• 
• 
• 

You are here: Agriculture » About Agriculture » Publications and Resources » Policy and Strategy » Policy Choice Framework

Policy Choice Framework

July 2007

Acknowledgements

This information was originally developed by Geoff Kaine, John Ford, Melinda Leth and Fiona Johnson Department of Primary Industries, Victoria

ISBN 978-1-74199-541-1 

The research reported here was funded by the Victorian Water Trust. We would like to thank David Lawler, Chris Nicholson, Ken Sampson, Stuart Gemmel, Geoff Coburn, Peter Lahey, Rabi Maskey and Abdi Quassim for sharing their key informant knowledge and expertise in helping us to apply the Policy Choice Framework to the Water Smart Farms Program in the Shepparton Irrigation Region. We have interpreted the information they provided for the purpose of testing the Policy Choice Framework and we take full responsibility for our interpretation.

Table of Contents

EXECUTIVE SUMMARY

INTRODUCTION

OVERVIEW OF FRAMEWORK

COMPONENT FRAMEWORKS

• PRIVATE AND PUBLIC NET BENEFIT
• TECHNICAL FEASIBILITY
• LANDHOLDER INVOLVEMENT - I3 RESPONSE FRAMEWORK
• USE VARIETY
• SCALE AND RATE OF CHANGE
• POLICY INNOVATION FRAMEWORK
• CONCLUSION

CASE STUDIES

1. INCENTIVES FOR AUTOMATIC IRRIGATION
2. REGULATION FOR WHOLE FARM PLANNING
3. RESEARCH INTO CENTRE PIVOT IRRIGATION TECHNOLOGIES
4. A CAP AND TRADE MARKET FOR WATER

DISCUSSION

• ACCOUNTING FOR VARIABILITY IN LANDHOLDER BEHAVIOUR
• ACCOUNTING FOR VARIABILITY IN INSTITUTIONAL BEHAVIOUR
• IMPLICATIONS OF POLICY INSTRUMENT SELECTION

CONCLUSION

REFERENCES

APPENDIX 1: SEQUENCE OF DECISION TREES USED TO WORK THROUGH POLICY CHOICE FRAMEWORK FOR AUTOMATIC IRRIGATION CASE STUDY

APPENDIX 2: SEQUENCE OF DECISION TREES USED TO WORK THROUGH POLICY CHOICE FRAMEWORK FOR WHOLE FARM PLAN CASE STUDY

APPENDIX 3: SEQUENCE OF DECISION TREES USED TO WORK THROUGH POLICY CHOICE FRAMEWORK FOR RESEARCH CASE STUDY

APPENDIX 4: SEQUENCE OF DECISION TREES USED TO WORK THROUGH POLICY CHOICE FRAMEWORK FOR WATER MARKET CASE STUDY

Executive Summary

The Policy instrument Choice Framework (PCF) contributes to existing approaches for choosing between NRM policy instruments by providing a means of incorporating the responses of landholders and NRM organisations into the policy instrument design and selection process. It incorporates conceptual frameworks from marketing and organisational change theory with existing approaches to the selection and design of suasive, market based and regulatory policy instruments.

Incorporating conceptual frameworks from marketing and organisational change theory builds on existing approaches in that it reveals:

• The aspects of the farm system that are important to landholders when they are considering their response to an environmental issue and the policy instruments addressing the issue.
• The potential for an undesirable behavioural response from groups of landholders, such as outrage, that could lead to policy failure or reduced efficacy of policy instruments. 
• Whether the distribution and rate of behavioural response by landholders to a policy instrument will achieve the environmental objective of the policy instrument. 
• The likely behavioural response of NRM organisations to the new policy instrument and what this means for the successful implementation of the new policy instrument.
• The nature of the change that would occur in the organisations responsible for implementing a policy instrument if implementation went ahead.

The PCF is made up of a sequence of six conceptual frameworks that are linked together through a series of decision trees. Although the decision trees present the appearance of a series of sequential steps, they contain feedback loops that support an iterative approach to policy instrument selection.

Figure 1. Stages in the Policy Instrument Choice Framework, including feedback loops.

Figure 1. Stages in the Policy Instrument Choice Framework, including feedback loops.

The PCF contains six separate conceptual frameworks that are grouped in three broad stages (see figure I). The first of these stages contains frameworks that address the justification for government intervention and the technical feasibility issues that impact on the selection of policy instruments. The frameworks in the second stage of the PCF reveal and incorporate the behavioural responses of landholders to a policy instrument into the design and selection process. In the final stage of the PCF the policy instruments that have been selected and adjusted in the previous stages are evaluated using a conceptual framework that predicts the response of institutions to the proposed policy instruments and the likely consequences of this response.

Four case studies were selected in the Shepparton Irrigation Region (SIR) to trial the PCF. These case studies analysed the decisions that led to the selection of some of the policy instruments that are part of the implementation program of the Goulburn Broken Regional Catchment Strategy in the SIR. These included regulatory controls associated with whole farm planning, financial incentives for automatic irrigation, research into efficient irrigation technologies and water trading.

The analysis of the decisions by policy makers in each of these case studies suggested that the PCF was able to reveal the complexity of decision making by policy makers that occurs in practice. In particular it was able to reveal how policy makers accounted for a variety of landholder responses to a policy instrument, as well as the causes of that variety. In each of the case studies the decisions taken by policy makers were consistent with the principles underpinning the various conceptual frameworks that constitute the PCF.

In summary, the four case studies suggested that the PCF had the ability to:

• Indicate broad selection of policy instruments on the balance between public net benefit and private net benefit.
• Refine the broad selection of policy instruments based on technical feasibility.
• Reveal the variety in landholder responses to the policy instrument and the causes of that variety.
• Reveal whether a policy instrument addresses the full range of landholder behaviours.
• Make explicit the need to address unfavourable attitudes landholders might have toward a policy instrument.
• Provide a sound basis for suggesting modifications to a proposed policy instrument to account for differences in landholder's responses.
• Provide a sound basis for suggesting when complementary policy instruments were needed to account for unfavourable attitudes by landholders.
• Provide a systematic method for demonstrating that individual organisations are affected in unique ways by the introduction of a new policy instrument.
• Be used as a means of revealing information gaps that are strategically critical to the successful implementation of policy instruments.

In summary, the case studies suggest that the PCF would be useful in assisting policy-makers to systematically select and customise policy instruments to achieve natural resource outcomes. They also suggested that a wide range of factors influence the behaviour of landholders which can have an impact on the efficiency or effectiveness of a policy instrument and that these factors can change over time. In principle, the PCF could be used in an iterative fashion to take into account for changes in behaviour and knowledge over time and thereby support double loop organisational learning processes.

Introduction

Traditionally, natural resource policy has relied on regulatory, statutory and legal instruments to direct behaviour in regard to the environment (Ward et al 2005). Examples of such instruments include prohibitions on the use of specific herbicides; the use of permits and licencing to control activities such as hunting and fishing, and planning restrictions on the location of new irrigation developments. Over the past decade or so there has been an increasing interest among policy makers in the potential for market-based instruments (MBIs) to influence environmental behaviour and contribute to natural resource policy (Hatton MacDonald et al 2004). Associated with this increased interest in market-based instruments has been a growing recognition of a need for frameworks to assist policy makers in choosing between different types of policy instruments – regulatory, suasive, and market-based (Hatton MacDonald et al 2004; National Action Plan for Salinity 2005; Ward et al 2005).

A variety of frameworks have been proposed to assist policy makers in choosing between policy instruments (Connor and Bright 2003; Hatton MacDonald et al 2004; Martin and Verbeek 2006; Ridley and Pannell 2005; Tietenburg and Johnstone 2004; Whitten et al 2006; Ward et al 2005; Young et al 1996). While these frameworks differ in their scope and detail, broadly speaking, they commonly suggest the following criteria are critical to choosing between policy instruments (Hatton MacDonald et al 2004):

1. The potential of different policy instruments to modify the behaviour of individuals enough to ensure environmental policy objectives are met.
2. The potential of different instruments to reduce the cost to government of meeting environmental policy objectives.
3. The need to reveal and account for behavioural responses of landholders in the process of selection and design of policy instruments.
4. The need to reveal and account for the institutional context and the responses of organisations operating in this context to new policy instruments in the process of selection and design of policy instruments.

Our concern in this paper is with the last two points.  The review of existing approaches to the selection of policy instruments completed as part of this project (Kaine et al. 2006) indicated that existing frameworks do not have rigorous, practical methods for revealing and accounting for the behavioural responses of landholders and organisations toward a policy instrument.

This has important implications for policy makers, for example, variations in behavioural response by landholders could limit the applicability of a policy instrument to a small group of individuals within a larger population (Hatton MacDonald et al 2004; Martin and Verbeek 2006). Similarly a means for anticipating potentially damaging behavioural responses such as outrage and suggesting a means for avoiding them could reduce both the political and implementation cost associated with controversial policy instruments (Gunningham et al 1998; Hatton MacDonald et al 2004).

Slightly different but given equal importance in the literature is understanding and accounting for the institutional context and the behaviour of organisations, since not doing so is likely to increase the cost of implementation or reduce the efficacy of a policy instrument (Gunningham et al 1998; Kaine and Higson 2006; Hatton MacDonald et al 2004).

The absence of procedures for assessing the variety of behavioural and institutional responses to policy instruments makes the task of selecting an instrument problematic. Instruments may be selected, for example, that appear efficient but this is of little help if their effectiveness is threatened by unanticipated behavioural responses of landholders affected by the policy instrument. Similarly, instruments may be selected because they appear effective but this is of little help if their effectiveness is threatened by unanticipated responses of those organisations whose responsibility it is to implement them.

Our intention is to contribute to the literature on frameworks for choosing between policy instruments by providing a framework that integrates with existing complementary frameworks and that addresses the third and fourth points will assist policy makers to choose between regulatory, suasive and market-based instruments by:

• Identifying a systematic and practical process for anticipating variety in the reactions of different groups to policy instruments and accounting for them in the choice and design of policy instrument.
• Identifying a systematic and practical process for anticipating the variety of reactions public, private and community organisations charged with implementing policy to different policy instruments and accounting for it in the choice and design of policy instrument.

In the next section we integrate frameworks for revealing variety in the responses of individuals and institutions to policy instruments (Kaine and Johnson 2004; Kaine and Higson 2006a, Kaine and Higson 2006b, Kaine et al 2006, Murdoch et al 2006) with complementary frameworks for choosing policy instruments from the literature (Hatton MacDonald et al 2004; Pannell 2006; Whitten 2006; Gunningham et al 1998; Gunningham and Sinclair 1999; Gunningham and Sinclair 2002) to create a Policy instrument Choice Framework (PCF).

Overview of Framework

The Policy instrument Choice Framework (PCF) is intended to assist policy makers in the field of natural resource policy to choose between policy instruments when the focus of the policy is on changing the behaviour of agricultural landholders. The PCF contains a series of conceptual frameworks that link systematically together through a series of decision trees (see Appendices 1 through to 6). Although the decision trees present the appearance of a series of sequential steps, they also contain feedback loops that support an iterative approach to policy instrument selection.

The PCF contains six separate conceptual frameworks that are grouped in three broad stages (see figure 1). The first of these stages contains frameworks that address the justification for government intervention and the technical feasibility issues that impact on the selection of policy instruments. The first of these frameworks justifies government intervention through broad categories of policy instruments on the basis of the public and private net benefits associated with landuse (Pannell 2006). The second of these frameworks further refines the selection process by evaluating the choices of policy instrument on the basis of technical considerations such as, ability to measure, certainty about biophysical processes, and distribution of impact (Gunningham and Sinclair 1999; Hatton MacDonald et al 2004; Whitten et al 2006).

