Productivity and Efficiency in Australian Agriculture

This report presents a comprehensive analysis of total factor productivity (TFP) in Australian broadacre agriculture, utilizing state-level data from 1990 to 2011. The study employs the Färe-Primont index to estimate and decompose TFP changes into measures of technical change and technical efficiency change. The findings reveal an average TFP growth rate of 1.36% per annum during the specified period, with variations across states and time. The research further identifies declining growth in technical possibilities (technological progress) as the primary driver of the observed slowdown in productivity growth. The paper reviews theoretical issues related to TFP indexes, including the advantages of the Färe-Primont index over the Hicks-Moorsteen index for multi-lateral comparisons. It also discusses the empirical literature on productivity growth in Australian agriculture, highlighting the limitations of previous studies and the importance of state-level analysis. The methodology includes the use of non-parametric data envelopment analysis (DEA) to estimate the production frontier. The report contributes insights for policy formulation by exploring the various components of productivity growth and providing a detailed overview of the Australian agriculture sector.

Nonparametric estimates of productivity and
efficiency change in Australian Broadacre
Agriculture*
Farid Khan, Ruhul Salim and Harry Bloch†
This paper computes and decomposes F€are-Primont indexes of totalfactor produc-
tivity of Australian broadacre agriculture by estimating distance functions.Using
state-leveldata from 1990 to 2011,the empiricalresults show that TFP grew at an
average rate of 1.36 per cent per annum in the broadacre agriculture over the period
1990–2011.There are variations oftotal factor productivity (TFP)growth across
states and fluctuations over time within each state and territory.However,overall,
there is a clear movementtowards slower TFP growth across the sample period.
Further decomposition of TFP growth shows that it is declining growth in technical
possibilities (technologicalprogress) that is the main driver of the declining trend in
productivity growth in broadacre agriculture in Australia.
Key words:Australia, broadacre agriculture, distance function, F€are-Primont Index,
technical efficiency, total factor productivity.
1. Introduction
Over the lastfew decades,efficiency and productivity growth analysis in
agriculture has attracted attention of economic researchers and policy makers
in both developed and developing countries (Battese and Coelli 1995; Bravo-
Ureta et al.2007;Samarajeewa et al.2011;O’Donnell2012b;Van Beveren
2012). It is not easy for a country to advance prosperity without attaining a
considerablegrowth in productivity.Recently,in the global context,
agriculturalproductivity growth has been falling,particularly in developed
economies.This also has implicationsfor food security in developing
countries, where growing populations will continue to raise demand for food
in the coming decades (Pardey et al. 2006).
There is limited empiricalevidence concerning the drivers of totalfactor
productivity (TFP) growth and its componentsin Australian broadacre
agriculture.Previous empiricalstudies ofAustralian broadacre agriculture
make limited useof decomposition analysisto find the componentsof
* We are gratefulto the editor and an associate editor ofthis journal,as wellas two
anonymousrefereesfor useful commentsand suggestionsto improve the quality and
presentation of the paper. However, we solely remain responsible for any remaining errors.
†Farid Khan is at Department of Economics, Rajshahi University, Rajshahi, Bangladesh.
Ruhul Salim (email: [e-mail]) and Harry Bloch are with the School of
Economics & Finance,Curtin Business School, Curtin University,Perth, Western Australia,
6845, Australia.
© 2014 Australian Agricultural and Resource Economics Society Inc.
doi: 10.1111/1467-8489.12076
Australian Journal of Agricultural and Resource Economics, 59, pp. 393–411
The Australian Journal of
Journal of the Australian
Agricultural and Resource
Economics Society

productivityand efficiencychanges.They are mainly concernedwith
estimating the growth oftotal factor productivity and technicalefficiency
change. However, productivity researchers have recognized the importance
measuring differenttypes ofefficiency change in both the agriculture and
manufacturing sectors.
