Enhancing the environmental benefits of agroforestry through government policy mechanisms

David J. Pannell

CRC for Plant-Based Management of Dryland Salinity and School of Agricultural and Resource Economics, University of Western Australia, c/- WA Department of Agriculture, 444 Albany Hwy, Albany WA 6330, Australia


Earlier chapters have considered both the economic and environmental benefits and costs of agroforestry. The market provides a ready-made mechanism for capturing the available economic benefits, but without additional interventions from governments, the community may miss out on a range of potential environmental benefits from agroforestry. Reasons for this include that some of the environmental benefits are relatively intangible and not able to be marketed. In addition, resource managers making decisions about agroforestry are unlikely to fully consider the range of environmental benefits accruing to others in the community, since many of those benefits are not experienced or captured by the resource managers.

This chapter focuses on the government’s use of policy mechanisms to ensure that environmental benefits are not neglected. This is not to say that environmental benefits would take precedence over economic benefits, but rather that benefits and costs in both categories would be considered and weighed up. We will discuss the circumstances in which government intervention to enhance environmental outcomes from agroforestry would and would not be appropriate. The mere existence of an environmental benefit is not sufficient, as we will see. A wide variety of policy approaches and mechanisms is available. These are briefly described, and some of their pros and cons are discussed. Finally we consider the vexed question of who should pay for the public environmental benefits generated on private land.


It has long been recognised that a range of environmental benefits can result from the introduction of agroforestry systems into extensive farming systems, including benefits related to dryland salinity, biodiversity, carbon sequestration, renewable energy, and flood mitigation. One of the key motivations for government to develop and promote agroforestry is that it can generate these benefits in addition to financial benefits from the sale of commercial products.

Major government programs for environmental protection and natural resource management in Australia have sought to support agroforestry. Current examples include the Natural Heritage Trust (NHT) and the National Action Plan for Salinity and Water Quality (NAP). They each use a variety of mechanisms to encourage measures, including agroforestry, that would not otherwise be adopted by farmers to the same extent.

This chapter provides a big-picture view on these issues. It outlines ideas from economic theory that provide helpful ways to consider a wide range of questions about the government’s policy decisions on agroforestry, focusing on environmental aspects. The questions include the following.

Although some of the material in the chapter may seem rather conceptual and a long way from the practical reality of growing trees on farms, you should not underestimate its importance in influencing what actually happens on real farms. The theories presented here are very influential high up in government agencies, particularly in departments of finance and treasury, which hold the purse strings for any proposed government policy initiative.

Economic rationale for policy intervention

In the public policy environment that currently dominates in Australia (as well as many other developed counties), government agencies are increasingly required to ask a difficult question about the policies that they would like to put in place. The question is, in effect, why bother? Why would these issues not be adequately dealt with if the government left decisions about them entirely in the hands of individuals and businesses in the community? Perhaps the best strategy for the government is to do nothing, other than provide the indirect support that comes from an effective legal and economic framework. Economists refer to this "do nothing" option as the free-market approach, and they have a wealth of evidence to draw on showing that in many cases it can be a pretty good option. So much so, in fact, that they believe that the burden of proof should rest on those who would have the government do something. They would start from a position that doing nothing is the best option, until proven otherwise.

The appropriateness of the free market depends on two considerations:

  1. does it result in the most efficient outcomes? (or is there "market failure?") and
  2. is the resulting distribution of costs and benefits fair and acceptable? (or who pays?).

Later sections deal with market failure and with the issue of who pays; efficiency and fairness. Try to bear in mind that an attempt to overcome market failure is aimed at ensuring the efficient operation of the economy (maximising total economic welfare) whereas redistributional activities are attempts to ensure equitable outcomes for all groups in the economy (achieving a fair distribution of the wealth).

The causes of market failure

Economic theory indicates that in the absence of a number of clear causes of market failure, intervention by governments in human affairs is likely to reduce economic efficiency.

Further, the mere existence of market failure is not a sufficient justification for government involvement. The failure has to be great enough to offset the direct cost of involvement and the risk of "government failure"; that is, the risk that even with the best of intentions, government will formulate laws or undertake measures that make things worse, rather than better. The cost of government involvement includes the administrative cost of collecting, holding, distributing and spending tax payers' money. For Australia this cost has been estimated as being of the order of $0.40 per dollar collected by governments (Findlay and Jones 1982). In other words, to spend $1.00, the government has to collect $1.66. (This applies generally, not just to environmental programs). For this to be efficient, the government's use of the money would have to be substantially more productive than the taxpayer's – at least 66 percent more.

