Papers prepared for 2002 conferences on NAFTA Chapter 11 held in Mexico (March 13), Ottawa (March 18), and

at the Institute for International Economics in Washington, DC (April 11)

Economic Issues Raised by Treatment of Takings p. 2 Under NAFTA Chapter 11 By Edward M. Graham, Senior Fellow, Institute for International Economics Protecting Investor Rights and the Public Good: p. 30 Assessing NAFTA’s Chapter 11 By Howard Mann, Associate, International Institute for Sustainable Development, and Konrad von Moltke, Senior Fellow, International Institute for Sustainable Development

Economic Issues Raised by Treatment of Takings Under NAFTA Chapter 11 By Edward M. Graham, Senior Fellow Institute for International Economics Paper prepared for 2002 conferences on NAFTA chapter 11 held in Mexico (March 13), Ottawa (March 18), and at the Institute for International Economics in Washington, DC (April 11). 1. Introduction

This paper examines, from an economic perspective, the treatment of takings (or property

rights) under NAFTA chapter 11 (to be more precise, the treatment of takings as environmental

groups fear might be established as the result of investor dispute settlement under this chapter; as

of the date of this writing, most of the cases that have the potential to be precedent-setting have

not been finally decided, albeit one—the Metalclad case has been decided in a way that is

unsettling to environmentalists1). In adopting an economic perspective, the author attempts to be

as value-free as possible but recognizes, however, that no analysis is ever completely value-free.

Thus, the attempt is made to determine whether requiring public compensation of private

investors for diminishment of value resulting from government regulatory action has the

potential of achieving anything close to an “optimal” outcome from a societal cost-benefit point

1 A number of disputes, e.g., most importantly the cases commonly called “Metalclad” and “Methanex”, have been lodged under NAFTA Chapter 11 part 2 (establishing the investor to state dispute settlement procedures) where the complaint of the investor is that regulations promulgated by relevant governments or government agencies have resulted in diminished value of investments that is tantamount to expropriation as per NAFTA article 1110.1.b and hence subject to requirements for compensation under articles 1110.1.d and 1110.2 - 1110.6. A final decision has not been rendered on Methanex, arguably the most important of the cases. In the case of Metalclad, the dispute settlement tribunal decided in favor of the company. Although the decision per se is unsettling to the environmental community, what worries the environmentalists most of all is the reasoning of the tribunal, which held that the decision by the government of Sonora (a state of Mexico) to disallow the company from operating a toxic waste disposal site there was tantamount to an expropriation because of “significance of impact” (or, otherwise put, because the company lost all or most of the value of its investment because of the disallowance, without taking into account that the motive of the decision was to prevent environmental harm, not necessarily to expropriate the property). On these issues, see Mann and von Moltke 2002. Complete descriptions of outstanding disputes submitted to dispute settlement under NAFTA 11 are found at www.NAFTALAW.org.

2

of view (defined below). This determination makes use of tools of economic analysis and, in

particular, Coase’s theorem regarding achieving optimal outcomes where negative externalities

are present.2 The overall conclusion is that, although Coase’s theorem can be invoked to argue

that such an outcome can be achieved either via a “polluter pays” approach or a “public pays” (or

“public must compensate”) approach, as a matter of practical application, the first approach is

preferable to the second for a number of reasons, including government “fiscal illusion” and

“moral hazard”.

Section 2 of this paper reviews Coase’s theorem and establishes the main result that

follows from it: if a bargaining process can be established and properly managed under the right

circumstances, either of the two approaches noted above can in principle yield the same outcome

in terms of achievement of a goal to reduce an external cost. As is well known, the two

approaches do yield differing results with respect to who actually bears the costs associated with

this reduction. Sections 3 and 4 then discuss respectively the issues posed by fiscal illusion and

moral hazard; the conclusion in each is the same—that in spite of the neutrality in principle of

Coase’s result regarding the best direction for public policy to take with respect to whether to

assess the polluter or the public for costs of pollution abatement, the former dominates the latter

when issues of practicability are considered. The overall conclusion then is that, although the

case for public compensation of investors for diminished value of investments induced by

environmentally motivated regulations is not wholly without merit, as a practical matter,

application of the “polluter pays” principle is preferable. To the extent that this is correct, it is

also arguable that use of NAFTA chapter 11 as a vehicle to force such compensation for such

2 Coase 1960.

3

diminished value is likely to lead to nonoptimal results.3 This of course would suggest that the

expropriation provisions of NAFTA chapter 11 should be interpreted, or perhaps even amended,

so that these would not cover this type of taking; this matter is discussed in the concluding

section (section 5).

2. Coase’s Theorem on External Costs, Somewhat Simplified but not Simplistic From an economic perspective, environmental problems originate, or at least to a large

extent, from “market failure”, i.e., a situation where the working of a market, even one in which

most of the standard conditions for optimality would seem to be met, fails to create an outcome

that meets a test for social optimality. These standard conditions for optimality include that the

market be characterized by something approaching perfect competition, which in practical terms

means that no seller or buyer in the market has market power (i.e., can unilaterally affect the

price or quantity of the good or service being sold in the market), and that both buyers and sellers

are well- informed with respect to the market, such that no buyer or seller possesses info rmation

bearing on the market not possessed by other buyers and sellers, which can be used as the basis

for gaining advantage.

Under most circumstances, if these conditions are present, a market will produce an

optimal outcome in the sense that consumer surplus minus total costs will be maximized when a

market-clearing price and quantity are achieved. This is equivalent to saying that, on a social

3 This remark does not apply when the clear intent of the government measure is to expropriate a property but, rather, it applies only to cases where the intent is to achieve a legitimate environmental goal. Admittedly, there could be “grey” cases, e.g., where the intent of a government is not clear or where expropriatory intent is disguised as a legitimate environmental goal. My own view is that it should be up to the NAFTA tribunal to determine true intent in a dispute case where intent is not clear and thus that intent of a measure is a legitimate issue for consideration in a NAFTA Chapter 11 case. For more on this, see von Moltke and Mann 2002, op. cit.

4

benefit-cost basis, total benefit is maximized. The details of how and why this happens are

explained in any standard textbook on microeconomics and thus further explanation is omitted

here. It is also true that this optimality is violated if the conditions are not met, e.g., that

competition is “imperfect” because one or a small number of sellers in the market do possess

market power. However, market failure resulting from market power of sellers (or indeed

buyers) is not the topic here. Rather, we are concerned about market failure that can occur even

when market power is absent.

What goes wrong so as to create a market failure even where these conditions hold? The

main culprit, and the one that concerns us most, is a so-called “negative externality” associated

with use of a public good.4 If present, such an externality results in total costs associated with the

buying and selling of a good to exceed the costs that are borne privately by sellers. Such an

externality (or, equivalently, an “external cost”) could occur in the form of harm to the

environment, e.g., as the result of an increase in air pollution that causes degradation in human

health (or, indeed, degradation in the health of forests or wildlife).

Suppose, for example, that air pollution, measured in units of total emission of pollutant u

(where u is interpreted as a total of u units of pollution), creates total external costs as

represented in curve PP in diagram 1. Note that these are indeed external costs and not the costs

associated with the production of some product; costs of the latter sort also are incurred, but are

private (“internal”) costs. We shall consider momentarily how these fit into the total picture.

4 For purposes here, we shall define a public good as one that is available for use by any user such that, in the absence of specific governmental measures to the contrary, its use is free of charge. More precisely, economists note that public goods are characterized by (1) lack of “rivalry”, i.e., the use and enjoyment of the good by one person does not diminish the supply and enjoyment available to any other person and (2) no party can appropriate the good for personal gain, or at least not in the absence of a government measure enabling such an appropriation. For the particular public good used to illustrate much of what follows, the air, these two conditions are only approximately met but this approximation is “close enough” (the supply of air is finite, but for all practical purposes my use of the air does not diminish the supply of air needed by my neighbor, and I certainly am not able to sell this air for profit).

5

Diagram 1. External costs associated with use of a public good (unpolluted air), and private costs associated with reducing the use of this good (reducing the amount of pollution)

The PP curve not only rises with the total amount of emission, but it also rises with increasing

slope. This “rise at an increasing rate” occurs because, as total emission of the pollutant grows,

the total external costs created by it increase at an increasing rate (if the amount of pollutant

released into the air is doubled, for example, this is likely to result in more than double the

6

number of pollution-related illnesses). Removal of these costs creates a benefit to society that is,

in magnitude, exactly the same as the cost.

Another curve—the AA curve—is indicated on the same diagram. It indicates the total

cost of reducing the pollution to level u, given some initial level of pollution. This cost would be

expected to rise as a function of how much pollutant is removed from the air and, thus, because

removal of pollutant is the inverse of emission of pollutant, on this diagram the AA curve falls as

the total amount of such emission increases. This fall is at a decreasing rate (the slope gets flatter

as total emission increases) because the marginal cost of cutting out an additional unit of

pollution rises as total emission of pollution is cut back.

What is desired is to minimize total cost, where this total cost is the sum of the AA and

the PP curves. This sum is given by the TT curve, and it is, as drawn, U-shaped.5 As can be seen

from the diagram, this minimum is reached at point u*, which is in fact where the slopes of the

AA and the PP curves are equal in magnitude.6 To the right of this point, e.g., at point uH, adding

a unit of pollution increases external costs by an amount greater than the cost of abating that unit.

But, to the left of u*, e.g., at point uL, reducing a unit of pollution costs more than the benefit that

is gained from a reduction in the external cost associated with this unit of pollution.

But, if u* indeed is the optimum level of pollution, in the sense that net benefit is

maximized at this point, the question remains—how to get to u*? Nobel-prize winning economist

5 This shape is likely, given the shapes of the PP and AA curves, but is not inevitable. See footnote 4. 6 Let total costs, as a function of total amount of pollution u, be given by T(u) = A(u) + P(u), where A and P are the AA and PP curves described in the text, which we assume to be C2 functions (continuous and differentiable, with continuous derivatives). A necessary condition for T(u) to be a minimum at some u that is not at either of the end-points of the domain of values taken on by u is that dT/du = 0, which can be true only if dA/du = -dP/du. Given that A and P slope in opposite directions and that the magnitude of dA/du increases with increasing u while the magnitude of dP/du diminishes with increasing u, this condition is likely to be met. Even if it is not met, the strict convexity of T (it is strictly convex because the second derivative of T with respect to u, d2T/du2, is strictly positive) guarantees that a local minimum does exist. If the condition dA/du = -dP/du is not met, the minimum will occur at one (but not both) of the end points of the domain of u.

7

Ronald Coase noted, in an article published more than 40 years ago (see footnote 2), that two

approaches would work. In the first approach, property rights to the air would be assigned to

society at large, so that holders of these rights, members of the general public, could in effect sell

the right to pollute to producers. Suppose that, at the moment that these rights were established,

the actual level of pollution was uH. In order to continue in production, polluters must either (1)

acquire property rights for use of the air and, by buying these rights, compensate the public for

the costs that the pollution forced citizens to incur, or (2) undertake pollution abatement, so that

the public’s property rights are not violated. As long as pollution levels were currently above u*

(which is in fact so, by assumption), it would in fact be cheaper for the polluters to abate their

pollution than to buy property rights, assuming that the public would sell these rights for no less

than the costs created by the pollution. But, once pollution is reduced to u*, a crossover point is

reached, where further abatement costs more than purchase of property rights, or at least so if the

rights are priced at the same level as the external costs. The least cost strategy for polluters

would thus be to reduce pollution to u* and then to buy from the public the right to pollute at this

level.

But, alternatively, property rights to use of the air, even as a garbage dump, could be

assigned to the polluters, but polluters could be required to forego pollution if these rights were

bought by the public. Under this approach, if at the time that this right is assigned the actual level

of pollution were again to be uH, citizens could organize to buy from the polluters enough rights

to force the polluters to reduce emissions to u*. As long as the price paid for these rights were at

least as great as the cost of the abatement, the firm should be eager (or at least indifferent) to

selling the rights and incurring the abatement costs. Because the cost of reducing pollution from

uH to u* remains less than the external cost borne by citizens at uH, citizens should be willing to

8

pay a price equal to the cost of abatement. But once u* is attained, the cost of further abatement

(which now must be borne by the public) becomes greater than the costs associated with

enduring this level of pollution. Therefore, the public might be expected to buy rights to clean air

until the level u* is attained, but not to buy rights to drive pollution below u*.

Thus, by Coase’s logic, the optimum u* can be reached by assigning a property right to

air either to the public or to the polluter. The two approaches of course do have differing

implications for the one who bears the cost of pollution abatement, the polluter or the public.

But, the point established by Coase is that, under either approach, the optimum can be achieved.

The issue of who pays the costs of abatement is important, of course. If property rights to

the air belong to the public, it is the polluter who must pay these costs (and this is essentially the

same as the “polluter pays” principle and variants on it, such as “cap and trade”).7 If, by contrast,

the property rights to the air belong to the users of the air (the polluters), the public must pay

them in exchange for them taking action that reduces external costs. Given that someone must

pay for the reduction of pollution, it becomes a matter of social choice as to who this someone is.