Using aggregate data,O’Donnell(2010) computes TFP indexes and the
components ofTFP change in Australian agriculture for the period from
1970 to 2001.One of the major limitations of this study is the use of the
Hicks-Moorsteen TFP index thatfails the transitivity testand is thus
unsuitablefor multi-lateraland multi-temporalcomparisons(O’Donnell
2012b).O’Donnell (2014)also providesargumentthat the F€are-Primont
index is preferred to the Hicks-Moorsteen index in estimating productivity
changes and its components.
Other previous studies in Australian agriculture mainly focus on aggregate
(Mullen and Cox 1996) or regionaland industry-specific (Fraser and Hone
2001) productivity growth.However,state-levelproductivity analysis in the
agriculturalsector is reported in studies conducted in other countries (Ball
et al. 2004;Laurenceson and O’Donnell2011;O’Donnell2012b;Rahman
and Salim 2013).These studies suggest that analysis of state-leveldata can
provide useful insights into the drivers of productivity growth.
The main objective of this paper is to estimate totalfactor productivity
changes in Australian broadacre agriculture and to decompose these change
into measures oftechnicalchange and technicalefficiency change.This is
done using the F€are-Primont index of total factor productivity, which satisfies
all axioms ofindex number theory,including the identity and transitivity
axioms.Further,the paper uses a new linear programming methodology
developed by O’Donnell(2014) for exhaustively decomposing TFP change
into measures of technical change and technical efficiency change. Finally, b
exploringthe differentcomponentsof productivitygrowth, this paper
contributes information for policy formulation, as different policies generally
affect different components of productivity change.
The rest of the paper proceedsas follows. The next section reviews
theoreticalissuesand previousempiricalstudies.Section 3 outlinesthe
empirical methodology to be used, followed by a discussion on data sources
in Section 4.Section 5 presents the empiricalestimates and an analysis of
results. Finally, Section 6 concludes the paper.
2. Review of theoretical and empirical literature
2.1. Theoretical issues: total factor productivity index
The change in the level of TFP can be measured as the ratio of an aggregate
output quantity index to an aggregate input quantity index. There are severa
formulas available for constructing such indexes in the productivity litera-
ture. The Tornqvistindex,the Fisher index and the Malmquistindex of
© 2014 Australian Agricultural and Resource Economics Society Inc.
394 F.U. Khan et al.

Caves et al. (1982) are some of the widely used indexes in empirical research
in agriculture.
Both the Tornqvist index and the Fisher index satisfy the identity axiom,
which says that if two firms produce the same outputs using the same inputs
the relative index value is one. However, neither of these two indexes satisfies
the circularity(transitivity)axiom, which requiresthat both a direct
comparison and an indirectcomparison oftwo firms/periods through an
intermediate firm/period will yield the same estimate of productivity change.
Intransitivity makes indexes inappropriate to be used to make multi-lateral or
multi-temporal comparisons (O’Donnell 2012b, 2014).
Malmquist productivity indexes are one of the standard approaches in the
productivity literature (Lovell2003),that can be decomposed exhaustively
(F€are et al. 1994), especially in nonparametric specifications and for translog
technologies (Bjurek 1996). However, the data envelopment analysis (DEA)
estimatesof Malmquistindexesare incomplete measuresof productivity
change as they fail to capture productivity changes associated with changes in
scale (Grifell-Tatje and Lovell 1995; O’Donnell 2012b).In fact, the
Malmquist index is not a productivity index; rather, it is only a measure of
technical change and technical efficiency change (F€are et al. 1994). Except in
specialcases,the MalmquistTFP index may not reliably measure TFP
changeand its decompositions.It generallyyields biased estimatesof
technical change and efficiency change (O’Donnell 2012a).
Recently,two other indexes,namely the Hicks-Moorsteen TFP index
proposed by Bjurek (1996)and the F€are-Primontindex proposed by
O’Donnell(2014)are used in constructing productivity indexes.They can
be broken into recognizable components without requiring data on prices and
any restrictive assumptions concerning statisticalnoise.However,between
the two indexes, O’Donnell (2014) argues that the F€are-Primont index is more
reliable than the Hicks-Moorsteen index, as the former can be used to make
reliable multi-lateral and multi-temporal comparisons. The Hicks-Moorsteen
index can validly only be used to make a single binary comparison, as it fails
the transitivity test.