As well as the issue of whether any Government involvement is justified, there is also the issue of the optimal level of involvement - the level of Government expenditure that maximises benefits to society. This depends on the biological, technical and economic characteristics of each issue and can be very difficult to determine.

There are four types of market failure: public goods, externalities, monopolies and ignorance or uncertainty, as briefly outlined below. All of these are of some relevance to agroforestry. For further details and theoretical background, see Randall (1981), Pearce and Turner (1990) or Tietenberg (1996).

Public Goods

The terms "public goods" and "private goods" in economics have very particular meanings that do not always coincide with popular usage. For example, a resource may be in private hands but still have public good characteristics. A good may be provided to the broad community (the public) but not have public good characteristics in the economic sense.

There are two broad types of public goods: non-rival and non-price excludable (Randall 1981). In each case the economic problem is that in a free market the good is provided at a lower level than would be socially optimal. Thus there is a prima facie case for government stepping in, in one of a variety of ways (which we will examine later) to ensure that the good is provided at an appropriate level.

(i). Non-rival goods

For a non-rival good, consumption by one person does not reduce the quantity or quality available to others. Most economic goods are rival in nature. Consumption by one person of bioenergy or wood products derived from agroforestry means that they are not available to be consumed by other people, unless the original purchaser is willing to give them up. However there is a range of examples where outputs from or inputs to agroforestry are non-rival.

Examples of non-rival goods related to agroforestry

1. Preservation of a threatened species that the community values. Agroforestry might contribute to such an outcome in some cases. The fact that one person benefits from (i.e. "consumes") the knowledge that the species has been conserved does not reduce the benefit attained by others who care about the issue.

2. Information. Widely applicable information about agroforestry production methods produced from research and development may be non-rival. Many producers can make use of the information without reducing its availability or usefulness to others.

3. Aesthetic benefits. The community may value the aesthetic appearance of a landscape that includes more agroforestry. The enjoyment of these aesthetic benefits by any one community member does not reduce their availability to others (provided that everyone does not go to enjoy them in the same place at the same time!)

From the examples you can see that non-rivalry applies to relatively intangible output from agroforestry, not to tangible products that you can hold in your hands.

The reason a non-rival good results in market failure is that there is no cost of providing the non-rival good to an additional consumer (the "marginal cost" equals zero). As a consequence of this, the socially optimal price to charge for the good is zero. The socially optimal price is the price that would maximise overall net benefits to the community. For a rival good, the socially optimal price reflects a balance between the marginal benefits from consumption and the marginal costs of supply. For a non-rival good, at any price above zero there will be consumers who would have consumed and benefited from the good if it were free but who now choose not to do so. Since the benefit they would have experienced would not have cost the community anything, charging a non-zero price for a non-rival good results in an overall loss of benefits to the community.

On the other hand, if government intervenes (e.g. by regulation) to require that private firms must not change a price for access to the non-rival good, we take away at least some of their incentive to supply the good. Some of the good may be provided anyway, as a spin-off benefit from commercial decisions (e.g. landholders practising commercial agroforestry may generate aesthetic benefits for which they are not paid), but the point is that because these wider non-rival benefits are not factored into the production decisions of landholders, the agroforestry may not be established over a wide enough area, or in the best places, from the perspective of the community as a whole.

(ii). Non-price excludable goods

A non-price excludable good is one that consumers cannot be prevented from consuming, leading to problems of "free-riders" and under-provision or over-exploitation. Because consumers have access to the good without constraint, the provider of the good is unable to charge a fee for access.

Non-price excludable goods can often be considered a problem of poorly defined property rights. If a private firm was able to assert and enforce a right to exclude people who did not pay for the good or service, it would be possible for the private sector to charge a price and consequently to efficiently provide the good or service. The assignment and enforcement of property rights also gives the holders of the rights an incentive to preserve the resources on which the rights depend.

Examples of non-price excludable goods related to agroforestry

1. Off-site benefits from water-table control. An agroforestry system may help to lower the watertable and reduce off site impacts from waterlogging and dryland salinity. However the producer may not be able to charge the off-site beneficiaries for these benefits.

