Environmental Policy Bonds: Injecting market incentives into the achievement of society’s environmental goals


This paper was presented to delegates of the Organisation of Economic Cooperation and Development (OECD) at a meeting in Paris in April 2002 of the Joint Working Party on Agriculture and the Environment.


The impacts of agriculture on the environment are of major public concern. This paper introduces a new financial instrument, intended to channel market forces into the achievement of environmental goals. Agriculture, of course, is not the only source of environmental degradation. In our ever more populated and complex world solutions to environmental problems are needed that transcend sectors, and even nation states. Environmental Policy Bonds, unlike many other policy instruments, would not prejudge the origins of environmentally damaging behaviour. Instead, they would focus on desirable outcomes.


In economic theory, and on the evidence of recent history, it seems clear that a market economy is essential for prosperity. Markets are the most efficient means yet discovered of allocating society’s limited resources, but many believe that market forces inevitably accentuate extremes of wealth and poverty. So it is important to remind ourselves that a market economy is consistent with many different outcomes and that market forces can serve public, as well as private, goals. Environmental Policy Bonds would be an entirely new financial instrument, intended to channel the market’s incentives and efficiencies into the achievement of society’s environmental objectives.


The first chapter of this paper outlines the Environmental Policy Bond concept using the example of nitrate pollution. Chapter 2 compares Environmental Policy Bonds with other more-market approaches to policy.  Chapter 3 discusses characteristics of policy areas in which Environmental Policy Bonds may have advantages, and looks in detail at how they could be used to address climate change. Chapter 4 discusses some of the practical aspects of the bonds, including potential problems and how these might be mitigated. The fifth and final chapter briefly discusses when the bonds might best be deployed, then looks in more detail at the use of the concept to stabilise the world’s climate. It closes by setting the scene for further discussion and trials of the bond concept.

Chapter 1: Environmental Policy Bonds

Outline of the concept


Environmental Policy Bonds are a new, and as yet unused, financial instrument that could be deployed to help inject market incentives into the achievement of environmental goals.

Central or local government would auction Environmental Policy Bonds on the open market. Government would undertake to redeem the Bonds for a fixed sum only when a specified environmental objective had been achieved.


Normal bonds are redeemable for a fixed sum, at a fixed date, and they yield a fixed rate of interest. Environmental Policy Bonds would be entirely different. They would have an uncertain redemption date, which, in combination with a fixed redemption value, implies an uncertain yield. Unlike conventional bonds, Environmental Policy Bonds would not bear interest: holders gain by ensuring that the targeted objective is achieved quickly.[*] As with conventional bonds, Environmental Policy Bonds would be issued by government and redeemed by government. But they would differ from conventional bonds in that all returns to their holders would take the form of capital appreciation.


Once issued, Environmental Policy Bonds would be freely tradable on the open market. Owning such Bonds, whose value would rise when an environmental objective became closer, would give bondholders an incentive to try to achieve that objective. Under a Bond regime bondholders would have every motivation to choose and invent ways of solving environmental problems that are cost-effective and efficient. Environmental Policy Bonds would have the effect of contracting out the achievement of environmental objectives to the private sector, while government would continue to set these objectives, and to be the ultimate source of finance for their achievement.


An example can elaborate the principles involved.

Example 1: Water quality


The OECD has carried out a stocktaking of the environmental performance of agriculture in its member countries (OECD, 2001). While in recent years there have been some positive developments, there have also been cases of environmental degradation. This has been a result partly of the intensification of farm production in some areas, and the regional concentration of activities such as livestock farming. These have generated higher levels of nutrient surpluses, ammonia and greenhouse gas emissions, with consequent increases in water and air pollution. There is also growing competition for scarce water resources both between agriculture and other users, and in meeting the water needs of aquatic ecosystems for recreational and environmental purposes.


One of the indicators the OECD uses to measure the environmental impact of agriculture on water quality is the water quality state indicator. This can be defined as “the proportion of surface water and groundwater above a national threshold value of nitrate concentration (NO3 mg/l)” at various measuring points (OECD 2001). Agriculture accounts for more than 40 per cent of nitrogen emissions, and many OECD countries are trying to reduce the concentration of nitrates in their water. Let us assume that a particular government wishes to reduce the existing level of nitrates, calculated as an average of the level in its rivers and groundwater, from 50 mg/l to 40 mg/l.


The government decides to target this level by issuing Environmental Policy Bonds that become redeemable for, say $10 only when the average nitrate water concentration has fallen to 40 mg/l as certified by an independent scientific body. It calculates that achievement of this target level is worth a maximum of, say, $10 million. By issuing 1 000 000 Bonds it therefore ensures that the net cost of achieving the objective will not exceed this figure. These Bonds would be floated at an auction. Those who bid the highest price for the limited number of Bonds will be successful in buying them.


What will determine the float price of the Bonds? Most obviously the market's assessment of how likely, and when, the targeted level of nitrate concentration will be reached. People will differ in their valuation of the Bonds, and their views will change as events occur that make achievement of the targeted objective a more or less distant prospect. The Bonds might sell for as little as a fraction of a cent if people thought there were virtually no chance of the nitrate concentration reaching the target level in their lifetime. However, let us assume that the Bonds fetch an average price of $5. The Bondholders now hold an asset that can appreciate by 100 per cent once nitrate pollution levels have fallen to the target level. They therefore have a powerful incentive to do what they can to help bring about such a reduction. The Bonds, once issued would be freely tradable, and their price publicly quoted just like those of ordinary bonds or shares.


Who would buy the Bonds, and what would they do?  Some people might buy the Bonds as they would a lottery ticket, or a publicly quoted company share. They would think that the Bonds’ value will rise, even if they do nothing. Such passive investors would want to become ‘free-riders’ hoping to benefit from any increase in the Bond price without actually participating in any pollution reduction activities. But the way the market works would limit the opportunities for such speculation. The more Bonds they collectively own, the more remote the targeted pollution will become, and so the more they stand to lose as the aggregate value of their Bond holdings fall. At some point, then, it would become worthwhile for these passive investors either to become, or to sell their Bonds to, active investors. (Chapter 4 will look at this, and other potential pitfalls arising from perverse incentives, in more detail.)


Active investors would finance initiatives aimed at reducing nitrate pollution. They could use their own capital, or borrow on the strength of the redemption value of their Bonds, or on the strength of any increase in the value of their Bonds, to support projects that help reduce the level of nitrates in water. They would have every incentive to co-operate with each other, to help achieve the targeted objective, and to do so as cost-effectively as possible. Their motivation arises from the expected capital gain they will experience as the value of their Bonds rises in line with the enhanced probability of the objective being achieved quickly.


Such active investors in the Bonds are likely to undertake a range of initiatives, including:


·         helping to finance farms’ water pollution reduction programmes;


·         paying farmers to reduce their pollution-generating output,


·         defraying the costs of production, or consumption, of less polluting alternatives to existing products or services; and


·         contributing to research into ways in which the same agricultural output can be produced with less pollution.


Bondholders can also be expected to discover and finance other pollution-reducing initiatives, the precise nature of which need not be known in advance.


Note that large polluters might buy Bonds themselves. They would do so knowing that their holdings would rise in value and help to offset the cost of any measures – whether compulsory or voluntary – that they take to reduce their pollution.


In some countries, central, regional or local government bodies might already be involved in some way with some of these initiatives. However, there is a crucial difference: under an Environmental Policy Bond regime participants (bondholders) have the incentive to seek out those ways of reducing pollution that will give them the best return on their outlay. Without government planning, therefore, bondholders finance the most cost-effective ways of reducing pollution.

Trading the bonds


Environmental Policy Bonds, once floated, must be readily tradable at any time until redemption. The operation of such a ‘secondary market’ is critical to the way the Bonds work. Many Bond purchasers will want, or need, to sell their Bonds before redemption—which may be a long time in the future. With a secondary market, these holders will be able to realise any capital appreciation experienced by their holdings of Environmental Policy Bonds, whenever they choose to do so. Tradability would make the Bonds a more attractive investment in the first place.


As the Bonds are traded, they would tend to flow towards those who are most able to help solve the targeted problem. In fact, though, it is not necessary for there to be any actual flow of Bonds. Large bondholders might simply decide to subcontract out the required work to many different agents, while they themselves hold the Bonds from issue to redemption. The important point is that the Bond mechanism ensures that the people who allocate the finance have an incentive to do so efficiently and to reward successful outcomes, rather than merely to pay people for undertaking an activity. At the limit we can conceive of just one single buyer of all the Bonds. If this buyer were determined to hold on to the Bonds until redemption, then the Bonds would function as a sort of performance-related contract, with the government paying only when the objective has been achieved. The buyer could contract out most, or all, of the work required to achieve the objective, with the incentives given by the Bonds for speedy accomplishment cascading down from the bondholder to those subcontracted to do the work.


