Article body

In his November 2008 Throne speech, Prime Minister Stephen Harper expressed his intention to preserve Canada’s environment and “to ban all bulk water […] exports[1]”. The Throne speech is a clear reminder that debates about water exports are all but over, largely because water is still a contentious source of hydro-nationalism and a highly sensitive topic. Over the last two decades, tremendous efforts have been dedicated to assessing the impacts of NAFTA[2] upon Canada’s water resources, mainly in legal journals[3]. Despite the widespread attention, it is surprising that little empirical work has been dedicated to the only currently existing type of bulk water transfers[4] between the two countries, that is to say, transboundary local water supplies between communities located along the Canadian-American border. These decades-old water transfers involve limited inter-local flows that are not just unidirectional towards the United States ; some are also directed towards Canada, while others are bidirectional.

The writing of this paper was prompted by the lack of information surrounding these water transfers, particularly since the entry into force of NAFTA. Many factors have contributed to this situation. First, research prior to NAFTA was mostly conducted by non-legal scholars, who were not overly concerned with legal issues. Second, when compared to continental-scale water diversions[5], these inter-local water transfers were considered to be Lilliputian and of little interest in terms of size and scale[6], and also because they were perceived as representing very little threat to Canada’s national interest[7]. Third, the literature published after the entry into force of NAFTA has highlighted the relevance of having a better understanding of Canada’s water resources, especially its transboundary waters. But it has poorly assessed the legal implications of inter-community water transfers across the border. Nonetheless, their relatively small scale should not be misinterpreted, since events in recent years have shown that the local scale does matter. Such local transfers have the ability to draw national and international attention, as was the case with Devils Lake[8], or Detroit’s water removals from Canada’s territorial waters[9].

The hypothesis supported by this paper is that transboundary local water supplies agreements between borderland communities are not commercial transactions through which water is traded as a product or as a service, and that they do not lead to the commodification of water. Instead, water is considered here as a public good[10]. These transfers constitute a type of inter-local collaboration between public entities that are dedicated to providing freshwater to their respective constituencies. The aim of these agreements is not strictly limited to water, but includes the public delivery service that allows this very water to be transferred from one side of the border to the other.

This paper is based on a geo-legal approach[11], and is divided into five sections. The first one introduces the concept of transboundary local water supplies and provides some factual data about their spatial, legal, economic, and social organization. The second section looks at the legal status of the water transferred between the borderland communities within the context of NAFTA, as well as the potential role of the North American Agreement on Environmental Cooperation. The third one questions whether or not transboundary local water supplies provide grounds for a precedent. The fourth section focuses upon government procurements in relation to the “buy American” clause in the U.S.’s stimulus package. The last section refers to the legal status of the infrastructures that allow the transboundary circulation of water.

1 Introducing the concept of transboundary local water supplies

One of the main tasks of any Canadian municipality is to provide high quality freshwater to its constituents at a reasonable cost and in a sufficient volume. To achieve that purpose, it is frequent for communities to combine their resources, for instance within metropolitan areas. But it is rather unusual to hear about agreements in the realm of water bringing together Canadian and American borderland communities[12]. The geographical situation of some of these communities has been particularly helpful in such endeavours. Indeed, in many places, the border is a mere legal convention cutting across urban settings that is almost indistinguishable from the air[13]. Thanks to the nearby presence of main transportation axes (highways and railroads), many of these communities are important points of entry for transiting goods and services between the two countries. It could be added that prior to 9/11, crossing the border was relatively easy and there was very little surveillance of the locals’ comings and goings, thus allowing the establishment of close relations between them. That physical proximity, coupled with a border that is mostly unprotected, has allowed for decades of in-depth collaboration and interaction. These communities have literally grown up together, and the realms of cooperation are very diverse : social (both formal, such as healthcare and firefighting, and informal, primarily interpersonal relations), economic (trade and investment), sporting (sports tournaments and infrastructure), and political (formal/informal agreements). Hence the water sharing agreements were just another layer of cooperation. Moreover, rather than being considered as a commodity, water has been seen by these communities as a public good being shared through the extension of the underground pipes up to, and even across, the border. These agreements are involving the sharing of a public service with another public body, albeit one that is across the border (fig. below).

