The adoption of sustainable practices: Some new insights. An analysis of drivers and constraints for the adoption of sustainable practices derived from research

Full text

The adoption of sustainable
practices: Some new insights
An analysis of drivers and constraints for the
adoption of sustainable practices derived from
research
July 2001
John Cary, Trevor Webb1 and Neil Barr2
1Social Sciences Centre
Bureau of Rural Sciences
Canberra
2Department of Natural Resources and Environment
Victoria

Land & Water Australia Project Reference Number:BRR19
Principal Investigator:
Ass/Prof John Cary
Social Sciences Centre
Bureau of Rural Sciences
PO Box E11
KINGSTON ACT 2604
Collaborators:
Dr Trevor Webb
Social Sciences Centre
Bureau of Rural Sciences
PO Box E11
KINGSTON ACT 2604
Dr Neil Barr
Department of Natural Resources and Environment
BENDIGO VIC 3550
Preferred way to cite this report:
Cary, J.W., Webb, T. and Barr N.F. (2001) The adoption of
sustainable practices: Some new insights. An analysis of
drivers and constraints for the adoption of sustainable
practices derived from research. Land & Water Australia,
Canberra.
This report does not represent professional advice given by
the Commonwealth or any person acting for the
Commonwealth for any particular purpose. It should not be
relied on as the basis for any decision to take action or not
to take action on any matter which it covers. Readers
should make their own further enquiries, and obtain
professional advice where appropriate, before making any
such decision.
The Commonwealth and all persons acting for the
Commonwealth in preparing this booklet disclaim all
responsibility and liability to any person arising directly or
indirectly from any person taking or not taking action based
upon the information in this booklet.

iii
Contents
CONTENTS III
LIST OF TABLES V
LIST OF FIGURES V
EXECUTIVE SUMMARY 1
INTRODUCTION 3
Project objectives 3
Background 3
The principal issues for sustainable practice adoption 4
SUSTAINABLE PRACTICES 5
Sustainable resource management outcomes 6
A FRAMEWORK FOR APPRAISING SUSTAINABLE PRACTICES 7
FORGOTTEN FOCUS – ATTRIBUTES OF SUSTAINABLE PRACTICES 10
How landholders see NRM practices – the key issues 10
The attributes of sustainable agriculture practices 10
Categorising NRM practices 12
A case example: phase farming with dryland lucerne 16
LEARNING ABOUT SUSTAINABLE PRACTICES 19
Categorising the learning focus 19
Reasons for learning 19
Styles of learning 20
SOME RECENT AUSTRALIAN FINDINGS ON FACTORS ASSOCIATED WITH THE ADOPTION OF
SUSTAINABLE PRACTICES 22
Age 24
Education 25
Property size 25
Farm business 25
Stewardship 25
Recent findings regarding attitudes 26
Other characteristics 26
MODELING ADOPTION BEHAVIOUR FROM THE 1998-99 RESOURCE MANAGEMENT SURVEY 28
Introduction 28
Modelling farmer behaviour 28
Findings 30
Summary 34
ATTITUDES AND VALUES AND THE ADOPTION OF SUSTAINABLE PRACTICES 35
Attitudes 35
Beliefs 35
Values 35
A framework of environmental concern 36
Values and the appraisal of sustainable management practices 37
INTERVENTIONS TO PROMOTE ADOPTION OF NRM PRACTICES 39
Consequences for adoption of sustainable practices 40
IMPLICATIONS FOR THE FOCUS OF R & D 41
PERFORMANCE INDICATORS AND COMMUNICATION ACTION PLAN 43
Performance indicators for assessing the effectiveness of adoption of R&D results 43
Communication action plan 44

iv
REFERENCES 45
APPENDIX A ANALYSIS OF THE 1998-99 RESOURCE MANAGEMENT SURVEY 49
Logistic regression 49
Model estimation 49
Results 49
APPENDIX B DESCRIPTION OF VARIABLES USED IN LOGISTIC REGRESSION ANALYSES 63

v
List of tables
Table 1 Characteristics of sustainable practices..................................................................................................................................... 13
Table 2 Solutions’ National Indicators .....................................................................................................................................................24
Table 3 Landholder and property characteristics with significant relationships with adoption of best practices .................................27
Table 4 Variables explored in analysis of Resource Management Supplementary survey..................................................................29
Table 5 Resource management practices investigated.......................................................................................................................... 30
Table 6 Characteristics significantly associated with practice adoption ................................................................................................31
Table 7 Factors which are associated with the adoption of sustainable management practices (shaded cells indicate association
relationships in predicted direction).................................................................................................................................................. 32
Table A1 Logit regression results for the adoption of controlled flow bores in the pastoral zone. ....................................................... 50
Table A2 Logit regression results for the control of grazing pressure by excluding access to water in the pastoral zone.................51
Table A3 Logit regression results for the adoption of monitoring pasture and vegetation condition in the pastoral zone..................52
Table A4 Logit regression results for the adoption of deep rooted perennial pasture in the wheat-sheep and high rainfall zones
(broadacre industries only)............................................................................................................................................................... 53
Table A5 Logit regression results for the adoption of soil/plant tests to determine fertiliser needs in the wheat-sheep and high
rainfall zones (broadacre industries only)........................................................................................................................................ 54
Table A6 Logit regression results for the establishment of trees and shrubs in the wheat-sheep and high rainfall zones (including
dairy industries).................................................................................................................................................................................55
Table A7 Logit regression results for the regular monitoring of watertables in the wheat-sheep and high rainfall zones (including
dairy industries).................................................................................................................................................................................55
Table A8 Logit regression results for the collection of dairy effluent (dairy industry only).................................................................... 56
Table A9 Logit regression results for the pumping of dairy shed effluent onto pasture (dairy industry only)...................................... 57
Table A10 Logit regression results for laser graded layout on irrigated farms...................................................................................... 57
Table A11 Logit regression results for the use of irrigation scheduling tools on irrigated farms.......................................................... 58
Table A12 Logit regression results for monitoring of pasture and vegetation condition (all farms). ....................................................59
Table A13 Logit regression results for preservation or enhancement of areas of conservation value (all farms)............................... 60
Table A14 Logit regression results for the exclusion of stock from degraded areas (all farms)........................................................... 61
Table A15 Logit regression results for the percentage of the farm under conservation tillage (all farms)........................................... 61
List of figures
Figure 1 Model of Adoption of Sustainable Land Management Practices [Modified from Fenton, Macgregor & Cary (2000)]............ 7
Figure 2 Area of mixed lucerne pasture and adoption of mixed lucerne pasture in the North Central catchment region of Victoria:
1996-99 (Source: ABS)..................................................................................................................................................................... 18
Figure 3 A framework of environmental concern (after Stern et al. 1995). ...........................................................................................36
Figure 4 Conditions for maximum influence of environmental values or attitudes on individual’s decision to adopt sustainable
practices ............................................................................................................................................................................................38

1
Executive Summary
Effective R&D intervention means designing practices to provide external benefits to
make environment-sustaining behaviour more likely. It is the inherent
characteristics, or attributes, of practices derived from research which largely
determine their rate of adoption by producers. Adoption of recommended
sustainable practices depends largely on whether landholders think they are
profitable. Sustainable NRM practices which provide economic and other
advantages will be adopted more rapidly. Recent low commodity prices in the
broadacre industries reduced the attractiveness of adoption of many practices.
Landholders generally seek to reduce the risk of adopting a new practice.
Sustainable NRM practices which are observable, trialable, and less complex are
generally more quickly adopted than NRM practices which are unobservable,
untrialable, and complex.
Humans are adaptive in implementing NRM practices rather than simply reactive to
information, promotional appeals or exhortations to farm sustainably. The use of
sustainable practices will depend on how landholders assess the value of
recommended practices and their own and others’ experience with use of such
practices. Characteristics of the practices – and their overwhelming influence on
adoption – often confound the influence of other factors such as social
characteristics.
It is often difficult for landholders to see the connection between recommended
NRM practices and sustainability. The difficulty, for landholders, of observing
linkages between many recommended NRM practices and desired sustainable
outcomes reduces positive appraisals of NRM practices by landholders. As a result,
they are often lukewarm about NRM practices that are promoted predominantly on
the basis of making land use more sustainable.
There are obvious advantages in being able to promote sustainable practices with
more universal or global applicability. However, given Australia’s diverse
environment, there are few sustainable practices that meet the test of global
applicability. Universally applicable practices are often less likely to have large
impacts on reducing local land degradation problems.
Increased effort needs to be applied to identify and develop locally applicable
sustainable practices and effort made to resist the temptation to promote them
beyond localities where their advantage has been established.
It is more effective, in the first instance, to look for sustainable practices with
characteristics influencing more rapid adoption behavior rather than depending on
pro-environmental values of landholders or on individual feelings, preferences, and
perceptions for improving the land environment.
Factors related to landholder characteristics that potentially influence capacity to
change are: level of farm income, landholder age, landholder participation in
training, having a documented farm plan and membership of landcare. There are
often interactions between these characteristics; and the relationships with adoption
behaviour are not always unequivocal. Personal financial capacity has been
observed to be an important component in determining the capacity of landholders
to adopt new practices. Landholders’ perceptions of their future financial situation
were more often associated with practice adoption than were objectively measured
indicators of current financial position. Landholders who feel secure in their financial
The rate of adoption of sustainable
practices derived from research
depends on practices being
economically attractive to adopt.
Rational self-interest predominates in
human assessment of sustainable
p
ractices.
Linkages between eventual
sustainable outcomes and sustainable
practices are often distant and
uncertain thus reducing incentive to
act.
Think locall
y
and act locall
y
.
Select sustainable practices on the
basis of attributes which enhance
likelihood of ado
p
tion.
Social and perceptual factors
influence adoption rates of
sustainable practices.

2
future are more likely to invest resources in adopting new resource management
practices.
Pro-environmental or ‘green’ values and attitudes have a relatively minor influence
on the adoption of sustainable practices. The effect of positive attitudes towards the
environment is constrained by the influence of prevailing incentives or disincentives
to adopt a sustainable practice. Positive attitudes towards the environment act in
combination with external incentives or disincentives (such as costs, benefits,
convenience, or uncertainty of outcome of a given practice) to determine adoption
behaviour.
The effect of strongly positive environmental attitudes on sustainable practice
adoption tends to be influential when there are no strong external incentives
(rewards) or disincentives (punishments) for undertaking the practice. Positive
environmental attitudes have little effect on behaviour when external incentives are
strongly positive or negative. In such cases it is the external factors which
effectively compel or prohibit the behaviour in question. The strength of the external
conditions determines the bounds of influence of positive environmental attitudes
and values.
When assessing the characteristics of potential new sustainable practices, and
when seeking to promote the use of such practices, there are a number of human
behaviour principles that should be considered. If the ‘behaviour’ associated with
the practice cannot be readily seen (ie it is not observable) by the individual and by
others it will be ineffective to encourage it. It will be difficult to be monitored, to be
seen as rewarded (or penalised, for its absence). As most human behaviour is
undertaken to gain a positive consequence or avoid a negative consequence,
humans learn more from their successes (which provide positive reinforcement)
than they learn from their mistakes. The significant negative consequences of
unsustainable consequences will not be experienced until long into the future.
Research and development of on-farm sustainable practices needs to identify
practices with relatively immediate positive consequences rather than less
immediate, diffused, or short-term negative, consequences. Practices that have
outcomes that are ‘soon’ and ‘certain’ will have the most powerful drivers for rapid
adoption.
Pro-environmental or ‘green’ values
and attitudes have a relatively minor
influence on the adoption of
sustainable practices.
Principles which predict likely human
behaviour can assist in selecting and
promoting sustainable practices.

3
Introduction
Land & Water Australia (LWA) established the Social and
Institutional Research Program (SIRP) to ensure that
biophysical research and development (R&D) also takes
into account social and institutional factors. LWA has
identified a considerable gap in understanding the social,
economic and institutional factors that are conducive to
implementation and adoption of sustainable resource
management practices (Mobbs & Dovers 1999). The LWA
Strategic R&D Plan places greater emphasis on
researching the social, institutional and economic issues
which may be constraining the development and adoption
of more sustainable natural resource management. The
Strategic R&D Plan also seeks to identify opportunities to
create a more enabling environment for sustainable natural
resource management.
This review and analysis of the drivers of, and constraints
to, producer adoption of sustainable practices derived from
research presents current knowledge on adoption
influences, risks and processes as they relate to resource
management R&D. We develop some new insights that
should ensure more realistic assessments of the likely
response of landholders to the problems of sustainable
resource use. An important purpose of the project is to
increase the effectiveness of future research and the
adoption of research results by producers and landholders.
The framework proposed by LWA for analysis of the drivers
of, and impediments to, the adoption on R&D results
includes:
• how producers and other land holders perceive and
value their natural resources on a property level and
wider catchment and regional scales
• how they learn about, and understand, those resources
and the management of them
• how they actually manage them and incorporate new
management approaches arising from research into
their practices
• how they are influenced by the wider economic, social,
legal, commercial, policy and institutional environment.
The project embraces an integrated approach to focus on
the relationships between developing awareness, values
and understanding, acquiring required knowledge and
skills, and adopting sustainable natural resource
management (NRM) practices as part of integrated
management systems.
Project objectives
The major project objective was to review and analyse the
drivers and constraints to adoption of sustainable practices
in agriculture derived from research. More specifically the
following objectives were to be achieved:
• review existing knowledge
• identify impediments and drivers for the adoption of
sustainable practices from R&D
• provide advice and options on strategies for
overcoming impediments
• key performance indicators for assessing the
effectiveness of producer adoption of R&D
• prepare a communication action plan for the project.
Background
In 2000 the Bureau of Rural Sciences (BRS) completed a
review of factors influential in determining individual NRM
decisions (Barr & Cary 2000). That review established a
number of important findings that provided a basis for the
development of this Report:
• Encouraging the adoption of more sustainable practices
by appealing to farmers’ stewardship ethic or altruism
will have only limited impact. The presence of factors
like the relative financial benefit or cost of the NRM
practice, farm financial capacity, farmer skills and
motivation are the necessary determinants as to
whether sustainable management practices are likely
to be adopted.
• Policies aimed at changing motivation in the absence of
meeting the other enabling conditions will achieve
little.
• Responses to messages about future threats of land
degradation are likely to be limited.
• Australian research regarding how landholders
perceive their environment and the threat of land
degradation shows landholders generally
underestimate the land degradation problems on their
own farm.
This project builds on these findings to develop a
conceptual approach regarding the relationships between
factors that are relevant in any assessment of likely
adoption of sustainable practices derived from research
and producer initiated and managed R&D.

4
The principal issues for sustainable
practice adoption
The idea that there is a key to the adoption of sustainable
practices is somewhat simplistic and unidimensional. It is
implied in assertions such as “We just need to convince
them to change”. The assumption is, if sustainable
outcomes are to be achieved and appropriate sustainable
practices are available, an understanding of human
motivation will provide the touchstone to unlock human
capacity to change. In fact, land managers differ
significantly in different localities in Australia and, for many
localities, there are few appropriate sustainable practices
that meet criteria which would lead to ready adoption.
Many of the desired outcomes of sustainable NRM
programs do not come about autonomously. NRM
programs present a policy challenge because of the range
of constraints that discourage individual uptake of NRM
practices. Constraints to change in NRM systems can be
assessed from the perspectives of individual landholders,
the characteristics of desirable management practices, the
socio-economic structure of catchment communities and
the broader institutional settings.
A significant issue is that economic costs to a landholder of
at least some NRM practices (particularly those which
provide benefits desired by the wider community) may
exceed the on-farm benefits on a short or long-term basis.
The lack of immediate financial incentive in a dynamic farm
economy may result in many landholders not adopting
these practices.
Identification of the social and economic factors that
constrain the participation of individual land managers
recognises that the significant decisions about land and
farm management are made by ‘individual farmers, not by
catchment groups or regional river management bodies’
(Pannell 2001a). Understanding some, if not all, of the
factors that determine individual landholder decisions will
ensure more realistic and more effective catchment and
regional plans.

5
Sustainable practices
Sustainable land management practices are defined here
as those which ameliorate unsustainable land use by
rectifying biophysical constraints to agricultural production
and conserve the resource base (SCARM 1998). The
following list of sustainable management practices has
been developed from SCARM (1998), Hamblin (1999), SCA
(1991), management practice indicators for State of the
Environment reporting (Saunders, Margules & Hill 1998)
and other sources. It should be recognised that this is an
incomplete list of sustainable management practices. Many
of these practices were identified in the National
Collaborative Project on Indicators for Sustainable
Agriculture based on then available ABS and ABARE
statistics.1 Measures of the level of landholder adoption of
sustainable management practices available from the
current ABARE Australian Resource Management
Supplementary surveys are identified with the superscript a.
• maintenance of soil cover
• establishing and monitoring ground cover targets –
monitoring of pasture and vegetation condition a
• nutrient balance accounting (soil and plant sampling)
• soil and plant tissue tests to determine fertiliser needs a
• regular soil testing
• fertilising of pastures
• agricultural lands treated with gypsum
• agricultural lands treated with lime
• regularly monitor water tables a
• use of deep-rooted perennial pastures a
• non-commercial tree and shrub planting a
• commercial tree and shrub planting (farm forestry) a
• preserve or enhance areas of conservation value a
• retention of vegetation along drainage lines a
• protection of land from stock by fencing –
exclude stock from degraded areas a
• protection of waterways from stock by fencing a
• animal pest or weed control to control land degradation
a
• pest and disease control in pastures
1 While ABARE farm surveys provide more reliable, in-depth
information than ABS agricultural census data they are selective
in industry coverage and geographic spread.
• use of integrated pest management (reducing pesticide
use)
• slashing and burning of pastures
Cropping farms
• use of reduced or zero tillage –
minimum tillage a
• stubble or pasture retention in ploughing –
direct drilling a
• use of crop or pasture legumes in rotations a
• use of contour banks in cropland a
• strip cropping a
• adjusting crop sequences in response to seasonal
conditions
Irrigation farms
• irrigation scheduling a
• laser graded layout a
• storage and reuse of drainage water a
• automated irrigation a
Rangelands
• control grazing pressure by excluding access to water a
• control of water flow from bores a
• piped water supplies for stock a
• pastoral land stocked at recommended rates
• degraded pastoral land converted to less damaging use
• pastoral land destocked in low feed conditions
Dairy farms
• use of effluent disposal systems–
collection of dairy effluent (ponds or drainage sump) a
• pump dairy shed effluent onto pasture a
Many of these farming practices are specific to particular
environments or to particular farming systems. The SCA
(1991) report identified the potential relevance of many of
these practices for the sustainable management for 46
agro-ecological regions of Australia.
Not all the NRM practices listed above will, in isolation, lead
to sustainable resource management (for example,
fertilising of pastures). What might be sustainable on a farm
might be unsustainable for rivers etc. Hence, the practices
which effectively contribute to sustainability will depend on
the context and the locality of their use. If one farmer

6
adopts a ‘sustainable’ practice, it could be totally ineffective
if neighboring landholders do not adopt complementary
practices.
Sustainable resource management
outcomes
The use of the sustainable practices listed above are
contended to lead to more sustainable resource
management. The association is often constrained – it is
likely to vary for different localities. The impact of use of a
practice may also have long time lags before a more
sustainable outcome is achieved. Broader conceptions of
sustainable management embrace the need for strategies
for sustaining both food security and the need to conserve
natural resources.
Definitions of sustainable resource management in
agriculture are generally concerned with the need for
agricultural practices to be economically viable, to meet
human needs for food, to be environmentally benign or
positive, and to be concerned with quality of life. Since
these objectives can be achieved in a number of different
ways, sustainable resource management is unlikely to be
linked to any particular management practice. Rather,
sustainable agriculture is thought of in terms of its
adaptability and flexibility over time to respond to the
demands for food and fibre, its demands on natural
resources for production, and its ability to protect the soil,
water and other natural resources. This goal requires an
efficient use of technology in a manner conducive to
sustainability (Wilson & Tyrchniewicz 1995). Because
agriculture is affected by changes in markets and resource
decisions in other sectors and regions, such changes often
provide additional pressures leading to depletion of local
agricultural resource bases.
Assessments of the sustainability of a production system
involve looking forward, to a future that is often not
universally agreed. It is often easier to look backward, and
assess the progress of production systems as they evolve
from unsustainable states. The process is further
complicated because a sustainable state of resource
management is not a fixed or ideal steady state, but rather
an evolutionary process of attempting to improve the
management of systems, through improved understanding
and knowledge. The process is not deterministic as the
end point is not known in advance (Wilkinson & Cary 2001).
Sustainable resource management is often an abstract
state – which occurs in the future and may be hard to
identify or measure. Sustainable practices (which lead to
sustainable states or outcomes) are used as ‘indicators’ or
proxies for sustainable management.

