SEA Working Paper 98/03

On the use of individual-farm models for analysis of resource conservation problems

David J. Pannell

Agricultural and Resource Economics, The University of W.A., Nedlands, 6907

Abstract

Resource conservation issues are widely held to be larger in scope than individual farms. Partly as a consequence of this, there has been an increasing emphasis on catchments and groups of farmers in agricultural extension. For these reasons, it has mistakenly been concluded by some that economic decision models at the individual-farm level have little or no contribution to make to evaluation of resource conservation issues. There are several reasons why this is not true. (a) Notwithstanding the reality of catchment processes (especially water-related) beyond farm boundaries, and the undoubted value of group-based extension, final decision making still rests with individual farmers. For resource conservation practices, as for others, private financial considerations are key drivers of management decisions. Individual farm models provide useful information about economic incentives facing farmers. (b) Even where the model is to provide direct support to decision makers who are concerned with aggregate rather than individual effects, individual farm models can contribute in a number of ways. These include providing information about the costs of reducing land degradation, and the likely responses of farmers to potential policies. (c) For several major land conservation issues, the central concerns are not with spillover effects, but with poor decision making due to poor information. In these cases, individual farm models are perfectly appropriate. (d) Even where externalities exist, they do not necessarily lead to market failure. There are reasons why the optimal farming strategy from a private farmer’s perspective may still correspond to the socially optimal strategy, and these reasons are likely to apply in practice, at least in some cases. The various potential uses and contibutions of individual farm models are outlined and illustrated with examples. However, it is important to recognise the broader context within which individual farms sit and that this context places limitations on the generality of results from individual-farm models.

Introduction

There is a family of detailed whole-farm bioeconomic models in use in different states of Australia. The success of the original MIDAS model in Western Australia (Kingwell and Pannell, 1987; Pannell, 1996) has inspired the development of MUDAS (Kingwell, 1993) in Western Australia, PRISM in Victoria and New South Wales and MIDAS-EP in South Australia. These models represent individual representative farms in different regions, and are designed to identify whole-farm management strategies that maximise profit subject to a detailed set of constraints. They are characterised by a high level of biological and technical as well as economic detail, and are all solved using mathematical programming software.

Although these models were originally developed primarily for application to short-term productivity-related issues, they are now starting to be used to contribute to analyses of longer-term resource conservation issues. Some doubts have been expressed about the legitimacy of their use for these purposes. The aim of this paper is to explore the benefits and limitations of using farm-level models to analyse the range of resource conservation problems related to agriculture in Australia. It addresses questions such as:

• Under what circumstances can these models capture all or most of the factors relevant to resource conservation issues?
• Where factors external to the farm-level models are important, what scope is there for such models to contribute usefully? What limitations need to be recognised in these cases?

Why Might Individual-Farm Models be Inadequate?

There are several reasons why the use of farm-level models for analysis of resource conservation issues might be questioned.

Some land degradation processes occur over scales larger than individual farms. In particular, dryland salinity can arise from processes occurring over catchments that encompass several or more farms.

At least partly inspired by point (a) there has been a growing emphasis on the importance of farmers operating in groups to combat land degradation. This has contributed to a dramatic shift in the emphasis in agricultural extension in Australia towards group-based approaches, and at least in some quarters there has been a major de-emphasis of the individual farmer’s role in resource conservation.

There is a growing interest in the potential for policy initiatives to combat land degradation (e.g. Industry Commission, 1997; Hayes, 1997), and such policies will have industry-wide impacts, in pursuit of broader objectives than held by individual farmers.

In the following section, it will be argued that none of these points invalidates the use of individual farm models to contribute valuable information to promote better decision making about resource conservation. This contribution may be to decision making by individual farmers, groups of farmers, extension agents, scientists and policy makers.

In Defence of Individual-Farm Models

The source of decisions

As noted earlier, there has been an increased focus on catchments and farmer groups for extension and certain aspects of planning. Although this clearly has positive aspects (Marsh and Pannell 1998), there has been a tendancy towards unrealistic expectations about the extent to which decision making can be collectivised simply by bringing together groups of farmers to learn about and discuss their resource conservation problems.

