This is a very basic question but after doing some research I could not find the answer to this. I keep hearing that FE dummies in OLS regression or FE estimator models give the same results. In terms of coefficients of independent variables it does. However, when it comes to the actual fixed effects of the entity in question I am not clear.

If I run OLS regression like this and my entity of interest is the hospital:
Code:
reg clinical_measure $controls ibn.Hospital_ID, base vce(robust) noconst
I get the following hospital coefficients which I am assuming to be the same as hospital fixed effects:
Code:
                         141  |   1.608901
                                     143  |   1.684427
                                     144  |   1.501123
                                     145  |   1.426459
                                     146  |   1.658545 
                                     147  |   1.808902  
                                     148  |   1.644542 
                                     149  |   1.726485 
                                    1410  |   1.756854   
                                    1411  |   1.597826   
                                    1412  |   1.874025
                                    1413  |   1.788815  
                                    1414  |   1.586877   
                                    1415  |   1.733997 
                                    1418  |   1.350842   
                                    1419  |   2.342925  
                                    1420  |   1.570274 
                                    1431  |    1.42267
                                    1432  |    2.28881
                                   14103  |   1.836362
                                   14105  |    1.52336
                                   14110  |   1.723062
                                   14115  |    1.75736 
                                   14117  |   2.039847
                                   14118  |    1.72009 
                                   14139  |   1.090784
If I run FE estimator:
Code:
xtset Hospital_ID
xtreg clinical_measure $controls, fe base vce(robust)
and then run the command below to get hospital-fixed effects:
Code:
predict individual_effect_fe, u
I get completely different hospital-fixed effect values for the same hospitals:
Code:
u[Centre_ID ] 
------------+-----
-.2092153
-.1345512
-.026773
.0487535
-.0378475
.0088683
.0228709
.0908111
.1211798
.1732282
.238351
-.2848315
-.2130032
-.0654002
-.0487969
.0983237
.1531409
.6531358
.7072516
-.1123142
.0873884
.1216858
.2006881
.4041728
Would anyone be able to clarify? Many thanks in advance.