ON EFFICIENCY OF HIGHER EDUCATION WITH DATA ENVELOPMENT ANALYSIS AND REGRESSION

The efficiency of higher education systems is a constant concern at the level of each country. At the national policy level, governments are interested to know the determinant factors to enhance the higher education efficiency and whether increasing the state's contribution to the financing of the sector are a significant factor in increasing the efficiency in Higher Education. Our paper addresses so-called output-oriented Data Envelopment Analysis (DEA) requiring mathematical programming as a first stage and the regression of the obtained efficiency scores on explanatory factors as a second step. These allow us a cross-country analysis of the efficiency of higher education sector in different states with different GDP but as homogeneous as possible in the sample, on models that combine as input the level of state contribution to the financing of the sector from the total expenditure, and as output the percentage of graduates with diploma from the total population of the same age and employability rate. The analysis in the two steps, respectively the DEA followed by the regression, examines which of the non-discretionary inputs is statistically significant, respectively to what extent the state's contribution to the financing of systems or other factors such as GDP per capita influences the efficiency of the higher education systems chosen for comparison.