Version classiqueVersion mobile

Les espèces envahissantes dans l’archipel néo-calédonien

Marie-Laure Beauvais
Alain Coléno
Hervé Jourdan

Part one. Synthesis and recommendations

Evaluating economic impacts

Texte intégral

1Should a value be put on nature? On what basis can it be evaluated? Although disagreements persist as to the legitimacy of such evaluations, many economists think that economics can usefully be applied to the mechanisms leading to destruction of natural environments. Ecological and human systems are closely intertwined. The aim of an evaluation is to pro-vide tools on which to base environmental policy. But there has to be agreement as to the methods and criteria to apply. For the United States, different studies estimate the annual cost entailed by alien species at values ranging between US$ 1.1bn and US$ 138bn (though it is reasonable to suppose that the real cost is nearer the top end of this range). Nonetheless, such a wide margin shows the need for a proper understanding of the principles and limits of each type of economic analysis in environmental matters.

2Traditionally, this kind of analysis distinguishes between two major types of value for ecosystems: use value and non-use or passive value. Use value applies to ecosystems that provide goods and services, either for direct use, like production of a resource, or indirect use, like carbon storage capacity. Non-use value reflects the importance of resources that are not destined for use: existence value (e.g. the Bengal tiger), or heritage value (for future generations). A third type of value is option value, which provides a way of reflecting the idea of value for future use. The total economic value (TEV) is a combination of these three. Even so, it is practically impossible to give a value to a species or ecosystem. For example, not all species are considered equal: the last surviving members of a species may be attributed a higher value than a more abundant species. It must also be borne in mind that the economic value of an ecosystem is always limited to a utilitarian and anthropocentric approach.


3One of the difficulties encountered in assessing the economic impact of an invasive species is that often there are two sides to such impacts: a species may be detrimental to biodiversity yet a source of income for the country’s economy. Examples are some species introduced for agricultural or horticultural uses. The fact that a species is introduced intentionally or unintentionally cannot tell us whether it will have a positive or negative impact. Some deliberately introduced species cause considerable damage (e.g. the Nile perch in Lake Victoria in Africa), while accidental arrivals may prove to have an economic value greater than the problems they cause.

Table 9 – Estimated annual cost generated by introduction of non-native micro-organisms into the United States (source: Pimentel et al., 2000)

Table 9 – Estimated annual cost generated by introduction of non-native micro-organisms into the United States (source: Pimentel et al., 2000)

4Economic impact can be split into two categories: market impact (variations in price and associated surplus) and non-market effects. The cost of non-market effects can be considerable, as Table 9 shows.

5Impacts on ecosystems and public health are part of non-market costs, as are impacts on economic activities that exploit natural resources (crop losses, fall in land values, impact on water resources, livestock diseases, genetic contamination, management and eradication costs). Damage to infrastructure is another example, as with the zebra mussel’s colonisation of hydro-electric power plant and pipes in the Great Lakes region of North America.


6The purpose of cost-benefit analysis (CBA) is to study the relevance of a set of regulations. It weighs up against each other the economic consequences of the regulations (in terms of market functioning) and the benefits generated by the resulting risk reduction. Once scientists have assessed the risks? the probability of invasion and the expected environmental consequences? CBA offers policy makers a choice between several forms of intervention. The aim is to quantify in monetary terms the total of gains and losses connected with risks of invasion and policies for controlling these risks and to provide information on the relative efficiency of different risk management instruments. This tool naturally has its limits, and circumspection is called for when moving from economic results to prescribing environmental policy (see below, “strengths and weaknesses of CBA”). Although many international organisations and government agencies recommend the use of CBA it remains open to criticism, its relevance to biological invasions particularly.

7Cost-benefit analysis in environmental matters consists of evaluating the expected environmental impact of a particular policy scenario (e.g. inaction, an information campaign or introducing regulations). The aim is to put in perspective the expected benefits of reducing the risk and the cost of introducing a public policy. This assumes that it is possible to estimate, for each scenario, the probability of an invasion (depending on the natural, economic and regulator parameters, see p. 189), the nature and extent of possible environmental damage (depending on the same parameters), the value this damage represents for society and the benefits of preventing or reducing the damage. If the species introduced has beneficial effects, these must either be included in the environmental impact assessment (stabilisation of soils, recreational value of forests, etc.) or accounted for separately if they are marketable benefits.