The second stage of the framework reveals and incorporates the behavioural responses of landholders to a policy instrument into the design and selection process. The first of the conceptual frameworks in this stage evaluates policy instruments against the variety of landholder's responses to them. This evaluation is based on assessments of landholder's level and source of involvement in the policy issue¹  and the likely level and source of their involvement with the policy instrument²  (Murdoch et al 2006).

Figure 1. Stages in the Policy Instrument Choice Framework, including feedback loops.

Figure 1. Stages in the Policy Instrument Choice Framework, including feedback loops.

The framework developed by Murdoch et al (2006) provides a systematic basis for classifying landholders into groups or segments on the basis of differential behavioural responses to policy instruments as advocated by Hatton MacDonald et al (2004). At this point policy instruments may be eliminated from consideration because they are likely to provoke unfavourable reactions among landholders. Alternatively, instruments may be modified or complemented by other measures to facilitate acceptance among landholders.

In the fourth conceptual framework the potential for landholders to comply with a policy instrument in ways that are counter-productive to the policy outcome are assessed (Kaine and Higson 2006b). While the fifth conceptual framework evaluates the potential for the instrument to change the behaviour of landholders on a scale, and at a rate, that is consistent with the policy outcome (Kaine and Johnson 2006).  This completes the second stage of the PCF, at this point policy makers have a choice between policy instruments (or mix of policy instruments) that account for the full range of behavioural responses from landholders. They are also effective in that they elicit a behavioural response at a scale and rate capable of achieving the policy objective.

In the final stage of the PCF the degree of fit between the remaining policy instruments (or mix of policy instruments) is evaluated against both the institutional context and the likely response of institutions to the proposed policy instruments.  This assessment involves considering the consistency of the principles underpinning the policy instrument with the principles underpinning the policy culture of the organisations that would be implementing the instrument. Substantial inconsistencies in these principles would indicate that substantial changes would be required in the structures, procedures and technical characteristics of these organisations (Kaine and Higson 2006a; Kaine et al 2006).

In summary, the PCF consists of a series of conceptual frameworks that are linked together through decision trees. Unlike other frameworks for choosing between policy instruments the PCF contains components that reveal detail about the behaviour of landholders and institutions that can be used to choose between policy instruments and suggest alterations in design or the inclusion of other policy instruments.

Component Frameworks

Private and public net benefit

Pannell (2006) has developed a private and public net benefits framework derived from economic theory that can be readily applied to select between broad categories of policy instruments. The principle aim of the framework is to choose policy instruments on the basis of allocating public funds efficiently. The framework is based on relative levels of total public and total private net benefits of changing land management.

Private net benefits refer to benefits minus costs accruing to the private land manager as a result of the proposed changes in land management (Pannell 2006). Public net benefits means benefits minus costs accruing to everyone other than the private land manager (Pannell 2006). The private net benefit dimension provides insights into the behaviour of the landholder, while the public net benefit dimension relates to the effects on everyone else that flow from the landholder's behaviour. Logically, different policy instruments are required to influence behaviour for different combinations of total private and total public net benefit.

The different relative levels of private and public net benefit that are theoretically possible are described in Table 1 and presented graphically in Figure 2. Each letter in the first column of Table 1, and the corresponding section in Figure 2, indicates a unique combination of public and private net benefit. Where the total private net benefit is positive for an activity we assume the activity is currently part of the commercial operations of private landholders. Where the total private net benefit is negative for an activity we assume the activity is not currently part of the commercial operations of private landholders.

Table 1 Summary derived from Pannell (2006) of rules for government intervention in natural resource management³

Section of diagram 1	Nature of Private and Public Benefit that policy  instrument must address
A -	Public Net Benefits  > Private Net Costs
B -	Public Net Benefits,  Private Net Benefits > 0
C -	Private Net Benefits  > Public Net Costs
D -	Public Net Costs >  Private Net Benefits
E -	Public Net Costs,  Private Net Costs >  0
F -	Private Net Costs  > Public Net Benefits

Figure 2 Combinations of private and public net benefit resulting from rules described in table 1.

Figure 2 Combinations of private and public net benefit resulting from rules described in table 1.

Section A in Figure 2 represents circumstances where an activity would create private net costs but these are exceeded by public net benefits. Since the public net benefit of the activity exceeds the private net cost then, in principle, the community will be better off if the activity were supported by publicly funded incentives. The conservation of remnant vegetation may be an example of such an activity. There are substantial public net benefits from conserving remnant vegetation but landholders may experience commercial losses that outweigh the aesthetic and other benefits of conservation they may experience.  Since the activity creates a private net cost a positive incentive is required to induce landholders to undertake the activity (see Figure 3). A publicly funded incentive compensates landholders for their net private losses. Positive incentives include incentive programs and tender mechanisms (Pannell 2006).

Section B in Figure 2 represents circumstances where an activity creates both private net benefits and public net benefits. The adoption of water saving technology such as spray irrigation may be an example of such an activity. It is assumed that where private net benefits arise from an activity that the technology will be adopted and that access to information about the technology limits the rate of adoption. Given this a publicly funded extension program that improves access to relevant information can be justified since it increases the rate which public net benefits can be realised (see Figure 3).

Section C in Figure 2 represents circumstances where an activity is creating public net costs but these are exceeded by private net benefits. Since the activity creates private net benefits that exceed the public net cost then, in principle, the community would be worse off if public resources were invested in efforts to restrict the activity (see Figure 3).

Figure 3 Policy responses to combinations of private and public net benefit

Figure 3 Policy responses to combinations of private and public net benefit

Section D in Figure 2 represents circumstances where an activity creates private net benefits but these are exceeded by public net costs. Since the public net cost of the activity exceeds the private net benefit then, in principle, there is a net cost to the community as a whole if the activity is allowed to continue.  To address this to ensure that on the whole community will be better off, the activity needs to be restricted using negative incentives, that is disincentives. Negative incentives include taxes, charges, regulations and cap and trade mechanisms (Pannell 2006). The obstruction of floodways by irrigation remodelling is an example of such an activity in the Shepparton Irrigation Region (SIR). There are considerable environmental costs associated with floodway obstruction and these outweigh the commercial returns associated with obstructions to floodwaters.  Since the activity creates a public net cost a disincentive is required to prevent landholders from engaging in the activity (see Figure 3).

Section F in Figure 2 represents circumstances where an activity would create public net benefits but these would be exceeded by private net cost. Since the private net cost of the activity exceeds the public net benefit then, in principle, the community would be better off if public funds were invested to reduce the private net cost of the activity. The investment in research to lower the private net cost of revegetating salt affected land is an example of such an activity. Pannell (2006) suggests that the investment of public funds in research programs to reduce the private costs associated with a environmentally desirable activity is the most appropriate policy response for section F (see Figure 3).

Placing the natural resource policy issue in the context of relative public and private net benefits provides a basis for choosing between four types of policy instruments – positive incentives, negative incentives or disincentives, research and extension. Where the combination of private and public net benefits suggests that either positive or negative incentives are appropriate the next step is to refine the choice within each of these policy types by considering the technical feasibility of different instruments.

Technical Feasibility

Where the application of the Private and Public Net Benefits framework indicates that positive incentives are appropriate the choice of policy instrument lies between incentive programs and tender programs. We have drawn on Whitten et al (2006) and Hatton MacDonald et al (2004) to create a decision tree that considers the technical feasibility of policy instruments on the basis of technical issues such as the nature of distribution of public benefits (environmental impact) and private benefits (benefits to landholders of publicly funded works) between landholders (see figure 4).

Where the application of Pannell's (2006) framework indicates that negative incentives are appropriate the choice of policy instrument lies between regulations, charges and markets. We have drawn on Whitten et al (2006), Hatton MacDonald et al (2004), Gunningham and Gabrosky (1998), and Gunningham and Sinclair (1999) to create decision trees identifying the factors that influence the technical feasibility of these instruments. These are presented in Figures 5 and 6. The primary factors influencing the technical feasibility of charges and markets are the measuring of individual use, and aggregate supply, of a resource and whether private net benefit differs across landholders. The primary factors influencing the technical feasibility of regulations are the measurement of individual use of the resource, or processes that entail use of the resource.

Figure 4 Technical feasibility of positive incentives. Based on Whitten et al. (2006) and Hatton McDonald et al. (2004)

Figure 4 Technical feasibility of positive incentives. Based on Whitten et al. (2006) and Hatton McDonald et al. (2004)

Figure 5 Technical feasibility of negative incentives – charges and markets. Based on Whitten et al. (2006) and Hatton McDonald et al. (2004)

Figure 5 Technical feasibility of negative incentives – charges and markets. Based on Whitten et al. (2006) and Hatton McDonald et al. (2004)

Figure 6 Technical feasibility of negative incentives – regulation. Based on Gunningham and Sinclair (1999)

Figure 6 Technical feasibility of negative incentives – regulation. Based on Gunningham and Sinclair (1999)

Landholder Involvement - I₃ Response Framework

The next stage in choosing between policy instruments is to consider the likely reactions of landholders to different instruments. Following Murdoch et al (2006) and Lourey et al (2006) we propose that the response of landholders to an instrument will depend on their likely level of involvement with the policy issue and their likely level of involvement with the policy instrument.

Central to the I₃ Response framework of Murdoch et al (2006) is the psychological concept of involvement. Murdoch et al (2006) use involvement to predict the likely behavioural responses of landholders to regulations that impose obligations on them to take specific actions. The I₃ response framework hypothesises that the landholders' responses to a regulation is product of the level of their involvement in the policy issue, their level of involvement with, and attitudes towards, the regulations and other interventions employed by government to address the policy issue (Murdoch et al 2006).

Involvement is a motivational state, the strength of which depends on the relevance of a product or activity to a person's utilitarian, social or hedonic goals (Mittel and Lee 1989). Involvement predicts the level of effort an individual will invest in relation to an activity such as extensiveness of decision-making, the number of factors evaluated in a decision, the number of alternative actions considered, and the time spent on reaching a decision (Dholakia 2001; Kapferer and Laurent 1986; Mittal and Lee 1989; Poisesz and de Bont 1995; Verbeke and Vackier 2004; Zaichkowsky 1986). In short, Murdoch et al (2006) argue an individual's level of involvement with a policy issue and associated government intervention will determine the attention a landholder gives to the policy instrument and the effort they will invest in information processing and decision making in regard to it.

The two dimensions of involvement, involvement with the issue and involvement with the intervention, can be combined to predict different categories of behavioural responses among landholders to a policy instrument. Where involvement in the issue and the policy instrument is low then landholders are likely to have limited awareness of the policy instrument. The landholder is unlikely to have devoted time and effort to considering the issue. Their response to the introduction of the policy instrument is predicted to be unintentional compliance or non-compliance – see quadrant 1 in Figure 7. Potential actions that may be taken to improve compliance with the instrument are to increase involvement or to reduce the effort required by landholders to be in compliance – see quadrant 1 in Figure 8.

Where involvement in the issue is high but involvement with the policy instrument is low then landholders are likely to have some awareness of the policy issue. Their high involvement with the policy issue suggests the issue is important enough to the landholder for them to devote time and effort to deciding on a course of action about the issue. Low involvement with the policy instrument is more likely in situations where the policy instrument is either irrelevant or easily compatible with existing farming practices, in both cases it does not disrupt farming businesses or impose a substantial financial costs on the business. Murdoch et al (2006) documented an example of this in relation to phylloxera quarantine regulations. Where existing regulations were consistent with vineyard practices, high rates of compliance resulted with minimum enforcement effort.