Transboundary local water supplies : location and direction of the water transfers*

Transboundary local water supplies : location and direction of the water transfers*

* ESRI Data and Maps [DVD] (1 May 2008). The arrows indicate the direction of the flows. The dark arrows refer to existing water supplies, and the pale ones to the transfers that have ended.

-> See the list of figures

A transboundary local water supplies agreement can be defined as an inter-local collaboration between borderland communities aiming at sharing their water supplies[14]. It results from “interlocal diplomacy” or “municipal diplomacy[15]”. These water transfers were developed endogenously by the locals in response to local needs. Therefore, each agreement is singular ; none has evolved nor is structured or designed according to the same model. In all cases, the transboundary circulation of freshwater has been limited to local domestic consumption. As a matter of fact, the largest of the water pipes going across the border has a diameter of 26 inches. Their capacity is thus clearly not enough to serve other communities, farms or industrial facilities located close by in the hinterland. Also, they do not involve any inter-basin water transfers as the water is coming from nearby sources — often shared — like transboundary aquifers[16]. Spatially, these inter-local bulk water supplies are scattered all along the Canadian-­American border, within all the provinces sharing a border with the contiguous United States. Data obtained through my doctoral research[17] shows that close to a quarter of all twin borderland communities are involved in water transfers, out of the 59 twin communities contacted. While the literature has previously referred to four agreements[18], my research has located a further nine agreements (cf. fig. above).

In most cases, water flows are directed from a supplying community to a benefitting one, either towards Canada or the United States. Some bidirectional flows also exist, for instance between North Portal, SK and Portal, MT (for maintenance purposes), Akwesasne, QC and St. Regis, NY (for emergency purposes), and Stanstead, QC and Derby Line, VT (for regular consumption). The latter example is of particular interest because of its longevity ; it has been in place since 1906. Initially, the International Water Company (IWC), incorporated in Vermont, was the owner of the network. It was in charge of the growth and development of the water supply system, which was designed as a single entity despite being binational. Eventually, the IWC became a transnational public utility as the three local communities[19] gradually bought the company’s shares, taking complete ownership by the end of the 1950s. Today, the binational infrastructure is critical for keeping the whole system working. The wells and the wastewater facilities are located on the Québec side of the border, while the water reservoirs are situated in Vermont[20]. In comparison with the other existing agreements, this inter-local water system is the most integrated one, with the water being circulated from one side of the border to the other thanks to each community’s involvement. The other transfers surveyed were mostly limited to a unidirectional flow connecting two distinct water distribution systems.

What factors have contributed to the establishment of these inter-local water supplies ? First, the communities had to deal with the existence or expected imminence of water scarcity/stress, both qualitative (through pollution or salinization of the streams or aquifers) and quantitative (due to water output being insufficient to accommodate a growing population). Second, money issues were an important factor since water distribution systems (wells, water treatment plants, wastewater treatment plants, reservoirs, pipes) are costly to build and to maintain. Since many borderland communities have a relatively small population, their ability to borrow money or to get access to governmental funding is quite restricted. When such funding has not been easily available or accessible, a partnership with a sister city has proven to be the most convenient way to provide water to citizens.

2 What is the legal status of transboundary local water supplies ?

Within the literature, transboundary local water supplies have been variously described as “exportation of water[21]”, “minor transfers[22]”, “diversions[23]”, “local scale exportation[24]” or “exportation at the local scale[25]”. In her doctoral dissertation, Bakenova affirmed that transboundary local water supplies are the only known exception, with the “interbasin diversion of shared boundary waters[26]”, to the claim that “there have never been any significant water exports from this country[27]”. Furthermore, Johansen has asserted that “Canada exports its water resources towards the United States when water is transferred, through canalisation, from communities located along the border to communities located in the United States[28]”. A further wrinkle in the literature is Fritz & McKinney’s[29] characterization of the water pipes running through the border as “pipelines”. The latter have also disputed the conception of treated water as a water export, based upon the argument that the pipes have a limited impact upon society and nature as compared to “continental schemes typically considered as exportation and which received most of the attention”.

