I want to create a line graph that shows the ratio of means of numerical variable(wage) by a bicategorial variable(sex) for different regions over time. So for example in year x in region y the ratio (male wage/female wage).
collapse (mean) wage = p_wage, by(year sex region)
by year sex region, sort: gen wage_m= wage if sex==1
by year sex region, sort: gen wage_f = wage if sex==2
by year sex region, sort: gen wage_ratio= (wage_m/wage_f)
twoway line wage_ratio year if region==1 ///
|| line wage_ratio year if region==2 ///
etc.
The graph comes up empty. Apparently I cant create the variables in the ways needed so that stata understands. What am I doing wrong?
collapse (mean) wage = p_wage, by(year sex region)
by year sex region, sort: gen wage_m= wage if sex==1
by year sex region, sort: gen wage_f = wage if sex==2
by year sex region, sort: gen wage_ratio= (wage_m/wage_f)
twoway line wage_ratio year if region==1 ///
|| line wage_ratio year if region==2 ///
etc.
The graph comes up empty. Apparently I cant create the variables in the ways needed so that stata understands. What am I doing wrong?
Comment