AEBC

AEBC/02/07

LIABILITY SUB GROUP

DRAFT PAPER ON ISSUES FOR DISCUSSION

1. This paper identifies a number of issues which the AEBC liability sub-group is grappling with, and sets out some of the background. The group’s thinking is now beginning to crystallise on a number of aspects, and group members would value the input of the full Commission as it starts to do so. Some issues for debate are identified below as a series of questions.

2. The sub-group may also suggest some early thoughts on possible scenarios as part of the debate at the AEBC Cardiff meeting.

Why study liability?
3. In making decisions about possible commercialisation of the GMHT crops in the Farm Scale Evaluations (FSEs), the Government will need to consider whether or not a specific liability regime is needed for unforeseen future effects. There is no specific regime at present, and taking as given that for any deliberate release of GMOs there would be regulation in accordance with the current understanding of risks, the future options include no changes in framing liability, or developing a specific system to deal with conditions arising from any future commercialisation of genetically modified crops.

4. AEBC has concluded the subject of liability is of sufficient importance and priority to study it specifically, to analyse options, and to advise Government on whether a regime is needed. The context for the study is the emergence of public concern about the environment and GMOs in the UK and elsewhere in Europe in recent years, the continuing development of scientific understanding of evidence and proof, the further development of an EU regime for environmental liability, and the development of precautionary approaches to new technologies.

What is liability? What are its limits? Functions of liability regulation and management
5. Provisions for liability apportion risks: they determine where risks lie in a range of circumstances.

6. Liability alone cannot deal with all issues, so the group will consider the limits to liability regimes: what they can and cannot do. Liability has to be seen alongside the functions of regulation and of management. The starting point may be to identify what should be protected and then to consider how to do so.

What is loss?
7. There are several types of loss which might give rise to harm, damage and liability, and be relevant to GMOs. For example:

Economic loss. An example of economic loss would be if farmers or beekeepers became unable to sell their produce for some reason related to GM crops.
Environmental loss or damage. An example would be effects on habitats or biodiversity.
Loss of or damage to property. For example physical damage to farms.
Damage caused by defective products is covered by an EC Directive: which also includes damage to private property but not to the environment.

What is the current position?
8. At present there is no specific liability regime for any form of agriculture. There are a number of relevant provisions in common/civil law and in criminal law, which might apply if GM crops were commercialised. These are outlined in Annex A.

What are the GMO issues for liability?
9. Particular issues for GMOs and liability include:

Unpredictability - and whether GMOs are different from other novel technologies in this respect.
Diffuse effects. Damage to biodiversity would be most likely from diffuse, cumulative impact from a number of sources rather than from a clearly identifiable single source. This might have analogies with, for example, pesticides and herbicides.
Co-existence of different farming types. The potential impact on other farmers and on beekeepers.
(The group is not considering issues like allergies or toxicity, as human health issues are beyond AEBC’s remit.)

GMOs: are they different - from what and how?
10. A key question, and a starting point for the study is whether GMOs are different from other potentially damaging activities in agriculture and the environment when looking at liability issues. It is important to debate and establish underlying principles at an early stage. GMOs are treated differently from other agricultural activities in the legal/regulatory system (through Directive 2001/18/EC for releases to the environment), which may /may not also be relevant for liability.

11. The group has started to consider whether GM crops throw up fundamentally different issues or orders of magnitude of potential harms from other new or existing activities. This raises questions for example about whether the issues are similar to transporting waste of different types, to manufacture and release of chemicals, to production or use of pesticides; and about how these areas are regulated. If the issues are broadly similar to those from a range of other activities, should GM liability be handled in the same way across that group of activities? If there is something specifically different for GMOs – or for some GMOs in some circumstances – what is it, and what does that mean for the requirements of a liability regime?

12. The emerging conclusions of the sub-group are that for liability purposes, GMOs are not very different from conventionally bred varieties or other forms of intervention that could be damaging, and thus that the issues on liability need to be seen from that perspective. The group considers that scientific, political and economic aspects might all have relevance, and need separate consideration. While GMOs pose new issues, so do other agricultural technologies, and a regime covering all hazards to the environment might be most appropriate. If there were damage, meansof redress could be more important than the way the damage occurred.