These divergent portrayals of transboundary local water supplies reveal the grey zone in which they are situated : do they constitute water exports or not ? Yet in considering this question, it is simultaneously important to recognize that most of these assertions were made before the entry into force of NAFTA or CUFTA[30], at a time when the concept of water export was not freighted with the heavy legal-commercial baggage that it has today. Moreover, most of the papers were written by non-legal scholars and thus involved very little legal analysis. Generally speaking, prior to the 1990s, the literature showed a propensity for interchangeably using the terms “water export” and “water transfer”. During this period, the concept of water export referred to a predominantly geographical phenomenon, and was conceived as a) the physical transfer of water, b) across the international border, c) through human intervention (canal, pipeline, rail, truck, tanker, bags)[31]. Quinn, a long-time expert in the field, proposed a definition of the concept of “water export” based on the public’s perception : “For most Canadians, water export means the artificial and massive diversion of lakes and rivers from their basins of origin for delivery by canal, pipeline or other means to the United States[32].”

But with the entry into force of CUFTA and NAFTA, a definitional shift took place. From that point on, the concept of water exports in the bulk water transfer literature became increasingly associated with a commercial and legal meaning. The Canadian government, as well as many scholars — particularly legal ones — had become aware that any reference to water as an “export” might have legal implications for Canada under NAFTA. A new set of concepts with environmental connotations then evolved around the notion of “transfer” : removal, extraction, transfer, or diversion. For example, the International Boundary Waters Regulations of the Canadian Department of Foreign Affairs and International Trade refer to the “removal of boundary waters in bulk[33]”, which is defined as “the removal of water from boundary waters and taking the water, whether it has been treated or not, outside the water basin in which the boundary waters are located[34]”. Environment Canada uses the expression “bulk water removal[35]”, which is defined as the “removal and transfer of water out of its basin of origin by man-made diversions (e.g., canals), tanker ships or trucks, and pipelines”. Significantly, the expression “water export” is nowhere in sight. In sum, choosing between “transfer” (in its geographical-environmental meaning) and “export” (in its commercial meaning) has become important, since each term refers to a different set of thoughts, one geographical and environmental, the other legal and commercial.

Where do transboundary local water supplies fit into this picture ? This paper argues that they are bulk water transfers. They do not belong to the “water export” category even if the transferred waters are no longer in their natural state. These water supplies are also not meant for commercial purposes — indeed, if this was the case, they would be covered by NAFTA[36]. Instead, they are inter-local public partnerships that value water based on its intrinsic value rather than its exchange value. It can also be argued that similar intra-basin water transfers currently exist within Canada. Municipalities are sharing common water distribution networks, for instance within metropolitan areas. The water itself is not subject to a charge, but rather the whole supply chain as a service — which involves the extraction, the treatment and the distribution of water — is charged for. However, because of the presence of the international border, transboundary local water supplies are significantly different from their Canada-only counterparts. The very presence of the border renders the situation even more complex, since it involves two different territorial jurisdictions and raises the issue of national sovereignty. These transfers are also both transnational and multiscalar : they involve the local scale, as well as the national and the international scale, since Canada and the United States are the only actors recognized under international law.

When asked about their perception of these transboundary local water supplies, the local and national actors contacted during the course of this research (71 taped interviews) articulated different views. Many argued that these inter-local water supplies are causing water to morph into a commodity since it is extracted, treated, distributed, sold, and transferred across the border. Water is thus transformed through society’s actions. It is no longer in its natural state. As was noted by two respondents :

We may consider that [water] is transformed as it is controlled. The chemistry is not changed, but dams, pipelines, canals […] I think it is becoming a commodity[37].

Water is not a good until you do something. Water is a fungible good : you can own it only when you capture it. That is the same with petroleum […]. If it is put into a pipeline, then, it is captured[38].

