Socioeconomic LCA of Milk Production in Canada

Abstract

Over the years, the agricultural sector, and the livestock and dairy sectors in particular, have been increasingly criticized for their environmental impacts, especially with regard to greenhouse gas emissions. At the same time, there has been a growing awareness that farm activities equally induce significant social and economic impacts over a wide range of stakeholders. In order to face the new challenges arising from this context and to clarify the path towards sustainable milk production in Canada, the Dairy Farmers of Canada (DFC) commissioned the realization of a Social and Environmental Life Cycle Assessment (SELCA) of Canadian Milk. Launched in 2010, this project, which ended in September 2012, was conducted as part of the Dairy Research Cluster. The study was conducted by three partners, two consulting firms (Groupe AGECO and Quantis) and a research center (CIRAIG), based at the Montreal Polytechnic, with a section dedicated to socioecomic life cycle assessment based at the University of Quebec in Montreal. It aimed at providing a comprehensive assessment of the Canadian milk production sector with respect to sustainability. The main deliverables include an environmental profile of the average kilogram of milk produced in Canada, as well as an evaluation of the socioeconomic performance of the Canadian dairy sector. This chapter addresses the social and socioeconomic dimensions of the global project. It presents the methodological choices made, such as combining a specific analysis and a potential hotspots analysis (PHA) for two parts of the system under study. It then presents the economic contributions of the Canadian dairy sector, which has generated over 127,000 direct, indirect and induced jobs in 2009, contributed approximately $7.2 billion to the national GDP, and procured almost $1.4 billion in total tax revenue. Canadian dairy farmers are also corporate citizens whose behaviors—individually and collectively—impact their stakeholders. This SLCA provides a detailed picture of this socioeconomic performance. It appears from this assessment that the Canadian dairy farms perform positively overall. The dairy farmers’ engagement towards their local communities is significant, with the vast majority involved in their communities in many different ways. However, more could be done in terms of cohabitation, with producers adopting practices minimizing the spreading of odors, for example. The picture is also contrasted with regard to farm workers. Although dairy farmers provide overall working conditions that go beyond labor standards—to which they are mostly not legally subjected—there is room for improvement regarding various issues, such as professional training and communication of working conditions. The same holds true with respect to their suppliers and business partners, given that a majority of dairy producers do not usually consider their suppliers’ performance in regards to social responsibility in their procurement decisions.

AGECO is a Quebec (Canada)-based consulting firm established in 2000 as a spin-off from Laval University in Quebec City by a group of professors well recognized in Quebec and Canada in the domain of socioeconomic analysis applied to the agrifood sector, natural resources, and the environment. AGECO performs impact assessment studies, policy and regulatory analyses, socioeconomic studies, surveys, structural analyses, studies of management tools as well as strategic channel planning. First and foremost, AGECO is a team trained in economics and the social sciences, specializing in agrifood, and natural and environmental resources. The team is known for its ability to understand the socioeconomic, political and strategic situations. Over the last 5 years, AGECO has developed expertise in SLCA, both in theory and practice.

The Interuniversity Research Centre for the Life Cycle of Products, Processes, and Services (CIRAIG) was founded initially by the École Polytechnique de Montréal, in collaboration with the Université de Montréal and HEC Montréal. The CIRAIG was created to meet the demands of industry and governments to develop leading-edge academic expertise on sustainable development tools. The CIRAIG now includes a team from the Department of Strategy, Social and Environmental Responsibility that is located within the School of Management Sciences of the Université du Québec à Montréal (UQÀM). This team deals specifically with the social and socioeconomic dimensions of life cycle assessment. The CIRAIG is the only university research centre on life cycle in Canada; it is also one of the largest internationally. It hosts the International Life Cycle Chair, supported by 14 industrial partners.

1 Introduction

In an effort to clarify the path towards sustainable milk production in Canada, the Dairy Farmers of Canada (DFC), through the Dairy Research Cluster, a part of the Canadian Agri-Science Clusters Initiative of Agriculture and Agri-Food Canada (AAC), commissioned an environmental and social life cycle assessment (SLCA) of Canadian milk. This study was carried out by Quantis Canada, AGECO, in collaboration with The Interuniversity Research Centre for the Life Cycle of Products, Processes and Services (CIRAIG), and the results were published in 2012 (Quantis Canada, Ageco and CIRAIG, 2012, Environmental and Socioeconomic Life Cycle Assessment of Canadian Milk, DFC, 285 pages). This project is the basis of the case study that we are presenting in this chapter and, as we will see, there was an exploratory dimension in the objectives as it was the first time that the DFC were commissioning an LCA, and, furthermore, an integrated environmental and socioeconomic LCA.

The project’s objectives were threefold:

• (1) To evaluate the environmental and socioeconomic impacts of dairy production in Canada;

• (2) To identify potential areas of focus for further improvements of the dairy sector’s sustainability; and

• (3) To provide the framework and the building blocks to support comparison and benchmarking (in reference to milk production in other countries, for example).

The results of this environmental and socioeconomic life cycle assessment were meant to be used by DFC for decision-making at a macro level, but also for communication purposes with all stakeholders (dairy farmers, policy makers, processors, consumers, media, etc.). The results will also serve as a basis for the sustainability agenda of the farmers’ association.

This initiative took place within a context where many relevant actors of the industry have been active on the international scene. At the international level, the International Dairy Federation (IDF) promotes the sustainable production of milk and milk-based products through its Dairy Sustainability Framework and the production of a methodology for the lifecycle assessment for the dairy sector.

Many associations of milk producers and governments have already reported the results of LCAs of milk production, including the European Dairy Association, which commissioned a carbon footprint across the EU dairy sector (Sevenster and De Jong 2008), as well as the Swedish Dairy Association, the Australian Dairy, and the US Dairy Management Inc. In France, an upcoming policy towards environmental labelling of products under the “Grenelle Environment Forum” has accelerated the implementation of LCA in various consumption products, including food and dairy. Furthermore, the FAO also completed a carbon footprint in 2010 with a global perspective over the entire supply chain, and there is a continuous process for improvement in place. Because of the wide scope of the study however, numerous assumptions and generalizations were needed.

Companies having performed and communicated on LCAs of their dairy products include Danone in France, Arla in Sweden and Denmark, Fonterra in New Zealand through a national investigation (Lundie et al. 2009), Aurora Organics in the US in 2007, and Cadbury in England in 2008. In Canada, Liberté has been active in LCA for many years and publishes information on their Web site (Liberté 2012). These studies are sometimes limited to a few farms only, which does not imply a small herd, as the Aurora Organics study involved six farms only and a total herd of close to 12,000 cows.

