4.

Sample selection Model : Heckman

Model

Empirical Example
 4. Sample selection Model : Heckman Model (Empirical Example)

 Some initial statistics: Relating income and expenditure of beer

 Full Sample versus sample of consumers

25 20
expenditure in beer
10 5
015

0 2000 4000 6000 8000 10000

Renta Mensual pc

all
buyers

61
 4. Sample selection Model : Heckman Model (Empirical Example)

 The outcome variable for the discrete choice model: to buy or not to buy
decision
.
. *Dummy que identifique el hogar en el que se consume cerveza
. gen byte consumidor=(gtot_cer>0 & gtot_cer<.)

. *Modelo 1: decisión de consumo de cerveza

. *Dummys que identifica al hogar que consume cerveza frente al que no consume
. tab consumidor, missing //1 hogar que consume cerveza - 0 hogar que no consume

consumidor Freq. Percent Cum.

0 10,817 57.15 57.15

1 8,112 42.85 100.00

Total 18,929 100.00

. sum consumidor

Variable Obs Mean Std. dev. Min Max

consumidor 18,929 .4285488 .4948815 0 1

. heckman gtot_cer $x, select(consumidor=$z) nolog twostep
∗
Heckman selection model -- two-step estimates Number of obs = 18,929

(regression model with sample selection) Selected = 8,112
Nonselected = 10,817

Wald chi2(10) = 417.65

Prob > chi2 = 0.0000

Coefficient Std. err. z P>|z| [95% conf. interval]

The Heckman model views this sample
gtot_cer
selection process as a form of omitted
renta
rentac
20.73475
-6.812333
1.717961
.7557411
12.07
-9.01
0.000
0.000
17.36761
-8.293558
24.10189
-5.331107 variable bias.
hombresp 2.760026 .4456736 6.19 0.000 1.886521 3.63353
adult 2.678986 .5132346 5.22 0.000 1.673065 3.684908
edadsp
edadsp2
.7575324
-.007226
.1153621
.0010042
6.57
-7.20
0.000
0.000
.5314267
-.0091942
.9836381
-.0052578
 𝒊+
diaria0 -4.037678 1.028675 -3.93 0.000 -6.053844 -2.021511
diaria1 -1.42442 .8189 -1.74 0.082 -3.029435 .1805943
𝛗 𝒁𝒊 𝜸
diaria2
diaria3
.8537074
-.6342487
.6944884
.7540386
1.23
-0.84
0.219
0.400
-.5074648
-2.112137
2.21488
.8436398  𝒊 : selection term
_cons -23.20353 3.269747 -7.10 0.000 -29.61212 -16.79495 𝚽 𝒁𝒊 𝜸

Should I estimate both

consumidor
renta .5857992 .0820532 7.14 0.000 .4249779 .7466204
rentac -.1898944 .0368576 -5.15 0.000 -.262134 -.1176549
hombresp
adult
.0941907
.0854871
.0216825
.027841
4.34
3.07
0.000
0.002
.0516938
.0309197
.1366876
.1400545
equations/decisions jointly?
menores .0265691 .0291151 0.91 0.361 -.0304955 .0836337
diaria0
diaria1
-.4509204
-.3098566
.0508238
.0441941
-8.87
-7.01
0.000
0.000
-.5505332
-.3964755
-.3513075
-.2232376 Rho=0.53 and it is statistically
diaria2
diaria3
-.1064416
-.0300077
.0375657
.0407837
-2.83
-0.74
0.005
0.462
-.1800691
-.1099423
-.0328141
.0499269 significant…….then it means that both
edadsp
edadsp2
.026171
-.0002815
.005488
.0000473
4.77
-5.95
0.000
0.000
.0154146
-.0003742
.0369273
-.0001888 equations should not be estimated
separately in order to get consistent
caprov -.1564166 .0333308 -4.69 0.000 -.2217439 -.0910893

tamamu
2 -.175105 .0366985 -4.77 0.000 -.2470328 -.1031772 estimates
3 -.1681606 .0398352 -4.22 0.000 -.2462362 -.090085

