• Community Benefits Agreements (CBAs) can be useful tools in providing greater opportunities for underrepresented groups to participate in the construction industry.
• However, British Columbia’s CBA (the “Agreement”) contains problematic elements that are not typical of traditional CBAs. These include:
  • A requirement that all workers on Agreement projects join one of the nineteen unions affiliated with the Agreement.
  • A provision that British Columbia Infrastructure Benefits, a Crown corporation, employs all workers, supplanting contractors in their role as employers.
• A case study of the Cowichan District Hospital Replacement Project demonstrates the effect of these problems:
  • Only two firms bid on the project.
  • Cost overruns are $559 million so far. This is the largest cost overrun in dollar terms—and the third-largest in percentage terms—among large Infrastructure BC projects.
  • The government blamed cost overruns in part on labour shortages, despite having restricted the labour pool through the Agreement.
  • An Indigenous contractor in Cowichan was refused permission to work on the project because of the unionization requirement, despite the fact that Indigenous workers and local residents are priority groups in the Agreement.
  • In providing an exemption from the workforce rules to the Indigenous contractor, the government has effectively admitted the Agreement’s fundamental flaw without fixing the Agreement itself.
• The government should therefore overhaul British Columbia’s CBA to (1) respect the right of workers to affiliate with unions as they wish, (2) re-establish contractors in their roles as employers, and (3) re-focus on expanding opportunities for underrepresented groups to participate in the construction industry.

In 2018, the Government of British Columbia, through a newly established Crown corporation called British Columbia Infrastructure Benefits Inc. (BCIB), signed an accord with an alliance of nineteen labour unions. The purpose of this agreement—the British Columbia Community Benefits Agreement (called “British Columbia’s CBA” or simply “the Agreement” in this report)—was to govern the workforce of certain public infrastructure projects under the jurisdiction of BCIB.

The Agreement, which remains in force today, has been controversial, culminating in a highly publicized dispute involving an Indigenous-owned construction company that was denied permission to work on the Cowichan District Hospital Replacement Project in 2022.

In two previous reports published prior to the Cowichan incident, Cardus recommended a framework for successfully implementing Community Benefits Agreements (CBAs). The purpose of this paper is to analyze how well British Columbia’s CBA matches these best practices.

This report finds that the Agreement contains serious flaws that put it at odds with core principles of supplier and workforce diversity. It finds that some of its provisions can lead to perverse effects that can undermine its intended results.

The first section of this report provides an outline of CBAs in general, including a description of what they are and a summary of previous Cardus research on how they should best be implemented. The second section examines the contents of British Columbia’s CBA itself. The third section is a case study on the Cowichan District Hospital Replacement Project. The final section contains recommendations on how the Government of British Columbia should revise the current Agreement.

Community Benefits Agreements 1 1 This section is drawn primarily from “Community Benefits Agreements: Toward a Fair, Open, and Inclusive Framework for Canada,” Cardus, 2021, https://www.cardus.ca/research/work-economics/reports/community-benefits-agreements-toward-a-fair-open-and-inclusive-framework-for-canada/; and A. Flootman, “A Framework for Implementing Community Benefits Agreements,” Cardus, 2022, https://www.cardus.ca/research/work-economics/reports/a-framework-for-implementing-community-benefits-agreements/. are contracts signed alongside an infrastructure construction contract to provide additional benefits to the community beyond the specific piece of infrastructure to be built. The idea of CBAs is to leverage public procurement that is already being contracted, to receive some extra advantage to the community.

The community benefits that are supposed to accrue to the community can be quite wide-ranging. Occasionally, they can include the construction of an additional piece of infrastructure, such as the construction of bike paths alongside a new road. More commonly, community benefits constitute the engagement of local businesses or employment and training opportunities for members of groups that are disadvantaged in the labour market. For example, a contractor might agree to certain targets or quotas for hiring and training of Indigenous workers, or members of other underrepresented groups.

CBAs are usually signed between a developer or construction company and a community group or government. 2 2 D. Graser, “Community Benefits: FAQs,” October 22, 2018, https://dinagraser.ca/wp-content/uploads/2018/10/Community-Benefits-FAQs-2018.pdf. In these cases, the government effectively acts as the agent of the community seeking some benefit, as well as the contracting authority that is seeking the construction of the main infrastructure project. Community benefits need not be outlined in a separate agreement; they can be included as community-benefit clauses within the main contract for a construction project.

