Distributional and consumptive water demand impacts of different types of economic growth in two northern Australian river catchments.
Stoeckl, Natalie ; Esparon, Michelle ; Farr, Marina 等
1. INTRODUCTION
Australia is the driest inhabited continent on earth--boasting the
lowest average rainfall, stream flow and run-off (Preston, 2009; State
of the Environment Advisory Council, 1996). On average, only 12 percent
of the nation's precipitation enters the rivers, although this
varies from less than 3 percent to almost 24 percent in drier and wetter
areas respectively (Australian Bureau of Statistics, 2003a). Moreover,
many of the country's water resources are being used at, or have
already exceeded sustainable extraction rates (Roberts et al., 2006).
Northern rivers and groundwater systems are estimated to contain
roughly 70 percent of Australia's fresh water resources (Land and
Water Australia, 2005), and it is in these regions that the majority (65
percent) of run-off occurs (Australian State of the Environment
Committee, 2006; Chartres and Williams, 2006). In comparison, the south,
which comprises most of the large urban centers and agricultural
activities, receives a meager 6.1 percent of the country's run-off
(Chartres and Williams, 2006).
Nevertheless, very little perennial water exists in the north
(CSIRO 2009a & b). At least part of the reason for this is because
rainfall in this part of the world is both highly seasonal and highly
variable. Australian river systems are the most flow variable in the
world and in the North this is largely due to the fact that many areas
receive no rain at all for 6-9 months each year during the winter dry
(Kennard et al 2010). Few northern rivers flow all year round, most are
but dry, sandy creek beds for extensive periods each year,
flooding--sometimes extensively--during the wet (Kennard et al., 2010).
That Australia's largest Europeans settlements emerged where they
did (i.e. predominantly in the south-east corner where perennial water
exists) is not a mere accident of fate.
Nevertheless, the temporal scarcity of water has not prevented
Indigenous owners from occupying lands in the North for thousands of
years. Neither has it prevented more recent, European, migrants from
settling in the region. Settlement has been possible at least partially
because some perennial surface waters do exist (e.g. as billabongs). But
that is an incomplete story: there are many underground aquifers
throughout Australia which offer themselves as a viable alternative to
surface water and they are often used as such (e.g. for stock, for urban
irrigation, and even for human consumption). That said, many of the
aquifers in Australia's north have been 'fully
exploited', particularly those located in the Queensland Gulf area
(Department of the Environment and Heritage, 2001). The region may be
rich in some resources, but clearly not in all. Evidently, the
"temporal and geographic scarcity of water [has served] as a
constraint to development" (Bennett, 2005, p.1) in this part of the
world.
The research presented in this paper is but one of many studies
seeking to provide information to support the sustainable use,
protection and management of Australia's tropical rivers. The
formally defined focus area--termed the Tropical Rivers (TR) region (see
Figure 1)--covers more than 1.3 million [km.sup.2] from the east side of
Cape York in Queensland to the Kimberley in Western Australia. It
includes 55 river basins, but this paper focuses on just two: the
Mitchell in Queensland and the Daly in the Northern Territory
(highlighted, Figure 1).
[FIGURE 1 OMITTED]
These catchments were chosen for intensive study for two key
reasons. First, both were in the formative stages of water policy and
planning, so a study such as this was well-timed to provide information
that might assist those involved in the planning process. Second, both
catchments were not as economically developed (in that residents had
generally lower incomes and access to fewer economic resources) as the
area in and around Darwin, but they were facing more development
pressures than other catchments in the TR region (Larson and
Alexandridis, 2009). Indeed, in their efforts to identify catchments
that were socio-economically 'similar', Larson and
Alexandridis (2009) found that:
1) The socio-economic characteristics of the Mitchell River were
similar to those of the Flinders.
2) The socio-economic characteristics of the Daly were similar to
those of the Flinders, the Ord and several of the northern Gulf
catchments.
3) Loosely speaking, there is a development 'spectrum',
with residents of the Darwin/Finiss Catchment having generally higher
incomes and access to more socioeconomic resources than residents of the
Mitchell (and Flinders) who in turn are relatively more
'developed' then residents of the Daly (and other
'similar' catchments) who in turn, are more
'developed' than other catchments in the TR region (Figure 2).
As such, at least some of the development issues confronting those
in the Mitchell and the Daly River Catchments are likely to: (a)
post-date those facing residents in and around the Darwin area; (b)
mimic those faced by other socioeconomically 'similar'
catchments; and (c) precede those in other, less developed TR
catchments. Lessons learned from these case-studies were thus deemed
likely to be useful in other regions today, and in the future.
[FIGURE 2 OMITTED]
Encompassing 53 197 [km.sup.2] and 71 471 [km.sup.2] respectively,
the Daly and the Mitchell, are relatively unique to the North, in that
both have perennial rivers--although the Daly's dry-season flows
are fed by underground aquifers whereas the Mitchell's are
sustained by relatively high rainfall in the upper reaches (an unusual
facet of northern rivers), supplemented in the late dry by discharge
from the Artesian basin (CSIRO, 2009a; 2009b). Importantly, both rivers
experience considerable seasonal variation in river flows (Kennard et
al., 2010); in the Daly, dry season flows are a very small fraction of
total annual flow (CSIRO, 2009a) and although the Mitchell is perennial,
other main water courses within that catchment cease to flow towards the
end of the dry season, or in times of drought (CSIRO, 2009b).
In terms of demographics, their relatively small populations (of
approximately 10 000 in the Daly, and 5 500 in the Mitchell), comprise
about 27.6 percent and 22.6 percent Indigenous persons, and Indigenous
populations are growing more rapidly than non-Indigenous populations
(Carson et al., 2009). In the Mitchell river catchment, the two most
important industries--in terms of employment and income--are agriculture
and government administration/defence, each sector contributing about 27
percent of the region's jobs (Larson and Alexandridis, 2009).
Ninety-five percent of land use is directed towards production from
unchanged land (predominantly grazing, but the Mareeba-Dimbulah
Irrigation Scheme also enables the upper catchment to be viable for
agriculture, horticulture and small scale cattle fattening projects).
Three percent of the Mitchell catchment has land that is still in its
natural condition and almost exclusively under conservation while land
under intensive use (including urban, mining, industrial) is minimal at
just 0.03 percent (Mitchell River Watershed Management Group, np; Larson
and Alexandridis, 2009). The predominant crops grown are sugarcane,
coffee, stone-fruit and a variety of tropical fruits (Connor et al. ,
2009). Further agricultural developments in the Mitchell catchment have
been discussed for many years, and several projects to supply water to
these developments have either already been implemented (e.g. the
construction of Lake Tinaroo, and the diversion of water from the Baron
river for agricultural developments in the upper Mitchell) or have been
deemed unsuitable (e.g. potential of installing a dam at the Pinnacles
which could have stored 158 000 ML) (Connor et al., 2009).
