Sunteți pe pagina 1din 11

Energy Economics 28 (2006) 444 454

www.elsevier.com/locate/eneco

The productivity and efficiency of the Australian


electricity supply industry
Malcolm Abbott
C/-PO Box 2995, Auckland, 0001, New Zealand
Received 27 September 2005; accepted 18 October 2005
Available online 19 December 2005

Abstract
Australias electricity supply industry has been through a period of reform over the last 10 years. The
purpose of this paper is to analyse the changes that have occurred to the Australian electricity supply
industry over the past 30 years, in order to evaluate to what degree these reforms have improved the
productivity and efficiency performance of the industry.
D 2005 Elsevier B.V. All rights reserved.
JEL classification: L94 Industry studies Electric utilities; Q48 Energy government policy
Keywords: Technical efficiency; Scale efficiency; Data envelopment analysis; Total factor productivity

1. Introduction
Over the past 20 years, the Australian Federal and State governments have made considerable
efforts to improve the efficiency and productivity of the electricity supply industry. Before 1991,
the Australian electricity supply industry consisted of a series of state-based, government owned,
mainly vertically integrated electricity authorities. Since the 1991 publication of the Industry
Commissions (1991) report on energy generation and distribution, these entities have been
broken up into their constituent parts (generation, transmission, distribution and retail): the
majority of states linked into a national market, competition introduced into the wholesale
electricity market, and retail competition introduced for large consumers. In two states,
electricity assets have also been privatised.
The basic purpose of the restructuring and introduction of competition into the electricity
supply industry has been to promote the more efficient operation of the industry, which in turn
E-mail address: malcolma@ais.ac.nz.
0140-9883/$ - see front matter D 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.eneco.2005.10.007

M. Abbott / Energy Economics 28 (2006) 444454

445

will lead to lower prices for electricity. The purpose of this paper is to use the data
envelopment analysis (DEA) Malmquist approach to estimate total factor productivity (TFP) of
the electricity supply industry, broken down into its state-based constituent parts over the long
term (1969 to 1999). By looking at the long-term changes in productivity levels, it will be
possible to get a more realistic view of the degree of change that has occurred to electricity
industry productivity and efficiency that has been brought about by the restructuring that
occurred during the 1990s. The paper is outlined as follows. In the Section 2, a description of
the broad structure of the industry over the past 30 years is presented. Past studies are reviewed,
the data employed and the estimation methodology are outlined in Section 3. The results are
analysed in Section 4 and some conclusions are made in Section 5.
2. Industry structure
Between 1969 and 1990, the Australian electricity supply industry was dominated in each
state by a single vertically integrated State government-owned authority or a collection of State
government-owned authorities. These bodies were also responsible for regulating electrical
safety of the industry in such areas. Under these arrangements, investment in the new
generation was largely the responsibility of State governments and their electricity authorities.
Electricity prices were set by the State governments and were designed to cover the industrys
costs, plus any return required by them as owners. Often politically motivated cross-subsidies
were built into price structures. The six main electricity supply companies for the six states
were, respectively, the Electricity Commission of New South Wales, the State Electricity
Commission of Victoria, the Queensland Electricity Commission, the Electricity Trust of South
Australia, the State Energy Commission of Western Australia (included gas as well as
electricity), and the Hydro-Electric Commission of Tasmania. With the exception of the States
of New South Wales and Queensland, each company run a vertically integrated operation
including generation, transmission and distribution of electricity. In New South Wales and
Queensland, distribution was carried out by a series of local power boards. Moreover, the
Federal Government operated the Snowy Mountains Hydro-Electric Authority as a purely
generation company.
After 1991, State governments began working to restructure their electricity authorities,
although the pace of this reform has varied across the different States.1 The first state to
introduce a wholesale electricity market was Victoria, which opened the Victorian Power
Exchange in 1994. This was followed in 1996 when Transgrid began operating the New South
Wales wholesale electricity market. These two major markets were subsequently joined and, in
1998, the National Electricity Market (NEM) commenced. The NEM is a wholesale market for
the supply and purchase of electricity, combined with an open access regime for use of
transmission and distribution networks in the participating jurisdictions of the Australian
Capital Territory, New South Wales, Queensland, South Australia and Victoria. Two
companies, the National Electricity Code Administrator Limited (NECA) and the National
Electricity Market Management Company Limited (NEMMCO), were formed in March 1996
by the participating jurisdictions to implement the NEM. NECA supervises, administers and
1

This process was initiated in May 1991 when the Industry Commission delivered its report entitled Energy generation
and distribution which recommended a major restructure of the electricity industry by disaggregating existing utilities
into generation, transmission and distribution components. It also recommended that each element be corporatised and a
competitive market established.

