Dear statalist users,

I am running a regression in which I have 2 fixed effects: year and 2-digit SIC code. I made these fixed effects as follows:

sum fyear
local a=r(min)
local b=r(max)
forvalues i=`a'(1)`b'{
qui gen yeardum_`i'=0
qui replace yeardum_`i'=1 if fyear==`i'
}

sum SIC2
local a=r(min)
local b=r(max)
forvalues i=`a'(1)`b'{
qui gen sicdum_`i'=0
qui replace sicdum_`i'=1 if SIC2==`i'
}

Now, when I run the regression I get a note that my yeardum_2014 and yeardum_2019 have been omitted because of collinearity. What does this mean? Does this mean that all observations from 2014 and 2019 have been excluded or only the fixed effects for these years?

Hope someone knows.

Kind regards,
Sarah