Agriculture Reference
In-Depth Information
2. THE ANALYSIS OF ENERGY INTENSITY AT A
SECTORAL LEVEL
2.1 Decomposition of intensity indexes
The first approach is the formulation of ad-hoc decompositions, which at
best have the desirable properties of being mutually exclusive components
(without interaction terms) and completely exhaustive. A next step is the
aggregation of sector intensities using several index number formulae.
These formulations separate changes in energy use into intensity (input of
energy per unit of output of a given sector) and mix (differential change, in
the gross output of groups or subgroups of industries). Recent advances in
this approach are in Greening
et al.
(1997). These techniques are generally
criticised for their lack of theoretical foundation.
A foundation for intensity indexes decompositions is based on
growth accounting. Applications of growth accounting involve the use of
an aggregate production function in which the effects of changing energy,
material, capital and labour inputs and productivity are translated into
changes in output growth. A simple reformulating to measure changes in
energy productivity, as in Chan and Mountain (1990), is related to the
economic theory of index numbers. Using the translog index number
formula, and simple arithmetic, changes in energy intensity are explained
by weighted changes in other inputs to energy ratios (substitution) and
technical change. The main operational advantage in this approach and its
main interpretation problem is that it assumes long-run equilibrium for all
inputs and Hicks-neutral technical progress. Given that full equilibrium in
agriculture is not expected, that quasi-fixed input valuation is difficult, and
that technical change is not usually neutral in agriculture, the growth
accounting approach is not followed in this chapter.
2.2 Derived demand approaches
Under appropriate assumptions, the econometric analysis of input demand
can measure input substitution and several other effects, including
technological change. The properties of these constructs, such as non-
negativity, linear homogeneity and concavity, should be examined in the
case of cost functions, and other structural properties, such as
homotheticity and separability can also be explored. However, it is usual
for production or cost functions to be used for empirical study without fully
exploiting the capabilities of the econometric structure for this empirical
analysis.
Search WWH ::
Custom Search