For others, the extraction, treatment, and distribution processes do not modify the status of water, which remains a public good, and they do not commodify it : “What does water treatment do ? Is it incorporating [water] into a product ? No, it is just drinking water. […] In most municipalities you pay for the service of treatment. You’re not buying the water[39].”

Thompson and others[40] and Anderson[41] have both affirmed that to be commodified, water must be treated as a tradable good of commerce within the context of a free market, where transactions between buyers and sellers are setting the price of any given commodity. Currently, such markets do not exist along the Canadian-American border. The water that is extracted by the Canadian borderland communities is used for the public good, is not considered as a commodity, and is not part of any sort of commercial transaction. Water transfers are made possible through partnership agreements between public actors, such as municipalities, water districts or public entities. These actors are committed, within the constraints of their territorial jurisdiction and responsibilities, to provide a public service to their respective constituencies. Even if the water is crossing the border, its purpose remains the same. In fact, transboundary local water supplies constitute an extension of a public service across the border. They are not a source of profit. It could also be argued that Canadian municipalities are not per se the owner of that water, but rather the provinces are, within the limits of their territorial jurisdiction[42]. Moreover, from a technical perspective, the water pipes connecting the borderland communities are being used for a water transfer from one nation to another ; however, in many cases, the water is extracted from transboundary aquifers[43]. Hence the source itself is transnational[44], and thus undermines the idea of a “water export”.

Following a joint reference made by Canada and the United States to the International Joint Committee (IJC, Feb. 10th 1999), the Canadian Department of Foreign Affairs and International Trade (DFAIT) submitted a document to the IJC on Nov. 16th, 1999, which unambiguously reiterated Canada’s position[45] concerning water : “Water does not become a good until it is removed from its natural state and enters into commerce as a saleable commodity, such as in bottles or in bulk containers[46].” However, the same document does not present the same degree of certitude in regards to transboundary local water supplies, which are not even expressly mentioned. On the one hand, it asserts that water is not a good when delivered by “municipalities or a province for domestic, industrial and agricultural uses[47]”. The charge for that water merely reflects “the cost of supplying it rather than a price for it as a commodity[48]”. The document also makes clear that water delivered by public entities is not considered by Canada to be a good covered by NAFTA. Nevertheless, it is unclear whether or not DFAIT had transboundary local water supplies in mind when it wrote the Canadian position, or if it was only referring to municipalities located within Canada’s boundaries, since in the very same paragraph, it stipulates that “[e]ven if that water [delivered by municipalities or a province] were considered a good, it would […] be in respect of that particular water and not water remaining in its natural state[49].” This quote points to an ambiguity in Canada’s position, since the use of the words “even if” opens up the possibility of considering the delivery of water by municipalities or a province as a good. If transboundary local water supplies fall under the rubric of ‘water delivered by municipalities or a province’ — as seems reasonable to assume in the absence of their being specifically referred to — then the water that they are transferring could potentially be considered as a good. And recognizing the water that is transferred through transboundary local water supplies as a good would logically put it under the jurisdiction of NAFTA, which could then potentially lead to a precedent being set.

In my view, DFAIT did not properly address the issue of transboundary local water supplies when it produced the document defending the Canadian position. The latter was written in the context of a joint reference with the US that aimed at examining water uses and diversions in the Great Lakes. DFAIT’s core concern was to make sure that water in its natural state is not treated as a good under NAFTA. If the water delivered by municipalities or a province was to be considered as a good, DFAIT claimed that it would not change the status of the remaining water in its natural state. In comparison to the Great Lakes water export controversy, transboundary local water supplies were a secondary issue that did not match the perceived threats to the Great Lakes’ water. Still, DFAIT missed a chance to clarify the legal status of the water that is transferred through inter-local water agreements.

“Even if” that water was considered to be a good, the General Agreement on Tariffs and Trade (GATT) could not prevent Canada or the provinces from legislating and restraining any transboundary local water supplies. According to Girouard’s interpretation[50], article XI relating to the elimination of quantitative restrictions other than taxes, duties or other charges, affirms that : “[I]f the supply of water is best understood as a usufructuary right or as a utility service, then regulating the supply of water constitutes neither an “export restriction” maintained in violation of the literal language of Article XI :1, nor a pattern of behaviour that the world trading system seeks to proscribe.”