2. Carbon sequestration. Agroforestry may make a contribution to prevention of adverse changes to global climate by sequestering carbon. Without government intervention, producers cannot charge the individual beneficiaries for this service. They are widely dispersed, impossible to identify and may not even be born yet.

3. Flood mitigation. In Western Australia, rising water-tables are increasing the flood risk in large areas of the agricultural region of the south-west. By lowering water-tables locally, agroforestry can reduce run-off and thereby benefit downstream towns and infrastructure. It is not possible for the producers to charge the downstream beneficiaries for this service.

Note that a good can be both non-rival and non-price excludable. For example, this is true for all three of the examples of non-rival agroforestry benefits provided earlier.

Note also that being non-price excludable is not necessarily an intrinsic characteristic of the good. A good may be non-price excludable in one time or place but price excludable in another, depending on the laws in place. For example, consumption of water (rainfall, surface water and groundwater) by trees in the Murray-Darling Basin in eastern Australia is currently non-price excludable, but there have been proposals that tree producers should be required to participate in the water market to purchase water flows that would have ended up in the Murray-Darling river system. Such a change would effectively convert the water into a price excludable good, not by physically changing the water, but by changing the rules under which it can be used.

Not all goods that are non-price excludable can be converted to price excludable goods by a change of law. Some of the intangible benefits (aesthetics, conservation) are intrinsically non-price excludable. No change in the legal or administrative system could convert them into a price excludable good.

Non-rival goods are always intrinsically non-rival and cannot be converted into rival goods by a change of law.

Although the problems of non-rival and non-price excludable goods both relate to zero prices, there is an important difference. For non-price excludable goods, it may be desirable but it is not possible for suppliers of a good to charge consumers for access, whereas for non-rival goods, it may be possible but it is not desirable for firms to charge a non-zero fee.

Of the two causes of public goods, non-price excludability is usually the more serious problem. Non-rivalry usually means that a proportion of potential consumers are missing out on modest benefits from access to the good, whereas non-price excludability, if it results in major over-exploitation of a resource, can mean that the entire resource stock is degraded, potentially to the point where it is lost to the whole community.


An externality occurs when an activity undertaken by an individual has side effects on others and the first individual does not take these side effects into consideration. There are two types of externality: negative and positive (also called external costs and external benefits).

The classic textbook example of a negative externality is pollution. Suppose that pollution is generated as a side effect of an economic activity. In a free market, a negative externality such as this pollution is a problem because the level of the activity chosen by the polluter is too great (in the sense that there exists the potential to improve the welfare of both the polluter and the sufferer). If the external costs could be factored into the polluter's business decision, the polluting activity would not be undertaken or would be undertaken at a lower level. In the absence of regulation or some form of government imposed incentive, polluters generate more pollution than is socially desirable because they do not consider the costs it imposes on others.

Examples of negative externalities in agricultural production can include drift of chemical sprays, failure to prevent the spread of insect pests or weeds, dust or sand deposits from wind erosion, or some cases of dryland salinity. In each of these cases a farmer, whether by action or inaction, increases the costs for neighbours or others in the community. The cost does not have to be a financial cost to be considered an externality. Effects on health or mental well-being are also relevant.

A positive externality is also a problem, but this time it is because the level of the activity is too low. For example, if establishing agroforestry on a farm would create off-site benefits (e.g. reduced dryland salinity, habitat for biodiversity, carbon sequestration) for which the farmer is not compensated, it may be that the area of agroforestry is too low from a community wide perspective. If the off-site benefits were factored into production decisions, the area planted may be increased (although not necessarily, as we shall see below).

An externality is only a social problem (in the sense of reducing the aggregate benefits of the community) if it is uncompensated – i.e. if the off-site benefits or costs are not taken into account by the decision maker. For example, in the case of trees using water that would otherwise be available to downstream users, if the use of such water if free and unregulated (i.e. non-price excludable) then the downstream consequences are unlikely to be considered, and it may be that the water is being used for lower-value uses than in could be. If tree growers are charged appropriately for the water they use but, having factored in those additional costs, still choose to use the water for tree production, then the fact that the water is not available to downstream users is not a social problem. It is already being allocated to its highest value use. Randall (1981) points out that it is not externalities per se that are the problem, but externalities that are associated with a non-rival or non-price excludable good. Public good are the more fundamental problem.