Alas, the analysis thus far does not address several problems that can be readily

identified. One of these is how to determine the external cost borne by a member of society as

the result of pollution. In practical terms, this is at best difficult and in fact the determination

might rest upon values of members of society that differ from member to member.8 Coase had in

7 Note that in the example given above, if property rights to the air are given to the public, the polluter not only must pay for abatement to bring pollution to level u*, but also must compensate the public for creating even this amount of pollution. However, as is elaborated upon shortly, if it is the polluter who pays, the cost is in fact likely to be reflected in a higher price of the relevant product if the pollution is created by production of the product. In this instance, at least some of the cost of pollution reduction is passed on to buyers of the product, who are themselves members of the public. Thus, in this latter instance, the issue of “who pays” comes down, to some large extent, to “exactly which members of the public pay”. 8 For example, although a resident of Washington, DC, the author is a native of California. As such, he highly values pristine redwood forests and assesses that the external cost of loss of these forests to timbering far exceeds the value of the picnic tables that would be made from the felled trees. Ronald Reagan, a non-native of California, was some

9

mind that each member of society would know with some precision what the cost to her or him

of pollution actually is, such that he or she could then, armed with this knowledge, bargain with

the polluters over the price to be paid for exchange of property rights. The polluters also are

assumed to be knowledgeable with respect to the precise external costs associated with pollution.

If both sides are thus knowledgeable, a bargaining process might indeed lead to an outcome

whereby the “right” price would be established by which exchange of property rights would

result in an optimal level of pollution. 9

However, in reality, most persons probably have no real idea of the cost to them of

breathing polluted air; indeed, if they did, many people might choose to live someplace other

than where they currently do! And, also, it is quite plausible that polluters are similarly ignorant

about the external costs created by their pollution, even if they might claim otherwise (but, on

this, one should see the movie Erin Brockovich).

A second problem is that, even if both the external costs of pollution and the costs of

abatement were accurately known both to polluters and the public, to organize the trading of

pollution rights so as to achieve the optimum, irrespective of who initially held these, might itself

prove to be quite costly or even ineffective. This is why this requires a little more than

imagination: suppose that property rights were given to the polluters for the air over Los

Angeles, California, and then a public auction were to be held whereby (it was hoped) trading of

rights would lead to an optimal outcome. All polluters and all citizens would be required to come

to this auction. A very large tent would be needed. Collective action problems would almost time ago elected governor of that state. It is clear that he held different values, as he allowed vast stands of redwood trees to be cut in order to be made into picnic tables. A proper valuation of the external costs associated with loss of redwood forests must take into account both the preferences of this author and those of former Governor Reagan and, indeed, all others who might be affected by the loss. While this is theoretically possible to do, in practice it is at best very difficult. 9 This assumes that the bargaining process itself is costless, which might not be the case. Rather, bargaining transactions costs could be significant. See paragraph that follows.

10

surely occur (e.g., citizen A might be willing to pay for cleaner air, but might hold back from

bidding for clean air rights in the hope that these would be bought by his neighbors. Of course, if

the neighbors did so, citizen A would get the benefit without having to pay his share of the cost.

In order to prevent such “free riding” by citizen A—and citizens B, C, D, ad infinitum—some

sort of complex mechanism would have to be agreed upon whereby the cost of property rights

acquired by the public was shared by all members of the public, where account was taken of the

fact that the value of clean air might not be the same for all such members. Similar problems

could be imagined if the property rights were held by the public initially. Given these problems,

it has been suggested that, for anything like a Coasian bargaining process to work, an agent

representing the public (e.g., the government) must do the bargaining and, where appropriate, the

assessing of the public. It has been further suggested that, if property rights for clean air rest

initially with the public, for an efficient outcome to be achieved, the best approach might be for

the government to assess an effluent tax on polluters, or to sell pollution rights in some form of a

“cap and trade” scheme.10 The effluent tax is discussed below. Likewise, if the property rights

rest initially with the polluters, the best approach might be for the government to require

pollution to be reduced to the optimal level (assuming that this can be determined) and then to

compensate relevant parties for abatement.

A third problem is that the analysis thus far does not take into direct account what effects

moving from uH to u* might have on the price of the relevant products (those whose production

or use creates the pollution in the first place). The analysis thus far is of the nature of a “partial

10 Cap and trade schemes have been much discussed in the literature on environmental economics. For a recent contribution that discusses the outcome of implementation of such a scheme for emission of sulfur dioxide in the United States, see Ellerman et al. 2000.

11

equilibrium” analysis, i.e., the effect of pollution abatement on product prices is not considered.11

But, if these prices change as the result of the trading of property rights to use of the air, these

changes do have implications for social benefit.

This latter matter is of some importance, so let us now explore this effect using some

simplifications that are necessary to keep things manageable. Diagram 2 is a modified version of

a simple supply and demand schedule, where the DD curve represents demand for some product.

Total social benefit created by sale of this product is measured by “consumer surplus”, the area

under the DD curve bounded by the Q axis and Q*, which represents the total amount of the

product bought and sold where the market clears. (Why this area measures total benefit is again

explained in any standard microeconomics text, and this explanation is not repeated here). The

SS curve represents total supply under an assumption that competition is “perfect” and marginal

cost of production is constant with respect to total output; thus the price/cost axis intercept of this

curve (at point C) is equal to the unit cost of production as borne by sellers of the product, and

the SS curve is a horizontal line extending from this intercept.12 The intersection of the SS and

DD curves (at point A) represents the price and quantity (P* and Q*) that both clear the market

and maximize consumer surplus. Thus, also, in the absence of an external cost, these are the

price and quantity at which net social benefit, i.e., total social benefits minus total social costs, is

maximized. The latter, total social costs, are in this instance given by the area under the SS

curve, bounded on the right by point A, and thus are given by the area of the rectangle OCAQ*.

Net benefits equal total benefits minus total costs and are given by the area of the triangle CDA.

11 These costs might be embodied in the AA curve, but we will now consider them more explicitly. 12 As is standard in this type of analysis, this cost includes components to give sellers a competitive rate of return on any capital they have invested in the selling of this product and to compensate them for their own, as well as their employee’s, labor.

12

However, what again is at issue is that there is an external cost—one that is borne by the

society rather than by sellers of the products—that must be accounted for. To account for this

additional cost in a way that keeps things simple, let us assume that this cost increases by an

equal increment E for every unit of the product sold in this market. In other words, we assume

that each unit of production produces a constant amount of pollution and total external costs of

the pollution increase linearly with pollution. (And, thus, we have dropped, in the interests of

keeping things simple, the assumption used earlier that total external costs rise, as a function of

total pollution, at an increasing rate. Also, the reader should note that, in this example, to reduce

pollution, it is necessary to reduce output. This is by design; the goal here after all is to

demonstrate welfare effects of changes in output.) Thus, the unit social cost is no longer equal to

private cost but rather is equal to the marginal (private) cost C plus the unit external cost E, so

that the social cost per unit of output is B, where B = C + E. Marginal social cost thus is given by

the curve BB, which by virtue of the simplifying assumptions is constant with respect to output.

Total social cost is given by the area under this curve, again bounded by the Q axis and the

amount of the good that is bought and sold.

To maximize social benefits (which still equal consumer surplus) net of social cost, the

quantity of the good bought and sold should now be QB, the quantity at which DD equals BB.

One should note that QB is less than Q*, reflecting that the true cost of the good is not C but the

higher B. By reducing the amount of the good bought and sold from Q* to QB, the amount of

pollution is reduced, and external costs are reduced accordingly, to an optimal level.

13

Diagram 2. Reaching the optimum price and quantity where an external cost is present

An environmental purist might argue that, if sale of this good creates an environmental

harm, societal interests dictate that its sale ought to be banned. In fact, this possibility is not

precluded from this analysis. Diagram 3 depicts a case where such a ban is warranted but where,

in the absence of a ban or a measure equivalent to a ban, sales will proceed. In this instance, the

14

total social cost of produc ing even one unit of the good, where that cost takes into account the

environmental harm, exceeds any possible benefit achieved from sale and use of the good.

(The activist must recognize that, in the absence of the external cost, there is a net benefit to

producing the good; if there were no benefit, there would be no production under any

circumstance). But production of the good should not take place if the true costs of such

production at any level of sales exceed any possible benefit. Picnic tables made from the wood of

first-growth redwood trees come to mind (see footnote 6 above). There may be some benefit

from such tables (they are attractive and durable), but this benefit is scant compared to the cost to

society of loss of magnificent redwood groves.

But how do we achieve the socially optimal quantity QB rather than the market-clearing

quantity Q*? After all, left to its own devices, the market will deliver Q* and not QB; again, Q*

is the market clearing quantity, and this is not changed by the presence of the external cost E.

Again, we can make an appeal to the reasoning behind Coase’s theorem.

As before, following the reasoning of Coase, there are in fact two ways to get to the

desired equilibrium. Let us, in what follows and consistent with the first example, assume that

the external cost in our example is attributable to air pollution. One approach to getting to the

equilibrium then is, again, to assign property rights to clean air to society at large, such that

society can charge suppliers for use of this air as a place to dump their gaseous wastes. The other

approach once again is to assign property rights for use of the air to the suppliers, but to allow

society at large to buy back some of these rights. Will either approach still work to get to the

equilibrium, albeit that the two routes will again have differing implications with respect to

exactly who bears the costs?

15

Diagram 3. If total costs B, including both internal private costs and external social costs, exceed the reservation price D, there will be no output of the relevant product. Internalization of the external cost is equivalent to a ban on the production and sale of this product.

Consider first the case where property rights are assigned to society at large. In order to

use the clean air (in this case, to “use” the air means to dump effluent into it), the suppliers of the

product must pay to acquire the right to pollute. As suggested above, in fact, a more practical

means of achieving this payment and transfer of rights than Coasian bargaining might be that the

16

government, acting on behalf of society, impose a tax on the suppliers for use of the air, and then

to distribute revenues gained from this tax to members of society affected by the pollution. 13 An

issue that then presents itself is, what price per unit of output (or rate of tax) is right? Clearly, to

get to the equilibrium, this price/tax should be equal to (E-C), the external cost, so that (E-C)

becomes fully internalized by the producers and an ordinary market clearing process yields the

desired outcome. Fortuitously, for the holders of the property rights, (E-C) in fact is exactly the

price that maximizes benefit to members of society. This becomes evident from the following

considerations: If a price/tax less than (E-C) is charged, members of society are under-

compensated for the external costs that they must bear (see diagram 2A). But, also, if a price/tax

greater than (E-C) is charged, members of society will lose consumer surplus that is in excess of

the revenue gained from the tax (see diagram 2B), such that total benefit to society is reduced.

Only when (E-C) is charged is the sum of net benefit, i.e., consumer surplus plus revenue from

the sale of the property right minus total social costs, maximized.

But a similar case can be made if property rights are granted to the suppliers. As long as

the amount of the product offered on the market exceeds QB in diagram 2, external costs borne

by the public will exceed additional consumer surplus created by a lower price being offered

than the price which clears the market at QB (see diagram 2B). Thus, there would be a net gain to

13 Such a tax was in fact proposed by noted British economist Arthur C. Pigou early in the 20th century and hence is now commonly known as a “Pigovian tax”.

17

Diagram 2A. If output QP is set above the socially optimal quantity QB, uncompensated external cost will exceed appropriable consumer surplus.

18

Diagram 2B. If output QP is set below the socially optimal QB, appropriable consumer surplus will be lost in excess of compensation of external cost.

society to paying suppliers an amount equal to QB x (E-C) to limit output to QB.14 This could be

done in two ways: by society paying the suppliers the amount QB x (E-C) in a lump sum, in

exchange for which the suppliers would agree to limit supply to QB but to hold price at C.

14 If quantity were to fall below QB, there would be a net loss of benefit due to lost consumer surplus, exactly as in the case described previously (see diagram 2A).

19

Alternatively, suppliers could agree to limit supply to QB, causing the price to rise to the new

equilibrium (C+E), and to appropriate the amount QB x (E-C) as a rent.15

Thus, by Coase’s reasoning, the same (and desired) outcome again is achieved either by

assigning a property right to clean air to the general public and allowing this to be sold to

suppliers who need this air in order to make a product which is desired by society or,

alternatively, by assigning the property right to the air to the suppliers but allowing society to

compensate these in some manner for reduction of use of the right in order to meet a socially

desired end (i.e., reduction of effluent).

But, as in the earlier example of diagram 1, there is a big difference between the two

assignments in terms of who ultimately pays in order to achieve the optimum level of output QB.

