Introduction

In Chapter 1, we introduced a series of questions that the tools discussed in this book are designed to help answer. In this Chapter, we focus on the examination of technical inefficiency in the context of production frontier models using cross-sectional data. That is, based on observations across a number of units (be it countries, companies, schools, hospitals, bank branches, retail outlets, water treatment works, farms, etc.) for one given point in time, this chapter introduces tools to model a production frontier. In modeling the impact of technical inefficiency on production it is assumed, at least implicitly, that inputs are exogenously given and that the objective is to maximize output (i.e., output is the only choice variable). Only quantities are modeled (e.g., the number of staff or total hours employed, the amount of output produced) and no price information is included in the modeling, so only technical efficiency is measured and allocative efficiency and economic or cost efficiency cannot be considered (see Chapter 8). As such, the approaches set out in this chapter can answer questions such as those set out below.

Which schools, or which universities, are the best in terms of graduation rates, staff numbers per student graduated, and so on (controlling for characteristics such as the level of student educational attainment upon entry, etc.)? That is, given the quality of the students upon entry, the socioeconomic characteristics of the local catchment area, and/or the parents of the students and other characteristics, what is the maximum graduation rate per staff member (or minimum number of total staff per graduated student) that a given school should be expected to achieve and how far away from this maximum (minimum) is each school? The former might help the parents/students in deciding which college/university to attend, whereas the minimum number of total staff per graduated student is of more interest to budget holders.