The undesirability of monopolies is well known. A monopolist faces a strong incentive to exploit its capacity to artificially restrict supply and increase the price it charges to consumers. The "simple" solution is to regulate to curb or avoid monopolies, and this is clearly the aim of bodies such as the Australian Competition and Consumer Commission, and is partly relevant to policies such as the National Competition Policy. However, the issue is complicated by the potential for "natural" monopolies. A natural monopoly occurs in an industry with very high fixed costs and/or very low costs of provision at the margin (e.g. quarantine). In the case of natural monopolies, regulations to prevent monopoly pricing can be counter-productive, because enforcing low sale prices may mean that the monopoly firm is unable to cover its fixed costs and so will shut down. The most common government responses to this situation are to take on responsibility for provision of the good or to allow the monopolist to charge a price that is higher than their marginal cost.

If agroforestry is used for production of bioenergy, it may become associated with natural monopolies in the form of power suppliers. The fixed costs associated with power generation are so high that there may not be scope for more than one supplier to be viable within each market (e.g. each state). Fixed costs in establishing processing plants for products from agroforestry may also be high. Whether or not this results in natural monopolies depends on the markets into which the products are sold. If they are international markets, competition from overseas suppliers would inhibit the potential for monopoly pricing. New processing firms may struggle to make profits in such cases.

Ignorance, uncertainty

Individuals sometimes behave in a socially undesirable way simply through ignorance. The ignorance may be over damage that they are unwittingly causing to others (e.g. off-site salinity) or over matters that affect them personally (e.g. on-site salinity). Where governments feel that they have better information than the individuals they may choose to provide that information as a public service. Examples include public education programs on health, law, pollution, safety and environmental management.

A potential danger is that government officers may be incorrect in their belief that they have better information than the individuals they are targeting with advice. Agroforestry has been promoted widely in Australia, particularly by programs seeking environmental outcomes (e.g. the National Landcare Program, the Natural Heritage Trust). Uptake to date has been relatively low. It seems likely that the main explanation for this has been the economic performance of the agroforestry systems being promoted. Those responsible for promoting agroforestry for environmental benefits have not adequately considered the private economics of their advice before proceeding with their efforts to promote agroforestry.

The net benefit test for market failure

The fact that there are various public good, externality and information problems associated with agroforestry suggests that there may be a case for government intervening to ensure that benefits to the community are maximised. However, there is an important additional test to consider: are the net benefits of intervening greater than the net costs? There is a risk, for example, that if the government steps in to reduce an external (indirect or public) cost, they may inadvertently cause an even greater internal (direct or private) cost. To avoid this risk, it is important to weigh up all of the relevant costs and benefits, both external and internal, indirect and direct, to check that the proposed change is actually beneficial overall. If it is not beneficial overall, then it is not actually addressing a market failure, despite the existence of public goods, externalities, monopoly or ignorance.

To illustrate, suppose that the government wishes to pay incentives to landholders to convert their land from traditional agricultural production to an agroforestry system. Suppose that the agroforestry system is less profitable than traditional agriculture, and that the farmers are not willing to make the change voluntarily. The income sacrifice (the net private cost) that they would have to bear varies from case to case. We will consider examples where it is high or low. Similarly the net external benefits of making the change to agroforestry may be high or low. Figure 1 shows the four relevant scenarios, combining high and low net private cost (reflected in the shortfall of private benefits below the farmer’s break-even requirement) and high and low net external benefit. For simplicity, it is assumed in the diagram that the farmer’s break-even requirement is the same in all scenarios.


Figure 1. The net benefit test for market failure. Conversion of traditional agriculture to agroforestry would generate positive net benefits in scenario A, but not on the other scenarios. In Scenario A, the unrealised potential for external benefits results in a case of market failure, but in the other scenarios the external benefits are not sufficient to cause market failure.


Scenarios A and B are where agroforestry is somewhat profitable, although not sufficiently so to be more attractive to farmers than their existing farming systems. In scenarios C and D the practices are much less profitable than existing systems. The levels of non-agricultural (external) benefits resulting from the treatments are relatively high in scenarios A and C and low for B and D. In scenario A, the combination of agricultural and non-agricultural benefits is such that it is possible for the proposed payment to farmers to change the land use and to be beneficial overall. The payment could provide sufficient incentive to exceed the farmer’s break-even requirement (mainly determined by the profitability of their existing land use) and prompt a change of management without needing to be so great that it outweighed the resulting external benefits.