However, in the current case, the specifics of who pays are somewhat different. In fact, in both

assignments, the achievement of the optimum is accomplished via a rise in price of the product

and a consequent scale-back of production. Thus, in both assignments, in fact the ultimate payer

is the user of the product, who must pay a higher price for a more constrained quantity of this

product. But, if property rights are assigned to suppliers, all consumer surplus appropriated by

them is retained by them. By contrast, where property rights are assigned to the general public,

consumer surplus appropriated by suppliers is returned to the general public in the form of a tax

revenue (or, in a cap and trade scheme, public revenue generated by initial sale of effluent

rights). Whether the revenue ultimately finds its way back to the individual members of society

is an issue worthy of some consideration. But, at least in principle, the consumer surplus is

15 The first possibility creates a condition of excess demand; that is, some consumers would be willing to buy the produce at the price C, but would be unable to obtain it. Under these circumstances, the product would have to be rationed. For this reason, economists would prefer the second possibility.

20

returned to society at large, which of course includes the users of the product, the parties who

ultimately pay the tax. Which of these two routes is chosen is, again, a matter of public choice.

This paper could end here with a conclusion to the effect “polluter pays or public pays,

the public must take its pick. Either approach can get society to a desired outcome, but someone

must pay to get to this outcome, and it is a public choice as to exactly who this is.” But we will

not end here. Rather, we will examine some additional argumentations. Both, if one accepts them

(one might not do so!), tend to lead to the conclusion that, Coasian reasoning notwithstanding,

the better path to achieving the desired outcome is “polluter pays”.

3. Fiscal Illusion

An argumentation made by those who oppose the implicit granting of ownership rights to

what otherwise are considered “public goods” (e.g., clean air, the example that is used

throughout this essay) to suppliers is as follows: if diminishment of value of an investment

resulting from a regulatory action by a government requires compensation to the investor by the

government, the effect of this requirement will be to reduce significantly the willingness of

governments to pass and enforce needed environmental and health regulations. What we wish to

explore here is whether this argumentation has any economic validity.

The argumentation is, in effect, an appeal to the idea of “fiscal illusion”. Boiled to its

essence, fiscal illusion occurs when a government evaluates whether to pursue an action on the

basis of budget cost to the government or on the basis of whether the action will create net social

benefit. A very real example where fiscal illusion is, in fact, official policy is to be found in US

trade policy. Under the Gramm-Rudman Act, any law or measure that affects the revenues or

expenses to the US government must be either revenue neutral or, if the law reduces revenue or

21

increases expenses, must make provisions to offset these gains or losses. This applies to tariff

reduction, which almost always will result in revenue reduction (although the case might be

found where elasticity of demand for a particular import is high enough that tariff reduction

actually causes demand to increase enough to cause total revenues to expand, this in practice

happens rarely if ever). But, by standard economic analysis, reduction of tariffs almost always

results in net social welfare gains for the United States, even after taking into account loss of

tariff revenue; in fact, this loss almost always is recaptured in the form of additional consumer

surplus, which is, as noted, a social benefit. Thus, the fact that US law requires that tariff

reductions be accompanied by other measures to recover lost tariff revenue does indicate that this

law embodies some element of fiscal illusion.

But, to be fair and balanced, it must be noted that fiscal illusion can cut two ways. For

example, in the United States, business/industry groups have long argued that some

environmental regulation has been passed that has created costs of compliance in excess of

benefits generated. In evaluating this regulation, or so it is argued, the government took account

of benefits but ignored costs (in effect, the government analysts would have taken into account

movement to the left along the PP curve in diagram 1 but ignored the AA curve) in deciding

what should be the amount of a particular pollutant allowed to be emitted. Whether or not this

argument is actually correct, the point here is that if the government did behave in this way, then

this is a form of fiscal illusion in the following sense: private costs are ignored but, had these

been government budgetary costs, they would have been accounted for. Fiscal illusion therefore

does not always necessarily always lead to under-regulation; it can lead to over-regulation as

well. If such over-regulation in fact does occur as a consequence of fiscal illusion, then

“socialization” of private costs via public compensation of private entities for costs these entities

22

must bear in order to comply with the regulation will, as per Coase’s theorem, lead to better

outcomes than will decisions made without such compensation. This is because, in effect, the

requirement for compensation implicitly does assign a property right for the public good to

private entities that use that good, such that government then must compensate them for what

amounts to a taking away of some of the right to use that public good. In determining the optimal

level of taking, if the government does face a requirement to compensate, it will balance social

benefits achieved by the taking against the (now public) costs of doing so. In principle at least,

the outcome should be something akin to attaining u* on diagram 1.

Having noted this last point, it does seem to the author that, in the current era in the

United States, government agencies at all levels are under pressure to avoid budget deficits that

might require tax increases (and hence every agency, except those whose mission has to do with

security, is under pressure not to do anything that increases expenditure). Also the costs of

regulation have been well- recognized by US politicians of all stripes, with the result that

momentum is toward deregulation rather than additional regulation. Given this, it seems

plausible that fiscal illusion of the sort that impedes governments from making expenditures that

are socially desirable is more likely than fiscal illusion that leads to over-regulation. (Admittedly,

this is a judgment by this author, and hence the author admits that he has violated the promise

made in the beginning to attempt to keep this whole analysis value-free to as great an extent as

possible.) If this is indeed so, the case for granting property rights for public goods to the public

rather than to suppliers, and hence not requiring public compensation to suppliers for regulatory

actions affecting public goods, would seem more compelling than the case for granting these

rights to the suppliers.

23

4. Considerations of Moral Hazard

“Moral hazard” has become a term with differing usage, even within economics.

Originally the term was associated with hidden (and often illegal) contracts that could have

adverse economic effects. The classical example is of the owner of a building who insures it for

greater than market value and then hires an arsonist to set it on fire to collect on the insurance.

(The hiring of the arsonist is the “hidden contract”, and the adverse outcome is of course the loss

of the building combined with the overpayment to the owner.) In contemporary times, the term

moral hazard is also used to depict a situation where a government (or supranational agency such

as the International Monetary Fund) commits to cover a potential loss, where that commitment

conveys some element of subsidy16, leads to an adverse outcome, e.g., a taking of an otherwise

unacceptable risk by the insured party knowing that, if the outcome is unfavorable, the loss will

be covered. Thus, for example, some economists believe that government-subsidized bank-

deposit insurance leads to moral hazard (banks will make imprudently risky investments) or even

that IMF stand-by credit to countries experiencing international financial crises will lead these

countries to do things that increase unacceptably the risk that such a crisis will occur.

How does “moral hazard” figure in the issues that have been discussed thus far? Suppose

that the property right to a public good is granted to a firm that is about to commence making a

product under the following circumstances: (1) the firm must decide how much to invest in

capacity for this product; (2) the manufacture of the product requires use of the public good

which, for the moment is unrestricted; however, there is some chance that for public health

reasons, this use will in the future be restricted. The claim to be made here is that the requirement 16 Including government-provided insurance where the price of the insurance is below what the market would charge in like circumstances.

24

for compensation created by the property right in fact leads to moral hazard in the following

sense: the firm will invest in more capacity than is prudent, or at least more than it would if the

compensation were absent.

This claim is made by way of a simple example. Suppose the cost of the capacity is given

by T(q), where q is the output when the capacity is fully utilized. To keep things as simple as

possible, let us assume what in the industrial economics literature is called a “two-period” model,

where in the first period capacity is set and then in the second period the product is produced and

sold; in other words, there is no third, or fourth, or subsequent period. In the second period, then,

the firm sells to a market where total demand is given by D(q) at a price P(q), such that D(q) is

simply equal to q (i.e., the firm is able to sell all of its output) but where P(q) diminishes to q (the

more output offered on the market, the lower the price—this is of course a standard assumption

regarding demand). Let us further assume, in the name of simplicity, that the only cost incurred

by the firm is that of creation of capacity T(q), where this cost of course increases as q increases

but at a constant rate (there are constant returns to scale). This last assumption is not, of course,

highly realistic; we could make a more realistic assumption, e.g., that there is a constant marginal

cost of output, but in fact to do so only complicates the calculations that follow while altering

nothing in terms of the result we obtain.

Absent the risk that the product will be subject to some form of regulation, the firm can

then maximize profits as follows: total profits are equal to revenue minus cost or, letting π

indicate profits,

)()( qTqqp −⋅=π .

Profits are maximized when the first derivative of π with respect to q is set to zero, or

25

0)(

)()(

=−+=dq

qdTqp

dqqdp

qdqdπ

implying of course that )()()(

qpdq

qdpq

dqqdT

+= , which simply is the standard condition that

marginal revenue must be equal to marginal cost (in this case, however, the marginal cost of

additional capacity) in order for profits to be maximized. The solution of this equation yields the

capacity that would be installed by the firm in order to maximize profits. Given the assumption

of constant returns to scale, this last equation can be restated as PdqdP

qC +⋅= , where C is the

unit cost of capacity and the explicit dependence of P on Q is dropped (although this dependence

is still there).

But, would the firm install the same capacity if there were some chance that, for reasons

of health, the output would be regulated in the future? The answer is no: suppose that, if it is

determined that the output of the product produced a health risk, this output were to be banned,

and that the firm assesses the probability of this happening to be p. (We assume that the

regulation is an outright ban again to keep things as simple as possible.) Given this, the

probability is (1-p) that the firm can produce the product but, with probability p, the firm will

lose entirely any investment it had made in capacity to produce the product. If the firm has risk-

neutral preferences, it is willing to make decisions based on expected value of outcomes. In this

case, the expected profit is given by the following

TPqpTpTPqp −⋅−=⋅−−⋅−= )1()()1(π .

In other words, the expected revenue falls from P⋅q to (1 - p)⋅P⋅q , where the latter is less than the

former because 0 < p < 1. But the expected costs at any given level of output remain the same

26

(these are incurred whether or not the production of the product is banned). Given that expected

revenue at any level of output is now less than before, the expected marginal revenue at any level

of output is also less than before. Given that the marginal cost of additional output is constant, to

achieve equality of marginal cost to marginal revenue, the firm would now reduce capacity from

that which maximizes profits where there is no chance of regulation being imposed.

Furthermore, this is a socially optimal decision.

But what if the firm could expect compensation for any expenditure on capacity if the

product were to be regulated (in our example, banned)? In this instance, there would be no net

cost to the firm if the product were banned. The expected profit would then be

))1()1(0)()1( TpPqppTPqp −−⋅−=⋅−−⋅−=π ; in effect, expected marginal cost would be

reduced from T to (1 - p)⋅T, and the firm would expand capacity over that which it would have

set had there been no requirement for compensation.

This is a non-optimal result because the government is in effect subsidizing the firm to

take more risk than it should from the point of view of social optimality.

5. Conclusions

From the point of view of economics, the issue of what to do to reduce the stress on the

environment created by economic activity largely boils down to how to deal with external costs,

i.e., ones that are borne by society at large rather than by whatever economic agents are involved

in transactions that give rise to these costs. The further issue that then arises is how to correct for

the market failure that is created by these costs. Coase’s theorem is very often invoked in this

latter matter; to repeat the earlier analysis, the essence of this theorem is that the market failure

can be corrected by assigning a property right to the underlying public good either to the user of

27

this good (the “polluter” in the case where the public good is the air) or the public at large, such

that a market can then operate whereby the polluter can buy rights for use of the public good or

the public can buy rights for non-use of it, depending upon who holds the rights initially. The

two different assignments have differing implications for who bears the cost of the action

necessary to achieve the optimal outcome; but to whom the costs are assigned is a matter of

social choice and does not affect an efficient solution of the market failure.

Under this reasoning, if NAFTA tribunals in dispute settlement cases where the dispute is

over the diminishment of value of an investment resulting from imposition of a governmental

regulation, side with plaintiffs, the effect is essentially equivalent to the assigning of a property

right to a public good to the investor.17 Under Coase’s reasoning, such an awarding does not

necessarily lead to a bad outcome from a social cost/benefit perspective; rather, it could lead to

an optimal outcome. What such an awarding does is to assign the costs associated with achieving

this outcome to the public.

However, issues of fiscal illusion and moral hazard do call into question whether this

assigning of a property right to the user of the public good is as likely to lead to a desirable

outcome as Coase’s reasoning would suggest. Rather, the possibility looms large that such an

assignment will lead to a non-optimal outcome. To be sure, the risk also looms, albeit in this

author’s judgment a lower risk, that fiscal illusion could also negatively affect the outcome of

assigning a property right to the public good to the general public. However, the same does not

seem to be true for moral hazard; if moral hazard is a factor, it only negatively affects the

outcome if a property right to a public good is assigned to the user, and not if this right is

17 If the outcome of NAFTA chapter 11 dispute settlement decisions were to be a dual system, whereby property rights to public goods are assigned to investors if these investors are foreign but to the public if the investors are domestic, this duality might in itself create some problems with respect to efficient solution of market failure created by external costs. This possibility is not explored here.

28

assigned to the public (the author did try to think of an instance where assignment of the right to

the public would create a problem associated with moral hazard, but came up empty-handed). On

balance, if either fiscal illusion or moral hazard exerts itself, the balance is tilted towards

assigning property rights to the public and not to the user of the public good, in the sense that the

former assignment seems likely to result in a more desirable outcome than the latter.