The Daly river catchment is heavily dependent upon the government
sector--it provides close to 48 percent of all employment (Larson and
Alexandridis, 2009). Agriculture (primarily irrigated) has been
identified as having much prospect for further development. However,
concerns have been raised: assessments of water availability have shown
that while there is room for additional growth, water availability is
likely to be a limiting factor (Daly River Management Advisory
Committee, 2009). Accordingly, those charged with managing water
resources in those catchments will have to be cognizant of the fact that
further developments will place increasing pressures on the catchments
resources. Moreover, as populations rise, these pressures may intensify.
Dependence upon the government sector is common in this region.
Indeed across northern Australia, three sectors which include: (i)
government administration and defence; (ii) Health and (iii) Education
are responsible for more than 25 percent of employment in
Australia's north (Stoeckl and Stanley, 2007). Those seeking to
become less dependent upon the government, rightfully, look towards
developing industries that capitalize on the region's comparative
advantage: namely, abundant natural resources. As such, there is much
interest in fostering the growth of agricultural, mining, and tourism
enterprises.
The key problem here, however, is that all of these industries use
and rely on the region's water resources. Of all these industries,
agriculture has been identified as having vital importance to the future
economic development of the region (Connor et al., 2009; Daly River
Management Advisory Committee, 2009; Stoeckl et al., 2011). The Northern
Australia Land and Water Taskforce (2009) notes that the sector could
double in size within the next 15 years. However, there has been limited
research into the use of water by this, or indeed any other industry in
Tropical Australia. Several studies have looked at water use and demand
by households (Australian Bureau of Statistics, 2010; Loh and Coghlan,
2003; Turner et al., 2005) and industries (Australian Bureau of
Statistics, 2010; Economics Consulting Services, 2004; Khan et al.,
2010). Some studies have explored the potential for sustainable use of
water in northern Australia and its tropical rivers (Northern Australia
Land and Water Taskforce, 2009; Stoeckl et al., 2006) and some have even
briefly looked at sectoral water use (Connor et al., 2009; Daly River
Management Advisory Committee, 2009). However, none have examined water
use by households in this region, and although the Australian Bureau of
Statistics (ABS) Water Account (ABS, 2001) reports on the sectoral water
use at the state level, similar information is not available at a finer
geographic scale in Australia's north. As such there is very
limited information about the potential 'consequences' on
water resources--conceptualized, here, as potential increases in the
demand for scarce water resources--of the expansion of any (or all) of
these key industries.
Moreover, the northern part of Australia contains a significant
number of Indigenous people (~one-third compared to just two percent
nationally --Carson et al., 2009); a group of people who are at a
significant socioeconomic disadvantage (Hunter, 1999; Banks, 2007,
Australian Institute of Health and Welfare 2010). However, to the best
of our knowledge, no study has compared the financial
'benefits' accruing to Indigenous and non-Indigenous people
(loosely interpreted here to be those associated with employment and
income) of different types economic growth with at least some of the
environmental 'costs' of that growth (e.g. increases in
consumptive water demand).
This paper thus seeks to at least partially fill those information
gaps. Specifically, it describes the way in which an extensive array of
primary and secondary data, were compiled in a manner that facilitated
the construction of water-use-input-output (WIO) models for both the
Daly River (NT), and the Mitchell River (QLD) catchments of northern
Australia. It then presents results from several simulations that look
at the way in which the expansion of different types of industries
affect (a) the incomes and employment of Indigenous and non-Indigenous
households and (b) consumptive water demand in each catchment. The
analysis thus provides insights into some of the distributional (i.e.
Indigenous versus non-Indigenous) and environmental (specifically,
changes in water demand) consequences of different types of economic
growth--insights which could potentially be used by a wide range of
government departments, NGO's and private enterprises when
attempting to assess the desirability, or otherwise, of development
proposals in Northern Australia.
Following this introduction (section 1) the paper is structured as
follows: section 2 describes the way in which the WIO models were built,
whilst section 3 presents the results of our simulations. Section 4
offers some concluding remarks.
2. THE MODELS
Although Australia is host to many world-class general equilibrium
models that could, theoretically at least, be extended to include
water-use variables, none provide information at a fine geographic scale
in the North. The economic structure of remote northern economies
differs, sometimes substantially, from that of urban and/or regional
centres (Stoeckl and Stanley 2007). In addition a clustering analysis
undertaken by Larson and Alexandridis (2009) suggests that there are
significant socio-economic differences between the Daly, the Mitchell
and Darwin. So information produced from any of the currently available
models that describe more urbanised economies (e.g. Darwin in the
Northern Territory) is unlikely to be relevant to those living in our
key focal catchments: regionally relevant models are clearly required.
One option is to build a regionally specific "Green"
computable general equilibrium (CGE) model. Unfortunately it can be
extremely costly, in terms of both time and money, to develop such
models. For example, the ORANI-NT model (based upon ORANI--a widely used
Australian model developed by Peter Dixon in the 1970s--cited in Breece
et al., 1994), comprised more than 7983 variables, in 3249 equations
(Knapman et al., 1991) and the Monash model (which used ORANI as its
base) took nine years to develop. The time frame associated with this
project precluded that as an option, but it did NOT rule out the option
of developing an IO model. Furthermore, CGE's use IO tables as
their base. In fact, most of the CGE models that are in existence today,
started 'life' as simple IO models; they were subsequently
refined and embellished upon over the course of time. It was thus
decided to build an IO model, reasoning that it could be refined and/or
'embellished' in future projects, perhaps subsequently
transforming it into a genuine CGE that could consider price effects,
alternative technologies and other more complex issues.
Conceptualisation
IO models are based on transactions tables which describe the
economic structure of an economy. Set out in matrix format, the columns
of the table show how a particular industry spends its money, whilst the
rows indicate where an industry sells its output. Each element
[x.sub.ij] shows how much industry j (the column) spends with industry i
(the row). By adding all elements in a column, the total expenditure of
a particular industry j can be estimated. Looked at the other way, each
element of each row [x.sub.ij] shows how much industry i (the row) earns
from (or sells to) industry j (the column). By adding all the elements
of a row, the total value of sales for a particular industry i can be
estimated. By definition, total expenditure (which includes provisions
for profits) equals total income (sales). Hence, for any given industry,
the sum of its column equals the sum of its row.
