THE ROLE OF SCIENCE IN CONTEMPORARY CANADIAN ENVIRONMENTAL DECISION MAKING: THE EXAMPLE OF ENVIRONMENTAL ASSESSMENT.

• Author: Westwood, A.R.
• Position: Special Section: Canada and Climate Change

1. INTRODUCTION

   In the fall of 2015, Canada adopted the United Nations' 2030 Agenda for Sustainable Development (the "Agenda"). (1) The Agenda consists of 17 Sustainable Development Goals ("SDGs", see Figure 1) and envisions "a secure world free of poverty and hunger, with full and productive employment, access to quality education and universal health coverage, the achievement of gender equality and the empowerment of all women and girls, and an end to environmental degradation." (2) Shortly thereafter, Canadians elected a new political party to government whose campaign included a promise to strengthen Canada* environmental laws after nearly a decade of weakening. (3) One of the government* first official acts was to sign and champion the 2015 Paris Agreement on climate change. (4)

   As of spring 2018, however, the status of these initiatives appears uncertain. With respect to the SDGs, the Commissioner of the Environment and Sustainable Development reported on 24 April 2018 that Canada had not developed a formal approach to implement the 2030 Agenda and the sustainable development goals.... [and] despite some specific action at the departmental level, there was still no federal governance structure... [and] no implementation plan or system to measure, monitor, and report on the progress in achieving the goals." (6) Similarly, and notwithstanding over two years of study, Canada's federal environmental-law regime appears on track to retain many of its dominant--and negative--characteristics, including excessively discretionary government powers, a failure to meaningfully tackle the challenge of cumulative environmental impacts, and a lack of a strong commitment to science that is open, rigorous, and impartial. (7)

   In this article, we examine the role of science in Canada's federal environmental assessment (EA) (8) regime to illustrate opportunities for improvement. We do not address the application of science in EA practices (that is, how to do good science within EA processes), which has been thoroughly reviewed by MacKinnon et al. (9) Instead, we examine the context for science in EA law: we examine the components of a regulatory regime, enshrined by law, that would allow for scientifically defensible assessments and evidence-based decision making. We have four objectives: (1) to provide a recent history of the role(s) of science in Canada's legislated EA regimes, including public support for science in EA law; (2) to propose five components necessary in an EA regime to ensure strong inclusion of science; (3) to evaluate if new proposed legislation meets scientific standards for modern EA, particularly cumulative effects and climate change; and (4) to encourage collaboration between scholars and practitioners in law and the natural and social sciences to work towards stronger scientific foundations in Canada's EA regimes at all levels.

2. SCIENCE AND EA IN CANADA

   1. THE FUNDAMENTAL ROLE OF SCIENCE IN EA

      EA laws in Canada have changed over time and among jurisdictions, but rarely have they been used to prevent negative environmental impacts outright. Rather, and as is the case with the western world's first modern environmental assessment law (the United States' National Environmental Policy Act (10)) the primary goal of EA laws has been considerably less ambitious: "to force agencies to consider the environmental effects of their actions and to provide a means to involve and inform the public in federal agency decision-making." (11) In simpler terms, laws like the NEPA "merely prohibit uninformed--rather than unwise--agency action." (12)

      Arguably, Canada's succession of EA laws has not been vastly different. As was noted by the Supreme Court of Canada (SCC) in its landmark 1992 decision, Friends of the Oldman River,, (13) in which it interpreted the then-applicable Environmental Assessment and Review Process Guidelines Order: (14)

      Environmental impact assessment is, in its simplest form, a planning tool that is now generally regarded as an integral component of sound decision-making.... As a planning tool it has both an information-gathering and a decision-making component which provide the decision maker with an objective basis for granting or denying approval for a proposed development. In short, environmental impact assessment is simply descriptive of a process of decision making. (15) The critical assumption underlying such EA regimes is that identifying potential environmental (and other) impacts in a transparent and rigorous manner should lead to decision making that better accounts for environmental, social, and health impacts, including through the potential for political or democratic accountability for such decisions. (16) The following description by the Federal Court of Canada, which can be applied equally to all of Canada's EA regimes (past, present, and future), captures this intended dynamic well: "In short, Parliament has designed a decision-making process... that is, when it functions properly, both evidence-based and democratically accountable." (17)