It must be stressed that this result is independent of the issue of who pays to achieve a

socially desired environmental goal, an issue that is, as stressed before, one of public choice. The

argument here for assignment of the property right is, in other words, based on considerations of

efficiency, not equity. But, if indeed considerations of equity are considered as valid, and if

society deems that it is in fact more equitable that “the polluter pays” than that “the public pays,”

then the result here is aligned properly with this choice.18

18 Again, it must be stressed that this does not imply that governments would not have to compensate investors for expropriations or measures that truly were tantamount to expropriations. See footnote 3 above.

29

BIBLIOGRAPHY

Ronald Coase. 1960. “The Problem of Social Cost.” Journal of Law and Economics 3: 1-44. A. D. Ellerman, P. L. Joskow, R. Schmalansee, J. P. Montero, and E. M. Bailey. 2000. Market for Clean Air: the U. S. Acid Rain Program. Cambridge, UK, and New York: Cambridge University Press. Howard Mann and Konrad von Moltke. 2002. “Protecting Investor Rights and the Public Good: Assessing NAFTA’s Chapter 11”, International Institute for Sustained Development. Paper prepared for conferences on NAFTA chapter 11 held in Mexico (March 13), Ottawa (March 18), and Washington, DC (April 11).

1

Protecting Investor Rights and the Public Good: Assessing NAFTA’s Chapter 11

By

Howard Mann Associate at International Institute for Sustainable Development

and

Konrad von Moltke

Senior Fellow at International Institute for Sustainable Development

Background Paper to the 2002 ILSD Tri-National Policy Workshops Mexico City, March 13; Ottawa, March 18; and

Washington (at the Institute for International Economics), April 11 http://www.iisd.org/trade/ILSDWorkshop

2

1. INTRODUCTION: FROM INVESTOR PROTECTION TO CHAPTER 11 The controversy surrounding NAFTA Chapter 11 might lead one to think that international foreign investment law is brand new. This is not the case. Rather, it has been developing as an increasingly specialized area of international law for over half a century. Rooted in the philosophy that aliens and their property were subject to the protection of their sovereign states under international law, cases on expropriation began to appear in the 1930s and saw significant growth between the 1950s and 1970s.1 Since these early roots, investment law has grown in scope. NAFTA’s Chapter 11 provides the quintessential model of the key elements that have now found their way into an ever-increasing number of investment agreements:2

• Protection from direct and indirect expropriation; • Rights to establish an investment; • Rights to national and most favored nation treatment; • Guaranteed minimum international standards of treatment for foreign investors; • Special protections against performance and personnel requirements; and • Rights to repatriate all monies.

In addition to this move to broaden investment rights, the ability to enforce these rights has also expanded. Where such enforcement used to take place only between states, and was thus very rare, now foreign investors have the ability to enforce their international rights themselves through the “investor-state” arbitration process. This development reflects the reality that states do not invest in foreign countries, so that international investment law deals with individual investor rights in relation to public authorities of the host country. Accompanying this remedy are “choice of forum” provisions allowing an investor to choose international remedies in place of host country domestic remedies. As the rights and remedies have expanded, so has the willingness of foreign investors to use these tools. Where state-to-state remedies used to be a last resort to enforce an investor’s rights, the investor-state arbitrations have now become a preferred venue.3 Protecting the environment and promoting sustainable development typically require government action. They also entail a high degree of uncertainty concerning the most effective path of action, and consequent ly involve a significant exercise of discretion on the part of public authorities. In most developed countries, environmental law has become the largest body of law; in all countries, implementation of environmental mandates is subject to controversy and litigation as societies seek to balance the need for environmental protection and the reduction of poverty against the prospects for individuals to prosper. All of these areas of public law are potentially affected by international investment law. It is hardly surprising that the environment has become the

1 Sornajah, M., The International Law on Foreign Investment, Cambridge University Press, 1994. 2 By the end of 2001, the number of bilateral or multilateral investment agreements exceeded 2000. 3 Luke Eric Peterson, "Changing Investment Litigation, BIT by BIT", Bridges Between Trade and Sustainable Development, May 2001, Year 5, No.4, pp 11-12.

3

stalking horse for the broader public debate about the impact of expanded international investment law. This debate has advanced rapidly from the consideration of specific environmental issues to a discussion of investment and sustainable development to a reconsideration of some of the premises of international investment law. The concerns raised in this paper can be divided into three related themes:

• The impact on promoting sustainable development; • The impact on the ability of governments to make necessary and effective

governance decisions, including but going beyond sustainability issues; and • The appropriateness of the institutional framework of NAFTA.

Discussion of the first theme begins with the recognition that investment is vital to promoting more sustainable development. But it revolves around a critical assessment of the foundations of current international investment law, which are largely derived from trade liberalization theory and practice. The second theme has two streams: the case law that has developed under Chapter 11 and the impacts on the investor-state process as a model for international dispute resolution. The third dimension considers whether the institutional framework of NAFTA and in particular of the Chapter 11 dispute settlement provisions are adequate to undertake a legitimate balancing of investor rights and public goods. The paper will conclude with a discussion of options to address the identified problems. These options are developed first within the NAFTA Chapter 11 context, but are then expanded to serve as a guidepost on investment issues for negotiations in the FTAA, WTO and elsewhere. 2. THE QUESTIONABLE FOUNDATIONS OF TODAY’S INVESTMENT AGREEMENTS The benefits of economic liberalization appear so obvious to those who promote it that negotiations are generally launched without much analysis. In part this is due to the difficulties that are always encountered when new negotiations are launched; the struggle to construct an agenda acceptable to all overrides discussion of whether the agenda that emerges is the “right” one by some agreed standard. This section goes back to that fundamental discussion, pursuing the seldom-asked question: what objectives should we be pursuing in constructing investment agreements? Three years after the conclusion of NAFTA, its approach to investor rights and remedies became the foundation for the OECD negotiations on the Multilateral Agreement on Investment (MAI)--negotiations that were aborted in 1998. More recently, Chapter 11 has provided the template for the initial stages of the investment negotiations in the Free Trade Area of the Americas process.

4

What rationale underlies the broadening of international investment law – a broadening that began over fifty years ago, and has culminated in NAFTA4? The question is especially salient given the history of the linkages between trade liberalization and investment liberalization. Both were part of the initial efforts to establish the International Trade Organization as part of the reordering of international economic institutions launched at the 1944 Bretton Woods Conference. The Havana Charter that followed in 1948 never entered into force for want of ratification. That left only the trade liberalization process of the GATT in place.5 As tariff barriers were removed under successive rounds of GATT negotiation, trade negotiators came to focus their efforts on other possible barriers to trade. In doing so, they found themselves drawn deeper and deeper into the details of domestic regulation. In turn their remit broadened to include all manner of new issues, from intellectual property rights to trade in services, and from investment to protection of the environment. This has been done through a new combination of traditional negative rules--states shall not discriminate or impose quotas--joined to a much deeper and broader set of positive obligations--requiring government measures to conform to disciplines that define how, and often when, governments may regulate. Given this broader scope of impact, we have now moved beyond the point where the promotion of public welfare associated with liberalization theory can be taken as a given. The dilemma is clearly identified by some trade theorists:

Under the assumption that markets are competitive, economic theory suggests that the reduction of border barriers such as tariffs will benefit both the importing and the exporting nation. Accordingly traditional trade agreements can be presumed to be win-win. But theory does not suggest that this will necessarily be the case for deeper agreements that deal with behind-the-border policies. Indeed, such agreements could well be win-lose. For example, an international agreement to enforce intellectual property rights could on balance harm a country that has little or no domestic innovation and has previously simply copied foreign innovations.6

The unquestioned extension of the rationale for trade liberalization to other areas of international economic governance must be examined critically. For fifty years, trade negotiators have fought the battle against special interests and the bastions of hidden protectionism. Covered by the principle of comparative advantage, they comfortably

4 For more on the full breadth of Chapter 11 of NAFTA see Howard Mann and Konrad von Moltke, NAFTA’s Chapter 11 and the Environment: Addressing the Impacts of the Investor-State Process on the Environment, International Institute for Sustainable Development, Working Paper, 1999. Hereinafter, Mann and von Moltke, 1999. 5 Konrad von Moltke, An International Investment Regime? Issues of Sustainability, International Institute for Sustainable Development, 2000, p. 9-10. 6 Nancy Birdsall and Robert Z. Lawrence, “Deep Integration and Trade Agreements: Good for Developing Countries?” in: Inge Kaul, et al., eds., Global Public Goods. International Cooperation in the 21st Century. New York: Oxford University Press, 1999, p.147.

5

assumed that their every effort promoted the public good, even when this inflicted hardship on some constituency. The assumption, in its most simplified form, was: imposing rules that prevent barriers to achieving market efficiency is a good thing to do. Gradually, this reasoning was extended from tariff to non-tariff issues, and into other areas such as intellectual property rights, services and investment. Today, this reasoning has been extended to support the expansion of foreign investors’ rights from protection of an investment after it is made to include the right to make investments (“right of establishment”), and rights precluding the imposition of national economic policies (such as local purchasing or hiring requirements) to the operation of the investment. Now, however, faced with increasing proposals to liberalize international markets that implicate significant behind-the-border policies, negotiators will have to justify their actions in terms of the broad public welfare that is being promoted. Their global impact as well as their differential impact on different countries must be assessed carefully so as to avoid win- lose situations, let alone lose- lose ones. The public welfare justification to expand traditional international investment law protections to private investment rights has yet to be provided. What ultimately tripped up the Multilateral Agreement on Investment was not just environmental opposition but the cumulative defects that emerged as the draft was submitted to broad public scrutiny as well as the fact that the public welfare benefits to be expected from the agreement were never properly articulated, if indeed they existed. In so far as a public welfare justification has been made, it has focused on the role of the protection of investors under international law as a means to expand and diversify foreign investment into more states, especially developing states. Whether this objective has ever been achieved is, however, open to some doubt.7 The empirical evidence that does exist suggests that the risk reduction element associated with traditional investment protection is at best a marginal factor in business decision-making on FDI. For lasting changes in the risk perceptions of investors a number of factors must come together. Strong domestic institutions rank high, as do resources in all areas that significantly affect an investment, and access to markets for the goods produced. Risk perception will deeply influence the rates of return that an investor requires, but risk itself is a many-faceted phenomenon and the creation of an international framework for investment will not by itself change investment flows. 8

7 K. Scott Gudgeon, “United States Bilateral Investment Treaties: Comments on their Origin, Purposes and General Treatment Standards” International Tax and Business Lawyer, Vol.4, 1986, at pp. 111-112; Thomas Walde and Stephen Dow, “Treaties and Regulatory Risk in Infrastructure Investment”, 34 Journal of World Trade, Vol.2, (2000); A. Perry, “Effective Legal Systems and Foreign Direct Investment: In Search of the Evidence”, International and Comparative Law Quarterly, Vol. 49, pp. 779-799, 2000. 8 Factors such as available resources, educated workforce, market potential in the host country/region, political stability, banking, administrative and physical infrastructures, etc., play a much larger role in these decisions. Indeed, up until the beginning of the Chapter 11 litigation, few investors appear to have even known of the web of bilateral agreements in this field. This appears to be confirmed in recent UNECLAC and UNCTAD studies, where risk management is clearly identified as one, but only one, investment factor. While lawyers focus their advice on risk and remedies, this does not make it the principal focus of the business investor itself. Foreign Investment Flows in Latin America, 1999, United Nations Economic

6

Mexico has seen significant growth in foreign direct investment since NAFTA came into force in 1994. But the additional investment flows to Mexico following NAFTA are the result of the entire agreement, including the critical rules of origin provisions that allow market access to the United States, together with supporting domestic measures and institutional changes in Mexico. It is impossible to determine whether NAFTA Chapter 11 had a positive role in promoting investment into Mexico, just as it is impossible to determine the opposite. Because of the complexities surrounding investment decisions the easy generalizations that have supported trade liberalization do not apply to investment agreements. Investment law initially developed to provide an international law route to obtain redress for egregious governmental actions relating to foreign property. The basic rationale was to overcome deficiencies in national legal regimes as they relate to foreign private capital by supplementing domestic regimes with an international law backstop. This rationale was established in the context of developed country investors and developing country host states during an era when nationalization was occurring in many countries, in particular former colonies or dependent territories. Recent investment agreements have expanded far beyond this limited purpose and can now impact on any regulatory decisions that investors may consider undesirable. In addition, the supplemental role of investment agreements has now become one of substitution: investors can choose one forum and set of rules over the other, as circumstances suit them. Moreover, investor protections have, as already noted, been expanded to include rights of establishment and to restrict locally imposed performance requirements, even those applicable to domestic investors. Thus, investment agreements now create a series of international law economic rights for private actors, enforceable as a matter of international law under international processes. Again, these new rights seek to overcome domestic law deficiencies or to actually override domestic law barriers, now substituting the domestic regime completely with international law rules. These newer international investment rights, combined with ongoing unilateral investment liberalization policies in many countries, do create new investment opportunities. It is well understood that the same regimes that may be deficient in the protection of foreign capital are also likely to be deficient in the protection of public welfare concerns: environment, health, resource management, and others. Yet, investment regimes to date have not included any legally binding provisions designed to address these deficiencies by supplementing or substituting the domestic regimes with international law obligations for the investors. This is despite the fact the agreements are knowingly and purposefully designed to promote investment into such countries, and that investment in the absence of such protections can lead to net welfare losses. Promoters of investment agreements argue that the deals do not give license to ignore local laws, so international protections are not necessary. To some extent this proposition

Commission for Latin America and the Caribbean, Santiago, 2000; World Investment Report 2000, United Nations Conference on Trade and Development , Geneva, 2000.