In matrix algebra:
(Ax) + (f) = (x) (1)
Where:
A is a block matrix of direct input coefficients
f is a vector of final demands
x is a vector of sectoral outputs
As such, final demand can be characterised (F) as:
(f) = (x) - (Ax) = (I - A) x) (2)
Where:
I is the identity matrix
Hence, estimates of the total change in final demands that would
occur in response to a change in demand for the final output of just one
sector can be generated as follows:
[DELTA](f) = (1 - A) [DELTA] (x) (3)
Which means that the total regional change in output ([DELTA]x)
that occurs as a result of the change in final demand ([DELTA]f) can be
calculated as:
[DELTA](x) = [(1-A).sup.-1] [DELTA] (f) (4)
Where:
[(1-A).sup.-1] is often referred to as the Leontief (inverse)
matrix
However, if the results of IO analysis are to be used to draw
inferences about the population in general, an assumption that each
sector within the model is essentially homogenous needs to be made.
There is clear evidence to suggest that this is not the case for
Indigenous and Non-Indigenous communities. Indeed, as highlighted by
Altman (2001), the economic structure of Indigenous communities is quite
different from that of Non-Indigenous communities. As such, Indigenous
and Non-Indigenous householders are not expected to have similar earning
and spending behaviors. This leads to questions regarding the efficacy
of models which fail to differentiate between the groups--particularly
models in regions like these, where Indigenous people comprise close to
25 percent of the population.
Fortunately, there are numerous techniques for adapting traditional
IO analysis to suit a variety of different circumstances, and the one
which is most pertinent in this instance is Miyazawa's extended
[IO] framework. Miyazawa's model allows analysis of the structure
of income distributions, by endogenising consumption demands in the
standard Leontief model (Miyazawa, 1976). Conceptually, this is
equivalent to the idea of 'enlarging' the matrix of technical
coefficients described above, to include coefficients that describe the
earning and consumption patterns of different types of households.
More formally, the model can be depicted by re-writing Equation 1:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (5)
Where:
x is a vector of output
y is a vector of total income for the different household groups
(Indigenous and Non-Indigenous, in this instance)
A is a block matrix of direct input coefficients
V is a matrix of value-added ratios for the different household
groups
C is a corresponding matrix of consumption coefficients for the
household groups
f is a vector of final demands--except for household consumption
g is a vector of exogenous income for the household groups
Solving this system yields:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (6)
Where:
B = [(I-A).sup.-1] is the Leontief matrix
BC is a matrix of production induced by endogenous consumption
VB(= VxB) is a matrix of endogenous income earned from production
L = VBC is a matrix of expenditures from endogenous income
K = [(1-L).sup.-1] is a matrix of the Miyazawa inter-relational
income multipliers
Researchers involved in this project, thus used this approach,
since it allowed them to explicitly consider the effect on both industry
and household incomes (Indigenous and Non-Indigenous) of changes in
final demand. In addition the models were populated with data collected
from a variety of different sources.
First, during 2006, a large-scale survey of the purchasing and
import behaviours of almost 1000 private businesses and government
organisations located across Australia's far north was conducted.
This information was used in the manner described by Stoeckl (2007;
2011) to construct the matrix of coefficients for each catchment.
Next, ABS census data on the sector of employment for Indigenous
and non-Indigenous workers was used to supplement the matrix of
coefficients--adding a value-added matrix. Specifically, collection
districts that lay either partially, or entirely within each focal
catchment were identified, and specialized tables were ordered from the
ABS, detailing the number of Indigenous and non-Indigenous people
employed in each sector ([E.sup.I.sub.j] and [E.sup.NI.sub.j]) as well
as the median incomes obtained ([Y.sup.I.sub.j] and [Y.sup.NI.sub.j]).
An estimate of the total annual income going to each household group in
each sector was then generated by multiplying the number of employees,
by the weekly median income, by 52, and then that information was used
to calculate the share of total income going to each household type from
each sector ([S.sup.I.sub.j] and [S.sup.NI.sub.j]). Estimates of the
proportion of total sectoral income paid to Indigenous (and
non-Indigenous) households in the form of wages within each
industry/sector, j,--i.e. the elements of the value-added matrix--were
then obtained by multiplying [S.sup.I.sub.j] (and [S.sup.NI.sub.j]) with
the corresponding technical coefficient (from the preceding section).
Finally, during 2009, a large-scale survey of the purchasing and
import behaviours of households in the Mitchell and Daly River
Catchments was conducted, so that a matrix of consumption coefficients
could be added to the other matrices. The 318 mail-out surveys received
from residents of the Mitchell River Catchment provided information
about the expenditure patterns of 775 people, covering approximately 18
percent of the population of non-Indigenous people and almost 31 percent
of all Indigenous people in that catchment. In the Daly River,
information was collected from 219 householders, covering approximately
6.42 percent and 8.70 percent, respectively, of the non-Indigenous and
Indigenous population in this catchment (NB: our estimates for
Indigenous people are likely to overstate the true representativeness of
the sample since ABS Census counts tend to underestimate the actual
number of Indigenous residents There are significant problems with the
quality of data relating to Indigenous people (Australian Human Rights
Commission, 2008). For a good discussion of these issues, see
http://www.hreoc.gov.au/social justice/statistics/index.html).
Incorporating Water Use
The ABS publishes data on the national and state-wide water use of
sectors within the economy (which, for the most part, coincide with the
ANZSIC sectors + the Household sector). These data clearly show that
some sectors - for example the Agricultural sector--are higher
'consumers' of water than other sectors, say Retail, or
Household. However, these figures do not give a complete story. To see
why, note that some households use water to grow their own fruit and
vegetables. But many household do not--instead choosing to purchase
their fruit and vegetables from a store. While these households are not
direct consumers of water for vegetable gardens, they are, nonetheless,
indirect consumers of water for this purpose. So if only the direct uses
of water (like those reported in the ABS accounts) are considered, some
important pieces of information will be omitted. Fortunately, IO models
allow both these types of water uses (direct and indirect) to be taken
into account.
As noted above, Equation 6 can be used to calculate the total
regional change in output (and household incomes) that occurs as a
result of the change in final demand. In a similar vein, it is possible
to calculate both the direct and the indirect changes to water demand
(AW) that are likely to occur in response to a change in final demand by
multiplying the TOTAL change in regional output by a vector that
describes sectoral (direct) water use (w):
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (7)
Where:
w is a vector of direct sectoral water use requirements (w' is
the transpose of w),
W is a vector of total sectoral water use requirements
There is clear guidance on methods for incorporating water-use into
input-output (IO) models (e.g. Kondo, 2005; O'Doherty and Tol,
2007). Guan and Hubacek (2008), for example, provide a very good
framework for considering both water consumption and water availability
within an IO model and, closer to home, Lenzen and Foran (2001) have
published an IO analysis of Australian water usage. However, there is no
regionally specific information available on the water use of different
industries and households for the catchments considered here. Thus for
this project, water use coefficients had to be estimated 'from
scratch'. This was carried out differently for households and
industry, as described below.