It is important to emphasize that a community benefit represents an additional benefit. The nomenclature of CBAs can make it easy to forget that the primary benefit to the community of a major infrastructure project is usually the infrastructure itself. It is, after all, for the benefit of the community that most infrastructure projects, whether a road, a community centre, or a hospital, are procured in the first place.

It is also important to distinguish CBAs from project labour agreements (PLAs). A PLA is an agreement with a labour union or a group of labour unions that sets out the work conditions on a project. This may include hiring procedures, wages and benefits, and limitations on strikes and lockouts. By setting out the conditions of employment in advance, PLAs seek to reduce labour disruptions, contribute to predictable and timely completion of construction projects, improve safety, and increase the diversity of the workforce. 3 3 US Department of Labor, Project Labor Agreement Resource Guide, https://www.dol.gov/general/good-jobs/project-labor-agreement-resource-guide. As such, PLAs can have positive effects. Problems can arise, however, when PLAs are used to restrict the workforce of a project, as discussed later in this report. CBAs typically do not contain such stipulations about the details of the workplace, although a more diverse labour force may be a goal of both CBAs and PLAs.

As with any public policy, how well a CBA performs will depend a great deal on details and implementation. If poorly designed or poorly implemented, even the best intentions of governments can lead to serious problems.

Cardus examined this issue at length in two previous reports, which looked at different models to draw lessons learned and establish best practices. 4 4 “Community Benefits Agreements: Toward a Fair, Open, and Inclusive Framework for Canada”; and Flootman, “A Framework for Implementing Community Benefits Agreements.” The first set out some general principles for what CBAs should and should not contain. The second examined what an implementation strategy would look like for CBAs and provided a checklist (appended again to this report), for governments to follow in setting them up. This research identified four general areas of concern in implementing CBAs 5 5 What follows is a summary of A. Flootman, “A Framework for Implementing Community Benefits Agreements.” :

When governments implement CBAs, it is important that they do so in a way that does not artificially reduce the number of firms that can compete for the project. Not only can this violate principles of fairness, it can also reduce competition for government procurement, resulting in higher prices for taxpayers.

Unfortunately, it is very easy for governments to inadvertently make it more difficult for certain businesses to compete. Rules concerning the gender, ethnic origin, union affiliation, or other characteristics of the firm or its workers can be used to prevent certain kinds of bidders from participating in a competition. This can be avoided by eliminating any criteria that rule out participants on the basis of the above characteristics.

Governments also sometimes bundle procurement for multiple projects together so as to make it more economically feasible to attach a CBA to them. The idea behind this is that larger projects can more easily absorb the costs and operational requirements associated with CBAs. However, bundling projects together can have the effect of putting these contracts out of reach of smaller businesses that do not have the capacity to take on multiple projects but may have been able to bid on individual projects if they hadn’t been bundled. Bundling may therefore have the unintended consequence of shutting out smaller local businesses, even though promoting local business is often a specific goal of CBAs. This has the further effect of reducing the overall number of firms that can bid on projects, reducing competition and leading to higher prices.

Government also has the responsibility to ensure that the procurement process is reasonably well advertised to potential qualified bidders so that the taxpayer gets the best possible deal. This could involve proactive outreach to potential suppliers and associations in the business community. These actions promote competition, increase the likelihood of getting a larger number of qualified bidders, and thereby seek a better price for taxpayers. While this may not be a provision of the CBA itself, it is an essential policy action that must occur alongside the CBA to ensure it is successful.

At the end of the day, the goal of supplier diversity is to ensure that a maximum number of qualified companies can bid on government procurement projects. Promoting healthy competition among bidders will help to ensure that taxpayers get the best deal for these projects. Rules that unnecessarily restrict competition should therefore be eliminated.

The diversity of the workforce takes many forms. CBAs must take account of and balance all these different kinds of diversity.

The most obvious form is the demographic makeup of the workforce, particularly the representation of traditionally underrepresented groups. This is, after all, one of the primary goals of a CBA: to increase the employment of underrepresented groups within the construction industry. In this regard, it is incumbent on governments, as part of the CBA, to ensure that employers have the tools and supports they need to hire members of these groups. The government should proactively engage with civil-society actors such as labour unions and social enterprises to support businesses that want to develop their workforces in this way.