7
A framework for appraising sustainable practices
The development of a model of conceptual relationships
helps to understand the place of sustainable practices that
lead to more sustainable outcomes. Additionally it helps to
focus on important factors influencing how landholders, or
other decision-makers, might perceive these relationships.
Any attempt to derive a complete predictive model which
encompasses all possible environmental, behavioural,
social and economic indicators, and which identifies the
inter-relationships amongst possible relevant variables
would be counter-productive and lead to confusion given
the current state of knowledge and research in this area.
Furthermore, given the heterogeneity of resource
management situations, the development and application of
a generalised predictive model at a national scale, which is
meaningful in relation to all farming practices, would require
more extensive knowledge and data than is currently
available.
The model in Figure 1 is a conceptual model rather than a
predictive model. Figure 1 shows some broad groups of
factors that influence the adoption of NRM practices that
are proposed to bring about more sustainable land
management. The characteristics of locality and
environment, and the characteristics of specific adoption
practices, which are both extremely significant in landholder
appraisal of NRM practices are specifically identified. The
model also shows that there is usually more than one NRM
practice that needs to be embraced to bring about more
sustainable land management. Institutional characteristics
incorporate the more formal structures that determine the
‘social’ environment in which landholders decide or
anticipate decisions regarding adoption of sustainable
practices. Institutional characteristics incorporate the
regulatory environment, government agency support
structures, and government policy reflected in incentive
schemes and taxation arrangements. Individual and social
characteristics include many factors such as age and
education and cognitive factors that are largely instilled and
maintained through social processes. No attempt is made
here to further elaborate all the elements that might
comprise individual and social characteristics.
Figure 1 Model of Adoption of Sustainable Land Management Practices [Modified from Fenton, Macgregor & Cary
(2000)]
Locality and
Environmental
Characteristics
Characteristics
of Practice j NRM Practice j
Appraisal
Adoption of
Recommended
NRM Practice 1
Sustainable
Land
Management
NRM Practice n
Institutional
Characteristics
Individual and
Social
Characteristics
PROCESSES OUTPUTS OUTCOMES

8
The model emphasises that adoption of sustainable NRM
practices is not uni-dimensional, consisting of a potentially
wide range of practices that are dependent upon appraisals
by landholders. These appraisals are mediated by
environmental, institutional, individual and social factors
prior to any implementation.
Central to the model presented in Figure 1 is the appraisal
process undertaken by the individual adopter or group of
adopters. Appraisal often involves a complicated
psychological calculus by an individual to arrive at a
decision. Appraisal has the elements of a ‘black box’ – it
may be objectively difficult to know the relative influence of
the factors that may determine an adoption or non-adoption
decision. Social factors such as land manager attitudes
and beliefs about specific NRM practices, and about
broader natural resource management, will influence
adoption of specific practices. Appraisal will also be
influenced by land manager attitudes towards those
organisations and institutions that may be promoting
sustainable land management practices.
Differences in appraisal are determined by a range of
individual, institutional and contextual variables and by
complex interactions amongst these variables. For
example, negative attitudes towards the land, as in the
belief that the land is ‘rubbish country’, is a component of
appraisal which includes beliefs about land and land
management which may act as a barrier to the adoption of
sustainable land management practices. The existence of
this belief may be due to specific individual characteristics,
historical relationships between the farmer and those
agencies promoting sustainable land management and
specific environmental and locality characteristics in which
farming occurs. To date there is limited understanding or
research on the appraisal component and its relationship to
the adoption of sustainable land management practices
and, because appraisal is a complex process, there are no
existing indicators of appraisal.
Human appraisal as an adaptive system
Adaptation in biological usage is the process by which an
organism fits itself to its environment. Complex adaptive
systems are systems comprised of interacting agents who
change their ‘rules of behaviour’ as their experience
accumulates (Holland 1995). In a complex adaptive system
a major part of the environment of any given adaptive agent
(in this case landholders) consists not only of the
biophysical environment but of other adaptive agents
including institutions (Figure 1). The focus of an adaptive
system is on improvement rather than optimisation (or the
attainment of some equilibrium).
The other focus of an adaptive system in areas such as
evolutionary systems theory is the idea of an iterating
process of stimulus and response (or learning). Human
behaviour is characterised by continuous human learning
and complex responses to stimuli that rarely produce
observable constancy. This is because most human
behaviour occurs in environments where humans interact
and respond – by actively changing environmental states –
rather than simply reacting to them. This process can be
described as reflexivity.
Reflexivity emphasises the uncertainties involved in
seeking to achieve more sustainable resource management
(uncertainties which are generally inadequately
acknowledged). Soros (2000), who has applied the
concept of reflexivity to explain the failure of equilibrium
theory in describing human behaviour in financial markets,
provides a simple description of reflexivity:
. . . our understanding of the world in which we
live is inherently imperfect. We are part of a
world we seek to understand, and our
imperfect understanding plays an active role in
shaping the events in which we participate.
There is a two-way interaction between our
understanding and these events that
introduces an element of uncertainty into both.
It ensures that we cannot base our decisions
on (perfect) knowledge and that our actions are
liable to have unintended consequences. The
two effects feed on each other. I call this two-
way feedback mechanism reflexivity . . .(p. xxii)
Another way of describing reflexivity is that thinking
participants seek to understand the situation in which they
participate and, as well, they participate in the situation that
they seek to understand (Soros 2000, p. 7).
Figure 1 is an example of an adaptive behavioural system
which has been kept as simple as possible. Solid lines
indicate more certain associations; broken lines indicate
associations about which less is likely to be known, or
where the association may be problematic or intermittent.
Single arrows indicate a likely one-way or recursive
relationship; double arrows indicate a likely two-way or non-
recursive (reflexive) relationship.
There are important reflexive, or feedback, loops between
appraisal and the adoption of given sustainable practices.
Landholders assess such practices for potential adoption
and any adoption of a practice either by the landholder, or
by others who’s experience can be observed by the
landholder, will influence how the appraising landholder
subsequently views (appraises) the adoption of that
practice and related practices. More importantly, there are
few feedback loops between ultimate states of sustainable
land management and NRM practices because there are
usually long time lags from the implementation of an NRM
practice to the outcome of a ‘sustainable state’. Thus
landholders cannot be readily assured with reasonable
feedback from their own observations that a practice
produces a desired ultimate outcome.
The process of appraisal deliberately subsumes the
complex and differing human motivations that may
influence NRM behaviour. The ability to choose one’s
motivations distinguishes humans from other animals. As a

9
consequence there can be no certainty about human
motivations (Soros 2000).
Sustainable outcomes and NRM outputs
The problem of the usually long time lag from the
implementation of an NRM practice to the outcome of a
sustainable state can be thought of as being represented
by different levels of abstraction – the desired sustainable
state is more difficult to observe (and to measure). This
‘end’ state can be considered an outcome. The means of
approaching the end state is easier to observe (and to
measure) and can be labeled an output. This
acknowledges that states can be observed (and measured)
at different levels of abstraction. Usually, the less abstract
the state the easier is its measurement. Typically,
outcomes will be represented by biophysical and ecological
attributes that characterise sustainable systems. Outputs,
such as appropriate vegetation cover, are posited to lead to
desired outcomes. NRM practices are typically directed to
producing outputs that subsequently lead to desired
outcomes (Figure 1).
Additionally, we can identify processes which (often more
tenuously) contribute to outcomes. Processes include
behaviours (eg participation in landcare) that contribute to
desired outcomes, and also include attitudes and social
learning, which are clearly social in nature. It should be
noted that the distinctions between these categories may
not always be clear-cut.
Understanding some, if not all, of the factors that determine
individual landholder decisions will ensure more realistic
and more effective catchment and regional plans. In the
next section we consider the factors which influence
landholder capacity to change to more sustainable
resource management practices.

10
Forgotten focus – attributes of sustainable practices
Sustainable practices have not been tried and
found wanting, rather
many have been found difficult and not tried.
(apologies to G.K. Chesterton)
How landholders see NRM practices –
the key issues
In order to understand the key influences determining
whether sustainable land management practices are
adopted it is necessary to understand both the nature of
NRM practices and, more particularly, how landholders see
particular NRM practices.
Normally, adoption of a given sustainable practice is
determined to a large degree by a landholder’s perceived
self-interest. Profitability of a practice is an important
element of self-interest, even for practices intended to
improve land and resource conservation (Cary & Wilkinson
1997; Riley 1999; Drake, Bergstrom & Svedsater 1999;
Marsh & Pannell 2000; Curtis et al. 2000). For different
localities a particular natural resource management
practice varies in terms of its relative profitability and
appropriateness for a given farm situation. In other words,
a given practice will have different profitability and differing
attractiveness to farmers in different regions or localities
(Barr & Cary 2000). This will largely reflect different
technical, soil, climatic endowments and, probably, the
level of land degradation in different localities (Cary 2000).2
But it may also reflect that a management technology (such
as a modified deep-rooted perennial) may be developed
and elaborated for one area but not for another. Many
practices to ameliorate salinity, for example, are not
universally applicable and hence have different profitability
in different localities.
Additionally, the economic environment for a given farm
activity (which influences on-farm implementation of
recommended management practices) will in turn be
influenced by both local conditions (drought or good
seasons) and the external marketplace (expressed in
product and commodity prices). Many broadacre farm
businesses do not produce sufficient surpluses to allow for
reasonable living standards, investments in the farm
2 Perversly, the value of some sustainable management practices
is likely to be greater in situations where other factor
endowments are high and likely to be less in situations of land
degradation where one or other factor endowments are likely to
be low.
business and investment in resource protection and the
environment. In some regions current adjustment patterns
are only slowly creating aggregated businesses more
capable of generating appropriate surpluses.
The attributes of sustainable
agriculture practices
Rogers has summarised the results of the many adoption
and diffusion studies conducted in the 1950s, 60s and 70s
(Rogers 1962; Rogers & Shoemaker 1971; Rogers 1983).
The general conclusions provide a means of analysing
environmental innovations and exploring the reasons for
the difficulties of promoting certain forms of sustainable
agriculture. The importance of innovation characteristics
was highlighted in major review of innovation adoption in
Australian agriculture by Guerin and Guerin (1994).
Important attributes influencing the rate of adoption of NRM
practices are the relative advantage, the complexity, the
compatibility, the trialability and the observability of a given
practice (see Barr & Cary 2000). These attributes together
with two other attributes – locality differentials in relative
advantage and risk characteristics of a practice – are
considered below.
Relative Advantage
Relative advantage is normally interpreted in terms of
financial advantage to the farm business or the adopter.
The perceived financial advantages of environmental
innovations (where they exist) have consistently been
shown to be one of the best indicators of their subsequent
adoption. There is little evidence to suggest that
sustainable practices are any different to other agricultural
practices in this respect. The nature of limited interaction of
pro-environmental attitudes or stewardship values
overriding, or compensating for, deficiencies in relative
financial advantage of an NRM practice will be developed
later in this review.
In a review of the history of environmental innovations on
Australian farms, Barr and Cary (1992) concluded that the
clear lesson was that environmental innovations that were
believed to be profitable were usually readily adopted.
Innovations with a net financial cost were rarely adopted.
The most studied adoption of an environmental innovation
is the progress of conservation cropping on the US corn
belt. In a review of Ohio research Carboni and Napier
(1993) concluded economic factors were the greatest
predictors of adoption.

11
Locality differentials in relative
advantage
Frequently it is assumed that the relative advantage of an
environment-enhancing practice, if positive, is of the same
order of magnitude in different localities. Generally, this is
unlikely to be the case. While little empirical evidence for
improved resource management practices has yet been
collected in Australia to support this common sense
assumption, the early work of Griliches (1957, 1960) on the
diffusion of the productive innovation of hybrid corn is
clearly indicative. Griliches contended that the differences
in rates of adoption of hybrid corn for different American
states were largely explained by the relative advantage
possessed by different geographic regions for growing
corn. This reflected productivity of soils, consequential
differential profitability of the crop, and differential
possession of harvesting and handling resources. As a
consequence, hybrid corn was ‘an innovation which was
more profitable in the “good” areas than in the “poor” areas’
(Griliches 1960, p. 280). These geographical differences in
relative advantage, and consequent differences in rates of
adoption, when expressed as diffusion curves (for the same
‘innovation’) have different shapes and, more importantly,
different slopes or rates of diffusion (see Cary 2000; Barr &
Cary 2000).
The important conclusion for the adoption of NRM practices
is that the appropriateness and relative advantage of given
NRM practices will vary in geographic space to a very large
extent.
Risk
The motivation of human behaviour is more complex than
being simply profit driven. While there is much research
demonstrating relationships between beliefs about
profitability and adoption behaviour this is mediated by a
great variation in attitudes towards business profit and a
consideration of the risks that characterise much Australian
agriculture. There is strong evidence that many Australian
farmers are motivated by the balance between the need for
profit and a satisfaction with a comfortable living which
minimises risk (Dunn, Gray & Phillips forthcoming; Rendell,
O’Callaghan & Clark 1996). Different attitudes to income
needs, risk perception, dynastic expectations and cultural
expectations of farming mean there are quite distinct
groups of farmers. Many farms trade off profit
maximisation for risk reduction (Howden et al. 1997; Marks
& O’Keefe 1996; Reeve & Black 1993). For many farm
operators relative advantage may be strongly moderated by
minimisation of complexity and minimisation of risk. As a
consequence the differing risk implications of different
sustainable practices will be an important consideration in
their adoption.
Complexity
Sometimes innovations which appear simple may in fact
imply significant and complex changes to the farm
production system. Such innovations are less likely to be
adopted. Complexity increases the risk of failure; and it
introduces increased costs in gaining knowledge (Vanclay
& Lawrence 1995).
Integrated pest management is an innovation that is
constrained by the management complexity of its practise.
Farmers often explain non-adoption of integrated pest
management as being based upon concerns about its ease
of use, speed and reliability (Bodnaruk & Frank 1997).
Another example of this complexity characteristic is the
planting of dryland lucerne. This is promoted in many
catchment plans across Australia as a means of reducing
watertable recharge. What appears to be a simple change
to a system can imply major restructuring of the farm
system. The complexity of adopting the sustainable
practice of dryland lucerne and phase farming is explored
as a case example later in this section.
Compatibility
Compatibility refers to the extent to which a new idea fits in
with existing knowledge and existing social practice. If a
new idea fits easily into an existing system it will be
adopted more quickly. There are usually two ‘systems’
against which the compatibility of a practice will be judged –
the current system of farming on a given property and the
social system embracing a farming community or broader
cultural beliefs and values. An apparent example of a
sustainability innovation failing this test can be seen in the
low adoption of perennial pasture sowing amongst a
substantial core of wool producers in the Western District of
Victoria. Pasture renovation in this region can be profitable
if combined with an increase in stocking rate. Local culture
has held that higher stocking rates are incompatible with
the region’s reputation as a producer of fine wool. This
opposition is documented as early as the 1920s when
subterranean clover was first promoted in the district (Barr
& Cary 1992). These beliefs are now complemented by
beliefs that improved perennial pastures and higher
stocking rates are ecologically unsustainable (Marks &
O’Keefe 1996). The promotion of pasture improvement has
often been incompatible with the values of this cultural
group.
For many broadacre farmers beliefs about ‘good farming’
tend to encompass matters such as tidiness, having fences
and gates well maintained and having good looking crops
or stock (Nassauer 1995). Profitability and sustainable
farming practices are less commonly seen as being
indicative of good farming (Dunn, Gray & Phillips
forthcoming; Phillips forthcoming; Wilkinson 1996;
Wilkinson & Cary 1992). While these cultural values may
be causing increasing frustration in industry bodies and the
agribusiness sector (Clancy 1999), there is evidence that
Australian agriculture is undergoing a period of
detraditionalisation in which traditional agricultural
occupational identities are being replaced by more complex
and diverse cultures (Bryant 1999; Dunn, Gray & Phillips
forthcoming). Current research gives little indication of the

12
impact of detraditionalisation upon changes in farm
management practice.
Trialability
Innovations which can be trialed on a small scale prior to
full implementation are more likely to be adopted. Trialing
enables decisions about the utility of an innovation with
minimal risk. Typically, farmers can easily assess a new
crop variety by sowing one paddock to the new variety
before deciding upon more extensive adoption. The
successful promotion of conservation cropping practices
which is dependent upon major machinery changes has
been encouraged by providing hire trash combines, thus
allowing trialing without significant investment in machinery.
In contrast, dryland salinity control is clearly not amenable
to trialing. Because the benefits of salinity control may not
be achievable for up to 50 years, a trial process will delay
more extensive salinity control for a century. Trialability is
in turn dependent upon observability.
Observability
NRM practices whose advantages are observable are more
likely to be adopted. Traditionally, new variety or crop is
often quite visible to passing observers and this visibility
has been used to advantage. Irrigation watertable control
is not normally an observable achievement. The
development of well flags (to indicate water levels) as part
of water-table watch was an innovative method of making
watertable levels visible to the passing observer. Many
Landcare programs have attempted to locate
demonstrations along major roads to enhance visibility.
Categorising NRM practices
An inventory of recommended NRM practices is presented
below (Table 1). The management practices in this
inventory are categorised in terms of attributes that have
been found to be important in determining whether
management practices are readily adopted or not. Such an
approach provides a method for assessing likely
adoptability in given farm situations and provides a
conceptualisation and categorisation of relevant NRM
practices. The appropriateness and relative advantage of
given NRM practices will vary in geographic space to a very
large extent.
In Table 1 the sustainable practices listed above are scored
on their level of possession of the following attributes:
• Geographic applicability– refers to relative
appropriateness of a practice, in terms of whether it is
effective or adapted to only specific localities or, more
universally, across many localities.
• Relative Advantage – the financial advantage or other
convenience or personal advantage to the farm
business or the adopter.
• Risk – refers to uncertainty about likely benefits or
costs associated with a sustainable practice,
uncertainty about the effectiveness of the practice,
uncertainty as to when the benefits might be realised
and uncertainty regarding the social acceptability of
the practice.
• Complexity – implies that a practice comprises more
than one or two simple elements and that its elements
interact with each other and, in sometimes
complicated ways, with elements of the farming
system into which it is to be incorporated.
• Compatibility – the extent to which a practice fits in with
existing farm practices, or with existing knowledge or
existing social practice.
• Trialability – where practices can be implemented on a
small, or pilot, scale decisions can be more easily
made about the value of a new practice without the
risks associated with full implementation.
• Observability – practices whose impact or advantage is
easily observable, or whose outcome is quickly
realised, are more likely to be adopted.

Table 1 Characteristics of sustainable practices
Sustainable practice Geographic
applicability
Relative
advantage
Risk Complexity Compatibility Trialability Observability
(Ideal rating) (Hi) (Hi) (Lo) (Lo) (Hi) (Hi) (Hi)`
Maintenance of soil cover Hi Hi (temporal) Lo M-Hi (locality) M M M-Lo
Establishing and monitoring ground cover targets
(monitoring of pasture and vegetation condition) aHi M Lo M-Hi M M M-Lo
Nutrient balance accounting
(soil and plant sampling) Lo Lo Lo Hi M Lo Lo
Soil and plant tissue tests to determine fertiliser needsa
Lo Lo Lo Hi M Lo Lo
Regular soil testing M M Lo Lo M Lo Lo
Fertilising of pastures M Hi-M (locality) M Lo Hi Hi Hi-M
Agricultural lands treated with gypsum M Lo M-Hi Lo Hi M M
Agricultural lands treated with lime M Lo M-Hi Lo Hi M M
Regularly monitor water tables a
M M (locality) Lo Lo Lo Hi M
Use of deep-rooted perennial pastures a
Hi M M-Hi M-Hi M (locality) M Lo
Non-commercial tree and shrub planting a
M – Hi Lo Lo Lo M-Hi Hi Hi
Commercial tree and shrub planting (farm forestry) a
Lo Lo (locality) Hi M Lo Lo Hi
Preserve, enhance areas of conservation value a
M Lo Lo M Lo M M-Hi
Retention of vegetation along drainage lines a
M Lo Lo M M-Lo M M-Hi
Protection of land from stock by fencing
(exclude stock from degraded areas) aLo Lo Lo M M Hi Hi

14
Sustainable practice Geographic
applicability
Relative
advantage
Risk Complexity Compatibility Trialability Observability
(Ideal rating) (Hi) (Hi) (Lo) (Lo) (Hi) (Hi) (Hi)`
Protection of waterways from stock by fencing a
Lo Lo M-Hi Lo M Hi Hi
Animal pest or weed control to control land degradation a
Hi M M M M-Hi M M
Pest and disease control in pastures M M-Hi (locality) M M M-Hi M-Lo M
Use of integrated pest management (reducing pesticide use) Lo M-Lo M-Hi Hi M M-Lo M-Lo
Slashing and burning of pastures Lo M-Lo M Lo M Hi-M Hi
Cropping farms
Use of reduced or zero tillage (minimum tillage) a
Hi M M M M-Hi Hi M
Stubble or pasture retention in ploughing
(direct drilling) aM M M-Hi M-Hi M Hi-M M
Use of crop or pasture legumes in rotations a
Hi M-Hi M-Lo M-Lo M-Hi M M-Lo
Use of contour banks in cropland a
M M-Lo M-Lo M-Hi M-Lo M-Lo M-Hi
Strip cropping a
M
Adjusting crop sequences in response to seasonal conditions Hi M-Hi M M M-Lo M-Lo Lo
Irrigation farms
Irrigation scheduling a
M M Lo M-Hi M-Lo M-Lo Lo
Laser graded layout a
Hi M-Hi Lo-M M M-Lo M Hi
Storage and reuse of drainage water a
M M-Hi M M M M-Lo M
Automated irrigation a
M M-Lo M-Hi Hi M-Lo Lo Hi

15
Sustainable practice Geographic
applicability
Relative
advantage
Risk Complexity Compatibility Trialability Observability
(Ideal rating) (Hi) (Hi) (Lo) (Lo) (Hi) (Hi) (Hi)`
Rangelands
Control grazing pressure by excluding access to water a
M M Lo M-Hi M M-Lo M-Hi
Controlled of water flow from bores a
Hi M-Lo Lo Lo Hi Hi Hi
Piped water supplies for stocka
Hi M-Lo M Lo Hi M Hi
Pastoral land stocked at recommended rates Hi M M M Hi M-Lo M-Hi
Degraded pastoral land converted to less damaging use M Lo Lo M M M-Lo M
Pastoral land destocked in low feed conditions Hi M-Hi M-Hi Hi-M Hi M-Lo M
Dairy farms
Use of effluent disposal systems
(collection of effluent; ponds or drainage sump) aHi M-Lo M M M M Hi
Pump dairy shed effluent onto pasture a
M M-Lo Lo Lo M Hi Hi
a Some measure of the level of landholder adoption of the practice available from the ABARE Australian Resource Management Supplementary survey.
(Comments in brackets refer to locality or temporal constraints on expression of attribute.)
Hi = High
M = Medium
Lo = Low

16
Observations on the characteristics of
sustainable practices
The following features and conclusions regarding
sustainable practices and their attributes can be identified:
• There is no one sustainable practice which optimally
comprises all the attributes by being widely applicable,
having high relative advantage to the landholder, low
complexity, high compatibility, high trialability and
observability, and low risk.
• Very few sustainable practices have widespread or
universal geographic applicability. As a consequence,
the identification, development and promotion of
relevant sustainable practices needs to be locality or
catchment specific.
• The sustainable practices with wider geographic
applicability (such as deep-rooted perennials) often
provide only moderate relative advantage to the
landholder. The relative advantage will be different in
different localities.
• The level of relative advantage is rarely independent of
commodity prices. The relative advantage of many
sustainable practices (such as deep-rooted
perennials) will be temporally dependent on the value
of rural commodities produced as a result of using the
practice. Low commodity prices in the broadacre
industries have reduced the relative advantage of
many sustainable practices.
• The relative advantage and risk attributes are the least
mutable in terms of feasible policy interventions.
Where relative advantage is low and risk is high,
attempts to achieve wide-scale adoption will require
large levels of external subsidy or insurance
intervention. It will be more feasible to promote those
sustainable practices which have higher relative
advantage (and preferably lower risk) and to use
policy interventions (such as extension and education
programs) to overcome or ameliorate complexity and
low compatibility and observability.
A case example: phase farming with
dryland lucerne
The watertable under the Murray Darling riverine plains has
been rising since the last century. The long term solution
for much of the plains is to develop a system of farming
based on a productive and profitable, deep-rooted
perennial crop. The most appropriate commercial plant is
lucerne. Dryland lucerne has been known of for many
years, yet only a few farmers grow significant areas of
lucerne (Ransom & Barr 1993; Whittet 1929).
The use of lucerne, a deep-rooted perennial species, is an
example of an apparently simple sustainable management
practice that has not been widely adopted. In most
circumstances of land degradation lucerne has a medium to
low relative advantage, reflecting low prices for pastoral
commodities (see Curtis et al. 2000). Lucerne is relatively
complex to introduce into a pastoral management system,
and there are considerable risks in its successful
establishment. Farmers sowing lucerne do not have a
guarantee they will successfully produce a crop of lucerne.
The chance of failure is greater than most other pasture
species. One way to minimise the financial risk of
establishing lucerne, and to make up for time a paddock
may be out of production, is to sow lucerne with a faster
growing crop such as safflower. Farmers following this
strategy may have to learn to grow new crops which are
more compatible with lucerne (Barker 1992).
Lucerne requires rotational grazing management. The
majority of farms are currently managed with a regime of
set stocking. Wool-producing farms typically run three
flocks: ewes, weaners and wethers. Some run an
additional flock of maiden ewes. Under the four-paddock
rotation system, such a farm would need 12 or 16
paddocks. For farms previously ‘set-stocked’ this implies
additional expensive fencing and more dams and
reticulation to provide watering points in each paddock.
Fencing at this intensity is likely to impede the easy
management of cropping activity on the farm.
Lucerne pasture is more productive than normal pasture,
but wool producers will not make money merely by growing
more pasture. There are complex ramifications in the farm
system. More sheep will be required to utilise the extra
pasture (Ransom 1992). The increased flock size will
require extra capital, more work in sheep handling and an
increased workload of rotational grazing. Higher sheep
densities in paddocks may mean a greater need for control
of intestinal parasites and increased use of veterinary
chemicals or greater attention to rotational grazing systems
to minimise parasite infestation (Coffey 1992).
One means of maximising the benefit of lucerne is to
abandon lambing in autumn in favour of spring lambing.
This may mean a need to further re-arrange the farm
timetable. Shearing will probably be moved to after the
harvest season and before sowing. The risk of grass seed
contamination will be higher. Grazing rotation strategies to
minimise this risk will be needed. To maximise the benefits
of prime lamb production, the farmer will often need to
develop new marketing skills and develop relationships with
export abattoirs.
These changes have to be worked in with the continuing
cropping enterprise. Lucerne can imply major changes in
crop management. How does the farmer combine the new
grazing rotation with the crop rotation side of the business?
Whereas an annual pasture may have been grazed for a
couple of years before cropping, there are good reasons to
maintain a lucerne paddock for its full eight-year life after
successful establishment. Consequently, the farmer may
have to crop paddocks elsewhere on the farm for a longer
period before putting them back into pasture. Forestalling