In the three other scenarios, either the treatment is not sufficiently profitable at the farm level, or the non-agricultural benefits are too small or both. In these scenarios, a payment to farmers sufficient to prompt a change in land use would need to be greater than the value of the resulting external benefits. It clearly makes no sense to pay more for a good than the good is worth. To do so would result in a net reduction in the total benefits that the community gains from this land, even after the external benefits are factored in.

Ideally, one would wish to evaluate proposals for new government policies using a broad benefit:cost analysis, considering both public and private benefits and costs. Where some of the benefits are relatively intangible, this is particularly difficult, although it may still be possible (see box).

Valuation of intangible environmental benefits from agroforestry

Economists refer to intangible environmental benefits (such as the satisfaction one gets from knowledge that a species has been protected from extinction) as "non-market" benefits, as they cannot be bought or sold in the same way that more tangible products can be.

Several methods for estimating non-market values have been developed and applied by economists. The most well-established techniques include three indirect methods ("travel cost recreation demand", "hedonic property value (or wage) equation", and "averting behaviour or household production model") and two direct methods ("contingent valuation" and "choice modelling") (Smith, 1996; Bennett and Blamey, 2001). Contingent valuation (CV) and choice modelling (CM) are usually the most relevant techniques for measuring non-market values of agroforestry (particularly non-use values). Both are based on social surveys of samples of the population. In CV, people are asked to state their willingness to pay for a hypothetical improvement in environmental quality or their willingness to accept compensation for a hypothetical deterioration in environmental quality. In CM, people are asked to rank hypothetical options that involve trade-offs between environmental and other outcomes. Values attributable to the environmental outcomes are inferred from their responses. In both cases, values are aggregated up to the level of the whole population.

There has been a spirited academic debate about the validity and usefulness of non-market valuation methods, particularly for CV. Arguments put forward by advocates of the techniques have included the following:

(a) When done well, the techniques give plausible and realistic results; and

(b) Even though the techniques are not perfect, it is important to attempt to measure non-market values using the best available methods because it assists in having them fully and properly considered in public planning and policy making.

On the other hand, some economists reject argument (a), particularly for CV. Much of the debate is technical (mostly based on arguments that results from actual CV studies are illogical in a variety of ways). CM avoids some of the problems associated with CV and its advocates make reasonable claims that it is a superior technique. However there are some more general concerns about CV that would also affect CM. See Pannell (2004) for further discussion.

To conduct a meaningful non-market valuation study, we need to be able to answer basic (non-monetary) questions such as:

Sometimes our knowledge of these areas is weak, and research to improve these aspects would be the priority.

Evidence for and against market failure related to agroforestry

It is one thing to identify potential causes of market failure related agroforestry, and quite another to determine that there is market failure in reality. To do so, we need information about private costs and benefits of agroforestry as well as public external benefits and costs.

In the case of dryland salinity, recent evidence is indicating that non-adoption of agroforestry is unlikely to be a general market failure, but that non-adoption in particular locations probably is a market failure. In much of the landscape, the situation is like Scenario B, C or D of Figure 1, but in certain locations, Scenario A probably applies.

This conclusion needs further explanation. Even though the off-farm costs of dryland salinity are obviously high, the off-farm benefits from on-farm treatments are often much smaller than the on-farm costs. Particularly in drier regions, the treatments are often only partly effective at preventing salinity off-site and the positive off-site effects tend to be very long delayed (Bell et al. 2000; George et al. 1999; Hatton and Nulsen 1999; Heaney et al. 2000; National Land and Water Resources Audit 2001; Dawes et al. 2002).

In recent years, we have lost earlier hopes that large-scale preventative impacts on salinity could be achieved by clever selection and placement of relatively small-scale plantings of deep-rooted perennial plants (shrubs, perennial pastures or trees). The new consensus is that large proportions of land in most catchments would need to be revegetated with perennials.

Even with major revegetation efforts, the degree of salinity prevention in the long run will probably be less than we would like, particularly in relatively flat landscapes. Figure 2 shows the results of hydrological modelling for several catchments in Western Australia (George et al. 1999). These results indicate that if recharge across a catchment were reduced by 50 per cent, implying perennials on more than 50 per cent of the land, the predicted eventual area of salinity in the catchment would be reduced by an amount somewhere between 3 and 12 per cent of the catchment.