The water-use IO (WIO) model requires data on the water use per $
of output for each sector. Thus it was decided that estimates of these
coefficients would be generated by dividing ABS estimates of total water
consumption per industry within Queensland and the Northern Territory,
by the associated Gross value added (GVA) for each sector (see Table 1).
For consistency (i.e. to match water use data), we used 2000-01
estimates of GVA for these calculations.
Evidently, water use per dollar of income varies greatly by
industry sector. There are also differences across states--an
observation that is likely to be at least partially attributable to
different climatic and rainfall conditions and also partially
attributable to differences in agricultural practices since different
types of agriculture have vastly different water-use requirements (see,
for example, Lenzen and Foran 2001). As such, it is clear that it cannot
simply be assumed that the water-use coefficients which apply to
Queensland as a whole will apply to the Mitchell, nor that those which
apply to the Northern Territory as a whole will apply to the Daly. For
quite legitimate reasons, water use coefficients will vary over time,
and in response to a wide range of external drivers such as climate,
policy, and technology. Therefore when conducting simulations, the
state-wide water-use vectors from Table 1 were chosen to define a
'plausible' minimum and maximum water-use coefficient for each
sector, within each catchment. In most cases, the minimum water-use
coefficient was that of the Northern Territory estimates, the three
exceptions being for the Retail, Transport and government sectors.
These, same, minimum and maximum coefficients were used in both WIO
models.
Household water use data was collected in conjunction with the
household expenditure data in the 2009 survey described above. When
collecting data, the researchers involved in this study were cognizant
of the fact that few respondents would be able to provide precise
information about the water used by their household each year
(particularly those without water meters). So a series of questions was
designed to elicit information about the extent to which various
water-using appliances were used, as illustrated in Figure 3.
This information was combined with information about the average
water used by a range of different appliances compiled from the
Melbourne's Household Water Use Calculator, Water Wise and Brisbane
Water (Melbourne City Council, 2003; Waterwise Brisbane, 2008) to
generate an estimate of total household water consumption. For example,
if the respondent indicated that their washing machine was a
front-loader and that they did approximately 3 loads of washing each
week, then researchers were able to conclude that the household used
approximately 300 litres per week of water for washing (3 x 100 litres).
This information was combined with other information about the number of
people living in the house and the type (and use) of other appliances to
generate an estimate of the total water used per household per week
inside the home. Consequently, estimates of inside water use, are a
function of (a) the number of householders; (b) the number of
water-saving appliances; and (c) the use of water and water-saving
appliances.
Householders were also asked about their water usage outside the
home which differentiated between wet season and dry season use, and
responses to this question were then combined with information about
internal water use to generate an estimate of the total quantity of
water used per week by each household during the wet and the dry season.
Table 2 shows data on household water consumption in the Mitchell
and Daly catchments. It is in the order of 200-260 litres per person per
day during the wet season (with most water consumption for internal
household use). In the dry season, this increases to between 370 and 790
litres per person--the extra consumption largely due to the extra water
used outside the house (in the garden, for the swimming pool, etc).
These estimates seem 'plausible' in so much as our lower,
wet-season estimates roughly accord with household water consumption
figures from the ABS's Water Accounts for Australia's south
east--where rainfall has a more even temporal dispersion than
Australia's north and may entice fewer householders to use
significant quantities of water outside (e.g. Victorian water
consumption was approximately 220 litres per person per day in 2001).
Our upper estimates of household water consumption relate to the dry
season in a hot climate (the Daly) and exceed the ABS's estimates
of the average estimate of household water consumption in the Northern
Territory (420 litres per person per day). This is not surprising, since
the ABS's figures are a 'whole of year' estimate; it
would be expected that dry-season consumption exceed that of the wet
season.
Interestingly, in the Mitchell River Catchment daily inside water
use is higher in Indigenous households than in non-Indigenous households
and most of the 'excess' is attributable to the use of water
for showers. One possible explanation for this difference can be found
in the qualitative information collected during interviews: Indigenous
householders generally earn much less than non-Indigenous householders
(quantifiably verifiable) and are thus not wealthy enough to pay large
electricity bills--instead some choose to shower many times a day during
the hot summer months as a way of keeping cool (in lieu of
air-conditioning).
Recognising that household water demand is every bit as likely to
vary across a range of factors as industry water demand, researchers
used data from Table 2 to generate a range of per-person water
consumption estimates:
Minimum annual water consumption = (Total daily water use during
the wet) * 365
Maximum annual water consumption = (Total daily water use during
the dry) * 365
Dividing these minimum and maximum estimates of household water
consumption by per-person income, thus allowed researchers to estimate
minimum and maximum water-use coefficients for each individual. This
information was then grouped by Indigeneity, and averaged, to generate
appropriate lower and upper bound estimates of household water-use
coefficients for use in the WIO model--in line with the minimum and
maximum estimates derived for industry.
Allowing for Employment
Just as it is possible to define a vector of direct sectoral
water-use requirements from which the total water requirements of a
change in final demand can be calculated, so too is it possible to do
this for employment. Specifically, it is possible to define a direct
vector of sectoral employment requirements (e) which can be used to
estimate the total change in employment ([DELTA]E) likely to arise in
response to change in final demand:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (8)
Where:
e is a vector of direct sectoral employment requirements, (e'
is the transpose of e),
E is a vector of total sectoral employment requirements
This general approach was used here, although researchers
distinguished between Indigenous and Non-Indigenous employment, thus
working with a matrix of employment requirements, rather than a vector
(as is done with water).
When populating the vector with data, state-wide data was used to
generate an estimate of the average number of employees per dollar
earned within each sector for each state (specifically, they divided the
total number of employees within each sector by each sector's
GVA)--since both employment and GVA data was not available at the
catchment level. These estimates were then converted into estimates of
the number of Indigenous and Non-Indigenous employees per dollar of
output using data supplied by the ABS to apportion the total number of
employees per $M of GVA across household types. For example, the number
of Indigenous employees per dollar of GVA in the Agricultural sector in
the Daly was calculated as:
[No of Indigenous employees working in the Agricultural
sector/Total no of employees working in the Agricultural sector] x
Employees per $ GVA in the NT
To ensure that employment and GVA data all related to the same
period, 2006 output data were used.