Where involvement in the policy instrument is low it is predicted that landholders the landholders will respond to the introduction of negative incentives by compliance – see quadrant 2 in Figure 7. Compliance with the instrument may be improved by promotion of the policy instrument – see quadrant 2 in Figure 8.

Figure 7. I₃ Response Framework with behavioural predictions (Adapted from Murdoch et al. 2006)

Figure 7. Response Framework with behavioural predictions (Adapted from Murdoch et al. 2006)

Figure 8. I₃ Response Framework with selected policy interventions (Adapted from Murdoch et al.2006)

Figure 8.I₃ Response Framework with selected policy interventions (Adapted from Murdoch et al.2006)

Where involvement in the policy issue and the policy instrument is high then landholders are likely to have some awareness of both the policy issue and the instrument. Consequently, the landholder is likely to have devoted time and effort to considering the issue and the instrument. Their response to the introduction of the policy instrument is predicted to be compliance or non-compliance with conflict depending on whether their attitude towards the instrument is favourable or unfavourable– see quadrant 3 in Figure 7. Non-compliance with conflict arises because landholders have devoted considerable effort to considering both the policy issue and policy instrument and have formed a preference for a different policy instrument. Potential actions that may be taken to improve compliance with the instrument are to incorporate complementary policy instruments that address some of the concerns of landholders into the policy or invest in monitoring and enforcement – see quadrant 3 in Figure 8.

Where involvement in the policy issue is low but involvement in the policy instrument is high landholders are likely to have little awareness of the issue but some awareness of the instrument. Consequently, the landholder is likely to have devoted time and effort to considering the policy instrument. Their response to the introduction of the policy instrument is predicted to be compliance or non-compliance with outrage depending on whether their attitude towards the instrument is favourable or unfavourable– see quadrant 4 in Figure 7. Non-compliance with outrage arises because landholders regard the policy issue as personally unimportant or irrelevant and may have taken the view that the instrument will impose unjustifiable costs on them. Potential actions that may be taken to improve compliance for target groups in this quadrant are to change involvement in the policy issue or invest in monitoring and enforcement – see quadrant 4 in Figure 8.

The application of the framework developed by Murdoch et al (2006) provides a systematic basis for classifying landholders into segments on the basis of differential responses to policy instruments as advocated by Hatton MacDonald et al (2004). The application of this framework provides a means of eliminating policy instruments from consideration because they are likely to provoke unfavourable reactions among landholders. Alternatively, the application of this framework provides a systematic means for determining how policy  instruments may need to be modified or supplemented by other measures so as to facilitate acceptance among landholders.

Use variety

In the next stage in the PCF, the potential for landholders to comply with a policy instrument in ways that are counter-productive in terms of the policy is assessed using Kaine and Higson (2006b). Using marketing theory Kaine and Higson argue that landholders have characteristics that will lead them to respond to a policy instrument in ways that are consistent with the policy instrument but are potentially counter-productive with regard to the policy objective. Examples of this behaviour are described in Kaine and Johnson (2004).

The following characteristics were identified by Kaine and Higson (2006b) as indicators of counter-productive responses:

• Landholders possess extensive and sometimes diverse business and social networks.
• Landholders express high levels of mastery and control over their farming systems.
• Policy instruments that seriously affect the management and performance of a farming system create a high involvement situation for landholders. This motivates them to invest the time and effort needed to identify responses to mediate the impact.
• Farming systems are highly complex and a change in one part of the system tends to create a cascade of changes throughout the system. Even though policy instruments may be directed at only one of many aspects of the farming system, the policy instrument may have implications for the entire farming system.

The application of Kaine and Higson (2006b) provides a basis for eliminating a policy instrument from consideration because there is a high potential that landholders may respond to the policy instrument in counter-productive ways. Alternatively, this framework may be employed to identify the need to modify an instrument or supplement an instrument with other measures in order to reduce the potential for counter-productive responses.

An important qualifier: The application of Kaine and Higson (2006b) will not identify policy risks associated with policy objectives other than the policy objective being addressed by the proposed policy instrument/s.

Scale and rate of change

At this point the instruments (or mix of instruments) that remain are those that are technically feasible, are feasible in terms of eliciting responses from landholders that are consistent with the policy outcome, and carry an acceptable level of risk in terms of potential variety in landholder responses.  The next stage is to consider the potential for these instruments to change the behaviour of landholders on a scale, and at a rate, that is consistent with the policy outcome.  That is, the effectiveness of the policy instrument.

Kaine and Johnson (2004) describe one approach to distinguishing scale and rate changes in the behaviour of landholders that is based on marketing theory and farming system theory.  Their approach draws, in part, on the psychological concept of involvement and is consistent with the approach used by Murdoch et al (2006) to identify variety in landholders' responses to policy instruments described earlier.

Kaine and Johnson (2004) characterise the decision-making of landholders in high involvement situations as complex decision-making (Assael 1998). In this framing policy instruments that provoke scale changes in the behaviour of landholders can be identified as those that change the decision making criteria used by landholders to evaluate changes to agricultural enterprises, practices and technologies (Kaine and Johnson 2004).

Some policy instruments are better than others at influencing the rate which the behaviour of landholder changes. If rate of behavioural change is important to the policy objective it is necessary to evaluate the proposed policy instrument or mix of policy instruments on the on the rate of behavioural change. There are two ways of doing this. The first is to establish whether the proposed policy instrument/s modifies the decision making criteria used by landholders to evaluate agricultural enterprises, practices and technologies. The other is to establish whether the proposed policy instrument/s change the amount of effort landholders expend when evaluating potential change to agricultural enterprises, practices and technologies. For example, including educating and training in the mix of proposed policy instruments has the potential to reduce the amount of effort landholders' need to expend to evaluate the change associated with a desirable behaviour (Kaine and Johnson 2004, 11).

This completes the evaluation of policy instruments with respect to the responses of landholders.

Policy Innovation Framework

In the final stage of the PCF proposed policy instrument/s are evaluated on the basis of their fit with the existing institutional context. There are two aspects to current approaches to assessing whether a proposed policy instrument/s fit within the existing institutional context (Gunningham et al 1995; Whitten et al 2006; Hatton MacDonald et al 2004). The first is to establish whether appropriate legislative and/or regulatory frameworks exist that the proposed policy instrument/s can work within. The second is to consider whether the institutions responsible for implementation have the required competencies to implement a proposed policy instrument/s. The assessment process in the PCF complements these approaches by providing a systematic method to describe and classify the effects of a new policy instrument on each organisation in a larger institutional context responsible for implementing the new policy instrument.

Introducing new policy instruments can have completely different consequences for different organisations involved in designing and implementing natural resource policy. Using the organisational management literature on product innovation Kaine and Higson (2006b) and Kaine et al (2006) argued that the response of organisations to a policy instrument depends on the consistency of the set of principles underpinning the policy instrument with the set of principles underpinning the policy culture of the organisations. Inconsistencies between these sets of principles indicate that changes will be required in the structures, procedures and technical characteristics of these organisations (Kaine and Higson 2006b).

Kaine and Higson (2006b) proposed that changes in policy instruments could be viewed as types of policy innovations. Given this proposition, they developed a framework to classify innovations in policy instruments into four types; incremental, modular, architectural and radical (see Figure 9). The fundamental elements of the policy innovation framework are described in Table 2. Each type has different implications for the organisational knowledge and skills that are required to implement the policy instrument. In this way the policy innovation framework provides a systematic way of anticipating the likely consequences of introducing a new policy instrument to the organisations responsible for its implementation and the most likely behavioural responses by organisations to these consequences.

This framework has been applied to policy instruments in salinity and biodiversity policy in Australian and water allocation and water quality policy in New Zealand (Kaine et al 2006). In each of these case studies the framework revealed the consequences of introducing new policy instruments for the organisations involved and was used to identify specific competencies needed for the successful implementation of the policy instruments.

Conclusion

In this section the various components of the PCF have been described. The PCF integrates economic, regulatory and behavioural frameworks to provide a systematic process that:

1. Chooses between broad categories of policy instruments.
2. Modifies policy instruments to account for anticipated variety in the behaviour of both landholders the policy instrument seeks to influence and organisations responsible for implementing the policy instrument.

The PCF itself has been formulated as a series of linked decision trees. Each of these trees is contained in the Appendices to this report.

In the next section we present the application of the PCF in four case studies.

Table 2. Fundamental elements of the Policy Innovation Framework

Adapted from Kaine et al2006

Instrument Concept	A generic description of the way that the policy instrument achieves the policy objective.
Components	The individual rules, processes and procedures that form the policy instrument.
Component Principles	The fundamental  principles that guide the design and functioning of a component.
Architecture	The way that the components are arranged or integrated to form the policy instrument.
Architectural Principles	The fundamental principles that underpin the arrangement and combined functioning of the  components that form the policy instrument.

Figure 9. Examples of policy instruments as types of innovations Adapted from Kaine et al2006

Figure 9. Examples of policy instruments as types of innovations Adapted from Kaine et al2006

Case Studies

To evaluate the PCF four case studies from the Water Smart Farms Program of the Goulburn-Broken Catchment Management Authority were selected for analysis using the PCF.  The case studies were selected to represent the different combinations of public and private net benefit outlined by Pannell (2006) and different policy instruments. The case studies selected were:

1. Incentives for Automatic Irrigation (Public Benefit > Private Cost)
2. Regulation for Whole Farm Planning (Public Costs > Private Benefit)
3. Research into Centre Pivot Irrigation Technologies (Private Cost>Public Benefit)
4. Cap and Trade Market for Water (Public Costs > Private Benefit)

 

Data for testing the PCF was obtained by interviewing key informants for each case study. The key informants provided expert knowledge and insights into the selection of policy instruments based on their extensive involvement in the implementation of policy in each case study. The key informants were interviewed individually or in pairs using a relatively unstructured interview approach. In the interviews open-ended questions were used to elicit responses from the key informants that we then interpreted as answers to the decision points in each of the PCF trees.  Where possible, uncertainty concerning informant's recall of events or information was clarified using relevant documents.  Importantly, the authors conducted the analysis and interpretation of the information supplied by the key informants and supporting documents. Consequently, any errors or omissions are our responsibility.

1. Incentives for Automatic Irrigation

Three key assumptions underpin the 1989 Salinity Management Plan for the Shepparton Irrigation Region (Sampson K personal communication 2007). These were:

• The implementation of effective regional drainage would deliver one third of the catchment targets for reducing salinity.
• The implementation of effective whole farm planning would deliver another third of the catchment targets for reducing salinity.
• Improvement in irrigation management practices of individual farmers would deliver the final third of the catchment targets for reducing salinity.

While the first and second of these points was relatively straightforward to address through the Regional Catchment Drainage Program and Whole Farm Plans, the third was found to be more difficult and complex to manage.

Automatic irrigation was seen by the GBCMA as one means of improving the irrigation management practices of individual farmers. Rough calculations indicated that 2 ML/ha of applied irrigation water could potentially be saved through a combination of laser grading, recycle dams and automatic irrigation.  Each of these technological innovations was estimated to generate 0.66 ML/ha of water savings.

However, in 1997 only one percent of flood irrigated properties in the SIR used automatic irrigation (Douglass and Poulton 1998, cited in Maskey 2002). At that time, the poor reliability of the available systems and their high capital cost were major barriers to the adoption of automatic irrigation.

To increase the adoption of automatic irrigation the GBCMA conducted a study of the reliability of the automatic irrigation systems that were available and sought funding from the National Heritage Trust (NHT) for an incentive program.  A range of extension materials providing detailed information on the available systems was produced. The GBCMA bid to NHT was successful and the incentive program began on 1^st July 2001.