Even where the physical source of a problem is at a scale greater than an individual farm, actual decision making still occurs by individual farmers based on their own objectives, perceptions, and circumstances. Among their prime objectives are those relating to the economic viability of their farm business. Notwithstanding comments by some commentators, economic objectives are of paramount importance in farmers’ judgements about the desirability of resource conservation practices. For example, much has been made of the need to promote an ethic of stewardship among farmers, but based on statistical analysis of actual farmer behaviour Sinden and King (1990) concluded that,

"While the stewardship motivation and personal factors encourage perception and recognition of a problem, economic factors promote actual adoption." (p. 179)

Similarly, Cary and Wilkinson (1997) found that,

"Generally, the best way to increase the use of conservation practices to overcome land degradation … will be to ensure the practices are economically profitable." (p. 20)

Therefore, in considering farmers’ adoption of conservation practices, it is critical to understand the economic incentives they face at the individual farm level. This, then, points to a key role for individual-farm models that encompass economics, biology and physical resources.

This is not to say that there are not other influences on farmers’ management decisions, but it is to strongly assert that the off-farm influences do not replace or subsume the influence of whole-farm profitability as an objective.

The creation and nurturing of farmer groups may help to exert peer pressure for management in the common good, but it cannot take final decision making power away from individual farmers, and it cannot remove the need of each farmer for economic viability. Indeed, the potential of groups to wield influence in the common good are constrained because the farmers who most need to be influenced can have considerable economic incentives not to be influenced. And it is easy for them to avoid this influence by simply opting out of the group.

The bottom line is that formation of groups per se does not change the underlying economic incentives. It does not overcome the common-property problem of negative spillovers because it does not provide a mechanism by which gains and losses can be redistributed to create the incentives needed.

To restate the key point of this discussion, decision making power continues to reside in the hands of individual farmers, so models of individual farms can play a valuable role in analysing the economic incentives faced by farmers to manage or not manage in particular ways. Thus it is appropriate to apply individual farm models to derive:

1. information for farmers about the economic attractiveness of particular practices or technologies, or
2. information for researchers, extension agents or policy makers about the likely level of adoption of particular practices or technologies,

The relationship to comprehensive models

Another important related area is the evaluation of policies and incentive systems intended to overcome spillover problems. An example of a model designed to represent spillover effects for dryland salinity is that of Greiner (1998). Here it might be considered that individual farm models have no role to play, but in fact they can contribute greatly. Consider the range of factors that impinge on an evaluation of this type. Figure 1 shows one representation.

Figure 1. Broad overview of framework within which society chooses policies and institutional arrangements for agricultural resource management.

Box A in Figure 1 represents the range of policy instruments available to government to influence farm management practices. These would include:

1. Extension

• One to one
• Groups
• Demonstrations
• Mass media

2. Community involvement
3. Research
4. Development
5. Carrots and sticks

• Subsidies for adoption
• Subsidies for interest rates
• Subsidies for information/monitoring
• Pollution taxes
• Tradeable pollution permits
• Safe minimum standards/penalties
• Cross compliance requirements

6. Altering property rights
7. Public acquisition and management
8. Initiatives to reduce transaction costs

These policy instruments affect the farmer’s chosen set of management options (Box E) both directly and indirectly via their impact on the economic and social environments (Boxes B and C). These linkages are presented in slightly more detail in Figure 2.

Figure 2. Linkages between policies, institutional arrangements, the economic and social environment and the farmers’ decision problem

As a result of his or her management strategy, each farmer generates a level of profit, but also possibly a set of off-farm impacts, potentially including outcomes such as increased salinity on neighbouring farms, increased salinity in waterways, and damage to native flora and fauna (Box F).

These outcomes can then be evaluated in terms of their consistency with broad government policy objectives, which might include economic, social and environmental objectives (Box G). The links between farm management and broad government objectives are depicted in Figure 3.