17
the depletion of soil nitrogen will inevitably mean
introducing grain legumes into a rotation system that was
predominantly based on wheat and pasture. This will
require improved cropping skills, marketing skills and
probably investment in cropping machinery.
Lucerne will also introduce greater risk into cropping
systems. The environmental advantage of lucerne is its
ability to remove water from the soil profile to reduce
recharge of the watertable. Traditional long fallow crop
systems were successful in minimising risk by conserving
soil moisture before a crop phase. Entering a crop phase
after drying the soil moisture may increase crop production
risk if the following season’s rainfall is below average.
Currently in southern Australia climate forecasters are
unable to provide useful forecasts to guide phase farming
decision-making.
Finally, a farmer considering integrating lucerne into the
farming system may need to borrow capital in the early
stages of the project. A bank is likely to require a business
plan to analyse the financial implications of the plan before
agreeing to the provision of loan finance.
The adoption of dryland lucerne in
central Victoria
Dryland lucerne has been promoted as a farming system
for over a decade and the adoption of this system has been
relatively well monitored. There has been a significant
increase in the adoption dryland lucerne during this period.
However, the rate of adoption is mediated by a number of
factors which lead to the conclusion that full adoption is
unlikely to be attained.
In 1991 Agriculture Victoria conducted a survey of dryland
lucerne adoption in north central Victoria. Using very
conservative assumptions about the non-respondents, the
investigators concluded that the area of lucerne had
increased from 4.2 percent of farmland in 1984 to 7.6
percent in 1991. A segmentation analysis revealed
adoption was limited to a small number of producers, but
that there was a very high latent interest in growing dryland
lucerne.
• Es ta blishe d luc ern e gro wers. T his grou p c ons is ted of
six farmers who ha d a h istory o f g oo d luce rn e
es ta blishmen t, man ag eme nt an d a ve ry po sitiv e a ttitu de
to luce rne . Th e e stablish ed lu cerne growe rs ha d o n
av erage 43 p erc ent o f their farm u nd er luc erne. F ifty per
ce nt of th e luc ern e in the c atc hme nt will be fo und o n this
re la tiv ely s mall p ercen tag e of farms . Farm siz es we re
high er tha n ave rag e.
• Lu ce rne plan ners. T his grou p o f 2 9 farmers believ ed
lu ce rne ha d a majo r role in the ir fu ture farmin g p la ns and
be lieve d the practic al pro blems as so cia ted with lu ce rne
co uld b e o ve rco me.
• De te rre d g ro wers. T his grou p o f 3 1 farmers had little
lu ce rne on their farms. T he y would lik e to hav e a larg er
area , b ut be lie ved there were too ma ny pra ctica l
prob lems to mak e this a wo rthwh ile g oal.
• Disinte res te d. T his grou p c ons is ted of 2 4 farmers who
mo stly belie ved th ere was little p la ce for luce rne o n their
fa rms. Th is group were also le ss likely to be membe rs of
la nd care typ e g rou ps . Farm siz es we re sma ller tha n
av erage .
This segmentation revealed the importance of finding easily
adoptable solutions to the technical and management
challenges posed by the integration of dryland lucerne into
the then traditional cropping/annual pasture mixed farming
system used in the district. Lucerne is relatively complex to
introduce into a pastoral management system, and there
are considerable risks in its successful adoption. The risks
and concerns revealed by the survey and informal
interviewing were considered previously and are listed
below:
• establishment failure
• lucerne requires rotational grazing management
• heavier stocking densities
• significant changes to farm management systems
• competition with cropping program
• drought risk management.
The degree of interest in lucerne was demonstrated by the
high degree of enthusiasm for an increased area of
lucerne. It was estimated that 36 per cent of the region was
suitable for dryland lucerne. Farmers indicated that they
would like to see 25 per cent of farm area under lucerne if
particular technical and management problems could be
overcome. Even at the existing sowing rates, it was
estimated the area of lucerne would rise to 11 per cent with
no change in the current establishment success rate.
Improvements in the establishment success rate to that
obtained by the more experienced farmers would see the
area of lucerne increase to 17 per cent of the farm area.
A follow up survey in 1996 revealed some contradictory
results (Oxley 1997). The overall adoption rate for dryland
lucerne increased significantly. The number of farmers
sowing more than 5 per cent of their farm area to lucerne
rose from 27 per cent to 48 per cent. The average
establishment success rate rose from 36 per cent to 60 per
cent. However, there was no change in the total area of
lucerne in the catchment. While more and more farmers
had been sowing lucerne, farmers with existing paddocks of
lucerne had been converting them to cropping in response
to the more attractive returns from cropping. The obviously
successful extension effort to promote the benefits of
dryland lucerne had merely managed to maintain the
existing area of dryland lucerne.
Since the 1996 survey, there has been a steady increase in
the area of lucerne sown in the North Central region (see
Figure 2), reflecting a gradual improvement in the relative

18
returns to livestock enterprises, particular prime lambs, in
comparison to returns to cropping enterprises (Karunaratne
& Barr 2001). The benefits of the lucerne extension
program of the previous decade are still being reaped,
underlining the difficulties of evaluating the adoption
outcomes of an extension program based upon data from a
short time period.
Figure 2 Area of mixed lucerne pasture and adoption of mixed lucerne pasture in the North Central catchment region of
Victoria: 1996-99 (Source: ABS)
A key conclusion is that dryland lucerne will not be adopted
when its comparative advantage is less than that provided
by cropping. Cropping has been a more profitable
enterprise for the past decade. The need or desire to
generate income has overridden any salinity control
benefits.
Over the past decade many farmers in north central Victoria
have gained the necessary managerial skills to give them
confidence in growing dryland lucerne. There has also
been a recognition of its recharge benefits. This has been
potentially a very positive story in the promotion of a
sustainable agricultural practice. However, while the current
relativity of cropping and grazing gross margins prevails,
there is a very strong constraint on the extent to which
lucerne can be expected to be adopted. The main
motivation for adopting dryland lucerne is its potential for
improving the profitability of grazing enterprises. This same
profitability motivation will ensure that the needs of the
cropping program will generally take precedence over
recharge control objectives.
30000
32000
34000
36000
38000
40000
42000
44000
46000
48000
50000
1996 1997 1998 1999
Pasture mix area (ha)
100
150
200
250
300
350
No. of farmers
Mixture of lucerne and other pasture area
Farmers reporting mixture of lucerne and other pasture

19
Learning about sustainable practices
As is the case in most occupational groupings, there is a
wide range of abilities and knowledge among farmers.
There is also a wide range of formal education and
knowledge about sustainable farm practices. These factors
suggest that to encourage better understanding and
implementation of sustainable management practices it is
more important to focus on how farmers might learn about
using these practices rather than to rely on exiting formal
levels of education.
Categorising the learning focus
Kilpatrick et al. (1999) recently carried out a major research
project exploring how farmers’ learning for management
and marketing can be improved. The research, funded by
the Rural Industries Research and Development Fund
(RIRDC), was motivated by the perception of experts that
farmers did not participate in training, particularly in
marketing and management, to their best advantage.
Building upon previous research into the learning needs
and styles of Australian farmers (eg Kilpatrick 1997;
Kilpatrick & Williamson 1996; Reeve & Black 1998;
Synapse Consulting Pty Ltd 1998), Kilpatrick et al. (1999)
conducted two separate studies. The first was a national
study involving qualitative interviews with 85
representatives of ‘farm management teams’ (the couple or
group involved in on-farm management decisions) in five
states. The second study was a Western Australia study
involving interviews with a random sample of 197 farmers
from eight agricultural regions of the south west.
Kilpatrick (1996) highlighted the important role that
education and training3 play in assisting farmers to make
changes in their farming practice. However not all farm
managers learn in the same manner. Farm managers often
differ in the learning sources they accessed, the manner in
which information was available to them, and their
motivations for learning. Kilpatrick et al. (1999) investigated
both learning-for-change and on-going learning. On the
basis of one specific change (learning-for-change) that the
farm management team had implemented, and previous
research, four learning pattern groups were developed.
These were:
3 Education and training ‘includes all organised education and
training activities, both non-formal and formal. . . . field days,
farmer-directed groups, seminars, conferences and workshops,
and non-accredited courses as well as formal education and
training, all are included as education and training activities’
(Kilpatrick et al. 1999:xi).
• Local focused The local focussed group seeks
information and advice only from local experts4 and
local farmers. They do not participate in training,
except for attendance at field days.
• People focused Such farm businesses consult two or
more people and use no more than one other learning
source when making changes (eg training, media and
observation).
• Outward looking These farm businesses use a variety
of sources, usually including one of training, media
and observation in addition to one-on-one learning
from other farmers, experts or agricultural
associations/organisations.
• Extensive networking These farm businesses consult
a wide range of sources when learning for change,
typically more than four sources including experts,
training, other farmers, media, agricultural
associations/organisations, and observations
(Kilpatrick et al. 1999:33).
Outward looking farm business dominated the national
study sample (40%) followed by people focussed (23.5%),
local focussed (18.8%) and extensive networking (17.7%)
(Kilpatrick et al. 1999). While these categories were
established on the basis of learning about farm
management, they seem equally appropriate for the
process of learning about more sustainable management
practices.
Farm management teams were also categorised according
to their farm management skills. Three levels were
developed with the farm businesses that exhibited higher
levels of management skills and experience given Level A,
and lower levels given Level C (Kilpatrick et al. 1999:29).
There were no significant differences between the learning
pattern and management category. However there were no
Level A farm business that were local focused and no Level
C farm businesses that were extensive networking in
learning focus.
Reasons for learning
Traditionally there has been a low level of formal education
among Australia’s farmers, however levels of education
have increased from 23 per cent with post-school
4 An expert ‘includes those who have specialised information and
skills of use to the farm business. Examples are government
extension officers, accountants, buyers of farm product,
company field officers, researchers, lawyers, rural counsellors,
suppliers of inputs (such as rural merchants) and private farm
consultants’ (Kilpatrick et al. 1999:xi).

20
qualifications in 1983 to 32 per cent in 1995. Though this is
still less than the 49 per cent of the Australian labour force
that has post-school qualifications (Synapse Consulting
1998). Importantly, those farmers with higher levels of
formal education are more likely to seek out and participate
in further education and training. However there is mixed
evidence concerning the link between formal farmer
education and good farm management (Bamberry, Dunn &
Lamont 1997; Kilpatrick et al. 1999). Though Gould, Saupe
and Kleme (1989) report that better educated farmers were
more likely to adopt conservation practices and Reeve and
Black (1993) found that they had more favourable attitudes
towards using outside expertise in conservation practices.
The motivations given by farmers for their learning were as
follows:
• improved farm business efficiency (52.9%)
• improved farm business viability (29.4%)
• acquisition of marketing information and skills (23.5%)
• compliance with legal requirements (15.3%)
• learning to better manage risk (14.1%)
• environmental awareness (10.6%)
• personal development (7.1%) (Kilpatrick et al. 1999)
Importantly, environmental management motivates only a
relative minority of farmers to learn, in contrast to business
efficiency and viability.
Styles of learning
Farmers draw upon a wide range of sources in their
learning, and changes to farm management are typically
influenced by a number of sources (Phillips 1985). Informal
interaction with others and social networks are very
important in farmer learning. Such interactions provide
opportunities for farmers to compare views on how
information could be applied to their own situations and to
test each other’s values and attitudes towards making
changes as a result of the information (Kilpatrick et al.
1999). People were cited as the most important sources (of
support and information) for both learning-for-change and
on-going learning.
Informal sources of learning were preferred by farmers as
they tended to have a:
• preference for independence
• familiarity with highly contextual learning mode
• lack of confidence in working in training settings
• preference for information from known sources
• fear of being exposed to new knowledge and skills
(Kilpatrick & Rosenblatt 1998).
Typically, farmers choose learning sources according to the
need; thus other farmers were often sought out for
background information and for information on practical
issues related to farming, extension officers and
consultants for detailed technical advice, and family and
employees for support during change.
Farmers learning to make a specific management change
used a variety of sources (6 types) with experts being the
most frequently accessed in learning-for-change situations.
Of the experts that were sources for learning by farmers,
government consultants were the expert source most often
used (Kilpatrick et al. 1999). Experts were the most
frequently used type of learning source for all learning
pattern types, except for the extensive networkers where
they were equal first with training. The manner in which
experts were perceived was contrasted between farm
management Levels A and B, where experts were
perceived as a resource to aide decision-making about
some change, and management Level C where experts
were seen as decision-makers.
Kilpatrick et al. (1999) differentiated between four types of
change:
• starting a new enterprise
• other strategic change
• record keeping
• tactical or technical changes.
Farmers sought access to different learning opportunities
for different types of change. Training was most frequently
sought for record keeping changes, while experts (primarily
government consultants) dominated other types of change.
Other farmers were also a frequently used source of
learning for tactical and technical changes as they were
seen as having good local knowledge.
Education and training, including field days, seminars,
farmer-directed groups, and both accredited and non-
accredited courses, were also important sources of learning
for some sections of the farming community. Field days
and accredited courses were useful to approximately 75 per
cent of farmers (Kilpatrick et al. 1999). Farmers with no
post-school qualifications were most likely to draw upon
field days, whilst those with agricultural qualifications drew
upon accredited and non-accredited courses (Kilpatrick et
al. 1999). When considering a specific change in farm
management, non-accredited courses followed by
accredited courses and field days were the most frequently
used (Kilpatrick et al. 1999).
Those farmers who identified environmental management
as a motivation for learning drew upon two or three learning
sources. In all cases these always included a farmer-
directed group (such as community landcare). Community

21
landcare and other similar groups have been highlighted as
an important source of information concerning sustainable
farming practices (Cary & Webb 2000).
Recent studies have highlighted the role that women play in
Australian agriculture (Alston 1995; RIRDC/DPIE 1998),
and learning for increased adoption of conservation
practices on farms should be cognisant of the roles that
women play in agriculture. The learning styles generally
preferred by men and women may be different (Kilpatrick et
al. 1999). Recent initiatives of the Women in Rural
Industries Section of AFFA have highlighted the
advantages of specifically targeting rural women in
education and information programs (Webb 2000a, 2000b).

22
Some recent Australian findings on factors associated with the adoption of
sustainable practices
In recent years there have been a number of major reviews
and studies that have explored the social aspects of
adoption of best practices in Australian agriculture (eg.
Fenton, MacGregor & Cary 2000; Barr & Cary 2000; Guerin
& Guerin 1994; Reeve & Black 1993). Two more recent
studies warrant further exploration. The first is a study
carried out by Curtis et al. (2000) in the Goulburn Broken
catchment of Victoria, the second is a benchmarking study
carried out by Solutions Marketing and Research (1999) to
monitor the achievement of goals in AFFA’s Agriculture –
Advancing Australia (AAA) policy package, hereafter
termed the ‘Solutions’ study.
The Curtis et al. (2000) study was in response to the
realisation that the adoption of best practices in the
Goulburn Broken Catchment was slower than is required to
arrest dryland salinity. Using a mail survey Curtis et al.
(2000) explored the key social factors affecting adoption of
best practices among a random sample of rural properties
covering the 14 land management units5 of the catchment.
Four hundred and eighty landholders completed the survey,
which explored the relationships between the use, and non-
use, of best practices and a range of landholder, business
and property characteristics. The practices explored in the
survey were:
• area sown to introduced perennial pastures
• area of changed grazing/fertiliser regimes to encourage
native perennial pastures
• area of remaining native bush and waterways fenced
• area of trees planted
• number of ground water pumps installed
• area of high density/intensive grazing.
Curtis et al. ( 2000) found that the lack of financial capacity
(level of net farm income) was a major constraint to
adoption. Landholder age was not seen as a constraint to
adoption, while property size, and its links to financial
capacity, was likely to be a major influence on the area over
which best practice was implemented. There is a more
detailed discussion of the characteristics of those
landholders who have adopted best practices below.
5 A land management unit is a area of land with common
geological and hydrogeological characteristics. The impacts of
salinity, its causes and downstream impacts, and the options for
control are common to each land management unit (Curtis et al.
2000).
The Solutions research involved a telephone survey of a
representative sample of 2,043 Australian agricultural
producers and a separate study utilising in-depth interviews
with a non-probability sample of key community people.
Here only the producer survey is discussed. Data were
collected on producer behaviour (by measuring reported
current usage and likely adoption within two years),
producer skill (by measuring producer confidence in their
current expertise to meet their needs now and in two
years), producer awareness, knowledge and usage of AAA
initiatives, producer attitudes, and demographic data
(Solutions 1999).
Solutions developed a series of five national indicators,
linked to the goals of AAA, each comprising a series of
between five and ten individual measures. Table 2 gives
the five indicators and a general description of the types of
measures that comprise each indicator. The current
utilisation score is the average score across all measures
for an indicator, and the score on a measure represents the
percentage of respondents that respond positively to that
measure. The maximum possible indicator score is 100 per
cent, which means that ALL respondents were utilising ALL
monitoring measures within an indicator.
The natural resource management indicator comprises a
number of measures of NRM behaviour; however most of
these are not measures of direct on-ground conservation
practice. Two measures of sustainable practice are
included, these are:
• action to reduce soil erosion in the past two years
• the strategic planting of trees in the past two years
(Solutions 1999).
As an additional monitoring indicator of capacity for change
and adoption of innovation, the adoption of new agronomic
practices was recorded. This was a self-elicited response
from which the following practices were identified:
• minimum till/no tillage
• new/improved/alternative water conservation/irrigation
• farming resource conservation techniques/holistic
management.
While the Solutions research is limited in its selection of
best practices, and other NRM measures, its
representativeness and the development of a longitudinal

23
data set6 make it an important source of data on agricultural
producer’s NRM behaviour and capacity to adopt
innovations.
6 The Solutions survey has been repeated in 2000, and will be
conducted again in 2002.

24
Table 2 Solutions’ National Indicators
Indicator Description Current
Utilisation
Score
Skill
Level
Strategic planning for future Existence and contents of a yearly farm plan, succession
plan and involvement in co-operative planning
46 40
Natural resource management Comprising a range of NRM measures including fire
insurance and prevention strategies, noxious weed
control, land/water resource plan, tree planting, landcare
activity, and climate monitoring
70 68
Financial self reliance Monitoring the financial performance of farm activities,
internal and external financial comparison of activities,
superannuation, off-farm investments, and presence of
retirement plan
59 54
Market competitiveness Calculation of production costs, knowledge and
investigation of marketing opportunities (domestic and
export) and buyer specifications, QA systems,
investigation of new on- and off-farm activities
26 28
Capacity for change and
adoption of innovation
Attendance at field days, training activities, use of soil
testing and advisers, and adoption of innovations, new
technologies
62 72
(Source: Solutions 1999)
As indicated in the above table, producers scored highest
on Solution’s NRM indicator. The strong performance is
linked to the widespread adoption of key measures
included in the indicator such as fire insurance (92% of
respondents), control of noxious weeds (90%), land and
water resource management plan (89%), fire prevention
strategies (80%) and activity in landcare (80%). There
were lower levels of adoption of the two best practices, with
63 per cent of respondents taking action to reduce soil
erosion and 61 per cent planting trees in the two past years
(Solutions 1999). For self-reported agronomic practices, 7
per cent identified their adoption of minimum tillage/no till, 3
per cent new/improved/alternative water
conservation/irrigation, and 2 per cent farming resource
conservation techniques/holistic management over the past
two years. The first two practices are best practices that
are sector specific and this explains in part the much lower
overall levels of adoption. Adoption of minimum tillage/no
till was at 24 per cent for cereal growers, while 10 per cent
of cotton growers had adopted some water conservation.
Furthermore these practices are self-reported and the
categories were generated after the survey had been
carried out. This is in contrast to the items in the indicators
that were known to the respondents.
Both Curtis et al. (2000) and Solutions (1999) record a
range of farmer, farm business, property and other
characteristics that may be linked to the self-reported NRM
behaviour. The following deals with the major
characteristics of those individuals who adopt best
practices.
Age
Younger farmers tend be more aware of land degradation
on their farms. It is also hypothesised that age is likely
have an impact upon the adoption of best practice in
agriculture. In particular, the aging rural population linked
with increasing out-migration from rural areas (Haberkorn et
al. 1999) suggests a reduction in family farm succession,
which in turn may lead to a reduced willingness to invest in
best practice (Curtis et al. 2000). However there appears
to be no clear correlation between age and best practice
adoption. Curtis et al. (2000) found no significant
relationship between adoption and age7, and thus the aging
of the rural population was not considered a major
constraint to adoption. However they found that younger
farmers were more likely to have prepared farm
management plans and budgets. Considering the
composite NRM behaviour indicator that Solutions (1999)
develop there was no significant relationship between age
cohorts and their scores on the indicator (_2=1.446, df=5,
p=0.919). However at the level of the two best practices
there were significant age effects for both tree planting
7 Curtis et al. (2000) performed both bivariate and multivariate
tests for relationships. In the discussion of Curtis et al.’s work
only significant multivariate tests are reported. Where there is
no significant relationship, a significant bivariate relationships
may exist, for example in the case of age there were significant
but very weak correlations between age and the area sown to
introduced perennial pastures, and the area of native bush and
waterways fenced.

