top of page

Preliminary Empirical Findings - Adoption of Water Sharing Agreements
Authors: Professor Brady Deaton Jr. and Bethany Lipka 

PDF version of this document available here: 

Key Empirical Research Question: are First Nations less likely to be engaged in water sharing arrangements (WSAs) than Ontario municipalities, after controlling for important considerations? 

Map showing the extent of water sharing taking place around the city of London through two regional networks: Lake Huron Primary Water Supply System and Elgin Area Primary Water Supply System. Both of these regional supply networks service the City of London. The Lake Huron Primary Water Supply System also services: Bluewater, South Huron, Lambton Shores, North Middlesex, Lucan Biddulph, Middlesex Centre and Strathroy-Caradoc. The Elgin Area Primary Water Supply System also services Bayham, Aylmer, Malahide, St. Thomas, Central Elgin, Southwolld, and Dutton/Dunwich. Notably, three First Nations fall within the catchement area of both of these networks are are not serviced by either: Chippewas of the Thames First Nation, Oneida Nation of the Thames and Munsee-Delaware Nation.

This map provides a specific example of water sharing taking place through two large distribution networks around the City of London, Ontario. This unique case example helps to motivate our research question. 

Several municipalities in this area are serviced by two large regional water distribution networks: the Lake Huron Primary Water Supply System (LHPWSS) and the Elgin Area Primary Water Supply system (EAPWSS).


Situated at the centre of these two regional supply networks are three First Nations: Chippewas of the Thames First Nation, Oneida Nation of the Thames, and Munsee-Delaware Nation. Despite being situated within the service areas of these regional supply networks, these First Nations were not engaged in water sharing through these networks during the study period (and are currently not engaged in water sharing through these networks). 

Our empirical dataset is comprised of 710 water systems in Ontario: 565 municipal water systems, and 145 First Nations water systems. Of those 710 water systems, 202 were engaging in water sharing with at least one neighbour during our study period of 2009-2010.


Some communities have more than one water system servicing their population. After aggregating to the community level, we found that 134 (approximately 32%) of communities in our data set were participating in water sharing during our study period. This map highlights communities participating in these arrangements. Only 12 of these communities are First Nation communities.

Map of Ontario census subdivisions. Red polygons are communities recieving water from a neighbour in the years 2009-10. Most of the extent of this sharing is concentrated in southern Ontario.

Water Sharing Arrangements (WSAs) in Ontario

In Ontario, municipalities are local governing units created by the Province, with rights and duties defined by provincial legislation: namely, the Municipal Act (2001). The Municipal Act defines municipal responsibilities with respect to community services like drinking water provision. Some municipalities in Ontario are placed within WSAs per force of law - we refer to these arrangements as rationing arrangements. Other municipalities have to negotiate in order to enter into WSAs - we refer to these arrangements as bargaining arrangements.

The figure below provides a summary of the number (and proportion) of First Nations and municipalities in our data set participating in WSAs during our study period.

Transaction Costs

Importantly, First Nations have a relationship with the federal government, and are not subject to Provincial legislation when it comes to the provision of services in their communities. For this reason, First Nations must always bargain to enter into a WSA. This process of bargaining comes with many costs, which are commonly referred to as transaction costs. We expect the transaction costs of 

bargaining to engage in WSAs to be higher for First Nations communities than for municipalities, for two key reasons:

  • Trust and social capital: while there is great diversity among First Nations across Ontario, all First Nations differ from neighbouring settler communities in important ways - with respect to nationhood, culture, systems of governance, and the experience of historic and contemporary colonial violence. These factors can impact the process of relationship building between First Nations and their neighbours. In contrast, there is much greater homogeneity among municipalities, which serve only as local governing units created by the province, with rights and duties defined through a similar provincial legislative process. This homogeneity among municipalities may make relationship building between municipal governments less costly. 

  • Governing institutions: the Municipal Act (2001) grants Ontario municipalities the rights of "natural persons" within their spheres of jurisdiction, including water services. In contrast, First Nations must operate within the constraints of the Indian Act (1985), and the federal oversight and many restrictions it imposes on community decisions making. 

Image of a bar graph showing the relative count and percentage of First Nations vs. Ontario municipalities sharing water as of 2009-10. There were 12 First Nations sharing water at this time, 10.2% of First Nations. There were 122 municipalities sharing water at this time, 40.5% of municipalities.

To best assess our research question, we include a number of key community characteristics in our analysis. Summary statistics for these community characteristics are provided in the table to the right. 

Table showing summary statistics for key variables in the analysis. Community located in northern ontario: mean is 0.428, standard deviation is 0.495, minimum is 0 and maximum is 1. Population density in 100s of people per square kilometer: mean is 1.9, standard deviation is 4.08, minimum is 0.003 and maximum is 39.72. Distance to closest neighbour with a water system (kilometers): mean is 11.14, standard deviation is 9.07, minimum is 0.132 and maximum is 178.42. Elevation in 10s of meters above sea level: mean is 25.06, standard deviation is 9.07, minimum is 1 and maximum is 49. Census division median income in $1000s per person: mean is 25.80, standard deviation is 2.6, minimum is 19.89 and maximum is 35.43.

Key Findings

We find that across several different model specifications, First Nations are consistently less likely to be engaged in a water sharing arrangement compared to municipalities.¹ One explanation of these results is that First Nations do in fact have relatively high transaction costs of bargaining for WSAs compared to municipalities. 


Another important observation is that WSAs are much less likely to emerge in northern Ontario, where 82% of First Nations communities are located. Additionally, in our examination of municipal water services across the province we identified 54 municipalities (approximately 18%) that were placed within rationing WSAs by the province. We attribute these rationing WSAs to a provincial institutional innovation designed to significantly reduced transactions costs associated with shared water provision among municipalities.


¹The magnitude of this result is sensitive to model specification, and may be influenced by continued research. However, the sign has remained consistent across all models estimated. 

Limitations and Importance of Qualitative Research

There are numerous critical complexities surrounding the issue of WSAs that we can only provide context to with these findings, and we cannot fully address. As outlined above, there are key institutional differences between how municipalities and First Nations are governed, producing different constraints and costs of transacting with neighbouring communities. Additionally, existing social capital and relationship building between First Nations and neighbouring municipalities can vary significantly from place to place, impacting the perceived benefits and desirability of collaborative relationships. The qualitative case study component of this research plays an essential role in shedding light on these crucial situational factors that cannot be captured by this broader empirical analysis.


We find evidence that First Nations are less likely to engage in WSAs compared to municipalities. As discussed above, transaction costs – e.g., the costs of gathering information, negotiating, etc. – may be a barrier to WSAs. There are many ways that these transaction costs could be reduced. Most simply, facilitating increased interactions between communities may reduce information costs associated with pursuing opportunities. More focused policy innovations might seek to reduce the burden of bargaining for WSAs that First Nations face, while respecting their autonomy over community water provision. For example, a policy innovation that requires municipalities to provide water to First Nations if requested could result in a shift in municipal planning decisions that would take into account First Nations access rights to potable water from shared water sources. While we are not in a position to evaluate a legal change such as this, we recognize there are a spectrum of ways to reduce transaction costs that are worthy of consideration.


Importantly, transaction costs are not the only costs associated with WSAs that are deserving of more careful exploration. There is a need for more localized analysis of the short and long term financial costs to First Nations associated with WSAs, compared to alternative water provision options. These costs will vary by community. Therefore, more focused research might focus on the relative costs associated with WSAs for First Nations in close proximity to municipalities.

bottom of page