Figure 3. The influences of farmers’ choices on broader social objectives related to salinity.

This framework, then, outlines the linkages between potential policy instruments and government objectives. It is clear that an ability to predict farmer decision making is a pivotal aspect of any evaluation of potential policies. For a complex farming system, this is a much more difficult task than often recognised. For example, Pannell (1995) presents a list of 28 types of factors that influence the profitability of legume-based farming systems (grouped into short-term profit factors, inter-year dynamic factors, longer term-dynamic factors, risk factors and whole-farm factors). It should be clear from this extensive list that simple economic assessments of complex farming systems are likely to be of little value, or even of negative value if they mislead. Detailed individual farm models have a considerable potential to contribute in this area. In a model that attempts to be comprehensive (in the sense of covering several farms in a region or catchment), it is very difficult to represent the farmer’s situation in sufficient detail to predict with reasonable accuracy.

There are, therefore, outcomes in this area where individual-farm models can play a key role in providing better quality information than otherwise available. These include:

1. information about the levels of resource degradation or spillover effects that result when an individual farmer pursues management practices that are optimal from the perspective of financial self interest.
2. information about the shadow price of land of different soil types in different regions, which indicates the opportunity cost of land converted to other uses for conservation purposes, and the value of land lost to degradation processes. For example, Table 1 shows annual shadow prices of land from the eastern wheatbelt of Western Australia.
3. information about the private financial costs of reducing resource degradation or reducing spillover effects (in other words, the trade offs between profit and environmental objectives at the farm level). For example, Figure 4 shows illustrative results from MIDAS of the costs of different levels of recharge abatement for a farm in the low rainfall wheatbelt of Western Australia.
4. information about the likely impacts of particular policies or incentive systems on farmer behaviour and consequently on resource degradation or spillovers. For example, Figure 5 shows an illustrative result from a relatively simple dynamic model used to evaluate the private benefits and costs of planting trees to reduce salinity.

Table 1. Shadow prices of land near Merredin, Western Australia.

Source: MIDAS version Mer98r4d

Figure 4. Recharge abatement cost curve for a farm in Western Australia’s eastern wheatbelt.

Figure 5. Payoff to individual farmer from different areas of trees with and without subsidy of $188/ha/year.

Thus an individual farm model contributes in crucial ways to the more comprehensive analyses required to provide:

1. information for policy makers, researchers or extension agents about the overall social benefits and costs (including spillovers) from adoption of particular practices or technologies,
2. the extent to which these benefits and costs differ from those faced by the individual adopters, and the extent of market failure.
3. information for policy makers about the large-scale (e.g. catchment-scale) benefits and costs of policy measures designed to overcome "externality" or spillover problems.

The individual-farm model can either be used to provide data and relationships needed for a comprehensive assessment, or it may be a stepping stone on the way to development of a more comprehensive model.

The marginal benefits of a comprehensive model

Although individual farm models do not provide a comprehensive assessment of resource conservation policies, they capture a large part of the system. In Figure 1, if Boxes A, B and C are treated as exogenous, then a model like MIDAS can provide information about the impacts of A and B on E and F. The aspects not represented are:

• social factors, which would probably be missing from higher level economic models as well,
• the dynamic linkages between farms, although exports of negative externalities from farms can be estimated,
• an endogenous link between the outcomes and governement objectives.

I believe that as long as policy makers are aware of the context outlined here, and of the omissions dot-pointed above, the output of carefully conducted analyses with individual farm models can be highly valuable to policy makers in their own right, even without being fed through a more comprehensive model. It is possible that with the application of careful judgement, the additional improvements to decision making that are possible with a comprehensive model are not large.

It should also be noted that there important resource conservation issues where the main reasons for government intervention are perceived information failures rather than externalities. Examples of this include soil acidification, soil fertility decline and herbicide resistance. These are primarily private rather than public problems, but they are slow acting and relatively new concerns for most farmers, so current understanding of the problems and their treatments among the farming community is low enough to warrant government-funded extension efforts. In these cases, there is clearly no cause for catchment or regional models; individual-farm or even paddock-level models provide all of the economic analyses needed.

