Fixed effects with three time periods with panel data

[Ivan Ivanov]

I am doing fixed effects with 3 time periods with panel data. Can I keep my data in long format, add 2 dummy variables d1, d2 for two years and then add fe at the end?
For example: xtreg Y1 X1...Xt d1 d2, fe. The second option would be to take the differences between periods, leaving actually two time periods. Or something third?
Anyways, thanks

[Jeff Wooldridge]

This is typically preferred. Probably dummies for period two and three. And use vce(cluster id).

[George Ford]

or reghdfe and absorb both fixed effects, unless you are interested in the time coefficients.

[Ivan Ivanov]

Thanks to both of you guys.

[Carlo Lazzaro]

Ivan:
why creating N-1 dummies for -i.year- when you can have the T of your panel dataset in one variable only if you adopt the highly recommended -long- format?

use https://www.stata-press.com/data/r19/nlswork.dta
. xtreg ln_wage i.year c.age##c.age if year<=70, fe vce(cluster idcode)

Fixed-effects (within) regression               Number of obs     =      4,284
Group variable: idcode                          Number of groups  =      2,230

R-squared:                                      Obs per group:
     Within  = 0.0986                                         min =          1
     Between = 0.0935                                         avg =        1.9
     Overall = 0.0943                                         max =          3

                                                F(4, 2229)        =      42.46
corr(u_i, Xb) = 0.0558                          Prob > F          =     0.0000

                             (Std. err. adjusted for 2,230 clusters in idcode)
------------------------------------------------------------------------------
             |               Robust
     ln_wage | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        year |
         69  |   .0642213   .0256044     2.51   0.012     .0140104    .1144322
         70  |   .0543866   .0518297     1.05   0.294     -.047253    .1560261
             |
         age |   .1968173   .0584201     3.37   0.001     .0822539    .3113808
             |
 c.age#c.age |  -.0036788    .001181    -3.12   0.002    -.0059948   -.0013629
             |
       _cons |  -1.055726    .778866    -1.36   0.175    -2.583105    .4716529
-------------+----------------------------------------------------------------
     sigma_u |  .36753152
     sigma_e |   .2050529
         rho |   .7626172   (fraction of variance due to u_i)
------------------------------------------------------------------------------

.

[Ivan Ivanov]

Grazie mille Carlo. I did my undergraduate studies at Bocconi. Now writing my masters thesis at another university.

[Carlo Lazzaro]

Grazie a Te, Ivan:
we share the same alma mater and we both move to another university for different purposes.
Perfect Italian, by the way.
Best of luck for what lies ahead.