An Efficiency Evaluation Model for Academic Faculties of a Leading University by Data Envelopment Analysis

Data envelopment analysis (DEA) is a linear programming based data analytic method for measuring the relative efficiency of organisational units where the presence of multiple inputs and outputs makes comparisons difficult. Academic departments have critical importance for a university so we agree to research and compare about academic faculties in a leading university in Turkey. The aim of the study is to measure the relative efficiency of the academic faculties and determine the efficient/inefficient ones in the studied university. 12 faculties of the university are investigated within the scope of this study. The input variables are considered as total number of academic staff, total number of non-academic staff, number of students and the output variables are as number of publications, number of projects and the percentage of budget used. While results of CCR model show an average of 90.5% relative efficiency value, five faculties are found 100% efficient according to the CCR model. According to BCC model, the results have an average of 93.7% and 6 faculties are 100% efficient. In terms of the potential improvements evaluated for each inefficient faculty, it is seen that faculty of mechanical engineering and faculty of civil engineering are the ones requiring the most improvement. This paper contributes to the literature a lot however it is a new and proper study on efficiency analysis of faculties of a Turkish university. On conclusion of the DEA efficiency scores, the existence of misallocation of resources or/and inefficient applications to the faculties’ academic development are uncovered. Keywords: data envelopment analysis, efficiency, academic faculties 1. Introduction With increasing number of students enrolling in Turkish universities, leading universities has faced with the problem of providing higher education in a more effective manner that enables existing resources to be used to meet increasing demand M. Gül – M. Yücesan – O Duman 9/3 (2017) 60-71 İşletme Araştırmaları Dergisi Journal of Business Research-­‐Türk 61 for education. As the most advanced city of Turkey, Istanbul has the most leading public universities. Increasing competition and university management’ desire to reach a better place in overall ranking by utilizing scarce resources indicate that efficiency evaluation may become more common among these universities. Each university has a certain number of faculties. The evaluation of the efficiency of these academic faculties is part of the process of resource allocation within a university (Lopes and Lanzer, 2002). Academic faculties compete and cooperate with others within a university to demonstrate their capabilities to the stakeholders inside and outside the university. This prompts university management to use a permanent process of cross-evaluation of departments within the university. DEA as a data oriented approach is frequently applied by researchers for evaluating the efficiency of a set of decisions making units (DMUs) which convert multiple inputs into multiple outputs. Especially in the recent literature, several studies are carried out in academic environments to measure performance and efficiency using DEA method. Each study is distinguished from its scope, DMUs, and input/output variables. A brief explanation of these related studies is provided in the following. It is noted that these studies can deal with efficiencies of universities, academic sub units within universities and their environments. The focus of this study is related with the faculties as sub units of the universities. Avkiran (2001) focuses on the evaluation of the relative efficiency of Australian universities using DEA. They propose three performance models as overall performance, performance on delivery of educational services, and performance on feepaying enrolments. They conclude based on 1995 data that the universities perform well on technical and scale efficiency but there is room for improving performance on feepaying enrolments. Lopes and Lanzer (2002) deal with the issue of performance evaluation of fifty-eight academic departments at a Brazilian university using DEA. The results of DEA in the dimensions of teaching, research, service and quality are modelled under fuzzy environment and then a single index of performance for each department is generated. Tauer et al. (2007) study for technical and allocative efficiencies of academic departments in the College of Agriculture and Life Sciences at Cornell University using DEA. They use various specifications of outputs and inputs to determine sensitivity of results to specification. It is concluded that allocations of faculty time between teaching, research, and extension vary by department and are used as unique prices in calculating allocative efficiencies. Kao and Hung (2008) apply DEA to assess the relative efficiency of the academic departments at a university in Taiwan. They