25
(_2=78.608, df=5, p<0.001) and for action taken to treat
erosion (_2=24.732, df=5, p<0.001). For tree planting there
is an increase in adoption as age increases to a maximum
adoption rate in the 45-55 age group, and then a decrease
for age cohorts beyond this. The pattern is repeated for
action to treat erosion, though the peak of adoption occurs
in a younger cohort, the 35-44 age group, and declines as
age increases.
Education
There is mixed evidence regarding the relationship between
farmers’ educational levels and their adoption of
sustainable land management practices. In many Australian
studies there are no direct relationships between adoption
of best practices and the level of formal education. Curtis
et al. (2000) found no significant relationship between
formal education and adoption. Likewise analysis of the
Solutions data shows no relationship between education
and the NRM behaviour indicator (_2=3.772, df=4,
p=0.438). However for the two best practices there is a
significant effect. For tree planting, adoption increases with
higher levels of formal education (_2=12.821, df=4,
p=0.012). For the treatment of erosion, increased adoption
occurs with increased formal education to a TAFE level and
then subsequently decreases with tertiary education
(_2=19.281, df=4, p<0.001).
Property size
Property size was a major influence on the adoption of best
practice in Curtis et al’s (2000) study. When the total area
of the best practice (eg absolute area sown to introduced
perennial pastures) was used there was a significant
positive relationship between the adoption of best practice
and property size. When the proportion of a property under
a best practice was used there was a significant negative
relationship between the adoption of best practice and
property size. Curtis et al. (2000) explain that smaller
properties were adopting some practices at levels
representing higher proportions of their total property, while
larger property owners had implemented most best
practices over a larger area. Thus while property size
influences adoption of best practices, there is interest and
adoption among both large and small property owners
(Curtis et al. 2000). Cary (1992) reported a similar inverse
relationship between proportion of property planted to trees
and property size, though no relationship with absolute
number of trees and property size. Cary suggested that
farmers plant a symbolic number of trees, thus those with
smaller properties plant a greater proportion of their
property to trees. Increased property sized is usually linked
to increased financial viability, which in turn removes a
major constraint to adoption. Curtis et al. (2000) did find a
significant relationship between increased property size and
the likelihood of returning an on-property profit and the level
of farm profit. The impact of financial viability on adoption
is considered later.
The Solutions research also recorded property size, though
analysis of the adoption by property size was not
performed.
Farm business
Poor or low financial viability is a major constraint on the
adoption of best practice (Barr & Cary 2000; Barr et al.
2000; Riley 1999). Curtis et al. (2000) found similar
relationships. There was a significant positive relationship
between reported on-farm profitability and the adoption of
some best practices, namely the total area sown to
introduced perennial pasture, changed grazing/fertiliser
regimes and high density or intensive grazing. However
there were no relationships between the proportion of the
property under best practice and on-farm profitability.
Similarly there were no significant relationships between
total or proportion of property under best practice and either
off-farm income or total income.
The only financial data recorded in the Solutions study
were estimates of on- and off-farm assets and gross value
of production. Analysis of adoption by these variables was
not available.
Stewardship
The often tenuous link between pro-environmental attitudes
(often subsumed in the value of stewardship) and pro-
environmental behaviour has been observed in a number of
studies in Australia, as well as in other countries. Attempts
to establish links between measures linked to stewardship
and NRM behaviour related to crop farming practices in
Australian research studies have generally been
unsuccessful. Harvey and Hurley (1990), in a stu dy o f c ro p
farme rs in Victoria fo un d n o sta tistica l re latio n sh ip b e twee n
p erce ption o f e ro sio n, o r c on ce rn ab o ut e ro s io n, an d us e o f
the c on s erva tio n tilla ge te ch niq ue s. A s tu d y of wh ea t
p r o d u c e r s i n N e w S o u t h W a l e s f o u n d t h a t a d h e r e n c e t o a
c on se rv a tion ethic d id n o t sign ifica n tly diffe r b etwe en th e
a do pters a nd no n-ad o pters o f co n se rv a tion c rop pin g (Sin d en
& Kin g, 19 88 ). Van c la y (19 88 ) fou nd fa rmers a dh e ring to a
c on se rv a tion ethic were les s lik ely to ad op t c on s erva tio n
farming . Th e se fin d in gs ma y re fle ct th e me a su re men t
d i f f i c u l t i e s a s s o c i a t e d w i t h m e a s u r i n g ‘ s t e w a r d s h i p ’ i n s u r v e y
intervie ws ; h owev er it is more lik ely , fo r c ro pp ing farmers,
c ro pp in g man a ge me nt de cis io ns a re es s en tially de termine d
o n th e ins tru me ntal grou n ds o f c on ve n ie nc e a nd c o st.
One of the few Australian longitudinal studies of NRM
‘behaviour’ and community landcare membership provides
some evidence of influence of membership on the complex
relationship between attitudes and behaviour with respect
to sustainable land management. The study of landholders

26
in central Victoria between 1988 and 1991 investigated the
influence of economic, psychological and social factors on
the adoption of the sustainable land management practices
of tree planting and the planting of deep-rooted pasture
(Wilkinson & Cary 1992; Cary & Wilkinson 1997; Cary
1999). Perception of long-term profit was an important
predictor of the decision to use both practices, but was
more important for the decision to plant trees. Trees were
planted irrespective of whether salinity was perceived by
the landholder as an environmental problem. In the case of
phalaris pasture, recognition of salinity as an environmental
problem increased the likelihood of planting.
Recent findings regarding attitudes
Despite the implied simplicity of assumed relationships
between attitudes and behaviour, the link is tenuous and
complex. The tenuous nature of the link between attitudes
and behaviour has been discussed above. Curtis et al.
(2000) did not directly measure attitudes in their study,
however respondents were asked a number of questions
that act as proxy indicators of a stewardship ethic. These
were:
• taking nature conservation values into account when
planning work
• willingness to work with government
• importance of community cooperation
• landcare participation.
They found no significant relationship between taking into
account nature conservation values and the adoption of
best practice. However with 78 per cent of respondents
indicating that they did take nature conservation values into
account when planning there is likely to be a high level of
support for a stewardship ethic. There were no significant
relationships between adoption of best practice and either
willingness to work with government or the importance of
community cooperation. However respondents did
acknowledge the importance of cooperation and were often
willing to work with government to tackle salinity. There
was a significant negative relationship between landcare
participation and the area sown to introduced perennial
pastures. This conflicts with earlier research where
landcare participation has been linked to significantly higher
levels of adoption (Curtis & De Lacy 1996).
Reeve and Black (1993) carried out an extensive survey to
explore farmer’s environmental attitudes. While not
restricted to attitudes to chemical use in Australian
agriculture, this was a major focus of the survey. The
survey did not record attitudes to specific best practices,
though it included a range of statements to gain insights
into, inter alia, land degradation and farmer’s conservation
orientation. Unfortunately Reeve and Black did not collect
data on the adoption or implementation of best practices
and thus links between attitudes and adoption cannot be
identified from this data.
The Solutions research incorporated a series of 23
attitudinal statements that related to motivation, planning,
advice, debt, marketing, time management, information
requirements and commitment to the industry. The
responses were cluster analysed resulting in five attitudinal
groups. These were:
• Committed, doing it tough, off farm income supported
(26% of the population)
• The older farmer prepared or preparing to leave (21%)
• Questioning their long term involvement in farming
(18%)
• Confident established older farmer (15%)
• The business person (19%).
On the NRM indicator the business person group scores
higher than the remaining groups. This group is
characterised by having more formal education, younger
average age group, highest value of on-farm assets and
gross value of production, but lower than average off-farm
assets and not reliant on off-farm income. They tend to be
committed to, and satisfied by, farming, and would like
more accurate market information and are generally willing
to pay for advice. The business person group scored
highest on the best practice of strategic planting of trees,
and equal highest with the questioning their long term
involvement group on the best practice of action to reduce
soil erosion.
Other characteristics
Curtis et al. (2000) found significant relationships between
best practice adoption and a number of other
characteristics. Generally these relationships were
significant with the adoption of just one best practice;
except for farming as an occupation where there were
significant negative relationships with the area of trees
planted and the area of river and native vegetation fenced.
These characteristics are shown in Table 3.

27
Table 3 Landholder and property characteristics with significant relationships with adoption of best
practices
Characteristic Number of best
practices
Property size (total area of best practice) 5
On-property profitability 3
Concern about salinity impacts 1
Concern about rising water tables being a threat to pasture production on respondent’s
property
1
Property having plants showing signs of salinity 1
Work on property partially funded by Federal or State government 1
Value of all government contributions to work on property 1
Hours worked off-property 1
Property size (proportion of property under best practice) 1
Having a written property plan that involved a map or other documents 1
Having written personal and/or family goals to be achieved on property 1
Including the cost of work to address salinity in property budget 1
The property will be sold 1
All or most of the property will be leased 1
Number of children 1
Farming as an occupation (total area of best practice) 1
Landcare membership 1
(Source: Curtis et al. 2000)

28
Modeling adoption behaviour from the 1998-99 Resource Management
Survey
Introduction
The relationships between the adoption of sustainable
farming practices and the available range of farm family,
farm property and farm business characteristics are
complex and sometimes tenuous. While some
relationships have been found in empirical studies, (eg. CIE
2001; Curtis et al. 2000; Drake, Bergstrom & Svedsater
1999; Mues, Chapman & Van Hilst 1998; Cary & Wilkinson
1997), there is no widely accepted theoretical model of
human adoption behaviour that can guide and direct
empirical studies. Consequently research tends to be
atheoretical and exploratory in nature.
The Australian Bureau of Agriculture and Resource
Economics (ABARE) was commissioned to undertake an
analysis of data from the Resource Management
Supplementary (RMS) survey for the 1998-99 financial
year. While ABARE performed the analysis of unit data
records and provided statistical advice, the determination of
models to be tested and the interpretation of results was
undertaken by the Social Sciences Centre. The general
approach in the analysis was to model the association
between a range of farm family, farm business and farm
property characteristics and the reported adoption of
various resource management practices. The testing of a
priori hypotheses was not the intention of the analysis,
rather we have inductively explored the associations
between practice adoption and the range of characteristics
where previous research has suggested relationships.
Modelling farmer behaviour
The previous chapter reviewed recent studies concerning
landholder’s characteristics in relation to the adoption of
more sustainable land management practices. Building
upon this review a number of landholder characteristics
collected as part of ABARE’s RMS were selected for further
analysis. While the following gives an overview of the
analysis and findings, a more detailed discussion can be
found in Appendix A.
Data covering a range of landholder characteristics, and
the reported adoption of sustainable land management
practices were selected from the 1998-99 ABARE
Australian Agriculture and Grazing Industries Survey
(AAGIS) and the Australian Dairy Industry Survey (ADIS)
(see Box 1). The surveys incorporated the RMS which
collected data regarding resource management practices
and land management, including the:
• presence, extent and costs of degradation
• participation in training
• landcare membership and involvement in landcare
activities
• content and use of farm plans
• cost of landcare capital works
• adoption of best practices in farm management
• area of crops sown with different tillage practices
• recent changes to tillage practices
• attitudes to degradation and conservation
• farm forestry and functions of trees on farms.

29
Box 1: ABARE Farm Surveys
A
BARE undertakes an annual survey of Australian farms throughout the broadacre industries AAGIS, and the dairy industry.
A
AGIS covers five industry types, namely: 1) wheat and other crops industry, 2) mixed livestock-crops industry, 3) sheep
industry, 4) beef industry, and 5) sheep-beef industry. Data collected as part of these surveys covers a broad range of socio-
economic data concerning farm family, farm business and farm property characteristics. Data collected for the five
broadacre industries in AAGIS are presented for three ABARE defined zones. These are the:
• Pastoral zone: which includes most of the northern tropical areas and the arid and semiarid regions of Australia.
Agricultural land use in this zone is characterised by extensive grazing of native pastures. Although some cropping is
undertaken, it is impractical on most farms because of inadequate rainfall.
• Wheat-sheep zone: which has a climate and topography that generally allows the regular cropping of grains in addition
to the grazing of sheep and beef cattle on a more intensive basis than in the pastoral zone. Rainfall is generally
adequate for producing a variety of pasture species, usually as part of a crop-grazing rotation.
• High rainfall zone: which forms the greater part of the coastal belt and adjacent tablelands of the three eastern mainland
states, small areas in south eastern South Australia and south western Western Australia, and the whole of Tasmania.
Higher rainfall, steeper topography, more adequate surface water and greater humidity make the high rainfall zone less
suitable than the wheat-sheep zone for grains based cropping but more suitable for grazing and producing other crops
(ABARE 2000).
Based upon their utility in understanding farmer adoption
behaviour 16 landholder characteristics were selected for
analysis. These characteristics are listed in Table 4, and
definitions of each variable are provided in Appendix B.
The potential range of sustainable resource management
practices that could be adopted will be determined by the
nature of the farming enterprise and its location (see Table
1, Chapter 3). Some practices will only be relevant to
particular types of industries, or to particular locations.
Table 5 details the practices for which adoption was
explored.
Table 4 Variables explored in analysis of Resource Management Supplementary survey
Farm family characteristics: state of residence
Farm financial characteristics: farm cash income
profit at full equity
closing equity ratio
Social and institutional contact: landcare membership
length of landcare membership
Education and training: recent training
PMP participation
Farm structure: farm size
land use intensity
Identification of problems: farm plan
Age and experience: age
Attitudes: financial concern attitude
financial outlook attitude
technical concern attitude
environmental concern attitude

30
Table 5 Resource management practices investigated
Resource management practice Pastoral Wheat-
Sheep
High
Rainfall
Dairy
Farms
Irrigated
Farms
controlled flow bores
x
controlling grazing pressure by
excluding access to water
x
monitoring of pasture and vegetation
condition
x
deep rooted perennial pasture x x
tree and shrub establishment x x
x
regularly monitor water tables x x
x
soil/plant tissue test to determine
fertiliser needs
x x
x
collection of dairy effluent (ponds or
drainage sump)
x
pump dairy shed effluent onto pasture
x
laser graded layout x
use irrigation scheduling tools x
monitoring of pasture and vegetation
condition
x
x x
x
x
preserve/enhance areas of
conservation value
x
x x
x
x
exclude stock from degraded areas
x
x x
x
x
Percentage conservation tillage
x
x x
x
x
Findings
The significant findings from the analysis are summarised
in Tables 6 and 7. Table 6 presents the frequencies with
which family, farm property and farm business
characteristics were associated with adoption of the
investigated management practices. In the majority of
cases the direction of the relationship between practice
adoption and characteristic as suggested by past studies
and theoretical propositions was correctly predicted.
Table 7 summarises the relationship between
characteristics and individual practices. Symbols in a box
indicate a significant association between the relevant
characteristic and practice adoption. An addition (+) symbol
indicates a positive relationship; as the characteristic
increases so too does the likelihood of practice adoption. A
minus (-) symbol indicates the converse. In the case of the
state of residence an asterisk (*) indicates a significant
association, though there is no inherent direction in the
variable. Where the direction of a significant relationship is
confirmed the cell is shaded.
The significant findings are considered in the context of
other recent studies. In particular we draw upon recent
work carried out in the rangelands for the Audit by the CIE
(2001), work by Curtis et al. (2000) in Victoria’s Goulburn-
Broken catchment and work by Mues, Chapman and Van
Hilst (1998).

31
Table 6 Characteristics significantly associated with practice adoption
Characteristic Frequency of significant
associations
predicted not predicte
d
state of residence 9
financial outlook attitude 7 1
farm plan 6 0
recent training 6 0
Environmental concern attitude 6 0
land use intensity 4 2
technical concern attitude 4 1
closing equity ratio 1 3
landcare membership (1998-99) 3 0
length of landcare membership 1 1
financial concern attitude 1 1
PMP participation in last 3 years 2 0
age 2 0
farm cash income 1 0
farm size 0 1
profit at full equity 1 0

32
Table 7 Factors which are associated with the adoption of sustainable management practices (shaded cells indicate association relationships in predicted
direction)
Pastoral zone Wheat-sheep and high rainfall zones Dairy farms Irrigation farms All farms
Farm family, farm property and farm
business characteristics
controlled flow
bores
controlling
grazing
pressure by
excluding
access to
water
monitoring of
pasture and
vegetation
condition
deep rooted
perennial
pasture1
soil/plant
tissue tests to
determine
fertiliser
needs1
tree and
shrub
establis hment2
regularly
monitor water
tables2
collection of
dairy effluent
pump dairy
shed effluent
onto pasture
laser graded
layout
use irrigation
scheduling
tools
monitoring of
pasture and
vegetation
condition
preserve/enha
nce areas of
conservation
value
exclude stock
from
degraded
areas
percentage
conservation
tillage
Age - -
Environmental concern attitude + + + + + +
Technical concern attitude - - +- -
Financial concern attitude + -
Financial outlook attitude -+- - - - - -
Landcare membership (1998-99) + + +
Length of landcare membership +-
Recent training + + + + + +
Farm cash income +
Closing equity ratio +-- -
Profit at full equity +
Farm plan + + + + + +
Farm size -
State * * * * * * * * *
Land use intensity + + + + --
PMP participation in last 3 years + +
+ significant positive association at the 95% confidence level or higher
- significant negative association at the 95% confidence level of higher
* see individual models for nature of relationship
1 broadacre farms only
2 including dairy farms

33
State of residence
Though clearly not a driver of practice adoption, state of
residence was commonly a significant explanatory variable
in adoption behaviour. Farmers based in Queensland were
typically less likely to adopt the practices explored than the
Australian average. This may be due to structural and
institutional arrangements that are unique to Queensland
farmers, alternatively it may be a consequence of
Queensland farming systems and the applicability of the
practices explored. The finding certainly suggests the need
for more detailed exploration of the Queensland situation,
particularly as much previous research has focussed on the
south-eastern states.
Age
Age was significant for the adoption of two practices, with
younger farmers more likely to adopt than older farmers.
This is consistent with findings that suggest that younger
farmers tend to be more aware of land degradation and
recognise the need for the adoption of conservation
practices (Fenton, Macgregor & Cary 2000).
Farm financial characteristics
The financial outlook attitude variable reflects
owner/operators’ perceptions about the future profitability of
their farms. With the exception of one practice, the direction
of the significant association was as expected. Those
individuals who thought their future profitability would fall in
the next five to 10 years were less likely to adopt the
practice. To a much lesser extent, owner/managers who felt
more able to afford to address land and water degradation
were more likely to adopt practices. The findings confirm
that financial capacity is an important component in
determining the capacity of individuals to adopt new
practices. Farmers who feel secure in their financial future
are more likely to invest resources in adopting new
resource management practices. Absolute measures of
financial capacity (farm cash income and profit at full
equity) were each positively related to adoption of only one
practice. Those owner/managers who had higher equity,
and thus greater financial flexibility to operate their farm
businesses, were less likely to adopt particular practices.
However in the case of controlling flowing bores the
converse relationship was found. Overall, the relationship
between equity and adoption was not as might be
predicted. This suggests the possibility of other
confounding effects associated with high equity and the
need for more careful analysis of the adoption habits of
those farming enterprises with high equity. It is possible
that landholders with high equity in their properties are
more risk averse and thus less inclined to adopt risky
resource management technologies.
Education and training
Recent involvement in training courses was consistently
significantly associated with adoption in a positive direction.
Those landholders who had attended more training courses
were more likely to adopt practices than those who had
attended less. Training is clearly an important contributor to
an individual’s capacity to change. Training was the
characteristic most frequently linked to practice adoption in
Mues, Chapman and Van Hilst (1998) study, though of less
importance in the CIE (2001) study. In addition to the
number of recent training courses attended, the more
specific involvement in a Property Management Planning
course or program was positively associated with the
adoption of two practices.
Training and participation in PMP may alleviate technical
concerns that owner/managers have about resource
management practices. While this direct relationship was
not explored, those owner/managers who felt they did not
have the technical resources to adequately address land
and water degradation on their property were less likely to
adopt resource management practices. This concern, and
the impact that training has on adoption behaviour, suggest
that training and overcoming any fears that
owner/managers may have about resource management
practices are an important aspect of an individual’s capacity
to adopt. Technical concerns about the resources required
to adequately address land and water degradation is likely
to reflect the fact that many sustainable management
practices are complex to integrate into farming systems and
are often not adapted or appropriate for use in many
localities.
It should be noted that a generic education level variable
was not incorporated in the models estimated due to
difficulties in developing a meaningful measure and, more
importantly, due to likely confounding effects with age.
Furthermore the specific measures, just discussed, are
likely to be a more accurate reflection of relevant training.
Farm structure and farm plan
Land use intensity measured in sheep equivalents per
hectare was associated with adoption for six practices. For
four of these practices as the intensity of land use
increased so too did the adoption of practices, while in the
case of preserving or enhancing areas of conservation
value and conservation tillage the association was
negative. Mues, Chapman and Van Hilst (1998) also found
a significant positive association with two practices they
investigated. In general it may be argued that those
owner/managers who farm most intensively would need to
adopt resource management practices in order to maintain
the productive capacity of their property. Those who did not
may find the resource base is unable to withstand
increased intensity of production.
While Curtis et al. (2000) highlighted the importance of farm
size in their work on practice adoption, this analysis found
that farm size was only significantly associated with
adoption of deep rooted perennial pasture in the wheat-
sheep and high rainfall zones.

34
The existence of a farm plan was significantly positively
associated with the adoption of six practices. The
importance of a farm plan was not demonstrated in either
Mues, Chapman and Van Hilst (1998) or Curtis’ et al.
(2000) studies, however it was the most frequently
associated characteristic in the CIE (2001) study of the
rangelands. The presence of a farm plan would suggest a
more pro-active and prepared owner/manager who may
take greater advantage of new farming techniques and
approaches.
Landcare and environmental attitude
Landcare membership has long been associated with
greater adoption of resource management practices (eg
Cary & Webb 2000; Mues, Chapman and Van Hilst 1998;
Curtis & DeLacy 1996) and these findings, at least in some
cases, also support that association. However the length of
landcare membership was anomalous with both a positive
and negative association found. Landcare membership and
the length of that membership are sometimes used as a
surrogate for commitment to a stewardship ethic or pro-
environmental value position (eg Curtis et al. 2000). In this
study an environmental concern attitude was able to tap
this stewardship dimension, and the expected relationship
was commonly found. Those owner/managers who
considered land and water degradation as critical concerns
in farm planning were also more likely to have adopted
resource management practices.
Summary
Personal financial capacity was an important component in
determining the capacity of landholders to adopt new
practices. Farmers who felt secure in their financial future
were more likely to invest resources in adopting new
resource management practices.
Landholders’ perceptions of their financial situation were
more often associated with practice adoption than were
objectively measured indicators of financial position. An
individual’s subjective assessment of their financial
situation may be a better predictor of adoption than
objective measures. This highlights the importance of
perceived reality in adoption behaviour; similar associations
between financial perceptions and business behaviour can
be observed in the wider economy.
More frequent landholder involvement in training courses
and having a farm plan were commonly associated with
adoption of resource management practices. Participation
in training courses related to management and skills is an
important contributor to an individual’s capacity to adopt
sustainable practices.
Landholders’ who considered they did not have the
technical resources to adequately address land and water
degradation on their property were less likely to adopt
resource management practices. Training may reduce this
concern and increase human capacity to implement
resource management practices. Technical concern about
the resources required to adequately address land and
water degradation is also likely to reflect the fact that many
sustainable management practices are complex to integrate
into farming systems and are often not adapted or
appropriate for use in many localities.
The analysis produced some statistically significant findings
which compare favourably with other studies, the
differences between studies and some anomalous findings
suggest more consideration needs to be given to the types
of characteristics that will usefully explain the adoption of
sustainable practices. Furthermore analysis such as these
that tend to be based on a large zonal, or whole of
Australia, approach which will frequently be confounded by
the large variability that exists in Australian agriculture and
the often locality-specific nature of many sustainable
management practices. Understanding of farmer adoption
behaviour could be advanced through more specific studies
focussing on particular localities and industries.