446

M. Abbott / Energy Economics 28 (2006) 444454

8
7
6
5

%4
3
2
1

1995

1993

1991

1989

1987

1985

1983

1981

1979

1977

1975

1973

1971

1969

Source: Electricity Supply Association of Australia


Fig. 1. Rate of return of the Australian electricity industry (earnings before interest and tax/total assets).

enforces an industry code of conduct and NEMMCO manages the wholesale electricity market
in accordance with the code. NEMMCO is responsible for the day to day management of the
power system.
As well as establishing a wholesale market for electricity, the vertically integrated electricity
companies were broken up into constituent parts and corporatised, and, in the Victorian and
South Australian cases, privatised. In each case, the basic division of the electricity authorities
was into the four distinct components: competitive elements (generation and retailing) and
monopoly elements (high voltage transmission lines and lower voltage local distribution).2 The
basic purpose of this separation has been to foster competition between generators of electricity
while leaving the monopoly transmission and distribution elements regulated. In each of the
NEM States (New South Wales, Victoria, Queensland and South Australia) generation was not
just separated from transmission and distribution but was also broken up into competing
generator companies. At the other end of the electricity chain, retail competition has been
introduced in each of the NEM states for large consumers in the late 1990s and is scheduled to
be progressively introduced in each state for all users in the years 2002 to 2004.
3. Productivity and efficiency measures
In determining the performance of a firm or industry a range of indicators can be used. The
conventional indicators of performance include such things as the level of rates of return and
prices. In Fig. 1, the rate of return for the Australian electricity supply industry is given for the
2
This model has occurred in all states except in Western Australia, which is outside the NEM and where a vertically
integrated government-owned company has been maintained, although gas supply was separated from this organisation
in 1994/1995.

M. Abbott / Energy Economics 28 (2006) 444454

447

period 1969 to 1996 (earnings before interest and tax/total assets). From Fig. 1, it can be seen
that the rates of return for the industry were quite modest in the 1970s but improved in the late
1980s and early 1990s. Another indicator of the industrys performance is the level of prices of
electricity. Fig. 2 gives an index of the average price of electricity (revenue from electricity
sales/electricity generated and sold) over the period 1969 to 1999. As can be seen in Fig. 2, the
average real retail electricity price declined through the early 1970s and 1990s but rose during
the 1980s. These two indicators would suggest that the electricity industry improved its
performance over the longer term and passed on some of these improvements to final
customers.
Utilities such as electricity supply, however, often operate in markets, which lack prices and
costs determined under competitive conditions. This was certainly the case for the electricity
supply industry before the reforms of the 1990s and this still exists for important components of
the industry such as distribution and transmission. In cases such as these, the usual market
indicators of performance, such as profitability and rates of return, cannot be used to gauge an
industrys economic performance accurately. It is possible that these financial indicators will be
more an indication of the distortions themselves rather than of the performance of the industry in
question. In these circumstances, indicators of the level and change of productivity and
efficiency are more appropriate indicator of an industrys performance.
Efficiency can be defined as being the degree to which resources are being used in an optimal
fashion to produce outputs of a given quantity. There are three main aspects of economic
efficiency: technical, allocative and scale (Farrell, 1957). Productivity on the other hand is a
measure of the physical output produced from the use of a given quantity of inputs. In the past,
the common method of determining levels of productivity and efficiency for the Australian
electricity supply industry has been to construct index numbers that indicate partial factor
productivities. For example, million kW h produced or sold per employee is a labour-based
partial productivity measure (see Fig. 3). In the electricity supply industry, capital productivity
measures are difficult to calculate given the difficulty in measuring capital inputs. Three
120
100

$ per MW

80
60
40
20
0

Fig. 2. Average real revenue for electricity sold in 1989/1989 $.

