Dual clustering for xtlogit

[Chris Warmington]

Dear all,
I am trying to run an xtlogit random effects model on panel data (89 countries, 45 years) with errors dually clustered on id and year. I have tried using vcemway, in the model
vcemway logit crisisbanking creditbonanzal1 meancreditbonanza var55 var57 var70, cluster(id year)
however it returns the error message "option cluster() not allowed". I have also tried generating a double cluster variable (egen dbl_clstr = group(id year)) and running
xtlogit crisisbanking creditbonanzal1 meancreditbonanza var55 var57 var70, vce(dbl_clst)
however this also doesn't work, returning the error message "panels are not nested within clusters".

How can I dually cluster my errors sucessfully?

For reference, my model is vcemway logit crisisbanking creditbonanzal1 meancreditbonanza var55 var57 var70, cluster(id year)
My data looks

id

year

creditbonanzal1

meancreditbonanza

var55

var57

var70

Thanks in advance,
Chris

[Chris Warmington]

Forgot to include the (dependent) crisis dummy variable in the data table and to mention that creditbonanzal1 is also a dummy, with meancreditbonanza its country mean.