13. This links with the discussion in the animals and biotechnology sub-group. The view that has emerged from the animals work is that general regulation/law in relation to animals should be used to cover modern biotechnology applications wherever possible. This, in accordance with the principles of better regulation, is to ensure consistency and coherence and to avoid double standards ie the regulatory burden for biotechnology applications should not generally be greater than the burden of regulation relating to other uses of animals. Within this framework there may be a need for special provisions that deal with specific issues around biotechnology or biotechnology in the context of novel processes etc. Where new legislation/regulation is being devised this would be specific, but where legislation is being reviewed it could be as part of framework legislation. However, the animals group consider that uncertainty and public perceptions of potential risk point to a degree of special care when considering biotechnology in this broader legal context. This does not necessarily imply different law or regulation, but does imply a ‘spotlight’ of focussed attention, at least for a time.

14. A major task for the liability group is to study the potential for future damage from GMOs (and perhaps to GMOs) from unpredictable /unforeseen/ unforeseeable future events. The people affected might not be in a contractual relationship with either the producer or the user of the GMO. The extent of risk will be very hard to foresee.

Where are there gaps at present?
15. The group will be looking at this in detail, but coverage is clearly not complete, and for some aspects is in civil law and some aspects in criminal law. The draft European Directive on environmental liability would cover only a small proportion of farmland – that which falls within special sites such as SSSIs – and some species/habitats. There would be a number of circumstances within those areas where liability would not apply. For example companies taking all reasonable precautions would probably not be liable for later unforeseen outcomes. It seems unclear how the civil law provisions on ‘traditional damage’ eg to property would apply to GMOs. (The group has had briefing from a legal philosopher who considers civil law has potentially wide scope for application.) Product liability seems to apply effectively and across the board.

Do we need a liability regime?
16. Providing a clear framework and an understanding of where liability will lie might help to create incentives to minimise the risks. It could stimulate people to try to assess potential risk of exposure to harm and - conversely – risk of liability for damage. It could give a framework for insurance and other financial mechanisms to develop. It could help people form an idea of potential costs and uncertainty levels in pursuing cases.

17. If a liability regime were established for GMOs, it could for example specify that liability lay with the person responsible for the act or omission claimed to have caused the damage. This could be the source of the GMO eg seed company, industry producer or importer. Farmers or other users might only be liable if they did not meet requirements that they had undertaken to meet.

How is damage to be defined?
18. Damage would need to be defineable for any liability regime. In addition to damage directly caused, there is a question of whether diffuse damage should - or could - be included. Diffuse damage If there were commercialisation of GM crops, any figures for thresholds and separation distances for co-existence would be relevant triggers in any liability regime. Thresholds would be part of defining when ‘damage’ to a crop is caused.

Options? Scenarios to explore options
19. The group will be discussing the likely consequences of a range of options, and proposes to use a range of scenarios. The group will develop scenarios as examples, in order to work through and test out what might happen under different liability and other provisions.

20. When the group’s thinking crystallises, a major focus is likely to be whether the present structures for liability decisions should be left largely unchanged (and why), or whether a special regime is needed - and if so the range of options and preferences.

If there’s a problem, what are the ways of tackling it? If a separate liability regime for GMOs seems needed, what sort of regime should it be? What activities to cover? How to prove a causal link? Who would be liable? How would co-existence be relevant?

21. The whole area of looking at options and coming to conclusions is a very large one. Some initial points, relevant to a new framework, are outlined here. Probably the first option for the group to explore in detail is having no specific regime. Even if there were new provisions, the lead time would mean their introduction after Government decisions and their implementation following the FSEs.

22. One framework might be based on the concept of ‘novelty’ of product rather than of GM/non-GM, as it is in Canada, and this will be explored further to see whether or not it could be a useful way forward in the UK.

23. As liability spans environmental liability, economic liability and traditional damage, all of which concepts have different origins and principles, these need to be considered separately, but also in the way they do and would fit together.

24. For all of these there are options for ‘strict’ liability or a less stringent fault based liability. With strict liability, the person carrying out the activity, the producer – and/or farmer – would have to compensate for all damage. There would be no requirement to show that anyone was at fault. The test would be whether the act or omission caused the damage. The burden of proof of causation in civil cases is generally on the balance of probabilities (ie ‘more likely than not’). The more stringent the burden of proof test, the more the potential ‘victims’ are left bearing the cost. Strict liability could be qualified by conferring defences (for which the burden of proof switches to the defendant), such as ‘state of the art’, operating in line with permits, and negligence.