The resulting incentive program for automatic irrigation provided information to landholders on the available automatic irrigation technology and a once-off payment of up to $6,000 per property.

The application of the PCF to the development and implementation of the incentive program is summarised in Table 3. We concluded from the results of this case study that:

• The application of the PCF to the development and implementation of financial incentives for automatic irrigation revealed limitations in institutional capacity that reduced the feasible choice of policy instrument.
• By revealing the limitations in institutional capacity that constrained the choice of policy instrument the PCF described the logic used by local natural resource managers to support the selection of the matrix based financial incentive over alternatives.

Table 3: Application of the PCF to incentives for automatic irrigation

Public Net Benefit, Private Net Benefit Framework
According to key  informants, not having automatic irrigation is likely to lead to over-irrigation of difficult to irrigate paddocks, or those paddocks that were irrigated at  night. This suggests that the public benefits of automatic irrigation were  reduced waterlogging and nutrient run-off. The GB-CMA, in their NHT bid, estimated that 0.66 ML/ha of water could be saved through the use of automatic  irrigation by farmers. The public net benefits achieved by preventing these  losses were thought to be significant. The key informants  believed that, in 2001, the private costs associated with the risk of the  technology failing and the initial cost of investing in the technology probably  outweighed the private benefits to be had from automatic irrigation. Resulting  in a negative private net benefit. The evidence available at that time  indicated that farmers were generally prepared to risk an initial investment of  only $10,000 in the technology. These conditions  imply that automatic irrigation lies within section A of the private/public net  benefits framework, where Public Benefits exceed Private Cost (negative private  net benefit). Conclusion: Go to positive incentives decision tree.
Incentives or Auctions
The key informants described two key reasons for landholder interest in automatic irrigation. These were to achieve labour savings and to have a less stressful lifestyle. They indicated that the level of private benefit varies considerably between  landholders using automatic irrigation. Examples of this variety included one landholder who had saved one full time labour unit by adopting automatic irrigation  on their property and another landholder who put automatic irrigation on paddocks he only irrigated at night. The evident  variety in private net costs suggests that an auction or tender instrument would be the most efficient policy instrument to allocate public funds to support the adoption of automatic irrigation. Conclusion: Consider a tender or auction. Go to I3 Response Framework.
I3 Response Framework
The information supplied by key informants gave no indication that landholders were highly involved in the policy objectives of reducing waterlogging and nutrient run-off. There  was no evidence to suggest that landholders were concerned about reducing the amount of water reaching the regional water table. Nor were they concerned  about the potential for automatic irrigation to reduce their water usage. However the key  informants did stress that landholders investing in automatic irrigation were  very interested in using it to better manage labour on their farms.  Consequently, we inferred that landholders reacted favourably to the  possibility of receiving financial support for investing in automatic  irrigation (positive attitude). Conclusion: A tender or auction is feasible. Go to Use Variety Tree
Use Variety Tree
The key informants suggested that there was some potential for use variety with automatic irrigation. Evidence was presented to decision makers at the time of instrument  design and implementation suggesting that automatic irrigation was not  necessarily installed on those paddocks where there was the highest level of  public benefit in terms of reduced waterlogging and run-off (Kaine and Bewsell  2002). However, the potential for use variety was not considered to be  sufficiently large to endanger the objectives of the policy. Conclusion: A tender or auction is feasible. Go to Scoping Tree
Scoping Tree
Specific targets  for the adoption of automatic irrigation were not set in the Land and Water  Management Plan. Instead annual Regional Catchment Investment Plan targets depend on the availability of funding.  Based on $100 000 annually allocated to fund  the incentive for automatic irrigation, the GB-CMA aims to achieve the  installation of around 30 systems per year on average.    Key informants  suggested that the population of potential adopters were those landholders who  had fully implemented their Whole Farm Plans and were looking to achieve further productivity gains. This suggests that all landholders that had  undertaken a WFP would eventually be targeted for automatic irrigation. Given the original estimate that 0.66 ML/ha/year could be saved through the use of automatic irrigation on farms large quantities of water could be saved if every  landholder that had implemented their WFP eventually adopted this technology. According to key informants the rate of adoption of automatic irrigation systems depends on general economic and climatic conditions indicating that in times of financial stress farm finances are invested elsewhere. This suggests that the use of a financial incentive is appropriate since it maintains the rate of adoption of  the technology. Conclusion: The scope and rate of adoption was acceptable. Go to Policy innovation tree.
Policy Innovation Tree
In 2001, there was  little capacity in the organisations involved in the design and implementation of a tender or auction instrument. As such, we considered the introduction of a tender or auction as a radical policy change that would have required  substantial resources and skills to proceed. Given the implementation of a tender or auction system was not feasible then an incentive program should be  considered. The design and implementation of an incentive program for automatic irrigation represented an incremental innovation for GB-CMA since such a program would utilise existing organisational skills and processes. However there is some evidence to suggest that the program was a modular change to DPI, since DPI had to acquire specific expertise in automatic irrigation design and product performance.  Conclusion: Tender or auction is not feasible. Incentive is feasible with some capacity building in DPI. Implement financial incentive for automatic irrigation.

2. Regulation for Whole Farm Planning

The world has changed markedly since a collection of regional stakeholders introduced Whole Farm Plans (WFP) into the planning regulations of Local Government in 1991. The following is a brief description, provided by key informants and relevant documents, of the issues and events leading up to the introduction of WFP in 1991.

According to the key informants the introduction of laser grading technology from the construction industry led to widespread redevelopment of the irrigation region in the 1980s. There were a number of features associated with redevelopment that made the technology attractive to landholders in the SIR:

• Relatively large areas of land could be redeveloped.
• Water could be more effectively and efficiently managed during irrigation.
• Irrigating land became less labour intensive.
• Properties could be redeveloped to use more water than they had in the past.

However with the new technology large volumes of water could be moved quickly across irrigation bays. This sometimes resulted in unwanted waterlogging or flooding. In addition the technology in siting, designing and constructing farm delivery channels and drains lagged behind that of laser levelling land. As a consequence the following difficult issues around irrigation redevelopment arose for the community in the SIR namely:

• Large volumes of irrigation tail-water flooded neighbouring properties or roadsides.
• Regional water tables rose due to widespread water logging.
• Floodwaters were unable to drain naturally because of obstructions associated with irrigation redevelopment.

The response by government and the community to these issues led to a number of linked initiatives between 1987 and 1991. These were:

• The introduction of Irrigation Management Grants Program in 1987. This program included a financial incentive for WFPs.
• Incorporation of minimum design standards for WFPs in the eligibility criteria for Irrigation Management Grants in 1988.
• The development of community and stakeholder consultation and coordination processes through the activities of Salinity Pilot Program Advisory Council.
• Development and implementation of the Water Act (1989), that described the rights and responsibilities of landholders and government.
• Development and implementation of the Local Government Act (1989) that described the rights and responsibilities of Local Government to other tiers of government as well as the community.
• Incorporation of regulatory guidelines for earth movement into Local Government planning controls in 1991.

The application of the PCF to the decision to implement a regulatory program is summarised in Table 4. We concluded from the results of this case study that:

• The application of the PCF to the development and implementation of regulations for WFPs enabled a complex decision making process that included an assortment of stakeholders, each of whom played a unique role in the process, to be described systematically.
• The application of the PCF to the initial regulation for WFPs suggested changes, in particular the use of financial incentives, to reduce the risk associated with unfavourable community reactions while still maintaining effectiveness. These suggestions mirrored the actions made by decision makers at the time.
• Given the complexity of both the issues involved and the policy instrument, it was necessary to incorporate multiple perspectives from a variety of sources to get the information needed to work through the PCF (eg irrigation designers, Local Government, DPI staff and CMA staff).
• In terms of institutional change it became evident in the regulation for WFPs case study that each organisation involved in the implementation of a policy instrument should be considered separately as they all faced different challenges. The failure of any one of the organisations involved to successfully perform their role in the design and implementation process would have endangered the implementation of the policy instrument.

Table 4: Application of PCF to the regulation of Whole Farm Planning

                               

Public Net Benefit, Private  Net Benefit Framework
There were large  variations in public costs associated with irrigation redevelopment in the  1980s. Irrigation redevelopment carried out with the newly introduced 'laser grading technology' often led to the obstruction of natural drainage lines and  very long bay lengths leading to waterlogging. It follows that the public net  benefit associated with unplanned irrigation redevelopment in the 1980s was  characterised by widespread damage to farmland and public infrastructure  resulting from: Redirection of floodwaters onto neighbouring properties, roads and public land. Water logging and high accessions to water table. On the other hand the private net benefits of redeveloping a farm using the new technology in the 1980s could be characterised as labour saving and being able to irrigate large  areas of land more effectively. This was because longer and fewer irrigation  bays meant that a farmer could irrigate larger areas of land in a given time  and better siting of channels and less undulating bays meant that it was easier  to effectively irrigate more land than in the past. Redeveloping a property  also meant that waterlogging associated with flood events could be avoided by  strategically placing channels or check banks to divert water. Given that these  actions directly increase farm productivity and reduce labour costs it can be  assumed that there was a high level of private benefit associated with these  actions. The nature of the  public costs and private benefits associated with redeveloping irrigation farms  during the 1980s suggests that the public costs (public net benefits) were  greater than the private net benefits associated with increases in farm  productivity and profitability. This places the  redevelopment of irrigation farms in the SIR in the 1980s in section D of the private/public net benefits framework. Conclusion: Go to negative incentives decision tree.
Regulation, Charges or  Markets
The two main  public cost issues addressed by whole farm plans are flood management and  irrigation induced salinity. The resources associated with these issues are  functional drainage lines and a water table more than two metres from the  surface. Although easily conceptualised and broadly mapped at the time, it was  not possible in a practical sense to define these resources in terms of their  'supply'. Hence a market instrument was infeasible. However it was  still important to control the impact individuals had on flooding and salinity  since the actions of one individual could affect many others, particular in the  case of flooding. The precise impact of individual's actions was, if not  impossible, then prohibitively expensive to measure, as they were generally  confounded with the actions of others. Hence, both a market instrument and a  charge instrument were unlikely to be feasible. This means a regulatory  instrument was likely to be the most practical instrument to implement. Conclusion: Go to Regulation Tree.

