Research methods

Why does surveying every member of a small, clearly listed population eliminate ordinary sampling error?