Thus article XI allows the government to monitor and regulate transboundary local water supplies. Any measure aiming at regulating these utility services would not constitute an “export restriction.” If we accept this paper’s interpretation that transboundary local water supplies are not water exports but water transfers that do not involve any commercial meaning, and that they are not transforming water into a good, could it be argued that these water transfers are a service covered by NAFTA ?

Chapter 11 (Investment) relates to investors from another Party, or investments made in Canada by investors from another Party, that are leading to the production of a good or a service in Canada. The chapter is of little importance since no American borderland community has ever made such investments regarding their water supplies in Canada. Even if they had, the article 1139 restricts the definition of an “investment” to the commercial, economic, and corporative spheres[51]. It excludes governments de facto as investors. Moreover, article 1108 specifically rules out local administrations[52] from the application of articles 1102 (National Treatment), 1103 (Most-Favoured-Nation Treatment), 1106 (Performance Requirements) and 1107 (Senior Management and Boards of Directors). Also, annex II-C-8 allows Canada to exclude water distribution services from the application of chapter 11. But as the latter does not apply to local governments, it is not surprising that the federal government did not make any use of the annex.

Chapter 12 relates to measures maintained or adopted by a Party “relating to cross-border trade in services by service providers of another Party” (art. 1201). This chapter is restricted to the commercial trade of services, which does not apply to transboundary local water supplies. Furthermore, article 1201.3 (b) specifies that : “Nothing in this Chapter shall be construed to : […] (b) prevent a Party from providing a service or performing a function such as law enforcement, correctional services, income security or insurance, social security or insurance, social welfare, public education, public training, health, and child care, in a manner that is not inconsistent with this Chapter.”

The Vienna Convention on the Law of Treaties[53] could be helpful in interpreting the legal status of the transboundary local water supply agreements within the context of NAFTA. Article 31 (1) of the convention stipulates that “[a] treaty shall be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose.” Therefore, the description of these transferred waters as a “product” seems unlikely. They do not involve a commercial transaction in which private parties are freely establishing the price of a commodity in the context of a free market. Instead, the transferred waters should be depicted as a public good (often extracted from a transnational aquifer) that is shared through public service delivery thanks to long term inter-local agreements.

Also, the North American Agreement on Environmental Cooperation (NAAEC) was created as an environmental side agreement to NAFTA[54]. The treaty’s preamble states that the three governments are “convinced of the importance of the conservation, protection and enhancement of the environment in their territories and the essential role of cooperation[55]”. To that end, the Parties are invited to “enhance compliance with, and enforcement of, environmental laws and regulations” (art. 1 (g)). Does the NAAEC have anything to say about transboundary local water supplies ?

First, this agreement relates to NAFTA, which is dedicated to easing the trade of goods and services between the Parties. Article 1 (e) of the NAAEC specifically affirms that the environment cannot be used to “creat[e] trade distortions or new trade barriers”. As was argued previously in this paper, transboundary local water supplies are not covered under NAFTA, and they cannot be assimilated to commercial transactions. NAAEC would probably not be enforceable upon these transfers. In fact, under the dispute resolution mechanism, article 24 (1) allows a Party to request an Arbitral Panel to consider any “pattern of failure by the Party complained against to effectively enforce its environmental law”. However, article 24 limits the jurisdiction of the Arbitral Panel to “a situation involving workplaces, firms, companies or sectors that produce goods or provide services : (a) traded between the territories of the Parties ; or (b) that compete, in the territory of the Party complained against, with goods or services produced or provided by persons of another Party.”