A Preliminary Caution about the Interpretation of Results
When IO tables are used to estimate the impact of an increase in
demand in one sector, it is implicitly assumed that the extra revenues
received by that sector will be distributed according to the current,
observed (average) expenditure patterns.
From the perspective of a householder, using observed expenditure
patterns to predict changes in expenditure that may result from changes
in income is tantamount to assuming that the marginal propensity to
consume (MPC) is equal to the average propensity to consume (APC).
Ceteris paribus, if consumption (C) is a linear function of income (Y),
comprised of both an autonomous (CA) and an induced component that
increases with income (CI), then the higher is the MPC and/or the
smaller is CA relative to Y, the closer will the APC be to the MPC, and
the more 'palatable' will be the assumptions underlying IO
analysis.
From the perspective of businesses, this is equivalent to assuming
that inputs are always used in fixed proportions (i.e. Leontief
technologies) and that production technologies are constant across time.
IO analysis also assumes (even if only implicitly) that prices are
constant. Conceptually, it is as if these 'limitations' mean
that IO models provide information about the maximum, likely, outward
shift of a demand curve. IO analysis is unable to allow for the fact
that subsequent changes in price and/or production methods may
'erode' some of that initial impact with the economy. In other
words, IO models are demand-driven. Without supply-side information
(like that collected for full-scale CGE models), 'a supply
curve' cannot be added to the model, so IO cannot be used to make
accurate predictions about the 'final' impact of a change on
either prices or quantity. Although some argue that these limitations
mean that IO analysis is more suited to short-term analysis than to
long-term analysis, such an interpretation is not strictly correct. As
clearly argued by Wilting et al., (2004), valid long-term projections
can still be produced with IO, providing that (a) exogenous changes (the
development 'scenarios') are being modeled, and (b) a
reference base is used--e.g. comparing the likely change in incomes
after 2006 from growth scenario A with the changes from scenario B.
3. SIMULATION RESULTS
Establishing a Base-Line
The first step of the modeling exercise required researchers to
establish a starting-point (or base year) for key variables. In all
cases, this was assumed to be 2006, since that is the year during which
most of the data that populate these models were collected.
The 2006 ABS census data referred to above were used to estimate
total employment in each industry/sector for each catchment
(differentiated by Indigeneity). In each sector, estimates of income
(GVA) were generated by multiplying ABS state-level estimates of $GVA
per employee by ABS census estimates of the number of employees within
that sector in that catchment (also differentiated by Indigeneity).
Survey data were used to estimate baseline aggregate household income
(household income = average per-person income x estimated resident
population). We did this because the ABS income data only provided
information about the income which householders earn from industry, and
may, therefore, have omitted income from other sources.
The upper and lower-bound industry water use coefficients were
multiplied by estimates of GVA during 2006 to generate upper and
lower-bound estimates of the total amount of water consumed by each
industry/sector within each catchment. For householders, upper and lower
bound estimates of total annual water use were generated by multiplying
daily dry-season (upper-bound) and daily wet-season (lower-bound)
estimates of per-person water use (see Table 6) by 365 and by estimates
of the total population for each household type (i.e. Indigenous and
non-Indigenous). These estimates are presented in Table 3.
The Department of Natural Resources, Environment, Arts and Sport
(2009, p.13) reports that in 2006/07 an estimated 1 085 ML of water was
used from the Tindall Aquifer for 'public water supply'; an
additional 12 456 GL was used for agriculture (including Horticulture),
with 1 195 GL used for industry, and 1 128GL used for rural stock and
domestic purposes. At close to 16GL in total, this is higher than our
lower bound estimates of water use (9.7GL) and just under one-half of
our upper bound estimates of water use (34.2GL) for the entire Daly
Catchment during 2006. As such, it seems that the actual quantity of
water used in this catchment during 2006 is likely to be between the
upper and lower bound estimates presented here. Evidently, the figures
presented in Table 3 may not be 'precise', but they are, at
least 'plausible'. Baseline water consumption was thus taken
as the mid-point of water use from Table 3--amounting to a total (across
all industry and household sectors) of 22GL and 36GL per annum in the
Daly and the Mitchell, respectively.
The Growth Scenarios
The 2009-10 budget forecasted economic growth of approximately 1.5
percent over the 2010-11 financial year (Commonwealth of Australia,
2009). So in the first instance, a 1.5 percent growth scenario across
all industries was performed. This was followed by other scenarios
investigating a 5 percent growth in agriculture, mining and tourism
respectively--the three main industries thought to offer prospects for
development. The 5% growth rate was chosen because it is in line with
the Northern Australia Land and Water Taskforce's (2009)
observations that Agricultural production could double within the next
15 years. The models were used to calculate results for one year, and
then extrapolated for the next 20 years.
Figure 3 and Figure 4 depict the potential impact of each of the
respective growth scenarios in each of the catchments. In each chart
there are separate lines showing the projected growth in Indigenous and
Non-Indigenous incomes and employment from the base year (2006). Each
chart also shows the projected increase in water demand--although there
are two 'water demand' lines on each chart: one derived from
the low water use coefficients, and one from the high water use
coefficients. These lines can thus be loosely interpreted as showing a
range of water demand estimates--that range dependent upon the
water-using habits of each community. From these charts, several
observations can be made:
[FIGURE 4a OMITTED]
[FIGURE 4b OMITTED]
[FIGURE 5a OMITTED]
[FIGURE 5b OMITTED]
(a) 1.5 percent growth per annum across all industries
The 'balanced' growth scenario (of 1.5 percent per annum
across all industries) significantly out-performed all other scenarios
for employment and income in the Daly. It was one of the top two
generators of income and employment in the Mitchell (alongside the 5
percent growth in agriculture scenario). Within 20 years, this scenario
increased industry income and non-Indigenous employment to levels that
were close to 1.6 times greater than in 2006. Indigenous employment
outcomes were more modest--rising to between 1.4 and 1.5 times the 2006
levels. This balanced growth scenario was also associated with moderate
increases in consumptive water demand--rising to between 1.2 and 1.7
times 2006 levels depending upon whether lower or upper bound estimates
were used.
(b) 5 percent growth in agriculture
In the Mitchell River, growth in the agricultural sector generated
substantial increases in business/industry incomes and in non-Indigenous
employment. Outcomes for Indigenous people were much more modest. If
growth in agriculture is achieved using water-efficient techniques
('mimicked' here, with the lower-bound water use
coefficients), then in 2026, our models predict that consumptive water
demand would be just 1.6 times greater than 2006 levels; but consumptive
water demand could be more than double 2006 levels in less than a decade
if higher water-use coefficients prevail.