However, there is little to no literature surveying the social or socioeconomic aspects of sustainability in dairy. The need to do so has been noted in certain documents, such as in the Life Cycle Initiative Program for the United Nations (Grießhammer et al. 2006) and the IDF review of literature, which noted that “Future research will possibly enable inclusion of social issues in LCA to create a new impact category. The social conditions of workers could be accounted for at farms as well as dairies or retail phase” (IDF 2009).

This chapter is directly derived from the full report, with a formal authorization of representatives of the DFC, but it will concentrate only on the social and socioeconomic LCA part of the study. In particular, we wish to stress that all tables and figures come from the report and therefore are not referenced individually to this report (For a detailed presentation of the environmental LCA and the socioeconomic LCA, please refer to the full integrated report at http://www.groupeageco.ca/PLC_EnvironmentalAndSocioeconomicLCA_FullReport.pdf).

After this introduction, the chapter is divided into four main sections. In Sect. 2 we consider a series of definitions related to SLCA and qualify the approach selected for the study. This will expectedly cover the boundaries, the system under study, and the assumptions made in defining the approach. Then we present the two different types of analysis that we will use for two components of the Canadian Milk Production System. First, the “specific analysis” that will apply to the farm level, for which we have gathered primary data (Sect. 2.2). Then we will present the various stakeholder categories used and the impact of the categories that we considered for these different stakeholders, continuing with the impact assessment methodology and the data collection process. Secondly, in Sect. 2.3 we deal with the generic part of the study—that is, the potential hotspot analysis. The same elements will be considered in this subsection as that in the previous one. In Sect. 3 we present the results of both assessments and discuss them as well as the challenges met in the study in Sect. 4. Section 5 deals with the main conclusions and possible future steps.

2 Social and Socioeconomic Life Cycle Assessment: Definition and Approach

SLCA is a “technique that aims to assess the social and socioeconomic aspects of products and their potential positive and negative impacts along their life cycle” (UNEP/SETAC 2009, p. 37). The main features of this tool are its broad scope, which encompasses a product’s entire life cycle, and its assessment method, which relies on benchmarks to assess the relative social performance of the organizations (private, public, or non-profit) involved in the product’s life cycle.

The SLCA methodology relies on the recently developed Guidelines for Social Life Cycle Assessment of Products (hereinafter the Guidelines). Published in 2009 by the United Nations Environment Programme (UNEP) in collaboration with the Society of Environmental Toxicology and Chemistry (SETAC), these Guidelines provide the general framework needed to conduct such an assessment.

The Guidelines propose a classification of the main socially significant themes to assess, as well as a categorization of the main stakeholder categories potentially affected by the socioeconomic impacts induced by the activities and behaviors of the organizations involved in the product’s life cycle. Six main impact categories are listed in the Guidelines, each one related to a number of impact subcategories, or specific issues of concern, which are “socially significant themes or attributes” to assess (UNEP/SETAC 2009, p. 44). These impact categories are: human rights, working conditions, health and safety, governance, cultural heritage, and socioeconomic repercussions. As for the stakeholder categories, the Guidelines list the following five groups: workers, local communities, society, consumers, and value chain actors.

In addition to this general framework, the Guidelines also specify the steps to follow and the requirements to fulfill in order to conduct a rigorous and transparent assessment. However, the Guidelines are a work in progress towards the elaboration of a comprehensive assessment framework. Adaptations are admittedly needed in order to perform an SLCA (UNEP/SETAC 2009, p. 82). For instance, the Guidelines do not define any particular assessment methodology, so it was necessary to develop an “assessment framework,” compatible with the Guidelines in order to perform the SLCA of milk production in Canada. The following sections thus describe this framework and present the methodological underpinnings on which it is based. When needed, the adjustments made to the general framework provided by the Guidelines are discussed.

The first step of an SLCA aims to describe the intended application and the reasons for carrying out the study (goal) and to define its depth and breadth (scope). As highlighted in the Guidelines, “the ultimate objective for conducting an SLCA is to promote improvement of social conditions and of the overall socioeconomic performance of a product throughout its life cycle for all of its stakeholders” (UNEP/SETAC 2009, p. 50). This is also the project’s main objective: assessing the socioeconomic performance of the Canadian milk production sector and identifying potential social hotspots to provide some recommendations in order to improve the system’s overall socioeconomic performance towards its stakeholders.

As for an ELCA, this implies identifying the functional unit, the product system, and its boundaries (UNEP/SETAC 2009, pp. 51–57). The UNEP/SETAC Guidelines do not provide any particular direction on how the scope of an SLCA should be adapted to fit that of an ELCA when both assessments are conducted together. It is acknowledged, however, that given the SLCA’s specificities, the scope might not necessarily be the same or totally integrated.

As the objective of the Canadian Dairy Farmers is to study not only the production of the milk but also its transportation at the gate of the processing facility, excluding the transformation, the functional unit for the ELCA part of the study is:

1 kg of fat and protein corrected milk (FPCM) from a Canadian farm, to the processing facility

We used it for the SLCA as well as for the sake of similarity in the development of the two components—social and environmental—of the project.

2.1 Boundaries and Assumptions

For the purposes of this analysis, the system was grouped into five principal life cycle stages, as presented in Fig. 1.

Fig. 1

Life cycle system

• (1) Feed Production: includes manure spreading, pesticide and fertilizer production and spreading, any energy required (diesel) for field manipulations, irrigation water.

• (2) Livestock Management: includes bedding, drinking water, milking equipment, cleaning products and water, ammonia emissions from housing, and methane emissions from enteric fermentation.

• (3) Manure Management: limited to emissions of nitrous oxide, methane and ammonia from storage.

• (4) Energy and Buildings: includes electricity for dairying, cattle housing and milk parlor equipment and buildings, and gasoline for regular operations.

• (5) Transportation: includes only purchased feed transportation, purchased animal transportation and raw milk transportation to processor.

Within each of these stages, the LCA considers all identifiable “upstream” inputs to provide as comprehensive a view as is practical of the product system. For example, when considering the environmental impact of transportation, not only are the emissions of the truck considered, but also included are the impact of additional processes and inputs needed to produce the fuel, as well as truck and tire manufacturing. In this way, the production chains of all inputs are traced back to the original extraction of raw materials, within feasible limits.