Too large a number of small bondholders would probably do little to help solve targeted environmental problems by themselves. If there were many small holders, it is likely that the value of their Bonds would fall until there were aggregation of holdings by people or institutions large enough to initiate effective problem-solving projects. As with shares in newly privatised companies the world over, Bonds would mainly end up in the hands of large holders—individuals or institutions. Between them, these large holders would probably account for the majority of Bond holding. Even these bodies might not be big enough, on their own, to achieve much without the co-operation of other bondholders. They might also resist initiating projects until they were assured that other holders would not be free riders. So there would be a powerful incentive for all bondholders to co-operate with each other to help solve the targeted problem. They share the same interest in seeing targeted objectives achieved quickly. So they would share information, trade Bonds with each other and collaborate on objective-achieving projects. They would also set up payment systems to ensure that people, bondholders or not, were mobilised to help achieve targeted objectives in ways that stimulate success. Bondholders would either trade Bonds, or make incentive payments to ensure that any proceeds from higher Bond prices, or from redemption, would be channelled in ways most likely to stimulate speedy achievement of the targeted objective. Large bondholders, in co-operation with each other, would be able to set up such systems cost-effectively.


Regardless of who actually owns the Bonds, aggregation of holdings, and the co-operation of large bondholders, would ensure that those who help achieve targeted goals are rewarded in ways

that maximise society’s environmental benefit per unit outlay.

What happens once an objective has been achieved, and the Bonds redeemed?


Once an objective is close to achievement, the issuing body can float a new set of Environmental Policy Bonds aimed at maintaining the achieved outcome, or at further improvements. Sustaining the outcome beyond the period specified in the original Bond issue is likely to be cheaper than achieving it, while further improvements targeted by a second Bond issue are likely to cost less, in terms of benefit per unit outlay, than those achieved by the first issue. Assume again that a Bond issue aims to reduce nitrate levels from 50 to 40 mg/l, and to sustain this lower level for, say three years. It is very likely that sustaining this lower level for an extended three-year period would cost less than whatever was the net cost of the initial Bond issue. There are three main reasons for this, the first two of which are linked:


1.      Bondholders may have invested in systems or capital assets that cost less, per unit benefit, to keep running than they did to set up.


2.      Bondholders, in a similar fashion, would have learned from their experience of achieving the objective targeted by the first Bond issue. They would have looked for, and experimented with, different methods of solving the targeted problem, and be able to choose the most efficient ones for subsequent Bond issues. If water quality were targeted, is likely that any know-how about monitoring systems or equipment installation would be more cheaply available once an initial targeted lower level had already been achieved.


3.      Less specifically, it is probable that general improvements in productivity, mainly arising from technology (including information technology), will continue to occur in our economies, and that bondholders will continue to adopt and adapt them.


New issues of Environmental Policy Bonds might not be the most cost-effective way of maintaining the achieved outcome. Instead alternative government actions, such as legislation or institutional monitoring, could be preferable.



The main advantage of Environmental Policy Bonds is that, because they would inject self-interest into all stages necessary for solving environmental problems, they would be more cost-effective than current, activity-based programmes. And because they target outcomes rather than activities or institutions, they do not embody fixed assumptions about the best ways of achieving society’s environmental objectives.

Chapter 2: Comparison with other ‘more-market’ approaches


This chapter briefly looks at three ‘more-market’ approaches to environmental policy, in relation to an Environmental Policy Bond regime.

1 Tradable permits to pollute


A tradable permit regime specifies the maximum amount of pollutant that can be discharged. It then issues tradable permits to emit amounts of pollutant making up this total. In the US, markets for permits to emit sulphur dioxide have been in operation for several years. Markets decide the price and allocation of these permits. Tradable permits are especially useful in allocating unpriced resources, such as the assimilative capacity of the environment, and also for targeting pollutants that have marked thresholds.


Tradable permits can work well with intrinsically large-scale processes, or for controlling emissions that have no polluting substitutes. Such processes and substances can be monitored quite easily, because there would be no fear that doing so would lead to offsetting increases in pollution via the setting up of difficult-to-monitor small-scale processes, or the emission of polluting substitutes that are not being monitored. But technological and ecological complexities mean that these processes and substances are a minority. Air pollution, for example, results from many sources and many different processes. Immense quantities of information would be needed to establish, monitor and enforce a comprehensive system of pollution control using tradable permits to pollute. An Environmental Policy Bond regime, however could be more flexible. It could target an index of, say ‘air pollution’, embodying many pollutants, whose weights in the index would differ according to their adverse effects on the environment. In general, it is air or water pollution as a whole, or the adverse effects of such pollution, that need control, not the concentration of single pollutants. Tradable permits to pollute can therefore play only a limited role in environmental protection.

2 Contracting out of existing services


Contracting out usually involves government specifying the outputs it requires, in terms of the nature and level of service required, and inviting the private sector to bid for the contract to supply these outputs. Specifying outputs in this way can be helpful when there is a definite and fixed relationship between a small number of outputs and the desired outcome. But for most environmental objectives such relationships are many and obscure. Contracting out of output-supplying services generally means that the required output must be specified to a high degree. This imposes its own costs, and means that contracting out tends:


·  to be limited to particular stages of outcome-delivery; and


·  to reinforce established ways of doing things.


Of course, government could contract out outcome-supplying services. (In the late 1980s the New Zealand Government did in fact contemplate this: see Schick (1996).) Doing so would stimulate some of the advantages of Environmental Policy Bonds, but not all. An important feature of Environmental Policy Bonds is their tradability. This makes them more attractive to potential purchasers, and enables active investors to benefit from involvement in only one, or a few, of the processes necessary for the targeted objective to be achieved. Tradability differentiates Environmental Policy Bonds from the contracting out of environmental services to private operators in other ways. It




o   transfers risk of breach of contract from the taxpayer to bondholders. If, under a contract system, the successful bidders fail to do what they were legally obliged to do, then it is up to the aggrieved party—the central or local government agency—to take proceedings against them. Even if such actions are successful, they can be protracted and costly. Under an Environmental Policy Bond regime, underperforming bondholders will find a ready market for their Bonds in people who believe they can be more efficient.


Tradable contracts


What if contracts were made tradable, so that the winner of a tendered contract could sell the right to supply an environmental service? This would be helpful if a company, having been successful in bidding for the right to supply this service, has done what it can to achieve a targeted objective. If it has done so efficiently and quickly, the value of the contract rises, and being tradable, can be sold at a profit. The new contractor has the same incentive as the original to perform efficiently.


Such tradable contracts would be similar to Environmental Policy Bonds, as long as the terms of the contract stipulated not that a certain output be supplied, but that a specified outcome be achieved. There would be some small differences though. Ownership of Environmental Policy Bonds would be more fluid, which means more market liquidity, more transparency and an enhanced ability for the government to fine tune its priorities after the outcome has been specified and the Bonds issued. The market prices of the Bonds are extremely helpful in telling the government, and any potential supplier of objective-achieving services: how close the objective is to being achieved, the potential rewards from buying the Bonds and participating in objective-achieving projects, and the likely costs of marginal improvements beyond those already targeted. This is discussed in the next chapter.


Another significant difference is that the Bond concept makes more feasible the targeting of remote objectives; ones that might take years or decades to achieve. Many businesses would be reluctant to take on such goals without the possibility that they could benefit in the shorter term. The Bonds allow them to do what they can to achieve a target, and then benefit from selling their Bonds at a higher price, so letting the new bondholders continue the advance toward the goal.


When governments target remote objectives, it is possible that the winners of a contract to supply a service underestimate their costs, and so fail to deliver. This often takes a long time to become apparent either to government or to other potential suppliers of the service. Under an Environmental Policy Bond regime, however, prices of the relevant Bonds would fall, enabling others to buy them earlier and continue to try to achieve the objective.




Tradable pollution permits can work well with inherently large-scale processes, which can be monitored relatively easily, but suffer from informational disadvantages when there are large numbers of polluters. Similarly, contractual arrangements between government ministers and government agencies, or between government and private suppliers of environmental services, suffer because of the need for government to specify in detail what is required. Another problem is that the most efficient suppliers of environmental services will not always be identifiable at the outset of an environmental improvement programme, nor will their identity remain constant, especially during the lifetime of a long-term programme. Tradable contracts mitigate some of these deficiencies, but the enhanced marketability of Environmental Policy Bonds might supply still more advantages, especially when targeting broad, long-term environmental goals.