Figure 2. Responsiveness of dryland salinity to reduced recharge (e.g. from perennials or drainage) in a range of catchment types in Western Australia, assuming that "business as usual" would result in salinisation of 30 percent of the catchment. (source: based on George et al. 1999)


Applying standard discounting methods to convert distant future benefits into present values further reduces the magnitude of the benefits. The significance of this for policy interventions is as follows. The level of off-farm benefits from on-farm treatments sets an upper limit on what it could be worthwhile for the community to provide in financial support to farmers to encourage adoption of agroforestry. Small off-site benefits warrant only small financial support. For similar reasons, they warrant only small financial penalties for non-compliance, when a regulatory or tax-based approach is used.

This prompts the question, would small financial support be sufficient? This depends on the on-farm economics of agroforestry. This varies widely from location to location, but there is general agreement that few of the current agroforestry options are economically attractive to farmers other than in high rainfall regions.

Overall, from a salinity perspective, a market failure argument for government intervention to support agroforestry to address dryland salinity can be made, but it requires particular circumstances.

In the case of terrestrial impacts from salinity, a further condition is that there should be downstream impacts on assets of exceptionally high value. In the case of salinity in water courses, market failure is most likely in locations where salt stores and salt discharge into waterways is outstandingly high.

Note that these problems can be side stepped if it is possible to identify or develop agroforestry systems that are profitable in their own right. In many cases, it may be more beneficial to pursue R&D for development of profitable agroforestry systems than through measures to support adoption of unprofitable systems.

For the other categories of environmental impacts (e.g. biodiversity, carbon sequestration, erosion mitigation), evidence for or against the proposition that lack of adoption of agroforestry reflects market failure is much less clear. It is true that there are ongoing efforts to promote agroforestry ostensibly for achievement of these, but we have little clear evidence on whether or not this is a good use of public funds, in the sense that it addresses a market failure.

Types of policy intervention

There are many different mechanisms that governments can use to encourage farmers to take up agroforestry systems (Pannell 2001). Table 1 shows some of the main categories of policy approach relevant to agroforestry and indicates some key pros and cons of each approach. Space does not permit detailed explanation of these many alternatives. The main point to note is that there are many options available to governments wishing to influence uptake of agroforestry, that their applicability varies widely in different situations, and that some care and effort is needed to ensure that the policy approaches used are the most appropriate options for the case in hand.


Table 1. Pros and cons of different types of policy approaches



Pros and cons

Subsidies and other positive financial incentives

Payments to farmers for adopting agroforestry

Carbon credits

Pros: Politically popular.

Targets specific activities.

Cons: Financial impact on government budgets.

May stimulate too much activity.

Uncertain effects.

Marketable permits

Emissions permits

Pros: Provides clear limits on emissions pollution.

Effective when sources are numerous and damage per unit of pollution varies with the quantity of pollution.

Provides stimulus to technological change.

Cons: Potentially high transaction costs.

Requires variation among the polluters in the cost of pollution abatement.

Pollution charges and taxes

Emission charges for saline discharges

Pros: Useful when damage per unit of pollution varies little with the quantity of pollution.

Provides stimulus to technological change.

Cons: Uncertain environmental effects.

Potentially large distributional effects (winners and losers).

High monitoring costs.

Input or output taxes and charges

Carbon tax

Fertilizer tax

Levy on polluting land uses

Pros: Administratively simple.

Relatively low monitoring costs.

Effective when sources are numerous and damage per unit of pollution varies little with the quantity of pollution.

Cons: Often weakly linked to pollution.

Uncertain environmental effects.

High monitoring costs.

Legal liability

Nuisance, trespass

Duty of care

Natural resource damage assessment

Pros: Can provide strong incentive (provided legal recognition of liability and enforcement are high – probably not the case with salinity)

Cons: Assessment and litigation costs high.

Burden of proof large.

Voluntary programs


Education and communication

Pros: Low cost.

Cons: Uncertain participation (for salinity, participation much too low).

Presumes viable technologies are available.

Support for industry development

Research and development

Provision of physical infrastructure

Reform of administrative barriers to the industry

Pros: Potentially applicable over very large areas.

Can generate environmental benefits in cases where direct subsidies would not be warranted.

Can improve the leverage of public subsidies.

If resulting industries are profitable, few problems encouraging adoption.

Additional social benefits from employment in new industries.

Relatively low draw on public funds.

Cons: Time lags of R&D


In Australia, the National Action Plan for Salinity and Water Quality has supported a pilot scheme investigating a range of innovative policy mechanisms intended to encourage environmental management. One of the innovative approaches draws on the age-old idea of an auction to ensure that public funds achieve the highest possible environmental benefits per dollar (see box).

