In the last few years, Medicaid has attracted more than casual attention, one reflection of which is the fact that JHPPL has published five papers on the program in its last few issues. This paper, a sixth, takes a broader view of the program than is typically the case. After a critique of the five recent articles, I discuss several questions raised by them and reach the following conclusions: First, the states do not invest enough in producing program data suitable for policy analysis and research. One lesson: Better data and analysis can help the states to avoid expensive mistakes. Second, those policy analyses that have been offered fail to give sufficient attention to the political dimension of policy. That is one reason why policy choices produce unexpected effects. Third, since Medicaid is a relatively small player in the vast medical care market, incentives adopted by Medicaid officials throughout the country rarely have the desired effects. Finally, as long as Medicaid remains the principal mechanism to provide access to health care for the poor, it must be made as efficient and effective as possible. Yet, for both political and economic reasons, Medicaid can never be what its original planners had hoped, the vehicle for providing the poor with reliable access to mainstream medical care.