1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

1977

1975

1973

1971

1969

Source: Electricity Supply Association of Australia. Australian Bureau of Statistics.

448

M. Abbott / Energy Economics 28 (2006) 444454

million KW hours per person

1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

1977

1975

1973

1971

1969

Source: Electricity Supply Association of Australia.


Fig. 3. Labour productivity in the Australian electricity supply industry, million kW h per person employed.

commonly used indicators of generation capital utilisation, and therefore capital productivity, are
the load factor, the capacity factor and the reserve plant margin, respectively.3 The capacity
factor of the Australian electricity industry is given in Fig. 4. Further examples of labour
productivity and capital productivity measures can be found in the Bureau of Industry
Economics International benchmarking1996.
Although relatively easy to calculate the partial factor approach has a disadvantage in that it
can be misleading when looking at the change in productivity of an industry. If the process has
simply involved a substitution of capital for labour, then a TFP indicator that indicates a more
modest increase in overall productivity would be a more appropriate measurement of
productivity. A TFP index is the ratio of a total aggregate output index to a total aggregate
input index. Examples of this approach in the Australian case include the work conducted by
Lawrence, Swan and Zeitsch (Swan Consultants, 1991) and the Industries Assistance
Commission (1989). The method used in this study was a multilateral one, which compares
the performance of different firms over time. This study included one output (GW h of electricity
supplied) and four inputs: capital stock (calculated using the perpetual inventory method), labour
(persons employed), fuel (TJ) and other materials and services. During the 1990s, the Steering
Committee on National Performance Monitoring of Government Trading Enterprises used the
same method for calculating the TFP of the various government-owned electricity companies
operating in Australia (Steering Committee, 1992, 1998). Coelli (1996a) in another study
measured the technical efficiency of coal fired electricity generators in Australia between 1981/
82 and 1990/91.
3
Load factorratio of annual generation to the peak generational load. Capacity factorratio of electricity generated
to effective plant capacity. Reserve plant margin is the difference between generating capacity and peak load expressed as
proportion of peak load; Bureau of Industry Economics (1996).

M. Abbott / Energy Economics 28 (2006) 444454

449

60

50

40

% 30
20

10

1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

1977

1975

1973

1971

1969

Source: Electricity Supply Association of Australia.


Fig. 4. Capacity usage in the Australian electricity supply industry %.

Another method of determining TFP is to use DEA to apply the Malmquist procedure. This
approach can be applied to panel data to calculate indices of total factor productivity change,
technological change, technical efficiency change and scale efficiency change. These methods
are discussed in Fare, Grosskopf, Norris and Zhang (1994). The idea behind efficiency analysis
is to use data collected for the state-based electricity sectors, and to derive what is known as the
dbest-practice frontierT. What constitutes a best-practice frontier can change over time;
therefore, it is important to incorporate this aspect of the production process. The Malmquist
total factor productivity index is one method of doing so. In effect, the Malmquist index derives
an efficiency measure for 1 year relative to the previous year, while allowing the technical
progress frontier to shift.4 This approach allows the decomposition of productivity change into
technical change and technical efficiency change. The Bureau of Industry Economics used this
approach in its report on the electricity industry in 1996 (Bureau of Industry Economics, 1996).
The previously mentioned exercises in the estimation of electricity supply industry
productivity approaches were completed too early in the process of electricity market industry
restructuring during the 1990s to make any conclusive remarks about the success of the process
at promoting greater efficiency and productivity. In more recent times, DEA has been used to
benchmark the performance of Australian electricity companies against overseas ones rather
than determine changes in productivity over time. Interest in the use of DEA in assessing the
performance of government bodies in this fashion in Australia rose throughout the 1990s. In
1997, the Committee of the Review of Government Services Provision, which had been
established in July 1993, published its work on DEA pointing out its usefulness in analysing
the efficiency of the provision of government services. In Australia, DEA has been used in a
variety of fields besides electricity including telecommunications (Whiteman and Pearson,

Logically, the frontier may shift outwards and reflect technical progress, or it may shift inwards and reflect technical
regression.