Coefficient estimates are NOT the same

4 -.1482927 .0433588 -3.42 0.001 -.2332743 -.0633111
5 -.1905596 .0374727 -5.09 0.000 -.2640046 -.1171145

cons_vino 1.250868 .0206027 60.71 0.000 1.210487 1.291248 whether we estimate the equation jointly
_cons -1.296076 .1634957 -7.93 0.000 -1.616522 -.9756303
using the heckman selection model or
/mills
lambda 10.26995 .5403055 19.01 0.000 9.210974 11.32893 separately using MCO and logit.
rho 0.53755
sigma 19.105058

63
. heckman gtot_cer $x, select(consumidor=$z) nolog twostep 4. Sample selection Model : Heckman
Heckman selection model -- two-step estimates Number of obs = 18,929 Model : The case of the beer....
(regression model with sample selection) Selected = 8,112
Nonselected = 10,817

Wald chi2(10)
Prob > chi2
=
=
417.65
0.0000
 +
𝛗 𝒁𝒊 𝜸
Coefficient Std. err. z P>|z| [95% conf. interval]
 𝒊 : selection term
gtot_cer
𝚽 𝒁𝒊 𝜸
renta 20.73475 1.717961 12.07 0.000 17.36761 24.10189
rentac
hombresp
-6.812333
2.760026
.7557411
.4456736
-9.01
6.19
0.000
0.000
-8.293558
1.886521
-5.331107
3.63353 An example of analysis for covariates:
adult 2.678986 .5132346 5.22 0.000 1.673065 3.684908
edadsp
edadsp2
.7575324
-.007226
.1153621
.0010042
6.57
-7.20
0.000
0.000
.5314267
-.0091942
.9836381
-.0052578
Income is present in {X} and {Z}
diaria0 -4.037678 1.028675 -3.93 0.000 -6.053844 -2.021511
diaria1
diaria2
-1.42442
.8537074
.8189
.6944884
-1.74
1.23
0.082
0.219
-3.029435
-.5074648
.1805943
2.21488
Influences the decision of whether to buy
diaria3
_cons
-.6342487
-23.20353
.7540386
3.269747
-0.84
-7.10
0.400
0.000
-2.112137
-29.61212
.8436398
-16.79495
beer and the decision of the amount of
consumidor
beer....
renta .5857992 .0820532 7.14 0.000 .4249779 .7466204
rentac
hombresp
-.1898944
.0941907
.0368576
.0216825
-5.15
4.34
0.000
0.000
-.262134
.0516938
-.1176549
.1366876
We can interpret the sign of the effect:
adult .0854871 .027841 3.07 0.002 .0309197 .1400545
menores .0265691 .0291151 0.91 0.361 -.0304955 .0836337  In both cases is positive and statistically
diaria0 -.4509204 .0508238 -8.87 0.000 -.5505332 -.3513075
diaria1 -.3098566 .0441941 -7.01 0.000 -.3964755 -.2232376 significant, hence, our model states that as the
diaria2 -.1064416 .0375657 -2.83 0.005 -.1800691 -.0328141
diaria3 -.0300077 .0407837 -0.74 0.462 -.1099423 .0499269 family income growths, the probability of
edadsp
edadsp2
.026171
-.0002815
.005488
.0000473
4.77
-5.95
0.000
0.000
.0154146
-.0003742
.0369273
-.0001888 consuming beer at home increases as well as
caprov -.1564166 .0333308 -4.69 0.000 -.2217439 -.0910893
the amount of beer.
tamamu
2 -.175105 .0366985 -4.77 0.000 -.2470328 -.1031772  ¿Can we say the same idea for age of the
3 -.1681606 .0398352 -4.22 0.000 -.2462362 -.090085
4 -.1482927 .0433588 -3.42 0.001 -.2332743 -.0633111 reference person?
5 -.1905596 .0374727 -5.09 0.000 -.2640046 -.1171145

cons_vino 1.250868 .0206027 60.71 0.000 1.210487 1.291248

_cons -1.296076 .1634957 -7.93 0.000 -1.616522 -.9756303

/mills
lambda 10.26995 .5403055 19.01 0.000 9.210974 11.32893

rho 0.53755
sigma 19.105058

64
. heckman gtot_cer $x, select(consumidor=$z) nolog twostep

Heckman selection model -- two-step estimates Number of obs = 18,929

(regression model with sample selection) Selected = 8,112
Nonselected = 10,817  Un ejemplo mas completo, lo
Wald chi2(10)
Prob > chi2
=
=
417.65
0.0000 visualizamos en Excel
. heckman gtot_cer $x, select(consumidor=$z) nolog twostep first

