models cannot explain the direction and magnitude of the change in school district
inefficiency arising out of a change in any of the socioeconomic variables. Contrary to the
above, the current study uses a model proposed by Battese and Coelli (1995) called
‘inefficiency effects model’ that estimated a stochastic production function and an
‘inefficiency effects function’ simultaneously. In this model conventional educational
inputs under the control of the district administration are used in the production function
while the socioeconomic variables are used in the inefficiency function. Considering the
importance of the socioeconomic factors affecting students’ achievement scores efficiency
estimates in this study are robust and reliable.
Except the study by Currier (2007) ‘inefficiency effects model’ has rarely been used in the
education production function literature to assess productive efficiency. In her study Currier
(2007) estimates technical efficiency for Oklahoma school districts using Battese and Coelli
(1995) model but the specification of the ‘inefficiency effects function’ is incorrect. For
example, instead of using socioeconomic variables in the ‘inefficiency effects function’ the
author uses them as inputs in the education production function. The current study contributes
to the literature by improving our existing knowledge on how the socioeconomic and other
variables affect simultaneously in the measurement of efficiency when ‘inefficiency effects
model’ is applied. The study is important because socioeconomic and environmental factors
play a significant role in students’ achievement score. The major advantage of this model is
its unique ability to estimate the impact of socioeconomic factors and other educational
inputs simultaneously using panel data. The major focus in this study is the results obtained
from Model-2. However for comparative analysis efficiency estimates from a stochastic
frontier model (Model-1) are also reported. These two models are discussed in detail in the
next section.