Some statistical and DEA evaluations of relative efficiencies of public and private institutions of higher learning

This paper uses the Charnes, Cooper and Rhodes (CCR) ratio form of Data Envelopment Analysis (DEA) (1) to examine how DEA can be utilized in analyzing different aspects of production behavior of institutions of higher learning (IHLs) as an alternative to more traditional approaches such as econometric-regression models, and (2) to compare “specifically” relative efficiencies of public and private doctoral-granting universities in the U.S. and to analyze technical and scale efficiencies of those universities. The separation of doctoral-granting universities into universities with and without medical colleges represents a departure from preceding studies. This division proved very important in uncovering substantial differences in behavior between the two groups even when using the “statistical averaging” approaches that were customary in previous studies. For both groups, public universities proved more efficient than private universities when managerial and program inefficiencies are present in the data. When managerial inefficiencies are disentangled from the data and medical schools are not present, private universities have more efficient programs. However, their managers are less efficient users of program opportunities than are managers in public universities. Another portion of the current study dealt with returns-to-scale possibilities. These differed markedly (even on average) between IHLs with and without medical schools. Moreover, analyses by DEA showed marked ranges of variation for returns-to-scale possibilities for individual IHLs within each group. These possibilities have generally been concealed by the statistical averaging utilized in previous econometric studies. Further, their identification is beyond the ability of the usual types of one-at-a-time ratio and trend analyses.