Iteration 0: log likelihood = -12926.644

 Al incluir las CCAA se

Iteration 1: log likelihood = -10406.044
Coefficient Std. err. z P>|z| [95% conf. interval] Iteration 2: log likelihood = -10377.315
Iteration 3: log likelihood = -10377.289

mantienen las ideas principales.

gtot_cer
renta 20.73475 1.717961 12.07 0.000 17.36761 24.10189 Probit regression Number of obs = 18,929
rentac -6.812333 .7557411 -9.01 0.000 -8.293558 -5.331107 LR chi2(17) = 5098.71
Prob > chi2 = 0.0000
hombresp 2.760026 .4456736 6.19 0.000 1.886521 3.63353
Log likelihood = -10377.289 Pseudo R2 = 0.1972
adult 2.678986 .5132346 5.22 0.000 1.673065 3.684908
edadsp .7575324 .1153621 6.57 0.000 .5314267 .9836381
edadsp2 -.007226 .0010042 -7.20 0.000 -.0091942 -.0052578 consumidor Coefficient Std. err. z P>|z| [95% conf. interval]
diaria0 -4.037678 1.028675 -3.93 0.000 -6.053844 -2.021511
diaria1 -1.42442 .8189 -1.74 0.082 -3.029435 .1805943 renta .5857992 .0820532 7.14 0.000 .4249779 .7466204
rentac -.1898944 .0368576 -5.15 0.000 -.262134 -.1176549
diaria2 .8537074 .6944884 1.23 0.219 -.5074648 2.21488
hombresp .0941907 .0216825 4.34 0.000 .0516938 .1366876
diaria3 -.6342487 .7540386 -0.84 0.400 -2.112137 .8436398 adult .0854871 .027841 3.07 0.002 .0309197 .1400545
_cons -23.20353 3.269747 -7.10 0.000 -29.61212 -16.79495 menores .0265691 .0291151 0.91 0.361 -.0304955 .0836337
diaria0 -.4509204 .0508238 -8.87 0.000 -.5505332 -.3513075
consumidor diaria1 -.3098566 .0441941 -7.01 0.000 -.3964755 -.2232376
renta .5857992 .0820532 7.14 0.000 .4249779 .7466204 diaria2 -.1064416 .0375657 -2.83 0.005 -.1800691 -.0328141
diaria3 -.0300077 .0407837 -0.74 0.462 -.1099423 .0499269
rentac -.1898944 .0368576 -5.15 0.000 -.262134 -.1176549
edadsp .026171 .005488 4.77 0.000 .0154146 .0369273
hombresp .0941907 .0216825 4.34 0.000 .0516938 .1366876 edadsp2 -.0002815 .0000473 -5.95 0.000 -.0003742 -.0001888
adult .0854871 .027841 3.07 0.002 .0309197 .1400545 caprov -.1564166 .0333308 -4.69 0.000 -.2217439 -.0910893
menores .0265691 .0291151 0.91 0.361 -.0304955 .0836337
diaria0 -.4509204 .0508238 -8.87 0.000 -.5505332 -.3513075 tamamu
diaria1 -.3098566 .0441941 -7.01 0.000 -.3964755 -.2232376 2 -.175105 .0366985 -4.77 0.000 -.2470328 -.1031772
diaria2 -.1064416 .0375657 -2.83 0.005 -.1800691 -.0328141 3 -.1681606 .0398352 -4.22 0.000 -.2462362 -.090085
4 -.1482927 .0433588 -3.42 0.001 -.2332743 -.0633111
diaria3 -.0300077 .0407837 -0.74 0.462 -.1099423 .0499269 5 -.1905596 .0374727 -5.09 0.000 -.2640046 -.1171145
edadsp .026171 .005488 4.77 0.000 .0154146 .0369273
edadsp2 -.0002815 .0000473 -5.95 0.000 -.0003742 -.0001888 cons_vino 1.250868 .0206027 60.71 0.000 1.210487 1.291248
caprov -.1564166 .0333308 -4.69 0.000 -.2217439 -.0910893 _cons -1.296076 .1634957 -7.93 0.000 -1.616522 -.9756303