As such, a Party could hardly make a case that the waters transferred through a transboundary local water supply agreement are goods or services that are traded or compete with “goods or services produced or provided by persons of another Party[56]”, even less produced by firms or companies. Second, the use of the dispute mechanism by Canada is quite restricted. The Canadian federal government is bound to the NAAEC, but only three provinces have agreed to sign the Canadian Intergovernmental Agreement[57], which extends the application of the NAAEC to their environmental jurisdiction, namely Québec, Alberta, and Manitoba[58]. Hence as long as no federal environmental laws are involved, only a few borderland communities, located in Québec and in Alberta, would be potentially affected by the NAEEC. Third, to call in the dispute mechanism, any Party[59] would have to prove the existence of a “persistent pattern of failure […] to effectively enforce its environmental law[60]” (art. 22 (1)) in regards to transboundary local water supplies. These inter-local water agreements, as I have observed, are spatially restricted to borderland communities and have limited or nonexistent impact upon their local environment. The applicability of the NAAEC to such agreements seems therefore to be restricted at most. Finally, although the NAAEC provides a dispute resolution mechanism (Part 5), under annex 41 (4), Canada “cannot [yet] fully benefit from the dispute resolution or arbitral panel process”. To be able to do so, “Canadian provinces representing 55 percent of GDP [would have to] become signatories to the [Canadian Intergovernmental Agreement][61]”.

3 Are transboundary local water supplies setting a precedent ?

Most transboundary local water supplies agreements were signed decades ago. They precede NAFTA by many years and set a precedent as the longest existing bulk water transfers between Canada and the United States. However, they do not represent a precedent under NAFTA that could be invoked by a private company or an investor to export bulk water. According to many experts who were consulted during the course of this research[62] existing transboundary local water supplies do benefit from vested interests, either from the grandfather clause confirmed in an agreement reached by the three Parties in 1996[63] or simply because they were established before NAFTA.

If a new transboundary local water supplies agreement that was similar in its characteristics to the existing ones was to be inaugurated, it would probably not be treated much differently than the existing ones, as it would remain a public service. On the other hand, if such a transfer was prompted by private interests and correlated with a commercial transaction involving water as a good or a service, then it might create a precedent under NAFTA[64].

4 Government procurement markets : the local and NAFTA

NAFTA’s chapter 10 relates to government procurements. It applies to “measures adopted or maintained by a Party relating to procurement : (a) by a federal government entity […], a government enterprise […], or a state or provincial government entity”. (art. 1001.1) These governmental entities and enterprises are all listed in annex 1001.1a-1, 1001.1a-2 and 1001.1a-3. Neither article 1001 nor the annexes relate to local governments or entities ; instead, they are only concerned with the upper levels of government. However, article 1024.1 did foresee further negotiations “no later than December 31, 1998, with a view to the further liberalization of their respective government procurement markets” notably for the purpose of “seeking to expand the coverage of this Chapter, including by adding (i) other government enterprises” (art. 1024.2 (b)). This article also urges “the Parties […] to consult their state and provincial governments with a view to obtaining commitments, on a voluntary and reciprocal basis, to include within this Chapter procurement by state and provincial government entities and enterprises” (art. 1024.3). Even though the Parties to NAFTA have agreed to scale down the application of this chapter to governmental entities and enterprises at the provincial/state level, these negotiations did not concern local governments[65], at least until now.

The recent American Recovery and Reinvestment Act[66], which was adopted by the American government in February 2009 ( $787 billion economic stimulus package), contains a “buy American” clause, which requires purchasing products made in the U.S. for all funded public works projects. Arguing against the protectionist nature of this measure, Canada negotiated an agreement with the United States to allow Canadian suppliers to bid on U.S. projects. This agreement entered into force on February 16, 2010[67]. It offers three main gains for Canada. The first one (art. 1) gives Canadian companies access to the procurement markets of 37 of the 50 states included in annex 2 of Appendix I of the Plurilateral Agreement on Government Procurement (GPA)[68]. The second gain is the exemption of Canada from the Buy American Clause for seven grant programs funded by the stimulus package (art. 7 and annex 3, list C). The third gain involves Canada and the US agreeing to negotiate an enlarged, permanent, and long-term agreement (art. 9). In exchange, Canada agreed to open the procurement markets of its sub-central government entities (provinces)[69] and some municipalities[70] to U.S. companies, which represents a market of over $100 billion[71] annually.

