Cross sectional data with year fixed effects

[Jattioui Bilal]

Hello, my model looks like this
SLL = a + B*GR + i.country + i.year + i.ID + Ei
the dependent variable is a dummy variable that takes 1 if a bank issued an SLL, otherwise 0, the problem is I'm dealing with a period of 3 years from 2020 to 2023 and i want to include Year fixed effects, how i can add it to my data table and interpret it in stata

Thank you

ID	SLL 2020	SLL 2021	SLL 2022
1	0	1	1
2	1	1	1
3	0	0	1
4	1	0	1

[Carlo Lazzaro]

Jattioui:
welcome to this forum.
I would -reshape- in -long- format my data, so to have an unique -SLL_new_ variable.
Then, If you have repeated cross-sectional data, you may want to consider -logit- -i.SLL_new-.

[Jattioui Bilal]

Thank you so much sir for your attention,

so i have to start the regression one SLL variable for one year (2020), then i repeat the regression for the other years ? could i get one table where i get consistent results over the years ? or it going to be unique results for each yer ?

thanks

[Carlo Lazzaro]

Jattioui:
I meant something along the following lines:

. set obs 4
Number of observations (_N) was 0, now 4.

. g SLL_2020=0

. g SLL_2021=0

. g SLL_2022=0

. replace SLL_2020 = 2020 in 2


. replace SLL_2020 = 2020 in 4


. replace SLL_2021 = 2021 in 1


. replace SLL_2021 = 2021 in 2


. replace SLL_2022=2022 if SLL_2022==0

.  stack SLL_* , into( SLL_new) clear

. drop if SLL_new==0

. bysort SLL_new: g id=_n

. drop _stack

. list

     +--------------+
     | SLL_new   id |
     |--------------|
  1. |    2020    1 |
  2. |    2020    2 |
  3. |    2021    1 |
  4. |    2021    2 |
  5. |    2022    1 |
     |--------------|
  6. |    2022    2 |
  7. |    2022    3 |
  8. |    2022    4 |
     +--------------+

. g y=runiform()*1000

. regress y i.SLL_new

      Source |       SS           df       MS      Number of obs   =         8
-------------+----------------------------------   F(2, 5)         =      1.00
       Model |  137548.527         2  68774.2633   Prob > F        =    0.4327
    Residual |  345509.669         5  69101.9339   R-squared       =    0.2847
-------------+----------------------------------   Adj R-squared   =   -0.0014
       Total |  483058.196         7  69008.3137   Root MSE        =    262.87

------------------------------------------------------------------------------
           y | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
     SLL_new |
       2021  |  -225.2771   262.8725    -0.86   0.431    -901.0123    450.4581
       2022  |    95.6866   227.6542     0.42   0.692    -489.5172    680.8904
             |
       _cons |   307.8787   185.8789     1.66   0.159    -169.9382    785.6956
------------------------------------------------------------------------------

.

[Jattioui Bilal]

Thanks you so much sir,

i think i solved it using excel formulas, just one question , i have a country variable containing ISO 2 letters code, should i uncode it to numerical variable ?

best regards

[Carlo Lazzaro]

Jattioui:
yes, you should, as in the following toy-example:

 use "C:\Program Files\Stata18\ado\base\a\auto.dta"
(1978 automobile data)

. encode make, gen(num_make)

. list make num_make in 1/10

     +-------------------------------+
     | make                 num_make |
     |-------------------------------|
  1. | AMC Concord       AMC Concord |
  2. | AMC Pacer           AMC Pacer |
  3. | AMC Spirit         AMC Spirit |
  4. | Buick Century   Buick Century |
  5. | Buick Electra   Buick Electra |
     |-------------------------------|
  6. | Buick LeSabre   Buick LeSabre |
  7. | Buick Opel         Buick Opel |
  8. | Buick Regal       Buick Regal |
  9. | Buick Riviera   Buick Riviera |
 10. | Buick Skylark   Buick Skylark |
     +-------------------------------+

.

Caveat emptor: take a look at the -encode- entry, Stata .pdf manual and focus your attention on the reported instances when -encode- behaves unexpectedly.