35
Attitudes and values and the adoption of sustainable practices
Because of the often tenuous nature of the link of attitudes
and values with behaviour it has always been difficult to
predict individual behaviour as a consequence of an
individual holding a particular value or having a particular
attitude. A weakly, or moderately, held attitude or value will
generally not result in a corresponding and consistent
behaviour where that behaviour requires ‘strong’
commitment of self or personal resources.
Attitudes
In many studies it has been observed that attitudes and
behaviours are related to an extent that ranges from a small
to a moderate degree. There is a general tendency for
individuals, in the absence of constraints, to seek
consistency between attitudes and behaviours. Another
way for individuals to achieve psychological consistency is
to publicly espouse ‘symbolic beliefs’ reflecting the relevant
social norms but engage only in token behaviour, sufficient
to provide apparent consistency (Cary 1991; Cary 1993).
Instrumental beliefs, related to self-interest, are likely to be
more powerful than (moderately held) symbolic beliefs in
influencing substantive environmental behaviour. The
attitude–behaviour relation is further complicated by the fact
that causation is not one-way: behaviour can also
determine attitudes.
The link between behaviour and attitudes is complex
because the relationship between attitudes and behaviour
is commonly many-to-one; i.e., many different attitudes – of
potentially differing strength and direction and including
attitudes towards complying with social norms regarding the
behaviour– may be associated with a particular desirable
behaviour. Thus an individual may have some attitudes
that are ‘positive’ towards a particular NRM behaviour (e.g.
it is good for the environment) and other attitudes that are
‘negative’ towards the particular behaviour (e.g. it costs me
money). An attitude exists within a personal knowledge
structure comprised of beliefs, linked in associative
networks. If all ‘attitudes’ in an individual’s belief system
with respect to the behaviour are not taken into account it is
unlikely behaviour can be successfully inferred.
Beliefs
Beliefs are the knowledge base upon which attitudes are
formed. The traditional, all-embracing tripartite conception
of an attitude asserted that an attitude comprised
emotional, cognitive and behavioural components. The
cognitive component is better thought of as the relevant
beliefs that underpin individuals’ attitudes. Beliefs can be
thought of as assertions about the degree of association
between objects which exist within, and comprise, a domain
of cognition. An assertion that planting trees will reduce
groundwater accessions is a belief. More complex beliefs
might embrace assertions about the length of time between
planting trees and subsequent reduction in ground water
accessions.
Belief systems have much in common with broader social
knowledge systems; but they are not identical. Belief
systems belong to an individual and are idiosyncratic. Belief
systems often include representations of ‘alternative’
worlds, typified as ideological beliefs (Abelson 1979). Belief
systems are likely to include a substantial amount of
episodic material from personal experience. And, beliefs
can be held with varying degrees of certainty.
Values
Values are more generalised aggregations of attitudes and
beliefs; values allow more generalised responses to a wider
range of entities. Values tend to be more strongly held and
to be more stable than attitudes and hence they are
changed less easily and less quickly. Changes in values
and subsequent associated changes in behaviour are thus
harder to observe. While there is mixed research evidence
in the literature, there is a body of evidence indicating
positive relationships between environmental values and
environmentally protective actions. Ross (1999:29) has
provided an assessment of the implications:
• Values are closely related to people’s priorities.
• They provide guidance – however loose – to people’s
likely behaviour, including their adoption of new
‘technologies’.
• They offer approaches for assessing what policy
options people will accept, or perhaps reject.
In considering the adoption of more sustainable resource
management practices the landcare ethic provides a useful
characterisation of environmental values. The landcare
ethic embraces a broad continuum of values. At one end –
the ‘deeper green’ end – is a concern for the health of the
land as an end in itself. At the other end is a more
utilitarian or instrumental focus of protecting the land to
ensure its continued productivity and thus economic benefit
to the farmer (Cary & Webb 2000).
Over the longer term aggregate changes in personal value
systems and more strongly held attitudes become
community norms. The resultant formation, or
reinforcement, of norms – such as the norms embracing a
landcare ethic – can lead to the strengthening of social
movements and reinforce feedback loops for socially
desirable personal (pro-environmental) behaviour and for

36
supporting social or legal regulation that prescribes or
proscribes such behaviour.
A framework of environmental
concern
It is useful to develop a simple framework to depict how
values and beliefs influence an individual’s appraisal
process and potentially influence individual behaviour. The
adaptive responses are complex and are mediated by wider
social assessments that are changing over time and are
multidimensional in nature. Values and beliefs which reflect
environmental concern and, potentially, adaptive behaviour
can be described in Stern et al.’s (1995) socio-
psychological framework (Figure 3). The framework we
adapted here is one developed from work by Stern (1992),
Stern et al. (1993), Stern and Dietz (1994) and Stern et al.
(1995).
The broad outline presented in Figure 3 shows interactions
which occur between the various elements. While the
interactions are often two-way, causation generally flows
from top to bottom. The factors identified at the top of the
framework are considered to be less mutable by the
individual or through the life course than those at the
bottom.
Figure 3 A framework of environmental concern (after Stern et al. 1995).
The framework highlights that individuals are located within
a social ‘culture’ which influences the development of
values, beliefs, attitudes and, ultimately, behaviours. Social
culture factors play a large role in the shaping of an
individual’s early life (and later life) experiences and,
general beliefs about the world.
social culture
institutional constraints
incentive structure
values
general beliefs
worldview
specific beliefs
specific attitudes
behavioural commitments and intentions
behaviour

37
Many values are formed through early family socialisation
processes and these values are thought to be relatively
stable in adults (Oppenheim 1992; Stern and Dietz 1994).
Such existing pre-formed values are likely to be resistant to
change; however ‘new’ (as opposed to earlier formed)
values can be more easily embraced.
At the next level Stern et al. (1995) place an individual’s
general beliefs and worldview. These encompass an
individual’s broad understanding of how the world operates.
In the context of environmental issues this level of the
framework comprises an individual’s understanding of the
biophysical environment and its function, and also how the
environment is affected by human actions. Stern et al.
(1995) consider worldview to be causally antecedent to
values. They argue that, in contrast to values that are
formed during early family life, worldviews are more likely to
be the result of broader experiences within the social and
political world. Furthermore, while values tend to be largely
immutable in adults, worldviews, being comprised of
beliefs, are vulnerable to empirical challenge and may
change. Notably worldviews and general beliefs at this
level of the framework are distinguished from more specific
or localised beliefs, such as those associated with a
particular location.
A key feature of Stern’s framework is the role that values
and worldview play in the assimilation of new information by
individuals. Values and worldview may operate as ‘social
amplifiers’ in that a particularly strong value orientation may
lead an individual to selectively seek information or attend
selectively to information about the consequences of some
action for the objects they value (Stern and Dietz 1994).
Likewise values and worldview may act as ‘filters’ for
information where individuals may more readily accept
information that is congruent with their values and
worldview. This function, particularly of value orientations,
may work to attenuate the potential impact that any
empirical challenge, through information provision, has
upon beliefs.
Specific beliefs and specific attitudes are located towards
the bottom of the framework in Figure 3. These
psychological variables are placed most proximate to
behavioural intent and actual behaviour, and are
considered to have greatest impact upon them. In the
framework this is the position where attitudes towards a
particular practice and beliefs about the impacts and
consequences of those practices would be located. Stern
and Dietz (1994) stress that the processes of construction
of an individual’s attitudes towards specific environmental
issues are important as these issues are, often, at least
initially, unfamiliar to those who form attitudes about them.
They argue that individuals tend to ignore details and issue-
specific information, but rather classify a topic and then
make reference to their more general beliefs and values in
forming their specific attitudes and beliefs about that issue
(Stern et al. 1995). When asked to express an attitude
about a particular environmental issue, an individual will
review their beliefs about the issue and assess the likely
impact upon the things they value.
The final two components of Stern’s framework relate to
behavioural intent and finally to actual behaviour. These
are of particular interest to policy makers attempting to
encourage more sustainable farming practices by farmers.
Behavioural intent has typically been operationalised by
Stern and colleagues through the use of responses to
scales about likelihood to take political action and
willingness to pay extra tax to ameliorate negative
environmental consequences (Stern et al. 1993; Stern and
Dietz 1994). The linking of actual behaviour to other
components of the framework, rather than self-reported
intentions, is arguably the least developed aspect of Stern’s
framework. The framework (Figure 3) highlights the
important role that social culture, values and beliefs play in
the formation of specific attitudes and specific beliefs
regarding a particular environmental issue.
Values and the appraisal of
sustainable management practices
Earlier a distinction was made between environmental or
deeper green values and those which were more utilitarian
or instrumental (see section on Values above). Typically
for most people, apart from the most environmentally
committed, utilitarian and instrumental values tend to
predominate over environmental and non-instrumental
values, in determining behaviour related to environment
and the use of sustainable practices. We will now explore
the conditions under which different values come into play.
Guagnano et al. (1995) have enunciated the conditions
under which positive environmental values and attitudes
are most likely to influence, or be associated with,
significant behaviour change. They found the environmental
attitude−behaviour association to be strongest when the
instrumental ‘context’ was weakest or neutral, for example
where the environmental behaviours were not obligatory or
tangibly rewarded. They found the environmental
attitude−behaviour association approached zero when the
instrumental or external forces were strongly positive or
negative, effectively compelling or prohibiting the behaviour
in question (Guagnano et al. 1995). We depict a schematic
diagram of the conditions for maximum influence of
environmental values or attitudes on individual’s decision to
adopt sustainable practices in Figure 4.
Guagnano et al.’s (1995) proposition implies that for
personal behaviours that are not strongly favoured by being
required or tangibly rewarded, the more difficult, time-
consuming, or costly the behaviour, the weaker is the
dependence on attitudinal or environmental value factors.
Values and attitudes have both direct and indirect effects
on individual behaviour. The analysis above applies to the
influence of personal attitudes and values on personal

38
behaviour and this can be considered a direct effect. The
impact of many individuals’ values when expressed in
social aggregates – as social norms – becomes more
complex and their influence is potentially more powerful.
This influence can be considered an indirect effect. Over
the longer term aggregate changes in personal value
systems and more strongly held attitudes become
community norms. The reinforcement of norms – such as
the norms embracing a landcare ethic – can lead to the
strengthening of social movements (such as the landcare
movement) and reinforce feedback loops for socially
desirable environmental behaviour (Stern, Dietz, Abel,
Guagnano and Kalof 1999).
Higher rewards or
incentives
Higher costs or
disincentives
Minimal effect of external
incentives or disincentives
on behaviour
Maximum effect of values and
attitudes on behaviour (when
incentives or disincentives minimal)
Extent of behaviour change
influenced by attitudes
Figure 4 Conditions for maximum influence of environmental values or attitudes on individual’s decision to adopt
sustainable practices

39
Interventions to promote adoption of NRM practices
It is important to understanding the mechanisms by which
values and attitudes may influence NRM behaviour and the
adoption of sustainable practices. It is also important to
recognise the relatively minor impact of ‘pro-environmental’
values and attitudes in bringing about sustainable practice
adoption when there are significant costs and considerable
uncertainty associated with the adoption of many NRM
practices. For significant behaviour changes the impact of
pro-environmental values and attitudes tends to be indirect
(and considerably delayed, through social influence) rather
than directly influencing individuals’ behaviour.
The analysis of the influence of values and attitudes on
behaviour in the previous chapter suggests that greater
influence on sustainable practice adoption will be achieved
by focusing on relevant behaviours and practices rather
than personal attitudes and values. Thus it will be more
useful to adopt a behaviour analysis approach to
encouraging appropriate NRM behaviours.
Geller (2001) has identified three behavioural principles that
are relevant to encouraging the adoption of sustainable
NRM practices. These principles focus on making any
intervention, to change behaviour, more effective. The
behaviour analysis approach is based on the behavioural
intervention principles developed by B. F. Skinner (1953,
1974). Skinner’s behaviourism has been long established
but, more recently, has been unfashionable because of its
exceptionally narrow view of human behaviour as being as
being a series of responses to external stimuli. Human
behaviour is clearly much more adaptive and internally
focused. However, a wider and more complex
understanding of human behaviour does not preclude
Skinner’s behavioural principles being relevant to
explaining human behaviour.
Principle 1: Focus Interventions on Observable
Behaviour
If the behaviour cannot be readily seen by the individual
(and by others) it will be ineffective to encourage it; it will be
difficult to be monitored, to be seen as rewarded (or
penalised, for its absence). Geller contends that behaviour-
based intervention acts people into thinking differently,
whereas person-based intervention thinks people into
acting differently (Geller 2001). Person-based approaches
are impractical for major interventions to change NRM
practices because they are not cost-effective in community
settings. Person-focused intervention requires extensive
one-on-one interaction between a client and a trained
intervention specialist (Geller 2001).
There are constraints to applying this principle to the
adoption of many recommended sustainable practices.
Many practices are not readily observable (see the section
on observability in the chapter Characteristics of
Sustainable Practices) and the outcomes from the practices
are not apparent until a considerable time after the
behaviour is initiated (see the discussion of Figure 1). The
success of farm tree planting, particularly along roadside
fences and in front paddocks, is an example of the
effectiveness of this principle. Where a policy choice exists
for intervention which involves observable behaviour it will
be useful to select such interventions.
Principle 2: Look for External Factors to Improve
Performance
Because specific attitudes, perceptions and beliefs related
to a given sustainable practice are difficult to identify and
change directly it is likely to be more effective, in the first
instance, to look for external factors influencing behaviour
independent of individual feelings, preferences, and
perceptions (Geller 2001). When interventions are
implemented which lead to changed individual behaviour,
indirectly, individuals change their attitude, commitment,
and internal motivation reflecting the reciprocity between
behavior and attitude (Bem 1970; Geller 2001). [See
reciprocal arrows in Figure 3.]
Principle 3: Focus on Positive Consequences to
Motivate Desired Behaviour
Most human behaviour is undertaken to gain a positive
consequence or to escape or avoid a negative
consequence. Humans learn more from their successes (ie
are more positively reinforced) than they learn from their
mistakes (Geller 2001). Geller contends that recognizing
people's environment-protective behavior will facilitate more
learning and positive motivation than will criticizing their
environment-damaging behavior. Ideally, to bring this
principle into play for increasing the adoption of sustainable
practices we need to identify NRM practices with relatively
immediate positive consequences rather than less
immediate, diffused, or short-term negative, consequences.
Practices which have outcomes that are ‘soon’ and ‘certain’
will have the most powerful motivating consequences
(Geller 2001). This suggests, that given the delay in
achieving sustainable environmental outcomes from many
NRM practices, the most effective practices will be those
with more immediate productive outcomes and
complementary (but more delayed) environmental
outcomes.

40
Consequences for adoption of
sustainable practices
A central tenet of the behaviourist or Skinnerian approach
is that behaviour is determined by its consequences and,
therefore, most people are unlikely to modify their
behaviour as the result of information or advice alone,
especially when the information pertains to a distant future
(Skinner 1987; Geller 2001). This is the conundrum (or
potential folly) for promoting sustainable practices, with low
or negative immediate benefits, on the basis of appeals to
future environmental sustainability.
Although people will often follow advice when
the advisor’s (or proponent’s) information
previously led to reinforcing consequences, this
situation requires people to experience the
reinforcing consequences of following the
advisor’s message. This type of learning . . .
is especially difficult when the future
consequences (reinforcing or punishing) are
unclear, uncertain, or remote. (Geller 2001)
Primary appeals to broad world views or abstract values
are unlikely to engender effective behaviour change
because such views are considerably removed, or often
disengaged, from everyday behaviour (see linkages in
Figure 3).8 Appeals such as ‘think globally, act locally’ are
not as psychologically powerful as appeals to ‘think locally,
act locally’.9 In the final chapter we discuss some
implications of these insights for research and development
and the adoption of sustainable practices derived from
research.
8 NRM practice promotional strategies, promoted primarily on the
basis of instrumental and more immediate benefits, can be
reinforced at an important secondary level by promotional
information regarding the environmental rationale for adopting
the practice. The secondary effect provides a long-term
reinforcement (Boyce & Geller (2001)).
9 And the global will obviously follow. The behaviourist view would
reverse the order of thinking and acting in the exhortation.

41
Implications for the focus of R & D
This chapter seeks to provide advice and options on
strategies for overcoming impediments to the adoption of
sustainable practices. The analysis in the previous two
chapters indicates that the most effective means for
encouraging sustainable practice adoption is to primarily
focus on the relevant behaviours and practices that
contribute to sustainable outcomes. Interventions to
change values and attitudes should be a secondary focus
because values and attitudes have indirect influence on
behaviour and their influence is commonly constrained
because NRM practices may often be complex or costly,
time consuming or characterised by delayed rewards.
Many practices may not be adapted or elaborated for local
conditions.
Adopters are adaptive
Human behaviour related to implementing NRM practices is
adaptive, rather than simply reactive, in its nature.
Landholders and farmers adapt their behaviour on the basis
of their experience. Appraisal and implementation of NRM
practices will depend on assessment of, and experience
with, the use of such practices. For landholders, the
difficulty of observing linkages between many
recommended NRM practices and desired sustainable
outcomes is likely to further reduce positive appraisals of
NRM practices by landholders.
Types of practices
It is the inherent characteristics of sustainable practices that
usually have the biggest influence on the rate of their
adoption by producers. Sustainable practices that provide
economic and other advantages will generally be adopted
more rapidly. There is a need to develop or identify the
practices that will produce desired sustainable outcomes
and be inherently attractive to potential adopters.
Landholders generally seek to reduce the risk of adopting a
new practice. Sustainable NRM practices which are
observable, trialable, and less complex are generally more
quickly adopted than NRM practices which are
unobservable, untrialable, and complex. Sustainable NRM
practices with environmental benefits are generally less
advantageous to the producer, more complex, harder to
trial and have benefits which are difficult to observe.
Pannell (2001b) has observed that the farm-level
economics of currently available management practices for
salinity prevention are adverse in many situations. As a
consequence, Pannell recommended both better targeting
of government programs, based on more rigorous analyses
of proposed public investments; and, more significantly in
the context of the discussion in this report, a greater
emphasis on the development of improved technologies,
both for salinity prevention and for adaptation to a saline
environment.
Research and development programs need to develop
NRM practices with relatively immediate positive
consequences rather than less immediate, diffused, or
short-term negative, consequences.
Effective R& D intervention means designing practices to
provide external benefits to make environment-sustaining
behavior more likely (Principle 2 in previous Chapter). The
most motivating consequences are ‘soon, certain, and
sizable’ (Geller 2001). The fall-back (and much more
expensive ) position is to otherwise change the external
conditions in order to make environment-sustaining
behavior more likely (Geller 2001). The latter strategy is
only feasible for government or institutional intervention
rather than R&D Corporations.
The attractiveness of practices is not independent
of the economic environment
The level of relative advantage is rarely independent of
commodity prices. The relative advantage of many
sustainable practices (such as deep-rooted perennials) will
be dependent on the value of rural commodities produced
as a result of using the practice. Low commodity prices in
the broadacre industries have reduced the relative
advantage of many sustainable practices. (See earlier case
study of adoption of dryland lucerne.)
Local adaptation
The relative advantage of sustainable practices varies in
different locations. It is dangerous to assume that a
practice with comparative advantages in one location will
yield the same level of advantage elsewhere. Few
sustainable practices have universal applicability. (See the
variation in geographic applicability for the sustainable
practices listed in Table 1.)
There are obvious advantages in being able to promote
sustainable practices with more universal or global
applicability. Firstly, messages can be simplified; secondly
and more important, where a given practice or
management behaviour is universally similar social
pressure can be more clearly brought to bear to ensure
behaviour maintenance. Social norms are easier to
establish for practices that are widely used and understood
than for locality specific behavior.
The advantage of generically global practices (for example,
small scale tree planting in higher rainfall areas), for
promoting and reinforcing sustainable practice adoption is
seductively attractive. However, given Australia’s diverse
environment, there are few sustainable practices which
meet the test of global applicability. And universally
applicable practices are often less likely to have large
impacts on reducing local land degradation problems. The
sustainable practices with wider geographic applicability,

42
such as currently available deep-rooted perennials, often
provide only moderate relative advantage to the landholder.
The relative advantage will be different in different localities.
As a consequence, every advantage should be taken of
sustainable practices that have widespread application.
But, more importantly, increased effort needs to be applied
to identify and develop locally applicable sustainable
practices and effort made to resist the temptation to
promote them beyond localities where their advantage has
been established. When, or if, local sustainable practices
are developed this approach can be thought of as locality
branding of practices in the same way that certain
agricultural products (such as wine) are locally branded.
This is an example of think locally act locally.
Assessing practices
To improve the likelihood that improved NRM practices will
be adopted by landholders R & D corporations have two
options. First, improved management systems can be
developed and adapted in conjunction with landholders in
relevant localities. Second, potential improved practices
being considered for promotion on farms, should be market
tested with typical farmers or landholders at any early stage
of their development. Such an approach encompasses
long established on-farm trials. However such assessment
should evaluate more than the traditional focus on technical
feasibility. Potential new NRM practices should be
assessed against each of the desired attributes of
geographic applicability, relative advantage, risk,
complexity, compatibility, observability and trialability
considered in Table 1. Ideally, a recommended NRM
practice should meet all these criteria for adoption in a
particular locality. Realistically, a recommended NRM
practice should meet as many of these criteria as is
possible, particularly relative advantage and observability.
After NRM practices are released and during early
promotion individual practices derived from research need
to be assessed – in terms of their attributes – as to why the
practices are, or are not, adopted.
Reinforcing learning
Given the outcomes from many NRM practices are not
readily observable (see observability characteristics in
Table 1) it is important to provide reinforcing learning
experiences in any promotional or extension campaigns.
Learning about, and adapting, NRM practices is especially
difficult when the future consequences (reinforcing or
punishing) are unclear, uncertain, or remote. Humans learn
more from their successes than they learn from their
mistakes and are more positively reinforced by their
successes (Principle 3 above). Therefore landholders often
need assistance to identify positive consequences of an
NRM practice as early as possible and to short circuit early
short-term failures.
An example of this approach has been the use of small on-
farm pilot demonstrations for the development (and local
elaboration) of conservation cropping techniques as part of
the Soil Care program (Wilkinson & Cary 1993; Barr & Cary
1992). The demonstrations and pilot development are
focused on local conditions, managed by a group of local
property holders and the responsibility for success or failure
is shared. Once the given practice is seen to be feasible
and advantageous to implement, individuals can do so
knowing the consequences are likely to be positive. Such
approaches help identify any immediate and positive
consequences of NRM practices which may not otherwise
be readily apparent.

