"Best in Show: Rethinking the Rankings Game"
U.S. News produced its first national rankings in 1983, basing its results on the composite scores of an opinion survey of college and university presidents. Over the next two decades, the magazine both refined and made more complex its methodology, adding “quantifiable” variables and shifting the weights assigned to items, ostensibly to fine-tune its process of assessing institutional quality. The resulting methodology all but ensures that a system characterized by glacial change will seem volatile enough to have a shake-up in the ordering of institutions every year. That reshuffling has included wide swings—variations that ultimately raise the question, “What exactly is it that the U.S. News rankings are measuring?”
U.S. News’ own answer is simple and direct: the rankings identify America’s best colleges and universities by measuring quality. Researchers at The Learning Alliance of the University of Pennsylvania have an alternate answer, arguing that what the rankings actually measure are the shape and preferences of the higher education market. In this issue of The Landscape, their taxonomy segmenting the collegiate marketplace goes head-to-head with U.S. News rankings. The results indicate what the rankings game ultimately captures, as well as how institutions’ standings compare to their ordering in the market taxonomy.
"An Integral Part: Placing the Two-Year College in the Higher Education Market."
Because of their traditionally unique role within the higher education enterprise, two-year institutions have always been considered in some way distinct from their four-year counterparts. However, because of the magnitude of two-year institutions, as well as their increasingly important role in higher education, many have begun to rethink their importance in the market. This issue of The Landscape informs that thinking, reporting on an analysis by the Institute for Research on Higher Education that proves the two-year market is not so distinct from its four-year counterpart.