Reproducing an output including all indicators of a categorial variable and a constant

[Torben Lieber]

Dear Stata-Listers,

I am currently trying to reproduce the output in the picture.
The model used in this paper is a SAR-model investigating an efficiency indicator with a value between 0 and 1 of German districts.
The paper is: Felder, S., & Tauchmann, H. (2013). Federal state differentials in the efficiency of health production in Germany: an artifact of spatial dependence?. The European Journal of Health Economics, 14(1), 21-39.

The problem I have with this output is that it includes all german states (13 states + 3 city-states) and has no reference variable, which would normally be the case when including a categorical variable and a constant ( by typing i.state for example).
Additionally, I can't really get my head around the "coefficients expressed as mean-deviations"-note.
The authors did not mention any helpful notes in their text either.

An idea might be that the authors calculated the state-mean and Germany-wide-mean and included the differences in the table.
However, the table explicitly states that these are "coefficients", which wouldn't be fitting to my hypothesis...

Does this output make sense to any of you?
And does someone have an idea of how to include all indicators without dropping the constant?


I would appreciate the help!

With kind regards,
Torben

[daniel klein]

There are many ways of looking at a set of indicator variables; the difference to one reference category is just one of them. I would guess that "mean-deviations" express differences in the expected values for each federal state from the (overall) mean expected value. For example, we would expect Hesse's efficiency score to lie 0.04 below the average of all federal states.

In Stata, see

help contrast

especially the g. operator to get such estimates.

Best
Daniel

[Torben Lieber]

Genius, thanks a lot!

reg y i.state
contrast g.state, pv

Works great for me!