Regulation
Two key opportunities emerged to introduce regulatory instruments as part of the WFP program to overcome the public costs associated with irrigation development. The first related to the introduction of minimum design standards for WFPs eligible for a CMA grant. The second was to link WFPs with the approval processes under the Planning regulations, Local Government Act, 1989. Conclusion: Incorporation of design standards into a Whole Farm Plan and link this to the approval processes under the Local Government Act. Go to I3 Response Framework.
I3 Response Framework
The involvement of landholders in the issue of flooding leading up to 1991 was high since large areas of the SIR had been regularly flooded throughout the 1980s. The issue of the involvement of landholders in salinity is more difficult to judge, particularly since much of the promotional material at the time referred to the "underground flood" instead of salinity. We assumed that the widespread community support for the 1989 Salinity Management Plan (SMP) indicated that the community was also highly involved in the issue of salinity at that time. Key informants suggested that there were two groups of landholders involved in the current WFP process. The first, and far more numerous, group was characterised by their positive attitude toward the WFP process. These landholders used the opportunity to develop a long-term irrigation development plan for their farm. The second, smaller, group was those landholders that had been forced to engage in the WFP process by local government. These landholders generally had a negative attitude about the process. Looking back to when the WFP policy instrument was being developed up until the introduction of planning controls in 1991 landholders had not been subjected to controls over earth works they carried out on their land. Hence, it is likely that the number of landholders in the second segment was much larger in 1991 than is the case now. Adjusting policy to account for unfavourable responses by landholders Given that there was likely to be a significant proportion of landholders with a negative attitude toward WFPs in 1991 when they were introduced an alternative route needed to be taken through the I3 Response Framework to understand how the PCF might have guided the decision-makers at the time. The decision makers at the time actively used a number of strong principles in their decision making, the two relevant here were as follows: The beneficiary pays principle - parties contribute proportionally based on the level of benefit they gain from changes in on farm practices. Prioritise actions and policies that are most likely to gain and maintain community support for the SMP. Given these strongly held principles it was probably considered that substituting annual planning permits for WFPs would not gain or maintain community support for the newly developed SMP. There was also an added risk that if WFPs were not supported throughout the community the SMP itself would be put in danger since the implementation of WFPs was to account for a third of the impact of the SMP. These risks were addressed by continuing the financial incentive for WFPs and giving WFPs an exemption from the need to annually renew the planning permit once gained. These modifications were consistent with both the beneficiary pays principle since government received benefits from WFPs and community values since they financially rewarded landholders for good works. Conclusion: Modify involvement in the regulation by including financial incentives for the development of the WFP, planning permit application fee and renewal exemption. Go to use variety tree.

Use Variety Tree
The key informants we interviewed suggested that the potential for use variety was high and could be described as follows: Not fully implementing a WFP. Making small undocumented changes to a WFP. Farm subdivision. Not implementing any works. During litigation, property sales and compulsory purchase of land. However, the key informants believed that, in reality, use variety was rare. For example, in practice the use of WFPs for litigation purposes has occurred in only a few cases. The most common form of use variety was thought to be using WFPs to cost works but not undertaking them (approximately 10% of WFPs). The key informants believed that currently 80% of landholders implement their WFPs to the point where 100% of the environmental benefit of implementing the WFP is realised. Documents dating from the introduction of the regulation suggest that policy makers had similar beliefs about the likely effectiveness of WFPs. Given this we believe the potential for use variety was not considered to be sufficiently large to endanger the objectives of the policy. Conclusion: Regulation is feasible.  Go to Scoping Tree.
Scoping Tree
The original 1989 SMP assumed that three percent of landholders would adopt a WFP on their property each year given existing levels of property redevelopment so that at the end of 30 years 90% of landholders had adopted a WFP to implement best management practices on their properties. Changing WFPs from a voluntary mechanism to one that is more or less compulsory would logically increase the rate and scope of adoption. This suggests that in 1991 it was reasonable for decision-makers to assume that the base line of three percent of landholders per annum was an achievable target. Conclusion: Regulation is feasible. Go to Policy Innovation tree.

     

Policy Innovation Tree
The move to incorporate WFPs into a regulatory framework, making them more or less compulsory for landholders, was a considerable change to the underlying principles by which government influenced on-farm irrigation development in comparison to its earlier voluntary approach.  However, if considered from the point of view of the different organisations this change is not as great as first appears. The regulatory process was split into modules and each module delegated to that organisation with the relevant expertise. The following is a list of organisations involved in the WFP process and the type of policy innovation the incorporation of WFPs into the Local Government planning process represented to them.
	Pre Implementation	Post Implementation	Type of change
Rural Water Authority (G-MW)	Played a role in designing and commenting on on-farm works from at least the 1960's to 1980's. Played an important role in developing guidelines for WFPs in late 80s.	Received referrals from local government to make recommendations on whether to accredit WFPs.	Incremental – technical skills and personnel already exist in organisation. Require funding to increase resources.
SPPAC (to become GB-CMA)	Coordinating stakeholders around: Catchment drainage. On farm works. Tree planting.	Manages payment of incentives to landholders.	Incremental – incorporation of more stakeholders into WFP coordination.
DARA (to evolve into DPI)	Promotion and administration of financial incentives for WFPs. Extension and advice on on-farm works.	Promotion and administration of financial incentives for WFPs. Extension and advice on on-farm works. However number increases due to compulsive nature.	Incremental – increase in demand for WFPs leading to more resources required.
CNR (to evolve into DSE)	Referral authority for local government planning processes.	Referral authority for WFPs where biodiversity is potentially affected.	Incremental – increase in demand for WFPs leading to more resources required.
Irrigation Designers	Already providing accredited plans to meet WFP incentive guidelines.	All WFPs have to be accredited and not just those subject to an incentive.	Incremental – increase in demand for accredited plans leads to industry expansion.
Local Government	Introduction of earth moving regulations in 1991 meant that planning permits were required for earth moving anyway.	Local Government planning decisions are guided by comments of referral authorities. Local Government plays collaborative role in SPPAC with other stakeholders.	Modular – Move from command/control to cooperative regulatory approach. Effort directed at collaboration instead of enforcement.
Ultimately the policy innovation brought about by the incorporation of WFPs into the Local Government planning process was either incremental or modular. This suggests that it would have been a relatively straightforward policy instrument to implement for the organisations involved. Interestingly where there was the greatest change, in Local Government and SPPAC, both organisations had considerable incentive to undertake the relevant change. The new policy would assist Local Government to protect public roads that were deteriorating rapidly, and increase the rate and scope of adoption of on-farm best management practices for SPPAC. Conclusion: Level of institutional change resulting from implementation of instrument is acceptable to the organisations involved, implement instrument.

3. Research into Centre Pivot Irrigation Technologies

The research project "Efficient Irrigation Technologies to Match Soils and Dairy Farming Systems" (EIT) commenced in 2004 and continued until the end of 2006. The objectives that were addressed in the project were (Department of Primary Industries 2006):

• Improve the access of information about the appropriate selection, design and management of Centre Pivot Technologies (CPT) on dairy farms.
• Develop measures of the economic and environmental benefits that the widespread adoption of CPT irrigation systems would have in the Goulburn Valley.

The research project was unique in that it combined market research with extension, on-farm experiments and traditional field station experiments. Ultimately, a web-based information source was produced. This was targeted at farmers, irrigation designers and suppliers of CPT. The project also produced valuable information suggesting that use of the technology would generate high levels of public benefit for only five percent of the region. This information suggests that widespread public support for the adoption of CPT was not warranted.

The project also provided the background for further research into alternative irrigation technologies. Support from the dairy industry has led to the development of another project investigating the use of sub surface irrigation on dairy farms.

The application of the PCF to the decision to implement a research program is summarised in Table 5. We concluded from the results of this case study that:

• The market research into CPT indicated that there were a range of landholder contexts each with a unique ratio of Private Benefit to Public Benefit. Given this a number of policy options, each reflecting the position of the landholder context in Pannell's framework, will be needed to achieve the policy objective. In this instance research into sub-surface and CPT was needed to reduce the private costs associated with limited information on more efficient irrigation technologies.
• Although representing architectural change to the organisations involved in the project they considered the project a success and well worth the extra effort and resources. This suggests that architectural change does not necessary endanger the successful implementation of a policy instrument if stakeholders actively collaborate and provide the support needed for the policy to be successful.
• To achieve its research goals the EIT project had to incorporate multiple perspectives into the development and adaptation of the research project to account for the complexity of landholder behaviour in relation to pressurised irrigation and automatic irrigation systems.
• Designing a policy instrument to best achieve desired outcomes might require an iterative process. In this case it was necessary to implement the research on CPT and then modify the research program when it became evident that it could be changed to better achieve the desired outcome.
• The EIT research project was important in defining the level of public benefit associated with alternative irrigation technologies. In particular, it defined how public benefit varied across the landscape and between landholder contexts.

Table 5: Application of the PCF to research into centre pivot technologies

Public Net Benefit, Private Net Benefit Framework

According to the market study conducted by Hill et al(2004) the benefits sought and achieved by dairy farmers through the adoption of centre pivot technologies were: To maximise fodder they could produce per megalitre of water. To save time and labour. Simplicity and the flexibility to adjust technology according to crop type and agronomy. Sale of technology when no longer needed. However there were significant private costs that had to be overcome before adoption took place: The large initial cost ($150-300k) of the system. The technology could not be implemented in stages as finance became available. The site had to be physically suited to Centre Pivot Technology (CPT). The site had to have access to sufficient quantity and quality of water supply. Despite being a technology that has a long history of use in the regions adjacent to the SIR information on the adaptation of the technology to Australian conditions was not available. Most sources originated from the US. Irrigation designers did not have the experience or local information to compensate for the lack of information on the use of the technology in Australian conditions. Manufacturers and suppliers of the technology could not be relied on to provide relevant information to dairy farmers. The first three private costs listed above varied considerably between farms across the region, while the last three private costs were constant across the region, suggesting that private benefits varied throughout the region. The EIT project aimed to reduce the cost of adoption by reducing the last three private costs to farmers. Government stakeholders in the project anticipated public net benefits in the form of reduced accessions of water to water tables through widespread adoption of CPT. Earlier studies indicated that the potential of reducing accessions is high on light soil types however these soils account for only five percent of the land in the Shepparton Irrigation Region (SIR). This suggests that public net benefits through the introduction of the technology are not large when considered across the entire region. In terms of the private/public net benefit framework we concluded that the private costs, or negative net private benefit, associated with adoption of CPT were higher than the potential public net benefits that might be achieved if the technology was only applied to light soils. This places the adoption of CPT by dairy farmers in quadrant F of Pannell's framework indicating that the appropriate policy instrument in this context is a research program aimed at reducing the cost of adoption to dairy farmers. Conclusion: Private cost of adoption of CPT is higher than the level of public benefits associated with the widespread adoption of the technology. Go to of I3 Response Framework.

I3 Response Framework
In the interviews with key informant informants and project documentation there was no information to suggest that dairy farmers were highly involved in the issue of efficient irrigation to reduce accessions to water tables. Given that salinity has not threatened the productive capacity of the region for the last seven years we assume that their involvement in this issue is low. The level of dairy farmers' involvement in the research project although not tested directly could be inferred from the level of involvement in the technology expressed in the marketing study conducted by Hill et al. (2004). Given the results of this study we suggest that the level of involvement of landholders in the research project was high since it made it easier to decide whether: The technology was applicable to their particular farming context. It reduced the likelihood of failure or the high cost of modifying the technology once in place. Attitudes towards the research project were positive since it addressed a need for relevant information. Conclusion: EIT research project is feasible. Go to Use Variety Tree.
Use Variety Tree
The potential for use variety was not considered to be a problem. Conclusion: EIT research project is feasible. Go to Scoping Tree.
Scoping Study Tree
The policy instrument contributes to the objective of reducing accessions to groundwater since five percent of the region is covered by light soil types where CPT irrigation can achieve high rates of efficiency in comparison to alternative types of irrigation. In the remaining area the efficiency of the technology is comparable to best practice border check irrigation. This suggests a two pronged approach. Information is still needed for those landholders considering CPT and in addition, alternative efficient irrigation technologies need to be investigated for the heavier soil types in the region. To take into account the need for alternative irrigation technologies on heavy soil types the research program was modified by developing a project on the management of sub-soil irrigation on dairy farms. This project takes into account the high level of interest in the dairy industry for this technology as well as its potential to avoid some of the limitations of CPT. This technology has also been widely adopted across the region in the tomato growing industry. This suggests that there is the potential for cost sharing arrangements between dairy and tomato industries for the establishment of this technology on dairy farms. This change in research direction does not reflect a change in involvement by landholders in regard to research into more efficient irrigation technologies. Conclusion: Instrument is feasible once modified to include alternative irrigation technologies relevant to heavy soil types. Go to policy innovation tree.

