Research performance is a critical component of the understanding of university prestige, yet measuring it is complex. This paper proposes using "real dollars of research expenditures per article published" as a measure of research productivity because it is more accurate than other quantitative measures of research productivity, including "articles published per capita," for certain applications, especially financial and economic analyses. In tracking this measure, a decline in the research productivity of US colleges and universities was observed during the period 1996 to 2002.