However, the product system differs slightly between an SLCA and an ELCA, firstly in its constituting parts: Since an SLCA primarily focuses on the behavior of the organizations involved in the product’s life cycle, an SLCA product system is made of those organizations, organized in value chains, rather than by the processes they perform as in an ELCA. Secondly, it differs in its scope: For a matter of simplification and access to data, the scope of an SLCA product system is usually circumscribed to include only the most important and relevant value chains and organizations, where the product system in ELCA is more exhaustive and usually extended until no more exchanges are made between processes inside the technosphere.

Hence, the definition of an SLCA product system first requires identifying the organizations involved in each value chain included in the product’s life cycle. In an SLCA perspective, a value chain can be defined as a set of businesses located whether upstream or downstream of an organization, providing the inputs and services needed for the production and the marketing of the product under assessment. Then, depending on the objectives of the project, criteria are set to delimit the scope and the range of the system under study.

The above considerations have been taken into account to specify the product system used to perform this SLCA of milk production in Canada. Based on the information provided by the Milk Cost of Production Database,¹ it was possible to define the main value chains involved in milk production according to the inputs and services they provide to the dairy farms.² Given the vast array of inputs and services involved, decisions were made to further circumscribe the scope of the system. First, inputs related to farm buildings are excluded from the system, because this group of expenses is related to various kinds of tools, materials and services of low individual significance. Cow replacement is also excluded, given that these animals are generally traded among dairy farmers. Items only related to services, such as salaries, joint marketing plan management fees and field equipment maintenance expenditures, and those not directly associated to milk production, such as interest fees and taxes, are also excluded. Although milk transportation is a service, it is left within the system since it is part of its scope. Finally, it was decided to exclude “electricity” from the system and to include “pesticides,” although it accounts only for 0.4 % of the average total cost. These choices are justified by the fact that electricity is a relatively minor and non-agricultural input from which suppliers are globally disconnected from the agricultural sector (Parent et al. 2012), whereas pesticides are an economically and socially sensitive product primarily used in agricultural production. According to these choices, the following inputs and services are therefore included in the SLCA system:

• Animal feed

• Farm inputs (fertilizers, seeds, pesticides)

• Milk transportation

• Veterinary services (drugs and bovine semen)

• Agricultural machinery

• Fuel and diesel

Each of these inputs and services is provided to dairy farms via a specific supply chain composed of a number of steps (from extraction of raw material to final distribution). Each step involves a vast number of businesses producing products or providing services. In order to simplify the system, cut-off criteria have also been used to limit the length and complexity of each of these value chains:

• For each value chain, only one to two representative inputs or services have been considered at each step, according to their relative importance at this step.

• The range of each value chain was extended, as long as it was possible to trace back a main input or service used in the production of the previous product or service.

Figure 2 shows the product system selected for the SLCA study. First tier suppliers, i.e., businesses or value chain actors directly interacting with dairy farmers for advice or commercial purposes related to the selected inputs, are shown to the left of dairy farms. They include advisers or representatives, such as feed and farm inputs dealers, whether or not affiliated to specific companies involved in the production or the handling of some inputs. Upstream are listed the selected inputs sold to dairy farmers (or used to supply the services) and the main auxiliary inputs needed to produce them. Taken together, these inputs, auxiliary inputs and the companies producing and handling them, shape the product system considered to perform this SLCA.

Fig. 2

Product system of the Canadian milk production

Although the aim of an SLCA is to provide, for a given product, a profile of the socioeconomic performance of the organizations involved in its entire life cycle, the assessment’s degree of details can vary across the system. It is not always readily possible, necessary or even relevant, to assess in detail the behavior of all the organizations throughout the life cycle of a product. While practical constraints such as data limitations, short delays or budget restrictions can impede in-depth analysis, the assessment’s focus is generally determined by the intended applications of the SLCA results by the commissioner (Parent et al. 2012).

In the case of this study, the objective of the SLCA is to give a socioeconomic profile of the product system with an emphasis on the Canadian milk production sector. Therefore, the socioeconomic performance of the Canadian dairy farms and their sectorial organizations are assessed through a specific analysis—which provides a high level of details on their degree of social responsibility based on the compilation of primary data collected on-site.

For the rest of the product system, a potential hotspots analysis (PHA) is performed—which offers an overview on the possibility of encountering risky behaviors among the supply companies/sectors based on the compilation of generic data collected from international and national databases, the Social Hotspots Database (SHDB), human rights reports, etc.

2.2 The Specific Analysis

The aim of the Specific Analysis is to provide a detailed analysis of the socioeconomic performance of a particular company/organization/sector by assessing its degree of social responsibility toward its stakeholders. Given the focus of this project, the Specific Analysis approach is used to assess the socioeconomic performance of the milk production sector in general, and of the dairy farms and dairy organizations in particular.

Because of the structure of the Canadian milk production sector, which involves about 13,000 dairy farms across Canada that are provincially and nationally organized, the assessment addresses more specifically the socioeconomic performance of the sector at three different levels—since the behaviors and practices encountered at each level do not necessarily affect the stakeholders in the same way or do not relate to the same issues of concern. The three assessment levels are:

• Dairy farms level. The dairy farms are at the center of the assessment. Their behavior and practices affect mostly the farm workers, the local communities where they are located, and their suppliers.

• Dairy boards level. All across Canada, dairy farms are organized in provincial dairy boards performing the administrative, marketing and communicative tasks assigned by the dairy farmers. By fulfilling these tasks, those organizations induce impacts on different stakeholders.

• Sector level Milk production takes place in a legal and institutional framework that shapes most of the sector’s characteristics, which in turn have significant implications on the entire sector’s stakeholders. Whereas this particular framework is not necessarily specific to the milk production sector, or dairy producers directly accountable for it, its implications still have to be assessed as producers have the ability to act upon it together.

In this chapter we will only present the detailed methodology and results for the dairy farm level, but neither for the dairy board nor sector levels.

It is important to stress that the SLCA approach in general, and the Specific Analysis in particular, exclusively addresses the relationships between a business/organization and its stakeholders, the former being the one inducing the socioeconomic impacts—positive or negative—on the surrounding groups of individuals. Accordingly, the impacts experienced by the dairy farmers or the dairy boards resulting from their own behavior are not addressed by this framework. Rather, the assessment framework assesses the degree to which the Canadian dairy farmers and dairy boards behave in a socially responsible manner towards their stakeholders.

2.2.1 Stakeholder Categories

Formally, stakeholders are “those groups and individuals that can affect, or are affected by, the accomplishment of organizational purpose” (Freeman 1984 cited by UNEP/SETAC 2009, p. 47). As pointed out earlier, the UNEP/SETAC’s Guidelines proposes a list of five main stakeholder categories potentially impacted by the life cycle of a product. These are the workers, the local communities, the society, the consumers, and the value chain actors. However, depending on the study’s boundaries and the sector’s particularities, it is possible to add, to exclude, to differentiate, or simply to define more precisely the proposed categories to get a clearer description, at each step of the value chain, of the stakeholders involved (UNEP/SETAC 2009, p. 46).