tamamu
2 -.175105 .0366985 -4.77 0.000 -.2470328 -.1031772
3 -.1681606 .0398352 -4.22 0.000 -.2462362 -.090085
4 -.1482927 .0433588 -3.42 0.001 -.2332743 -.0633111
5 -.1905596 .0374727 -5.09 0.000 -.2640046 -.1171145

cons_vino 1.250868 .0206027 60.71 0.000 1.210487 1.291248

_cons -1.296076 .1634957 -7.93 0.000 -1.616522 -.9756303

/mills
lambda 10.26995 .5403055 19.01 0.000 9.210974 11.32893

rho 0.53755
sigma 19.105058

65
 4. Sample selection Model : Heckman Model : The case of the beer....

 Postestimation análisis: Predictions

∗
 . predict gcer_cond, ycond

∗ . sum gtot_cer gcer_cond


Variable Obs Mean Std. dev. Min Max
∗
Prob gtot_cer 18,929 6.888089 14.31399 0 99.42825
gcer_cond 18,929 17.72152 4.96884 -.3451334 30.34419

¿Bad performnace of the model?

This “ycond” predictions are available for all observations in the dataset, but they
∗ ∗
are intended to compute : or

66
 4. Sample selection Model : Heckman Model : The case of the beer....

 Postestimation analysis: predictions

∗

∗

∗
Prob
Not so bad if you compare the right prediction with the observed consumption
The first conditional consumption predictions are available for all observations in the
dataset
For the subsample of consumers: We see that the average predictions from heckman are
close to the observed levels (they do have the same mean..)
. predict gcer_cond, ycond

. sum gtot_cer gcer_cond

Variable Obs Mean Std. dev. Min Max

gtot_cer 18,929 6.888089 14.31399 0 99.42825

gcer_cond 18,929 17.72152 4.96884 -.3451334 30.34419

.
. sum gtot_cer gcer_cond if gtot_cer>0

Variable Obs Mean Std. dev. Min Max

gtot_cer 8,112 16.07306 18.17931 .0021723 99.42825

gcer_cond 8,112 16.07306 5.059395 -.3451334 29.20117

67
 4. Sample selection Model : Heckman Model : The case of the beer....

 Postestimation analysis: predictions

 𝐸 Q |𝑋, 𝑦 = 1

 𝐸 Q |𝑋
 Prob Q∗ |𝑋, 𝑦 = 1
What if we were interested in making predictions about mean consumption for all
households? Here the expected expenditure is 0 for those who are not expected to buy
good Q, with expected participation (buy=yes) determined by the selection equation.
 These values can be obtained with the yexpected option of predict (“y” taken to be 0 where
unobserved)
. predict gcer_exp, yexpected

. sum gtot_cer gcer_exp

Variable Obs Mean Std. dev. Min Max

gtot_cer 18,929 6.888089 14.31399 0 99.42825

gcer_exp 18,929 6.885344 3.526196 -.1273255 18.18161

Again we note that the predictions from heckman are close to the observed mean
expenditure level for beer for all households.
 Why might be that predictions using ycond and yexpected are not equal to their observed sample
equivalents? For the Heckman model, unlike linear regression, the sample moments implied by the
optimal solution to the model likelihood do not require that these predictions match observed data.
Properly accounting for the additional variation from the selection equation requires that the model
use more information than just the sample moments of the observed expenditure 68
 4. Sample selection Model : Heckman Model : The case of the beer....

 Postestimation analysis: predictions

 𝐸 Q |𝑋, 𝑦 = 1

 𝐸 Q |𝑋
 Prob Q∗ |𝑋, 𝑦 = 1
Predict with the option “psel” calculates the probability of selection (or being observed).

. predict gtot_pr1, psel

. sum gtot_pr1 consumidor

Variable Obs Mean Std. dev. Min Max

gtot_pr1 18,929 .4279757 .2500402 .0443376 .8830402

consumidor 18,929 .4285488 .4948815 0 1

69
The issue of marginal effects
 4. Sample selection Model : Heckman Model : The case of the beer....

 What are the marginal effects in the Heckman model?

