GIVEN THE DECENTRALIZED ORGANIZAtion of higher education in the United States, the notion that there is a national system of graduate education will perhaps seem counterfactual. In the humanities, for example, there exists no governing body overseeing the conduct of graduate education above the level of individual departments and university administrations.' Nonetheless I would like to propose that graduate education does exhibit certain systemic features on a national scale. The most signal of these features is the status hierarchy of graduate schools, currently ratified by the several agencies who regularly rank graduate programs. This hierarchy roughly corresponds to the perceived distinction between elite and nonelite universities, and more loosely to the levels of the Carnegie classification, with its multiple categories of research and nonresearch institutions. Tertiary education in the United States comprehends a vast and complex hierarchy, extending from Yale University to Bob Jones, from the University of California, Berkeley, to the University of Phoenix. The hierarchical arrangement of these institutions constitutes an informal system of higher education at both the undergraduate and graduate levels. If higher education is organized primarily by the fact of hierarchy or status, this principle is obviously vulnerable to distortions of perception. Surveying the relation between status and the American university system, economists Robert H. Frank and Philip J. Cook cite an opinion survey in which many respondents asked to name the top law schools in the country list Princeton among them (149). Princeton, of course, does not have a law school. But the perception, as our pundits like to say, is the reality. Many consequences flow from the perception of prestige, including the distribution of research funds and the generation of private or philanthropic support. But the hierarchy of status is more than a collective illusion. Prestige is also the result of successful competition among institutions for tangible resources. Because the syst m of higher education is not organize on the basis of an imposed division of functions or the shared allocation of social resources, individual institutions are free to compete with one another, which is to say that institutions are not free not to compete. The decentralized state of higher education is the condition of its highly competitive nature and determines the preeminence of hierarchy as its principle of organization. The reality of the status hierarchy has to be seen in relation to another reality, with which it appears to come into conflict: the ongoing democratization of access to higher education. At the turn of the nineteenth century, only four percent of the population attended college. Currently more than half of college-age youth in the United States attend some sort of postsecondary school. Over the last several decades the rate of growth in the population of graduate students has in fact surpassed that of undergraduate education. And yet the democratization of access by no means ensures that degrees can be converted into commensurate careers. In the humanities, as we are now very aware, the process of democratization often encounters an insuperable barrier at that stage when new PhDs seek to enter the ranks of the professoriat. The disparity between the number of PhDs and the number of jobs, what is now loosely called the job crisis, is in turn the motor of intense competition among schools, and among candidates themselves. In these circumstances, the status hierarchy of graduate schools has assumed an increased importance as a mechanism of selection. Students