Given the scope of this study and the focus of the Specific Analysis, such adaptation of the basic stakeholder categories was necessary. The “consumers” category (seen as the “people who buy milk in different forms from a retail store”) was hence excluded from the framework. The issues of concern potentially affecting consumers have instead been assessed in relation with the “value chain actors” category, since raw milk is the main input used by dairy processors to elaborate the dairy products sold to consumers. The other four stakeholder categories adequately cover the various groups of individuals potentially impacted by milk production activities, as shown by a review of the existing literature. Based on the results of several focus groups conducted in the first stages of the study, each stakeholder category has been defined in more detail (Table 1). Given that the Specific Analysis was exclusively conducted on the dairy farms and their boards, the categories have been adapted only to the individuals impacted by dairy activities.

Table 1 Definition of the stakeholder categories impacted by milk production activities of the Canadian dairy farms and their boards

2.2.2 Issues of Concern or Impact Subcategories

Impact subcategories are the “socially relevant characteristic or attribute to be assessed” in an SLCA (UNEP/SETAC 2009, p. 71). Based on international agreements (conventions, treaties, etc.), the guidelines already propose a list of internationally recognized impact subcategories, each being related to a specific stakeholder category. While most of the listed impact subcategories are relevant in a Canadian context, some of them, such as “delocalization and migration” or “prevention of armed conflicts,” are not necessarily relevant.

In order to encompass comprehensively the issues of concern related to milk production in Canada, and as allowed by the guidelines, the list of subcategories was justifiedly adjusted on the basis of a review of the existing literature, experts’ opinions, and the results of three focus groups conducted among the sector’s stakeholders.

Table 2 presents the impact subcategories chosen for the study. Each one is explicitly defined to ensure a common understanding of the social issue it covers. These definitions do not necessarily follow those proposed in the methodological sheets published by the Life Cycle Initiative (LCI 2010), because they do not adequately describe the issues under assessment in this specific case.

Table 2 Impact subcategories according to the corresponding stakeholder categories

A scale of assessment level is also specified, as some issues of concern relate primarily to dairy farm activities while some others relate rather to their provincial boards, or even to the milk sector as a whole. One issue of concern can be related to more than one level of assessment as well.

2.2.3 Impact Assessment Methodology

The impact assessment phase of an SLCA involves translating inventory data into measured impacts by aggregating inventory indicators within subcategories and comparing them against a so-called “performance reference point” (PRP)—or benchmark. However, as the Guidelines point out, “impact assessment methodologies are under development and SLCA is an open field for future research” (UNEP/SETAC 2009, p. 69). For instance, unlike the ELCA methodology, there is no characterization model allowing the translation of inventory indicators into socioeconomic impacts using quantitative models.

Although the Guidelines do not provide any particular indications or suggestions regarding the impact assessment methodology to use in an SLCA, this issue is extensively discussed in the socioeconomic impact evaluation literature (Burdge 2004; Burdge and Vanclay 1995; Chadwick 2002; Becker and Vanclay 2003 ). Our assessment methodology thus relies not only on this literature, but also on our expertise in this field.³

Most social assessment methods, including the SLCA methodology, rely on socioeconomic indicators to measure and assess the social and economic impacts induced on stakeholders by a particular activity. But as pointed out in the Guidelines, “several inventory indicators and units of measurement/reporting types may be used to assess each of the subcategories. Inventory indicators and units of measurement may vary, depending of the context of the study” (UNEP/SETAC 2009, p. 44). Indeed, there is no formal or universally acknowledged set of indicators to which one can refer to assess the socioeconomic performance of a particular product or company. To carry out a particular assessment, a specific set of indicators thus has to be developed according to the project’s objectives and data availability.

Based on the multiple assessment frameworks suggested in the literature—many of which have been conceived to be used in an agricultural context—but also on expert judgments, a list of indicators has therefore been developed to assess the socioeconomic performance of the Canadian milk production sector. A four-level evaluation scale was created and they specify how each indicator can be declined practically, given the PRP used.

More specifically, these evaluation scales (Table 3) allow assessing, for a given issue of concern, the level of social responsibility of a dairy farmer.

Table 3 Specific analysis’s behavior evaluation scale

A risky behavior is considered to be a hazardous practice that can cause significant damages or create serious problems to the concerned stakeholders. Given that most hazardous practices are forbidden by law, they are generally related to illegal behaviors. Yet, in some cases, it is possible to consider a particular behavior as risky (even if it is not illegal) insofar as it can potentially have serious and negative implications for the individual or group of individuals it concerns, compared to its potential benefits. This is, for example, the case with the “working hours” subcategory, as there is generally no legal limit to the length of the work week or legal standard relating to work overload in the agricultural sector. Allowing a number of working hours beyond a certain threshold can, however, have negative implications for the workers’ health and safety—irrespective of the fact that they agree to work them.

A compliant behavior refers to a normal and expected practice. It generally corresponds to a minimal legal requirement or simply to an absence of initiative or commitment in situations where it is not required. In other words, a compliant behavior means that the organization, while not acting in a socially irresponsible manner, is not especially socially responsible either.

The two other levels refer to behaviors that go beyond compliant or minimal expectations to tend toward more socially responsible behaviors. Depending on the issue and the PRP identified, a committed behavior is hence considered to be the most socially responsible practice a leading organization can reach, while a proactive behavior translates an in-between engagement; the business goes beyond legal requirement, but has not yet reached a leading behavior.

Of course, this classification is relative, as the PRPs used to determine whether a particular behavior is more or less socially responsible can evolve in time and place. In other words, today a committed behavior could become a minimal expectation in the future, or could be considered a desired behavior in another region. This evaluation scale is also dependent on data availability. In order to assess a particular behavior according to this four-level scale, it is necessary to have access to detailed information both to establish the PRPs and to assess the behavior itself.

Table 4 presents a selection of indicators used to assess the socioeconomic performance of dairy farmers, but all indicators developed are presented later in the results of the study. They are classified according to the stakeholder categories and the related impact subcategories. To ensure that the assessment framework is both clear and transparent, each indicator is detailed, using a standardized approach. First, a brief description of what each indicator measures is given; then, the PRPs—or benchmarks—against which the performance is assessed are specified (UNEP/SETAC 2009, p. 69).

