Estimating the Efficiency of Four-year Public Master\u27s Universities in Arkansas Using Data Envelopment Analysis

For the fourth time in the past 20 years, the state of Arkansas has implemented a performance-funding program. Although directly referenced in the title of Act 1203 of 2011, the implemented performance-funding program does not include a measure of how efficiently institutions are converting their inputs into outputs. The purpose of this study was two-fold: (1) to examine, through the conceptual framework of agency theory, why the state of Arkansas adopted performance funding policy after experiencing three adopt-implement-abandon cycles in the past 20 years, and (2) to estimate the efficiency of four-year public Master\u27s universities in Arkansas in achieving one of the mandatory measures in the performance-funding program: number of bachelor\u27s degrees awarded. The institutions included in this study were Arkansas State University, Arkansas Tech University, and University of Central Arkansas in the larger group, and Henderson State University, Southern Arkansas University, and University of Arkansas at Monticello in the medium/smaller group. The analytical technique utilized in this study was based on Simar and Wilson\u27s (2007) double bootstrap truncated regression approach. This study found that the six Arkansas institutions were inefficient in the production of bachelor\u27s degrees during the 2011-2012 academic year, given the inputs included in the model and in relationship to the national sample of Master\u27s Colleges and Universities. Through the second-stage regression analysis, this study also found that a state\u27s per capita real gross domestic product (GDP) was a positive and statistically significant determinant of estimated efficiency for both groups of institutions (larger and medium/smaller). For the medium/smaller group, higher education\u27s share of total state expenditures was a negative and statistically significant determinant of estimated efficiency, while the share of operating revenues based on net tuition was a positive and statistically significant determinant of estimated efficiency for the larger group institutions. The existence of a performance-funding program was found to not be a significant determinant of institutional efficiency through the second-stage regression analysis. Using Eisenhardt\u27s (1989) taxonomy of behavior- versus outcome-based contracts, this study also found that agency theory served as a viable conceptualization of why the state of Arkansas chose to adopt performance-funding policy for the fourth time since its initial adoption in 1995