consider outputs as total credit-hours, publications, and external grants; and the inputs as personnel, operating expenses, and floor space. Tzeremes and Halkos (2010) apply bootstrapped DEA in order to determine the performance levels of 16 departments of a public owned university in Greece. They conclude that there are strong inefficiencies among the departments, indicating misallocation of resources or/and inefficient application of departments policy developments. Agha et al. (2011) study the evaluation of the relative technical efficiencies of academic departments at the Islamic University in Gaza during the years 2004-2006 using DEA. They use operating expenses, credit hours and training resources as inputs and number of graduates, promotions and public service activities as outputs variables. Results of their study show that the average efficiency score is 68.5% and that there are 10 efficient departments out of the 30 studied. Al-Shayea and Battal (2013) investigate the efficiency of eighteen faculties in a university in Saudi Arabia for M. Gül – M. Yücesan – O Duman 9/3 (2017) 60-71 İşletme Araştırmaları Dergisi Journal of Business Research-­‐Türk 62 the academic year 2011-2012 using DEA. They use the number of students enrolled, the number of teachers and staff as inputs, and the total number of students with a bachelor's degree and a number of research as outputs. The results show that 55.5% are efficient with average of 0.88 in terms of variable return to scale efficiency. The university obtains average scale efficiency 0.68 and only three faculties reach at the frontier. The aim of the paper is to estimate and analyse the efficiency of faculties of a leading university for the year 2014 using DEA. Although there are numerous studies focused on the efficiency of universities, university departments and so on in different countries around the world using various parametric and non-parametric methods (Kokkelenberg et al. 2008; Al-Shayea and Battal, 2013; Izadi et al. 2002; Glass et al. 2006; McMillan and Chan, 2006; Worthington and Lee, 2008; Abbott and Doucouliagos, 2003; Tzeremes and Halkos, 2010; Johnes and Johnes, 1993; Tauer et al. 2007; Kao and Hung, 2008; Colbert et al. 2000; Agha et al. 2011), it is limited in Turkey. Therefore, we aim at contributing the current literature by this way considering efficiency analysis of faculties of a Turkish university. 2. Material and Method In this section, we present the data used in evaluating the efficiency of faculties of the observed university and DEA methodology, respectively. 2.1. Data The observed university has 12 faculties. Considering all selection criteria, the research sample includes 12 DMUs spanning all of the faculties as shown in Table 1. Table 1 Decision making units # Faculties (DMUs) 1 Faculty of Education 2 Faculty of Electrical & Electronics 3 Faculty of Arts & Science 4 Faculty of Naval Architecture and Maritime 5 Faculty of Economic and Administrative Sciences 6 Faculty of Civil Engineering 7 Faculty of Chemical and Metallurgical Engineering 8 Faculty of Mechanical Engineering 9 Faculty of Architecture 10 Faculty of Art & Design 11 Technical Vocational School of Higher Education 12 School of Foreign Languages To ensure meaningful efficiency scores, the number of DMUs must be large enough relative to the number of input and output variables. We have 3 inputs and 3 outputs variables as in Table 2. We obtain input data from Academic Activity Report of the observed university for the year 2014. On the other hand, output data is provided by three various source. We provide data of number of publications from encourage publication list for the semester 2013-2014. While the data of number of projects is M. Gül – M. Yücesan – O Duman 9/3 (2017) 60-71 İşletme Araştırmaları Dergisi Journal of Business Research-­‐Türk 63 obtained from the university project support office, the percentage of budget used data is received from Directorate of Strategy Development of the university. After input and output variables were finalized, a data sheet is designed in such a way that the values of these variables are filled in by different departments and units. University website, publications, and brochures are used in data collection. The collected data within the scope of this study are shown in Table 3. All variables are belonged to the year 2014 except from the variable number of publication. NP variable is included by 2013-2014 academic year data (1.10.2013-30.09.2014) because the data after September 2014 has not published yet. It will be published after 2014-2015 academic year. Another reason that we used 2013-2014 academic year’s data is that most of data belongs to the year 2014. Table 2 Data type and source Variable Data source Total Number of Academic Staff (NA) Academic Activity Report of the year 2014 Total Number of Non-academic Staff (NN) Academic Activity Report of the year 2014 Number of Students (NS) Academic Activity Report of the year 2014 Number of Publications (PB): How many SCI, SSCI o