7

is even questionable, given the possible scope of certain provisions in investment agreements. (See section III, below.) But assuming it to be true, arguendo, this still does not answer the concern that the regimes in question are often underdeveloped and may lack effective enforcement. This is no different in kind than the concerns raised for the deficiencies in legal, administrative and judicial processes that exist in these countries in relation to the protection of capital. Simply put, one is deemed to require supplementing and substituting, the other is not. Studies by the World Bank and others are now documenting the results of the expansion of private economic rights and the exclusion of public welfare protections. They are beginning to identify trends towards net economic losses to host countries as a result of poorly controlled development sparked by investment liberalization in the absence of appropriate public welfare management regimes and processes. Environmentalists need to be careful in this area. We now know that “pollution havens”-- in the conventional sense of jurisdictions that actively seek to attract investment by offering low environmental standards--have not developed to the extent many feared. Investment decisions involve numerous factors, and environmental factors are rarely dominant for productive investments that do not rely directly on the environment for their product—even in the handful of industries, essentially petrochemicals, where environmental protection now represents as much as 10 percent of total investment.9 At the same time, however, there is a growing body of literature and empirical evidence that demonstrates a “stuck at the bottom” phenomenon for environmental protection: low levels of environmental protection and enforcement are not increased as needs increase.10 This observation is based on several case studies in highly liberalized developing countries. It clearly shows the risk of leading to a situation where short-term economic growth has been transformed into medium or long term net losses due to the environmental and human costs of environmental mismanagement. In short, investment liberalization in countries with lower environmental standards may dissipate any

9 The evidence against pollution haven investment patterns is compelling for major global corporations that generate and use state of the art technologies. The evidence is less empirically clear, however, in relation to smaller foreign investors, where data simply is less complete. The evidence of sustainable performance in relation to renewable resource industries such as forestry and fisheries is deeply disturbing. See for example the series of World Bank evaluation reports of forest regimes: Uma Lele, et al., The World Bank Forestry Strategy. Striking the Right Balance. Washington, D.C. The World Bank, 2000 (World Bank Operations Evaluation Department. B. Essama-Nash and James J. Gockowski, Cameroon. Forest Sector Development in a Difficult Political Economy . (Evaluation Country Case Study Series). Washington, D.C.: The World Bank, 2000. (World Bank Operations Evaluation Department). Nigel Sizer and Dominiek Plouvier, Increased Investment and Trade by Transnational Logging Companies in Africa, the Caribbean and the Pacific: Implications for the Sustainable Management and Conservation of Tropical Forests. Brussels: World Wide Fund for Nature, 2000 10 Ibid, and see Gareth Porter, “Trade Competition and Pollution Standards: “Race to the Bottom” or “Stuck at the Bottom”?, Journal of Environment and Development, Vol. 8, pp. 133150, 1999; Luba Zarsky, “Havens, Halos, and Spaghetti: Untangling the Evidence about Foreign Direct Investment and the Environment” In Foreign Direct Investment and the Environment . Organization for Economic and Cooperation and Development, Paris: OECD, 1999; Thomas, V. et al., The Quality of Growth. Washington, DC: World Bank and Oxford University Press, 2000; Per Fredriksson, ed., Trade, Global Policy, and the Environment, Washington, DC: World Bank, 1999 .

8

economic gains through insufficient attention to the sustainable development dimension of economic growth. In addition to the growing environmental concern, there are several other problems in the current pattern of investment from the international and public welfare perspectives. Some countries that desperately need investment are unable to attract it in significant amounts. Some sectors that require investment to promote public welfare, environmental protection among them, do not get additional investment at the level that is required. Yet there is nothing in existing investment agreements to suggest that they are designed to rectify such imbalances. They either assume that the act of liberalizing international markets for investment will somehow automatically contribute to solving these problems, or they assume that addressing such problems is beyond their remit. The above is not to argue that there is no international agenda for investment. Rather, it represents a powerful argument that a different international agenda for investment must be set. There are certain areas of public policy—environment and sustainable development most prominent among them—that require private investments to achieve well- identified public goods. Governments have long recognized the need to promote such investments, providing subsidies and many other incentives to influence investment decisions. The reduction of risk is one tool among others to support private investments that promote public goods and it has the advantage of not creating distortions that become so internalized into expectations that they cannot be removed. It is essential to identify the public goods that should be promoted through international investment agreements. If these cannot be achieved, governments may want to reassess their willingness to invest negotiating resources and limit their rights to regulate foreign investors or pursue their development objectives The importance of identifying the public good is twofold: it provides a standard against which the provisions of an investment agreement can be measured; and it provides a basis for assessing whether the benefits of an investment agreement justify the costs. Specific options in this regard will be returned to in Section 5 of this paper. 3. THE APPLICATION OF THE KEY PROVISIONS OF CHAPTER 11 In the introduction to this paper, two related categories of concern were noted in the relationship between investment law and sustainable development:

• The impact on promoting sustainable development; and • The impact on the ability of governments to make necessary and effective

governance decisions, including but going beyond sustainability issues. The NAFTA includes three preambular statements that are relevant to these concerns, affirming that the NAFTA partners are resolved to:

• Preserve their flexibility to safeguard the public welfare

9

• Promote sustainable development; • Strengthen the development and enforcement of environmental laws and

regulations.11 In practice these preambular statements have had little actual relevance to the design, interpretation and application of NAFTA’s Chapter 11. The promotion of sustainable development in NAFTA is largely based on the assumption that economic growth is good and will in time lead to the correction of any environmental problems it may create. As seen above, rather than enunciating explicit criteria to promote sustainable investments, Chapter 11 relies on the private welfare rationales for investor protections and investor rights. The remaining two preambular paragraphs relate directly to the ability of governments to make necessary and effective governance decisions to protect the environment. While apparently designed to ensure the ability of governments to take necessary decisions to protect the environment as a public good, these preambular paragraphs have not affected the interpretation of the investor protection provisions of Chapter 11. The result is a growing sense that the provisions of Chapter 11 put at risk the ability of governments to maintain a dynamic environmental management system that can anticipate and forestall environmental and human health damages, as well as ensure redress when such damage occurs. Thus, it is argued, a critical pre-condition identified by both the World Bank and the World Trade Organization for liberalization to contribute positively to sustainable development—the presence of a dynamic and effective environmental management system 12—is being jeopardized by the operation of Chapter 11. It is to this issue that we turn now. Chapter 11 contains four key provisions, or disciplines, that relate to the capacity of governments to effectively exercise their environmental management functions:

• National treatment and most favored nation treatment (nondiscrimination), Article 1102, 1103;

• Minimum international standards of treatment, Article 1105; • Prohibitions on performance requirements, Article 1106; and • Expropriations, Article 1110.

In addition, the character and legitimacy of the institutional provisions for dispute resolution are of critical importance. These are discussed separately in the next section of this paper. Before looking at each discipline briefly, 13 it is worth noting that the Chapter 11 cases to date have made it clear that each of these provisions applies to all government measures, 11 NAFTA, Preamble, paragraphs 12-14. 12 Fredrickson, supra , n. 9; H. Nordström and S.Vaughan, Trade and Environment: Special Studies 4, World Trade Organization, Geneva, 1999. 13 More extensive discussions of these provisions can be found in Mann and Von Moltke, 1999 and Howard Mann, Private Rights, Public Problems: A Guide to NAFTA’s Controversial Chapter on Investor Rights, International Institute for Sustainable Development/WWF, 2001. Recent literature taking different views

10

whether they are taken specifically in the context of an individual investment or are measures of general application that apply equally to foreign and domestic investors. Moreover, the range of measures covered is extremely broad, including laws, regulations, administrative decisions on licenses or permits, policies with a direct impact on businesses, or other possible government actions 14. The full range applies without exception to actions taken by local, state or provincial governments, as well as national governments. For federal actions, measures undertaken before Chapter 11 came into force are covered in terms of their continued application. 15 In addition, Chapter 11 includes extensive rights of establishment in the three Parties.16 As a result, the disciplines noted above apply both before and after an investment is made (to the extent the right of establishment exists or an investor otherwise is permitted to make the investment in question), ensuring that in the course of exercising the right of establishment a foreign investor is not discriminated against or otherwise deprived of any protections. The discussion of the disciplines below should be understood as applying to the decision-making and building phase of investments as well as their subsequent operation. A second general point of note is that investments and investors are defined about as broadly as possible, covering not just foreign owned production or resource harvesting facilities but also sales offices, portfolio investment, and other nonproductive forms of business or financial activity in the host states. As a consequence, it is difficult to establish with any certainty the minimum elements necessary to qualify as an investment. 3.1 . National treatment

A key objective of investor protections is to preclude discriminatory treatment against investors based on their country of origin. Yet, several key aspects of this discipline remain ill defined. For example, the threshold test of what type or extent of differences in treatment will trigger a breach of this article is unclear: Is it any difference, or a difference that has a significant impact on a business? Must it be a difference in law (de jure) or simply a difference in effect (de facto)? A related question is what standard of treatment is to be chosen against which to compare the investor: the lowest applicable in a jurisdiction, the highest, or some middle ground? How does fairness in an administrative process relate to the assessment of outcomes from that process, or is it only the outcomes that matter, for example in an environmental assessment or permitting process that requires a variety of inputs to be assessed? to the above can be found in e.g., Patrick Dumberry, “The NAFTA Investment Dispute Settlement Mechanism: A Review of the Latest Case-Law”, Journal of World Investment, Vol. 2, pp. 151-195, 2001; and Daniel Price, “NAFTA Chapter 11 – Investor-State Dispute Settlement: Frankenstein or Safety Valve”, Canada-United States Law Journal , Supplement, Vol. 26, pages 1-9, 2001. 14 There is only one exception to the definition of covered measures, namely those concerning interest rates for purposes of macroeconomic policy. In accordance with standard principles of treaty interpretation, by including a single exception the text effectively affirms that no other measure has been exempted. 15 Provincial and municipal measures taken before NAFTA came into force have now been grandfathered. However, new decisions under enabling laws or regulations would be covered as new measures. 16 Under NAFTA, the three countries had to list the sectors they wanted excluded from the right of establishment. In some cases these are fairly extensive, such as the energy sector in Mexico. Any sector not listed is covered by Chapter 11 and is open to investment by investors of the other two NAFTA parties without discrimination.

11

In the language of Article 1102, two main issues must be addressed to answer any of these questions: is the investor treated less favorably than another investor, and which investors are “in like circumstances” for the purposes of making this comparison? It is only when investors that are in like circumstances that they must be treated no less favorably one than the other. The cases are just beginning to interpret what these terms mean and how they should be applied. Although the existing case law is not consistent on this issue, context-sensitive analysis will, we believe, be essential to determine whether investors are in like circumstances with others they claim form an appropriate reference group. The absence of a context-sensitive approach means that there is no basis for including or excluding reasons for differences in treatment of investors in an analysis of outcomes. Some sample questions may help illustrate this concern. Would a factory that has to meet newer and tougher air emission standards than a preexisting factory be in like circumstances to a factory in the same area set up some years earlier? Is a factory located in a watershed that is heavily used in like circumstances to one where water is clean abundant and organic emissions can be readily absorbed? Is a lumber company that locates adjacent to a national park in like circumstances to a similar company located in an area with no such environmentally sensitive sites? Or is an investor who uses a hazardous component in his manufacturing process in like circumstances to an investor using a non-hazardous product towards a similar end? And would an investor with no other assets in a country be discriminated against if required to post a bond for environmental damage or abandoned site insurance purposes, when a company with millions in other domestic assets is not required to do so? All these types of situations could lead to differences in treatment, and some might even prevent entry into a market. Domestic institutions operate under rules that give guidance on such questions, or according to procedures that permit their open and accountable consideration. It seems likely that an international analysis without comparable resources faces almost insurmountable obstacles. We believe that the “in like circumstances” analysis requires the articulation of criteria to make legitimate distinctions, and the use of procedures that ensure accountability. Trade law has also struggled to deal with the added complications that have attached to concepts that seemed only pleasantly ambiguous before trade law went behind the border. The concept of “like product” is central to the operation of the central principle of “non-discrimination” in trade law. Trade law cases addressing “like products” have established a series of factors that focus primarily on the commercial substitutability of products. Only recently, in the Asbestos case, has WTO case law given an indication that the environmental or human health impacts of a product may be a relevant factor for direct or indirect consideration in this regard, either as a direct factor in assessing likeness or indirectly as a factor in assessing aspects of the commercial substitutability test.17 What is 17 European Communities--Measures Affecting Asbestos and Asbestos Containing Products, Report of the Appellate Body, WTO, AB -2000-11, March 2001. The extent to which environmental or human health factors can be brought to bear remains a somewhat open question, with a minority opinion being specifically addressed to this point in the Appellate Body judgment in the case. The differences between investment and trade, however, make it critical for such factors to be brought to bear in the investment analysis of like circumstances.