Policy Innovation Tree

The EIT project reflects a significant change in how research has been conducted at DPI, Tatura. First, it incorporated a marketing study to investigate the information needs of dairy farmers who have adopted or are considering adopting CPT. Second, it incorporated an extension component that worked closely with the scientists to develop an extension product that was comprehensive, accurate and relevant to the needs of dairy farmers.  Third, it involved the application of new methods for determining water use of crops and pastures under irrigation. In the past each of these things was separate and only interacted through more formal means. Instead, during the course of the project, staff from each component worked informally with other staff working on other components. This suggests that the nature of how the various components relate to each other was very different in this project to how other projects are normally conducted in the research and extension sections of DPI.  This suggests that the project represented an architectural policy innovation for DPI since it involved considerable change to how the various components (research, marketing and extension) related to each other. Given that this type of change can be highly disruptive to an organisation a number of allowances were made by the stakeholders supporting the project: CAS employees were made available to work on the project. Dairy Australia insisted that marketing research and extension be incorporated into the project and funded the project to account for the extra resources required. Active collaboration between traditionally separate areas of work and expertise was encouraged to deliver project objectives. Conclusion: Instrument is feasible even though considerable institutional change required. The collaborating institutions involved prepared to resource the changes required to facilitate the progress of the project.

4. A Cap and Trade Market for Water

Victoria has a long history of innovation associated with the management of its water resources. Beginning with Alfred Deakin's 1886 Irrigation Act up to and including the current round of water reform that separates water entitlement from land and explicitly defines property rights for the environment. From this long history of innovation we have chosen the development and implementation of transferable water entitlements, described here as a cap and trade market for water, between 1981 and 1997 as our case study for analysis. The reasons for this choice are as follows:

• It is the only cap and trade MBI aimed at managing natural resources implemented in the study area.
• The period between 1981 and 1997 covers the initial period of development through to final adjustments needed to account for use variety behaviour.
• The cap and trade water market has been widely considered as successful and is considered World's best practice.
• For individual irrigators a cap and trade water market was  very different to earlier processes of water allocation.
• The impact of this policy innovation has had far reaching and sometimes unforseen consequences beyond the case study period.

The application of the PCF to the development and implementation of the Victorian cap and trade water market is summarised in Table 6. We concluded from the results of this case study that:

• The net public benefit in this cap and trade water market case study was defined by the opportunity cost of misallocated and inefficient water use. This case study highlighted the importance of accounting for opportunity costs when choosing policy instruments.
• Similar to regulation for WFP, the cap and trade water market required considerable alteration and in particular the inclusion of other policy instruments. These included information provision, community consultation and gradual implementation.
• The design and implementation process for the cap and trade water market was lengthy. If full implementation is considered to be the point at which permanent water entitlement is able to be traded between irrigation districts then the implementation process took at least thirteen years from the publication of the first discussion papers.
• Given the need to reduce the risk of policy failure due to landholders negative attitude toward the proposed policy instrument, the introduction of principles underpinning trading followed by a gradual implementation of full policy instrument was an effective strategy.
• The sale and substitution of temporary entitlement with sales water by diverters in the mid-nineties highlighted the need to monitor policy instruments for use variety by landholders that put the policy objective at risk. 
• The cap and trade water market represented a radical change in policy for the allocation of water. This would suggest that the implementation of the water market could have led to disruptive change in the organisations involved. However this did not seem to be the case given the information provided by our key informants. There appears to be three factors that contributed to this. First, the innovation of the cap and trade water market utilised existing competencies. Second, the innovation did not introduce further complexity to the water allocation process and third, the institutions involved underwent most of the necessary organisational changes to enable the implementation of the market prior to establishment of the water market.

Table 6: Application of PCF to the Water Market

Public Net Benefit, Private Net Benefit Framework

Both key informants and reference documents suggested that there were numerous aspects of the institutional and industry context in the 1980s that led to the cap and trade water market, some of which can be characterised in terms of the private/public net benefits framework. They were as follows: The construction of new dams to provide water for irrigation was no longer politically acceptable. Available supplies of surface water were fully allocated. The introduction of new land levelling technology to the broadacre irrigation industry resulted in increases both farm productivity and demand for water by irrigators. The environmental impacts, in particular salinity, of irrigation were becoming apparent. Increasing importance of new public governance and economic principles in policy decision making. The net public benefits to justify the cap and trade water market were the avoidance of opportunity costs resulting from the inefficient allocation of water between uses, and the inefficient application of water in each particular use. Documents from the time (Nielsen & Associates 1981; ACIL 1984) describe these opportunity costs as being the costs to society of not encouraging the adoption of efficient irrigation technologies, ongoing irrigation of land unsuitable to irrigation and not developing high value irrigation industries such as horticulture. Given the ongoing nature of these opportunity costs and the increasing demand for water these public costs were judged to exceed the private benefits accruing from the users of water at that time.   Conclusion: Public Costs are greater than Private Benefits. Go to negative incentive decision tree.

Negative Incentive Tree

A number of reports to state government by consultants had recommended cap and trade mechanisms for the allocation of water (Nielson & Associates 1981; ACIL 1984). Support for a cap and trade approach to water allocation was to be found both within the bureaucracy and parliament (Parliamentary Salinity Committee cited in Barr 1999, Irrigation Management Study 1986 cited in Department of Natural Resources and Environment 2001). Following is a description of how decision makers may have worked through the negative incentive decision tree when considering a cap and trade mechanism for allocating water to irrigators. Given that the Victorian government was no longer prepared to construct dams the volume of water in Victorian dams defined the aggregate supply of water available to the irrigators of Northern Victoria. At the time decision makers had not predicted any externalities arising from individual water use that was allocated by a water market. Instead government supported the proposal since it was seen to address salinity problems caused by the existing allocating processes (Parliamentary Salinity Committee cited in Barr 1999).  Within the channel distribution system in Northern Victoria water use by irrigators has been measured using Detheridge wheels since the development of the irrigation regions. However flow meters for pumps on river diversions or aquifers were not widely used at the time. Given that the bulk of the water used by irrigators was distributed by the channel system for practical purposes it could be considered that water use was measured most of the time. The northern irrigation regions encompass a range of geomorphology, soil characteristics and agricultural / horticultural industries that strongly suggest that on biophysical grounds alone there will be large variations in private benefit between individuals for the use of a given quantity of irrigation water.  Also, net marginal benefits of water use vary between irrigators because of variable costs and prices for agricultural commodities depending on land use and industry. Conclusion: Given supply and use of the resource can me measured and marginal benefits will vary across uses over time a cap and trade market for water is recommended. Go to I3 Response Framework.

I3 Response Framework

Key informants and documents suggest that there was widespread support for water reform at the time since it had to be addressed to enable ongoing growth in irrigation based industries. The strong link between water reform and economic growth suggests that there was widespread high involvement in the issue of water reform. The level and type of involvement in the proposed cap and trade mechanism however is complex. Key informants suggested that there was considerable support for temporary water trading since it was a more straightforward way of accessing water for on farm development. It also partially offset the costs of owning water right where all or part of a water allocation was unused. However when it came to permanent trading of water rights both documents and key informants suggest that there was considerable opposition within the community for a number of reasons. Firstly it represented a move away from the agrarian principles on which the irrigation regions were founded and secondly there were potential third party impacts such as waterlogging, flooding, salinity, changes in channel capacity and delivery charges. On the whole this suggests that involvement in the policy was high and attitudes toward it were negative, although the temporary trading aspect of the policy instrument was generally high and positive. However to achieve government objectives permanent trading had to be implemented, which meant that a number of changes had to be introduced to reduce the number of irrigators that had negative attitudes toward the proposal. They were as follows: Community involvement and consultation in the design and implementation process. Through Irrigation Advisory Boards and community forums run by the VFF and State Rivers. Planning process to address third party impacts, this led to salinity, drainage assessments and channel capacity assessment. Introduce trading in a staged manner, first temporary trading, second permanent trading within district, and third permanent trading between districts. By the end of the consultation period, 80% of the irrigation community supported temporary trade with around 50% supporting permanent trade. Conclusion: Adjust design and implementation process to include community consultation and third party impacts. Government willing to act without consensus. Go to Use Variety tree

Use Variety Tree

Given that the design and implementation of the water market in the late 80s and early 90s was guided by a philosophy that believed trade in water rights should have as few restriction as possible (Department of Natural Resources and Environment 2001) it is of no surprise that there was considerable potential for use variety.  However, it was apparent early on in the implementation phase that there were a number of impacts resulting from trade that were counter-productive in terms of the economic and natural resource objectives of government policy.  In particular the substituting of sales water for other entitlement was an issue in the first few years of implementation. An increase in overall trading in 1994/95, partly due to dry conditions and low allocations, led to a jump in 'sales' trading. In particular 'sales' water from low value agriculture in the Pyramid Boort and Tragowal Plains. This prompted a rethink about the trading of 'sales' water.  The major concern was that private diverters, who paid little for their water licences and often did not use them, were suddenly cashing in on the 'sales' market.  This reduced the reliability of irrigation supplies generally and led to higher overall usage at the time of MDB Interim Cap implementation. In 1995 'sales' trading by private diverters was banned in northern Victoria. To prevent this occurring again private diverters were not allowed to use any 'sales' water if they temporarily transferred any of their other water allocation from 1995 onwards. From 1997 gravity irrigators lost access to all sales above 30% if they temporarily traded any water right or any of the first 30% sales. Overall we concluded the potential for use variety was high, however it could be accounted for by placing conditions on water trading and usage.  These changes did not result in changes in the communities' involvement with the policy instrument. Conclusion: Potential for use variety is high, but negative impacts of use variety can be accounted for by changes in rules for water trading and use. Go to Scoping tree

Scoping Tree

Temporary trade alone did not fully achieve the outcome of encouraging the most economically efficient use of water since it did not allow the permanent shift of water resources to high value uses. In particular it did not give the horticultural and dairy industries the permanent access to high security water entitlement they needed for ongoing investment. A survey of trade in 1990 found that 60% of water sold had previously been unused and 60% of water purchased was to improve reliability of supplies rather than expand irrigation. Key informants told us that community support for permanent trade increased throughout the early 1990s. By this stage irrigators had grown more confident with trading and were seeing the benefits trade generated. The volume of water traded on a permanent and temporary basis increased rapidly from the mid 90s onwards. The nature of this trading suggests that the initial policy outcomes of reallocating water resources to high value uses and lower environmental impact areas will be achieved. Both background documents and key informants attributed this acceleration in temporary and permanent water trading to four key factors. First, trading rules were widened significantly in 1994 allowing the trade out of irrigation districts, between supply systems (Goulburn to the Murray) and into the Sunraysia region. Second, since 1994/95 conditions have generally been dry this has limited the availability of water resulting in a trend toward lower allocations and little if any 'sales' water.  Third, the Cap on diversions in the Murray Darling Basin and Victoria's steps to implement the Cap combined with dry seasons demonstrated to many irrigators the need to manage risk by purchasing water entitlement. Fourth, irrigators had gradually gained more knowledge and confidence in trading and how to operate in the market.  Conclusion: Permanent trading was feasible when introduced in an gradual fashion, beginning with temporary trading. Go to Policy Innovation Tree.