43
Performance indicators and communication action plan
Performance indicators for assessing
the effectiveness of adoption of R&D
results
Key performance indicators are, ideally, direct measures of
desired producer behaviour related to sustainable NRM
practices. Alternatively, surrogate indicators may be used
which reflect correlations between selected indicators and
desired outcomes of more sustainable management
practices in relevant natural resources areas – including
biodiversity, salinity, soil erosion.
Useful indicators ideally should meet the following criteria:
• relate unambiguously to intended policy use
• the data are available at the required geographical
scale and frequency
• indicator can be easily understood
• be generalisable over wide geographical scale.
In practice selection of an indicator is usually determined by
the measures or indicators that are available. The non-
exhaustive list presented earlier in the section Sustainable
Practices identified potential performance indicators of
producer adoption.
Currently ABARE, through its Resource Management
Supplementary Survey, collects data, some of which
provide measures which, with some constraints, may be
appropriate indicators. This survey is ancillary to the
Australian agricultural and grazing industry and Australian
dairy industry surveys. The most recent Resource
Management Supplementary Survey was 1998-99;
however the Supplementary Surveys are irregular and
currently not undertaken on an on-going basis.
It should be recognised that the list presented earlier in the
section Sustainable Practices is an incomplete list. Many of
these practices were identified in the National Collaborative
Project on Indicators for Sustainable Agriculture based on
then available ABS and ABARE statistics. Measures of the
level of landholder adoption of sustainable management
practices available from the current ABARE Australian
Resource Management Supplementary surveys were
identified with the superscript a and cover generalised
situations, cropping farms, irrigation farms, the Rangelands
and dairy farms.
The above practice indicators are essentially output
indicators rather than outcome indicators (see Figure 1).
Research on social and other characteristics, based on
useful indicators, which might indicate sustainable practice
adoption is made difficult by the fact that sustainable
practices have differential advantage to landholders in
different localities and, often, differential advantage to
different landholders within given localities. We are faced
with the unsatisfactory choice of localised studies which
lack uniformity in approach or more extensive studies (the
ABARE RMS surveys) which currently collect simplified
measures of the presence or absence of selected practices
on properties rather than the extent of use of practices. As
yet there is a very limited theoretical basis for postulating
relationships, and guiding data collections, regarding
adoption behaviour. Further work is required to design a
suitable program of longitudinal data collection regarding
adoption behaviour across and within localities to take
account of the variability of Australian agriculture.
The ideal we should be seeking is independently measured
physical evidence of the extent of adoption of nominated
practices, rather than of use of practices self-reported by
landholders and collected by survey.
The following issues, identified as important in relation to
the adoption by producers of sustainable practices derived
from research, present significant obstacles to establishing
generalisable performance indicators:
• importance of locality specificity
• need to refine characteristics explaining adoption
• need to focus on locality and complexity of individual
NRM practices to explain adoption.
As a consequence NRM resource questions included in the
standard Australian Agricultural Census (AAC) question
suite need to be segmented for localities and industries. It
is likely that, in most cases, the industry specific questions
in the ABARE RMS survey on sustainable practices are too
broadly based to overcome the problem of locality
specificity.
The development of a suite of natural management
resource questions as part of the standard AAC question
suite is recommended. The inclusion of NRM questions
within the AAC is for the most part dependent upon external
user funding. NRM questions are not considered to be of
prime importance in the development of the AAC. The need
to capture this user funding, together with the relatively
short timelines allowed for the development of the census
form, means that the inclusion of a good suite of NRM
questions is potentially a matter of circumstance rather than
design. We recommend a reassessment of the relative
importance of NRM issues in the census question suite to

44
ensure that a solid longitudinal data set of NRM
management across Australia is achieved.
Communication action plan
A communication action plan should comprise the following
steps:
• recognise the barriers to implementation of findings of
and implications of this report
• identify areas where implications and findings from the
report might be implemented
• design a communication program to overcome these
barriers.
Barriers to implementation of findings:
The most important message to be communicated from this
report is the achievement of desired sustainable outcomes
depends on more widespread adoption and use of
sustainable practices. Relatively few current sustainable
practices possess attributes that make them attractive for
widespread adoption by landholders. The rate of adoption
of sustainable practices derived from research depends on
practices being economically attractive to adopt. Newly
developed sustainable practices need to provide
observable and positive consequences for landholders over
a short time frame.
Most research for the development of sustainable practices
is discipline driven. Thus the focus is on biological or
technical feasibility rather than complementary
consideration of adoptability. A broader focus requires one
or more of clear directives in research objectives, a broader
programmatic approach, or a systems approach to
sustainable practice research. Linkages for on-farm
commercial adaptation need to be established in such
research programs. The impact of longer term commodity
prices cannot be ignored in making assessments about
implementation of sustainable practices.
Areas where findings can be
implemented
Land & Water Australia with its recent broader thematic
R&D program with explicit social and institutional focus is
well placed to respond to the findings of this report.
However as much sustainable practice research and
development is conducted in other industry-based R&D
corporations it is important that the messages from this
research are communicated to these bodies and other
relevant research agencies.
Communication strategies
1. Preparation of simple check-lists for assessing the
attributes and characteristics required (for more rapid
adoption) when seeking to develop new sustainable
practices.
2. Presentation of a seminar or workshop for rural R&D
project managers to elaborate the findings of this
project.
3. Identify more clearly the localities where given
sustainable practices have a comparative advantage
for landholders.
4. Promote the idea of more integrated and formal
‘market testing’ of prototype sustainable practices in
the development and local adaptation stages.
5. Develop support systems to ensure more rapid ‘trial
and error’ learning for landholders by the use of small
on-farm pilot demonstrations, where appropriate, for
the development and local elaboration of new
developed sustainable practices.

45
References
ABARE (2000). Australian Farm Surveys Report 2000.
Canberra: ABARE.
Abelson, R.P. (1979). Differences between belief and
knowledge systems. Cognitive Science, 3: 355-366.
Alston, M. (1995). Women on the Land: The Hidden Heart
of Rural Australia. Kensington: University of New South
Wales Press.
Bamberry, G., Dunn,T., & Lamont, A. (1997). A Pilot Study
of the Relationship between Farmer Education and
Good Farm Management. Rural Industries Research
and Development Corporation (Publication No. 97/30),
Barton, ACT.
Barker, J. (1992). Lucerne intercropping. Paper presented
to the Water efficient farming workshop. Bendigo:
Centre for Land Protection Research.
Barr, N. F., & Cary, J. W. (1992). Greening a Brown Land:
The Australian Search for Sustainable Land Use,
Melbourne: Macmillan.
Barr, N.F., & Cary, J.W. (2000). Influencing Improved
Natural Resource Management on Farms: A guide to
understanding factors influencing the adoption of
sustainable resource practices. Canberra: Bureau of
Rural Sciences.
http://www.brs.gov.au/social_sciences/nat_resource_m
gt.pdf
Barr, N., Ridges, S., Anderson, N., Gray, I., Crockett, J.,
Watson, B., & Hall, N. (2000). Adjusting for Catchment
Management: Structural Adjustment & its Implications
for Catchment in the Murray-Darling Basin. Dryland
Technical Report No. 2. Canberra: Murray-Darling
Basin Commission.
Bem, D.J. (1970). Beliefs, Attitudes and Human Affairs.
Monterey CA: Brooks Cole.
Bodnaruk, K. & Frank, B. (1997). Factors influencing the
acceptance of IPM in the Australian apple industry. In
proceedings of the 2nd Australasia Pacific Extension
Conference. Albury: Australasia Pacific Extension
Network. pp. 232-8.
Boyce, T. E., & Geller, E. S. (2001). Applied behavior
analysis and occupational safety: The challenge of
response maintenance. Journal of Occupational
Behavior Management. (forthcoming)
Bryant, L. (1999). The detraditionalisation of occupational
identities in farming in South Australia. Sociologia
Ruralis, 39, 2: 236-61.
Carboni, S.M. & Napier, T.L. (1993). Factors affecting use
of conservation farming practices in East Central Ohio.
Agriculture, Ecosystems and Environment, 45, 1: 79-94.
Cary, J. W. (1991). Symbolic Beliefs and the Construction
of Environmental Knowledge. Procs. IX International
Congress on Personal Construct Psychology, Albany,
New York.
Cary, J.W. (1992). Lessons from past and present
attempts to develop sustainable land use systems.
Review of Marketing & Agricultural Economics, 60(2):
57-64.
Cary, J.W. (1993). The nature of symbolic beliefs and
environmental behaviour in a rural setting. Environment
& Behavior, 25, 555-576.
Cary, J.W. (1999). Successful soil and landscape
conservation in Australia. In T. L. Napier, S. M.
Camboni, & J. Tvrdon (Eds.) Soil and Water
Conservation Policies: Successes and Failures (pp.
658-671). Ankeny, Iowa: Soil and Water Conservation
Society of America.
Cary, J.W. (2000). Social Perspectives on Technology
Transfer to Meet Economic & Environmental
Objectives. In David White & Jim Walcott (Eds.)
Emerging Technologies in Agriculture: From Ideas to
Adoption. Canberra: Bureau of Rural Sciences.
http://www.brs.gov.au/agrifood/confEtiA7-
00/papers/JCary_paper.pdf
Cary, J.W., & Webb, T. (2000). Community landcare, the
National Landcare Program, & the landcare movement:
The social dimensions of landcare. Canberra: Bureau of
Rural Sciences.
http://www.affa.gov.au/corporate_docs/publications/pdf/
rural_science/social_science/landcare.pdf
Cary, J.W., & Wilkinson, R.L. (1997). Perceived profitability
& farmers’ conservation behaviour. Journal of
Agricultural Economics, 48 (1): 13-21.
CIE (2001). Indicators Within a Decision Framework:
Social, Economic Institutional Indicators for Sustainable
Management of the Rangelands. Report for the
National Land & Water Resources Audit, Canberra.
Clancy, B. (1999). ‘Elders wool plan’. The Weekly Times,
Melbourne: Herald and Weekly Times, pp. 1-2.
Coffey, M. (1992). Farmers’ perceptions of upgrading
pasture to realise the potential of their grazing land. 4th
year B.Ag.Sci. project, La Trobe University, Bundoora.

46
Curtis, A.L., & De Lacy, T. (1996). Landcare in Australia:
Does it make a difference? Journal of Environmental
Management, 46 (2), 119-137.
Curtis, A., MacKay, J., Van Nouhuys, M. Lockwood, M.,
Byron, I., & Graham, M. (2000). Exploring landholder
willingness & capacity to manage dryland salinity: The
Goulburn Broken Catchment. Albury: Johnstone
Centre, Charles Sturt University.
Drake, L., Bergstrom, P. & Svedsater, H. (1999). Farmers’
attitudes and uptake. In G. Van Huylenbroeck & M.
Whitby (Eds.) Countryside Stewardship: Farmers,
Policies and Markets. Amsterdam: Pergamon.
Dunn, T., Gray, I., & Phillips, E. (Forthcoming). From
personal barriers to community plans: A farm and
community planning approach to the extension of
sustainable agriculture. A. D. Schulman, & R. J. Price
Case Studies in Increasing the Adoption of Sustainable
Resource Management Practices (pp. 17-33).
Canberra: Land & Water Resources Research and
Development Corporation.
Fenton, D.M., MacGregor, C., & Cary, J.W. (2000).
Framework & Review of Capacity & Motivation for
Change to Sustainable Management Practices. (Land &
Water Resources Audit Report, Project 6.2.1).
Canberra: Bureau of Rural Sciences.
ftp://ftp.nlwra.gov.au/pub/public_v2/30_projects/48_the
mes/6_capacity_for_change/10_project_2_1/621_final_
report.pdf
Geller, E.S. (2001). The Challenge of Increasing Pro-
Environment Behavior. In R. B. Bechtel and A.
Churchman (Eds.) The New Environmental Psychology
Handbook, New York: John Wiley & Sons
Guagnano, G.A., Stern, P.C. & Dietz, T. (1995). Influences
on attitude-behavior relationships: A natural experiment
with curbside recycling. Environment and Behaviour 27:
669-718.
Griliches, Z. (1957). Hybrid corn: An exploration in the
economics of technological change. Econometrica, 25,
501-522.
Griliches, Z. (1960). Hybrid corn and the economics of
innovation. Science, 132, 275-280.
Gould, B., Saupe, W., & Kleme R. (1989). Conservation
tillage: The role of farm and operator characteristics
and the perception of soil erosion. Land Economics,
65(2), 167-182.
Guerin, L.J., & Guerin, T.F. (1994). Constraints to the
adoption of innovations in agricultural research &
environmental management: a review. Australian
Journal of Experimental Agriculture, 34, 4: 549-71.
Haberkorn, G., Hugo, G., Fisher, M., & Aylward, R. (1999).
Country Matters: Social Atlas of Rural & Regional
Australia. Bureau of Rural Sciences, Canberra.
http://www.affa.gov.au/docs/rural_science/social_scienc
e/atlas.html
Hamblin, A. (1999). Options to quantify the quality & range
soil, water & vegetation management practices
throughout Australia. National Land & Water Resources
Audit Scoping Report.
http://www.nlwra.gov.au/full/30_themes_and_projects/5
0_scoping_projects/04_methods_papers/12_Hamblin/M
gmt_Practices.html
Harvey, J. T., & Hurley, F. T. (1990). Cropping and
Conservation. Ballarat, Victoria: Regional Studies Unit,
Ballarat University College.
Holland, J.H. (1995). Hidden Order: How Adaptation Builds
Complexity. Reading, Massachusetts: Addison-Wesley.
Hosmer, D.W. & Lemeshow, S. (1989). Applied Logistic
Regression. New York: John Wiley & Sons.
Howden, P., Vanclay, F., Lemerle, D. & Kent, J. (1997).
Understanding farming styles. In proceedings of the
2nd Australasia Pacific Extension Conference. Albury:
Australia Pacific Extension Network. pp. 269-76.
Karunaratne, K. and Barr, N. (2001). A baseline of adoption
of pasture management practices: North Central
Region. DNRE Victoria.
Kilpatrick, S. (1996). Training Australia’s Farm Workforce.
Paper presented to the National Centre for Vocational
Education Research Conference, Launceston, 9-12
July.
Kilpatrick J. (1997). Effective Delivery Methodologies for
Education and Training to Rural Australia. Centre for
Research and Learning in Regional Australia.
University of Tasmania, Launceston.
Kilpatrick, S., Johns, S., Murray-Prior, R., & Hart, D. (1999).
Managing Farming: How Farmers Learn. Rural
Industries Research and Development Corporation
(Publication No. 99/74), Barton, ACT.
Kilpatrick, S. & Rosenblatt, T. (1998). Information vs
training: Issues in farmer learning. European Journal of
Agricultural Education and Extension, 51(1), 39-51.
Kilpatrick, S. & Williamson J. (1996). Farmer participation in
training. Rural Society, 6 (4), 3-12.
Marks, N. & O’Keefe, M. (1996). Pastures and farm
management practices: research report prepared for
Agriculture Victoria. [Unpublished].
Marsh S.P. & Pannell, D.J. (2000). Agricultural extension
policy in Australia: the good, the bad & the misguided.
Australian Journal of Agricultural & Resource
Economics, 44 (4): 605-627.

47
Mobbs, C. & Dovers, S. (Eds.) (1999). Social, Economic,
Legal, Policy and Institutional R&D for Natural
Resource Management: Issues and Directions for
LWRRDC. Canberra: LWRRDC.
Mues, C., Chapman, L. & Van Hilst, R. (1998). Landcare:
Promoting Improved Land Management Practices on
Australian Farms: A Survey of Landcare and Land
Management Related Programs. Canberra: ABARE.
Nassauer, J.I. (1995). Messy ecosystems, orderly frames.
Landscape Journal, 14 (2), 161-169.
Oppenheim, A.N. (1992). Questionnaire Design,
Interviewing and Attitude Measurement 2nd ed.
London: Pinter.
Oxley, T. (1997). Dryland Lucerne: Establishment and
Management in Northern Victoria, 4th year project,
University of Melbourne.
Pannell, D. (2001a). Cited in ‘Paying the penalty’ The
Bulletin, 9 January, p.24. See also Pannell, D.J.,
(2000). ‘Salt Levy? The Complex Case for Public
Funding of Salinity’, SEA Working Paper 2000/01,
Department of Agriculture & Resource Economics,
University of Western Australia.
Pannell, D.J. (2001b). Dryland Salinity: Economic,
Scientific, Social and Policy Dimensions, Australian
Journal of Agricultural and Resource Economics (in
press).
Phillips, E. (forthcoming). Social & Cultural Factors that
Influence the Adoption of Sustainable Farm Practices.
In: A.D Schulman & R.J Price. (Eds.) Case Studies in
Increasing the Adoption of Sustainable Resource
Management Practices. Canberra: Land & Water
Resources Research & Development Corporation, pp.
48-57.
Phillips, T.I. (1985). The Development of Methodologies
for the Determination & Facilitation of Learning for Dairy
Farmers. School of Agriculture & Forestry, University of
Melbourne: Melbourne.
Ransom, K. (1992). Dryland lucerne grazing systems:
development, demonstration and computer simulation
of grazing systems to utilise mixed pastures of sub
clover, annual grasses and dryland lucerne. Bendigo:
Department of Agriculture Victoria.
Ransom, K. & Barr, N. (1993). The adoption of dryland
lucerne in North-Central Victoria. Bendigo: Department
of Agriculture Victoria.
Reeve, I.J., & Black, A.W. (1993). Australian Farmers’
Attitudes to Rural Environmental Issues. Armidale:
University of New England, Rural Development Centre.
Reeve & Black (1998). Improving Farmers’ Management
Practices through Learning and Group Participation.
Barton, ACT: Rural Industries Research and
Development Corporation (Publication No. 98/71).
Rendell, R.J., O’Callaghan, P. & Clark, N. (1996).
Families, Farming and the Future: Business
performance indicators for farming systems in the
Wimmera and Mallee of Victoria. Bendigo: Agriculture
Victoria.
Riley, D. (1999). The Australian beef industry: Farmers’
attitudes & practices. Australian Farm Surveys Report.
Canberra: ABARE.
RIRDC/DPIE (1998). Missed Opportunities: Harnessing the
Potential of Women in Australian Agriculture. Volume 1.
Social Survey and Analysis. Canberra: RIRDC/DPIE.
Rogers, E.M. (1962). Diffusion of innovations. New York:
The Free Press of Glencoe.
Rogers, E.M. (1983). Diffusion of innovations (3rd edn.).
New York: The Free Press-Collier Macmillan.
Rogers, E.M. & Shoemaker, F.F. (1971). The
communication of innovations: a cross cultural
approach. New York: Collier-Macmillan.
Ross, H. (1999). Social R&D for sustainable natural
resource management in rural Australia: Issues for
LWRRDC. C. Mobbs, & S. Dovers (Eds.), Social,
Economic, Legal, policy and Institutional R&D for
Natural Resource Management: Issues and Directions
for LWRRDC (pp. 22-41). Canberra: LWRRDC.
Saunders, D., Margules, C., & Hill, B. (1998).
Environmental Indicators for State of the Environment
Reporting: Biodiversity. Australia: State of the
Environment (Environmental indicators reports).
Department of Environment, Canberra.
SCA (1991). Sustainable Agriculture. SCA Technical
Report No. 36. Melbourne: CSIRO Publishing.
SCARM (1998). Sustainable Agriculture: Assessing
Australia’s Recent Performance. Report of National
Collaborative Project on Indicators for Sustainable
Agriculture. SCARM Technical Report No. 70.
Melbourne: CSIRO Publishing.
Sinden, J.A. & King, D.A. (1990). Adoption of soil
conservation measures in Manilla Shire, New South
Wales. Review of Marketing and Agricultural
Economics, 58, 179-92.
Skinner, B.F. (1953). Science and Human Behavior. New
York: Macmillan.
Skinner, B.F. (1974). About Behaviorism. New York:
Alfred A. Knopf.
Skinner, B.F. (1987). Upon Further Reflection. Englewood
Cliffs, NJ: Prentice-Hall.

48
Solutions Marketing and Research Group (1999).
‘Agriculture - Advancing Australia’ Research report.
Prepared for Agriculture, Fisheries and Forestry -
Australia. Sydney: Solutions Marketing and Research
Group.
Soros, G. (2000). Open Society: Reforming Global
Capitalism. London: Little, Brown and Company.
Stern, P.C. (1992). Psychological Dimensions of Global
Environmental Change. Annual Review of Psychology
43: 269-302.
Stern, P.C. & Dietz, T. (1994). The Value Basis of
Environmental Concern. Journal of Social Issues 50:
65-84.
Stern, P.C., Dietz, T., Abel, T., Guagnano, G.A., & Kalof, L.
(1999). A value-belief-norm theory of support for social
movements: The case of environmentalism. Human
Ecology Review, 6(2), 81-97.
Stern, P.C., Dietz, T. & Guagnano, G.A.. (1995). The New
Ecological Paradigm in Social-Psychological Context.
Environment and Behaviour 27: 723-43.
Stern, P.C., Dietz, T. & Kalof, L. (1993). Value Orientations,
Gender and Environmental Concern. Environment and
Behaviour 25: 322-48.]
Synapse Consulting (1998). Farmers Education and
Training: Issues for Research and Development. Rural
Industries Research and Development Corporation
(Publication No. 98/26), Barton, ACT.
Vanclay, F. (1988). Socio-economic Characteristics of
Adoption of Soil Conservation. Unpublished Master of
Social Science thesis, University of Queensland, St
Lucia, Queensland.
Vanclay, F. & Lawrence, G. (1995). Farmer rationality and
the adoption of environmentally sound practices: A
critique of the assumptions of traditional agricultural
extension. European Journal of Agricultural Education
and Extension, 1, 1: 50-90.
Webb, T (2000a). Evaluation of the ‘Women as Clients’
Partnership Agreement Between WIRIS and NRM.
Bureau of Rural Sciences, Canberra.
Webb, T (2000b). Evaluation of the WIRIS-ATO
Communication Partnership for the new Tax System.
Bureau of Rural Sciences, Canberra.
Whittet, J.N. (1929). Lucerne as pasture in the western
districts: Recent investigations into establishing and
grazing the stands. Agricultural Gazette, 40, 1-16.
Wilkinson, R. (1996). Resource Monitoring by Hawke’s
Bay Farmers. Canterbury: Manaaki Whenua Press.
Wilkinson, R.L. & Cary, J.W. (1992). Monitoring Landcare in
Central Victoria. Parkville, Victoria: School of
Agriculture and Forestry, University of Melbourne.
Wilkinson, R.L. & Cary, J.W. (1993). Monitoring Soilcare in
Northeast Victoria, School of Agriculture and Forestry,
University of Melbourne.
Wilkinson, R.L. & Cary, J.W. (2001). Sustainability as an
Evolutionary Process. International Journal of
Sustainable Development. (forthcoming)
Wilson, A. & Tyrchniewicz, A. (1995). Agriculture and
sustainable development : policy analysis on the Great
Plains. Winnipeg: International Institute for Sustainable
Development.