Table 4 Selection of impact subcategories and the corresponding socioeconomic indicators per stakeholder categories documented at the dairy farm level

As mentioned earlier, PRPs are acknowledged social standards, norms or practices used as thresholds to distinguish, among the observed practices or behaviors, those that are socially responsible from those that are minimally expected from the organization. One indicator can be related to several PRPs, such as a national or international minimal legal standard, a “best available practice,” an average performance of a company or a group of businesses, etc. Given the Canadian milk production sector’s particularities, the PRPs have mostly been selected according to minimal legal requirements, sectorial standards and average performance, as well as best expected practices based on our own expertise of the sector. The choice of each PRP is justified for each indicator.

The Specific Analysis was conducted by scoring, at the level of each socioeconomic indicator, the behavior or practice of each participating farm. However, given that the project aimed at evaluating the socioeconomic performance of the milk production sector as a whole, and in order to preserve the respondents’ privacy, the individual scores have been compiled at the provincial level to get a weighted⁴ average score of the socioeconomic performance of the Canadian milk production sector.

2.2.4 Data Collection Process

Conducting a Specific Analysis requires a significant amount of data and information to document the PRPs and the organizations’ behaviors. Unfortunately, there are very few databases that cover and record, on a regular and systematic basis, the social and socioeconomic issues at the sector or organization level. Primary data, i.e., data collected directly from the participating businesses and organizations, are thus generally needed to undertake such an analysis.

Due to the scope of the Specific Analysis performed in this project, the data collection process was expectedly challenging. In addition to the large variety of undocumented information needed, it was also necessary to document this information in a standardized manner across all provinces in order to obtain consistent results at the Canadian level.

This challenge was first met through the PRPs. The lack of data and reliable documentation on most of the issues of concern under assessment made it difficult to assess not only these issues, but also to select standardized PRPs suited for the milk production context in each province. For that reason, most of the PRPs used have been based on experts’ judgement and on our own knowledge of the Canadian dairy sector and agricultural production.

Primary data were used to assess dairy farms’ behaviors and practices. To do so, questionnaires were sent to 817 milk producers located in six (6) provinces: Prince Edward Island, Nova Scotia, New Brunswick, Quebec, Ontario, and Alberta. The indicators in Table 4 above were at the basis of the questionnaire, together with traditional socioeconomics variables qualifying the farm. The participation in the survey was on a voluntary basis,⁵ and various techniques were used to distribute the questionnaires. In Quebec and New Brunswick, the producers participating in the annual cost of the production study carried out by the AGECO team were asked to complete a complementary questionnaire between September and November 2011. In Ontario, Nova Scotia, and Alberta, questionnaires were sent to all dairy producers, all of whom were offered $20 in compensation for sending back the completed form between March and June 2012. Three hundred (300) completed questionnaires were received. Both the sample’s size and the characteristics (number of cows, ownership, cultural practices, etc.) of the participating dairy farms in each province fairly well reflect the population they represent.

The data collected at the provincial level have been pooled and weighted at a national level to assess the average Canadian dairy farmers’ socioeconomic performance. Weighting was necessary because the provincial samples were not of relative equivalent size, and the Canadian average score has been determined by compiling, for each indicator, farmers’ individual answers. In case of a missing value for a particular question, this was taken into account by an adjustment of the size of the sample when calculating the mean. Then, the weight of each individual answer was established according to the relative size, in terms of number of dairy producers, of the respective province.

2.3 The Potential Hotspot Analysis: The Generic Part of the Study

The PHA aims to provide a screening of the socioeconomic performance of the companies involved in the product system. This assessment uses generic data, i.e., data that are not site-specific, and it is therefore easier to run than a Specific Analysis.

The PHA assesses the risk of encountering behaviors going against accepted social norms among businesses that are part of the system’s supply chains (upstream system). More specifically, this assessment method allows identifying potential socioeconomic hotspots,⁶ i.e., the presence of risky behaviors that might negatively impact groups of stakeholders. A PHA therefore provides a preliminary overview of the social issues found among a product’s supply chains to bring awareness of the socioeconomic risks related to current procurement practices and to point out issues requiring deeper analysis. It was carried out through the combination of literature survey, consultation of specific sources of information (such as Web sites) and of using the Social Hotspot Database (SHDB), a database that was under development when the study was being conducted.⁷

As for the Specific Analysis framework, the PHA framework is built upon the UNEP/SETAC’s Guidelines, which have been adjusted to be operationalized.

The stakeholder categories considered in the PHA framework are the same as those considered in the Specific Analysis: workers, local communities, society, and value chain actors. The “consumers” category is also excluded, as they are not significantly and directly impacted by the behavior of the assessed businesses operating upstream in the milk’s value chain.

2.3.1 Impact Subcategories

The PHA assesses the possibility of encountering risky behaviors according to a list of issues of concern (impact subcategories) related to a particular stakeholder category. While most issues are drawn from the UNEP/SETAC’s Guidelines, some adjustments have, however, been made in the context of the PHA.

Since the PHA framework is developed to cover a vast array of organizations operating in various countries, impact subcategories have not been adjusted to take into account specific sectorial or regional issues of concern. The reasons that subcategories have been removed or adjusted are rather related to methodological concerns. In some cases, it is due to the lack of relevant generic data necessary to assess a particular issue. Some subcategories have also been removed because they are not related to risky behaviors that could negatively impact individuals (e.g., social benefits and social security or end-of-life responsibility). When possible, those subcategories have been adjusted (or reworded) to cover social risks rather than benefits (e.g., “social benefits and social security” has been replaced by “employment insecurity”). Finally, some have been merged not only because of their similarities, but also because the subtlety between them could not be adequately captured by the PHA methodology (e.g., access to material resources, access to immaterial resources, delocalization, and migration and cultural heritage have been merged).

To perform a PHA it is first necessary to identify and localize the companies involved at each step in order to document their behaviors afterwards. The product system defined earlier identified nine (9) main supply chains associated with milk production. Each supply chain has been defined by identifying only one or two representative inputs and by limiting its range up to the last identifiable major auxiliary input.

In order to assess the presence of potential social hotspots, the PHA refers to proxies such as representative sectorial practices or frequently observed behaviors, informing on businesses’ behaviors. According to Macombe et al. (2010), “companies belonging to one industry tend to become similar with time.” Therefore, one can assume that the information gathered at a sector or industry level is a representative proxy of individual behaviors of the companies operating in that sector or industry.

