A A A

I am wondering why do we use a selective sample, instead of the census, in policy analysis such as social protection?