49
Appendix A Analysis of the 1998-99 Resource Management
Survey
Logistic regression
Using data collected from their 1998-99 annual survey of
Australia’s broadacre and dairy industries ABARE modelled
the relationship between a range of farm family, farm
property and farm business characteristics and reported
adoption of sustainable land management practices. An
overview of the analysis was presented earlier, with tables
4 and 5 indicating the independent (farm/farmer
characteristics) and dependent (practice adoption)
variables respectively.
The dependent variable was represented as binary data
(1=practice adoption, 0=no adoption). The independent
variables were represented as nominal (including binary)
and interval data (see Appendix B). In modelling a binary
dependent variable a logistic regression technique is the
most appropriate approach.
The logistic model models the log odds ratio or “logit” of the
dependent variable (that is ln[p/(1-p)] where p is the
probability that adoption of a practice occurs) as a linear
combination of a series of explanatory variables. This gives
the following function: ln[p/(1-p)] = b0 + b1X1 + b2X2 + b3X3 +
… + bkXk + e. Where Xi are the explanatory variables, bi are
the coefficients to be estimated and e is an error term. In
these analyses the logit of adoption is modelled as a linear
combination of a range of farm family, farm property and
farm business characteristics.
This approach is considered superior to ordinary least
squares regression (OLS) for modelling binary dependant
variables. The logistic model guarantees that the predicted
probability of adoption is between zero and one.
Furthermore with a binary dependent variable, the
assumptions of normally distributed and homoskedastic
residuals required for OLS are both broken (Hosmer &
Lemeshow 1989).
Model estimation
ABARE surveys a sample of farms out of the total number
of farms available. A survey sample of 1426 was selected
for the 1998-99 survey, of which 197 respondents were
interviewed by telephone with the remaining respondents
interviewed in person. The sample covers the five
broadacre industry types and the three farming zones. In
order to reflect the actual composition of Australian farms
the sample is weighted prior to data analysis. Sample
weights are calculated so that population estimates of farm
numbers, crop areas and livestock numbers in various
geographical regions and industries correspond as closely
as possible to known Australian Bureau of Statistics census
data (ABARE 2000).
Prior to model estimation the samples were weighted so
that the results could be inferred to the broader industries
rather than just the survey sample. The models include two
benchmark variables (sheep closing number and wheat
area sown) which were used to ensure appropriate sample
weighting.
The farm sample is restricted by selecting the sample from
those farms with an estimated value of agricultural
operations (EVAO) of $22,500 or greater. Farms with an
EVAO of less than this level were excluded from the
survey. Accordingly the results presented here will be
inaccurate to the extent that farms with an EVAO of less
than $22,500 are excluded.
The models were estimated using maximum likelihood
regression. Prior to model estimation correlation matrices of
continuous variables and frequency cross-tabulations of
discrete variables were checked for high levels of
correlation that may give rise to problems of
multicollinearity. Where potential problems existed
individual variables were removed from models to gauge
impact. There were some issues of multicollinearity,
however statistical advice from ABARE indicated it would
not unduly affect the models.
In estimating the models, Wald’s Chi-square and its
corresponding probability were used to indicate the
confidence level for which each independent variable is
significantly associated with the logit function. For the
purposes of this analysis a confidence level of 95% was
chosen to indicate a significant association.
The likelihood ratio test was used to indicate the
significance of the overall model. The association of
predicted probabilities and observed responses was also
calculated as a measure of model ‘fit’. The higher the
proportion of concordant associations the better the model
is at predicting farmer behaviour based upon the variables
included in the model.
Results
Rather than explore the findings in terms of dependent
variables as discussed earlier, this section details each
individual model estimated. For each practice the logit
regression results are provided. These indicate which
explanatory variables were included in each model and the

50
parameter estimates indicate the strength of the association
between that variable and the logit function. Those in bold
indicate a significant association between the characteristic
and practice adoption as indicated by Wald’s Chi-square.
Additionally we compare and contrast these findings with
other recent comparable studies. In particular we draw
upon recent work carried out in the rangelands for the Audit
by the CIE (2001), work by Curtis et al. (2000) in Victoria’s
Goulburn-Broken catchment and work by Mues, Chapman
and Van Hilst (1998). Note that the measures of practice
adoption may have been operationalised differently in each
study, and that subsequent analyses performed were often
upon different subsections of the farming community. Mues,
Chapman and Van Hilst’s (1998) work is the most directly
comparable, being based upon data collected during the
1995-1996 AAGIS/ADIS surveys and a landcare
supplementary survey, though their model specifications
were different. Thus while some relationships have been
confirmed in different studies, differences can, in part, be
explained as an artefact of study specificity and design.
Controlled flow bores in the pastoral
zone
Three variables were significantly associated with the
control of flowing bores in the pastoral zone (Table A1).
Overall the model was significant and gave 91.9%
concordance between predicted probabilities and observed
responses. Owner/managers with higher closing equity
ratios (ie with lower relative debt burdens) were more likely
to control flowing bores than those with lower equity levels.
Owner/managers who thought the future profitability of their
property was likely to drop were less likely to adopt than
those who considered their future profitability to increases
or remain the same in the next five to 10 years. Landcare
membership was also positively associated with adoption.
No other variables were found to be significantly associated
with the decision to control flowing bores in the pastoral
zone.
Importantly the capping of bores is of primary relevance to
Queensland pastoralists where uncontrolled bore flows are
more of an historical artefact. However in this model the
dummy variable for Queensland was found not to be
significantly associated with practice adoption.
CIE (2001) in its modelling of indicators of sustainable
practice in the rangelands, while recording the practice, did
not model this behaviour.
Table A1 Logit regression results for the adoption of controlled flow bores in the pastoral zone.
Effect Unit DF Parameter
Estimate
Std Error Chi-
square
Prob
Intercept 1 -11.7300 542.2000 0.00 0.9827
Age yrs 1 0.0302 0.0393 0.59 0.4428
Environmental concern attitude 1 0.4479 0.4261 1.11 0.2931
Financial concern attitude 1 0.0414 0.4945 0.01 0.9332
Technical concern attitude 1 0.3535 0.5435 0.42 0.5154
Financial outlook concern attitude 1 -1.0012 0.4330 5.35 0.0208
Landcare membership- yes 1 1.7601 0.8578 4.21 0.0402
Length of landcare membership yrs 1 0.3801 0.2126 3.20 0.0737
Recent training 1 0.2114 0.4389 0.23 0.6301
Farm cash income $ 1 0.0000 0.0000 0.07 0.7883
Profit at full equity $ 1 0.0000 0.0000 0.91 0.3396
Closing equity ratio % 1 0.0523 0.0264 3.94 0.0471
Farm plan / property manag plan - yes 1 0.1132 0.5084 0.05 0.8238
Farm size 000/ha 1 -0.0027 0.0037 0.51 0.4741
State 4 5.19 0.2684
-NSW -0.6409 542.2000 0.00 0.9991
-QLD 5.7524 542.2000 0.00 0.9915
-SA 4.8158 542.2000 0.00 0.9929
-WA -14.5163 2168.7000 0.00 0.9947
Land use intensity se/ha 1 -1.6158 0.8584 3.54 0.0598
PMP participation last 3 years - yes 1 0.02200 0.7973 0.00 0.9780
Wheat area sown ha 1 0.0027 0.0033 0.66 0.4151
Closing number of sheep no 1 -0.0001 0.0001 1.14 0.2841
Sample size=125, adoptors=33.
Model fit likelihood ratio test Chi-square=74.94, DF=21, P<0.0001
Association of predicted probabilities and observed responses: concordant=91.9%, discordant=7.9%, tied=0.2%

51
Controlling grazing pressure by
excluding access to water in the pastoral
zone
No variables were significantly associated with the control
of grazing pressure by excluding access to water (Table
A2).
Research concerning practice adoption in the rangelands
carried out by the CIE (2001) did not consider the control of
grazing pressure by excluding stock access to water.
Table A2 Logit regression results for the control of grazing pressure by excluding access to water in the
pastoral zone.
Effect Unit DF Parameter
Estimate
Std Error Chi-
square
Prob
Intercept 1 -0.7370 1.8381 0.16 0.6885
Age yrs 1 -0.0113 0.0191 0.35 0.5551
Environmental concern attitude 1 0.2931 0.2526 1.35 0.2458
Financial concern attitude 1 -0.1086 0.3106 0.12 0.7266
Technical concern attitude 1 0.3379 0.3206 1.11 0.2919
Financial outlook concern attitude 1 -0.4422 0.2321 3.63 0.0567
Landcare membership- yes 1 -0.4226 0.4229 1.00 0.3177
Length of landcare membership yrs 1 0.0246 0.0916 0.07 0.7884
Recent training 1 -0.1399 0.2478 0.32 0.5723
Farm cash income $ 1 0.0000 0.0000 0.87 0.3513
Profit at full equity $ 1 0.0000 0.0000 0.89 0.3466
Closing equity ratio % 1 -0.0003 0.0069 0.00 0.9687
Farm plan / property manag plan - yes 1 0.1586 0.2567 0.38 0.5366
Farm size 000/ha 1 0.0030 0.0024 1.47 0.2240
State 4 9.18 0.0568
-NSW 0.5355 0.4851 1.22 0.2696
-QLD -0.5937 0.4370 1.85 0.1742
-SA -1.3306 0.6734 3.90 0.0482
-WA 0.8751 0.7238 1.46 0.2267
Land use intensity se/ha 1 0.2104 0.2520 0.70 0.4037
PMP participation last 3 years - yes 1 0.1696 0.3345 0.26 0.6120
Wheat area sown ha 1 0.0005 0.0010 0.23 0.6308
Closing number of sheep no 1 0.0000 0.0000 0.08 0.7767
Sample size=186, adoptors=71.
Model fit Likelihood ratio test Chi-square=41.40, DF=21, P=0.0050
Association of predicted probabilities and observed responses: concordant=68.3%, discordant=31.4%, tied=0.3%
Monitoring pasture and vegetation
condition in the pastoral zone
Seven variables were significantly associated with
monitoring pasture and vegetation condition in the pastoral
zone at the 95% confidence level (Table A3). Three
financial variables, profit at full equity, closing equity ratio
and financial outlook concern attitude were significantly
associated with monitoring pasture and vegetation
condition. However not all of the relationships were in the
predicted direction. While owner/operators with increased
profit levels were more likely to adopt the practice, those
who had higher levels of debt and those that thought their
future farm profitability would decrease were also more
likely to adopt the practice.
Owner operators with a farm plan or a property
management plan were more likely, than those without
plans, to adopt the practice. State of residence was
significantly associated with pasture and vegetation
monitoring, with Queensland based owner/managers being
significantly less likely to adopt than the Australian average
and those in Western Australia more likely to adopt the
practice.
Two other attitude variables, the environmental concern
and technical concern attitudes were significantly
associated with adoption. Those owner/operators who
considered water and land degradation to be a critical issue
in their farm planning were more likely to adopt.
Owner/operators who considered they did not have the
technical resources to adequately address land and water
degradation were less likely to adopt than those that
considered they did have the technical resources.

52
Mues, Chapman and Van Hilst (1998) considered this
practice, however none of the variables they found to have
significant associations (recent training, degradation, land
use intensity and farm size) were significant in the model
tested here.
CIE (2001) considered pasture monitoring in its study in the
rangelands. CIE found confirming results with respect to
state of residence with graziers resident in Western
Australia more likely to adopt, and those from Queensland
less likely to adopt. They also found a positive association
between having a farm plan and practice adoption. CIE also
found that age was significantly and negatively associated
with practice adoption.
Table A3 Logit regression results for the adoption of monitoring pasture and vegetation condition in the
pastoral zone.
Effect Unit DF Parameter
Estimate
Std Error Chi-
square
Prob
Intercept 1 -0.8746 2.0165 0.19 0.6645
Age yrs 1 -0.0334 0.0212 2.48 0.1150
Environmental concern attitude 1 0.7963 0.2553 9.73 0.0018
Financial concern attitude 1 0.5074 0.3001 2.86 0.0909
Technical concern attitude 1 -0.6737 0.3070 4.82 0.0282
Financial outlook concern attitude 1 0.7454 0.2579 8.35 0.0038
Landcare membership- yes 1 -0.1692 0.4469 0.14 0.7049
Length of landcare membership yrs 1 0.0983 0.0972 1.02 0.3121
Recent training 1 0.2385 0.2692 0.79 0.3756
Farm cash income $ 1 0.0000 0.0000 2.50 0.1138
Profit at full equity $ 1 0.0000 0.0000 7.27 0.0070
Closing equity ratio % 1 -0.0199 0.0081 5.97 0.0145
Farm plan/property man plan - yes 1 1.3614 0.3286 17.17 <.0001
Farm size 000/ha 1 -0.0019 0.0021 0.74 0.3882
State 4 13.73 0.0082
-NSW -0.5199 0.5393 0.93 0.3351
-QLD -1.7715 0.5901 9.01 0.0027
-SA -0.5725 0.6702 0.73 0.3930
-WA 1.9204 0.6629 8.39 0.0038
Land use intensity se/ha 1 0.0461 0.4443 0.01 0.9174
PMP participation last 3 years - yes 1 -0.1794 0.3716 0.23 0.6294
Wheat area sown ha 1 -0.0229 0.0214 1.14 0.2848
Closing number of sheep no 1 0.0000 0.0001 0.04 0.8399
Sample size=203, adoptors=82
Model fit Likelihood ratio test Chi-square=116.53, DF=21, P<0.0001
Association of predicted probabilities and observed responses: concordant=81.4%, discordant=18.2%, tied=0.4%
Deep rooted perennial pasture in the
wheat-sheep and high rainfall zones
(broadacre industries only)
Seven variables were associated with the adoption of deep
rooted perennial pasture in the wheat-sheep and high
rainfall zone (Table A4). Farm size and closing equity ratio
were both negatively associated with adoption.
Owner/managers with larger farms or with lower debt ratios
were less likely to adopt than those with smaller farms or
higher debt ratios. Involvement in Property Management
Planning in the last three years was significantly associated
with higher levels of adoption.
Three attitude variables were also significantly associated
with adoption. Owner/managers who considered they did
not have the technical resources to deal with water and
land degradation or those who thought their farm
profitability would drop in the next five to 10 years were less
likely to plant deep rooted perennial pasture. However
those who considered they did not have the financial
resources to adopt were less likely to adopt than those who
considered they did have the financial resources.
State of residence was an important explanatory variable;
New South Wales was more likely to adopt than the
average for Australia. Victoria, Queensland and Western
Australia were all less likely to adopt than the average for
Australia.
Cary and Wilkinson (1997) found an association between
farm size and planting deep-rooted pasture species, and a
weak association with length of Landcare membership.
Curtis et al. (2000) found significant associations between

53
the area sown to introduced perennial pastures and a
number of explanatory variables. However the practice was
operationalised in a different manner, and this may explain
differences between the findings. Curtis et al. (2000) found
property size to be significantly related, however in their
case larger properties were more likely to have planted
greater areas to deep rooted perennial pasture, while in the
analysis of the RMS data here larger properties were more
likely not to have made the discrete decision to adopt the
practice. Curtis et al. (2000) also found landcare
membership to be negatively associated with adoption,
where the ABARE analysis did not find a significant
association.
Table A4 Logit regression results for the adoption of deep rooted perennial pasture in the wheat-sheep
and high rainfall zones (broadacre industries only).
Effect Unit DF Parameter
Estimate
Std Error Chi-
square
Prob
Intercept 1 1.5385 0.6976 4.86 0.0274
Age yrs 1 -0.0007 0.0070 0.01 0.9262
Environmental concern attitude 1 0.0060 0.0759 0.01 0.9372
Financial concern attitude 1 0.1968 0.0787 6.26 0.0124
Technical concern attitude 1 -0.3272 0.0857 14.58 0.0001
Financial outlook concern attitude 1 -0.3150 0.0727 18.77 <.0001
Landcare membership- yes 1 0.0942 0.1269 0.55 0.4582
Length of landcare membership yrs 1 0.0311 0.0316 0.97 0.3246
Recent training 1 -0.0274 0.0797 0.12 0.7310
Farm cash income $ 1 0.0000 0.0000 0.20 0.6533
Profit at full equity $ 1 0.0000 0.0000 0.51 0.4737
Closing equity ratio % 1 -0.0095 0.0031 9.59 0.0020
Farm plan / property manag plan - yes 1 -0.1173 0.0954 1.51 0.2190
Farm size 000/ha 1 -0.1341 0.0596 5.05 0.0246
State 5 85.62 <.0001
-NSW 1.0706 0.1536 48.60 <.0001
-VIC -0.4970 0.1736 8.19 0.0042
-QLD -0.4487 0.2256 3.96 0.0467
-SA -0.2240 0.2189 1.05 0.3061
-WA -0.4989 0.2184 5.22 0.0224
Land use intensity se/ha 1 0.0331 0.0217 2.33 0.1268
PMP participation last 3 years - yes 1 0.2792 0.1323 4.45 0.0348
Wheat area sown ha 1 -0.0001 0.0003 0.25 0.6172
Closing number of sheep no 1 0.0001 0.0000 4.20 0.0404
Sample size=894, adoptors=403
Model fit Likelihood ratio test Chi-square=175.13, DF=21, P<0.0001
Association of predicted probabilities and observed responses: concordant=71.3%, discordant=28.5%, tied=0.2%
Soil/plant tissue tests to determine
fertiliser needs in the wheat-sheep and
high rainfall zones (broadacre only)
Seven variables were associated with the use of soil or
plant tissue tests to determine fertiliser needs (Table A5).
Recent training, length of landcare membership, farm cash
income, land use intensity and having a farm plan were all
positively associated with practice adoption. While farm
cash income is significantly associated the parameter
estimate indicates the actual effect of farm cash income to
be minimal. State of residence was also associated with
adoption, with owner/managers based in Queensland less
likely to adopt the practice than the average for Australia.
Those owner/managers who considered land and water
degradation to be critical in farm planning were more likely
to adopt tests for fertiliser needs.
Mues, Chapman and Van Hilst (1998) investigated the
adoption of regular soil testing by cropping specialists and
found similar relationships, with training and land use
intensity both associated with adoption. Additionally, they
found closing equity ratio and farm size to be significantly
related to regular soil testing.

54
Table A5 Logit regression results for the adoption of soil/plant tests to determine fertiliser needs in the
wheat-sheep and high rainfall zones (broadacre industries only).
Effect Unit DF Parameter
Estimate
Std Error Chi-
square
Prob
Intercept 1 -0.5124 0.5731 0.80 0.3713
Age yrs 1 -0.0058 0.0060 0.91 0.3402
Environmental concern attitude 1 0.1750 0.0666 6.91 0.0086
Financial concern attitude 1 -0.0964 0.0719 1.80 0.1799
Technical concern attitude 1 -0.0281 0.0758 0.14 0.7110
Financial outlook concern attitude 1 -0.0532 0.0590 0.81 0.3670
Landcare membership- yes 1 -0.0539 0.1159 0.22 0.6422
Length of landcare membership yrs 1 0.0889 0.0331 7.21 0.0072
Recent training 1 0.3202 0.0718 19.88 <.0001
Farm cash income $ 1 0.0000 0.0000 7.01 0.0081
Profit at full equity $ 1 0.0000 0.0000 1.61 0.2050
Closing equity ratio % 1 -0.0028 0.0028 1.04 0.3068
Farm plan / property man plan - yes 1 0.2333 0.0841 7.70 0.0055
Farm size 000/ha 1 -0.0420 0.0392 1.15 0.2844
State 5 22.05 0.0005
-NSW -0.2513 0.1395 3.34 0.0676
-VIC -0.1940 0.1468 1.75 0.1863
-QLD -0.8055 0.1897 18.03 <.0001
-SA 0.2184 0.1922 1.29 0.2557
-WA -0.0080 0.2049 0.00 0.9690
Land use intensity se/ha 1 0.0658 0.0112 34.28 <.0001
PMP participation last 3 years - yes 1 -0.1486 0.1169 1.62 0.2038
Wheat area sown ha 1 0.0014 0.0004 13.46 0.0002
Closing number of sheep no 1 0.0000 0.0000 0.34 0.5581
Sample size=1224, adoptors=786.
Model fit Likelihood ratio test Chi-square=269.11, DF=22, P<0.0001
Association of predicted probabilities and observed responses: concordant=69.6%, discordant=29.9%, tied=0.5%.
Tree and shrub establishment in the
wheat-sheep and high rainfall zones
(including dairy industries)
Seven variables were associated with the establishment of
trees and shrubs in the wheat-sheep and high rainfall zones
(Table A6). Landcare membership, having a farm plan and
land use intensity were positively associated with adoption.
Three attitude variables: environmental concern, technical
concern and financial outlook were also significantly
associated. Owner/managers who considered land and
water degradation to be critical in farm planning were more
likely to establish trees and shrubs. Those with increased
levels of concern regarding their technical resources were
less likely to engage in the practice than those with lower
levels of technical concern. Similarly, those
owner/managers with a more negative financial outlook
were less likely to plant trees and shrubs than those with a
more positive financial outlook. Owner/managers in
Queensland were less likely to adopt than the Australian
average and those in Western Australia were more likely to
adopt.
Curtis et al. (2000) considered the total area of trees
planted in their study of the Goulburn–Broken catchment.
The only common variable they found to be significantly
associated with the practice was ‘having a written property
plan that involved a map or other documents’. They found
significant associations between the total area of trees
planted and farm size.
Mues, Chapman and Van Hilst (1998) considered the
adoption of tree planting by livestock specialists in the
wheat-sheep and high rainfall zones. They also found the
presence of a farm plan to be significantly associated with
practice adoption, however they also found training to be
significantly related, a result which was not confirmed in the
analysis here. The more selective industry analysis
performed by Mues, Chapman and Van Hilst (1998) is likely
to explain some differences in the findings.

55
Table A6 Logit regression results for the establishment of trees and shrubs in the wheat-sheep and high
rainfall zones (including dairy industries).
Effect Unit DF Parameter
Estimate
Std Error Chi-
square
Prob
Intercept 1 0.4799 0.5923 0.66 0.4179
Age yrs 1 -0.0071 0.0062 1.32 0.2507
Environmental concern attitude 1 0.1898 0.0673 7.95 0.0048
Financial concern attitude 1 -0.0400 0.0739 0.29 0.5882
Technical concern attitude 1 0.2188 0.0806 7.37 0.0066
Financial outlook concern attitude 1 -0.1663 0.0604 7.59 0.0059
Landcare membership- yes 1 0.4288 0.1264 11.50 0.0007
Length of landcare membership yrs 1 0.0184 0.0352 0.27 0.6012
Recent training 1 0.0253 0.0723 0.12 0.7260
Farm cash income $ 1 0.0000 0.0000 0.10 0.7528
Profit at full equity $ 1 0.0000 0.0000 0.25 0.6174
Closing equity ratio % 1 -0.0042 0.0028 2.16 0.1418
Farm plan / property man plan - yes 1 0.2250 0.0889 6.41 0.0113
Farm size 000/ha 1 -0.0274 0.0343 0.64 0.4234
State 5 42.91 <.0001
-NSW -0.0052 0.1336 0.00 0.9689
-VIC 0.2790 0.1453 3.69 0.0549
-QLD -0.9993 0.1878 28.30 <.0001
-SA -0.0209 0.1847 0.01 0.9097
-WA 0.8958 0.2199 16.59 <.0001
Land use intensity se/ha 1 0.0459 0.0113 16.51 <.0001
PMP participation last 3 years - yes 1 0.1529 0.1249 1.49 0.2209
Wheat area sown Ha 1 -0.0003 0.0002 1.77 0.1834
Closing number of sheep No 1 0.0000 0.0000 0.65 0.4189
Sample size=1201, adoptors=717.
Model fit Likelihood ratio test Chi-square=235.20, DF=22, P<0.0001
Association of predicted probabilities and observed responses: concordant=74.0%, discordant=25.8%, tied=0.2%.
Regularly monitor watertables in the
wheat-sheep and high rainfall zones
(including dairy industries)
Three variables were associated with the adoption of
regularly monitoring water tables (Table A7). These
variables were having a farm plan, closing equity ratio and
land use intensity. Owner/managers with a farm plan were
more likely to monitor their water tables than those that did
not have farm plans. As land-use intensity increased,
adoption also increased; and as equity increased the
likelihood of adoption decreased. In contrast to these
findings, Mues, Chapman and Van Hilst (1998) found
landcare membership and training to be significantly
associated with regular monitoring of water tables by mixed
livestock-cropping enterprises, while having a farm plan,
closing equity ratio and land use intensity were not
significantly associated with adoption of the practice.
Table A7 Logit regression results for the regular monitoring of watertables in the wheat-sheep and high
rainfall zones (including dairy industries).
Effect Unit DF Parameter
Estimate
Std Error Chi-
square
Prob
Intercept 1 -1.5965 0.7725 4.27 0.0388
Age yrs 1 -0.0021 0.0086 0.06 0.8030
Environmental concern attitude 1 0.0644 0.0929 0.48 0.4878
Financial concern attitude 1 0.0422 0.0961 0.19 0.6610
Technical concern attitude 1 -0.0895 0.1056 0.71 0.3971
Financial outlook concern attitude 1 0.1149 0.0808 2.02 0.1551
Landcare membership- yes 1 0.1391 0.1400 0.99 0.3203
Length of landcare membership yrs 1 0.0372 0.0336 1.22 0.2683