450

M. Abbott / Energy Economics 28 (2006) 444454

1993), correctional services, hospitals and police patrols (Steering Committee, 1997), rail
(Productivity Commission, 2000) and vocational education and training (Abbott and
Doucouliagos, 2000a,b).
Whiteman and Bell (1994) used data from the Electricity Supply Association of Australia
(1993) and data on international utilities from the Electricity Association of the United Kingdom
(1993) to benchmark the performance of Australian electricity companies with those overseas.
Two output measures were used (electricity generated and sales per customer). Inputs were
represented by thermal generating capacity (MW), other generating capacity (MW) and
employment. The results from this study indicated that after accounting for scale and other
external factors the technical efficiency of the Australian electricity supply industry was close to
the best-practice benchmark performance. The Electricity Supply Association of Australia also did
a similar study to measure the performance of the Australian electricity supply industry relative to
selected utilities for the period 19901991 (Electricity Supply Association of Australia, 1994).
Finally, Whiteman (1999) used the stochastic production frontier approach, as well as DEA, to
estimate potential efficiency gains for Australian and international electricity suppliers. These
studies give a good indication of the degree to which there is scope to improve the degree of
efficiency and productivity to worlds best practice but do not give any indication of the degree to
which the Australian industry has improved its performance over the past decade.
In contrast to other measures of productivity, DEA requires data only on the physical
quantities of inputs employed and output produced if only technical and scale efficiency
indicators are to be estimated. To estimate allocative efficiency, factor prices are also needed.
Hence, the information requirements for DEA are fewer and less cumbersome than conventional
TFP analysis and the problem of allowing for different accounting treatments across
organisations does not arise. For this reason, in this paper, the DEA Malmquist approach is
used to determine the total factor productivity measurements of the Australian State electricity
sectors over the period 1969 to 1999.
The primary source of data is the statistical publication of the Electricity Supply Association
of Australia (Electricity Supply Association of Australia, various issues). In terms of determining
the level of output for the industry, the amount of electricity consumed in each jurisdiction is
used. Inputs used include the capital stock, energy used (in TJ) and labour employed. Instead of
using some sort of accounting method to estimate the stock of capital, physical indicators of the
main capital assets used in electricity supply have been used (this follows the approach used by
Whiteman and Bell, 1994). The capital stock has been determined by taking the physical
amounts of lines, transmission station capacity and generation capacity.
Table 1
DEA results for the State electricity sectors
Malmquist index summary of State sectors

New South Wales


Victoria
Queensland
South Australia
Western Australia
Tasmania
Australian mean

Efficiency
change

Technical
change

Pure efficiency
change

Scale efficiency
change

Total factor
productivity change

1.010
1.011
1.009
1.007
1.005
1.000
1.007

1.027
1.008
1.018
1.011
1.005
1.041
1.018

1.000
1.000
1.005
1.006
0.995
1.000
1.001

1.010
1.011
1.004
1.001
1.009
1.000
1.006

1.037
1.019
1.028
1.018
1.010
1.041
1.025

M. Abbott / Energy Economics 28 (2006) 444454

451

4. Results
The results of the DEA modelling are derived from using the computer program DEAP (see
Coelli, 1996b). The results of Malmquist indices for the electricity industry in each State are
reported in Table 1. Estimations for annual performance of all state industries during the period
of 1969 and 1999 are presented in Table 2 and plotted in Figs. 5 and 6.
In Table 1, the efficiency indices for technical efficiency change, technical change, pure
efficiency change, scale efficiency change and total factor productivity change for electricity
companies in each state are listed. By assuming constant return to scale, technical efficiency
improved in five out the six states while leaving Tasmania unchanged. There was technical change
in the industry in all of the six states. On average, technical progress occurred at 1.8% annual
growth rate, with New South Wales recording 3% rate of growth and Tasmania 4.1%. If variable
returns to scale are assumed, technical efficiency in most states remained constant. However,

Table 2
DEA results for the Australian electricity industry 19691999
Malmquist index summary of firms over sample period

