Log-Linear Model

In this example the simple regression first considered in chapter 4 is modified to include more variables and an interaction. The model adds experience to the model

ln(wage) = ві + e2educ + вз exper + e (5.19)

In this model suppose that the marginal effect of another year of schooling depends on how much experience the worker has. This requires adding an interaction

ln(wage) = в1 + e2educ + e3exper + e4(educ x exper) + e (5.20)

image157 Подпись: в3 + в4 educ Подпись: (5.21)

The marginal effect of another year of experience is

In percentage terms the marginal effect of another year of experience is 100(в3 + e4educ). The model can be estimated and the marginal effect computed easily with hansl

1 open "@gretldirdatapoecps4_small. gdt"

2 logs wage

3 series ed_exp=educ*exper

4 ols l_wage const educ exper ed_exp

5 scalar me8 = $coeff(exper)+8*$coeff(ed_exp)

6 scalar me16 = $coeff(exper)+16*$coeff(ed_exp)

7 printf "nThe marginal effect of exper for one

8 with 8 years of schooling is %.3f%%n",100*me8

9 printf "nThe marginal effect of exper for one io with 16 years of schooling is %.3f%%.n",100*me16

The result is

The marginal effect of exper for one with 8 years of schooling is 0.604%. The marginal effect of exper for one with 16 years of schooling is 0.575%.

Leave a reply

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>