Esttab r-squared help

[Aleksandr Polusyan]

I created four regressions with the eststo command to put them all in the table with esttab. I used the following code ( esttab, se r2 label) for my specifications, however, the r2 appears blank, how can I fix this?

I used xtreg to create my regressions if that helps.

Also, I have these year variables that show when running esttab in the format (year=2005, year=2006, etc.) They appear as a result of me including time fixed effects in the regression (i.year). I tried applying the drop command in esttab, but this error returns
=exp not allowed
r(101);

[Andrew Musau]

estout is from SSC (FAQ Advice #12). xtreg outputs three R2 statistics; you probably want the within R2 with fixed effects models. You can use the -indicate- option to handle the year dummies. The following example illustrates:

webuse grunfeld, clear
xtset company year
xtreg invest mvalue kstock i.year, fe
esttab ., drop(_cons) stats(N r2_w, labels(Observations Within-R2)) indicate("Year effects=*.year") 

Res.:

. esttab ., drop(_cons) stats(N r2_w, labels(Observations Within-R2)) indicate("Year effects=*.year")

----------------------------
                      (1)  
                   invest  
----------------------------
mvalue              0.118***
                   (8.56)  

kstock              0.358***
                  (15.75)  

Year effects          Yes  
----------------------------
Observations          200  
Within-R2           0.799  
----------------------------
t statistics in parentheses
* p<0.05, ** p<0.01, *** p<0.001

[Aleksandr Polusyan]

Thank you for your reply. I have one more question though. When I run the xtreg command separately and then use ereturn list, all of the r2 variables are higher than if I use esttab for multiple xtreg regressions. For example, r2_w is 0.74 for an individual regression, but is 0.55 for that same regression when ran with multiple regressions using esttab. Thank you.

[Andrew Musau]

When I run the xtreg command separately and then use ereturn list, all of the r2 variables are higher than if I use esttab for multiple xtreg regressions. For example, r2_w is 0.74 for an individual regression, but is 0.55 for that same regression when ran with multiple regressions using esttab. Thank you.

I have never experienced this. You should present a reproducible example that illustrates this.