Chapter 3: Possible advantages of a Bond regime



Environmental Policy Bonds could have informational advantages over other forms of control when there are many sources of pollution. In this respect, regulation suffers from the same weaknesses as tradable permits: both can work efficiently when dealing with intrinsically large-scale processes. But in agriculture there are typically large numbers of polluters, which can make enforcement of environmental regulations expensive and intrusive. Establishing and monitoring a fully comprehensive water pollution regime, for example, would require immense quantities of information.


The Bonds, focused as they are on outcomes rather than activities, would target society's overall objective – less water pollution – rather than each of the numerous activities that generate water pollution. They would therefore require government to monitor comparatively few data: regular measurements of aggregate pollution, or of indicators or pollution, or of the adverse effects of pollution, need only be taken at selected sites.


Environmental Policy Bonds might have particular advantages too when a multiplicity of solutions is required. The Bonds specify and reward outcomes; they do not prejudge how these outcomes should be achieved. For broad objectives applicable to large areas, it may be preferable to encourage a full range of activities, adapted to local circumstances, than to specify how environmental outcomes are to be met. And, ideally, policies should not discourage research into, and application of, new, more efficient solutions than those that can currently be envisaged by policymakers. Policies that seek to constrain certain prescribed activities, or reward certain prescribed processes, may not be optimally efficient.


A related concern is that the adverse impacts of a particular agricultural activity on the environment will vary with natural circumstances. The composition of groundwater or river water, for example, varies according to natural climatic and geochemical conditions. In consecutive seasons, with the same intensity of production, nitrate levels, for example, may vary widely and inversely with rainfall, approaching dangerous levels when, say, rainfall is very low. Regulation mainly concerns itself with activities, rather than the results of these activities on the environment. When, as happens frequently in agriculture, the effects of production systems on the environment vary with (say) the weather, it may be more efficient to target outcomes rather than systems or activities. Environmental Policy Bonds transfer the risk of failing to meet environmental objectives from society to the polluter.


The same reasoning applies where there are uncertain relationships between processes and their effects on the environment. We may, for example, be concerned about the health of a river, say, as exhibited by the number and variety of fish within it. Perhaps species have been dying, for reasons that are not clear, and not obviously related to identified water pollutant levels. An Environmental Policy Bond issue targeting an increase in the fish population directly may be more efficient than one targeting a reduction in pollutants such as nitrates or phosphates. Bondholders may find it more worthwhile to investigate the causes of the decline of the fish population than to pursue pollution reduction schemes on lines envisaged by policymakers. Or they may decide on a combination of these and other measures. Their research may even lead them to conclude that the decline in the fish population results from a combination of circumstances that will resolve itself without any need for intervention. Whatever bondholders do, they will be motivated to restore the health of the river using the most cost-effective combination of measures possible.


For most environmental problems, not all the alternative solutions are known in advance. The optimal range of approaches is seldom a one-size fits all, government-dictated, inflexible, policy suite. More often, it is a matter for investigation and experimentation, and a wide variety of approaches is essential. Under an Environmental Policy Bond regime, diverse activities would be rewarded in proportion to their success. And once applied they, or an optimal combination of them, would become the new standard for future projects.


Environmental Policy Bonds targeting national goals would encourage investigation of local circumstances, on the basis that it could lead to more efficient solution of the environmental problem affecting a much larger region, or the whole country. In similar fashion they would stimulate diverse responses to a particular problem. Bondholders might find, after a bit of experimenting with different approaches, that certain activities work better than others under certain conditions. They would take the best of these approaches, and apply them where their contribution to achieving the targeted environmental outcome would be greatest, and they would recognise that, for certain objectives, a mosaic of diverse activities is most efficient.


Government policy tends to be uniform. It is often obliged to impose similar policies across whole sectors or large regions, regardless of whether such an approach is cost-effective. It focuses on pollution standards, rather than on the impacts of pollution, which can vary greatly over space and time. All this is not to say that government cannot be both innovative and adaptive in formulating environmental policy. But too often its response is ‘too much, too late’, being absent when longstanding environmental degradation accumulates, and overly prescriptive when, for instance, the same impacts attract media attention. Bondholders, having a financial incentive to do so, could help shape government policy, making it more responsible and effective. They could act as a counterweight to those whose self-interest impels them to lobby against effective environmental policy.



Environmental Policy Bonds would help make environmental policy objectives more transparent. Under an Environmental Policy Bond regime, environmental goals would be explicitly identified, while indirect, as well as direct, means of achieving them would be encouraged – so long as bondholders think them more efficient. Focusing on identifiable outcomes would encourage constructive participation in the political process, and means that measures taken to achieve them would be more likely to attract public support. At least as important, all outcomes would have to be costed. This means that the maximum value that society wishes to place on an outcome would have to be decided and publicly known before any programmes have begun. Once that has been determined, the issuing body would be able to decide on the Bonds’ redemption value and the number of Bonds to be issued. Costing outcomes in this way would make the tradeoffs between environmental outcomes more transparent, and make people's expectations of what government can do for the environment more realistic.


A clear expression of desired environmental outcomes would also mean that progress toward them could be accurately monitored.

Stable goals


Many environmental goals have a necessarily long lead-time. Under an Environmental Policy Bond regime bondholders would aim to achieve environmental goals as cost-effectively as possible. They would be free to choose how to achieve the targeted objective. Environmental goals can be expressed transparently, rather than in terms of technical indicators that embody assumed links between them, and the desired outcome. Society’s environmental goals are likely to be more stable than these links. For example, our ultimate objective in reducing nitrate levels in rivers and groundwater is likely to be one of maintaining or restoring wildlife habitats. Environmental Policy Bonds would allow this outcome to be targeted, rather than each of the many possible influences on such an objective. Of course, it would be unsatisfactory to redeem such Bonds as soon as a targeted improvement in water quality, however specified, had been achieved. The objective would be a sustained improvement in water quality, and this is how it would have to be defined when the Bonds are issued.


Stability of desired outcome makes it unlikely that investors will be deterred from taking measures to achieve them by fears of a reversal of government policy—or, indeed, a change of government. Environmental objectives will have to be carefully defined, so that future governments would be unlikely to repudiate them, even if the associated Environmental Policy Bonds had been issued by a ruling party with a different political outlook. The risk of that happening would be not much greater than that of a government refusing to redeem fixed interest stock issued by any of its predecessors. This risk, always present, is factored into ordinary bond prices, and in no way impedes the operation of bond markets.


Importantly, for the Bonds to be as successful as possible, governments would have to give some assurances as to their future behaviour. For maximum success, therefore, they would also have to choose their objectives in consultation with opposition political parties, as well as with the electorate.

Efficient costing of objectives


Many environmental objectives are difficult to value. Environmental Policy Bonds share with conventional policy instruments the need for some estimate of the value to society of a specified objective. But they have an advantage over most other instruments in that the cost of achieving the targeted outcome is minimised and capped. And if bondholders fail to perform, the cost to the taxpayer is zero. In maximising the efficiency with which the outcome is achieved, the market for the Bonds is elegantly efficient in conveying information about the cost of achieving objectives and, crucially for policymakers, how this cost varies with time and circumstances.


Take our example of lowering the nitrate concentration in groundwater from 50 mg/l to 40 mg/l.

Let us assume that the government issues 1 million ‘Nitrate Reduction Bonds’ (NRBs) redeemable for $10 each once this lower level has been attained. The maximum cost to the government of achieving this objective is then $10 million. But if the Bonds, when issued, fetch $5 each, then the market is saying that it thinks it can achieve this objective for just $5 million. It doesn't say when it thinks it can achieve that objective, but that can be inferred from market behaviour, and the market value of the Bonds compared with other financial indicators. But what if the Bonds sell for virtually nothing, and the market value of the Bonds fails to move from that floor? That would mean that the government had miscalculated: in the market's view there is no realistic chance of the objective being achieved for an outlay of $10 million in the foreseeable future. The government could respond in different ways:


·                     It can wait for new technology to arrive, or for circumstances to change in other ways, such that the market sees the objective as becoming more easily achievable, and the value of the Bonds consequently rises. Or


·                     It can issue more Bonds, with the same specification, also redeemable for $10. It can do this in stages, gauging the market reaction to each new tranche of Bonds, which will tell government the maximum cost of achieving the objective.


Either way, the government can be reasonably sure that it is getting a good deal, expressed as nitrate reduction per unit outlay. This important benefit is worth spelling out in some detail. Valuing the benefit of achieving an environmental outcome is bound to be an uncertain, and to some extent, subjective task, whichever policy instrument is used. But minimising the cost of whatever outcome is targeted is a different matter. Under an Environmental Policy Bond regime, it is the collective wisdom of those in the market for Bonds that determines how much the government (that is, taxpayers) will pay to achieve the targeted outcome. An issuing government can determine the maximum cost of achieving the objective because that would simply be the total number of Bonds issued multiplied by the redemption value plus administration costs minus any revenues gained on floating the Bonds.