Moreover, given that the legal and cultural context can influence businesses’ behavior, it is also important to specify where the companies, sectors or industries assessed carry their operations. As one product or input supplied to the Canadian market can come from several countries, only the main or outweighing sourcing countries for each input have been taken into consideration, in line with Bienge et al. (2010). As a consequence, the possibility of encountering businesses behaving inappropriately (or in a risky way in comparison with the commonly accepted social norms) has been assessed, at each step of each supply chain, at the sector level and in the different countries where the companies are supposed to carry out their activities.

For this purpose, the relevant representative sourcing regions have been specified. To do so, the relative weight of imports, compared to the domestic consumption level, has been calculated to make, first, an assumption on whether the supply of each input is mostly ensured by the domestic market or by a foreign one.⁸Then, countries supplying the Canadian market have been identified using a trade database.⁹

2.3.2 Data Collection Process

The PHA approach relies on generic data and is thus dependent on their availability. In order to document potential risky behaviors among supply chains, three complementary data collection techniques were therefore used, depending on the information needed.

First of all, when available, data on potential behaviors in a specific sector located in a specific country have been collected from national and international statistical databases, country-specific human rights reports, and from a variety of other sources identified through a Web search and a literature review.

While data collected at the sector level are relevant proxies to document behaviors of specific companies, they are generally scarce. To fill in this gap, another proxy was used; it involved documenting behaviors of a small sample of companies belonging to the sector and localized in the country under assessment. Samples were built by identifying the major businesses operating in the sector/country under assessment by using, for example, the Canadian Industry Statistic database.¹⁰ Information on those businesses’ behavior was also collected from human rights literature and other sources. The Business and Human Rights Resources Centre¹¹ collects articles on businesses’ practices related to human rights issues, and Wikipedia also compiles information on social issues related to specific companies; those two sources were systematically used. As the goal of the PHA is to highlight the risk of encountering potential hotspots, it was not necessary to validate the collected information at the field level.

Finally, when no data were available, either at the sector level or by referring to the sample of companies, the social performance of the country was used as a proxy. It is acknowledged that the national context in which a business carries out its activities greatly influences its behavior (Macombe et al. 2010).

In summary, for each step of each supply chain under assessment, three proxies were used to collect data giving insight on the potential behavior of companies:

• 1. Sectorial data;

• 2. Information related to the behavior of a sample of representative businesses; and

• 3. Country-level data.

2.3.3 Impact Assessment Method

This section details how the possibility of encountering companies not behaving in compliance with accepted social norms was assessed. As for the Specific Analysis, each issue of concern was assessed using an assessment method. Since the PHA relies on generic data, the method varies according to their availability. For some issues of concern, it was possible to document behaviors at a business or sectorial level. For others, information was only available at a national level. Depending on sources, quantitative, semi-qualitative and qualitative data have also been used. But in all cases, the assessment was carried out using a standardized three-level evaluation scale assessing the possibility (low, moderate, high) of encountering companies with risky behavior, i.e., not behaving in compliance with the accepted social norms (Table 5).

Table 5 Risk evaluation scale

The following tables describe the method used to assess the possibility of encountering enterprises with non-complying behaviors for each issue of concern, depending on how the indicators have been documented. When more than one source of data could have been used to assess the level of risk related for a same issue of concern, only the most relevant, i.e., the most closely related to the sector, was used. Sector- specific data, as well as data collected through a sample of companies, have been favored because they constitute better proxies of businesses’ behavior than country-level data. We relied on a country-level indicator only when no sectorial data were found using available statistical databases or a Web review. But given the current scarcity of information regarding companies’ or sectors’ behavior, the assessment relied mostly on country level indicators.

2.3.4 Sectorial Data

The issues of concern have first been documented using sectorial data collected from three different sources. In the case of fair salary, working hours and occupational health and safety, statistical data at the sector level have been used to assess the possibility of encountering social hotspots. Table 6 describes the indicators developed as well as the PRPs considered to assess the level of risk.

Table 6 Risk evaluation scale

The issues of freedom of association and collective bargaining, child labor, working hours, forced labor and occupational health and safety have also been assessed at the sector level using information found in two human rights reports: the US Department of State Country Report on Human Rights (U.S. Department of State 2011), and the Annual Survey of Violations of Trade Union Rights 2011 (ITUC et al. 2012). The qualitative information provided in those reports has been used to assess the possibility of encountering violations in the sectors under assessment.¹² The assessment method used to differentiate the risk level relied on our expert judgment. For a matter of transparency, this judgment is always justified in the “detailed justifications” sections found in Appendix J of the full report.

Finally, a web search has been conducted to document all issues of concern at a sector and country level. The collected information was assessed based on our expert judgment and transparently detailed in the “detailed justifications” sections found in Appendix J.

2.3.5 Data Collected from a Sample of Businesses

To complement the sectorial data, a review of the available publications was conducted to document, for each issue of concern, the potential risky behaviors of the main companies involved in the sectors and regions under review. This review focused on the criticisms directed towards the businesses included in the sample for practices going against accepted social norms. Here again, the collected information was assessed based on our expert judgment and transparently detailed in the results sections.

2.3.6 Country Level Data

Finally, for issues that could neither be documented through the sector-level assessment nor through the sample of businesses, country-level data were used. The possibility of encountering companies behaving inappropriately compared to accepted social norms was assessed using social indicators selected from several sources.¹³

Three main sources of data have been used:

• The World Economic Forum’s (WEF) annual Executive Opinion Survey, whose results are published in The Global Competitiveness Report 2011–2012 (WEF 2011);

• The SHDB; and

• a variety of other sources, such as the GINI and the Corruption Perception Index.

Some issues of concern were assessed using data collected from the WEF Annual Executive Opinion Survey. This survey, published in The Global Competitiveness Report 2011–2012 (WEF 2011), asks business executives about the situation in their respective countries regarding several socioeconomic issues, some of them similar to those addressed in the PHA. For each issue, the survey respondents’ opinion was scaled from 1 to 7, with 1 representing the worst situation and 7 the best; the score represents the average opinion.

Table 7 lists the WEF indicators to which we referred in the PHA. The assessment method, which is similar for all indicators, is described below.

Table 7 Indicators of the WEF annual executive opinion survey

Table 8 presents the list of indicators selected from the SHDB, which is being developed to support SLCA practice. Only the “workers” stakeholder category is evaluated using the SHDB indicators. The SHDB offers a risk assessment analysis at the country level. The evaluation scales come from the SHDB: Risk and Opportunity Table Development document (Benoît et al. 2010). Data sources are not listed in the present document but can be found in Benoît et al. (2010).