Policy Innovation Tree

At first glance a cap and trade market in water entitlements is radically different to the allocation processes of the mid 1980s that preceded it. However in considering the underlying principles of the water market that was implemented and the principles being introduced into the institutional environment prior to it's implementation a different picture emerges. The cap and trade policy instrument to allocate water resources was preceded by considerable institutional change that reflected the trends in governance of the time. Suggesting that the degree of institutional change the cap and trade policy instrument represented was confounded with the impacts of other institutional change at the time. What is clear though is that the introduction of these reforms went some way in preparing the ground for a policy instrument such as a cap and trade market in water entitlements. These enabling reforms were described by key informants as: Introduction of User Pays principles in the administration of irrigation infrastructure. Decentralisation of decision making leading to greater community consultation and participation in the decision making processes involved in administrating the irrigation regions. Separating responsibilities for providing policy advice to government from responsibilities associated with delivering water to irrigators by the creation of the Department of Water Resources (DWR) and Rural Water Commission (RWC) respectively (Barr 1999). Shift in influence away from engineering profession to economics, accountancy, ecology and environmental science (Considine 1994 cited in Bjornlund 1999). Each of these reforms represents either an architectural or radical change in the institutional environment in comparison to what had gone on before. Given this institutional context key informants suggested that the introduction of cap and trading in water entitlements did not disrupt existing institutional structures. Indeed one informant suggested that the new allocation process reduced the difficulty associated with managing water allocations when there was a large volume of unused water entitlement. The same informant also suggested that the introduction of this policy instrument at a time of considerable institutional change led to an institutional environment by the mid 1990s that was well adapted to working within the broad and flexible boundaries of the water market.

Policy  Innovation continued……
Impact of introduction of cap  and trade market for water entitlements
Organisation	Impact
Department of Water Resources	Policy is consistent with governance principles such as "new public  management" and "microeconomic rationalisation" of the time. Department of  Water Resources actively restructured government policy along these new  governance principles. Impact on this organisation is incremental.
RWC / G-MW	The cap and trade water market represented considerable change in policy  concept and architectural principles in how the RWC allocated water to  irrigators. However the impact of the change on the RWC/G-MW was limited to the  incorporation of an administrative arm of the organisations responsible for  transferring water entitlements between irrigators. Despite the change in policy  concept and architectural principles this suggests the cap and trade water  market is a modular innovation, and  was not considered disruptive to the organisation as a whole.
Industry	For irrigators the  introduction of the cap and trade market for water entitlements was a  continuation of micro economic reforms that had started ten years earlier.  These reforms introduced new principles into public policy, in particular the  importance of optimising economic efficiency. This was contrary to the old  rural paradigm with principles derived from agrarianism. However  the act of buying water under the new allocation process was far simpler than  before and was not different to purchasing other inputs needed to carry out a  farming business. This suggests that the new allocation process was more likely  to be competence enhancing than destroying for industry. Impact on industry is architectural.
Conclusion: A cap and trade  market in water entitlements was feasible, although it required considerable  adjustment to alter the negative attitude of irrigators towards the proposed  policy instrument. These adjustments included gradual implementation, community  consultation and information provision. Use variety was potentially an issue  and the policy instrument eventually had to be adjusted to account for it.  Changes to reduce negative involvement of irrigators were sufficient to  ameliorate the disruptive impact of the policy instrument on industry.

Discussion

In the introduction we argued that the frameworks that have been proposed in the literature for choosing between policy instruments lack rigorous, practical processes for evaluating policy instruments, particularly with respect to anticipating variety in the reactions of landholders and organisations to the introduction of new policy instruments. We developed the PCF to build on existing frameworks by providing a systematic process for incorporating the potential variety in the behaviour of landholders that are the target of a policy, and organisations implementing a policy, into the choice of policy instrument.

Applying the PCF to the four case studies suggests that the PCF can reveal relevant variety in the behaviour of both landholders and institutions that, when accounted for, increases the chances that the environmental outcome will be achieved. This is discussed in detail below.

Accounting for variability in landholder behaviour

Of particular value in the case studies was that the PCF focussed attention on systematically identifying and explaining variations in landholder and institutional behaviour. The presence of such variations raised important issues for the policy-makers involved in the case studies such as:

1. The importance of clearly describing variations among landholders in the levels of private net benefit associated with a proposed policy instrument and grouping landholders into clearly defined segments based on these variations. For example, in the application of the PCF to the incentive program for automatic irrigation segments of landholders were identified for whom the program was irrelevant. The PCF results supported policy-makers' expectations that the program would have a limited impact. This knowledge was critical in assessing the success of the program in terms of the scale and rate of shift to automatic irrigation among landholders.
2. The need to consider whether there are likely to be some groups of landholders that might react in ways that could endanger the success of a proposed policy instrument.

The results of the PCF indicated that some of the landholders that exhibited high involvement with the implementation of Whole Farm Plans had unfavourable attitudes towards this instrument. This potentially placed the successful implementation of WFP at risk. In this case policy-makers attempted to reverse unfavourable attitudes by incorporating complementary policy instruments, such as an incentive program and other measures, to address the concerns of landholders.

1. The potential for landholders to respond to policy instruments in ways that were counter-productive. In the case of the cap and trade water market, the potential for use variety to place achievement of the policy outcome at risk action was evident and action was quickly taken to prevent over-utilisation of the resource.
2. Considering whether a package of policy instruments targeting different groups, may be required to achieve a policy objective. For example, in the application of the PCF to the EIT research project segments of landholders were identified for whom the policy instrument was irrelevant. This knowledge gave the policy maker the opportunity to modify the policy instrument. In this instance the project was modified to include sub-surface irrigation to increase its relevance to landholders.

Generally speaking, the results of the case studies suggest that the PCF highlighted issues in regard to the potential responses of landholders that were critical in practice to the choice of policy instrument. The application of the PCF also appeared to direct policy-makers toward solutions to these issues that were consistent with those that were actually identified in practice. Finally, from the perspective of the case study participants, the PCF provided a systematic framework for describing the influence of landholder behaviour on their choice of policy instruments that was largely consistent with their reasoning in practice.

Accounting for variability in institutional behaviour

The PCF was able to reveal two important aspects of the institutional environment that decision makers had to take into account when developing policy instruments. These were:

1. Identifying which organisations involved in developing and implementing a policy instrument were most likely to experience disruptive organisational change. This is critical to successful policy implementation because organisations experiencing disruptive change are those most likely to undermine the effectiveness of the policy instrument if their issues are unaddressed (Kaine et al. 2006). The application of the PCF in the case study on the incentive program for automatic irrigation provides an example where the choice of policy instrument was influenced by limitations in institutional capacity of organisations charged with implementing policy.
2. Consider a variety of actions to address the different impacts of the policy innovation on the different organisations involved in implementing the policy instrument. The application of the PCF highlighted the need for a systematic method for assessing the impact of new policy instruments on organisations. Although the water market represented a significant change in the principles governing the allocation of water to irrigators, the policy change was incremental or modular for many of the organisations involved. Hence, the organisational consequences were relatively minor. On the other hand, the Centre Pivot technologies research project entailed radical changes in the principles governing research in the organisations involved. Hence, the organisational consequences were substantial and extra resources and changes to the research program were required to ensure successful implementation.

These results demonstrate the capacity of the PCF to reveal complexity in institutional responses to new policy instruments. The PCF represents an important step toward a systematic method for anticipating the institutional changes associated with implementing policy instruments.

Generally speaking, the results of the case studies suggest that the PCF highlighted issues in regard to the potential responses of organisations that were critical in practice to the choice of policy instrument. From the perspective of the case study participants, the PCF provided a systematic framework for describing the influence organisational behaviour had on their reasoning and practice leading up to the choice of a policy instrument.

Implications of policy instrument selection

A systematic process for revealing variety in the behaviour of both organisations and landholders has some important implications for policy instrument selection. First, such a process supports the logical development of a package of policy instruments to achieve a natural resource policy outcome. The existing literature describes the benefits of packages of policy instruments designed to complement each other when addressing a specific environmental problem (Gunningham and Gabrosky 1998; Gunningham and Sinclair 1999; Hatton MacDonald et al 2004; Whitten et al 2006). The whole farm plan case study, for example, highlighted the valuable contribution complementary policy instruments make to the successful implementation of a policy instrument. Yet the evidence from the case studies suggests that there are not  systematic processes available for policy-makers to use to develop packages of complementary policy instruments to achieve policy objectives. The PCF goes some way to address this since the application of the PCF highlights circumstances when a mix of policy instruments may be needed to accommodate variety in the behaviour of landholders. However the PCF may not necessarily assist policy-makers to identify what types of complementary policy instruments should be considered in all circumstances.

Second, the PCF, by systematically identifying variety in the behaviour of landholders, also supports the 'smart regulation' approach to policy instrument choice advocated by other policy researchers such as Gunningham and Gabrosky (1998) and Gunningham and Sinclair (1999). The PCF provides a mechanism for relating the behaviour of landholders to their context. This knowledge is critical to identifying the type of policy instrument that is most likely to address landholder concerns. For example, planning permits for on farm earth works originally expired within twelve months of issue. Changing permit conditions to allow an indefinite extension of the planning term recognised the reality that remodelling of irrigation farms generally takes place over ten years. This example indicates the potential for the PCF to provide practical direction on policy instrument design issues.

Third, experience with the case studies demonstrated that the usefulness and value of the PCF was proportional to the data that was used. As with any decision aid the PCF is more likely to generate reliable information when more reliable data is used.  We used a variety of information sources and multiple key informants to collect the information needed to apply the PCF. In principle, we suggest that policy-makers need to use a range of sources of information that is sufficient to account for all the potential variety in the behaviours of landholders and institutions. In practical terms this means that decision-makers should attempt to use a variety of unrelated information sources when applying the PCF.

Fourth, each of the case studies was relatively rich in information in that key informants and historical documents were able to provide important information on private and public net benefit, level of landholder involvement, potential for use variety, and the ability of the policy instrument to achieve its objective. This made it relatively straightforward for us to work through the PCF. In circumstances where limited information is available the PCF may be valuable as a means of revealing strategically critical information gaps.

Fifth, in conducting the case studies it became apparent that existing review processes could not be relied upon to identify strategically important changes in farm or institutional contexts. It was also evident that, while the PCF could contribute to the identification of important contextual changes, the PCF did not in itself contain a mechanism that could be reliably used to determine when such reviews should be triggered. This is an important policy issue since both landholder and institutional contexts are subject to both continual and sudden change. Occasionally these changes in context warrant changes in the choice of policy instrument. The inability to identify such trigger points results in the continuation of inappropriate policy instruments.

In the future it will be important to develop and incorporate processes into the PCF that can assist in identifying strategically critical changes in context that impact on the policy objective or the likelihood of successful implementation of a policy instrument.

Conclusion

The PCF was developed with a view to identifying variety in landholder and institutional responses to policy instruments. The PCF is intended to complement other frameworks for choosing policy instruments (see Connor and Bright 2003; Gunningham and Gabrosky 1998; Gunningham and Sinclair 2002; Hatton MacDonald et al 2004; Martin and Verbeek 2006; Ridley and Pannell 2006; Tietenburg and Johnstone 2004; Whitten et al 2006; Ward et al 2005; Young et al 1996). The four case studies suggest that the PCF was able to complement existing frameworks since the actions of landholders and institutions in each case study were consistent with the principles underpinning the various frameworks that constitute the PCF.  However, applying the PCF to contexts outside of the SIR may reveal limitations that were not apparent from the case studies we conducted.