12

merely complex in trade law appears almost insoluble in investment law, because the issues are so different, as pointed out above. The notion of “likeness”, however expressed, is equally central to investment law. But because investments involve numerous independent factors, and in particular a large time factor that is essentially absent in trade law, the interpretative task is orders of magnitude more difficult. There is nothing in NAFTA Chapter 11 that suggests that negotiators were particularly aware of these difficulties, and so there is nothing that is of much help to those who now struggle to implement it. From a sustainable development perspective, relevant factors in establishing “likeness” would include suitability of an investment to a receiving community and physical environment, the availability of sufficient resources and infrastructure for its operation, applicable legal processes and structures, and other factors relating to its establishment and operation. To date, the cases suggest a context sensitive analysis can and should be applied. Ensuring this is critical, as Chapter 11 has no associated general exclusion such as that found in Article XX of the GATT, which creates a framework for exploring government acts even when they appear to contravene core principles of the agreement.18 If this is not done in the context of defining like circumstances, it is unlikely it can be done effectively at all19. The second key interpretational issue under Article 1102 is the meaning of “no less favorable” as the standard for treatment for investors that are in like circumstances. It is well understood from its trade law origins that this does not mean identical treatment. But exactly what it does mean, or what differences in de jure or de facto treatment are acceptable is much less clear. Minor or inconsequential differences are unlikely to found a case,20 but how significant the differences must be in law or in effect is not clear. And, as already noted, the proper standard for comparison is most uncertain: is it the highest standard, the lowest, some middle ground? And in some cases, for example the application of environmental assessment procedures, is the treatment looked at to be process related or result s related, or necessarily both? There is little guidance at present on these issues in the cases. While the cases send a positive signal on this issue, and are insufficiently developed on the basis for comparing treatments, they are very troubling on another “like circumstances” issue. One would have thought that comparing like circumstances would involve comparing similar operations located in the host country. In S.D. Myers v. 18 Tim Burke, “Importing the ‘Aims and Effects’ Test into NAFTA Article 1102: S.D. Myers, Inc. v. Government of Canada”, Student term paper, Fall, 2001, University of Ottawa, Faculty of Law, on file with authors. 19 In the final phases of the MAI negotiations the chair proposed a footnote that would identify some environmental considerations that could be taken into account. This footnote points in the right direction without fully responding to the need that has been identified here. “Engering Paper on Labor and Environment”, reprinted in Inside US Trade, March 27, 1998, .p. 17-19. 20 In The Matter of an Arbitration Under Chapter Eleven of the North American Free Trade Agreement Between Pope & Talbot Inc. and the Government of Canada, Interim Award by Arbitral Tribunal, June 26, 2000. And see the comments on this issue in Anthony van Duzer, NAFTA Chapter 11 To Date: The Progress of a Work in Progress”, Centre for Trade Policy and Law, NAFTA Chapter 11 Conference, January 18, 2002, draft paper available at www.carleton.ca/ctpl, final paper to be published by CTPL.

13

Canada, however, this was not done. The investment in Canada by US based Myers21 (S.D. Myers Canada) operated as a waste broker service – a middleman that wanted to export waste generated in Canada to its parent company’s U.S.-based waste disposal operations. The Tribunal, however, went beyond comparing the investment located in Canada with the many other Canadian-based waste broker services, and asked whether the investor’s waste disposal operations located in the US were receiving less favorable treatment than that accorded to similar Canadian waste-disposal operations. In this way, the national treatment obligation was applied to the integrated business line of the investor including those elements that were not located in Canada.22 This has some appeal from a pure business perspective, but represents a remarkable extension of the scope of the “national” treatment requirement. Moreover, it is an extension that would override a true context-sensitive analysis in the host state by superimposing a business-focused context over all other issues. A final factor that should be noted here is how broadly an investment or investor is defined. It is clear that no actual productive process or facility is now required for protections and rights to be vested in a foreign investor. In S.D. Myers v. Canada, a sales office was held to be an investment under Chapter 11. This may have been a correct interpretation of the language of Chapter 11, but it must be asked whether this is an appropriate approach. It reflects one of the means by which the concept of national treatment has been greatly expanded from previous investment agreements, and from the origins of investment protection to create the beginnings of a new system of international economic rights. The consequences are real: if a sales office constitutes an investment that allows a suit to be initiated, and if market share is a protected property right, as the S.D. Myers and Pope & Talbot cases found, then any measure impacting that market share becomes subject to challenge by any foreign company with a sales office located in the host country. This could amount to nothing less than a wholesale privatization of international trade law, and without any of the environmental, human health or other legitimate regulatory dimensions of government being expressly protected under these provisions.23 This might prove destabilizing for governments, particularly since regulatory requirements will often increase over the life of an investment. Whether this fear will be justified remains to be seen, but the broad legal outlines for it to materialize are now present in the existing case law. 3.2. Minimum international standards

Like most bilateral investment agreements, Chapter 11 contains provisions requiring host countries to treat foreign investors in a way that meets minimum international standards.

21 Whether it was actually S.D. Myers that made the investment or the individual owners of the company remains an outstanding question being considered in a judicial review process. The language used here is shorthand for the description of the investment, not the investor. We believe the Tribunal erred in ruling this distinction is not relevant for founding its jurisdiction in this case. 22 These issues are discussed at pages 52-59 of the S.D. Myers decision, In A NAFTA Arbitration Under UNCITRAL Arbitration Rules, S.D. Myers, Inc. v. Government of Canada, Partial Award, November 13, 2000. 23 There are such exceptions built in for other parts of NAFTA, Chapter 7 and 9 for example, as well as for the agreement as a whole by a reference to Article XX of the GATT. However, this cross-reference is made inapplicable to Chapter 11.

14

This requirement is expressed in very general language as “treatment in accordance with international law, including fair and equitable treatment and full protection and security.”24 Exactly what this means is not spelled out in NAFTA, or in other investment agreements. Still, when investment provisions were used only as a shield against egregious acts this created little controversy; it was understood that the intention was to provide a floor of minimum standards of fair treatment, regardless of whether domestic firms were being treated equally badly. But with the change in the use of the provisions from a defensive shield into a sword to attack government measures,25 the lack of precision simply invites new scope for claims under this discipline, often coming from different areas of law. Chapter 11’s Metalclad case highlighted the increasing significance of this obligation to public welfare law making. In this case, Mexico was found to violate the minimum international standards based on breaches of transparency requirements not expressed in Chapter 11 but elsewhere in NAFTA, a breach of Mexican law established by the Tribunal but not by a Mexican court, and a failure to be notified of a meeting after years of intense negotiations with all levels of government in the region in question. 26 The Tribunal summed up its findings by saying that Mexico failed to provide a transparent, predictable framework for business planning and investment, and demonstrated a lack of orderly process and timely disposition in relation to an investor. This extremely broad reading of minimum international standards requirements was groundbreaking in international investment law. However, this breadth was rejected in a judicial review of the case by a court in Vancouver, British Columbia.27 A narrower scope, based on a traditional customary international law interpretation of minimum international standards, has now been put forward by the Free Trade Commission in an interpretive statement under Article 1131(2) of NAFTA, a process discussed in more detail below. 28 Arguments that this statement is in fact an amendment rather than an interpretation have been made, however, and this is a live issue now before a Chapter 11 Tribunal. Initial assessments of the minimum standards obligation in Chapter 11 suggested it was not an area of major concern from a broader public welfare perspective.29 But this chain of events, and the current uncertainty they have generated, indicate the need to carefully define the nature of all the obligations in an investment agreement, rather than leave them undefined and hence open to broad interpretation with unanticipated consequences. Again, some questions highlight the need for greater government certainty: Can it really 24 Article 1105 of NAFTA. 25 See the discussion on this in Mann, Private Rights, Public Problems, supra , n.13, p. 16, and Peterson, supra, n. 3. 26 See paras. 75-100 of the Metalclad decision, Metalclad Corporation v. United Mexican States, Award, International Center for Settlement of Investment Disputes (Additional Facility), Case No. ARB (AF)/97/1, 30 August, 2000, reprinted in ICSID Review- Foreign Investment Law Journal, Spring 2001 27 United Mexican States v. Metalclad Corporation , 2001 BCSC 664, S.C.B.C May 2, 2001. 28 The August 1, 2001 statement can be found attached to a press release at http://www.dfait-maeci.gc.ca/nafta-alena/menu-e.asp Under Article 1131(2) of NAFTA, this statement will legally bind any arbitration body set up under Chapter 11 as it relates to Article 1105. 29 E.g. Mann and von Moltke, 1999.

15

be, for example, that governments have an obligation to correct poor legal advice received by an investor as found in the Metalclad case? Does one bureaucrat’s representation that a certain event will or will not take place under the law of another level of government bind the country at all levels of government? Is every government act, official’s conversation, and any written communication from any level of government official sufficient to found a breach of Article 1105, as the final award in the Pope & Talbot case would suggest? These possibilities suggest standards never before made applicable in domestic law or international law, which may or may not be ruled out by the July 31, 2001 interpretive statement.

3.3. Performance requirements

Article 1106 of NAFTA prohibits the NAFTA parties from imposing what are known as “performance requirements.” This means that states cannot impose certain rules on the operation of a foreign investment—irrespective of whether such rules are imposed on domestic investors. Article 1106 prohibits host governments from requiring an investor to:

• Export a given portion of production; • Use a given level of local inputs or services in business operations, or otherwise

show a preference for domestic goods or services; • Generate foreign exchange flows based on the firm’s levels of imports or exports; • Use or transfer certain technologies (with some exceptions); or • Employ specified types or levels of personnel.

Prohibitions on performance requirements are a classic example of the shift from investor protections to the creation of international economic rights for private companies. They aim to prevent a host government from imposing conditions on an investor that may limit its ability to achieve economic efficiency and profits. It was originally anticipated that this provision would apply only to measures specifically targeted at a foreign investor or its investment. Therefore, even though the provision covered all stages of the investment cycle--from initiating to operating to terminating the investment--it was thought that only a narrow range of measures would be captured. The early cases have shown otherwise. It is now clear that under Chapter 11 even non-discriminatory measures of general application (that is, measures not targeted at a specific investor or sector), both new and preexisting, can be considered to be performance requirements.30 Using this reasoning, it can and has been argued that an import ban on a product used by manufacturers is in effect a requirement to use local substitute products. The result is that foreign investors that might be affected by such a ban are able to bypass the traditional state-to-state process for challenging such trade measures--a process that has been the hallmark of the development of trade law in NAFTA and the WTO--and themselves directly challenge the measure. This issue was raised and accepted as a matter for proper ruling on the merits in the Ethyl v. Canada 30 S.D. Myers v. Canada, paras. 289-300; Pope & Talbot v. Canada, paras. 74; and in the Ethyl Corp. v. Canada, Award on Jurisdiction in the NAFTA/UNCITRAL Case between Ethyl Corporation and the Government of Canada, 24 June, 1998, paras. 62-64. There is no jurisprudence to the contrary.

