Parametric and non-parametric methods of measuring departmental performance: An application to higher education

To evaluate the departmental performance in higher education system, we take an example of departments in a large public university in the United States. We include 19 departments for six consecutive academic years (2002-2008). Using both parametric distance function approach and non-parametric Data Envelopment Analysis approach, we estimate the technical efficiency for each department in each academic year. The result shows that social science departments on average are more efficient than natural science departments in terms of saving on input usage. We then estimate the marginal rate of substitution between inputs. We find that holding everything else constant, one faculty member's contribution to the output production is approximately equivalent to that of four graduate assistants. Both parametric and non-parametric analyses provide us with reasonably consistent result.