Apart from choosing an index formula,decomposing TFP indexes into
measuresof technicalchangeand other measuresof efficiency change
involves estimating the production frontier. A range of approaches has been
proposed in the literature on how to estimate the production technology. The
two competingapproachesto obtain potential or frontier output are
stochastic frontier analysis (SFA) and DEA.
The SFA approach is a stochastic parametric approach,which parame-
terizesthe production frontierundersome distributionalassumptionsof
random error terms.This approach uses a two-component error term – a
stochastic random error component and a technicalinefficiency component
(Aigner and Chu 1968;Aigner et al.1977;Meeusen and van den Broeck
1977). The main weaknesses of this approach are that results may be sensitiv
to the choice offunctionalform of the unknown production frontier and
© 2014 Australian Agricultural and Resource Economics Society Inc.
Productivity growth in the broadacre agriculture 395

assumptions concerning the distributions of error terms, and the estimates o
unknown parameters may be statistically unreliable if sample sizes are small
(O’Donnell 2014). The issue of endogeneity is also likely to be associated wit
estimating multiple-inputand multiple-outputproduction technologiesin
SFA model (Mutter et al. 2013; O’Donnell 2014). Besides, the SFA approach
has difficulties in identifying some components of TFP change, such as pure
scale efficiency change and pure mix efficiency change.
Data envelopmentanalysisis a non-parametric deterministic approach
popularly employed to estimatethe production frontier.This approach
primarily involves mathematicalprogramming and requires no assumption
about the error term and the distributions of the parameters (e.g., means an
variances)(Farrell 1957).Moreover, it does not require any explicit
assumptionsregarding the functionalform of the production frontieror
any structure to compute relative efficiency scores (Banker 1993). However,
limitation of assuming away the statistical noise is that it leads to an intrinsic
bias with all deviations from the estimated frontier attributed to inefficiency
(Coelli et al. 2005). If there is substantial statistical noise in the data, then th
use of DEA becomes problematic and stochastic frontier analysis remains the
only choice as it allows statistical noise (Simar and Wilson 2000). Nonethe-
less, this paper uses a non-parametric DEA to estimate a production frontier
and then to compute and decompose the TFP index. This allows more direct
comparison to most other studies that have applied index number approache
to measuring productivity in Australian agriculture.
2.2. Empirical studies: productivity growth in agriculture
A substantialbody of literature has emerged over the past few decades on
efficiency and productivity measurementin Australian agriculture.At the
economy-wide level,Males et al. (1990)measure productivity growth of
broadacre agriculture and find that TFP growth averaged 2.2 per cent per
annum over the period 1978–1989.They also disaggregate the sample size
into different enterprise types and find that productivity growth rates vary
across enterprise types.Particularly,they report5.5 per centproductivity
growth per annum for specialist crops. Knopke et al. (1995) extend a similar
dataset to 1994 and find the productivity growth of the specialist crop slowe
to 4.6 per cent per annum,while productivity growth in broadacre agricul-
ture was at 2.7 per cent per annum for the period 1978–1994.Dividing the
farms into three groups,they also find thatscale matters significantly in
productivity growth.
Using a farm-level dataset covering the period from 1953 to 1994, Mullen
and Cox (1996) find an average rate of productivity growth of 2.5 per cent
per annum in Australian broadacre agriculture.They compare alternative
measuresof productivitygrowth including traditional index number
approaches, a scale-adjusted Christensen and Jorgenson index, nonparamet-
ric measures and an econometric estimate of a translog cost function.They
© 2014 Australian Agricultural and Resource Economics Society Inc.