16

case.31 The case was never heard on the merits, as it was subsequently settled. Nonetheless, it is now being raised again in a new case at the initial stage of a Notice of Intent to arbitrate brought by Crompton Corp. against Canada.32 While this reading of Article 1106 has not been seen yet in a final decision, the ingredients for it to emerge are in place, and this is disturbing. As in the possible combination of interpretations of investor, property right and expropriation (see below) that was discussed under the national treatment issue, this provision could now become a means for a foreign investor to challenge any measure impacting its operations as a prohibited performance requirement. Just the prospect that this can happen may seriously weaken the ability of governments to protect human health and the environment from the actions of foreign investors. Again, it is a case of the potential breadth of a provision not having been fully assessed during its drafting. 3.4. Expropriation

Chapter 11 sets out protections against expropriation that have now been expanded by the arbitration bodies to include an extreme reading of the American concept of “regulatory takings”: that a regulation impacting on the reasonable expectation of profit of a company constitutes an expropriation that requires compensation to be paid. The effect of this approach, if it is followed in other cases or other agreements, will be to require governments to pay foreign investors compensation if any regulation to protect the environment, human health, public welfare or community interests significantly interferes with their ability to make a profit from their investment. This interpretation of expropriation, which applies even if a regulation is completely non-discriminatory between any and all investors, was never seen before in international law, as recently acknowledged by a Canadian court that reviewed, but upheld, one such arbitral decision. 33 The Chapter 11 provisions on expropriation have been the most debated issue concerning the relation of investor protections to environmental and human welfare protection. Article 1110 of NAFTA requires that any expropriation of a foreign investor’s investment be for a public policy purpose and be accompanied by compensation. This is consistent with most OECD country approaches to government expropriation, where it is not sufficient for a government to expropriate property simply for a public purpose: it must still provide compensation. The critical question that triggers the provision is which government acts constitute an expropriation, or a “taking” in U.S. legal language, of property by a government in the first place, and therefore creates the need for compensation? 34 Different countries have evolved distinct approaches to this question. 31 Ethyl v. Canada, Award on Jurisdiction, ibid. 32 Notice of Intent to Submit a Claim to Arbitration Under Section B of Chapter Eleven of the North American Free Trade Agreement, Crompton Corp. v. Canada, November , 2001, paras. 37-39. 33 The United Mexican States v. Metalclad , Supreme Court of British Columbia, 2001 BCSC 664. 34The key issue of what constitutes a “taking” is widely debated in the United States in particular, where the full scope of a constitutional protection of private property rights remains unresolved. Most countries give public authorities wide latitude before recognizing a “taking.” In the United States, however, this issue has continued to be widely debated, and has a primary constitutional law dimension. One of the factors that has made Chapter 11 particularly disconcerting to U.S. environmentalists, and now increasingly

17

Lacking a principle of deference to national law, and lacking any criterion for its application in NAFTA disputes, these provisions are an invitation to companies to seek out arbitrators who espouse particularly pronounced views on the takings issue and who are willing to apply these in selecting a third member and in working within the dynamics of a three-person panel. The public welfare issues raised here are profound, and difficult. To what extent would Article 1110 be applied to laws and regulations that protect the environment and/or human health from hazardous products, from pollution and from dangerous activities? Can setting high environmental standards amount to expropriation if it impacts on business activities? The original fears of environmental groups and others centered around the inclusion of three “types” of expropriation--direct expropriation, indirect expropriation, and measures tantamount to expropriation in Article 1110 of NAFTA. The cases to date have held that these last two terms have the same meaning: measures that do not directly take investment property, but which amount to the same thing. A high enough business tax levied on a specific firm, for example, would eventually have the same effect as direct expropriation. This is also consistent with the more traditional development of the concept of “creeping” expropriation, whereby measures that incrementally create the effect of an expropriation are considered to be just that. While this particular issue no longer appears germane to the debate, a second concern raised early on by at least one critical review of Chapter 11 has materialized. This concerns the way Article 1110 relates to what is called the exercise of “police powers” by a country enacting a measure.35 Under the traditional international law concept of the exercise of police powers, when a state acted in a non-discriminatory manner to protect public goods such as its environment, the health of its people or other public welfare interests, such actions were understood to fall outside the scope of what was meant by expropriation. In trade law terms, this was “a carve out” from the applicable rules. Such acts were simply not covered by the concept of expropriation, were not a taking of property, and no compensation was payable as a matter of international law.

Stated simply, if the police powers rule is recognized under the expropriation provision in Chapter 11, then environmental and human health protection laws or regulations will not be considered expropriations.36 If the rule is not recognized, on the other hand, then even

environmentalists and other observers in all three NAFTA parties, is the growing prospect that a fundamental question of constitutional law in the United States and one of enormous practical implications for all environmental regulators—may now be decided not through the development of domestic case law but through the essentially unappealable rulings of ad hoc Chapter 11 tribunals meeting behind closed doors in a process modeled after private arbitration and based on non-domestic sources of law. 35 This issue was first raised and explored in Mann and von Moltke, 1999. 36 Measures within the normal scope of police powers rule would still be subject to review under the provisions of Chapter 11 on national treatment and minimum international standards, among others. The scope of a state’s legitimate police powers is, of course, not always simple to determine, and may depend on the type of law or regulation in question. For example, protective measures that limit polluting emissions, establish controls or bans on certain hazardous products, would generally be considered routine exercises of police power. Measures that take land to create a national park, on the other hand, are

18

normal exercises of a government’s regulatory authority may be considered expropriations requiring payment of compensation. It is on this issue that the interpretational “battle” over the scope of the expropriation provisions now centers. There do not appear to be any cases of this type on the record prior to NAFTA. Certainly, issues of creeping expropriation have been addressed, but never in a context of public welfare regulation. Thus, it is the Chapter 11 case law that is salient, and it is the Metalclad case that provides the clearest Chapter 11 decision on this issue. Rather than undertake an analysis of whether the Mexican government acted in a manner inconsistent with a normal (nonconfiscatory) exercise of its police powers, the Tribunal stated simply and concisely that “The Tribunal need not decide or consider the motivation or intent of the adoption” of the environmental measure in question in that case.37 Instead, the test that was used in Metalclad considered only the scale of impact of a challenged measure on an investment:

Thus, expropriation under NAFTA includes not only open, deliberate and acknowledged takings of property … but also covert or incidental interference with the use of property which has the effect of depriving the owner, in whole or in significant part, of the use or reasonably-to-be-expected economic benefit of property even if not necessarily to the obvious benefit of the host state.38

This same approach is repeated in the Pope & Talbot v. Canada decision.39 It is the identification of this singular test, combined with what appears to be a complete rejection of the police powers authority in Metalclad, which has raised so many concerns. Some questions remain unanswered regarding this test: what constitutes a significant impact?40 What is a reasonably-to-be-expected economic benefit if part of that benefit is based on creating environmental damage to others? And is that latter issue one for assessing quantum of damages or for assessing whether there is an expropriation in the

generally compensated for in most legal systems and may well not be excluded by the police powers rule. Still, the question of what constitutes a normal (or nonconfiscatory) exercise of police powers varies in some measure in accordance with national custom and practice. 37 Metalclad , para. 111. The measure in question was the creation of an ecological reserve that included the land owned by Metalclad, thereby ending its possible use for siting a hazardous waste management facility, as Metalclad intended. Had the Tribunal held that such a change of land use, while in the public interest, still requires compensation to be paid to the landowners, as is the case in almost all OECD and many other countries, one would have at least found some analysis of the issues relevant to the application of the police powers rule, and a much lower degree of concern flowing from this case. 38 Metalclad , para. 103. 39 Pope & Talbot, paras. 96-105 40 To constitute an expropriation, it is can be argued that traditional international legal analysis has required that an investor lose all or essentially all of the value of an investment through a government action that is so far beyond the normal exercise of police powers as to be confiscatory in nature. In the United States, the “loss of all value” doctrine and the question of normal police powers have been related, because the “value” of a property has been considered as subject to the routine exercise of local regulatory power. Thus, an investor has usually been held to enjoy a reasonable expectation to profit from his or her property only to the extent possible under legitimate applications of the police power. The full analysis of these types of issues under national and international law is complex and beyond the scope of this paper.

19

first place requiring any compensation? 41 Further, it is to be noted that the S.D. Myers tribunal did say that government regulatory action is not normally understood as being expropriation. Unfortunately, that decision creates a genuine ambiguity: it also states that the main reason the measure in that case does not amount to an expropriation is that it is a temporary one with a temporary impact, hence returning suspiciously close to the degree of impact test in Metalclad.42 These “loose ends” give some cause for hope that the Metalclad test may not be as cut and dried as it appears to be. But the situation still remains that the only consistent test across the NAFTA cases is the scale of impact test enunciated most explicitly in Metalclad. At best this approach significantly limits the scope of the police powers rule. More realistically, however, it appears to effectively eliminate this traditional international law test from consideration in the review of an expropriation claim. Under this reasoning, regardless of the purpose, compensation must be paid if there is a significant impact. This is alarming since any environmental law worth adopting will affect business operations and may substantially modify or even end the use of, or trade in, certain processes or products, and therefore will have a significant impact on the business in question. The ultimate effect of this nascent NAFTA doctrine would be to reverse a central tenet of sound environmental policy: that the environmental costs of economic activities should be internalized in prices so that polluters bear the costs of their pollution, rather than enjoying a right to pollute which they must be paid to cede. On the issue of expropriation, the concerns of many civil society groups appear to be well founded. No matter how needed or valuable a new piece of law or regulation, the odds against it will steadily stack up as regulators tally the costs of potential compensation claims from affected businesses under Chapter 11’s expropriation provisions. Indeed, if governments have to guess whether a measure to protect the environment or human health is covered by the concept of expropriation, it could have (and already appears to be having) a significant impact on the freedom of governments to enact strong regulations to protect the environment or other aspects of the public welfare. 4. DISPUTE SETTLEMENT AND SUSTAINABLE DEVELOPMENT So far we have considered the rationale and basic structure of investment agreements, drawing upon Chapter 11 as the leading example of such an agreement. We have considered some of the sustainability issues relating to specific disciplines in these agreements, again using the jurisprudence surrounding Chapter 11 to give detail and texture to them. In this section, we consider the issues concerning dispute settlement under the investor-state mechanism of Chapter 11.

41 This proposal, which IISD believes is fundamentally flawed, is raised as an approach in T. Weiler, ”A First Look at the Interim Award in S.D. Myers Inc. v. Canada: Is it Possible to Balance Legitimate Environmental Concerns with Investment Protection”, Hastings International & Comparative Law Review, Vol 24, pp. 173 et seq., at 187. 42 S.D. Myers v. Canada, paras. 279-287.

20

The principal issue considered here concerns the legitimacy of the investor-state process as a mechanism to resolve the type of public welfare disputes that are now a main feature of Chapter 11 litigation. The central focus of the legitimacy issue arises from the absence of full transparency in the proceedings on the one hand, and the absence of an institutional setting capable of managing the process in a publicly accessible way on the other. 4.1. Transparency

The investor-state process that is the main dispute resolution process for the protection of investor rights under Chapter 11 draws exclusively on three pre-existing arbitration mechanisms, the International Centre for the Settlement of Investment Disputes (ICSID), the ICSID Additional Facility, and the United Nations Centre for International Trade Law (UNCITRAL). Each of these bodies has a specified set of arbitration rules, drawn from a traditional commercial arbitration model. Each follows the basic steps:

• A private party submits the case; • The state responds; • Each side chooses its own arbitrator and a third is appointed in a neutral fashion

(usually from a designated list of international commercial arbitrators); • There is agreement on detailed rules of procedure; • Proceedings are engaged through a series of briefs and oral arguments; and • The arbitration panel reaches a decision.

With a few exceptions, it is these pre-existing rules of procedure that govern the arbitration process under Chapter 11. In addition to the procedural steps, these rules also share an antipathy towards public accessibility and transparency of the proceedings, drawing as they do on a traditional commercial arbitration model. As a result, the first Chapter 11 proceedings took place in almost complete secrecy. Indeed, the existence of many cases was not even discovered until well after they had begun. Further, none of the proceedings have been open to the public or the press, and written arguments of the investor and the state party have been made available in only one case to date.43 The only area of guaranteed public access is for final awards, and then only in the case of Canada and the United States. Mexico reserves the right to maintain awards secret in cases where it is a party, though it has not done so to date. In August 2000, the International Institute for Sustainable Development (IISD) initiated for the first time a petition to appear as an amicus in a Chapter 11 case, notably in Methanex v. United States. In arguments joined subsequently by three US NGOs IISD argued, in particular, that the public interest nature of the litigation in question, challenging the enactment of an environmental protection law, required a greater degree of public access and accessibility than in traditional commercial arbitration cases. Canada

43 So far as we are aware, the full arguments have been made publicly available only in the Methanex v. United States case, still at the jurisdiction phase.

21

and the United States supported the amicus petition, Mexico and Methanex Corp. opposed it. The Tribunal responded positively, noting that:

There is an undoubtedly public interest in this arbitration. The substantive issues extend far beyond those raised by the usual transnational arbitration between private parties. This is not merely because one of the Disputing Parties is a State… The public interest in this arbitration arises from its subject matter, as powerfully suggested in the Petitions. There is also a broader argument, as suggested by the [United States] and Canada: the Chapter 11 arbitral process could benefit from being perceived as more open or transparent, or conversely be harmed if seen as unduly secretive. (Para. 49)

While an order allowing written amicus submissions is anticipated if the jurisdiction proceedings result in a determination to proceed to the merits stage, permission to attend the hearings and to have ensured access to the written pleadings of the parties prior to submitting a brief were both denied. On these points, the secrecy of the proceedings is to be maintained, unless the parties expressly agree otherwise. The IISD recognizes the significant step taken by the tribunal in the Methanex case.44 However, this step in itself does not overcome the remaining transparency failings of the investor-state process. As the Tribunal noted, the importance of public credibility and acceptance is real, and an important value. Both are seriously lacking. Indeed, nowhere is the democratic deficit of NAFTA more readily apparent than in its investor-state dispute settlement procedures. By the summer of 2001, the transparency issue had risen to the top of NAFTA’s political agenda. The July 31, 2001 interpretive statement of the Free Trade Commission focused most of its terms on this question. Even so, the statement in reality moved the legal basis for transparency only marginally. Recent analysis of the statement highlights both the absence of significant parts of the arbitration process from its scope and the lack of binding force of this part of the statement on tribunals. This is because its actual wording defers to the decisions of individual tribunals taken under the arbitral rules to establish their own final rules of procedure, including rules relating to confidentiality of documents. In addition, nothing in the statement overrides the requirements for secret proceedings, closed to press and the public.45 In short while the statement recognizes the political dimension of transparency as a problem, it fails to address in a binding way the details of the problem.