1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
Annual mean

Efficiency
change

Technical
change

Pure efficiency
change

Scale efficiency
change

Total factor productivity


change

1.000
0.970
1.082
0.954
1.004
0.993
1.080
0.991
0.894
1.054
0.969
1.033
1.006
0.976
1.077
0.997
1.028
1.008
0.992
1.017
0.930
1.127
1.008
1.003
1.003
1.018
1.033
1.003
0.990
0.991
1.009
1.007

1.000
1.208
0.819
1.116
1.061
0.954
0.911
1.028
1.373
0.903
0.868
1.156
0.849
0.976
0.901
1.021
1.041
1.297
1.052
0.871
1.466
0.645
0.999
1.370
1.025
1.023
0.999
1.036
1.025
1.028
0.944
1.018

1.000
0.988
1.020
1.007
1.012
1.008
1.016
1.003
0.944
1.006
1.018
0.973
0.954
1.017
1.033
0.996
1.008
1.003
0.976
1.031
0.917
1.132
0.999
0.996
1.005
0.997
1.007
1.003
0.986
0.997
0.999
1.001

1.000
0.981
1.061
0.947
0.992
0.985
1.063
0.988
0.947
1.048
0.951
1.062
1.055
0.959
1.043
1.001
1.019
1.005
1.017
0.987
1.014
0.996
1.009
1.007
0.998
1.021
1.026
1.001
1.004
0.994
1.010
1.006

1.000
1.171
0.887
1.065
1.065
0.947
0.983
1.019
1.228
0.952
0.841
1.195
0.854
0.953
0.971
1.018
1.069
1.308
1.043
0.886
1.363
0.727
1.007
1.374
1.028
1.041
1.032
1.040
1.014
1.019
0.952
1.025

M. Abbott / Energy Economics 28 (2006) 444454

Index

452

2.4
2.3
2.2
2.1
2
1.9
1.8
1.7
1.6
1.5
1.4
1.3
1.2
1.1
1
0.9
0.8

EffCh
TechCh
TFPCh

1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

1977

1975

1973

1971

1969

Year
Fig. 5. Accumulative indices of efficiency charge, technical charge and total factor productivity charge for the Australian
electricity industry, 1969 to 1999.

Western Australia recorded a slight fall in efficiency, while Queensland and South Australia had
experienced a small improvement in efficiency performance. Scale efficiency shows a similar
pattern to technical efficiency change. Accordingly, total factor productivity increased in all the
states with average annual growth rate of 2.5%. Tasmania and New South Wales are the two states
experiencing rapid productivity changes, mainly due to technical progress.
The annual means of the Malmquist indices are presented in Table 2. They are consistent with
those from the Australian means of Malmquist indices for each state. However, we observe wide
fluctuations in all the DEA indices over the 31-year period. Such fluctuations are not unusual in
DEA studies, since the DEA method is much sensitive to year-to-year changes in inputs and
outputs. In the case of the electricity industry, firms have fixed levels of capital and often fairly
fixed levels of labour endowment in the short run, but a fluctuating demand for their output
3
NSW
VIC
QLD
SA
WA

2.5

TFP Index

2
1.5
1
0.5
0

1999

1997

1995

1993

1991

1989

1987

1985

1983

1981

1979

1977

1975

1973

1971

1969

Fig. 6. Accumulative TPF for each state during 19691999.