396 F.U. Khan et al.

find a smallvariation in average TFP growth from 2.4 to 2.6 per cent over
the differentestimation approaches.These robustresults from parametric
and nonparametric methodologies suggestconfidence for traditionalindex
number approaches,such as the Fisher index. However, when they
disaggregatethe study periods into three sub-periods,they find that
productivityin Australian broadacreagriculturedeclined from 2.0 to
1.8 per cent between the sub-periods 1953–1968 and 1969–1984.
Recently,using country-levelagriculture data for 88 countries over the
period 1970–2001,O’Donnell(2010) computes indexes of TFP change and
decomposes them into economically meaningful components. Particularly in
Australia,O’Donnellshows that over the period agriculture experienced a
15 per centdecline in productivity and he explainsthat increasesin net
returns to agriculture are associated with falls in productivity. However, this
study uses the Hicks-Moorsteen TFP index,which is only valid for binary
comparisons. Moreover, it uses only two outputs and is for overall country-
level agriculture data which fails to capture regional variations in agricultural
productivity.
One major drawback of the previous studies of productivity for Australian
broadacre farmsis that it is difficultto disentangle changesin technical
efficiency and scale-mix efficiency from the contribution of technical change
to productivity growth.Studiesthat use the conventionalmeasuresof
productivity do not take the multiple sources of the productivity growth into
account.For example,the previous studies of Australian broadacre farms
cannotproperly assesswhetherthe productivity change issourced from
improving the rate of technicalprogress or from improving levels of either
technical or scale and mix efficiency. Further, most of the previous studies use
imputed prices for the broadacre outputs or inputs to construct the indexes,
which may bias the estimates due to measurement error.
3. Empirical methodology
3.1. Total factor productivity indexes
This article uses the F€are-Primont index to compute and decompose total
factor productivity (TFP)growth into a measure oftechnicalchange and
severalfiner measures of efficiency change of Australian broadacre agricul-
ture.Index number approaches to measuring totalfactor productivity as a
ratio of aggregate outputsover aggregate inputscan be traced back to
Jorgenson and Griliches(1967),Nadiri (1970)and Good et al. (1996).
However, these early studies rely on market prices to form aggregates in case
of multiple outputs and multiple inputs farms.
Recently,O’Donnell(2012b)defines TFP asthe ratio of an aggregate
output to an aggregateinput wherethe aggregatorfunctionsare non-
negative, non-decreasing and linearly homogeneous. These properties of the
aggregator functions are crucial to construct a TFP index that satisfies basic
© 2014 Australian Agricultural and Resource Economics Society Inc.
Productivity growth in the broadacre agriculture 397

axioms from index theory. Let qit 2 <J
þ and xit 2 <K
þ denote vectors of output
and input quantities for firm i in period t. Following O’Donnell (2012b), TFP
is defined as TFPit = Qit/X it, where TFPit indicates the TFP of firm i in period
t, and Qit = Q(qit) and Xit = X(xit) are aggregate output and aggregate input,
respectively.
Using this TFP definition, the productivity index that compares the TFP of
firm i in period t with the TFP of firm h in period s is (O’Donnell 2014):
TFP hs;it¼ TFP it
TFP hs
¼ Qit=Xit
Qhs=Xhs
¼Qit=Qhs
Xit=Xhs
¼Qhs;it
Xhs;it
ð1Þ
Where Qhs,itand Xhs,itare the output quantity index and input quantity index,
respectively.Equation (1) showsthat TFP change can be obtained by
dividing an index of output growth by an index of input growth. The index
number formed in this way as a measure of relative productivity is said to be
multiplicatively complete (O’Donnell 2012a).