Another form of workforce diversity is achieved by ensuring that government is maximizing the pool of labour that is available for its procurement projects. In a complex economy like Canada’s, there are usually many different pools of labour, including unionized labour and non-unionized labour. Even among unionized labour, there are often different kinds of labour unions. Governments that restrict their procurement processes to one kind of labour miss out on the diversity of the unionized and non-unionized workforce. In periods of labour shortages, especially, governments cannot afford to leave certain kinds of labour on the sidelines.

This principle is closely connected to supplier diversity, as policies that restrict certain kinds of labour will have the effect of restricting the suppliers, if their workers are not members of the favoured labour pool.

A successful CBA requires active engagement on the part of government, from the early stages of developing the CBA through to the on-the-ground implementation. The costs associated with community benefits must be included early in the planning phase so as to avoid surprise cost overruns. Similarly, the practicalities of implementing a CBA need to be considered early in its development in order to ensure an available pool of labour for the contractor to hire between the signing of the contract and the anticipated start of construction.

Governments should then act as partners with contractors in connecting new workers with them, including by providing them with a list of resources that exist in the community to identify sources of labour. As part of this, the government should make a realistic assessment of a community’s ability to provide the workforce diversity that the government is seeking to encourage. Understanding a community’s capacity in advance is essential to avoiding conflicting requirements in CBAs. For example, in some communities, establishing certain demographic quotas while also requiring the use of local labour may simply be impossible. Careful assessments can ensure that the provisions of CBAs are internally consistent and work in harmony.

It may be obvious that governments must measure the success of their CBAs, but the challenge is in determining what to measure. If, for example, the CBA is focused on increasing the employment of underrepresented groups in construction, then the government will have to consider how to measure this representation. It may be insufficient simply to count the number of members of underrepresented groups working on a project, since companies could hire workers just to meet a quota but not give them meaningful work that employs or develops their skills.

On the other hand, it is important that governments provide a certain amount of flexibility to contractors to meet the requirements. In the same vein, it is also important for governments to track aggregate data on what is happening with employment and training rates for target groups at a community level. Placing too much emphasis on what is happening with a single employer may miss the big picture.

Finally, governments must recognize that there is a threshold below which a CBA might simply not make sense. For small procurement projects, the number of workers on a project could be so small that it is not practical to establish a target for the employment of underrepresented groups. In this sense, a certain amount of measurement is necessary even in the process of contemplating whether a CBA is feasible. It is difficult to establish a threshold that would apply to all contexts. Previous attempts to establish a threshold have ranged quite substantially, with some advocates having suggested a threshold as high as $100 million. Others have suggested a lower threshold, with the federal government establishing a threshold of $10 million for its Community Employment Benefits requirements. 6 6 Infrastructure Canada, Backgrounder: Applying a Community Employment Benefits Requirement to Major Infrastructure Projects (June 22, 2018), https://www.canada.ca/en/office-infrastructure/news/2018/06/backgrounder-applying-a-community-employment-benefits-requirement-to-major-infrastructure-projects.html. The reality is that the ideal threshold is likely to be dependent on context and may vary from community to community.

If governments follow the framework for implementing CBAs outlined above and summarized in the checklist appended to this report, they will be well positioned to meet their goals with minimal unintended consequences and with minimal extra costs for their taxpayers.

The previous section considered CBAs in general and summarized a framework to guide governments in implementing them. British Columbia’s CBA differs from this framework and other CBAs in significant ways. This section examines some of these points of departure and considers their effects.

As part of instituting the Agreement, the Government of British Columbia established British Columbia Infrastructure Benefits Inc., a Crown corporation that would go on to become the employer of all workers involved in the construction of certain infrastructure projects.

The Agreement itself is an agreement between BCIB and the Allied Infrastructure and Related Construction Council of British Columbia, an alliance of nineteen building-trade unions, collectively known in this context as the “affiliated unions.” 7 7 British Columbia Infrastructure Benefits, Inc. (BCIB), Community Benefits Agreement (July 17, 2018; Amended and Restated March 1, 2022), https://bcib.ca/wp-content/uploads/2023/05/Community-Benefits-Agreement.pdf.