8The quantification work must express all the effects of a biological invasion, both market and non-market, in terms of the same unit of value. For the sake of simplicity, a discounted monetary value is usually used so that the costs engendered by a project at different time horizons can be compared. The idea is to bring all values down to a single unit of measurement, which is usually the current value. However, there is some controversy over this practice in environmental projects because it tends to underestimate the future beneficial effects of a conservation programme (see paper by A. Thomas et al.). These benefits are often expected in the medium or long terms whereas the costs are immediate.

9A cost-benefit analysis is conducted in stages. The agents to be considered in the analysis must be decided on (costs and benefits for whom?). The regulatory measures envisaged must be identified. An inventory of all the potential impacts of these measures must be drawn up (environmental effects, market effects), using suitable indicators to measure them. The next stage is to give an economic value to each impact, discount the future costs and benefits, and add the discounted values. Then comes the “sensitivity” study to test the robustness of the analysis. It is then possible to recommend the regulatory option that will provide the greatest net social benefit.

10The inventory is a tricky stage, especially as regards measuring market effects. Market effects take account of the fact that public intervention to control an environmental risk is not neutral but has repercussions on supply and demand and hence on variations in the surpluses of agents, which must be included in the cost-benefit analysis (see paper by A. Thomas et al.). Some risk situations may be analysed as market failures which public policy instruments can correct (obligatory standards for producers, information policy to improve perception of risk, ecotaxes, civil or environmental responsibility policies). These instruments affect producers’ cost structures and consumer demand.

11Most relatively complete studies giving values for cost-benefit ratios of policies to control invasive species are from South Africa or the United States. Biological pest control in South America is a leading example, where the cost of research amounted to US$ 3 million between 1997 and 2000. But the ratio between the benefits of the programme and its cost, for a particular invasive species or region, is high: close to 1,130/1 for keeping the water fern Azolla filiculoides under control, for example. Other examples are the cactus Opuntia aurantiaca and control measures for Acacia saligna. The cost-benefit ratio of eradication programmes in watershed catchments are between 6:1 and 12:1, while those associated with plantations that cannot be economically exploited are between 360:1 and 382:1 (see description of all the examples in paper by A. Thomas et al.). Such values should be treated with caution, because these analyses tend to overestimate the likelihood of success of an eradication policy.

12Incentive measures play a crucial role. The failure of the 1999-2001 eradication campaign against the banana bunchy top virus in New Caledonia was probably largely due to the lack of sufficient incentives. Whatever the benefits expected from a risk management policy in theory, its effectiveness depends on efficient implementation. In the case of bunchy top, the population clearly did not play the game. Individual incentives did not persuade the public that it was worth taking part: the compensation per plant was considered small, there was a mandatory waiting period before people could replant, families whose plantations were destroyed were uncertain about the conditions on which they could obtain healthy plants, etc. Further, plantation destruction deprived farming families of a staple food for a long period. This example illustrates the importance of more thorough examination of the socio-economic effects of invasive species eradication and management policies in general.

13There is also the matter of cost sharing. Official action will not succeed unless polluters have a genuine incentive to take social costs into account (cost internalisation). This is the aim of such economic instruments as ecotaxes and civil and environmental responsibility. Different variants of the polluter pays principle may be envisaged, though attention must be paid to market effects. One of the aims must be to set up instruments that discourage neither trade nor tourism, but are sources of income intended to finance inspection, prevention and control measures. Biological invasions are perhaps the only externality in international trade which, to be properly controlled, would require a reduction in trade. This argues for a tax. Some authors argue that the best policy would be to calculate an ecotax. This would be similar to what economists call a Pigouvian tax. Part of it would be a contribution corresponding to the marginal damage of an introduction, the rest a contribution to inspection costs.


14Before any cost-benefit analysis is carried out, agreement must be reached on what policies are to be considered. These are then compared with scenarios of probable events. The scenarios give greater or lesser probabilities of a species becoming established or proliferating in the case of species already introduced, and greater or lesser probabilities of introduction in cases where this has not yet occurred. Given the many types of invasion possible? different species, different stages of establishment with different probabilities, etc.? it is generally wiser to plan for a combination of environ-mental policy instruments (see Table 10). Some may serve to help finance others, e.g. income from taxing imports or passengers can provide funds for inspection and quarantine services.