Income and employment outcomes associated with the agricultural
scenario were more modest in the Daly than in the Mitchell, but
pressures on consumptive water demand were similar in both catchments.
Outcomes for Indigenous people (incomes and employment) were also very
modest in both regions--rising by less than 10 percent, in total, over a
20 year period.
(c) 5 percent growth in Tourism
The tourism scenario delivered the smallest 'returns' to
income and employment for both Indigenous and non-Indigenous households,
in both catchments. This is a consequence of the fact that tourism
currently makes a relatively small contribution to these economies (just
3 and 2.3 percent of the Mitchell and Daly River's GVA,
respectively). Consequently, 5 percent growth in tourism represents a
very small increase in economic activity (5 percent of 3 percent).
(d) 5 percent growth in Mining
The mining scenario delivered marginally better household income
and employment outcomes to both Indigenous and non-Indigenous households
than did the tourism scenario, but the returns were still quite small.
In contrast, the associated increases in industry output/incomes were
relatively good and even out-performed those of the agricultural
scenario in the Daly River. The predicted increases in consumptive water
demand were similar for the mining and tourism scenarios. However, care
must be taken when interpreting data relating to the Mining and
Manufacturing sector: as noted by the NLAW taskforce (2009, p 23)
"mining and resource projects are generally excluded from water
resource accounting, exact water use estimates for this industry are not
readily available". Consequently, the estimates presented here may
understate perhaps substantially--consumptive water demand in the mining
sector.
4. CONCLUDING COMMENTS
Some of the significant water problems confronting those in the
south of Australia have served to increase development pressures on
those in the North, perhaps at least partially because the region is
perceived to be relatively water abundant. Not only is that perception
incorrect (taking into account the fact that few of Australia's
Tropical rivers are perennial), but it is possible that water may serve
to constrain development in this region in the not-too-distant future.
Our research clearly highlights that in both the Mitchell and the
Daly river catchments--like elsewhere in Australia--it is the
agricultural sector that uses most water. Moreover, our simulations
indicate that if agriculture were to grow at 5 percent per annum (i.e.
if the sector were to double in 15 years), and if the water use
coefficients that applied in Queensland during 2001 were to prevail,
then total consumptive water demand would double in less than 10 years
in both the Daly and the Mitchell River Catchments. This is clearly of
concern because consumptive water demand cannot grow indefinitely:
sooner or later water will 'run out'.
CSIRO (2009b) notes that in the Mitchell, current average water
uses amount to less than 1% of total annual flows. But without storage,
total annual flows are not the relevant factor to consider; it is the
availability of water during the dry season that serves as the binding
constraint. This has been recognized as a significant issue in the Daly
(CSIRO, 2009a), and may also affect parts of the Mitchell (although in
this region there are significant knowledge gaps surrounding groundwater
storages and recharge options--as noted by CSIRO, 2009b). Evidently,
unless more efficient ways of using water are adopted, dry-season flows
may soon start to constrain development in some northern regions. Our
simulations serve to highlight the importance of water-saving
technologies and research, particularly in the agricultural sector.
Moreover, our scenarios also highlight the fact that development
does not benefit all equally. Our simulations clearly show that most
forms of development serve to generate larger absolute increases in
incomes and employment for non-Indigenous people than for Indigenous
people. As such these types of development will widen, rather than
'close' the 'gap' (unless there are changes to the
underlying structure of these economies).
Evidently, development in northern Australia does not just involve
potential tradeoffs between income and environment (water); equity
issues abound.
ACKNOWLEDGEMENTS: The research described in this paper is one of
the outcomes of several projects that were funded by JCU, the Tropical
Savannas CRC, and the Tropical Rivers and Coastal Knowledge (TRaCK)
Commonwealth Environmental Research Hub. TRaCK received major funding
for its research through the Australian Government's Commonwealth
Environment Research Facilities initiative; the Australian
Government's Raising National Water Standards Program; Land and
Water Australia; the Fisheries Research and Development Corporation and
the Queensland Government's Smart State Innovation Fund.
Researchers also gratefully acknowledge and appreciate the contribution
to this research made by Mitchell River Catchment Traditional Owners
(The Olgol, the Yir Yoront, The Western Gugu Yalanji; The Mulliridgee;
The Barbarum, The Kuku Djunkan and Gugu Mini, and by the Daly River
Catchment Traditional Owners living in Kybrook Farm & Pine Creek and
Nauiyu Nambiyu (Daly River). A special thank you is due to The community
of Kowanyama, particularly:
* Viv Sinnamon--Kowanyama Aboriginal Land and Natural Resource
Management Office
* Michael Yam-Shire--Councillor, Former chair of Mitchell River
Watershed Management Group, Member Mitchell River Traditional Custodians
Advisory Group and Olgol TO
* Anzac Frank--Kowanyama Aboriginal Ranger and Yir Yoront TO
* Ravin Greenwool--Kowanyama Aboriginal Ranger: Cultural Heritage
and Kin Kopol/Yir Yoront TO
* Phillip Mango and Stanley Budby--Kowanyama Aboriginal Rangers
The Mitchell River Traditional Custodian Advisory Group,
particularly
* Ruth Link--Chair and Western Gugu Yalanji TO
* Gerry Turpin--Treasurer and Mbabaram TO
* The people of Western Gugu Yalanji, Kuku Juungan, Mbabaram and
Wokomin.
The Daly River Aboriginal Reference Group (ARG), particularly:
* Mona Liddy
* Valemina White
The TO's who worked with us so diligently when collecting
data:
* Darren Birchley--Kowanyama Aboriginal Ranger and Kokoberra TO
* Sharon Brady--Western Gugu Yalanji TO
* John Grainer--Kuku Djungan TO
* Eddie Turpin--Mbabaram TO
* Eddie Thomas--Wokomin TO
* Agnes Page
* Kathleen Perry
* Bridget Kikitin; and
* Lizzie Sullivan
The Tropical Savannas CRC--for their in-kind data contribution.
Finally, we wish to extend our sincere appreciation to the hundreds
of anonymous householders who took the time and effort to complete our
survey--without such input, the project could not have gone ahead.
REFERENCES
Altman, J. C. (2001). Sustainable development options on Aboriginal
land: The hybrid economy in the twenty-first century, CAEPR Discussion
Paper No. 226/2001. Canberra: Centre for Aboriginal Economic Policy
Research.
