Problems when generating dummy

[Angela Cojoc]

Hello everybody,
I am new to Stata, but I need to use it for my bachelor thesis. I have recently come across a problem with my code in which I am trying to create a dummy value if one country is part of the eurozone in order to see the effects of a monetary union on trade. I am using the CEPII gravity database plus the years in which the eurozone countries have adopted the euro. Both the time variables are correct yet when I try to do the following code
"gen EMU_one_sided = ((eurozone_join_o <= year & eurozone_join_d > year) | (eurozone_join_d <= year & eurozone_join_o > year))" I only get 0s.

I have also tried to do it using this:

"gen eurozone_o = eurozone_join_o <= year
gen eurozone_d = eurozone_join_d <= year
gen EMU_one_sided = (eurozone_o != eurozone_d)"

Does anyone knows what I am doing wrong or what I can do in order to get a normal variation since there have to be countries that are in the eurozone trading with non-members.

(I have replaced all the missing values for the eurozone_join_o and eurozone_join_d variables with 2042 for countries not part of the eurozone)

[George Ford]

With,

gen EMU_one_sided = ((eurozone_join_o <= year & eurozone_join_d > year) | (eurozone_join_d <= year & eurozone_join_o > year))

the second part is exactly the opposite of the first, and you've used "or".

how are these variables defined? eurozone_join_o , eurozone_join_d

[Angela Cojoc]

So eurozone_join_o is the year when the origin country adopted the euro and eurozone_join_d is. the year for the destination country. Both are int
And yes I used or because I have already done the analysis for both countries being a part of the eurozone at the same time. So i am only interestedin either the origin or destination country being in the eurozone

Eurozone_join_o and eurozone_join_d are the same it was just easier to duplicate it

[George Ford]

gen eurozone_o = year  >= eurozone_join_o  // (1 if orig)
gen eurozone_d = year > = eurozone_join_d  // (1 if dest)
g eurozone_both = eurozone_o + eurozone_d  == 2 // (1 if both)
g eurozone_one_side = eurozone_o + eurozone_d == 1 // 1 if only one is eurozone
g eurozone_niether = eurozone_o + eurozone_d == 0  // 1 if neither in eurozone
g eurozone_o_not_d = eurozone_o==1 & eurozone_d == 0  
g eurozone_d_not_o = eurozone_o==0 & eurozone_d == 1

[Angela Cojoc]

the eurozone_one_side is still just 0s

[George Ford]

tab2 eurozone_o eurozone_d

[Angela Cojoc]

does this look right?

[George Ford]

It's telling you why one-sided is always zero.

If one-sided is not always zero (in reality), then there's a problem with the variables used to construct eurozone_o/d.

[Angela Cojoc]

yes at first I also thought there was something wrong with the eurozone_join variables but they look fine when i tabulate them

[George Ford]

switch 2042 to -999. it's catching the 2042 as specified.

[Angela Cojoc]

It worked now! Thank you very much!!