dummy variable

[sarah rzzm]

I want to enter COVID-19 as a dummy variable that takes the number 1 from (2020-2022) and 0 in the rest of the years from 2013-2019, as I have 32 countries and ten years . my dependent variable is food security X index and my independent variables: GDP, unemployment, and the exchange rate. But when I do FMOLS, I find that the effect of Covid and unemployment are positive and GDP IS NEGATIVE , which is illogical.
I think I'm not good at using dummy variables, should I introduce other dummy variables?
What SHOULD I do?
I ENTER THE LOG AND I USE THE COMMAND/ cointreg lx lgdp ler lunme dcovid

[Nick Cox]

The part I understand here would seem to imply using the variable created by

gen covid = inrange(year, 2020, 2022)

or in your case

gen covid = year > 2019

in your model fitting.