25. Liability might rest with those who developed the GMOs eg the company or seed importer. Farmers might be liable if they acted against the requirements of the company/seed importer, or the Government might be liable for giving permission. Would this approach pit farmers against producers in an ineffective way, in cases where a problem could arise from any of many sources if there was widespread release? Would a ‘single source’ approach give incentives for accountability through the supply chain?

What is the role of insurance?
26. Insurance provides a way of spreading risks over time and over a wider grouping. It increases the likelihood of compensation being available to potential ‘victims’. Re-insurance by insurers pools their risks further.

27. Insurance for GMOs is currently not well developed. Insurance is generally focussed on sudden and accidental damage, for which the risks can be actuarially calculated. Insurance against diffuse environmental pollution such as GMOs might cause, is far more difficult to come by. Insurance firms do not have claims histories to help them assess risks, or to assess the extent of precautions being put in place by companies, thus making it hard initially to set premiums. Initially the industry would aim to limit its own potential liability. Experience tends to build up over time, and the insurance market tends not to move into new areas rapidly. It would probably not be feasible – even if desirable – to make insurance cover compulsory for those using GMOs.

28. At present there is little ‘first party’ insurance for aspects such as crop value, although this is apparently common in the USA, where it is said to be subsidised by Government. It covers a range of reasons for the crop being of less value in the market place than expected – including storm damage, crop disease etc. This area might develop in the future. The Association of British Insurers have said that there may also be scope for development of financial packages.

29. It would be possible through ‘capping’, to limit liability of firms and for the state or individuals to take on the remaining liability. This was an option considered in preparation of the EC environmental liability draft Directive’s development, but not included in the draft Directive published in January 2002. The European Commission notes that capping is likely to improve the chances of early development of the insurance market but would erode the effective application of the ‘polluter pays’ principle.

30. In the UK there are some ‘financial security’ schemes - which go beyond insurance – for example for licensed waste sites and imports and exports of hazardous wastes. GMO firms might be required to demonstrate financial security, which would avoid the potential for firms becoming insolvent and not paying damages, and could be relevant in an industry where the company ownership changes are fairly common.

How to deal with unforeseeable damage or where no-one is at fault?
31. The group has considered the injustice that could be done to bio-tech companies behaving responsibly and obeying all relevant laws and protocols, and also potential victims who might not receive compensation for unforeseeable loss.

32. A compensation fund provided by the industry, or a levy, might be a way of ensuring funds were available, and it could also demonstrate confidence by the industry in their product and at the same time help to allay public concerns.

What about timescale?
33. The provisional timetable for the remainder of the FSE programme is at Annex C. If the Government concluded that a liability regime is needed, for example for economic liability, establishing a regime would require primary legislation. Because of the lead-in times, getting any legislation agreed and implemented could take some years. Government decisions on commercialisation may well be needed in a shorter time frame. The group will consider whether this would matter in relation to co-existence questions, in formulating its advice to Government.

34. The EU Directive on environmental liability is being discussed by Member States, will be debated and will probably evolve. Legislation is unlikely to be in operation for several years.

Do the same principles apply across the UK?
35. The same principles should apply across the UK, and AEBC will as always seek to provide advice relevant to all the administrations.

AEBC Secretariat
April 2002

ANNEX A

1. At present there is no specific liability regime for any form of agriculture. This Annex outlines briefly a number of relevant provisions in common/ civil law and in criminal law, which might apply if GM crops were commercialised.

Environmental liability: to restore/remedy damage to the environment.
2. In environmental law, there is extensive provision for criminal liability. The common pattern is that the emission of specified substances into specified environmental media is a crime, unless done in accordance with a permit. Part VI of the Environmental Protection Act 1990 also confers powers for remediation of environmental damage arising from release of GMOs if an offence is proved.

3. There is currently some domestic provision for civil liability for damage to SSSIs, from the Habitats Directive and Part 1 of the Wildlife and Countryside Act 1981 (as amended by the Countryside and Rights of Way Act 2000). Further details are at Annex B.