Even where externalities exist, it is not always the case that they have a major influence on the set of farm management practices that would be socially optimal. For example, Salerian (1991) found that internalising the externalities from salinity for a particular sub-catchment in Western Australia made little difference to the optimal farm strategy. The best farm practices from the farmer’s point of view were also optimal from a catchment perspective. In other words, the existence of externalities does not necessarily imply that there is an important market failure. It may be that the best available options advance both private and public interest, or that the cost of pursuing the public interest is so high that it is not worthwhile.

Cost

More comprehensive models, such as that of Greiner (1998), require a high level of expertise and resourcing. This clearly comes at a cost, although the additional cost once a farm-level model has been developed would not be as large as otherwise.

Conclusion

It is argued here that, far from being unsuitable, individual-farm models can make a number of important contributions to assessment of resource conservation issues. Reasons include:

• the fact that actual decision making occurs at the farm level, not the catchment level;
• some resource conservation problems are primarily private problems, constrained within the bounds of individual farms; and
• predicting individual farmer responses to policies and consequent environmental impacts are key elements of a more comprehensive assessment.

It is not claimed here that individual-farm models can provide all that is needed in the way of economic analysis of resource conservation issues. It is, however, argued that they can play a range of very valuable roles and that they should not be discounted on the basis that they may not be fully comprehensive. Indeed, in some cases, it is possible that once a farm-level model has been developed, the extra benefits of a comprehensive model do not exceed the costs. Nevertheless, when individual farm models are used, it is important to recognise the broader context within which individual farms sit and that this context places limitations on the generality of results from individual-farm models.

References

Anderson, J.R., Dillon, J.L. and Hardaker, J.B. (1977), Agricultural Decision Analysis, Iowa State University Press, Ames.

Cary, J.W. and Wilkinson, R.L. (1997). Perceived profitability and farmers’ conservation behaviour, Journal of Agricultural Economics 48: 13-21.

Greiner (1998). Catchment management for dryland salinity control: Model analysis for the Liverpool plains in New South Wales, Agricultural Systems 56(2): 225-251.

Hayes, G. (1997). An Assessment of the National Dryland Salinity R,D&E Program, LWRRDC Occasional Paper No. 16/97.

Industry Commission, (1997). A Full Repairing Lease: Inquiry into Ecologically Sustainable Land Management, Draft Report, Industry Commission, Canberra.

Kingwell, R.S. and Pannell, D.J. (Eds) (1987). MIDAS, A Bioeconomic Model of a Dryland Farm System, Pudoc, Wageningen, 207pp.

Marsh, S.P. and Pannell, D.J. (1998). Agricultural extension policy in Australia: The good, the bad and the misguided. Paper presented at the 42nd Annual Conference of the Australian Agricultural and Resource Economics Society, University of New England, Armidale, NSW Jan 19-21 1998.

Pannell, D.J. (1995). Economic aspects of legume management and legume research in dryland farming systems of southern Australia, Agricultural Systems 49: 217-236.

Pannell, D.J. (1996). Lessons from a decade of whole-farm modelling in Western Australia. Review of Agricultural Economics 18: 373-383.

Salerian, J.S. (1991). Economic Analysis of Soil Salinity in a Sub-Catchment in Western Australia, PhD thesis, University of Western Australia, Nedlands.

Sinden, J.A. and King, D.A. (1990). Adoption of soil conservation measures in Manilla Shire, New South Wales, Review of Marketing and Agricultural Economics 58: 179-192.

Citation: Pannell, D.J. (1998). On the use of individual-farm models for analysis of resource conservation problems, SEA Working Paper 98/03, Agricultural and Resource Economics, University of Western Australia, http://www.general.uwa.edu.au/u/dpannell/dpap9809f.htm

SEA News issue #2

The SEA News index is at http://welcome.to/seanews

Copyright © David J. Pannell 1998
Last revised: May 21, 2003.