On the infimum of quantum effects