4. For the European Union, a Directive is planned on aspects of environmental liability for certain activities, including GMOs, though it will only cover limited geographical areas. The draft Directive builds on earlier EC environment Directives, including waste management licensing, pollution prevention and control, and contaminated land. There are special provisions for civil liability for the remediation of contaminated land in the Environmental Protection Act 1990. Further detail is at Annex B.

‘Traditional damage’ liability: to compensate for damage to property
5. Civil law is based on the common law of tort, with rules that may be modified generally by statute or in specific instances by contract. Liability arises when a person’s acts or omissions harm the interests of another person and certain conditions are fulfilled: eg (1)harm of a character recognised by the law - to private property in the law of nuisance; (2) cause and effect can be established by evidence; (3) a culpability requirement (other than in cases where there is absolute liability). In England the general test for environmental damage under the law of nuisance is one of strict liability based on foreseeability – as in the lead cases of Rylands v. Fletcher and Cambridge Water Company v. Eastern Counties Leather.

6. For physical damage to property or land or personal injury, the laws of nuisance apply. Individuals take cases to court for compensation. It is unclear how the existing principles would apply to cases involving GMOs.

Product liability
7. Product liability is covered by an EC Directive, which imposes strict liability on the producer of a defective product for damage caused by it. Damage includes damage to private property, but not damage to the environment. Products of the soil (eg cereals, oilseeds, pulses) come within the scope of the Directive. The provisions include a time limit for claims, a definition of ‘defective’ products, and a ‘state of the art’ defence - for applying all appropriate precautions at the time.

Economic liability: to compensate for economic loss
8. (The draft environmental liability European Directive does not extend to economic loss.) At present, economic loss - eg if a farmer were unable to sell produce at a price expected – would rely substantially on being able to show damage to property, or damage caused by a product. It seems unclear how current law would apply specifically to GM crops.

ANNEX B

ENVIRONMENTAL LIABILITY

EU draft Directive on environmental liability
1. This outlines the EU proposal for a directive on environmental liability, current UK arrangements, and issues such as insurance.

2. The EU proposal is based on Article 175 (1) of the Treaty of Rome, together with Article 174.

3. The regime would establish strict (ie no-fault) liability for environmental damage caused by a wide range of activities already regulated by EC provisions. Waste, water, dangerous goods, plant protection and pesticides (referred to as biocidal products) and GMOs would be included. Separately, there would also be fault-based liability for biodiversity damage caused by other activities.

4. Those responsible for significant environmental damage would be responsible for remedying that damage or reimbursing regulatory authorities such as the Environment Agency or Statutory Conservation Agencies (eg English Nature in England) for doing so in default. The role of regulatory authorities would be strengthened.

5. There is currently domestic (UK) provision on biodiversity for some liability for damage to Sites of Special Scientific Interest (SSSIs). The proposed Directive – unlike the present UK regime – is aimed only at serious incidents of environmental damage, and for these it covers a wide range of causes of damage and contains more stringent provisions on remediation than currently.

6. If the regime reinforced the “polluter pays” principle, this could internalise costs and shift the burden of damage from society (including the exchequer) to the polluter. For those potentially liable, covering their risks through insurance might reduce environmental damage by encouraging precautionary measures.

7. The draft Directive does not include traditional damage – generally private property, personal injury and economic loss.

8. The draft Directive includes defences that environmental damage which results from compliance with a statutory permit (“compliance with permit”), and environmental damage arising from activities not at the time considered, on the basis of the best scientific and technical knowledge, to be capable of having that effect (“best available scientific and technical knowledge”), are not within its provisions.

9. Biodiversity is defined as “biodiversity means natural habitats and species listed in….Directive 79/409/EEC…or Directive 92/43/EEC, or habitats and species not covered by those Directives, for which areas of protection or conservation have been designated pursuant to the relevant national legislation on nature conservation” and “biodiversity damage…is any damage that has serious adverse effects on the conservation status of biodiversity”. DEFRA understand that the reference to national legislation is only to nationally designated sites.

UK legislation
10. As noted above, there is currently domestic provision on biodiversity for some liability for damage to SSSIs. (Scotland and Northern Ireland have their own measures in some cases, with many similarities to the England and Wales legislation.)

