Despite an attempt by its own authors, it is difficult to argue that the influential model of the size of government developed by Meltzer and Richard (1981) has had convincing empirical backing. In this paper, we adapt that model to a model of state government size. The main testable hypothesis is that as income inequality grows, government size (as measured by the percentage of income devoted to government redistribution) grows. We test the model using panel data from the US states from 1979-1991. In contrast to the results found by Meltzer and Richard (1983), we find little evidence to support the model. The results are robust to several model specifications and estimation techniques.