I am trying to run
and
to analyse a potential wage gap between men and women.
The idea is based on Chernozhukov, V., I. Fernández-Val, and B. Melly. 2013. Inference on counterfactual distributions. Econometrica 81(6): 2205-2268.
The results of the distribution process are displayed as I expected them to.
When calling
afterwards, Stata should plot the results into convenient graphs.
However, I do get the pop up of the plots, but they are all empty (see image below).
Education is split into 3 levels: low, medium and high. Child = whether having children that live at home. Country of origin = 1 if from the Netherlands, marital = 1 if living together or married. Gender = 1 if female.
The exact code I use is
And I end up with the following plots.
Can somebody explain what I am doing incorrectly?
I also do not know how to include the differences between men and women.
First, I had the idea to run
However, this result in the same blank plots. The example I have seen has the following graphs.
Which are based on the quantile regression, so they will differ from mine. However, I hope this gives an idea of what I am trying to achieve.
After running the distribution regression, I need to establish counterfactuals to compare men's wages having male characteristics with men's wages when having female characteristics.
That is what the second figure of the example shows.
Below are 50 of the 2658 observations of my data set.
Code:
drprocess
Code:
counterfactual
The idea is based on Chernozhukov, V., I. Fernández-Val, and B. Melly. 2013. Inference on counterfactual distributions. Econometrica 81(6): 2205-2268.
The results of the distribution process are displayed as I expected them to.
When calling
Code:
plotprocess
However, I do get the pop up of the plots, but they are all empty (see image below).
Education is split into 3 levels: low, medium and high. Child = whether having children that live at home. Country of origin = 1 if from the Netherlands, marital = 1 if living together or married. Gender = 1 if female.
The exact code I use is
Code:
drprocess log_income age marital origin low med high child plotprocess
Can somebody explain what I am doing incorrectly?
I also do not know how to include the differences between men and women.
First, I had the idea to run
Code:
drprocess log_income age marital origin low med high child if gender==0 drprocess log_income age marital origin low med high child if gender==1
Which are based on the quantile regression, so they will differ from mine. However, I hope this gives an idea of what I am trying to achieve.
After running the distribution regression, I need to establish counterfactuals to compare men's wages having male characteristics with men's wages when having female characteristics.
That is what the second figure of the example shows.
Below are 50 of the 2658 observations of my data set.
Code:
* Example generated by -dataex-. For more info, type help dataex clear input float(log_income gender) byte age float(marital origin low med high child) 7.313221 0 50 0 1 0 0 1 0 7.600903 0 60 1 1 1 0 0 0 6.993933 0 31 0 1 0 1 0 0 7.313221 1 38 0 1 0 1 0 0 7.408531 1 51 0 1 0 0 1 0 8.34284 0 55 1 1 1 0 0 1 7.901007 0 38 1 1 1 0 0 0 7.783224 0 41 0 1 0 0 1 0 7.740664 1 44 1 1 0 0 1 1 7.549609 1 40 0 1 1 0 0 0 8.411833 1 46 0 1 1 0 0 1 7.718686 1 42 1 1 0 0 1 0 7.718686 0 54 1 1 0 1 0 0 7.130899 1 51 1 1 1 0 0 0 7.600903 1 60 0 1 0 1 0 0 8.070906 0 45 0 1 0 0 1 0 7.824046 0 46 1 1 0 1 0 1 7.058758 1 46 0 0 1 0 0 1 7.438384 0 56 0 0 1 0 0 0 7.696213 0 60 1 1 1 0 0 0 7.495542 1 27 1 1 1 0 0 0 8.268732 0 44 1 1 0 1 0 1 6.745236 1 40 1 1 0 1 0 1 8.006368 0 43 1 1 0 1 0 1 7.244227 1 43 1 1 1 0 0 1 8.032685 0 56 1 1 1 0 0 0 7.299798 1 32 0 1 0 1 0 0 7.617268 0 32 1 1 1 0 0 1 8.022897 0 36 1 1 0 0 1 1 7.972466 0 36 1 1 0 1 0 1 7.495542 1 32 1 1 0 0 1 1 8.101678 0 37 1 1 0 1 0 1 7.824046 0 50 1 1 0 1 0 1 7.740664 0 44 1 1 0 1 0 1 7.783224 0 38 1 1 0 1 0 1 7.649693 0 32 1 1 1 0 0 1 6.813445 1 44 0 1 1 0 0 0 7.003066 0 24 1 1 1 0 0 0 7.776115 0 31 1 1 1 0 0 1 7.600903 0 28 1 1 1 0 0 1 end