But while it has to decide on the maximum cost to society of achieving the objective, a government wouldn't have to work out how much the actual cost will be with any accuracy. Potential bondholders would do that for the Bonds in the open market. Assume again that a government issues 1 million Nitrate Reduction Bonds of redemption value $10. If the market decides that the issue value of these Bonds is $1 each, the net cost of achieving the targeted objective (neglecting administration costs) would be $9 million. In other words, the market at the time of issue believes that the cost, including a profit margin, of achieving the objective will be $9 million.


But suppose the government is totally in the dark about how much it will cost to achieve the targeted reduction in nitrates, and instead of issuing one million Bonds, it issues 10 million with the same redemption value, $10. It would then be liable for a maximum cost of $100 million. However, the market will still reckon that it could achieve the objective for $9 million. So instead of valuing the Bonds at $1.00 it would bid up the issue price of the Bonds to $9.10. (Environmental Policy Bonds would be an unusual financial instrument, in that the more that are issued, the higher would be their market value!)


The Environmental Policy Bond mechanism ensures that people other than government employees decide roughly how much a targeted objective will cost to achieve. Potential investors do this when they bid for the Bonds at issue, and afterwards. This fact, and the would-be bondholders' incentive to minimise their costs, contrast with the current system, in which the costs of achieving particular outcomes, if they are calculated at all, are not widely known, nor subject to competitive bidding. Under the current system, in fact, many of the people employed or contracted by government to supply social or environmental services have every incentive to inflate their projected costs.


Note that the government could add to the number of Bonds in circulation after floating, at any time, if it wanted to boost the efforts going into achieving a particular environmental goal. If it wanted, for whatever reason, to reduce such efforts, the situation is a little more complicated. The government could buy Bonds back from holders, but doing so would reduce the total funds to be spent on achieving the targeted objective, and so would lower the value of all Bonds in circulation. People might therefore be unwilling to buy Bonds in the first place, if they thought there were a high probability of the government’s buying some of them back in this way. They would demand some sort of premium for taking that risk. Or the government could undertake either that it would never buy Environmental Policy Bonds back or, that if it did, it would buy them all back at the market value ruling before it announced its purchase intentions. This unlikely sequence of events should not obscure the fact that under a Bond regime competitive bidding would work to minimise the cost to taxpayers of achieving environmental goals.


But the Bond mechanism does not merely minimise the total cost of achieving a specified objective. It also indicates the marginal cost of achieving further improvements. Say the one million NRBs initially issued do actually sell for $5. This tells the government that the present value of the expected maximum cost of reducing groundwater nitrogen concentration from 50 mg/l by 10 mg/l is $5 million. The government may therefore judge that it is well worth being more ambitious, and aim for a further fall in nitrate pollution of 10 mg/l to 30 mg/l. It could issue a million additional NRBs redeemable when this new lower concentration is reached. These would (probably) have an initial market value of less than $5, reflecting the (probably) diminishing returns involved in lowering the nitrate concentration. The point is that, by letting the market do the pricing of the Bonds, the government is getting an informed view of the marginal cost of its objectives. So if the NRBs targeting the new level of 30 mg/l sell for $4, then the maximum cost of achieving that objective is $11 million, being equal to: $5 million (paid out when the level fell from 50 to 40 mg/l) plus $6 million (paid out when the level fell from 40 to 30 mg/l). The marginal cost of a 10 mg/l drop in the nitrate concentration is revealed to have risen from $5 million to $6 million. Should the government aim for a further fall to 20 mg/l? Under a NRB regime it would have robust information about the cost of doing so.


Of course, this is a simplified example. In fact, the Bond market will continuously update this sort of information. Say that improvements in technology, of the sort that might be stimulated by an initial nitrate-targeting Bond issue, mean that it becomes much cheaper to reduce nitrate levels. Bondholders may, for example, have financed successful research into new varieties of grasses that exhibit better fertiliser uptake. How would the market react to such a development? Once the new varieties’ effectiveness was revealed, the value of all the nitrate-targeting Bonds would rise. Instead of being priced at $5 and $4, the two NRB issues of our example might sell for $8 and $7. The total cost to the government of redeeming these Bonds will not change: it will remain $11 million (though redemption will occur earlier). But the market is providing new information as to the likely cost of future improvements in water quality. The market now expects reductions of 10 mg/l in nitrate concentration to cost $2 million (from 50 to 40 mg/l), and $3 million (from 40 to 30 mg/l). The new grass varieties have reduced the costs from $5 million and $6 million (respectively). So the cost of any further nitrate reductions will also fall, and by following market price movements we can gauge approximately by how much. 


These figures are made-up and simplified, but they do indicate the role that an Environmental Policy Bond market could play in helping the government, and taxpayers, decide on their spending priorities. The importance of this sort of market information can hardly be exaggerated. The failure in history of central planning can plausibly be attributed to the absence of market-generated information (Hayek, 1978). Market prices reflect all of the information used by all who transact, or choose not to transact, in the market. Central planning fails in comparison with a market economy because it encounters the limits of human beings’ calculating capacity: no individual or group of individual planners knows or feasibly can know all the dispersed information that is embodied in prices. Even with a sound incentive system in place—and the centrally planned economies had some fearsome systems—without the information that only markets can generate the computational task of organising an efficient allocation of resources is too great. Prices incorporate and simplify all of the dispersed information implicit in getting a product or service to the marketplace. Markets for Environmental Policy Bonds would continually generate and reveal this information to policymakers and all those involved in achieving social and environmental outcomes—probably for the first time on a systematic basis. In solving society’s environmental problems this combination could prove potent.  

More attractive money flows


A further advantage of Environmental Policy Bonds is that, in many cases, they will have more politically appealing money flows. Many current methods of pollution control inflict identifiable losses on certain people in pursuit of vague objectives. Environmental Policy Bonds, however, would reward people for achieving successful outcomes. The Bonds would of course be redeemed by funds from the government's general taxation revenue and taxes would still have to be levied to provide this, but there is, nevertheless, a presentational advantage.


The other money flow advantage of Environmental Policy Bonds is that the government would incur expenditure only when definite outcomes had actually been achieved. For this reason, the Bonds may attract greater political support for certain goals than agency- or activity- based programmes.


Chapter 4: Potential problems and practicalities


Environmental Policy Bonds would represent a radical change in the way in which our society does things. At first sight, a Bond regime might seem outlandish: it appears to mean that government hands over responsibility for achieving environmental goals to the private sector. It also allows private companies to profit from work done to achieve public goals. But it is important to realise that there is no necessary conflict between public and private goals. Under an Environmental Policy Bond regime government would merely contract out the achievement of environmental objectives. Government would still set these goals and, by undertaking to redeem the Bonds, would still be the ultimate source of finance for the projects that achieve them. Moreover, competitive bidding for Bonds would bid away excessive company profits. Nevertheless, the concept does raise some important questions. One such is whether a Bond regime would generate perverse financial incentives.


Perverse incentives


What if, following a Bond issue, polluters spurn bondholders’ blandishments and refuse to reduce their pollution? Bond prices might therefore fall, and polluters could collude to buy them at a lower price. They would then profit by reducing pollution and redeeming the Bonds. If a pattern of such behaviour were established, would not polluters then be the only investors in future issues of Environmental Policy Bonds? A quick answer would point out that the targeted objective would still have been achieved for a sum equal to, or less than, the maximum cost for which the issuers have allowed. True, the cost would be lower if there were no such collusion. But another answer is that Bonds are only one weapon in a government’s armoury. Regulation of pollution, or the threat of it, would work to raise the market value of the Bonds, and make such behaviour risky. This type of behaviour would be a threat only when there were a few big polluters who could collude. In such circumstances Environmental Policy Bonds might anyway not be the best pollution control mechanism, because their informational advantages over tradable permits for example (see Chapter 2) might not be so significant.


But what if, following a Bond issue, businesses pollute more than they otherwise would, and gain from bondholders paying them to stop? In effect they would pollute more on the expectation they will receive enhanced payment for reducing pollution in the future. This would also, however, have some risk attached, because bondholders may calculate that the most cost-effective reductions can be achieved by businesses other than such anti-social polluters, or that the objective can be achieved in other ways (such as, for example, removing air pollutants from the atmosphere). If pollution were a bye-product of production, then the output of these polluters would be at an above-optimal quantity, so their pollution increase would not be costless.