Table 10 – Invasive species control policies and scenarios

Table 10 – Invasive species control policies and scenarios

15A responsibility fund concerns exporters of goods who have provided a prior phytosanitary certificate. Funds are provided by payments prior to import, in proportion to the value of the goods. If a species is found to have been introduced and it is proven that it came from a particular importer, the payment is not refunded. This fund functions as a strong incentive to exporters. But there is the problem of proof, which has led to some rather more sophisticated plans (see paper by A. Thomas et al.). The advantage of such a system is that considerable funds are immediately available to support a more vigorous policy of quarantine or inspection, or to finance eradication measures.

16The idea of a solidarity fund assumes shared responsibility. Funds come from the public authorities and other concerned parties (importers, some categories of local producer, etc.). It is used for control measures against species already introduced, partly covering the cost of eradication and awareness/communication campaigns. It has no incentive effect, however.

17The tax on ornamental plants is an example of an ecotax or a special Pigouvian tax. It taxes local commerce. It embodies the idea that damage due to naturalisation or proliferation of non-native ornamentals should be paid for from the profits generated by the ornamentals trade. A particular problem is posed by the supply of seeds, bulbs and tubers by post, increasingly frequent with the development of e-commerce. The efficacy of these measures is therefore based on the authorities’ ability to inspect mail. Failing adequate human and technical resources to do that, an “indirect” tax could be applied, e.g. based on turnover, but this method may create severely disadvantageous distortions.

18Any policy on introduction of alien species must be adapted to each vector and associated entry pathway (import taxes, quarantine) and species (certification, inspection). On the other hand, a solidarity fund, responsibility fund or import tax does not depend a priori on the species concerned. For species already introduced, measures must be managed according to the species identified (monitoring, awareness campaigns). If the overall aim of policy is to prevent further invasions while eradicating or adapting to species that have already become established, the question is how to arbitrate between “traditional” prevention policies (border measures, quarantine etc.) and new measures. In particular, the probability of invasion may be significant for some access pathways (see Table 8).

19Once a set of policies has been selected, based on technical, legal and administrative feasibility, their cost must be assessed? mean annual cost for the first years based on net discounted values. These sums are then compared with the expected benefits, particularly in terms of protecting the local ecosystem and the various use values. One important parameter to consider is the probability of success: that is what will ultimately determine the damage prevented compared to the cost of the measures.

20Data gathering is a key point. This means finding all the necessary data for establishing the cost-benefit ratio of the policies under consideration. These data include data on damage caused by the invasive species (directly assessed or from the value of the natural resources used). The value of biodiversity is doubtless the most difficult and costly to obtain, which may mean that several levels of damage calculation must be used (with or without the value of the biodiversity). A second data set concerns policy costs. These must incorporate, among other factors, the impact on the final consumer if new instruments such as ecotaxes, solidarity funds etc. are introduced.

21Who bears the cost will vary according to the instrument chosen. Ecotaxes are mainly paid for by the end consumer, the direct cost to the authorities being very minor. For inspection and quarantine policies, New Caledonia will probably make major investments in equipment, and especially staff training. It is easier to quantify variations in cost to be expected from introducing an extended policy (as with pre-border or border controls). Also, given the scale that postal imports have reached, a major investment is required for mail inspection (X-rays, staff training). Private persons’ activities are harder to keep check on; there detection must be based on inspecting mail on arrival and checking import declarations.


22Cost-benefit analysis is the only approach with a theoretical grounding (in terms of theory of decision in situations of limited or imperfect information) and it involves seeking out all the scientific information available in order to improve decision aids. But the method has drawbacks as well as advantages. One problem is the need to provide a realistic distribution of the events the authorities are seeking to control. Similarly the probability of success of a control policy may be overestimated in the analysis, and this will give exaggerated benefit-cost ratios. CBA is certainly useful for its exhaustive a priori coverage, but in practice it rarely gives reliable and robust predictions in terms of policy choice. And it is very demanding of data (see box).