The Agreement governs the workplaces and hiring procedures for the projects that fall under its authority, of which there are or have been seven 8 8 BCIB, Network of Projects, https://bcib.ca/projects/. :

• The Illecillewaet brake check project
• The Pattullo Bridge Replacement Project
• The Trans Canada Highway #1 – Kamloops to Alberta 4-Laning Project
• The Broadway Subway Project
• The Cowichan District Hospital Replacement Project
• The Kicking Horse Canyon, Phase 4 Project
• The Centre for Clean Energy and Automotive Innovation at Vancouver Community College

With the exception of the now-completed Illecillewaet brake check project, the costs of each of these projects range from hundreds of millions to billions of dollars. In this respect, it is fair to conclude that the Agreement’s current projects all fall well above a reasonable threshold for applying CBAs.

The signatories of the Agreement are the first indication that it is an unusual arrangement for a CBA. Normally, a CBA is signed between a government and a contractor, but British Columbia’s CBA is an agreement between government and a number of labour unions. The contractors responsible for building the infrastructure projects associated with the Agreement are not signatories to it at all. In this respect, the Agreement more closely resembles a PLA, which is normally signed between the employer and a labour union, to govern the workplace.

The Agreement’s twenty-five articles cover a range of issues, including grievance procedures in the event of a workplace dispute (article 10), specifications about wages and benefits (article 13), the standard hours of work, including lunch breaks, rest breaks, overtime policies, and the standard work week (article 14), first aid and safety (article 20), and travel reimbursements for workers (article 22). These are provisions that would typically be found in a PLA or a collective agreement.

In fact, British Columbia’s CBA states outright that its “provisions constitute a Collective Agreement,” 9 9 BCIB, Community Benefits Agreement, at Master Section, p. 3. and BCIB’s website brands the document as “The CBA: A Collective Agreement.” 10 10 BCIB, BC’s Community Benefits Agreement, https://bcib.ca/about-us/cba/. In other places, the government describes it as a PLA. For example, the request for qualifications of the Cowichan District Hospital Replacement Project, a project that is highlighted in greater detail below, uses that very term to explain the nature of the Agreement to prospective bidders: “The CBA is a project labour agreement which sets out the employment terms and conditions for the supply of workers to be utilized by the Project Alliance and its subcontractors on the Project.” 11 11 Vancouver Island Health Authority, Request for Qualifications, December 18, 2020, at p. 14, https://www.infrastructurebc.com/wp2/wp-content/uploads/2021/04/2021-02-24-VIHA-CDH-RFQ-FINAL-conformed.pdf.

The fact that the Government of British Columbia calls the Agreement a “collective agreement” and a “project labour agreement” is not just a matter of nomenclature. It signals that the Agreement includes more than a traditional CBA. An examination of its key features will confirm this.

The defining features of British Columbia’s CBA are as follows:

Under article 2.1, BCIB is designated as the employer of all employees on projects that fall under the Agreement. Although BCIB delegates many responsibilities to contractors, it nevertheless remains the employer of all workers. The article also requires contractors, who are not parties to the Agreement, to agree that BCIB is the employer and to follow all the terms of it.

The term “employee” is defined in the Agreement as anyone hired to work on construction of a project under its scope. There are some explicit exceptions (such as professional engineers, consultants, and oversight personnel), but the vast majority of those working on an Agreement-covered project would fall under the definition of “employee.”

Article 2.1 represents a fundamental re-orientation of the contractor role. BCIB pitches this as a benefit to contractors, by providing them with a supply of workers and taking care of human-resources concerns. However, this could equally be seen by contractors as giving up management rights. It also undermines the purpose of tendering to an external contractor, which is to have a construction project managed by a non-governmental entity with more knowledge and experience in construction. Moreover, this is an uploading to government of some of the basic functions of employers, who are best positioned to know the needs of their firm, the requirements of the project, and the kinds of workers they need.

Comparing this feature with the best practices outlined in the previous section, it would seem that article 2.1 goes well beyond the reasonable scope of project management for a government. It is also unnecessary to achieve the goals of traditional CBAs. Ideally, the government’s role in recruitment and human resources is one of help and support for contractors in finding new and diverse pools of labour. It does not consist in taking over as employer.

Article 8.1 of the Agreement is clear that all employees working on projects subject to it must “be members of or secure membership in the Appropriate Affiliate and maintain such membership in good standing as a condition of employment.” Employees that are not yet members of an affiliated union must apply for membership within thirty days of their engagement to work.

This article has been the source of much controversy. Opponents have pointed out that union membership is required for employees working on projects under the scope of the Agreement. Supporters have argued that construction firms and their employees do not have to be unionized in order to access work associated with the Agreement: they argue that the workers simply have to join one of the nineteen affiliated unions upon being hired by BCIB.