Australian Bureau of Statistics (2003a). Year Book--Environment
Special Article: Australia's Rivers. Online version accessed 18
April 2011, http://www.abs.gov.au/Ausstats/abs@.nsf/0/a301124cdaa05be4ca2 56cae001599c5?OpenDocument
Australian Bureau of Statistics (2001). Water Account, Australia,
2000-2001 (ABS Catalogue No. 4610.0). Canberra: Australian Bureau of
Statistics.
Australian Bureau of Statistics (2010). Water Account, Australia,
2008-09 (ABS Catalogue No.4610.0). Canberra.
Australian Institute of Health and Welfare (AIHW). (2010).
Australia's health 2010. Australia's health series no. 12.
Cat. No. AUS 122. AIHW, Canberra, Australia.
Australian State of the Environment Committee (2006). Australia
State of the Environment 2006: Independent report to the Australian
Government Minister for the Environment and Heritage. Canberra:
Department of the Environment and Heritage.
Banks, G. (2007). Overcoming Indigenous disadvantage in Australia.
Proceedings of the Second OECD World Forum on "Statistics,
Knowledge and Policy", Measuring and Fostering the Progress of
Societies, Istanbul, Turkey, 27-30 June 2007.
Bennett, J. (2005). The evolution of markets for water: Theory and
Practice in Australia, Edward Elgar, UK.
Breece, J., Mclaren, K., Murphy, C. and Powell, A. (1994). Using
the Murphy Model to provide shot-run macroeconomic closure for ORANI.
Economic Record, 70(210), pp. 292-314.
Carson, D., Taylor, A. and Campbell, S. (2009). Demographic trends
and likely futures for Australia's Tropical Rivers. Darwin: School
for Social and Policy Research, Charles Darwin University.
Chartres, C. and Williams, J. (2006). Can Australia overcome its
water scarcity problems? Journal of Developments in Sustainable
Agriculture, 1, pp. 17-24.
Commonwealth of Australia (2009). Budget 2009-10: Part 2--Economic
Outlook. Online version accessed 19 July 2010,
<http://www.budget.gov.au/2009-10/content/myefo/html/part_2.htm>
Connor, S., Sokolich, B., Hoogwerf, T., Mackenzie, J. and Butler,
J. (2009). Mitchell river catchment: Regional perspective. Online
version accessed 30 Aug 2011,
http://www.nalwt.gov.au/files/Chapter_26-Mitchell_River_catchment_regional_perspective.pdf
CSIRO (2009a). Water in the Daly region, pp 273-361. In CSIRO
(2009) Water in the Timor Sea Drainage Division. A report to the
Autralian Government from the CSIRO Northern Australia Sustainable
Yields Project. CSIRO Water for a Healthy Country Flagship, Australia.
xi + 508pp.
CSIRO (2009b). Water in the Mitchell region, pp 347-416. In CSIRO
(2009) Water in the Gulf of Carpentaria Drainage Division. A report to
the Autralian Government from the CSIRO Northern Australia Sustainable
Yields Project. CSIRO Water for a Healthy Country Flagship, Australia.
xi + 479pp.
Daly River Management Advisory Committee (2009). Daly river
catchment: Regional perspective. Online version accessed 30 Aug
2011, http: //www.nalwt.gov.au/files/Chapter_27-Daly_River_catchment_regional_perspective.pdf
Department of the Environment and Heritage 2001, Australia State of
the Environment Report, CSIRO Publishing, Canberra.
Department of Natural Resources, Environment, The Arts and Sport,
(2009). Water allocation plan for the Tindall Limestone Aquifer,
Katherine 2009-2019. Online version accessed 26 February 2011,
http://www.nt.gov.au/nreta/water/kwac/pdf/final wap.pdf
Economics Consulting Services (2004). Water and the Western
Australian minerals and energy industry: Certainty of supply for future
growth. Applecross WA: Economics Consulting Services.
Guan, D. and Hubacek, K. (2008). A new and integrated
hydro-economic accounting and analytical framework for water resources:
A case study for North China. Journal of Environmental Management, 88,
pp. 1300-1313.
Hunter, B. (1999). Three nations, not one: Indigenous and other
Australian poverty. Canberra: Centre for Aboriginal Economic Policy
Research, The Australian National University.
Kennard, M., Pusey, B., Olden, J., Mackay, S., Stein, J. and Marsh,
N. (2010). Classification of natural flow regimes in Australia to
support environmental flow management. Freshwater Biology, 55,
pp.171-193.
Khan, S., Abbas, A., Rana, T. and Carroll, J. (2010). Dairy water
use in Australian dairy farms: Past trends and future prospects: CSIRO:
Water for a Healthy Country National Research Flagship.
Knapman, B., Stanley, O. and Lea, J. (1991). Tourism and gold in
Kakadu: The impact of current and potential natural resource use on the
Northern Territory economy. Darwin: Northern Australia Research Unit,
Australian National University.
Kondo, K. (2005). Economic analysis of water resources in Japan:
Using factor decomposition analysis based on input-output tables.
Environmental Economics and Policy Studies, 7, pp. 109-129.
Land and Water Australia (2005). The Tropical Rivers Program
Canberra: Land and Water Australia.
Larson, S. and K. Alexandridis. (2009). Socio-economic profiling of
tropical rivers. Townsville: CSIRO Sustainable Ecosystems.
Larson, S., Stoeckl, N. and Blanco-Martin, B. (2013) Use of
socioeconomic typologies for improved integrated management of data-poor
regions: explorations from the Australian north., Australasian Journal
of Environmental Management.
http://dx.doi.org/10.1080/14486563.2012.763145
Lenzen, M. and Foran, B. (2001). An input-output analysis of
Australian water usage. Water Policy, 3, pp. 321-340.
Loh, M. and Coghlan, P. (2003). Domestic water use study: In Perth,
Western Australia 1998-2001. Perth: Water Corporation.
Melbourne City Council (2003). Household water use calculator.
Online version accessed 14 December 2008,
http://www.melbourne.vic.gov.au/Environment/SavingWater/Documents /water
household calculator.pdf
Mitchell River Watershed Management Group (np). About the
catchment. Online version accessed 28 April 2011,
http://www.mitchell-river.com.au/about_the_catchment.html
Miyazawa, K. (1976). Input-Output analysis and the structure of
income distribution. New York: Springer-Verlag.
Northern Australia Land and Water Taskforce (2009). Sustainable
development in Northern Australia. Canberra: Northern Australia Land and
Water Taskforce.
O'Doherty, J. and Tol, R. (2007). An Environmental
Input-Output Model for Ireland. The Economic and Social Review, 38(2),
pp. 157-190.
Preston, B. (2009). Water and ecologically sustainable development
in the courts. MqJICEL 6, pp. 129-146.