The F€are-Primontindex is a memberof a class of ‘multiplicatively
complete’productivity indexes thatuses the following non-negative,non-
decreasing and linearly homogenous aggregator functions: Q(q) = DO(x0, q,
t0) and X(x) = DI(x, q0, t0), where DO(x0, q, t0) and DI(x, q0, t0) are the
Shephard output and input distance functions, respectively, representing the
production technology available in period t0. Here,q0 and x0 are arbitrary
vectors of representative outputs and inputs.Then O’Donnell(2011,2014)
shows that the F€are-Primont index that measures the TFP of firm i in period t
relative to the TFP of firm h in period s is:
TFP hs;it¼ DOðx0; qit; t0Þ
DOðx0; qhs; t0Þ
DI ðxhs; q0; t0Þ
DI ðxit; q0; t0Þ ð2Þ
3.2. Measures of efficiency
Following O’Donnell (2012a), several measures of efficiency are defined as:
Output-oriented technical efficiency;OTEit ¼Qit
Qit
; ð3:aÞ
Output-oriented scale efficiency; OSEit ¼Qit=Xit
~Qit=~Xit
; ð3:bÞ
Output-oriented mix efficiency; OMEit ¼Qit
^Qit
; ð3:cÞ
Residual output-oriented scale efficiency; ROSEit ¼ ^Qit=Xit
Qit=Xit
and ð3:dÞ
© 2014 Australian Agricultural and Resource Economics Society Inc.
398 F.U. Khan et al.

Residual mix efficiency; RMEit ¼ ~Qit=~Xit
Qit=Xit
: ð3:eÞ
where,Qit is the maximum aggregate output that is technically feasible to
produce a scalarmultiple of qit using xit; ^Qit is the maximum possible
aggregate output using xit to produce any output vector;~Qit and ~Xit denote
the aggregateoutput and input quantitiesat the point whereTFP is
maximized subject to the constraint that the output and input vectors are
scalarmultiplesof qit and xit respectively;and Qit and Xit denotethe
aggregate output and input quantities at the point of maximum productivity.
As an overallmeasure offirm performance,O’Donnell (2011)measures
TFP efficiency (TFPE) as the ratio of observed TFP to the maximum TFP
given the available technology.Mathematically,TFP efficiency of firm iin
period t is
TFPE it ¼TFP it
TFP t
¼Qit=Xit
Qit=Xit
ð4Þ
where TFPt indicates maximum TFP possible given the technology in period
t, and Qt and Xt are the TFP-maximizing aggregate output and aggregate
input,respectively.Following O’Donnell(2012b),the TFP decompositions
are as follows:
TFP it ¼ TFPt OTEit OME it ROSE itð Þ ¼ TFPt OTEit OSE it RME itð Þ:
ð5Þ
A similar decomposition holds for firm h in period s.Then,the relative
TFP index comparing TFP of firm iin period t with the TFP of firm h in
period s can be decomposed exhaustively in either of the two following ways:
TFP hs;it¼ TFP it
TFP hs
¼ TFP t
TFP s
OTEit
OTEhs
OME it
OME hs
ROSEit
ROSEhs
ð6:aÞ
TFP hs;it¼ TFP it
TFP hs
¼ TFP t
TFP s
OTEit
OTEhs
OSEit
OSEhs
RME it
RME hs
: ð6:bÞ
The first term in parentheses on the right-hand side of each of the above
equations is a measure of technicalchange,which compares the maximum
TFP possible in period t with the maximum TFP possible in period s.The
other terms on the right-hand sides of the equations are the different output-
oriented measures of relative efficiency, including relative technical efficiency
relative mix efficiency and relative residualscale efficiency.The other two
alternativecomponentsare output-oriented relativescale efficiency and
relative residual mix efficiency.
Further, Equations (6.a) or (6.b) can be written as
© 2014 Australian Agricultural and Resource Economics Society Inc.
Productivity growth in the broadacre agriculture 399

TFP hs;it¼ TFP it
TFP hs
¼ TFP t
TFP s
OTEit
OTEhs
OSME it
OSME hs
ð6:cÞ
where OSME it = OMEit 9 ROSE it = OSEit 9 RME it is the measureof
scale-mix efficiency defined by O’Donnell(2012b),which is a combined
measure of scale and mix efficiency. The output-oriented scale-mix efficiency
OSME, measures the increase in TFP between a technically efficient point
with the observed scale and input mix to the point of maximum productivity.
