Gender Income Difference Test

[badman127]

Hi everybody,
I have the data of about 5000 observations about the income factors of male and female. I want to generate OLS regression for each gender's income (meaning that I will get 2 separate equations for male and female income respectively)
But I have trouble in comparing the gap between the income of male and female.
Any suggestion about the Test or model that I can use to find the income difference in gender?
Thanks a lot.

[startz]

Hi everybody,
I have the data of about 5000 observations about the income factors of male and female. I want to generate OLS regression for each gender's income (meaning that I will get 2 separate equations for male and female income respectively)
But I have trouble in comparing the gap between the income of male and female.
Any suggestion about the Test or model that I can use to find the income difference in gender?
Thanks a lot.

Suppose that income is Y, there is an explanatory variable X, and D is a dummy variable coded 1 for women. You can run the equation

Code: Select all

ls Y C X D D*X


The coefficient on D tells you the difference in the intercept and the coefficient on D*X tells you the difference in the effect of X on Y.

[badman127]

Hi everybody,
I have the data of about 5000 observations about the income factors of male and female. I want to generate OLS regression for each gender's income (meaning that I will get 2 separate equations for male and female income respectively)
But I have trouble in comparing the gap between the income of male and female.
Any suggestion about the Test or model that I can use to find the income difference in gender?
Thanks a lot.

Suppose that income is Y, there is an explanatory variable X, and D is a dummy variable coded 1 for women. You can run the equation

Code: Select all

ls Y C X D D*X


The coefficient on D tells you the difference in the intercept and the coefficient on D*X tells you the difference in the effect of X on Y.

Dear Startz,
Thanks for your help but the main problem here is about my data.
Here is the description of the factors:
earns Wife’s weekly earnings
educ Wife’s years of schooling
age Wife’s age
exper Age – educ – 6
union =1 if wife is in union
kidge6 =1 if child>=6 in family
kidlt6 =1 if child<6
hours Wife’s weekly hours
inlf =1 if wife is in labour force
nwifeinc Non-wife income, $1000s
faminc Annual family income
husage Husband’s age
husearns Husband’s weekly earnings
husunion =1 if husband is in union
huseduc Husband’s years of schooling
husexp Husage – Huseduc – 6
hushrs Husband’s weekly hours
With this data I need to estimate Earnings Function separately for male and female, and assess the gender differences. Therefore I think it's hard to generate the model as you mentioned above
Please help!

[startz]

The equation I gave you does estimate the equations separately. But you can also go

Code: Select all

smpl if D
ls y c x
smpl if not D
ls y c x

[badman127]

Please take a look at the data, each row represents a household with a husband and a wife, and I don't know how to deal with it to generate the dummy variables you're saying
The model you said is really good, but how to achieve it is too hard for me

Thanks for your help in advance!

[startz]

Run one regression with earns as the dependent variable and one with husearns as the dependent variable?

[badman127]

That's exactly what I have done, I got the estimation for those 2.
And the thing I tried to ask is how to assess the gender difference in income?
Sorry if I couldn't clarify it from the beginning.

[startz]

What do you mean by "assess the difference?" If your assignment is what I think it is, you might google "Blinder-Oaxaca Decomposition"