∗


 + , and , j=1,…,N {all individuals}

 To correctly evaluate marginal effects we have to take into account

1. We have different “types” of regressors:
Those that only appear in the “buying” decision/equation (exclusion restrictions)
Those that appear in both equations
Those that appears only in the demand equation (i.e prices)
2. Whether we want to measure marginal effects on:
Marginal effects on the observed {unconditional}
 Q = 𝛼 + γp + X 𝛽 + θλ +ε
∗
Marginal effects on the “latent” variable

 𝐸 Q∗ |𝑋, 𝑦 = 1 = 𝛼 + X 𝛽 + 𝜌𝜎

71
 4. Sample selection Model : Heckman Model : The case of the beer....

∗
 Marginal effects on the “latent” variable

 + , and , j=1,…,N {all individuals}

The total marginal effect of the independent variables {Z (some X are also in Z)} on
in the observed sample may consists of two components.
First, there is the direct effect of the independent variable on the mean of which is
captured by .
Second, there is an indirect effect if the independent variable also appears in the selection
equation. This is because a change in some X not only changes the the mean of , but also
the probability that an observation is actually in the sample i.e. it will affect through

[ | ]
 ,

The main point is that if  and the independent variable appears in the selection and
outcome equation, then does NOT indicate the marginal effect of x on . It is quite
possible for the magnitude, sign, and statistical significance of the marginal effect to all be
different from the estimate of . This point is often ignored. Thus, it is not sufficient to
simply estimate the model and look at t-statistics to know if an independent variable (that
appears in the selection and outcome equation) has an effect on .

72
 4. Sample selection Model : Heckman Model : The case of the beer....

 What are the marginal effects in the Heckman model?

∗


λ = , j=1,…,N {all individuals}

Z≠𝑋
[R] heckman postestimation Postestimation tools for heckman
(View complete PDF manual entry)

Postestimation commands

The following postestimation commands are available after heckman:

Command Description

contrast contrasts and ANOVA-style joint tests of estimates

* estat ic Akaike's and Schwarz's Bayesian information criteria (AIC and BIC)
estat summarize summary statistics for the estimation sample
estat vce variance-covariance matrix of the estimators (VCE)
estat (svy) postestimation statistics for survey data
estimates cataloging estimation results
+ hausman Hausman's specification test
lincom point estimates, standard errors, testing, and inference for linear combinations of coefficients
+ lrtest likelihood-ratio test; not available with two-step estimator
margins marginal means, predictive margins, marginal effects, and average marginal effects
marginsplot graph the results from margins (profile plots, interaction plots, etc.)
nlcom point estimates, standard errors, testing, and inference for nonlinear combinations of coefficients
predict predictions, residuals, influence statistics, and other diagnostic measures
predictnl point estimates, standard errors, testing, and inference for generalized predictions
pwcompare pairwise comparisons of estimates
* suest seemingly unrelated estimation
test Wald tests of simple and composite linear hypotheses
testnl Wald tests of nonlinear hypotheses

73
 4. Sample selection Model : Heckman Model : The case of the beer....

 What are the marginal effects in the Heckman model?

heckman demand educ age, select(consumer=married children educ age)
 predict ycond, ycond /* Predicted Q conditional on it being observed*/
 Probability of wage being observed
 predict probseen, psel /*Probability of consumption being observed*/

Syntax for margins

margins [marginlist] [, options]

margins [marginlist] , predict(statistic ...) [predict(statistic ...) ...] [options]

statistic Description

xb linear prediction; the default

xbsel linear prediction for selection equation
pr(a,b) Pr(y a < y < b)
e(a,b) E(y a < y < b)
ystar(a,b) E(y*), y* = max{a,min(y,b)}
* ycond E(y y observed)
* yexpected E(y*), y taken to be 0 where unobserved
nshazard or mills nonselection hazard (also called inverse of Mills's ratio)
psel Pr(y observed)
stdp not allowed with margins
stdf not allowed with margins
stdpsel not allowed with margins

* ycond and yexpected are not allowed with margins after heckman, twostep.

74
 4. Sample selection Model : Heckman Model : The case of the beer....