Data required for analysis of one management policy on one invasive species (non-exhaustive list)
mean current value of biodiversity in the territory
value added use of natural resources by agriculture, trade and industry
value-added use of existing resources by tourism
geographical proportion of regions and territories potentially affected by the biological invasion
proportion of above value-added uses suffering loss in connection with the invasion (or likelihood of invasion)
fixed costs of policy (investment in buildings and other fixed factors, training and education)
variables costs of policy (staff, intermediate consumption)
data required for evaluating market effects (household and industrial demand, international trade flows, substitute products, etc.);
probability of success for the policy
discount rate to be uses for calculating net discounted value of costs and benefits
weighting coefficients to be used in calculating the sum of benefits for different actor categories (rural and urban populations, manufacturers, importers etc.).

23But the most serious difficulty concerns the reliability of the analysis for evaluating environmental benefits, which concerns non-market goods not usually involved in any monetary transaction (see details in the paper by E. Gozlan et al.). One solution, where possible, is to base calculations on the value of associated goods that do have a market value? identifying revealed preferences. This is called the revealed preferences approach. Without benchmarks of this kind, one can conduct consumer opinion polls, asking people to imagine themselves in a fictitious use situation involving prices and use costs. This type of fictitious market study gives information about declared preferences rather than preferences revealed by actual behaviour. Guidelines for such surveys have been adopted internationally, with a relatively strong consensus; these are the NOAA Contingent Valuation Panel guidelines of 1993.

24Another difficulty is the very concept of environmental benefits. Critics point out that the cost-benefit analysis concerns non-market goods. Other critics reproach CBA for failing to take account of the interests of future generations. But the most frequent criticism is that the reasons most likely to drive environmental policy are difficult to quantify: ethical, psychological or religious considerations can scarcely be given a monetary equivalent. Further, relying on an evaluation of society’s preferences may not always be the best way to ensure respect for these considerations. And finally some writers criticise CBA for its inability to take account of the redistributive aspects associated with official policy. This criticism is not well founded, however.

25A final point of ongoing debate among economists is the conceptualisation and valuation of non-use values. Many authors doubt the very existence of such values, or at least doubt that the values attributed are at all meaningful. The existence value of a resource should in their view only be considered in terms of their use by future generations: it is because a resource could be utilised in future that it has an existence value today. In the applied economics literature, non-use and existence value are generally treated as a use of natural resources that is not yet profitable, but which could become so if market conditions proved favourable.


26In practice, it is usually a matter of choosing between a method that is satisfactory but very demanding of data (CBA) and a more robust method. What is needed is a decision criterion that is affected as little as possible by the degree of imprecision in the methods. In other words, if the estimations used in the CBA are too imprecise, it may be better to consider criteria that are less restrictive but require fewer data.

27One alternative method is based on the precautionary principle (Rio Declaration of 1992), which defines a framework for behaviour with regard to an activity that endangers the environment. The principle is that one should not wait for scientific proof of a causal link or of the extent of the damage before taking preventive measures. This principle is sometimes confused with an attempt to achieve zero risk, which would be incompatible with economic growth and the well-being of societies, which partly depends on innovations and their dissemination. A less restrictive view is to select decisions that are reversible or can be corrected in future. This means introducing an important concept into dynamic models with uncertainty? that of option value (see paper by E. Gozlan et al.). The option value is the difference between hoped-for discounted value following a decision after resolution of an uncertainty, and ex ante discounted value in a situation of uncertainty.

28In practice, the precautionary principle should thus result in arbitrating between two risks: the risk of taking decisions too late (possibly resulting in major damage) and that of taking over-hasty decisions (and so holding back economic growth). The debate over CBA versus precautionary principle approaches is not yet settled. The partisans of CBA seem to favour the hypothesis that catastrophic risks (climate change etc.) are still very unlikely and that our societies have time to prepare for them. Its detractors consider that in the absence of reliable probabilities for the various scenarios, it is impossible to calculate the likely damage. A compromise of sorts involves sufficiently tempering the precautionary principle to make it equivalent to an extension of the CBA in which risk aversion is taken into account.