3.3. Estimation using the DEA approach
The F€are-Primont index is a distance-based index which can be estimated
relativelystraightforwardlyby DEA methodology, which assumesthe
frontier of a firm takes the linear form in the neighbourhood ofthe
technically efficient point (O’Donnell 2011). The distance function represent-
ing the production technology isalso locally linear.Then, according to
O’Donnell (2011)the (local)output distance function holdsonly in the
neighbourhood of the (technically efficient) point (xit, qit/OTE it) and takes the
form:
DOðxit; qit; tÞ ¼ ðq0
itaÞ=ðc þ x0
itbÞ ð7Þ
The standard output-oriented DEA problem involves finding the solutions
for the unknown parameters in Equation (7) to minimize technical efficiency:
OTE it = DO(xit, qit, t). If a and b are non-negative, then the only constraint
that needs to be satisfied is DO(xit, qit, t) ≤ 1. Setting an additional constraint
q0
ita ¼ 1, the DEA problem takes the following form of linear programming
(LP):
DOðxit; qit; tÞ1 ¼ OTE1
it ¼ min
a;c;b
fc þ x0
itb : cs þ X0
b Q 0
a; q0
ita
¼ 1; a 0; b 0g ð8Þ
where Q is a vector of observed outputs, X is a vector of observed inputs, an
s is a unit vector (for details, see O’Donnell 2011).
To compute the F€are-Primont aggregates, the variant of LP that needs to
be solved is:
DOðx0; q0; t0Þ1 ¼ min
a;c;b
fc þ x0
0b : cs þ X0
b Q 0
a; q0
0a ¼ 1; a 0; b 0g ð9Þ
Estimates of aggregate outputs, Qit and aggregate inputs, Xit for all i and t
are then estimated as:
Qit ¼ ðq0
ita0Þ=ðc0 þ x0
0b0Þ ð10Þ
© 2014 Australian Agricultural and Resource Economics Society Inc.
400 F.U. Khan et al.

Xit ¼ ðx0
itg0Þ=ðq0
0/ 0 d 0Þ ð11Þ
where a0, b0, c0 solve Equation 9. The computer software DPIN1 3.0 further
uses linear programming technique to decompose productivity into various
efficiency indexes.
4. Data sources
This paper makes use of a state-level panel dataset from the AgSurf covering
the period 1990–2011. The data in AgSurf are sourced from the annual farm
surveys of ABARES (Australian Bureau of Agricultural Resource Economics
and Sciences).The datasetconsistsof observationson quantitiesof
agriculturalinputs,outputs and corresponding values in each state in each
year. In the case of some outputs, quantity data are not available. This study
uses six major inputs:land,labour,capital,fertilizer,materials and services
and rainfall, and four outputs:crops, livestock,wool and other output
variables.
Rainfalldata are collected from Australian Bureau of Meteorology.This
study includesthe rainfall variable asan importantinput of broadacre
agriculture production assuming that seasonal conditions may have influence
on broadacre agriculture in Australia.The period for measuring rainfallis
chosen to match thegrowing season in each state.The detailsof the
construction of variables are given in the Appendix.
5. Analysis of empirical results
Table 1 presentsthe F€are-Primontestimatesof actual TFP, maximum
TFP and TFP efficiency and their relative changes between 1990 and 2011.
The F €are-Primontindexesare estimated assuming thatthe production
technologyexhibits variable returns to scale (VRS). The production
possibilities setalso allows both technicalprogress and technicalregress.
All the indexesreported in this table are meaningfully comparablein
performance,both spatially and inter-temporally,as the indexesare
transitive.
The estimates ofactualTFP relative to the DEA maximum in the first
column show that WA (Western Australia) was the most productive state and
QLD (Queensland) was the least productive state in 1990. The difference in
productivitybetween thetwo states was 72 per cent(TFP WA/TFP QLD
= 0.69/0.40 = 1.72),so that WA was 72 per centmore productivethan
QLD in 1990. The TFP estimates in the second column show that in 2011 WA
and QLD remained the most productive state and the least productive state,
respectively.The productivity difference between the highestand the least
1 DPIN 3.0 is computer software provided by the Centre for Efficiency and Productivity
Analysis, University of Queensland, Australia.