Table 8 Indicators selected from the social hotspots database

Finally, Table 9 presents a list of country level indicators selected from various sources. Issues of concern related to the stakeholder categories Local community and Society are evaluated using these indicators. The PRP and the scales of evaluation are also presented.

Table 9 Indicators selected from a variety of sources

All these indicators in Tables 7, 8 and 9 were documented and assessed. However, they were aggregated in Table 11, as indicated later, but the detailed results are published in the 50-page Annex J of the full report.

3 SLCA Results

The socioeconomic performance of the Canadian milk production sector will therefore be analyzed in two ways: (1) at a specific level by describing the dairy farms’ level of social engagement on the one hand, and (2) at a generic level by providing a preliminary overview of the social risks (potential hotspots) related to the sector’s supply chains on the other.

3.1 Socioeconomic Performance at the Dairy Farm Level

Figure 3 shows the average socioeconomic performance of the Canadian dairy farms towards their stakeholders, i.e., the farm workers, their local communities, the society and the value chain actors—i.e., their suppliers and business partners (including the consumers). Each circle represents a level of the social responsibility evaluation scale, going from “risky behavior” in red to “committed behavior” in dark green. The red line represents the average socioeconomic performance of the Canadian dairy farmers according to each indicator. The closer the red line is to the outermost circle, the better the sector’s average socioeconomic performance, with a leading socially responsible behavior.

Fig. 3

Socioeconomic performance of the Canadian dairy farms

Canadian dairy farms have a positive socioeconomic performance globally. That is the case, for instance, for the agroenvironmental practices, whether it concerns water source protection, manure storage, or soil conservation. If this commitment is obvious from an environmental point of view, it is also significant in a socioeconomic perspective, as it also meets the Canadian society’s expectations.

The dairy farmers’ engagement towards their local community is also significant, with the vast majority involved in their communities in many different ways. More could be done, however, in terms of cohabitation, with more producers adopting practices—for instance, in minimizing odor propagation.

The picture is also contrasted with regard to farm workers. Although dairy farmers provide overall working conditions that go beyond the labor standards—to which they are mostly not legally subjected—there is still room for improvements regarding various issues such as professional training and communication of working conditions. The same holds true with respect to their suppliers and business partners, since a majority of dairy producers do not usually consider their suppliers’ performance with regard to social responsibility in their procurement decisions.

The average performance of the Canadian dairy farmers, as seen before, can hide some variability within the sector. For a given issue of concern, some producers might have a proactive or committed behavior where others will only comply with the expected social norms, as is the case with the odors spread reduction practices. This variability suggests that there is always room for improvements, since the average socioeconomic performance can be improved and, when already committed, reinforced, as more dairy producers could adopt some more socially responsible practices. Moreover, since today a committed behavior could become a minimal expectation in the future, a continuous engagement from all the producers is also advisable in order to improve, but also to preserve, the sector’s socioeconomic performance over time.

This variability is shown in Table 10. For each indicator, the average score is presented according to the evaluation scale used, as well as the variability of the practices and behaviors documented.

Table 10 The average socioeconomic performance of the Canadian dairy farms

The “Variability” column presents the relative share of answers that correspond to each possible value and the “average performance” column shows where the “mean” value is, via an arrow. The color code is as described earlier in the document:

: risky behavior; : compliant behavior; : proactive behavior;

: committed behavior; : non-available evaluation level.

3.2 The Potential Hotspots Analysis Results

Although this SLCA is primarily aimed at assessing the socioeconomic performance of Canadian milk at the farm level, the study also looked at the potential social risk in the suppliers upstream of the dairy sector, such as manufacturers of machinery, fertilizers, pesticides, or pharmaceuticals.

The PHA has been conducted over nine supply chains in order to assess, by using generic data, the possibility of encountering risky behaviors among the businesses involved at each stage.

The detailed description and evaluation of these risks can be found in Appendix J of the full report. This section presents the overall results and discusses their implications for the Canadian dairy sector.

Table 11 presents the aggregated results as well as the main potential hotspots related to the Canadian dairy sector’s supply chains. The results have been aggregated for simplification, by measuring the average risk related to each stakeholder category, given the score attributed to each associated issue of concern. No weighting method has been used between the issues of concern or the regions, when it was applicable.

Table 11 Aggregated results and main potential hotspots related to the Canadian dairy sector’s supply chains^{a, b}

Globally, this preliminary overview indicates that most supply chains show low social risk. With the main suppliers located in Canada or the United States, the prevalence of social hotspots is generally lower than in countries such as China. Yet there are some socially troubling practices occurring upstream in the sector’s supply chains, beyond the first-tier suppliers (which were not covered in this study). Among the most troubling practices are corruption, unsafe working conditions, non-respect of indigenous rights, and unfair competition.

This is, for example, the case in the fertilizer and oil extraction industries, where it was possible to document disturbing practices of collusion as well as bankrolling techniques from subsidiary companies of some major players. Potential hotspots were also identified in the Canadian grain and oilseed sector with regard to working conditions, as workers are generally not protected by labor standards. The analysis also brought up public health issues, as well as conflicts of use of natural resources related to many industries, among them the pesticides and pharmaceutical sectors.

Unfortunately, the use of generic data does not allow having a precise and detailed analysis of the actual hotspots occurring in the supply chains. Manufacturing information is only available at a national level, for instance, and is hence characterized by a high level of uncertainty regarding the actual behaviors of the businesses operating there. Furthermore, many of the identified hotspots are related to companies, sectors or regions located far upstream and on which the Canadian dairy sector has little power to influence.

The objective of this PHA was, however, to provide a preliminary overview of the social issues found among a product’s supply chains in order to bring awareness of the socioeconomic risks related to current procurement practices and to point out issues requiring deeper analysis. In a social responsibility perspective, it is important for the Canadian dairy farmers—as well as for their organizations—to consider not only the risks but the potential socioeconomic consequences related to their sourcing practices as well. By getting involved and by considering environmentally and socially responsible criteria in their procurement practices, the Canadian dairy sector could improve the overall socioeconomic performance of milk production in a life cycle perspective. This assessment can be seen as a starting point in this direction.

4 Discussion: Advantages, Limitations and Challenges Faced

As one of the first SLCAs conducted after the publication of the UNEP-SETAC guidelines in 2009, this project has been the opportunity for developing an innovative methodology to assess the socioeconomic performance of a product in a life cycle perspective. We strove to develop a rigorous, transparent, replicable and evolutive assessment methodology to enhance the SLCA development and facilitate its wider use. To do so, we relied on a few guiding principles that we still follow today: operationalization, readability, usefulness, and participation.

