code (like margins) You can illustrate the problem with 3 variables (and increase your chances of a response). Instead of the first 14 lines, heckman ln_wage age ,select(olevel) would suffice.

There are a bunch of output routines - outreg2 etc. They all operate a little differently. I'm pretty sure outreg2 will give the first stage. I notice in your Heckman, you don't ask for the first stage results. Try asking for them. Alternatively, you can do the probit yourself (check if gives the same results as the first stage of the heckman).