56
Effect Unit DF Parameter
Estimate
Std Error Chi-
square
Prob
Recent training 1 0.0099 0.0871 0.01 0.9099
Farm cash income $ 1 0.0000 0.0000 0.04 0.8435
Profit at full equity $ 1 0.0000 0.0000 0.81 0.3685
Closing equity ratio % 1 -0.0109 0.0035 9.73 0.0018
Farm plan / property man plan - yes 1 0.4103 0.1035 15.72 <.0001
Farm size 000/ha 1 -0.2400 0.1274 3.55 0.0596
State 5 7.33 0.1971
-NSW -0.0895 0.2241 0.16 0.6895
-VIC 0.2993 0.2121 1.99 0.1582
-QLD 0.1046 0.3065 0.12 0.7330
-SA 0.5820 0.2692 4.68 0.0306
-WA 0.0498 0.2826 0.03 0.8601
Land use intensity se/ha 1 0.0268 0.0125 4.61 0.0318
PMP participation last 3 years - yes 1 0.0702 0.1429 0.24 0.6232
Wheat area sown ha 1 0.0004 0.0004 0.89 0.3461
Closing number of sheep no 1 0.0001 0.0001 2.59 0.1078
Sample size=1147, adoptors=190.
Model fit Likelihood ratio test Chi-square=94.44, DF=22, P<0.0001
Association of predicted probabilities and observed responses: concordant=69.6%, discordant=29.9%, tied=0.5%.
Collection of dairy effluent (dairy
industry only)
Two variables were associated with the collection of dairy
effluent by dairy farmers (Table A8). These variables were
recent training and technical concern. As training increased
the likelihood of collecting dairy effluent increased; and with
increasing technical concern about inadequate resources to
address land degradation the likelihood of collection of dairy
effluent decreased.
Table A8 Logit regression results for the collection of dairy effluent (dairy industry only).
Effect Unit DF Parameter
Estimate
Std Error Chi-
square
Prob
Intercept 1 1.0427 1.4824 0.49 0.4818
Age yrs 1 0.0018 0.0139 0.02 0.8990
Environmental concern attitude 1 0.0383 0.1908 0.04 0.8410
Financial concern attitude 1 -0.0034 0.2013 0.00 0.9867
Technical concern attitude 1 -0.5608 0.2032 7.62 0.0058
Financial outlook concern attitude 1 -0.0806 0.1389 0.34 0.5919
Landcare membership- yes 1 -0.5936 0.3243 3.35 0.0672
Length of landcare membership yrs 1 0.1183 0.1151 1.06 0.3042
Recent training 1 0.6553 0.2050 10.22 0.0014
Farm cash income $ 1 0.0000 0.0000 2.17 0.1411
Profit at full equity $ 1 0.0000 0.0000 1.98 0.1590
Closing equity ratio % 1 0.0026 0.0081 0.10 0.7483
Farm plan / property manag plan - yes 1 0.0898 0.2211 0.16 0.6845
Farm size 000/ha 1 0.4491 0.8234 0.30 0.5855
State 5 5.67 0.3402
-NSW -0.0005 0.3887 0.00 0.9902
-VIC 0.1845 0.3121 0.35 0.5545
-QLD 0.7661 0.3868 3.92 0.0476
-SA 0.4581 0.5204 0.77 0.3788
-WA 0.4367 0.7296 0.36 0.5495
Land use intensity se/ha 1 0.0330 0.0196 2.81 0.0935
PMP participation last 3 years - yes 1 -0.2293 0.2730 0.71 0.4009
Wheat area sown ha 1 0.0111 0.2000 0.30 0.5809
Closing number of sheep no 1 -0.0008 0.0010 0.55 0.4576

57
Sample size=307, adoptors=237.
Model fit Likelihood ratio test Chi-square=53.90, DF=22, P=0.0002
Association of predicted probabilities and observed responses: concordant=68.6%, discordant=31.0%, tied=0.3%.
Pump dairy shed effluent onto pasture
(dairy industry only)
Four variables were associated with the adoption of
pumping dairy shed effluent onto pasture (Table A9).
Owner/operators who considered their future farm
profitability likely to increase were more likely to adopt the
practice than those who thought farm profitability would fall.
Landcare membership was positively associated with
pumping effluent, however owner/managers with longer
membership of landcare were less likely to pump effluent
than those who had been landcare members for shorter
periods. State of residence was also significant with dairy
farmers in Victoria and Queensland both more likely to
adopt the practice than the Australian average.
Table A9 Logit regression results for the pumping of dairy shed effluent onto pasture (dairy industry only).
Effect Unit DF Parameter
Estimate
Std Error Chi-
square
Prob
Intercept 1 3.1346 1.3824 5.14 0.0234
Age yrs 1 -0.0095 0.0129 0.54 0.4645
Environmental concern attitude 1 -0.1904 0.1607 1.40 0.2360
Financial concern attitude 1 0.2537 0.1825 1.93 0.1644
Technical concern attitude 1 0.0266 0.1843 0.02 0.8851
Financial outlook concern attitude 1 -0.3404 0.1221 7.77 0.0053
Landcare membership- yes 1 0.9697 0.3425 8.02 0.0046
Length of landcare membership yrs 1 -0.2957 0.1129 6.86 0.0088
Recent training 1 0.1989 0.1516 1.72 0.1895
Farm cash income $ 1 0.0000 0.0000 0.32 0.5692
Profit at full equity $ 1 0.0000 0.0000 0.01 0.9191
Closing equity ratio % 1 -0.0067 0.0075 0.80 0.3724
Farm plan / property manag plan - yes 1 -0.2206 0.1827 1.46 0.2272
Farm size 000/ha 1 -0.9747 0.7842 1.54 0.2139
State 5 13.84 0.0166
-NSW -0.1943 0.3446 0.32 0.5729
-VIC 0.6949 0.2809 6.12 0.0134
-QLD 1.0134 0.4006 6.40 0.0114
-SA 0.0362 0.4855 0.01 0.9406
-WA -1.2845 0.6650 3.73 0.0534
Land use intensity se/ha 1 -0.0206 0.0175 1.39 0.2377
PMP participation last 3 years - yes 1 -0.2772 0.2479 1.25 0.2634
Wheat area sown ha 1 -0.0341 0.0303 1.26 0.2612
Closing number of sheep no 1 0.0014 0.0011 1.75 0.1858
Sample size=307, adoptors=187.
Model fit Likelihood ratio test Chi-square=51.12, DF=22, P=0.0004
Association of predicted probabilities and observed responses: concordant=67.8%, discordant=32.0%, tied=0.2%.
Laser graded layout (irrigated farms
only)
Three variables were associated with the use of laser
graded layout in irrigation farming (Table A10). Having a
farm plan was positively associated with adoption. Viewing
water and land degradation as a critical concern in farm
planning, the environmental concern attitude, was also
positively associated with practice adoption. State of
residence was also significantly associated with adoption,
with Victorian farmers more likely to adopt the practice than
the Australian average.
Table A10 Logit regression results for laser graded layout on irrigated farms.
Effect Unit DF Parameter
Estimate
Std Error Chi-
square
Prob
Intercept 1 -2.5899 2.0394 1.61 0.2041
Age yrs 1 0.0051 0.0172 0.09 0.7656

58
Effect Unit DF Parameter
Estimate
Std Error Chi-
square
Prob
Environmental concern attitude 1 0.4555 0.2174 4.39 0.0362
Financial concern attitude 1 -0.0901 0.2268 0.16 0.6912
Technical concern attitude 1 -0.3107 0.2071 2.25 0.1335
Financial outlook concern attitude 1 0.2349 0.1683 1.95 0.1629
Landcare membership- yes 1 -0.5664 0.3168 3.20 0.0738
Length of landcare membership yrs 1 0.0578 0.0826 0.49 0.4841
Recent training 1 0.0119 0.1989 0.00 0.9524
Farm cash income $ 1 0.0000 0.0000 0.30 0.5853
Profit at full equity $ 1 0.0000 0.0000 0.17 0.6772
Closing equity ratio % 1 -0.0093 0.0090 1.07 0.3007
Farm plan / property man plan - yes 1 0.4758 0.2359 4.07 0.0436
Farm size 000/ha 1 -0.2074 0.1590 1.70 0.1922
State 5 17.97 0.0030
-NSW 1.0609 0.8365 1.61 0.2047
-VIC 2.0051 0.8368 5.74 0.0166
-QLD -1.4671 1.0620 1.91 0.1671
-SA 0.5854 0.9866 0.35 0.5529
-WA 1.7234 1.2728 1.83 0.1757
Land use intensity se/ha 1 0.0303 0.0237 1.63 0.2013
PMP participation last 3 years - yes 1 -0.3706 0.3168 1.37 0.2420
Wheat area sown ha 1 0.0060 0.0030 4.01 0.0451
Closing number of sheep no 1 0.0001 0.0002 0.15 0.6943
Sample size=323, adoptors=151.
Model fit Likelihood ratio test Chi-square=72.67, DF=22, P<0.0001
Association of predicted probabilities and observed responses: concordant=78.6%, discordant=21.2%, tied=0.2%.
Use of irrigation scheduling tools
(irrigated farms only)
Two variables were associated with the use of irrigation
scheduling tools on irrigated farms (Table A11). Recent
training was positively related to practice adoption, as the
number of courses attended increased so too did the
likelihood of using irrigation scheduling tools. However,
participation in the Property Management Program in the
previous three years was negatively associated with
practice adoption. The results from this model should be
interpreted cautiously as the overall model was not found to
be significant.
Table A11 Logit regression results for the use of irrigation scheduling tools on irrigated farms.
Effect Unit DF Parameter
Estimate
Std Error Chi-
square
Prob
Intercept 1 -7.1687 313.3000 0.00 0.9817
Age yrs 1 0.0104 0.0236 0.19 0.6595
Environmental concern attitude 1 0.2359 0.2816 0.70 0.4021
Financial concern attitude 1 0.0035 0.2570 0.00 0.9892
Technical concern attitude 1 0.0061 0.2686 0.00 0.9818
Financial outlook concern attitude 1 0.2458 0.2266 1.18 0.2779
Landcare membership- yes 1 -0.1936 0.4403 0.19 0.6601
Length of landcare membership yrs 1 -0.0108 0.1235 0.01 0.9300
Recent training 1 0.8824 0.2363 13.94 0.0002
Farm cash income $ 1 0.0000 0.0000 0.09 0.7639
Profit at full equity $ 1 0.0000 0.0000 0.27 0.6052
Closing equity ratio % 1 -0.0138 0.0094 2.14 0.1431
Farm plan / property manag plan - yes 1 0.2570 0.2665 0.93 0.3349
Farm size 000/ha 1 0.0060 0.1172 0.00 0.9586
State 5 4.05 0.5418
-NSW 2.3606 313.3000 0.00 0.9940
-VIC 2.0809 313.3000 0.00 0.9947

59
Effect Unit DF Parameter
Estimate
Std Error Chi-
square
Prob
-QLD 2.1244 313.3000 0.00 0.9946
-SA 3.3594 313.3000 0.00 0.9914
-WA -13.8941 1566.3000 0.00 0.9929
Land use intensity se/ha 1 -0.0150 0.0309 0.24 0.6277
PMP participation last 3 years - yes 1 -1.2984 0.4895 7.04 0.0080
Wheat area sown ha 1 -0.0006 0.0018 0.10 0.7487
Closing number of sheep no 1 0.0000 0.0002 0.03 0.8532
Sample size=332, adoptors=53.
Model fit Likelihood ratio test Chi-square=32.26, DF=22, P=0.0731
Association of predicted probabilities and observed responses: concordant=74.7%, discordant=24.9%, tied=0.4%.
Monitoring of pasture and vegetation
condition (all farms)
Four variables were associated with the monitoring of
pasture and vegetation condition when all types of farms
are considered (Table A12). Recent training, land use
intensity and having a farm plan were all positively
associated with reported monitoring behaviour. Financial
outlook was also significantly associated. Owner/managers
who felt their farm profitability would fall in the next five to
10 years were less likely to adopt the practice. The earlier
consideration of monitoring of pasture and vegetation
condition in the pastoral zone only found significant
associations with several financial and attitude variables,
but no relationship with training and land use intensity.
Table A12 Logit regression results for monitoring of pasture and vegetation condition (all farms).
Effect Unit DF Parameter
Estimate
Std Error Chi-
square
Prob
Intercept 1 -0.0874 0.6030 0.02 0.8848
Age yrs 1 -0.0035 0.0064 0.29 0.5902
Environmental concern attitude 1 -0.0404 0.0679 0.35 0.5514
Financial concern attitude 1 -0.0882 0.0701 1.58 0.2086
Technical concern attitude 1 0.0256 0.0760 0.11 0.7361
Financial outlook concern attitude 1 -0.3023 0.0636 22.58 <.0001
Landcare membership- yes 1 0.1184 0.1112 1.13 0.2868
Length of landcare membership yrs 1 0.0492 0.0279 3.11 0.0779
Recent training 1 0.3175 0.0665 22.77 <.0001
Farm cash income $ 1 0.0000 0.0000 0.23 0.6325
Profit at full equity $ 1 0.0000 0.0000 0.54 0.4613
Closing equity ratio % 1 -0.0026 0.0026 0.96 0.3260
Farm plan / property man plan - yes 1 0.1921 0.0799 5.78 0.0162
Farm size 000/ha 1 0.0053 0.0031 2.90 0.0885
State 6 8.47 0.2059
-NSW 0.0211 0.2505 0.01 0.9330
-VIC -0.0984 0.2573 0.14 0.7021
-QLD -0.0002 0.2679 0.00 0.9995
-SA -0.2179 0.2939 0.55 0.4585
-WA -0.6161 0.3000 4.22 0.0400
-TAS 0.5266 0.4040 1.70 0.1924
Land use intensity se/ha 1 0.0225 0.0102 4.85 0.0276
PMP participation last 3 years - yes 1 0.0058 0.1067 0.00 0.9563
Wheat area sown ha 1 -0.0001 0.0002 0.19 0.6576
Closing number of sheep no 1 0.0000 0.0000 0.14 0.7086
Sample size=1402, adoptors=391.
Model fit Likelihood ratio test Chi-square=165.18, DF=23, P<0.0001
Association of predicted probabilities and observed responses: concordant=70.1%, discordant=29.5%, tied=0.3%.

60
Preserve/enhance areas of conservation
value (all farms)
Six variables were associated with the preservation or
enhancement of areas of conservation value (Table A13).
Recent training was positively associated with practice
adoption. Land use intensity was negatively associated, as
land use intensity increased adoption of the practice
decreased. Environmental concern, financial concern and
financial outlook attitude variables were significantly
associated with practice adoption. All were related to
adoption in the direction expected: as owner/managers
became more concerned with water and land degradation
in their farm planning they were more likely to preserve or
enhance areas of conservation value; owner/managers who
considered they did not have the financial resources to
address water and land degradation were less likely to
adopt; and owner/managers who felt future farm profitability
would increase were more likely to adopt the practice. The
owner/managers’ state of residence was significantly
associated with adoption; Tasmanian farmers were more
likely to adopt than the average for Australia and
Queensland-based farmers were less likely to adopt.
Mues, Chapman and Van Hilst (1998) found training,
landcare membership and farm size to have significant
positive associations with the preservation of areas of
conservation value for dairy farmers. In the Curtis et al.
(2000) study the ‘area of remaining native bush and
waterways fenced’ is the most comparable practice
operationalisation to the preservation or enhancement of
areas of conservation value. Of the variables in common,
only property size was found to be significantly associated
with the practice.
Table A13 Logit regression results for preservation or enhancement of areas of conservation value (all
farms).
Effect Unit DF Parameter
Estimate
Std Error Chi-
square
Prob
Intercept 1 -0.0172 0.5450 0.00 0.9748
Age yrs 1 0.0077 0.0056 1.87 0.1710
Environmental concern attitude 1 0.3375 0.0616 29.98 <.0001
Financial concern attitude 1 -0.1333 0.0644 4.29 0.0384
Technical concern attitude 1 -0.0703 0.0687 1.05 0.3062
Financial outlook concern attitude 1 -0.1782 0.0545 10.71 0.0011
Landcare membership- yes 1 0.0578 0.1035 0.31 0.5769
Length of landcare membership yrs 1 0.0121 0.0273 0.20 0.6581
Recent training 1 0.2112 0.0636 11.03 0.0009
Farm cash income $ 1 0.0000 0.0000 0.28 0.5978
Profit at full equity $ 1 0.0000 0.0000 0.14 0.7055
Closing equity ratio % 1 -0.0029 0.0023 1.61 0.2038
Farm plan / property manag plan - yes 1 0.0155 0.0756 0.04 0.8379
Farm size 000/ha 1 -0.0011 0.0022 0.23 0.6346
State 6 19.95 0.0028
-NSW -0.3022 0.2295 1.73 0.1879
-VIC -0.0579 0.2378 0.06 0.8078
-QLD -0.7075 0.2437 8.43 0.0037
-SA -0.2168 0.2646 0.67 0.4127
-WA 0.0056 0.2579 0.00 0.9828
-TAS 0.8661 0.3926 4.87 0.0274
Land use intensity se/ha 1 -0.0208 0.0094 4.91 0.0268
PMP participation last 3 years - yes 1 0.0194 0.1026 0.04 0.8501
Wheat area sown ha 1 0.0001 0.0002 0.49 0.4849
Closing number of sheep no 1 0.0000 0.0000 0.63 0.4285
Sample size=1306, adoptors=715.
Model fit Likelihood ratio test Chi-square=137.83, DF=23, P<0.0001
Association of predicted probabilities and observed responses: concordant=65.4%, discordant=34.3%, tied=0.3%.
Excluding stock from degraded areas (all
farms)
Four variables were associated with the practice of
excluding stock from degraded areas (Table A14). Age was
negatively associated with practice adoption: younger
farmers were more likely to exclude stock from degraded
areas than older farmers. Environmental concern was
positively associated with excluding stock; and poor
financial outlook was negatively associated with adopting
the practice. Queensland based farmers were less likely to

61
adopt the practice while Tasmanian farmers were more likely to exclude stock from degraded areas.
Table A14 Logit regression results for the exclusion of stock from degraded areas (all farms).
Effect Unit DF Parameter
Estimate
Std Error Chi-
square
Prob
Intercept 1 -0.6034 0.5819 1.08 0.2998
Age yrs 1 -0.0122 0.0060 4.17 0.0412
Environmental concern attitude 1 0.3013 0.0655 21.18 <.0001
Financial concern attitude 1 -0.0414 0.0661 0.39 0.5315
Technical concern attitude 1 0.0480 0.0701 0.47 0.4934
Financial outlook concern attitude 1 -0.1786 0.0579 9.51 0.0020
Landcare membership- yes 1 0.2002 0.1043 3.68 0.0549
Length of landcare membership yrs 1 0.0445 0.0274 2.63 0.1047
Recent training 1 0.1134 0.0642 3.12 0.0774
Farm cash income $ 1 0.0000 0.0000 0.00 0.9838
Profit at full equity $ 1 0.0000 0.0000 0.01 0.9373
Closing equity ratio % 1 -0.0014 0.0024 0.32 0.5724
Farm plan / property manag plan - yes 1 0.0953 0.0777 1.50 0.2202
Farm size 000/ha 1 -0.0023 0.0029 0.63 0.4269
State 6 45.04 <.0001
-NSW 0.1000 0.2557 0.15 0.6957
-VIC 0.0094 0.2642 0.00 0.9715
-QLD -1.0421 0.2851 3.36 0.0003
-SA -0.1697 0.2934 0.33 0.5630
-WA 0.5009 0.2799 3.20 0.0735
-TAS 1.0912 0.3929 7.71 0.0055
Land use intensity se/ha 1 0.0047 0.0097 0.24 0.6235
PMP participation last 3 years - yes 1 -0.1823 0.1056 2.98 0.0842
Wheat area sown ha 1 0.0000 0.0002 0.00 0.9677
Closing number of sheep no 1 0.0000 0.0000 0.02 0.8848
Sample size=1306, adoptors=488.
Model fit Likelihood ratio test Chi-square=179.44, DF=23, P<0.0001
Association of predicted probabilities and observed responses: concordant=69.9%, discordant=29.9%, tied=0.2%.
Conservation tillage (all farms with a
cropped area in 1998-99)
Five variables were associated with the adoption of
conservation tillage (Table A15). Age, technical concern
and land use intensity were negatively associated with the
adoption of conservation tillage. Older farmers were less
likely to adopt the practice than younger farmers, as were
those who operated their properties at more intensive levels
of production. Owner/operators who felt they did not have
the technical resources to address land and water
degradation were also less likely to adopt than those who
felt they had the technical resources. Recent training was
positively associated with practice adoption. State of
residence was also significantly associated with adoption,
however no single state was significantly different from the
Australian average.
Table A15 Logit regression results for the percentage of the farm under conservation tillage (all farms).
Effect Unit DF Parameter
Estimate
Std Error Chi-
square
Prob
Intercept 1 0.4042 1.2100 0.11 0.7402
Age yrs 1 -0.0268 0.0066 16.59 <.0001
Environmental concern attitude 1 0.0219 0.0691 0.10 0.7516
Financial concern attitude 1 0.0611 0.0747 0.67 0.4136
Technical concern attitude 1 -0.2121 0.0812 6.82 0.0090
Financial outlook concern attitude 1 0.0889 0.0620 2.05 0.1519
Landcare membership- yes 1 0.2027 0.1120 3.28 0.0703
Length of landcare membership yrs 1 -0.0036 0.0300 0.01 0.9037

62
Effect Unit DF Parameter
Estimate
Std Error Chi-
square
Prob
Recent training 1 0.2139 0.0683 9.79 0.0018
Farm cash income $ 1 0.0000 0.0000 2.51 0.1130
Profit at full equity $ 1 0.0000 0.0000 0.02 0.8937
Closing equity ratio % 1 0.0008 0.0029 0.09 0.7656
Farm plan / property manag plan - yes 1 0.1498 0.0828 3.27 0.0705
Farm size 000/ha 1 -0.0580 0.0328 3.12 0.0775
State 6 16.37 0.0119
-NSW 0.1618 1.0706 0.02 0.8799
-VIC 0.2578 1.0718 0.06 0.8099
-QLD 0.0352 1.0810 0.00 0.9740
-SA 0.7840 1.0766 0.53 0.4665
-WA -0.1711 1.0835 0.02 0.8745
-TAS -0.5823 1.1328 0.26 0.6072
Land use intensity se/ha 1 -0.0299 0.0110 7.35 0.0067
PMP participation last 3 years - yes 1 -0.1130 0.1175 0.93 0.3360
Wheat area sown ha 1 0.0027 0.0004 39.51 <.0001
Closing number of sheep no 1 0.0002 0.0000 14.24 0.0002
Sample size=1154, adoptors=553.
Model fit Likelihood ratio test Chi-square=305.47, DF=23, P<0.0001
Association of predicted probabilities and observed responses: concordant=78.5%, discordant=21.3%, tied=0.2%.

63
Appendix B Description of variables used in logistic
regression analyses
Farm family characteristics:
state of residence: categorical variable with between five
and seven levels as appropriate.
Farm financial characteristics:
farm cash income: difference between total cash receipts
and total cash costs.
profit at full equity: farm business profit, plus rent, interest
and finance lease payments less depreciation on leased
items. It is the return produced by all the resources used in
the farm business.
closing equity ratio: calculated as farm business equity
(value of owned capital, less farm business debt at 30
June) as a percentage of owned capital at 30 June.
financial concern attitude: ordinal variable with 5 levels
(strongly disagree, disagree, neither agree nor disagree,
agree and strongly agree) indicating response to the
statement “I don’t have the financial resources available to
adequately address land and water degradation on my
property”. Treated as an interval level variable in the
analyses.
financial outlook attitude: ordinal variable with 5 levels
(strongly disagree, disagree, neither agree nor disagree,
agree and strongly agree) indicating response to the
statement “I feel the profitability of my farm is likely to fall
from current levels over the next 5 to 10 years”. Treated as
an interval level variable in the analyses.
Voluntary participation:
landcare membership: binary variable indicating the
membership of landcare in the year 1998-99.
length of landcare membership: interval level variable
measured in years.
Education and training:
recent training: number of courses or training activities
undertaken in the last three years.
PMP participation: binary variable indicating involvement in
a Property Management Planning program.
Farm structure:
farm size: size of the farm measured in ‘000 Hectares.
land use intensity: intensity of land use measured in sheep
equivalents per hectare.
Farming or land management experience:
age: interval level variable indicating the operator’s age in
years.
farm plan: binary variable indicating the presence of a farm
plan or property management plan.
technical concern attitude: ordinal variable with 5 levels
(strongly disagree, disagree, neither agree nor disagree,
agree and strongly agree) indicating response to the
statement “I don’t have the technical resources available to
adequately address land and water degradation on my
property”. Treated as an interval level variable in the
analyses.
environmental concern attitude: ordinal variable with 5
levels (strongly disagree, disagree, neither agree nor
disagree, agree and strongly agree) indicating response to
the statement “Land and water degradation is a critical
concern to me in farm planning”. Treated as an interval
level variable in the analyses.