Abstract
This paper discusses the measure of biases of apportionment methods. Generally biases mean that large states are more favored than small states or the opposite. However it is not easy to define the large states or small states rigorously. In this paper a new bias measure is developed which does not require the rigorous definition of large or small states. We show by computer simulation that the new bias measure can measure bias in apportionment methods as correctly as the existing measures.