© 2014 Australian Agricultural and Resource Economics Society Inc.
Productivity growth in the broadacre agriculture 401

productive states remained almost the same, 70 per cent(TFP WA/
TFP QLD = 0.97/0.57 = 1.70).
The third column of Table 1 reveals that productivity increased in all states
over the sample period. Among them, SA (South Australia) experienced the
largest increase in productivity,which was 46 per cent between the period
1990 and 2011.The last row of the table showsaverageestimatesfor
Australia. It shows that, on average,Australian broadacreagriculture
experienced a 33 per cent productivity increase between the periods 1990 a
2011.
The maximum TFP (TFP*) estimates are obtained under the assumption
that in any given period allstates experience the same setof production
possibilities, which can be observed in the first and second column of TFP*
estimates in Table 1.The third column of TFP* estimates reveals that over
the period between 1990 and 2011 technicalpossibilities improved by 41 or
1.62 per cent per annum (ln(1.41)/(2011–1990) = 0.0162 or 1.62%)).
The TFP change is a combined effect of the maximum technically feasible
change and efficiency change (dTFP = dTFP* 9 dTFPE). The third column
of TFPE estimates reveals that efficiency has improved over the period in SA
QLD and WA, but fell for NSW (New South Wales), TAS (Tasmania) and,
less so, for VIC (Victoria). Estimates shown in Table 1 suggest that, in spite
of a fall in efficiency in few states,all states experienced TFP improvement
due to the more powerful common improvement in technology. In QLD, SA
and WA, both technicalpossibilities and efficiency increased,resulting in
large TFP increase.
When TFPE is further decomposed into OTE and OSME,the estimates
indicate that OTE is almost always equal to 1.0 and, in particular, equals 1.0
for all states in both 1990 and 2011. Other studies that have calculated OTE
also find most values are equalto 1.0 or at least very close,suggesting that
pure technical efficiency is commonly achieved (see, for example, O’Donnell
2010 and 2012b). This implies that the shortfall in TFP efficiency is due solely
to scale and mix efficiency,rather than pure technicalefficiency,with
Table 1 TFP Index and its components: 1990–2011
States TFP TFP* TFPE
1990 2011 Change 1990 2011 Change 1990 2011 Change
NSW 0.63 0.73 1.14 0.69 0.97 1.41 0.92 0.75 0.81
VIC 0.53 0.72 1.38 0.69 0.97 1.41 0.76 0.75 0.98
QLD 0.40 0.57 1.43 0.69 0.97 1.41 0.58 0.59 1.02
SA 0.59 0.86 1.46 0.69 0.97 1.41 0.85 0.89 1.04
WA 0.69 0.97 1.41 0.69 0.97 1.41 1.00 1.00 1.00
TAS 0.50 0.61 1.20 0.69 0.97 1.41 0.73 0.62 0.86
AUS 0.56 0.74 1.33 0.69 0.97 1.41 0.81 0.77 0.95
Note: Other output-oriented measures, namely OTE, OSE, OME, ROSE and RME, are not reported here
to conserve space.
© 2014 Australian Agricultural and Resource Economics Society Inc.
402 F.U. Khan et al.

TFPE = OSME in each state for both time periods. Thus, the results are not
reported in Table 1 to save space.2
Table 2 reportsestimated annualgrowth in TFP, maximum TFP and
technicalefficiency ofbroadacre agriculture in Australian states for three
cumulative periods of 1990–2000, 1990–2007, and 1990–2011 and also for tw
recent sub-periods of 2000–2007 and 2007–2011.The choices of these sub-
periods are made based on studies by ABARES in Australia and to facilitate
comparison with them (e.g., Sheng et al. 2011). The entries in the table can be
interpreted asthe average rate ofgrowth for the indicated periods.For
example,in the 1990s,WA experienced the highestaverage rate ofTFP
growth which wasestimated to be 3.78 per centper annum (ln(TFP2000/
TFP 1990)/(2000–1990) = ln(1.006/0.689)/10 = 0.0378).During this sub-per-
iod, broadacre agriculture experienced a 3.78 per cent annual average rate of
technological progress.