But the possibility still arises that these businesses could profit from such behaviour. Or even that people who previously generated no pollution whatsoever might begin to pollute so that they could benefit either from bondholders’ paying them to pollute less, or from buying the Bonds cheaply, and then reducing their pollution and selling their Bonds at a higher price. In all these cases there need be no collusion amongst bondholders. For ‘market fundamentalists’ contemplating using Environmental Policy Bonds as the sole means of achieving environmental goals, this might constitute a fatal flaw. But, again, Environmental Policy Bonds would almost certainly complement a government’s regulatory powers, including its powers to apply the Polluter Pays Principle. There is probably enough existing legal, as well as moral, sanction against cynical polluters to ensure that it need not happen. Governments would certainly retain its powers to tax or regulate in ways that would make perverse increases in pollution more risky, or criminal. And it bears repeating that, in a Bond regime, holders have powerful incentives to see that existing rules against pollution are enforced, or that new regulations on polluters are imposed.


Another possible source of perverse incentives could arise from the development of futures and options markets in Environmental Policy Bonds. These would enable people to benefit from a falling Bond price, so giving them an incentive to delay achievement of the targeted goal.


It is quite likely that there would be futures and options markets for the Bonds, and it is almost certain that the price of any particular Environmental Policy Bond would not always be increasing along an upward trend from its float price to its redemption value. It is right too that bondholders should be able to hedge against the consequent falls in the value of their asset, and it follows that people who do not hold Bonds will also participate in markets for derivatives of Bonds, some of which would rise in value as the targeted goal became more remote. This in turn means that speculators and short sellers could certainly profit by short-term Bond price falls, and the question is whether these people would then take steps to reduce climate stability. There are two main reasons why they would not. The first is that, in the long term, the weight of money would be against them. Provided sufficient funds are allocated to achieving the targeted objective, there would be a very large net positive sum of money payable if the climate is stabilised, and a net zero sum being paid if the climate is not stabilised. All the long-term incentive is to achieve the environmental objective. Those who, for whatever reason, would suffer from achievement of this goal could be compensated by bondholders, or bribed to change their ideas. Note also that for every buyer of a ‘put’ option there would be a seller, and that for every futures contract bought on the expectation that the Bond price would fall, there would be an equivalent futures contract sold on that basis, so that the net incentive generated by derivatives would be in line with the incentive created by the underlying financial instrument, the Environmental Policy Bond: in the long run, this would be in favour of achieving the targeted objective.


The other reason that short sellers, or holders of put options, in Environmental Policy Bonds might not take actions aimed at interfering with achievement of the goal is that such actions might well already be illegal or, again given the incentives that the Bonds would generate, be made illegal once the Bonds have been issued.


Might issuing governments themselves try to avoid expenditure by failing to redeem the Bonds, either by reneging on their commitment, or by doing what they can to stop the goal being achieved? If governments issue Bonds they would be doing so as representatives of their citizens. They would therefore be under strong moral pressure to comply with their commitment to supply funds for Bond redemption, and not to take actions impeding progress toward the targeted goal. But it is also in governments’ own interest to fulfil their obligation. If they did not, they would be discrediting the entire Bond principle, which they might well want to deploy again, either domestically or as participants in efforts to solve global social or environmental problems.

The Free Rider question


Could ‘free riders’ undermine the workings of the Environmental Policy Bond mechanism? If purchasers of a significant number of the Bonds hold them with no intention of doing anything to help reduce solve the targeted environmental problem that would undermine the potential of a Bond regime.


There are, though, grounds to believe that free riding would not be a serious problem, mainly because it is unlikely much free riding would occur, and partly because even if it did occur, it would not impede the operation of the Bonds mechanism. Free-riding would be a self-cancelling activity: if most of the Bonds were held by would-be free riders, then little would be done to help achieve the targeted objective. As the objective becomes more remote, the value of the Bonds would fall. And as the Bonds lose value, they would make a more attractive purchase for people who would be prepared to reduce water pollution. So free-riders would be tempted to sell, even at a loss, rather than see the value of their Bonds continue to fall. Some history of falling Bond prices would tend to make free riding on Environmental Policy Bonds less appealing with future issues. There are other reasons why passive bondholding would be unattractive to potential free riders:



·         As with other financial instruments, small players would have to pay higher transaction costs than the bigger institutions – the ones that can initiate objective-achieving projects.



Note also that even if free riders were to gain from holding Environmental Policy Bonds, they would do be doing so only because their Bonds rise in value when the targeted objective becomes closer to being achieved. Attempted free riding would have positive effects: it would add liquidity to the Environmental Policy Bond market.

Incentive to commit illegal acts


Environmental Policy Bonds would work by giving financial incentives for people to achieve particular environmental goals. Realistically, however, they could also encourage people to break the law to do so.


Examples of acts that would be illegal, but that particular Bond issues might encourage, are:


·         emitting pollutants that, while unspecified by Bonds targeting pollution, are still controlled or banned; and


·         falsifying pollution data.


Most such acts are already illegal but before issuing Environmental Policy Bonds, governments should be aware that there would be greater inducements to commit them.


The Bonds could also induce people to modify behaviour in ways that, while not illegal, would undermine what society is trying to achieve. So, for example, if Bonds aiming to reduce the area of intensively farmed land were issued, bondholders might lobby for changes in the definition of what constitutes an intensive farm. Or they might finance projects that shift production from some intensive farms onto others, thereby reducing the actual area intensively farmed, but increasing the pollution generated on that reduced area.


In this case judicious specification of the targeted objective would help: the Bonds could define precisely what shall be deemed an intensive farm for the purpose of the Bond issue. And they include a proviso making the Bonds redeemable on the condition that the maximum pollution level from farming did not exceed a specified level. In general, objectives that are complementary, and that if not pursued jointly are likely to conflict, should be targeted by a single Bond issue. (See box: “What to target?”) 


It is likely though, that even carefully specified objectives would not always eliminate or mitigate the kind of illegal, or negative-but-legal, activities that the Bonds may stimulate. So how can this potential problem be solved? The solutions have to do with the way in which the Bonds are introduced, and with the role of government.


A cautious, gradual, introduction of Environmental Policy Bonds is one means by which adjustment problems can be minimised. If, despite such an approach, bondholders behaved illegally, government could prosecute the perpetrators. If bondholders behaved in negative, but legal ways, government has other options. In ascending order of severity, government could:


·         persuade or cajole bondholders into toeing the line. It could do this publicly or privately - initially, at least, bondholdings could be registered in the same way as shares;


·         buy back as many of the Bonds as is necessary to eliminate the incentive to holders to take any action;


·         legislate against the unforeseen activity; or


·         declare the Bonds null and void, and offer compensation related to the Bonds’ issue price, or their current market price.

What to target?

Breadth of objective


Targeted objectives should, in principle, be as broad as possible. It would probably be unsatisfactory to make nitrates, for example, the sole target of an Environmental Policy Bond issue if it were likely that farmers would respond by increasing the use of phosphates. Instead, both could be made the target of a single Bond issue. Environmental Policy Bonds lend themselves to targeting combinations of objectives. They could, for example, target indices encompassing a range of pollutants, weighted according to their contribution to environmental degradation.

Pollution or effects of pollution?

In principle ends, rather than means to ends, make better targets for Environmental Policy Bonds. Thus it may be preferable to target not pollution, but such indicators of environmental status as wildlife biodiversity, landscape features, perhaps in conjunction with more subjective indicators like the views people have about the quality of the rural environment, as measured by questionnaire responses. Bonds could be issued whose redemption value is on a sliding scale, reflecting the perceived adverse environmental impacts of the targeted range of pollutants. 

Spatial distribution


Bonds aimed at improving national averages of such measures as water quality would be adequate sole policy instruments only if society is unconcerned about the distribution of pollution. Otherwise water pollution targets should be used in conjunction with existing sanctions against polluters. Or Bonds targeting pollution could be made redeemable only on the condition that minimum pollution thresholds shall not be breached anywhere in the country concerned.

Time period


Bond issues could provide bonus payments for achievement of the targeted goal by a specified date. Or issuers could stipulate that Bonds would not be redeemed unless the targeted objective were achieved by a certain date, or that they would be redeemed for a sum that diminishes over the time it takes for the objective to be achieved. The market would factor all such penalties or bonuses into the Bond price.