 What are the marginal effects in the Heckman model?

heckman demand educ age, select(consumer=married children educ age)
 predict ycond, ycond /* Predicted Q conditional on it being observed*/
 Probability of wage being observed
 predict probseen, psel /*Probability of consumption being observed*/

Syntax for margins

margins [marginlist] [, options]

margins [marginlist] , predict(statistic ...) [predict(statistic ...) ...] [options]

statistic Description
The partial effects on the ycond are the partial effects on
xb linear prediction; the default
xbsel linear prediction for selection equation the truncated mean, i.e. only for those who actually have
pr(a,b)
e(a,b)
Pr(y
E(y
a < y < b)
a < y < b)
an observed value. 𝐸 Q∗ |𝑋, Q > 0
ystar(a,b) E(y*), y* = max{a,min(y,b)}
* ycond E(y y observed)
* yexpected E(y*), y taken to be 0 where unobserved
nshazard or mills nonselection hazard (also called inverse of Mills's ratio)
psel Pr(y observed)
stdp
stdf
not allowed with margins
not allowed with margins
The partial effects on the yexpected value, are the partial
stdpsel not allowed with margins effects on the censored mean of the dependent variable.
* ycond and yexpected are not allowed with margins after heckman, 𝐸 Q∗ |𝑋, 𝑄 > 0 *Prob Q∗ > 0
twostep.
The censored mean is supposed to be equal to the
probability of being observed times the truncated
mean.

75
 4. Sample selection Model : Heckman Model : The case of the beer....

 Some remarks
margins, dydx(*) after heckmand estimation
will always get you the same coefficients as you have for the value equation when
using heckman because since you don't specify any predict() options, it assumes that you
want the marginal effects on the linear prediction of the value equation, i.e. the xb option is
the default.
Nothing wrong there, only that if that's not what you want to estimate you should mention
that by setting an appropriate predict() option.
See heckman postestimation to know what options are available and which one
matches what you want.

76
. heckman gtot_cer renta c.renta#c.renta edadsp c.edadsp#c.edadsp $x1, select(consumidor=renta c.renta#c.renta edadsp c.edadsp#c.edadsp $z1) nolog twostep

Heckman selection model -- two-step estimates Number of obs = 18,929

(regression model with sample selection) Selected = 8,112
Nonselected = 10,817

Wald chi2(10)
Prob > chi2
=
=
417.65
0.0000
 Fist step
Coefficient Std. err. z P>|z| [95% conf. interval]

gtot_cer
renta 20.73475 1.717961 12.07 0.000 17.36761 24.10189

c.renta#c.renta -6.812333 .7557411 -9.01 0.000 -8.293558 -5.331107

edadsp .7575324 .1153621 6.57 0.000 .5314267 .9836381

c.edadsp#c.edadsp -.007226 .0010042 -7.20 0.000 -.0091942 -.0052578

. margins, dydx(renta edad hombresp adult)
hombresp 2.760026 .4456736 6.19 0.000 1.886521 3.63353
adult 2.678986 .5132346 5.22 0.000 1.673065 3.684908 Average marginal effects Number of obs = 18,929
diaria0 -4.037678 1.028675 -3.93 0.000 -6.053844 -2.021511 Model VCE: Conventional
diaria1 -1.42442 .8189 -1.74 0.082 -3.029435 .1805943
diaria2 .8537074 .6944884 1.23 0.219 -.5074648 2.21488 Expression: Linear prediction, predict()
diaria3 -.6342487 .7540386 -0.84 0.400 -2.112137 .8436397 dy/dx wrt: renta edadsp hombresp adult
_cons -23.20353 3.269747 -7.10 0.000 -29.61212 -16.79495

consumidor Delta-method
renta .5857992 .0820532 7.14 0.000 .4249779 .7466204 dy/dx std. err. z P>|z| [95% conf. interval]

c.renta#c.renta -.1898944 .0368576 -5.15 0.000 -.262134 -.1176549 renta 10.04027 .6753112 14.87 0.000 8.716683 11.36385
edadsp -.0558587 .0161448 -3.46 0.001 -.0875018 -.0242155
edadsp .026171 .005488 4.77 0.000 .0154146 .0369273 hombresp 2.760026 .4456736 6.19 0.000 1.886521 3.63353
adult 2.678986 .5132346 5.22 0.000 1.673065 3.684908
c.edadsp#c.edadsp -.0002815 .0000473 -5.95 0.000 -.0003742 -.0001888