      Instead, Canada's EA regimes have garnered much criticism, including for weaknesses regarding the requirements for, and the quality of, science and evidence in the EA process. (18) In an effort to address these and other shortcomings, the Government of Canada is currently in the midst of reforming the existing federal EA regime, the Canadian Environmental Assessment Act, 2012. (19) The CEAA 2012 itself is the product of widely panned reforms made just six years ago by the previous Conservative government. (20) Within this process, there have been repeated calls from scientific, legal, environmental, and Indigenous communities to improve the scientific basis of EA. (21)

      Although each EA regime in Canada is unique, all can be understood as consisting of four general phases: (1) a planning phase where basic information about a project and its potential impacts, as well as the assessment that will be conducted, are shared and discussed with stakeholders; (2) an assessment phase, where baseline social, environmental, and health conditions are measured and potential project impacts are estimated (that is, "environmental impact statement"); (3) a decision-making phase, where the relevant authority considers environmental (and other) potential impacts and makes decisions (for example, to approve or reject a project or establish conditions of approval); and, ideally, (4) a learning or adaptive management phase, where project follow-up and monitoring occurs, and EA processes at large are evaluated and adjusted in a recursive manner.

      Conventionally, the scientific method includes identifying a hypothesis to be tested, including associated predictions and assumptions, making observations, and analyzing whether the observed results match the predictions. (22) In the EA context, hypothesis testing manifests through the predicted necessity or effectiveness of mitigation measures. For example, if mitigation is deemed unnecessary, it is because the proponent hypothesizes that the project will have negligible environmental impacts. If mitigation is prescribed, and the proponent deems a project to have negligible (that is, nonsignificant) residual impacts, it is because the mitigation is hypothesized to work in the manner predicted by the proponent. Equally relevant to hypothesis testing in both a purely scientific and EA context, are evaluations of data reliability (for example, accuracy, precision, independence, and efforts to reduce sampling biases), sample size, effect size vis-a-vis statistical significance, and statistical power. Scientific methods are also used to estimate the costs and consequences of negative environmental impacts on species, ecosystems, and communities, as well as the expected costs to humans.

      Scientific evidence (23) is a critical element in each of the four EA phases. During the planning and assessment phases, much of the evidence used to determine baseline conditions, identify environmental sensitivities (for example, species at risk and water quality), and make potential impact estimations is derived using approaches from the natural and social sciences. Though scientific evidence is only one of many considerations (for example, economic, social, political), it is used to inform decision making regarding whether an industrial project will be approved, and under what conditions. Scientific information, evidence, and theory are also used in the recursive component of adapting and improving EA processes.

   2. A RECENT HISTORY OF SCIENCE AND EA LAW IN CANADA

      Although some form of EA has taken place in Canada since the 1930s, federal legislation was not enshrined until 1995. In this section, we briefly describe the evolution of relevant federal statutes, (24) each of which contained different provisions and opportunities for including science. Importantly, none of the federal EA statutes implemented to date have ever contained explicit provisions with respect to science. This stands in contrast to several other federal environmental laws. As one example, the preamble to the Canadian Environmental Protection Act, 1999 (25) "recognizes the integral role of science, as well as the role of traditional aboriginal knowledge, in the process of making decisions relating to the protection of the environment and human health". (26) CEPA 1999 also imposes a duty on the federal government to "apply knowledge, including traditional aboriginal knowledge, science and technology, to identify and resolve environmental problems". (27) Most concretely, section 76.1 requires the Minister of Environment, when "conducting and interpreting the results of" various toxicity assessments pursuant to Part 5, to "apply a weight of evidence approach and the precautionary principle." (28) In Goodyear Canada Inc v Canada (Minister of the Environment), the Federal Court described one such assessment as "a statutorily mandated scientific evaluation of... a...