[River Huang]

Dear Robert, I see, and thanks.

[Ernesto Rodriguez]

Dear Colleagues,
I am estimating earnings with selection from a multinomial logit model with three alternatives in a first stage. In the second stage I regress earnings introducing IMR calculated in the first stage. It is quite close to Prof. Huang's post: https://www.statalist.org/forums/for...w-to-bootstrap
In my case, I am not working in a panel scenario. My case is a cross-section. I tried to replicate Prof. Samuel's code but it doesn't work:
My first quiestion is why shouldn't I just regress my second stage just as Prof. Huang did initially:
// second stage
reg ca ca0 imr1 if err == 1, robust
reg ca ca0 imr2 if err == 2, robust
reg ca ca0 imr3 if err == 3, robust

And second, in case any of you could have a look at my replication, I wounder if you could check if I am wrong and where...

Replication code:

   
 save d1, replace  
     
 cap program drop _all  
 scalar drop _all  
 macro drop _all  
     
 /* Convert CODE to PROGRAM  */  
 program define twostep, rclass  
     tempfile mlogit  
     // first stage  
     statsby _b, saving(`mlogit', replace): mlogit choice_mnl edad1 edad2 d_mujer d_educacion_compacta2 ///  
         d_educacion_compacta3 d_castellano d_departamentos41 d_departamentos42 d_departamentos43 d_ingNoLaboral d_vivpropia d_ayudaFliarGob ///  
         d_TrabajoInformal d_contratoIndefinido d_hogarPobre, baseoutcome(3)  
      
     merge m:1 using `mlogit'  
     sort id  
     
     gen del1 = _eq1_b_edad1*edad1 + _eq1_b_edad2*edad2 + _eq1_b_d_mujer*d_mujer + _eq1_b_d_educacion_compacta2*d_educacion_compacta2 + ///  
         _eq1_b_d_educacion_compacta3*d_educacion_compacta3 + _eq1_b_d_castellano*d_castellano + _eq1_b_d_departamentos41*d_departamentos41 + ///  
         _eq1_b_d_departamentos42*d_departamentos42 + _eq1_b_d_departamentos43*d_departamentos43 + _eq1_b_d_ingNoLaboral*d_ingNoLaboral + ///  
         _eq1_b_d_vivpropia*d_vivpropia + _eq1_b_d_ayudaFliarGob*d_ayudaFliarGob + _eq1_b_d_TrabajoInformal*d_TrabajoInformal + ///  
         _eq1_b_d_contratoIndefinido*d_contratoIndefinido + _eq1_b_d_hogarPobre*d_hogarPobre + _eq1_b_cons  
     gen del2 = _eq2_b_edad1*edad1 + _eq2_b_edad2*edad2 + _eq2_b_d_mujer*d_mujer + _eq2_b_d_educacion_compacta2*d_educacion_compacta2 + ///  
         _eq2_b_d_educacion_compacta3*d_educacion_compacta3 + _eq2_b_d_castellano*d_castellano + _eq2_b_d_departamentos41*d_departamentos41 + ///  
         _eq2_b_d_departamentos42*d_departamentos42 + _eq2_b_d_departamentos43*d_departamentos43 + _eq2_b_d_ingNoLaboral*d_ingNoLaboral + ///  
         _eq2_b_d_vivpropia*d_vivpropia + _eq2_b_d_ayudaFliarGob*d_ayudaFliarGob + _eq2_b_d_TrabajoInformal*d_TrabajoInformal + ///  
         _eq2_b_d_contratoIndefinido*d_contratoIndefinido + _eq2_b_d_hogarPobre*d_hogarPobre + _eq2_b_cons  
     
     gen F1 = exp(del1)/(1+exp(del2)+exp(del1))  
     gen F2 = exp(del2)/(1+exp(del2)+exp(del1))  
     gen F3 = 1/(1+exp(del1)+exp(del2))  
       
     gen J1 = -invnormal(F1)  
     gen J2 = -invnormal(F2)  
     gen J3 = -invnormal(F3)  
     
     gen imr1 = -normalden(J1)/F1  
     gen imr2 = -normalden(J2)/F2  
     gen imr3 = -normalden(J3)/F3  
     // second stage  
     tempfile t1  
     save `t1'  
     forvalues i = 1/3 {  
         use `t1', clear  
         regress ln_w edad1 imr`i' if choice_mnl==`i'  
         local xedad1 =  _b[edad1]  
         return scalar edad1_`i' = `xedad1'  
         local ximr =  _b[imr`i']  
         return scalar imr_`i' = `ximr'  
         local xcons =  _b[_cons]  
         return scalar cons_`i' = `xcons'  
         /* Now must drop variables created in the program  
            or generate will fail silently in replicates  
            because variables already exist */  
         }  
      drop _merge del2 del3 F1 F2 F3 J1 J2 J3 imr1 imr2 imr3  
 end  
     
 use d1, clear  
 /* Check Program twostep */  
 twostep  
 return list    
     
 /* Bootstrap program twostep */  
 bootstrap edad1_1 = r(edad1_1) imr1_1 = r(imr1_1)  ///  
           edad1_2 = r(edad1_2) imr1_2 = r(imr1_2)  ///  
           edad1_3 = r(edad1_3) imr1_3 = r(imr1_3), ///  
           nodrop reps(10): twostep  
     
 estat bootstrap, all  
 log close"

Many thanks indeed.
Kind regards,
Ernesto