Effects on government’s behaviour


Another possible problem arising from the integration of Environmental Policy Bonds into the current policy-making system concerns the government's role as creator of statutes. Government can pass laws affecting the Bond price, or its actions could influence the Bond price in other ways. For instance: government could come under great pressure to increase any taxes it imposes on nitrogen fertiliser from holders of Bonds targeting nitrate levels. Note, though, that the source of the pressure, and the motivation for it, would be easy to identify. And lobbying is a legitimate activity. There is no reason why bondholders, in common with other pressure groups, should not lobby politicians. They would be doing so mainly out of financial self-interest of course. But other pressure groups are also self-interested, and in the case of bondholders their self-interest is, or should be, congruent with society's interests. Bondholders will lobby for legislative change, and they will benefit in obvious, financial ways. In assessing the Bonds’ value potential investors will take into account the likely influence of bondholders on legislation, and the potential influence of changes in legislation on the speed at which the targeted objective is achieved.


These influences make it important for there to be some element of consultation when defining targeted objectives. People do become rich by using their influence on politicians under the current system, but they are less identifiable, and they do so in ways that are not always easy to identify. Politicians will always have to weigh the evidence for and against any course of action promoted by lobbyists, with due regard to the lobbyists’ motivation. In the end, though, it is up to potential investors in Environmental Policy Bonds to take into account likely or possible changes in the legislative environment when bidding for the Bonds.


The threat of bondholders lobbying governments for legislative changes can have a positive aspect. For Bond issues to be as successful as possible, governments would ideally give assurances as to their future behaviour. So, for example, prospective purchasers of Bonds targeting water pollution would want to know, for instance, the government’s fertiliser taxation (or subsidy), plans. Government would maximise interest in the Bonds by being as open as possible about its legislative and spending intentions as soon as possible. Government could also undertake not to do such things as eliminate taxes on fertiliser.


Of course, if the Bonds target only small changes in, say, pollution, the government’s long-range plans will not be so significant to prospective bondholders. It might emerge after trials of the Bond concept that targeting incremental improvements in environmental indicators would be the best way of dealing with any uncertainties over future government behaviour. Or it might be that these uncertainties turn out to be relatively insignificant when compared to the uncertainty involved in this, or any other, sort of investment. Markets routinely deal with low levels of risk and uncertainty by attaching lower values to riskier instruments.

Assessment of indicators and insider trading


Governments issuing Environmental Policy Bonds would have to organise reliable and accurate monitoring of the targeted problem so that progress towards the attainment of the environmental objective could be impartially assessed. This monitoring must also be seen to be independent of bondholders, who could benefit unfairly from dubious data collection. Naturally the information as to how close the objective is to being achieved would have value. It is not difficult, for instance, to imagine the latest water pollution figures being sought in advance of official publication and used for ‘insider trading’ purposes. If too much insider trading went on, it would increase the riskiness of the Bonds to those without access to this information and so tarnish their value as an investment. How could it be minimised?


·         In many cases data gathering and collation would have to be more transparent. There would be a ready market for this information, which would help to make these processes more robust. There would be more interest in more frequently updated information, so that progress toward achieving objectives could be more readily charted. If large sums of money were at stake, there would be a great deal of private information gathering: investors, bondholders, and financial commentators will be taking their own soundings during the lifetime of each Bond issue. All this would serve to keep the official information assessors honest.


·         Those involved in gathering, collating and processing information would be bound by contractual terms deterring or forbidding them from abusing privileged information.


·         Indicators for targeted objectives could be chosen with a view to minimising the possibility of insider trading being an important factor. Combinations of indicators could shorten the length of the information chain. The same effect could be achieved by a national government stipulating that Bonds targeting such objectives as water or air pollution would be redeemed on the basis of data from a random sample of sites, rather than all sites or any pre-arranged subset of sites.


·         The objectives themselves could be chosen to minimise the possibility of insider trading. Bonds targeting long range objectives such as cutting nitrate pollution by 50 per cent rather than 20 per cent, would probably be less sensitive to insider trading. With long range objectives, each datum illegally withheld from the Bond market would probably represent a smaller proportion of the total relevant information available to the Bond market, and so have a lesser effect on the Bond’s market value.


None of these ways of mitigating insider trading would always be fully effective. That said, there are already sensitive indicators, such as unemployment or retail sales figures, that are capable of moving markets, and so there are already in place mechanisms to control access of such information until it is time for publication. There are also sanctions against those who obtain, and act on, such information illegally. These mechanisms and sanctions might need to be strengthened under a Bond regime, but it remains to be seen how important abuse of insider information will be. While insider trading means that unscrupulous people benefit at the expense of the public, it does not generally impede the operation of markets. Markets continue to function and the possibility that a low level of insider trading goes on is generally discounted by the broader market.

Government as purchaser of Bonds


Government agencies could, as competitive suppliers of objective-achieving services, participate as active investors in Environmental Policy Bonds, under certain conditions. Unlike in industry the private sector would be unlikely to cry ‘unfair competition’, even if the operations of these agencies were heavily subsidised, because its own Bonds would appreciate as a result of the government, or government-inspired, activity. If government agencies were to participate in the Environmental Policy Bond market, they should not have privileged access to information. Also, it is important that any profits or losses they make as a result should accrue to that agency. The people who work for the agency must have the same incentives as the private sector bodies to perform efficiently. This would change the character of these agencies, and would probably lead to their ultimate divorce from the public sector.

Existing institutions and the transition to an Environmental Policy Bond regime


Few bodies charged with achieving environmental goals are currently paid in ways that correlate directly with their performance. Nevertheless these bodies are the main repositories of expertise for solving environmental problems, and some of them are bound to be efficient, or to be capable of becoming efficient. It would be unwise as well as unfair and unnecessary to divert their funding at short notice into funds allocated to redeem Environmental Policy Bonds. The answer, at least for those goals currently the responsibility of major government-funded institutions, is a gradual transition to an outcome-based, rather than institution- or activity- based funding programme. This would mean that funds from government allocated on an institutional basis would gradually decline, while funds allocated via Environmental Policy Bonds to the outcomes that these (and new) institutions are collectively trying to achieve – a cleaner, safer environment – would gradually rise.


One method would be for government to reduce its funding of environment agencies and research institutes by 1 per cent a year, in real terms. It could allocate the saved funding to the future redemption of the Environmental Policy Bonds it has issued. After five years, each body would be receiving directly from central government only 95 per cent of what it formerly received. But holders of Bonds targeting environmental outcomes may choose to supplement the income of some of these environmental bodies. They might judge some of them to be especially effective at converting the funds they receive into measurable environmental benefits, as defined by their Bonds’ redemption terms. It could be that the most effective environmental bodies, in terms of achievement per unit outlay, are working in especially polluted areas, where small outlays would probably bring about larger improvements in the environment. Or bondholders might judge that a particular research body is worthy of additional funding, because they judge its research most likely to yield pollution control measures, say, that would be most efficient in terms of converting funding to better environmental outcomes. In such cases, bondholders would supplement government funding. It might well be that these favoured bodies end up receiving considerably more than 100 per cent of their former income, throughout the lifetime of a Bond regime.


It could also happen that investors in Bonds targeting environmental outcomes look at completely new ways of achieving these objectives; ways that currently receive no, or very little, funding. To give a not entirely unbelievable example, they might be convinced that one of the best ways of achieving society’s long-term environmental objectives is to increase women’s access to family planning information. Following this logic, they might find that one of the most efficient ways of doing this is to support non-governmental organisations, including registered charities, that already have this objective. It is difficult to imagine how our current activity- or institution- based government fund allocation mechanisms could take this sort of approach.


Could Bonds targeting remote objectives be compatible with a gradual transition of the type described above, where funding to existing government-funded institutions reduces by 1 per cent annually? At first sight there is an apparent mismatch between small reductions in government spending, and the large sums that would be needed to motivate potential investors in Bonds that target remote objectives. The critical point here is that bondholders will be investing not on the basis of the annual reductions in government expenditure on existing environmental bodies, but on the basis of the redemption value of all the Bonds issued. To be more precise, it will be this total redemption value, minus the Bonds’ existing market value, that will drive bondholders’ investment decisions. This sum could be many times each year’s incremental reduction in government’s institution-based spending. One of the virtues of an Environmental Policy Bond regime is that even in the short term bondholders would begin to invest in projects with a long range objective, on the expectation of capital gains that might arise only in the distant future.


The accumulated reductions in spending to existing institutions would be one, but not the only, factor influencing how much government decides to spend on achieving a specified environmental goal. Also important would be the financial savings (if any) that achieving the objective would bring about, and the value society places on any nonfinancial benefits.


It might be possible to expand spending allocated via the Bonds at a faster rate than the 1 per cent suggested: expertise in the environment is still mobile, relative to that in other areas of government spending, which would make it easier quickly to establish new outcome-based institutions or to reorientate existing ones.