The BushTender trial of an auction-based policy mechanism

A variation on the theme of ignorance as a cause of market failure is so-called "information asymmetry". Farmers have information about important aspects of the economics of the land use options available to them, but the details of this information are unknown to policy makers. This limits their powers to design and implement efficient policies. Auction-based systems have been developed to directly address this problem. Such an approach was recently trialled in Victoria under the title of BushTender as a means of allocating public funds to protection of remnant vegetation, (Stoneham et al. 2003).

In simple terms, the idea in BushTender is that individual farmers bid for public funds to support environmentally beneficial activities on their farm. They specify the environmental works that they propose to do, and set a level of public support that they would require. The government agency evaluates the bids and select those that give the best environmental outcomes per dollar of public funding.

There is an element of competition between the bidding farmers that helps to ensure that bids offer the best possible value to the public funders. Importantly, the bids received reflect not only the direct financial costs and benefits that the farmers face, but also their personal satisfaction from contributing to environmental protection.

An evaluation of the BushTender trial found that the environmental outcomes per dollar were very substantially increased relative to a more traditional subsidy scheme based on fixed subsidies for particular types of action. There seems potential for wider application of this type of policy mechanism, including potentially to agroforestry.

Who should pay?

Unfortunately, there is nothing in economic theory that helps us to objectively evaluate the relative merits of different decisions about who pays. For such questions, the contribution of economists is limited to:

A commonly cited rule-of-thumb for distributional questions is the "user-pays principle" or the "beneficiary-pays principle", under which the beneficiary of a good or service should bear the costs of its provision. It is not a principle in the sense of a scientific principle, but rather a suggestion of what is fair.

Another commonly cited system for distributing costs is the "polluter-pays principle". Generally, this approach is in direct conflict with the user-pays approach. It is similarly lacking in any basis in economic theory, but may be considered a fair and reasonable norm by the community.

There are problems in trying to rigorously implement either rule. For many environmental issues, it is difficult to accurately identify and quantify the benefits and costs for either the polluters or the beneficiaries of environmental works. The user-pays approach would dictate that members of the community should pay in proportion to their benefits from establishment of agroforestry. The polluter-pays approach requires costs to be borne in proportion to the damage caused if agroforestry is not established. Meeting the information requirements of either rule is impossible in any practical sense, although approximations may be feasible.

Application of any simple rule may be compromised, as governments’ decisions about the distribution of benefits and costs are influenced by a range of considerations. These may include political gain, parochialism, the activities of lobby groups or a wish to benefit particular groups due to perceptions that they are disadvantaged in some way.

The market will also have an influence on the distribution of benefits and costs, irrespective of the government’s wishes. For example, if farmers’ production costs go up due to legal requirements to establish agroforestry, the farmers may or may not be able to pass on the increase to consumers of their products. It depends on how responsive consumers are to price changes. If consumers of their products are too responsive and dramatically cut their consumption as prices rise, farmers lose more than they gain by attempting to pass on the extra costs. In a free market, the distribution of costs between farmers and consumers is completely outside government control as it depends entirely on the responsiveness of supply and demand to price changes, and these depend on producers’ cost structures and consumers’ preferences, not on government policy.

In summary, economics offers less help with the question of who should pay than is often suggested. Sometimes there are efficiency dimensions to the question, and economics is certainly useful in addressing these, but usually more important to the community are questions of rights and fairness. These are somewhat flexible over time, driven by community attitudes, politics and power.

Should we have an environmental levy?

One aspect that governments can control is the level of public funds allocated to promote and support agroforestry. Public funding for environmental and resource management has increased over time, but calls by environmentalists for dramatic further increases are common.

The possibility of introducing an environmental levy, either on the price of food or on income tax, is increasingly being advocated. For example, the Prime Minister’s Science, Engineering and Innovation Council (2002) identifies as one of its four priorities for investment, "Redressing the absence of economic signals, to urban and rural Australians alike, connecting the underlying ecological condition of natural systems to our use of them for products and services" (p. 14). The "Wentworth Group of Concerned Scientists" suggests a one percent levy on income tax. While laudable in intent, there are a number of problems with these proposals:


"The government ought to do something about it." This chapter should have given you some insight that, for a responsible government, issues are not as simple as this makes it sound. A responsible government needs to be concerned about whether a particular policy proposal is warranted. This requires an understanding of the concept of market failure, careful consideration of the public and private benefits and costs, and consideration of the distributional consequences. Environmental benefits from agroforestry are clearly significant in some cases, but the government needs to consider more than just this in examining its policy options. Where policy actions are deemed to be warranted, there is a daunting array of policy mechanisms available, with a variety of strengths and weaknesses. The choice of policy mechanism also needs careful consideration.