44 For a review of the full decision see Howard Mann, “Opening the Doors: At Least a Little Comment on the Amicus Decision in Methanex v. United States”, Review of European Community and International Environmental Law, Vol. 10, 2001, pages 241-245; and Patrick Dumberry, “The Admissibility of Amicus Curiae Briefs in NAFTA Chapter 11 Proceedings: Some Remarks on the Methanex Case, A Precedent Likely to Be Followed by Other NAFTA Arbitral Tribunals”, Bulletin 1, 2001, Association Suisse de l’arbitrage, pp. 74-85. 45 Van Duzer, supra, n. 19, IISD, Note on NAFTA Commission’s July 31, 2001, Initiative to Clarify Chapter 11 Investment Provisions, at http://www.iisd.org/pdf/2001/trade_nafta_aug2001.pdf

22

4.2. The Institutional Void

Just as critical as the lack of transparency is the absence of an institutional capacity to manage the dispute settlement process and subsequent substantial issues that arise. The process is entirely ad-hoc, and problems begin with the very first steps of a dispute; there is not even an assured avenue to find out whether a dispute has been initiated. While this may be acceptable for commercial arbitration it contravenes the most basic principles of accountability when matters of public welfare are at issue. ICSID has recently begun to publish a list of current disputes but UNCITRAL has no comparable requirement. The negotiators recognized the significance of this issue and provided for a public register of arbitration notices filed, to be maintained by the NAFTA trade secretariat. This secretariat is, however, composed of three national sections with no central point of contact. The obligation to maintain a public register of cases has effectively fallen between the cracks since none of the national sections has overarching responsibilities that transcend its own interests. The choice of arbitrators is left to the parties to a dispute. That may be acceptable for commercial disputes but when matters of public welfare are at stake it again contravenes one of the most fundamental principles of jurisprudence, namely that parties to a dispute may not pick their own judges. Moreover, with no standing roster of arbitrators the accountability of persons who serve as arbitrators is severely attenuated. The protections afforded judges in most societies are usually balanced by stringent requirements concerning public access to and transparency of proceedings. In addition, arbitrators are not expected to be conversant, let alone expert in, Chapter 11 in particular or international investment law in general, and even less so in terms of the types of national and international environmental law matters that have already arisen in the cases. For the most part, the arbitrators chosen for NAFTA cases have been persons of high standing, some of whom have extensive experience as judges but not necessarily in North America. Some choices suggest calculated self- interest from the party concerned. The essential point about institutional safeguards such as the ones discussed in this section is that they are designed to provide reassurance about the legitimacy of the process, no matter how honorable—or even dishonorable—the arbitrators. Chapter 11 provides for no formal appeal process, leaving this to an ad hoc process of judicial review under various local statutes that vary in scope and quality of review. Given the importance of Chapter 11 in the scheme of NAFTA, and its now obvious linkage to critical public policy and public welfare issues, this lack of institutional framework is a critical gap. Some observers have suggested that Chapter 11 “case law” would find pragmatic solutions to many of the problems that had arisen. The institutional and procedural gaps described above reduce the likelihood of this occurring. Most importantly, lack of transparency makes the notion of “case law” a hit or miss proposition: there is never any assurance that all relevant information is available. Declarations from parties to a dispute that the necessary information has been released cannot be verified and there is no objective standard against which to judge such declarations. Also highly significant is the lack of any institutions that might promote a sense of collective responsibility on the part

23

of panel members. Moreover, their immediate obligations are to the specific aspects of the dispute before them and not to any overarching institutional purpose. Indeed, they are professionally enjoined from engaging in “strategic” behavior that may be required to remedy defects in the text of the agreement. It is by now fairly obvious that the negotiators—themselves honorable men and women—created a quasi-judicial process without giving much thought to some of the most fundamental principles that underpin the administration of justice in the countries of North America. That is inexplicable unless one assumes they remained unaware of the public welfare implications of their actions. There is a certain irony in the fact that the United States has resisted the creation of an International Criminal Court with a full panoply of institutional safeguards for fear of harm for its citizens. Yet it has accepted the jurisdiction of NAFTA arbitration panels that do not meet some of the most elementary standards of institutional probity when the public welfare is at stake. In most democratic countries, major investment decisions involve a significant institutional commitment on the part of public authorities, to ensure that essential public goods are promoted, to provide necessary support to investors, and to ensure that disputes between investors and public authorities at all levels can be considered under circumstances that all parties recognize as legitimate and equitable. These include administrative hearings (often at several levels and with varying membership), transparency, environmental assessments, public participation and judicial review, all within a complex framework of substantive and procedural law. An international process that substitutes itself for these domestic procedures must meet the same fundamental standards of legitimacy, accountability and equity that have been implemented domestically. Otherwise it will inevitably result in the erosion of central values of many of the countries in question. It needs to be emphasized that none of the above is an argument against investor-state dispute settlement per se. States do not invest in other states and it is not the function of governments to act as legal counsel to their citizens in private matters abroad. The logic of globalization with the resulting increase in private actions at the international level and the concomitant need for legal frameworks to promote fairness actually points strongly towards the creation of substantial international institutions. Indeed, this process is well under way. There are also a growing number of issues concerning international public welfare that can no longer be fully represented by individual states acting through their own institutions alone. These also create a dynamic towards the emergence of international jurisdiction. Yet none of this justifies abjuring the most fundamental principles of legitimacy, transparency and accountability that have been developed over long periods of time to ensure that private interests and the public welfare can be properly balanced. Moreover, none of this reduces the imperative—discussed below—to ensure that investment law incorporates the responsibilities of the investors themselves, as well as their rights.

24

5. OPTIONS FOR CHANGE The inclusion of provisions promoting sustainable development can be seen as a positive new direction for international investment law. At the same time, addressing the interpretation of the specific investor protection provisions (of Chapter 11 or other agreements) can be understood to some extent as “damage control”: ensuring that legitimate elements of investor protection do not come at the expense of the governmental role in protecting public goods and the public welfare. Options relating to both of these categories are discussed below. Options for change must also be considered with two other criteria in mind: options of specific relevance to Chapter 11, and options geared to the development of further investment agreements. Again, both are considered below. 5.1. Chapter 11 Options: Interpretive statement and amendment

In 1999, IISD proposed an interpretive statement under Article 1131(2) of Chapter 11 to remedy some defects of the Chapter 11 process that had become evident by then. This Article requires that Chapter 11 tribunals apply any interpretations of Chapter 11’s provisions as may be set out by the Free Trade Commission. In international law, this is known as an interpretive statement. While an interpretive statement formally issued by the parties to a treaty is not necessarily binding on a dispute resolution body under Article 31 of the Vienna Convention on the Law of Treaties, Article 1131(2) of NAFTA does make it fully binding. In essence, this was a safety valve inserted into Chapter 11 to allow some control by the parties over the life of the Chapter. There were two main reasons for the approach advocated by IISD. The first reason was that Mexico would not undertake any significant review of the provisions of Chapter 11 at that time, as it believed the cases would sort out any putative problems. An interpretive statement was thus seen as responding to this reluctance. In any event, further cases have not clarified the situation as was expected. Indeed some new issues of concern have emerged. The second reason was the hope that an interpretive statement could be developed in a relatively short period of time, compared to an amendment process. It was thus thought worthwhile to test its value as a tool for the Parties to resolve problems that arose. To date, the one statement issued in July 2001 was limited in scope, and remains uncertain in terms of its substantive and procedural impact. Negotiations on a second statement are on the table, but there are no reports of any significant progress. To date, there is no indication of its possible scope, nor any public discussion of proposals for such a statement. Finally, it was thought that suggesting a full amendment process might be premature. Given the apparent lack of additional issues that might be considered as part of an amendment process, it was seen as unlikely that an amendment process just for Chapter 11 was a realistic possibility.

25

As noted several times previously, the Free Trade Commission did issue an interpretive statement on 31 July 2001. It covered two issues, transparency and the interpretation of the minimum international standards provision. In both cases, as already noted, the scope of the statement is limited, and on transparency it is not legally binding on the tribunals. While the IISD believes that the interpretive statement process should continue, its inherent limitations, combined with the passage of an additional three years, suggest that it may be necessary to move towards an amendment. The NAFTA text was written ten years ago and it has now been in force for over eight years. There are other issues of substance that have developed, and other areas of process that need addressing. (In particular, public access to the working groups and other bodies that implement and further develop NAFTA’s obligations is becoming critical as these bodies move into areas of broader public interest.) Secondly, as regards Chapter 11 itself, it is apparent that an interpretive statement cannot provide the procedural legitimacy and institutional robustness that is required of the dispute settlement system. Only an amendment can do so. Finally, while an interpretive statement can address the “damage control” required to mitigate the uncertainties and regulatory chill created by the existing provisions of Chapter 11, it cannot expand the content of Chapter 11 to include the promotion of sustainable investments. Given these factors, the option of an amendment to Chapter 11, with or without other amendments to NAFTA, should now be considered. Promoting sustainable investments, and ensuring the practical ability of governments to maintain environmental management regimes that are dynamic, preventive and responsive to changing needs is central to the global capacity to achieve sustainable development. This recognition underscores the importance of getting Chapter 11, and all other investment agreements, right. 5.2. Future negotiations

Future negotiations on an amendment to Chapter 11 or on a new agreement should address two factors: the promotion of sustainable investments, and ensuring that provisions related to investor protections do not inhibit the ability of governments to take legitimate actions to protect the public welfare. The first step in relation to any negotiations must come at a conceptual level: investment negotiators can no longer remain in ignorance of the environment and sustainable development implications of their actions, as has been the case in the broadened investment negotiations since 1990. Environment and sustainable development are part of what they do--indeed are at the core of what they do--and only the narrowest neo-classical view of economics could pretend otherwise. It is no longer acceptable to simply promote investment and market liberalization while deferring the sustainability dimensions to the domestic legal systems, particularly when those systems are expressly overridden by the creation of internationally prescribed private economic rights. Yes, this would make already complex negotiations much more so. But any fear of addressing the necessary complexities only reinforces the need to broaden the negotiating base beyond trade negotiators.

26

Second, the process of negotiation must be much more transparent. The days of closed-door negotiations on matters of high public importance should be over. Open doors and access to evolving draft texts will also ensure a much sounder review of the potential consequences of the negotiations before they are concluded, rather than several years afterwards. Third, on substance, several elements can be contemplated for inclusion in an agreement that would work to promote sustainable investments. These include:

• Mandating environmental assessments of proposed projects, based on the higher of locally applicable or minimum international standards in this regard. The World Bank standards might be considered here.

• Requiring investors to establish certified environmental managements systems in their ongoing operations.

• Ending the private international law rules on forum non conveniens that allow foreign investors to shed legal liabilities in their home states for acts of their foreign investments. This measure--cost free to governments--would dramatically alter the corporate perception of responsibility for the activities of their subsidiaries, and create a significant legal incentive to improve performance levels.

• Capacity building from a legal, administrative and enforcement perspective are necessary adjuncts to this approach, and should be developed as an integral part of a broader investment agreement.

Avoiding unintended interference with the government duty to protect the public welfare is obviously critical. Here, again, several elements can be proposed: • Much closer attention must be paid to the negotiation of international investment

agreements than appears to have been the case in NAFTA’s Chapter 11. It cannot be assumed that provisions on investor protection are benign in this regard.

• A possible carve out provision or exclusionary clause for areas of “legitimate regulation” under NAFTA or areas covered by the general exceptions in Article XX of the GATT should be considered.

• The NAFTA experience must provoke a rethinking of the desired scope of investor protections and private sector economic rights. Disciplines must be spelled out clearly, not by simply utilizing terms brought over from trade law. It cannot be assumed that trade law is wholly or even in part applicable to investment. While trade and investment decisions may reflect decision-making on a business continuum, their footprints in environmental and social terms can be vastly different, and these differences must be reflected at the core of an investment regime.

• Disciplines must be clear, not vague. They must have a finite range of interpretation, rather than be open ended. To some extent, of course, lawyers will always seek the unforeseen. It is incumbent on governments to now take the risks associated with this

27

search seriously, and undertake the same effort before an agreement is concluded rather than afterwards.