11. The Habitats Regulations. The EU Habitats Directive (92/43/EC) is transposed by the Conservation (Natural Habitats &c.) Regulations 1994 (S.I. 1994/2716) “the 1994 Regulations” into law (England Wales and Scotland). Those Regulations and the Wildlife and Countryside Act 1981 (as amended by the Countryside and Rights of Way Act 2000) provide protection for threatened species and habitats.

12. The scope of damage covered is damage to the special interest for which the site has been identified. The liability regime is fault based. Owners or occupiers of SSSIs convicted of damaging sites via specified operations, and third parties convicted of reckless or intentional damage, can be ordered by the court to restore the protected site to its former condition where this is possible. The duty of the regulator is to maintain the status of EC protected habitats and species. The regulator has powers to order restitution.

13. Plans or projects that may have a significant effect on sites designated in accordance with the Directive have to be assessed. If they will have an adverse effect on such sites they can only proceed where there are no satisfactory alternatives and where there is an over-riding public interest, where measures are put in place to compensate for the damage.

14. All European sites in England and Wales have been or will be notified as SSSI sites. Legislation which came into effect in England and Wales in January 2001 introduced increased penalties for deliberate damage to SSSI’s:

Owners and occupiers of SSSIs are required to seek consent from the conservation agency (English Nature in England) for specified operations on an SSSI. The penalty for failure to obtain consent is up to £20,000 in the magistrate’s court and unlimited fine in the crown Court; the courts also have power to order restoration of the damaged special interest, where this is practicable.
Public bodies are required to notify the conservation agency when they plan to carry out operations which affect SSSIs, and are required to take any advice into account. They may be prosecuted if they cause more damage than necessary, and if they fail to restore the special interest where this is practicable.
There is also a general offence which applies to damage to the features of an SSSI, by any person, again, subject to a fine of up to £20,000. The agencies also have extended byelaw making powers on any SSSI.

15. Outside designated sites, species are also protected under the 1994 Regulations. A person found guilty of an offence (of taking, killing, injury or disturbance of a species, or of causing the damage or deterioration of a species’ place of rest or breeding) can be fined up to £5,000 per offence (per animal killed or injured, egg taken etc). A similar position exists for plant species at Regulation 43 of the 1994 regulations where the fine per plant is set at £2,500. Failure to comply with licensed exceptions carry similar fines.

16. There is considerable overlap between the species provisions of the Habitats Directive and Part 1 of the Wildlife and Countryside Act 1981 (as amended by the Countryside and Rights of Way Act 2000).

17. The UK has a range of arrangements for other areas covered by the draft Directive, often based on earlier EC environment Directives, including waste management licensing, pollution prevention and control, and contaminated land. The provisions on contaminated land in the Environmental Protection Act 1990 aim mainly at past rather than new or future damage. They tackle remediation in a different way from the proposed Directive. On water the provisions are similar in approach to the draft Directive but less far-reaching.

ANNEX C

FARM SCALE EVALUATIONS: PROVISIONAL TIMETABLE FOR REMAINDER OF PROGRAMME

March 2002	Routine news release announcing locations of spring maize sites
End April 2002	Agriculture and Environment Biotechnology Commission to provide further advice on how and when to promote public debate on possible commercialisation of FSE crops
July 2002	Routine news release announcing locations of winter oilseed rape sites
Autumn 2002	Field work (data collection) ends for FSE spring 2002 crops and researchers begin to collate results
Early 2003	Researchers send draft report to Scientific Steering Committee to validate research results for spring-sown crops
Summer 2003	Final results for spring-sown crops published in peer-reviewed scientific journal at same time as SSC provides advice to Government. Results considered by Advisory Committee on Releases to the Environment and the Advisory Committee on Pesticides. Results available for public comment
Autumn 2003	Field work (data collection) ends for winter rape 2002 crops and researchers begin to collate results
Early 2004	Researchers send draft report to Scientific Steering Committee to validate research results for winter rape crops
Summer 2004	Final results for winter rape crops published in peer-reviewed scientific journal at same time as SSC provides advice to Government. Results considered by Advisory Committee on Releases to the Environment and the Advisory Committee on Pesticides. Results available for public comment

Notes:
(i) Timings are approximate
(ii) The FSE results will be presented on an individual crop basis. There are three spring-sown crops (oilseed rape, beet and maize) and one autumn-sown crop (winter oilseed rape). The results will therefore be published in two tranches.

DEFRA GM Crops Policy Branch
February 2002