A careful reading of the text shows that both are technically correct but also that the latter argument fails to adequately address the former. The result of article 8.1 is that workers on projects falling under the Agreement must be members of a specific list of unions within one month of being hired. They are not permitted to choose not to be unionized or to join an alternative union. Likewise, contractors who wish to bid on projects under the Agreement are required to have their workers unionized by one of the chosen labour unions, whether their workers want that or not. Moreover, when combined with article 2.1, their workers are not only unionized but become employees of BCIB for the purposes of the projects.

Article 9.1 outlines a priority list for hiring qualified workers from four groups:

• Indigenous people
• Equity groups (“women in non-traditional work, people with disabilities, and other traditionally underrepresented groups”)
• Local residents (BC residents living within 100 km of the project)
• Existing members of an affiliated union

At the top of the priority list are qualified Indigenous people to whom the government has made specific commitments for the project. After that, the hiring proceeds by prioritizing the above-stated groups and combinations thereof. Generally, the more of these groups a potential worker belongs to, the higher they are on the priority hiring list. Once the pool of local residents is exhausted, the priority list expands to residents of Yukon and the other Western provinces who are members of the affiliated unions, then to all Canadian residents who are members of the affiliated unions, and only then to other Canadian residents.

There are a few things to note about this arrangement. First, although membership in an affiliated union is only one of four factors used to prioritize the hiring of a worker, we must recall that article 8.1 requires that any employee hired to work on the project must become a member of an affiliated union. It would therefore be wrong to interpret this article as allowing non-unionized or alternatively unionized workers to work on the project, even if they are deprioritized in the hiring process. These workers must join the union within thirty days of being hired.

Second, this article highlights that, despite a requirement for all employees to join one of the affiliated unions, there is nonetheless a preference in the hiring process for workers who are already members of one of these unions. Thus, there is a two-fold benefit for the affiliated unions in this respect: its existing members are among the priority hiring groups, and even if a previously nonunionized worker is hired, he or she must join. Both of these provisions entrench the monopoly that the affiliated unions have on projects under the Agreement.

Article 8.6 outlines the specific processes by which BCIB and the affiliated unions choose employees for a project. There are four processes outlined. The process that applies to a contractor depends on two factors: the number of workers they need (six or less, versus more than six), and whether the contractor needs workers from a single trade or multiple trades.

In any case, the hiring process follows a similar structure typical of collective agreements with building-trade unions:

1. BCIB name hires all supervisors 12 12 “Name hiring” is when the employer chooses a worker or requests of the union a particular worker.
2. BCIB name hires between two and four employees, depending on the applicable hiring process
3. A rotation begins and is repeated until all workers are chosen 13 13 All processes follow this basic rotational structure, although there can be minor deviations in narrow circumstances. :
   1. The affiliated union dispatches an employee
   2. BCIB requests an employee name from the affiliated union
   3. BCIB name hires an employee

The process is summarized for contractors as follows: “Subcontractors can name their existing workers to fill a portion of their crew on CBA projects. BCIB hires those workers and fills remaining positions with our candidates and with dispatches from the Affiliated Unions.” 14 14 BCIB, Labour Supply, https://bcib.ca/contractors/labour-supply/.

It is important to recall that, under article 8.1, all employees must be or become members of an affiliated union within thirty days of being hired. Thus, even the name hire must be or become a member of an affiliated union. In effect, article 8.6 is a corollary of article 8.1. In other words, the strict hiring process is a result of requiring workers to become unionized by the affiliated unions. Notably, however, this article has nothing to do with encouraging a more diverse workforce. It is solely about who gets to name an employee.

These four articles create a system that infringes the best practices identified in the previous section of this report. The requirement that all workers be or become members of one of the affiliated unions goes against the principle of workforce diversity by restricting the labour pool to one kind of labour. This in turn violates the principle of supplier diversity by discouraging potential bidders from bidding on government infrastructure projects. The combined effect is to restrict competition, leading to higher prices and less value for BC taxpayers.

The most controversial of the six projects that have fallen under the jurisdiction of British Columbia’s CBA is the Cowichan District Hospital Replacement Project. In late 2022, a dispute arose on the construction site, involving a company owned by a member of the Cowichan First Nation.