M. Abbott / Energy Economics 28 (2006) 444454

453

(electricity) over time. If a recession occurs, electricity sales fall but inputs will not change very
much, leading to a fall in levels of efficiency and productivity. Alternatively, if the industry has
under-utilised capacity and sales boom, then productivity levels will also rise significantly.
In the long run, investment in capital plant in the electricity industry takes place and technical
progress occurs; however, the process can be a bit lumpy. Generation capacity is constructed in
advance of increases in electricity sales, therefore firms experience a sizeable increase in inputs
before the increases in output, which then lowers efficiency and productivity during and
immediately after the construction year, but gives rise to a potential rise in efficiency and
productivity in the following years. What is important, therefore, is the trend changes in
productivity and efficiency.
In Fig. 5, the three accumulative indices of efficiency change, technical changes and TFP
show no certain patterns of rise and fall over the early part of the sample period, but a general
trend of rising since the mid-1980s. Over the whole sample period, there is a 2.5% TFP growth
per year in the electricity industry. Part of the reason behind the poor performance in
productivity change prior to the mid-1980s can be explained in terms of the large-scale
construction of new generation capacity by the industry in advance of demand. The subsequent
increases can be then partially explained by the growth in demand and fuller utilisation of the
new capacity. However, there appears an acceleration in productivity growth after 1990, which
can be linked to the changes that have occurred to the industry in the 1990s. Although the
improved performance cannot be entirely attributed to the reform process, as improvements
occurred in the late 1980s, a great deal of the improvement in industry performance can be
related to the reform of the electricity industry.
In Fig. 6, the accumulative total factor productivity indexes for five of the six states are given.
In each case, there has been a marked increase in the level of total factor productivity. Caution
should be taken when comparing the total factor productivity of the different states as they do
each have individual characteristics that make this difficult. The stand out state is the Tasmanian
system, which is predominately hydro-generation compared to coal-fired stations that
predominate in the other states. Even amongst the other states, there is some divergence in
that the Victorian industry is based on brown rather than black coal which has implications for
the capital costs of generation capacity. Nonetheless, productivity changes for each state over
time are still useful in displaying the performance of the industry in each state.
5. Conclusion
In this paper, a DEA Malmquist approach has been used to estimate total factor productivity
of the electricity supply industry, broken down into its state-based constituent parts over the
period 1969 to 1999. The results indicate that there has been a substantial improvement in the
performance of the industry since the mid-1980s. The beginning of this improvement pre-dates
the substantial restructuring of the industry in the early 1990s although the improvement in the
productivity performance of the industry did speed up after 1991. It should be noted, however,
that, according to the study by Whiteman (1999), there is still scope for further improvements.
References
Abbott, M., Doucouliagos, C., 2000a. Amalgamations and the efficiency of Queensland TAFE Institutes. Australian and
New Zealand Journal of Vocational Education Research 8 (2).
Abbott, M., Doucouliagos, C., 2000b. Interrelated factor demand in Queensland TAFE Institutions. Journal of Economic
and Social Policy 5 (1), 41 56.

454

M. Abbott / Energy Economics 28 (2006) 444454

Bureau of Industry Economics, 1996. Electricity 1996: international benchmarking, research report 96/16, Canberra.
Coelli, Tim, 1996a. Measurement and sources of technical inefficiency in Australian coal fired electricity generation,
working paper no. 1/96, Centre for Efficiency and Productivity Analysis, University of New England.
Coelli, Tim, 1996b. A guide to DEAP version 2.1: a data envelopment analysis (computer) program, working paper no.
8/96, Centre for Efficiency and Productivity Analysis, University of New England.
Electricity Supply Association of Australia, 1994. International performance measurement for the Australian electricity
supply industry 19901991, June.
Electricity Supply Association of Australia, various issues, Electricity Australia: the electricity supply industry in
Australia year. . ., Melbourne.
Farrell, M.J., 1957. The measurement of productive efficiency. Journal of the Royal Statistical Society 32, 237 243.
Industries Assistance Commission, 1989. Government (Non-Tax) Charges, Report no 422. AGPS, Canberra.
Industry Commission, 1991. Energy Generation and Distribution, Report No. 11, Three Volumes, 17 May. AGPS,
Canberra.
Productivity Commission, 2000. Progress in Rail Reform, Melbourne.
Steering Committee on, 1997. Data Envelopment Analysis; A Technique for Measuring the Efficiency of Government
Service Delivery, Canberra. April.
Steering Committee on National Performance Monitoring of Government Trading Enterprises, 1992. Measuring the Total
Factor Productivity of Government Trading Enterprises. Industry Commission, Belconnen, ACT.
Steering Committee on the National Performance Monitoring of Government Trading Enterprises, 1998. Performance of
Government Trading Enterprises, 199192 to 199697, Canberra.
Swan Consultants, 1991. An international benchmark for Australias electricity supply industry, report prepared for the
Business Council of Australia.
Whiteman, J., 1999. The potential benefits of Hilmer and related reforms: electricity supply. Australian Economic
Review 32 (1).
Whiteman, John, Bell, Christopher, 1994. Bench marking electricity using data envelopment analysis. Economic Papers,
63 73.