In the specific case of salinity-related benefits from agroforestry, we have seen that, in many cases, the private net costs of existing agroforestry option exceed the public net benefits, in which case direct policy intervention is probably not warranted. For these situations, the best agroforestry-related use of public funds is probably to support efforts to improve the profitability of agroforestry, such as through research and development into new plants, new products and new systems. However, there may be cases where direct measures to promote agroforestry are justified on the basis of off-site salinity benefits. These would generally be cases where groundwater systems are most responsive to changes in land management, where changing to agroforestry is of low cost to farmers, where the land in question is physically close to the discharge site affecting an asset, and where the asset is of high public value.


Bennett, J. and Blamey, R., 2001. The Choice Modelling Approach to Environmental Valuation. Edward Elgar, Cheltenham.

Dawes, W.R.; Gilfedder, M.; Stauffacher, M.; Coram, J.; Hajkowicz, S.; Walker, G.R.; and Young M., 2002. Assessing the viability of recharge reduction for dryland salinity control: Wanilla, Eyre Peninsula, Australian Journal of Soil Research 40: 1407-1424.

Findlay, C.C. and Jones, R.L. (1982). The marginal cost of Australian income taxation. Economic Record 58, 253-66.

George, R., Clarke, C., Hatton, T., Reggiani, P., Herbert, A., Ruprecht, J., Bowman, S. and Keighery, G. 1999b, ‘The effect of recharge management on the extent of dryland salinity, flood risk and biodiversity in Western Australia. Preliminary computer modelling, assessment and financial analysis’, Unpublished report to State Salinity Council of Western Australia.

Hatton, T.J. and Nulsen, R.A. 1999, ‘Towards achieving functional ecosystem mimicry with respect to water cycling in southern Australian agriculture’, Agroforestry Systems, vol. 45, pp. 203-14.

Heaney, A., Beare, S. and Bell, R. 2000, Targeting reforestation for salinity management, Australian Commodities, vol. 7, pp. 511-518.

National Land and Water Resources Audit 2001, Australian Dryland Salinity Assessment 2000, National Land and Water Resources Audit, Canberra.

OECD, 1997. Environmental Taxes and Green Tax Reform, OECD, Paris.

Pannell, D.J., 2001. Economic Tools to Tackle Dryland Salinity in Western Australia. Economic Research Paper, Department of Treasury and Finance, Perth. 38 pp.

Pannell, D.J. (2004). Heathens in the chapel? Application of economics to biodiversity, Pacific Conservation Biology (submitted).

Pearce, D. W. and Turner, R.K. (1990). Economics of natural resources and the Environment, Harvester Wheatsheaf, New York.

Prime Minister’s Science, Engineering and Innovation Council 1999, Dryland Salinity and its Impact on Rural Industries and the Landscape, Prime Minister’s Science, Engineering and Innovation Council, Occasional Paper Number 1, Department of Industry, Science and Resources, Canberra.

Randall, A. (1981). Resource Economics, An Economic Approach to Natural Resource and Environmental Policy, Wiley, New York.

Smith, V.K., 1996. Estimating Economic Values for Nature: Methods for Non-Market Valuation. Edward Elgar, Cheltenham.

Stoneham, G. Chaudhri, V., Ha, A., and Strappazzon, L. 2003. Auctions for conservation contracts: and empirical examination of Victoria’s BushTender trial, Australian Journal of Agricultural and Resource Economics 47(4), pp. 477-500.

Tietenberg, T. (1996). Environmental and Natural Resource Economics, 4th ed, HarperCollins, New York.


Citation: Pannell, D.J. (2004). Enhancing the environmental benefits of agroforestry through government policy mechanisms. In: I. Nuberg (ed.) Agroforestry for Natural Resource Management, (forthcoming).  http://www.general.uwa.edu.au/u/dpannell/dp0404.htm

David Pannell home page

Copyright © 2004 David J. Pannell
Last revised: June 16, 2013.