The situation is a useful case study of what can go wrong when a CBA is designed in ways that restrict supplier and workforce diversity. This section profiles the project and the issues that arose.

Cowichan Valley is a regional district on Vancouver Island in British Columbia, near Victoria. It has been the home of the Cowichan District Hospital since 1967. 15 15 Office of the Premier, Government of British Columbia, New Hospital Will Help People in the Cowichan Valley Access Quality Health Care (July 6, 2018),  https://news.gov.bc.ca/releases/2018PREM0050-001350.

The project was commissioned to replace the current hospital with a new one in North Cowichan, a few kilometres north of the current site. The plan would increase the community’s hospital capacity from 134 beds to 204 beds and in other ways, such as more operating rooms. 16 16 Infrastructure BC, Cowichan District Hospital Replacement Project (2024), https://www.infrastructurebc.com/projects/projects-under-construction/cowichan-district-hospital-replacement-project/. The new hospital would serve a large area spanning the entire Cowichan Valley Regional District, including the municipalities of North Cowichan, Duncan, Lake Cowichan, and Ladysmith.

The project was approved in June 2018 and was announced by the Premier’s Office in July 2018. 17 17 Office of the Premier, Government of British Columbia, New Hospital Will Help People in the Cowichan Valley Access Quality Health Care.

The Request for Qualifications (RFQ), a formal part of the government’s procurement process wherein contractors submit proposals to determine if they meet the technical qualifications for the project, was issued on December 18, 2020. The RFQ document laid out the conditions of the Agreement, including the provision that “BCIB will provide the labour force and manage labour relations for the Project Alliance and its subcontractors working on the Project.” The RFQ closed just over two months later, on February 25, 2021. According to the fairness reviewer, an independent examiner who verifies that a procurement process is fairly implemented, only two firms applied for the RFQ. 18 18 J. Shackell, “Report of the Fairness Reviewer,” March 23, 2021, Vancouver Island Health Authority, https://www.infrastructurebc.com/wp2/wp-content/uploads/2021/04/53275014_2_CDH-Report-of-the-Fairness-Reviewer-RFQ_-002.pdf. Both were deemed qualified to bid.

The Request for Proposals (RFP) was issued shortly thereafter on April 1, 2021, 19 19 Vancouver Island Health Authority, Request for Proposals: Alliance Development Agreement (April 1, 2021), https://www.infrastructurebc.com/wp2/wp-content/uploads/2021/04/VIHA-CDH-RFP-FINAL-conformed-April-27-2021.pdf. with a second phase issued on July 21, 2021. 20 20 Vancouver Island Health Authority, Request for Proposals: Alliance Development Phase (July 21, 2021), https://www.infrastructurebc.com/wp2/wp-content/uploads/2021/11/2021-07-21-VIHA-CDH-ADP-RFP.pdf. Under the RFP, those contractors that qualified through the RFQ were invited to submit a formal proposal. The two qualified contractors were Alliance Care Partners and a partnership between EllisDon Corporation and Parkin Architects Western Ltd. 21 21 Vancouver Island Health Authority, Request for Proposals: Alliance Development Agreement at p. 1. The RFP closed in December 2021. 22 22 Vancouver Island Health Authority, Request for Proposals: Alliance Development Phase at p. i.

The contract, dated September 9, 2022, was awarded to EllisDon and Parkin. 23 23 Vancouver Island Health Authority, Cowichan District Hospital Replacement Project: Project Alliance Agreement (September 9, 2022), https://www.infrastructurebc.com/wp2/wp-content/uploads/2023/08/Signed_CDHRP_PAA_FINAL_Redacted.pdf. The completion date was initially listed as 2026. 24 24 Ministry of Finance, Government of British Columbia, First Quarterly Report: 2020/21 Economic Outlook and Financial Forecast & Three Month Results April–June 2020 (September 2020), https://www2.gov.bc.ca/assets/gov/british-columbians-our-governments/government-finances/quarterly-reports/2020-21-q1-report.pdf. The government now lists the completion date as 2027. 25 25 Ministry of Finance, Government of British Columbia, 2023 British Columbia Financial and Economic Review, 83rd Edition (September 2023), https://www2.gov.bc.ca/assets/gov/british-columbians-our-governments/government-finances/financial-economic-review/financial-economic-review-2023.pdf.