Are transboundary local water supplies agreements affected by this agreement ? From what was observed at the border, most infrastructures are already in place and do not require more than regular maintenance. No major new investments are to be expected for many years, which lessens the impact of the agreement given that it is due to expire on September 30th 2011. Even if it was permanent, most provinces specifically listed the municipalities covered, usually the largest ones, with none being involved in inter-local water agreements. The only exception is British Columbia, which extended the application of the Treaty to “[a]ll Crown Corporations and all municipalities” (Appendix C). Hence the Greater Vancouver Regional District and the City of Surrey could be affected by the agreement, though only if their procurements are above the $8,500,000 (Cdn) threshold. But as mentioned above, no such investments are planned in the short- and medium-term.

5 The infrastructures behind the transfer of water

Transferring water across the border necessitates at least some vital infrastructures, such as water mains, water treatment plants, pump stations, water reservoirs, water storage tanks and disinfection facilities. What is the status of these infrastructures under NAFTA ? Following fieldwork and interviews with local actors, it has been observed that most infrastructures were separately owned by each of the communities involved. In most cases, the contiguous water distribution networks did not need any particular transnational infrastructures, except for the joint itself at the border. There are two major exceptions :

  1. In 1987, the Greater Vancouver Water District (GVWD) agreed to supply the Point Roberts Water District (PRWD) with freshwater from Seymour reservoir. At that time, the Canadian Transport Commission allowed the GVWD to install on Canadian soil a “26 inch diameter commodity pipeline to transfer 0.7 million gallons per day[72]” (3,182 m3), with an approximate length of 4,400 feet. It was deemed necessary by the GVWD to construct a reservoir near the Canadian side of the border. Article 3.1 of the contract specified that the construction costs of the main and the reservoir (22,730 m3 or 5 million imperial gallons)[73] were to be paid by PRWD[74] as connection fees, while the GVWD would retain ownership.

  2. The inter-local water supplies agreement between Stanstead, QC and Derby Line, VT is probably the most fully realized one. Established in 1906, the International Water Company (IWC) was initially a private entity owned by local interests. The villages gradually bought the company’s shares until they achieved total ownership by the end of the 1950s, thus transforming the IWC into a transnational public utility. The IWC owned the entire water distribution network, including the water mains and the reservoir. In 1996, new investments were needed and a new agreement was signed. A covered reservoir was built beside the old one, and two wells were dug in Stanstead. The two municipalities each gained ownership over the newly built infrastructure located on their own territory, and of their respective water pipe network. The IWC kept its ownership over the old reservoir while remaining in charge of water quality and the maintenance of the reservoirs and the wells.

Conclusion

This paper has stressed that transboundary local water supplies cannot be only restricted to water circulating across the border. The water is rather part of larger water distribution networks, including water plants, treatment plants, pipes, and reservoirs, that allow its circulation from the source to households, from one country to another. In fact, it could be said that water is acting as a lubricant[75] for social relations. It is delivered as a public service thanks to long-time inter-local partnerships. Water has become, for these communities, a symbol of transboundary collaboration. But in doing so, this very practice has raised many legal issues in relation to NAFTA.

This paper has argued that transboundary local water supplies are not covered by NAFTA. These inter-local agreements are not meant to be commercial transactions through which buyers and sellers freely establish the price of water as if it was a commodity, either as a good or a service. The communities are not paying for water itself ; instead, they are paying for the service of extracting, treating, and distributing the water, as well as for the availability and accessibility of the contiguous water system that allows the circulation of water across the border. Water is treated as a public good. The extension of a public service from one community to another is not intrinsically different whether it is performed across the border or within the hinterland. These transfers should be conceived as good neighbour policies between borderland communities that share tight relationships in the political, economic, professional, and social realms[76]. For these communities, the transferred water is vital to their survival. Finally, considering these transfers as exports under NAFTA makes little sense, especially in the cases where water is extracted from transboundary aquifers.