Roberts, R., Mitchell, N. and Douglas, J. (2006). Water and
Australia's Future Economic Growth. Economic Round-Up, Online
version accessed 18 April, 2011,
http://www.treasury.gov.au/documents/1087/PDF/05_Water.pdf
State of the Environment Advisory Council (1996). Australia: State
of the Environment 1996. Online version accessed 18 April 2011,
http://www.environment.gov.au/soe/1996/publications/report/pubs/
ex-summary.pdf
Stoeckl, N. (2007). Using Surveys of Business Expenditure to Draw
Inferences about the size of Regional Multipliers: A Case-study of
Tourism in Northern Australia. Regional Studies, 41(7), pp. 917-931.
Stoeckl, N. and Stanley, O. (2007). Key Industries in
Australia's Tropical Savanna. Australasian Journal of Regional
Studies, 13(3), 255-286.
Stoeckl, N. (2011). Comparing multipliers from survey and
non-survey based IO models: An empirical investigation from Australia.
International Regional Science Review,--published online before print
December 30, 2010, doi: 10.1177/0160017610385452.
Stoeckl, N., Esparon, M., Stanley, O., Farr, M., Delisle, A., and
Altai, Z., (2011) "Socio-Economic Activity and Water Use in
Australia's Tropical Rivers: A case Study in the Mitchell and Daly
River Catchments", Final report on project 3.1, submitted to the
Tropical Rivers and Coastal Knowledge Commonwealth Environmental
Research Facility, March 2011, 108 pp, available at:
http://www.track.gov.au/publications/registry/905
Stoeckl, N., Stanley, O., Brown, V., Jackson, S. and Straton, A.
(2006). An assessment of social and economic values of Australia's
tropical rivers. Towsnville: CSIRO & James Cook University.
Turner, A., White, S., Beatty, K. and Gregory, A. (2005). Results
of the largest residential demand management program in Australia.
Proceedings of the International Conference on the Efficient Use and
Management of Urban Water, Santiago, Chile, 15-17 March 2005.
Waterwise Brisbane (2008). Water efficiency calculator. Online
version accessed 14 December 2008, http://www .watercorporation.com
.au/_includes/calculator/calculator.asp
Wilting, H. C., Faber, A. and Idenburg, A. M. (2004). Exploring
Technology Scenarios with an Input-Output Model. Proceedings of the the
International Conference on Input-Output and General Equilibrium: Data,
Modelling and Policy Analysis.
Natalie Stoeckl
School of Business and Cairns Institute, James Cook University
(JCU), 4811, Australia. Email: Natalie. Stoeckl@jcu.edu.au
Michelle Esparon
School of Business and School of Earth and Environmental Sciences,
James Cook University, 4811, Australia.
Marina Farr
School of Business, James Cook University, 4811, Australia.
Aurelie Delisle
School of Business and School of Earth and Environmental Sciences,
James Cook University, 4811, Australia.
Owen Stanley
School of Business, James Cook University, 4811, Australia and
Research Institute for the Environment and Livelihoods, Charles Darwin
University
Table 1. Litres of Water Consumed Per
$GVA--by Sector and State (2001).
Sector Queensland Northern
Territory
Agriculture 664.52 185.20
Mining 15.54 3.90
Electricity 111.18 54.28
Construction 0.13 0.07
Retail 1.60 3.54
Accommodation 2.76 1.63
Transport 1.89 3.50
Finance 0.86 0.59
Government 1.73 9.14
Culture 21.55 6.32
Source: author calculations using ABS data.
Table 2. Average Litres of Water Used Per Person Per Day--by
Catchment and Indigenous Status.
Type of use Daly River Catchment
Indigenous non-Indigenous
General Water Use 20.00 20.00
Wash Water Use 24.49 32.65
Dishwater Use 11.46 14.67
Shower Water Use 104.90 68.96
Toilet Water Use 50.63 50.61
Leaking Toilet Water Use 27.50 14.26
Leaking Taps Water Use 84.78 125.17
Bath Water Use 3.99 4.10
Total Inside Water Use 152.53 183.11
Outside water use during the dry 285.33 604.07
Outside water use during the wet 51.01 75.94
Total daily water use during the dry 437.86 786.05
Total daily water use during the wet 203.54 259.05
Type of use Mitchell River Catchment
Indigenous non-Indigenous
General Water Use 20.00 20.00
Wash Water Use 29.93 33.31
Dishwater Use 16.12 18.74
Shower Water Use 133.82 62.87
Toilet Water Use 50.16 48.43
Leaking Toilet Water Use 13.29 19.90
Leaking Taps Water Use 93.00 118.61
Bath Water Use 3.87 1.59
Total Inside Water Use 215.49 165.11
Outside water use during the dry 156.59 393.55
Outside water use during the wet 9.23 66.39
Total daily water use during the dry 372.08 558.10
Total daily water use during the wet 224.72 230.94
Source: survey data collected by authors
Table 3. Estimated Total Water Consumption by Sector and by
Catchment (ML, 2006).
Sector Daly River Catchment Mitchell River Catchment
Lower Upper Lower Upper
bound bound bound bound
Agriculture 7 794 223 27 966 140 15 145 25 619
Mining 381 239 1 518 556 14 680 611 52 674 915
Electricity 168 394 344 933 1158 2239
Construction 2296 4441 44 783 152 651
Retail 54 282 120 261 258 951 530 429
Accommodation 17 507 29 616 12 350 17 996
Transport 45 110 83 766 106 316 561 417
Finance 27 361 39 867 163 742 652 220
Government 261 983 1 383 447 38 382 85 036
Culture 65 325 222 671 25 339 47 053
Indigenous 205 046 441 090 101 544 168 132
Households
Non-Indigenous 684 566 2 077 216 359 257 868 197
Households
Total 9 707 332 34 232 004 15 807 578 55 785 904
Source: author calculations using ABS data.
Figure 3. Excerpt from the Questionnaire.
What type of washing machine do you have? Please tick appropriate
box. If you do not have a washing machine, but use a Laundromat
instead, then please tell us about the type of washing machine at
the Laundromat
[] We do not have a washing machine--and wash our clothes by hand.
(Please go to question 9)
[] Twin Tub [] Front Loader [] Top Loader
How many times per week does your household use a washing machine?
Please tick appropriate box. If you do your washing at a
Laundromat, please tell us how many times you use a washing machine
at the Laundromat
[] We rarely use a washing machine (or do not have one)
[] Once a week [] 4 times a week [] 7 times a week (approx
once a day)
[] Twice a week [] 5 times a week [] 14 times a week (approx
twice a day)
[] 3 times a week [] 6 times a week [] More than 3 times a day