The project has faced some significant cost increases. When it was originally proposed, several years before its approval in 2018, the estimated cost was reportedly $350 million. 27 27 R. Barron, “Cowichan Valley to Have a New Hospital by 2024, Horgan Announces,” Cowichan Valley Citizen, July 6, 2018, https://www.cowichanvalleycitizen.com/news/cowichan-valley-to-have-a-new-hospital-by-2024-horgan-announces-784377. When the project first appeared in the government’s official financial records in 2020, the total anticipated cost was listed at $887 million. 28 28 Ministry of Finance, Government of British Columbia, First Quarterly Report. By 2022, the total anticipated cost had risen to $1.446 billion, an increase of $559 million, or 63 percent, in only two years. 29 29 Ministry of Finance, Government of British Columbia, Second Quarterly Report: 2022/23 Economic Outlook & Six Month Financial Results April–September 2022 (November 2022),  https://www2.gov.bc.ca/assets/gov/british-columbians-our-governments/government-finances/quarterly-reports/2022-23-q2-report.pdf. That is where the official cost estimate remains today. 30 30 Ministry of Finance, Government of British Columbia, Budget and Fiscal Plan, 2024/25–2026/27 (February 22, 2024),  https://www.bcbudget.gov.bc.ca/2024/pdf/2024_Budget_and_Fiscal_Plan.pdf.

This is one of the largest cost overruns for large infrastructure projects in British Columbia. Table 1 analyzes data from the “British Columbia Financial and Economic Review” for the years 2016–17 through 2022–23, the most recent available year. In particular, table 1 tracks the change in the total anticipated cost of projects started since 2017 under the jurisdiction of Infrastructure BC that have appeared in the tables entitled “Capital Expenditure Projects Greater Than $50 Million.” These results are presented in figures 1 and 2.

This analysis shows that the Cowichan District Hospital Replacement Project has experienced the highest cost overruns in absolute terms and the third-highest in percentage terms. In fact, in terms of dollar amount, it is well over double the next-highest cost overrun.

It should be noted that this is not a perfect comparison of all projects. It could be that some of the recently announced projects simply have not had time to experience significant cost overruns, and they may yet do so. It is also not reasonable to say that all of the cost overruns are owing to restrictive tendering practices. 31 31 It is also noteworthy that six of the top seven projects for cost overruns are hospitals. This warrants further examination.

Nevertheless, this analysis shows that the 63 percent increase in anticipated total cost of the Cowichan District Hospital Replacement Project is not a typical cost overrun for government infrastructure projects. It is exceptionally large, whether it is examined on an absolute or relative basis.

Projects undergo changes in their scope, either in reductions or expansions in the size of the project itself. In this case, the government has explained that part of the cost overrun is due to expanding the size of the planned hospital by 18 percent. This would not lead to a directly proportionate increase of 18 percent in the costs, owing to the presence of certain fixed costs. It is also difficult to compare the Cowichan hospital cost overruns with those of other projects without knowing the scope expansions or scope reductions of all of them. But even taking the scope increase into account, it remains the case that the Cowichan District Hospital Replacement Project has seen some of the highest cost overruns among large infrastructure projects in British Columbia.

When the cost increases were revealed in 2022, the BC Minister of Finance attributed them primarily to inflation, labour shortages, and supply chain disruptions. 32 32 R. Barron, “Costs of New Cowichan Hospital Increase to $1.45 Billion,” Cowichan Valley Citizen, December 5, 2022, https://www.cowichanvalleycitizen.com/news/costs-of-new-cowichan-hospital-increase-to-1-45-billion-840733. It is impossible to say with precision what role British Columbia’s CBA has played in these cost overruns, though it is noteworthy that the Minister of Finance flagged labour shortages as a key cause. Indeed, labour shortages have plagued the construction sector for a number of years, including during the period in which the cost overruns occurred. But far from absolving the government of the cost overruns, the labour shortages call further into question the government’s own decision to restrict the pool of labour. Rather than limiting the pool from which the Cowichan District Hospital Replacement Project—and other Agreement projects—could draw workers, the government could have responded to labour shortages by drawing from a wider workforce.

The government’s decisions went against the principle of workforce diversity, as outlined in the section above that dealt with the appropriate structure of CBAs. In particular, restricting the labour pool to members of the building-trade unions violated the second form of workforce diversity, which ensures that the government has access to as wide a population of workers as possible. It is precisely in order to avoid cost overruns that this principle is so important.