However, this context of novelty was also the first challenge that we faced. There were not many practical studies to use as a model, and the guidelines were vague on many aspects (Blom and Solar 2009; Revéret and Parent 2012, 2013). As Feschet (2014) explains, we could see that some researchers were promoting an impact pathway approach to SLCA and others a social responsibility approach and performance-based SLCA. Macombe and Falque (2013) identify this second category as “life cycle corporate social responsibility.” Although we understand and accept that measuring a company’s social performance is not an endpoint measurement of the social impact that we are interested in, we considered the fact that there are solid hypotheses on the causal relationship between adopting good practices and generating a positive impact. Therefore we made the choice to develop an approach based on PRP, which was later adopted in other studies conducted by our team and also by others. We note that the recently published Handbook for Product Social Impact Assessment (2014) also promotes an approach based on PRPs (Goedkoop 2014). However, it remains important to develop a better understanding of the pathways that link company behaviors to social impacts so that SLCA can make use of these relations to measure accordingly the positive and negative social impacts of products.

A second limitation of the guidelines that we faced was that not much was said about the linkages between environmental and SLCA when both were to be conducted simultaneously in a single study. Questions about the goal and scope, the limits of the system under study, and the still controversial question of the functional unit about whether or not it is relevant to use the same, when it is clear that the impacts as they are considered in an SLCA, are not quantified in a way that allows them to be reported per functional unit. This question of the quantitative versus qualitative nature of what is being observed is as present now as it was four years ago.

These differences—between the now classic environmental LCA and the newly born SLCA—were another source of debate. Although the development of SLCA was very much influenced by the parallel development of tools for CSR, such as ISO 26000 (performance-based, using similar categories of stakeholders, etc.), it was also firmly based on the ground of environmental LCA and on a similar logic. However, social issues are of a different nature than environmental ones, and are captured by different types of variables. Very often the more important aspects of social impacts are qualitative in nature, and the causal relationship with the product at the core of the study is not so clear; they are more related to the company’s behavior than to the product itself. All these elements do not facilitate the integration into a tool that is designed for quantitative data with solid causal chains that can be accurately and mathematically modeled—all of which create obstacles in the development of an efficient interdisciplinarity.

Understanding that we would not be able to develop integration at a conceptual level, we at least made sure to develop a procedural integration in the way we worked with the environmental team from QUANTIS and CIRAIG. We developed the social system under study from that used for environment and adjusted it to the fact that our social study was conducted, as we saw, at two levels of precision—the first, a specific study at the farm level using primary data, and the second, searching for potential hotspots in the supply using secondary and generic data.

5 Conclusion

The purpose of this study was to profile the socioeconomic performance of ordinary Canadian milk. Using data from over 300 farms as well as provincial and national statistics, a regionalized characterization of average provincial scenarios allowed for a nationwide understanding and assessment of milk production. While variability in farm practices and results were discussed at every stage of the life cycle steps, and for the different socioeconomic indicators, it is important to remind the reader that such variability was evaluated only between the provincial averages, and, as such, does not come close to evaluating and understanding the variability between various farms. As a result, the current study provides an understanding of how various scenarios and locations affect the environmental profile of milk—without, however, being able to assess the potential by which best practices within one type of management can contribute to reducing the overall burdens. With respect to the assessment of the average socioeconomic performance of Canadian milk production, the study evaluated the Canadian dairy farmers and their boards at a national level, based on their degree of social engagement, and was not intended to assess the performance at an individual level.

It is clear from this assessment that Canadian dairy farms have an overall positive performance. It is also obvious, with respect to the agroenvironmental practices, whether concerning water source protection, manure storage, or soil conservation. The engagement of dairy farmers with their local community is also significant, as he vast majority are involved in their communities in many different ways. More can be done, however, in terms of cohabitation, with producers adopting practices that minimize odors propagation.

The picture is also contrasted with regard to farm workers. Although dairy farmers provide overall working conditions that go beyond labor standards—to which they are mostly not legally subjected—there is room for improvements regarding various issues, such as professional training and communication of working conditions. The same holds true with respect to their suppliers and business partners, given that a majority of dairy producers do not usually consider their suppliers’ performance with regard to social responsibility in their procurement decisions.

This suggests that there is always room for improvement, both now and in the future. For example, with more producers adopting more socially responsible practices, the average socioeconomic performance could be enhanced. Moreover, since a committed behavior today can become standard in the future, continuous improvement from all producers is also required, not only to improve, but also to preserve the sector’s socioeconomic performance.

Finally, the study also looked at the social risk potentially present in the suppliers upstream of the dairy sector, such as manufacturers of machinery, fertilizers, pesticides, or pharmaceuticals. With the main suppliers located in Canada or the United States, the prevalence of social hotspots is generally lower than in countries such as China. The fact remains, however, that some risks seem present in a few links of the supply chains. This is the case in the fertilizer and oil extraction industries, for example, where it was possible to document disturbing practices of collusion as well as bankrolling techniques from the subsidiaries of some major players. Potential hotspots were also identified in the North American grain and oilseed sector with regard to working conditions, as they are generally not protected by labor standards. The analysis also brought up public health issues, as well as conflicts of use of natural resources related to many industries, among which are the pesticides and pharmaceutical sectors. Some links are also characterized by a lack of competition. Although the Canadian dairy sector has little power to influence these actors located far upstream, in a life cycle perspective, it falls under the responsibility of dairy farmers and their associations to get involved. This assessment can be seen as a starting point in this direction.

This SLCA and the environmental LCA were the first step towards engaging all stakeholders in a comprehensive sustainable development strategy. This assessment provides the Canadian dairy sector with an innovative, comprehensive and actionable roadmap to move in the direction of a more sustainable milk production in Canada. We should mention, as a practical recognition of the importance of the role of farmers, the Dairy Farm Sustainability Award, which was established in 2012.¹⁴ This competition promotes the recognition of Canadian dairy farmers that have adopted on-farm management practices that extend beyond standard industry practice and meet the objectives defined in the DFC’s sustainability strategy.

Moreover, capitalizing on these results also depends on the extent of communication with involved parties, which is a crucial part of the next steps. In addition to the various academic conferences and the numerous webinars organized for farmers and their boards’ representatives across the country, many communications documents based on the study were produced and are being used at the national and provincial levels. The model generated here can also serve as a basis for a self-assessment tool aimed at farmers, which could be improved to better identify best practices. Such a self-assessment tool is now under development as part of the next phase of the Dairy Research Cluster and will be soon implemented at the farm level.