Interaction with existing programmes and projects


Note that, while changes in the source of funds would be gradual, those involved in existing institutions might well react by quickly reviewing how all their existing programmes and projects operate. If bondholders see existing programmes as being particularly effective in achieving targeted outcomes they would be inclined to invest in them. On the one hand, the switch in funding would tell existing institutions that they could expect to see their relatively ineffective operations receive diminishing funds in the future. On the other hand, their effective operations could look forward to higher—possibly much higher—funding. Even a gradual transition, involving 1 per cent annual cuts in funds allocated to existing institutions, that was balanced by Environmental Policy Bonds could bring about a rapid change in the way existing bodies conduct all their programmes. They might have to devote some of their resources into persuading bondholders of the cost-effectiveness of their activities; but this would not represent a radical difference from the way these bodies lobby for government funding nowadays. Under a Bond regime, though, they would do their lobbying on a more transparent, outcome-oriented, basis.


It has to be said that some of today’s policies would not long survive the scrutiny of an Environmental Policy Bond regime, with its emphasis on clarity and transparency of outcomes. In agriculture, in particular, it would be difficult to imagine a wide and enthusiastic consensus in favour of the existing array of policies and their unstated, obscure or mutually contradictory objectives. Clarifying agricultural policy objectives could be expected to reduce costs to the consumers and taxpayers who, often unwittingly, pay for the panoply of current policies.

Chapter 5: Environmental Policy Bonds in context


Environmental Policy Bonds are unlikely to be the optimal solution to all environmental problems. This chapter begins by the circumstances under which they are likely to be most beneficial in comparison to other policy instruments. It then looks at one example of a global problem perhaps ideally amenable to a Bond-based solution: climate change. The chapter concludes with a brief discussion of the Bond concept, in the context of the current policy environment.


Key criteria


We should expect a Bond regime to show the most marked improvement over current approaches when:






For some environmental problems these concerns would not apply. The third concern, indeed, would imply that the Bonds would not be especially helpful in achieving local goals, and would be better deployed to solve large-scale problems, perhaps of regional, national or even global dimensions. Many agri-environmental problems are local in nature. Even then, there might be some scope to apply the Bond concept, for example, to clean up a lake fed by rivers polluted by many different farms. However, the most significant advantages of a Bond regime are likely to arise when dealing with larger scale problems. One such is climate change.


Example 2: Climate change


Climate change is an example of a process that might arise from many sources, natural and man-made. It is going to require a full range of policy approaches and it involves many scientific uncertainties. Globally backed Environmental Policy Bonds rewarding climate stability may well have significant advantages over the current suggested solution.


The evidence that the global climate is changing is large and growing. That said, scientists are divided as to (a) how fast climate is changing, (b) the effects of climate change (c) how much man can do about it, and (d) how much man should do about it. And there are still respectable scientists who argue that the climate is not changing at all in any meaningful way. Despite these uncertainties, climate change has the potential to inflict serious harm on many people, so there is a strong argument for doing what we can to minimise its adverse effects.


The December 1997 Kyoto Protocol (‘Kyoto’) saw 159 nations reach the world's first legally binding commitments to reduce the global output of carbon dioxide and five other gases thought to contribute to the ‘greenhouse’ effect. Thirty-eight industrialised countries agreed to reduce emissions by 2012 to an average of 5.2 per cent below their 1990 levels, and in July 2001 180 countries reached a broad political agreement on the operational rules that will govern the Protocol.


The Kyoto targets are far lower than those that some environmentalists had hoped for, and that some countries, most notably the European Union, had been advocating. Kyoto will only slow, but not stop, the build-up of carbon dioxide and other greenhouse gases in the atmosphere. (Carbon dioxide, which is given off by fossil fuel combustion, is thought to be by far the most important of the man-made greenhouse gases that form an insulating blanket around Earth.) But subsequent evaluations by leading scientists indicate that the environmental effects may be so small as to be almost unnoticeable in the near term.


Focusing on outcomes, not activities: Climate Stability Bonds


Kyoto embodies the assumption that controlling the targeted greenhouse gases is the best way of achieving climate stability. But with climate change, the biological and physical relationships involved are many and complex. Even specialists disagree about the degree to which the multitude of biological and physical variables influences climate change. Apart from the daunting uncertainties about the role of greenhouse gases in climate change, there is even less understanding of the role that agriculture and forestry can play as sinks for these gases.


Environmental Policy Bonds targeting climate stability would bypass these, and other, uncertainties, and encourage research into clarifying the relevant scientific relationships. Such Climate Stability Bonds (CSBs) would be issued by governments on the open market and would become redeemable for a fixed sum only when the climate had achieved an agreed and sustained level of stability. In this way there is no need for the targeting mechanism to make assumptions as to how to stabilise the world climate: that would be left to Bondholders.


Ideally, CSBs would be internationally backed. They could be issued by a world body, perhaps one supervised by the United Nations or World Bank. This body would undertake to redeem the Bonds using funds that could perhaps be obtained from all countries, in proportion to their Gross National Product. It would be up to individual countries to decide how to raise funds, presumably from taxation revenue. Importantly though, no Bonds would be redeemed until the objective of a more stable climate has been achieved and sustained. Climate Stability Bonds would be issued by open tender, as at an auction; those who bid the highest price for the limited number of Bonds would be successful in buying them. A fixed number of Bonds would be issued, redeemable for, say, $10, only when climate stability, as certified by objective measurements made by independent scientific bodies, has been achieved and sustained. Once issued, the Bonds will be freely tradable on the free market.


People will differ in their valuation of the Bonds, and their views will change as events occur that make achievement of a stable climate a more or less remote prospect. They would also change as new information about climate, and about the causes of climate change, is discovered.


There are obvious difficulties involved in defining what a stable climate actually is, but the same difficulties apply when attempting to monitor the success or otherwise of Kyoto. Presumably scientists will measure the effects of the cuts by monitoring such objectively verifiable indicators as temperature, change in temperature, rate of change of temperature, precipitation, and many others, at a wide range of locations. 


Under a CSB regime, the Bonds would be redeemed only when climate stability, as defined by such a similar set of indicators, had been achieved. A Bond regime might also explicitly target less scientific measures, such as the frequency and severity of adverse climatic events, the numbers of people killed or made homeless by such events, or the insurance payouts to which they give rise.


What would Bondholders do?


A CSB regime would not dictate how to achieve a stable climate. Bondholders could undertake a wide range of projects including:


·         helping countries or companies to set up and run greenhouse gas emission control programmes;


·         helping countries or companies to set up carbon sequestration plantations;


·         financing the production or consumption of energy that involves lower greenhouse gas emissions; or


·         carrying out, or supporting, research into increasing the reflectiveness of the Earth or its atmosphere.


Bondholders can also be expected to finance other research and initiatives, all aimed at stabilising climate as cost-effectively as possible.


Some governments, research institutes and others are already carrying out these, or similar, activities. But under a Climate Stability Bond regime, the Bondholders will have the incentive to seek out those ways of achieving a stable climate that would give them the best return on what is, in effect, the taxpayers’ outlay. Only when the targeted degree of climate stability is achieved would governments have to pay for it by redeeming the Bonds. Until then, it is bondholders who have to finance the initiatives that they think will achieve climate stability. The body that issues the Bonds will, in effect, contract out the achievement of climate stability to the private sector – having defined the degree of stability it wants and undertaken to pay Bondholders when it has been achieved.


Advantages of Climate Stability Bonds


Climate Stability Bonds have two critical advantages over Kyoto. One is that the Bonds do not rely on the robustness of our existing scientific knowledge even as to whether the climate is changing in the way that many scientists believe it is, let alone as to how best to stabilise it. Kyoto aims to reduce emissions of a small range of gases. But there may be other causes of climate change that are far more important, of which we are currently unaware. And these need not be man-made: natural variability of climate has had severe impacts on human life in the past. Kyoto, responding to effects whose causes are uncertain, embodies a limited number of fixed ideas about the nature of the relationships involved. A CSB regime, targeting climate change directly, might well lead to cuts in greenhouse gas emissions, but it would not assume that doing so is the best solution. CSBs improve on Kyoto, because they encourage behaviour leading to the desired outcome, rather than seeking to control activities whose effects on the climate stability are not fully known.


The other major advantage of a CSB regime is that Bondholders tackle whichever causes of climate change give them the best return for their outlay, whether they are natural or man-made. The more efficient Bondholders are in achieving climate stability the more they will gain from appreciation of the value of their Bonds. This efficiency maximises the degree of climate stability that can be achieved per dollar outlay. Because of the colossal sums involved, the benefits that CSBs offer in comparison activity-based regimes, such as Kyoto, are likely to be huge.


Further advantages of a CSB regime are:


·         The informational advantages that apply when there are large numbers of polluters. Greenhouse gases are emitted from many sources. About half of carbon dioxide emissions, for instance, come from dispersed sources, such as cars and home heating systems, and about 8.4 per cent of greenhouse gases (in the OECD) are emitted by the agriculture sector (OECD 2001, ).