hombresp .0941907 .0216825 4.34 0.000 .0516938 .1366876

adult .0854871 .027841 3.07 0.002 .0309197 .1400545
menores .0265691 .0291151 0.91 0.361 -.0304955 .0836337
diaria0 -.4509204 .0508238 -8.87 0.000 -.5505332 -.3513075
diaria1 -.3098566 .0441941 -7.01 0.000 -.3964755 -.2232376
diaria2 -.1064416 .0375657 -2.83 0.005 -.1800691 -.0328141
diaria3 -.0300077 .0407837 -0.74 0.462 -.1099423 .0499269
caprov -.1564166 .0333308 -4.69 0.000 -.2217439 -.0910893

tamamu
2 -.175105 .0366985 -4.77 0.000 -.2470328 -.1031772
3 -.1681606 .0398352 -4.22 0.000 -.2462362 -.090085
4 -.1482927 .0433588 -3.42 0.001 -.2332743 -.0633111
5 -.1905596 .0374727 -5.09 0.000 -.2640046 -.1171145

cons_vino 1.250868 .0206027 60.71 0.000 1.210487 1.291248

_cons -1.296076 .1634957 -7.93 0.000 -1.616522 -.9756303

/mills
lambda 10.26995 .5403055 19.01 0.000 9.210974 11.32893

rho 0.53755
sigma 19.105058
77
 Fist step

. qui heckman gtot_cer renta c.renta#c.renta edadsp c.edadsp#c.edadsp $x1, select(consumidor=renta c.renta#c.renta edadsp c.edadsp#c.edadsp $z

. *margins, dydx(*)
. margins, dydx(renta edad)

Average marginal effects Number of obs = 18,929

Model VCE: Conventional

Expression: Linear prediction, predict()

dy/dx wrt: renta edadsp

Delta-method
dy/dx std. err. z P>|z| [95% conf. interval]

renta 5.84647 12.54 0.47 0.641 -18.73147 30.42441

edadsp .0155905 .2446368 0.06 0.949 -.4638888 .4950699

78
 1. Corner Solutions: Heckman Model

[ | ] ( )
 Marginal effects: ,

An example for Income {this variable can also be observed for Q=0}
( )
 β <marginal effect in this case since >0

. margins, dydx(renta_pc) predict(ycond) atmean

Conditional marginal effects Number of obs = 11,344

Model VCE : OIM

Expression : E(qcerv_mes|Zg>0), predict(ycond)

dy/dx w.r.t. : renta_pc
at : precio = 1.841017 (mean)
renta_pc = .7871813 (mean)
parado = .3485543 (mean)
cvino = .6964915 (mean)
pvino = 2.810937 (mean)
edadsp = 55.06964 (mean)
hombresp = .7074224 (mean)
estudios1 = .1287024 (mean)
estudios2 = .2857898 (mean)
estudios3 = .1950811 (mean)
diaria2 = .3398272 (mean)
diaria3 = .5128702 (mean)

Delta-method
dy/dx Std. Err. z P>|z| [95% Conf. Interval]

renta_pc 1.352254 .222692 6.07 0.000 .9157857 1.788722

79
 1. Corner Solutions: Heckman Model

[ | ] ( )
 Marginal effects: ,

An example for Income {this variable can also be observed for Q=0}
( )
 β <marginal effect in this case since >0

. margins, dydx(renta_pc) predict(yexpected) atmean

Conditional marginal effects Number of obs = 11,344

Model VCE : OIM

Expression : E(qcerv_mes*|Pr(cantpos)), predict(yexpected)

dy/dx w.r.t. : renta_pc
at : precio = 1.841017 (mean)
renta_pc = .7871813 (mean)
parado = .3485543 (mean)
cvino = .6964915 (mean)
pvino = 2.810937 (mean)
edadsp = 55.06964 (mean)
hombresp = .7074224 (mean)
estudios1 = .1287024 (mean)
estudios2 = .2857898 (mean)
estudios3 = .1950811 (mean)
diaria2 = .3398272 (mean)
diaria3 = .5128702 (mean)

Delta-method
dy/dx Std. Err. z P>|z| [95% Conf. Interval]

renta_pc .4691453 .04939 9.50 0.000 .3723425 .565948

80