In summary, using the PCF to analyse landholder and organisational behaviour had the following benefits:

• The PCF revealed the variety in landholder responses to the policy instrument and the causes of that variety.
• The PCF revealed whether a policy instrument could address the full range of landholder behaviours.
• The PCF made explicit the need to address unfavourable attitudes landholders had toward a policy instrument.
• The PCF provided a sound basis for suggesting modifications to a proposed policy instrument to account for differences in landholder's responses. Similarly, the PCF provided a sound basis for suggesting when complementary policy instruments were needed to account for responses by landholders that increased the risk of not achieving the policy objective.
• The PCF provides a systematic method for demonstrating that individual organisations were affected in different ways by the introduction of a new policy instrument.
• The PCF could be used as a means of revealing information gaps that are strategically critical to the successful implementation of policy instruments.

The evidence from the case studies suggest that the PCF would be useful in assisting policy-makers to systematically select and customise policy instruments to achieve natural resource outcomes. In working through the case studies it became evident that a wide range of factors influence the behaviour of landholders and can impact on the efficiency or effectiveness of a policy instrument and that these factors can change over time. In principle, the PCF could be used in an iterative fashion to take into account changing circumstances and thereby support double loop organisational learning processes.

The PCF in its present form would support the development of methods for identifying strategically critical issues that will impact on the efficiency and effectiveness of a policy instrument. A key area for future work is the development of processes that can assist in identifying strategically critical changes that impact on the value of policy outcomes or the likelihood of successful implementation of a policy instrument.

References

ACIL (1984). Summary report to Department of Water Resources Victoria: Transferability of Water Entitlements. ACIL Australia PTY. LTD., Melbourne.

Assael H (1998). Consumer Behaviour and Marketing Action. South Western College Publishing, Cincinnati, Ohio.

Barr N (1999). Salinity control, water reform and structural adjustment: The Tragowel Plains Irrigation District. PhD thesis, Institute of Land and Food, University of Melbourne, Australia.

Bjornlund H (1999). Water trade policies as a component of environmentally, socially and economically sustainable water use in rural southeastern Australia. PhD Thesis. University of South Australia, School of Law and Legal Practice and School of International Business, Division of Business and Management.

Catchment and Agricultural Services (2004). Review of the Whole Farm Plan Program in the Shepparton Irrigation Region.  Department of Primary Industries Victoria, Tatura. 

Connor J and Bright M (2003). Which Market-based instrument for natural resource management? Paper prepared for presentation at the 47th Annual Conference of the Australian Agricultural and Resource Economics Society. CSIRO Land and Water.

Department of Natural Resources and Environment (2001).  The Value of Water: A guide to water trading in Victoria.  Victorian Government, Melbourne.

Department of Primary Industries (2006). Efficient Irrigation Technologies to Match Soils and Dairy Farming Systems (DAV11700). Final report to Dairy Australia. Department Primary Industries, Tatura.

Dholakia U (2001). A motivational process model of product involvement and consumer risk perception. European Journal of Marketing, Vol. 35 No.11/12, pp.1340-62.

Gunningham N, Grabosky P and Sinclair D (1998). Smart Regulation: Designing Environmental Policy. Oxford University Press Inc, New York, United States.

Gunningham N and Sinclair D (1999). Regulatory Pluralism: Designing Policy Mixes for Environmental Protection. Law and Policy, Volume 21, Number 1.

Gunningham N and Sinclair D (2002). Leaders and Laggards: Next Generation Environmental Regulation. Greenleaf, UK.

Hatton MacDonald D, Connor J and Morrison M (2004). Market-based instruments for managing water quality in New Zealand.  Final report prepared by CSIRO Land and Water for New Zealand Ministry for the Environment.

Hill M, Linehan C and Murdoch H (2004). Efficient irrigation technologies to match soils and dairy farming systems.  Practice Change Research Report, Department of Primary Industries, Tatura.

Johnson F, Kaine G, Ford J and Leth M (2006). A framework for selecting policy instruments in natural resource management. Practice Change Research Working Paper 07/06, Department of Primary Industries, Tatura.

Kaine G and Bewsell D (2000). Irrigation Systems, Irrigation Management and Dairy Farming (First and Second Reports). Reports to the Victorian Department of Natural Resources and Environment, School of Marketing and Management, University of New England, Armidale.

Kaine G and Higson M (2006a). Understanding variety in landholders' responses to resource policy. Australian Agribusiness Review, 14, Paper 6.

Kaine G and Higson M (2006b). Policy as an Innovation. Practice Change Research Working Paper 02/06, Department of Primary Industries, Tatura.

Kaine G, Higson M, Sandall J and Lourey R (2006). Policy as an Innovation: Case Studies in Australia and New Zealand. Practice Change Research Working Paper 07/06, DPI Victoria, Tatura.

Kaine G and Johnson F (2004a). Applying Marketing Principles to Policy Design and Implementation. Social Research Working Paper 02/04, AgResearch, Hamilton, New Zealand.

Kaine G and Johnson F (2004b). Water Policy in New Zealand and Australia: We're not playing for tiddlywinks. Proceedings of the New Zealand Grasslands Association (66), Ashburton, October.

Kaine G, Johnson F, Ford J and Leth M (2006). A review of approaches for selecting policy instruments. Practice Change Research Working Paper 06/06, Department of Primary Industries of Victoria, Tatura.

Kapferer J and Laurent G (1986). Consumer involvement profiles: a new practical approach to consumer involvement. Journal of Advertising Research, Volume 25, Number 6, pp.48-56

Lourey R, Kaine G and Murdoch H (2006). Understanding private landholders decisions about pest management. Practice Change Research Working Paper 08/06, Department of Primary Industries of Victoria, Tatura.

Martin P and Verbeek M (2006). Sustainability Strategy. Federation Press, Australia.

Maskey R (2002). Attitudes of farmers about automation of flood irrigation. Report to the Goulburn Broken Catchment Management Authority. Department Natural Resources and Environment, Tatura.

Mittel B, Lee MA (1987) A causal model of consumer involvement. Journal of Economic Psychology 10, 363-389

Murdoch H, Bewsell D, Lourey R and Kaine G (2006). Understanding people's response to biosecurity regulation. Paper presented to 3rd National Conference of the New Zealand Society for Risk Management Inc, Decision making in uncertain times, Auckland New Zealand, October 17.

National Action Plan for Salinity (2005). National Market-based Instruments Pilot Program: Round One. An Interim Report, by the National Market Based Instrument Working Group, Canberra, Australia.

Nielsen & Associates (1981). Study of irrigation management in Victoria. For the Public Bodies Review Committee, parliament of Victoria. Volume 1, October. 

Pannell DJ (2006). Public benefits, private benefits, and the choice of policy tool for land-use change. Available [online] at http://cyllene.uwa.edu.au/~dpannell/dp0601.htm. [June 15th 2006].

Poisesz T and deBont C (1995). Do we need involvement to understand consumer behaviour? Advances in consumer research, Voume 22, pp.448-452.

Quiggan J (2001). Environmental economics and the Murray Darling river system.  The Australian journal of Agricultural and Resource Economics. Volume 45, Number 1, pp.67-94.

Reeve I, Frost L, Musgrave W and Stayner R (2002).  Agriculture and natural resource management in the Murray-Darling Basin: A policy history and analysis – Overview report. Institute of Rural Futures, University of New England, Armidale, NSW.

Ridley A and Pannell D (2005). SIF3: An investment framework for managing dryland salinity in Australia. SEA working paper 1901. CRC for Plant-based Management of Dryland Salinity, University of Western Australia, Perth.

Sampson K, Wall R, and Norman C (2002). Efficient irrigation and prosperous farming - synchronising with the land and the environment. Project proposal for Water for Growth Initiative, pp.1-14.

Tietenberg T and Johnstone N (2004). The Management of Ex Post Public Policy Evaluations: Tradable Permits. Policy Evaluation, Design and Reform. Organization for Economic Co-operation and Development, Paris, pp.173-189.

Verbeke W, Vackier I (2004) Profile and effects of consumer involvement in fresh meat. Meat Science 67, 159-168.

Ward J, Bryan B, Gale G and Hobbs T (2005). Market-Based Instrument approaches to implementing priority revegetation in the South Australian Murray-Darling Basin. CSIRO Land and Water Client Report.

Whitten SM, Coggan A, Reeson A and Shelton D (2006) 'Market based instruments: Applying MBIs in a regional context', Report to the Joint Venture Agroforestry Program and Rural Industries Research and Development Corporation, CSIRO Sustainable Ecosystems, Canberra.

Young M, Gunningham N, Elix J, Lambert J, Howard B, Grabosky P and McCrone E (1996). Reimbursing the future: An evaluation of motivational, voluntary, price-based, property-right, and regulatory incentives for the conservation of biodiversity – Part 1.  Prepared by CSIRO Division of Wildlife and Ecology, the Australian Centre for Environmental Law, and Community Solutions January 1996. Biodiversity Series, Paper No. 9. Biodiversity Unit.  Available [Online]: http://ea.gov.au/biodiversity/publications/series/paper9/index.html. [June 2nd 2003].

Young M and Hatton MacDonald D (2003). An Opportunity to Improve Water Trading in the South East Catchment of South Australia. Water Policy Volume 5, Issue 2, 127-146

Zaichkowsky JL (1986) Conceptualizing involvement. Journal of Advertising 15, 4-14

Appendix 1: Sequence of decision trees used to work through Policy Choice Framework for automatic irrigation case study

Step 1. Private/Public Net Benefits

See page 39 for a description of this case study, including how the first step, the Private/Public Net Benefits Framework, is worked through. The sequence of decision trees begins with the second step, technical feasibility.

Step 2. Technical Feasibility Decision Tree for Positive Incentives

Step 3. I3 Behavioural Response for Positive Incentives

Step 4. Use Variety

Step 5. Scoping

Step 6. Policy Innovation

Appendix 2: Sequence of decision trees used to work through Policy Choice Framework for whole farm plan Case Study

Step 1. Private/Public Net Benefit

See page 43 for a description of this case study, including how the first step, the Private/Public Net Benefits Framework, is worked through. The sequence of decision trees begins with the second step, technical feasibility for negative incentives.

Step 2. Technical Feasibility Decision Tree for Negative Incentives

Step 2 continued. Technical Feasibility Decision for Selection of Regulation

Step 3. I3 Behavioural Response for Negative Incentives

Step 4. Use Variety

Step 5. Scoping

Step 6. Policy Innovation

Appendix 3: Sequence of decision trees used to work through Policy Choice Framework for research case study

Step 1. Private/Public Net Benefit

See page 50 for a description of this case study, including how the first step, the Private/Public Net Benefits Framework, is worked through. The sequence of decision trees begins with the second step, technical feasibility for negative incentives.

Step 2. A technical feasibility decision tree is not required for research.

Step 3. I3 Behavioural Response for Research

Step 4. Use Variety

Step 5. Scoping

Step 6. Policy Innovation

Appendix 4: Sequence of decision trees used to work through Policy Choice Framework for water market case study

Step 1. Private/Public Net Benefit

See page 55 for a description of this case study, including how the first step, the Private/Public Net Benefits Framework, is worked through. The sequence of decision trees begins with the second step, technical feasibility for negative incentives.

Step 2. Technical Feasibility Decision Tree for Negative Incentives

Step 3. I3 Behavioural Responses for Negative Incentives

Step 4. Use Variety

Step 5. Scoping

Step 6. Policy Innovation

Foot Notes

¹  The level and source of landholder's involvement in the policy issue reflects the importance, interest and concern landholders have in the policy issue and provides a basis for anticipating the likely effort they will invest in responding to the issue.

²  The level and source of landholder's involvement in the policy instrument reflects the degree of fit between the landholder's context and the policy instrument and provides a basis for anticipating the likely effort they will invest in responding, favourably or unfavourably, to the policy instrument.

³Where Private and Public Net Benefit < 0 they are described as Private and Public Net Costs.