·         That governments pay up only when a stable climate has been achieved: any risk of failure or of undershooting the climate stability target is borne by Bondholders, rather than taxpayers.


·         That funds for climate stability need not be used for scientifically approved projects. They could, for instance, be used to bribe corrupt or malicious governments. Appealing to these governments’ financial self-interest could be the most effective way of modifying their behaviour in favour of achieving climate stability.


·         That formulating the redemption terms for CSBs would entail clarifying of what is actually wanted. In global terms, climate change—as distinct from climate variability—could actually be a good thing. It could lead to longer growing seasons and higher yields in some regions. Others point to the boost to crop productivity given by higher levels of atmospheric carbon dioxide. ‘Climate Stability’ as targeted by Climate Stability Bonds could be defined such that Bondholders tackle only the negative effects of climate change.


Achieving a stable climate will probably require a huge range of different projects. Reducing greenhouse gas emissions or sequestering carbon might be helpful, but they are not necessarily going to be the most cost-effective. Other ways yet to be discovered could be far cheaper. Kyoto is deficient in that it offers no incentives to find out how to achieve a stable climate most cost-effectively. CSBs would encourage the most efficient solutions given the knowledge available at any time, and they would stimulate research into finding ever more cost-effective solutions. This occurs because of the nature of the Bond mechanism, and requires no presupposition as to the optimal set of solutions. The Bond issuers would dictate only the objective—climate stability—not the ways of achieving it. Crucially too, the objective would be one with which all shades of public opinion can support. Without such support no coherent policy addressing climate change is likely to succeed.

Next steps


Environmental Policy Bonds are one application of the principle of Social Policy Bonds, devised to inject market incentives into the solution of social problems, such as crime, unemployment, illiteracy and poor health standards (see bibliography). The Bond principle might though be especially suited to the environment, whose complexity means that it is difficult to identify and specify the full range of non-natural processes that generate positive or negative externalities, the degree to which they do so, and the relationships between these processes, natural processes and desired outcomes. This applies especially to larger-scale problems, including perhaps some agri-environmental problems.


Targeting outcomes rather than activities encourages diverse and untried solutions and bypasses scientific uncertainties. As well, existing policy instruments do not allocate resources rationally between controlling sources of environmental problems, and rewarding measures that solve them. And Environmental Policy Bonds need not be deployed in isolation: they can be used in conjunction with existing policies. Under a Bond regime it will still be desirable to apply existing regulatory sanctions against polluters, as well as the Polluter Pays Principle. A Bond regime would not take away the initiative for solving environmental problems from those with the expertise to solve them. The Bonds would simply provide a way in which that expertise could be effectively mobilised to achieve targeted outcomes. Environmental Policy Bonds can be used to reinforce, rather than substitute for, existing environmental policy instruments, by injecting incentives to make sure those instruments are used optimally. For all these reasons Environmental Policy Bonds may have advantages over conventional policy instruments.


Despite all these theoretical advantages, however, and the fact that it has been in the public arena for some years, governments have not yet applied the Bond concept either in agriculture or, indeed, in any other sector. However there are, at the time of writing, people outside government proposing to issue privately-backed Bonds for projects as diverse as boosting voter registration, raising literacy in developing countries, and developing open-source software. Governments appear to be more wary. Perhaps this is because under a Bond regime, they would surrender (within limits) their power to dictate how objectives shall be achieved, and which institutions shall be charged with achieving them. Government, along with those who administer current policies, might be reluctant to lose this power, even though they would continue to set and rank objectives. And elements of government, along with beneficiaries of some current policies would be likely to oppose a shift toward more transparent policymaking.


In any case, as an entirely new policy instrument, the Bonds will need to be introduced cautiously. Initial goals should be relatively small scale and uncontroversial, and the Bonds should complement, rather than replace, existing government or local authority programmes. Local authorities could begin by issuing Environmental Policy Bonds that target the water quality in particular rivers, for instance; indicators of success could be the number and variety of fish present. Such contained, easily identifiable, goals could help the Bonds gain acceptability amongst the public, and encourage policymakers to discuss and refine the concept. National governments could collate local experiences, and apply any lessons learned before issuing Environmental Policy Bonds addressing national concerns. Some deliberation would be essential, especially when Bonds target particular indicators for the first time. Then they are especially likely to encourage unanticipated negative behaviour by bondholders. Lessons learned from such initial issues could be applied to later issues targeting the same objective.


Trials of the Bond concept would be necessary to answer such basic questions as:





As well, there are technical aspects that need investigating; for example, which indicators can most reliably be used to measure society’s water or air pollution objectives?


The introduction of an Environmental Policy Bond regime might be accompanied by operational challenges and problems, not all of which can be anticipated. But these potential problems should not be overstated: existing laws, careful choice and specification of targeted objectives and perhaps more assurances as to how government will behave would probably circumvent or remedy most of them.


And the question of how well Environmental Policy Bonds would achieve society's environmental goals needs to be considered alongside existing policy-making methods. Currently policymakers, officials or lobbyists can escape or deflect censure because the adverse results of their policies are difficult to relate to their cause. If Environmental Policy Bonds were to lead to negative effects, the relationship between these effects and their cause would be easier to identify, and deterring such effects would be simpler, than doing so under current activity- or institution- based funding arrangements.


Resources are always going to be limited and Environmental Policy Bonds will not change that fact. Priorities and choices will always have to be made: under the Bond principle, governments will still decide on which environmental problems to solve, and on the sums allocated to their solution. In economic theory, and on all the evidence, markets are the best way of allocating scarce resources to achieve prescribed ends. Environmental Policy Bonds would allow both democratic governments and markets to do what each is best at doing: respectively: prescribing ends, and allocating resources to meet these ends as efficiently as possible.


Efficiency, of course, is not the only attribute of a successful policy instrument. Environmental Policy Bonds make transparent not only the goals of environmental policy, but also its returns, measured as environmental benefit per taxpayer dollar. They would also clarify the inevitable trade-offs that society will have to make between different environmental goals. In today’s complex economies, many people feel alienated from decision making, which to outsiders can appear remote and specialised. Such alienation can express itself in many forms, including apathy, cynicism and even violence. Too often government intervention in environmental matters is seen as bureaucratic, intrusive, and unrelated to relevant outcomes. With their focus on identifiable goals, Environmental Policy Bonds may, in a small way, help to bridge this gap, and encourage public interest and participation in policy formation. That could well prove to be as valuable to society as any efficiency gains.




Hayek, F A, ‘The Pretence of Knowledge’, New Studies in Philosophy, Politics, Economics and the History of Ideas, University of Chicago Press, Chicago, 1978.


OECD (2001) Environmental indicators for Agriculture, Volume 3: Methods and Results, OECD  Paris, France.


Schick, Allen The spirit of reform: managing the New Zealand state sector in a time of change, State Services Commission, Wellington, 1996.


Tangermann, S (1990) “A Bond Scheme for Supporting Farm Incomes.” In The Changing Role of the CAP, Marsh, J, Green, B, Kearney, B, Mahé, L, Tangermann, S, and Tarditi, S, Belhaven Press, London, UK.



Investing for the future (September–October 1991), Ronnie Horesh, UK CEED Bulletin No 35, Centre for Economic and Environmental Development, Cambridge, UK. (Presented as Room Document 3 to the December 1994 meeting of the OECD Joint Working Party of the Committee for Agriculture and the Environment Policy Committee.)


Social Policy Bonds: Injecting market incentives into the solution of social problems (August 1992), Ronnie Horesh, AEU Occasional Papers, University of Cambridge, Cambridge, UK.


Injecting incentives into the solution of social problems: Social Policy Bonds (September 2000), Ronnie Horesh, Economic Affairs, 20 (3), Institute of Economic Affairs, London, UK.


Better than Kyoto: how Climate Stability Bonds can inject market incentives into the achievement of a stable climate, (December 2001), Ronnie Horesh, Writers Club Press, USA. ISBN: 0-595-21164-X


Better than Kyoto: Climate Stability Bonds (September 2002), Ronnie Horesh, Economic Affairs, 22 (3), Institute of Economic Affairs, London, UK.


Injecting Incentives Into the Achievement of Social and Environmental Outcomes: Social Policy Bonds, (September 2002), Ronnie Horesh, ISBN: 0-595-24823-3, Writers Club Press, USA, September 2002.


© Ronnie Horesh April 2002


[*] Nor have Environmental Policy Bonds anything to do with the bond idea associated with Professor S Tangermann, which involves one-off payments to farmers in compensation for reduced agricultural support payments. See Tangermann et al (1990).