The last row of Table 2 presents estimates of the average annualrate of
growth of broadacre agriculture in Australia,which shows that on average
broadacre agriculture in Australia experienced an annual productivity growth
rate of 1.36 per centduring the study period of1990–2011.In spite of a
0.26 per cent per annum fall in overall efficiency, the main driver on back of
this productivity growth was a 1.62 per cent per annum technicalprogress
over the entire period of study.However,there has been a generally falling
average rate of TFP growth over the sub-periods. For example, average TFP
growth was estimated to be 2.40 per cent per annum in the 1990–2000 sub-
period, 1.65 per cent per annum in 2000–2007 and 1.74 per cent per annum
in the latest sub-period of 2007–2011.
The slowdown of total factor productivity growth in broadacre agriculture
has been largely driven by a slowing technicalchange during the past two
decades.Table 2 shows that technicalchange in broadacre agriculture fell
from 3.78 per cent per annum in 1990–2000 to 1.55 per cent per annum in
2000–2007.In the latest sub-period,2007–2011,it dropped as low as
3.64 per cent, implying technicalregress.However,like Coelli and Rao
(2005) and O’Donnell (2010, 2012b), this study allows technical regress as a
proxy for the adverse effects ofexternalshocks omitted from the model.
These include the poorenvironmentalconditionsexperienced during the
Millennium Droughtof the early 2000s.Also, the technicalregresswill
account for measurement errors or other sources of statistical noise, as DEA
makes no allowance for them (O’Donnell 2010).
The term technical change is viewed in a broad sense – the same way that
Solow expressed it,namely ‘any kind ofshift in the production function’
(Solow 1957,p. 312).It is the measure ofthe change in the production
possibilitiesset, which might be causedby any changesin external
environmental factors, including weather and climatic variations (O’Donnell
2010).A drought will cause production possibilities to contract(the same
2 Detailed results of the decomposition are available from the authors on request.
© 2014 Australian Agricultural and Resource Economics Society Inc.
Productivity growth in the broadacre agriculture 403

Table 2 Average annual rates of growth in TFP and efficiency (%)
States 1990–2000 1990–2007 1990–2011 2000–2007 2007–2011
TFP TFP* TFPE TFP TFP* TFPE TFP TFP* TFPE TFP TFP* TFPE TFP TFP* TFPE
NSW 0.38 3.78 3.40 0.96 2.86 1.90 0.64 1.62 0.98 1.79 1.55 0.24 0.71 3.64 2.93
VIC 2.44 3.78 1.34 2.01 2.86 0.85 1.52 1.62 0.10 1.38 1.55 0.16 0.54 3.64 3.10
QLD 1.89 3.78 1.89 1.83 2.86 1.03 1.71 1.62 0.09 1.75 1.55 0.20 1.18 3.64 4.82
SA 3.67 3.78 0.10 1.67 2.86 1.19 1.82 1.62 0.19 1.20 1.55 2.74 2.44 3.64 6.08
WA 3.78 3.78 0.00 2.86 2.86 0.00 1.62 1.62 0.00 1.55 1.55 0.00 3.64 3.64 0.00
TAS 2.26 3.78 1.52 3.24 2.86 0.38 0.88 1.62 0.75 4.64 1.55 3.09 9.17 3.64 5.54
AUS 2.40 3.78 1.37 2.09 2.86 0.76 1.36 1.62 0.26 1.65 1.55 0.11 1.74 3.64 1.90
Note Annual TFP indexes for each state are reported in Table A1.
© 2014 Australian Agricultural and Resource Economics Society Inc.
404 F.U. Khan et al.