29Another method, cost-efficiency analysis, unlike CBA, considers the environmental goal as an exogenous datum. Unlike CBA, which assumes that policy costs are known and their benefits uncertain, cost-efficiency analysis assumes the flexibility to lie in the cost, for a predetermined future benefit. One example of this approach concerns the goal of reducing greenhouse gas emissions (Kyoto protocol). In some cases this approach may be more robust that CBA, particularly if the benefits to be expected from a set of measures are poorly known.

30A last type of approach, based on reducing risk to a minimum, looks at extreme events in the distribution of probabilities connected with the risks studied (see paper by E. Gozlan et al.). Typically, the safety first strategy means deciding on a policy that minimises the probability of an unfavourable event.


31Uncertainty, in the case of invasive species, has less to do with the possibility of damage occurring than with its extent and the efficiency of any preventive measures applied. CBA can provide a transparent, objective framework. As a principle, this method has been widely adopted as a decision guide in matters of plant health. It is a fully-fledged risk management tool in the recommendations of the Intergovernmental Panel on Climate Change (IPCC), in the United States and in Australia. The New Zealand Biosecurity Act of 1993 describes the process of applying CBA to quarantine decisions. European regulations recommend it as an approach for decisions about damage to biodiversity (European Commission, 2000). The CBA principle is altogether compatible with the use of white lists or blacklists.

32Where data are lacking, several practical alternatives consistent with the CBA approach are possible. The White Book on environmental responsibility suggests using a “reasonableness” criterion. This means taking into account a subjective consensus on the reasonableness of managing a risk in view of what is known about the problem. This approach is very suitable where a quick decision is needed to avoid an out-of-control invasion. It is similar to a precautionary approach, prevention being generally less costly that cure. However, how best to divide a limited budget between preventive measures and post-introduction control measures is not a trivial question and is in itself a dilemma for the policy maker (see contribution by E. Gozlan et al.).

33Be this as it may, the search for an optimum prevention and control strategy for invasive species cannot be reduced to a problem of data availability. For example, the simplicity of the basic CBA principle (at least in its most basic version) masks an assumption that all interest groups carry the same weight, and that the policy maker is indifferent to their relative positions. This is a hypothesis about the policy maker’s preferences. CBA aims to classify different possible control policies regardless of whether the decision makers are in favour of them (it is a positive criterion, not a normative one). The other strategies we have discussed incorporate more sophisticated policy maker preferences. In fact each strategy reflects the implicit or explicit preferences of the authorities.

34In New Caledonia, the range of policies or measure considered is still narrow: a quarantine policy to be redefined with the establishment of white and black lists, inspection of shipping containers and postal parcels, ecotaxes. If the aim is to limit the risks due to inadequate application or control, a “safety first” approach should be taken. In this case garden shops and nurseries are no doubt the most affected. For example, a white list of species whose import is authorised means prohibiting all those whose innocuousness has not been proven. The cost of precaution is thus borne by those sectors that would benefit from their introduction, and no CBA is performed to justify their loss of income. One argument in favour of this approach would be that it encourages the parties concerned to get involved in producing the missing data. But if the authorities prefer the highest possible total unweighted benefit, they may prefer to spread the costs more widely (tourism, general imports), at the price of greater uncertainty as to the real benefits of the policy.

35Lastly, it is clear that the decisive factor must be the urgency of the issue. If the New Caledonian authorities want to combat efficiently (in terms of use of public funds) invasive species that have already been identified and ranked in order of dangerousness, then they should wait for more scientific information. In other words, a cost-benefit analysis worthy of the name takes time, and that time may or may not be precious. Further, the fact that species often have a latency period before proliferating sometimes makes it harder to identify which introductions should be prevented. On the other hand, if the situation is already considered sufficiently urgent for the precautionary principle to apply, criteria such as those presented above can be used to assess the suitability of short-term policies on the basis of limited scientific information. Though less efficient in the long term, these measures will make it possible to avoid an increased probability of ecological disaster due to irreversible consequences of an invasion.

Table des illustrations

Titre Table 9 – Estimated annual cost generated by introduction of non-native micro-organisms into the United States (source: Pimentel et al., 2000)
Fichier image/jpeg, 95k
Titre Table 10 – Invasive species control policies and scenarios
Fichier image/jpeg, 111k

© IRD Éditions, 2006

Conditions d’utilisation :

Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search