Some may argue that there is no reduction in the pool of labour because no workers are technically excluded as long as they sign up for an affiliated union upon hiring. But this ignores two key facts. First, this rule is likely to discourage some workers who prefer to be affiliated with a different kind of union or to remain unaffiliated. Second, the rule is likely to discourage some contractors or sub-contractors from participating in the construction project. This has the effect not only of depriving the project of the workers associated with those contractors but also of depriving the contract-bidding competition of those contractors and sub-contractors. In this sense, the rule violates the principle of supplier diversity. With less competition among suppliers, the proponent of the project fails to benefit from the lower prices that would have come with a larger set of potential suppliers.

The key point of the foregoing discussion is simply that the significant cost overrun associated with the Cowichan District Hospital Replacement Project was exacerbated by the restrictions on the pool of labour, and thus on the supplier diversity, that were put in place. The fact that the Minister attributed the cost overrun to a labour shortage makes the case even clearer that it was a mistake to restrict the labour in this way.

One of the goals of British Columbia’s CBA is the priority hiring of designated demographic groups. As such, BCIB maintains statistics on the workers involved in its projects. According to data released by BCIB, the project has had some success in meeting its goals. From the beginning of the project to December 31, 2023, 91 percent of workers were designated as local residents, while 98 percent were residents of British Columbia. 33 33 BCIB, Memo: BC Infrastructure Benefits Workforce Equity Statistics (May 1, 2024).

With respect to underrepresented workers, BCIB says that 18 percent of the workers on the project self-identified as Indigenous, 8 percent as women, and 15 percent as youth. The percentage of hours worked by each group was similar: 16 percent, 7 percent, and 16 percent, respectively.

Comparing these statistics to the whole construction industry in British Columbia, we can see that the project is doing better on some scores and not as well on others. On Indigenous hiring, it appears to outpace the average rate for the construction and manufacturing sectors in British Columbia, where Statistics Canada records that only about 5 percent of the labour force in those sectors self-identified as Indigenous in 2023. 34 34 Statistics Canada, Table 14-10-0367-01. Employment by Geography, Indigenous Group and Industry, https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=1410036701. However, it should be noted that Statistics Canada’s Labour Force Survey excludes anyone living on a reserve, which would reduce the number of Indigenous workers counted. That, and the fact that there are nine reserves among the Cowichan Tribes, may explain some of the difference.

With respect to women, the employment figures on the project lag behind the average for British Columbia. Although BCIB claims that BC’s construction-industry average for female workers is 4.5 percent, 35 35 BCIB, 2022/23 Annual Service Plan Report (August 2023), https://bcib.ca/wp-content/uploads/F23-Annual-Service-Plan-Report.pdf. Statistics Canada recorded that in 2023 nearly 14 percent of the labour force in British Columbia’s construction industry was female. 36 36 Statistics Canada, Table 14-10-0023-01. Labour Force Characteristics by Industry, Annual, https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=1410002301. In data linked to on BCIB’s website, BuildForce Canada recorded that nearly 12 percent of the industry’s labour force was female, as of May 2024. 37 37 BuildForce Canada, “Construction Industry Key Indicators,” https://www.buildforce.ca/en/key-indicators.

It should be noted that BCIB’s statistics focus on the Agreement’s projects and their employment of underrepresented groups. This focus is understandable, but it does not necessarily reveal what is happening in the wider construction industry in British Columbia. As discussed in the section above on measurement, the decision of what to measure is crucial, and policymakers should be attentive to what is happening at a societal level.

Construction on the new Cowichan District Hospital began in the autumn of 2022. 38 38 Infrastructure BC, Cowichan District Hospital Replacement Project. Shortly after, in December of the same year, it was halted by protests at the construction site by a contractor named Jon Coleman and other members of Cowichan Tribes.

Jon Coleman was the owner of Jon-co Contracting Ltd., a company that had been subcontracted to work on the hospital. He was also a member of the Cowichan First Nation, and his company was a member of the Khowutzun Development Corporation.

Under the terms of British Columbia’s CBA, the employees of Jon-co Contracting Ltd. and Khowutzun who were working on the project were required to become members of one of the affiliated unions. According to comments made by Cowichan Tribes at the time of the protests, they and Khowutzun had been seeking accommodation for members of the First Nation to work on the project “outside the BCIB process.” 39 39 Cowichan Tribes, “Work Stoppage at New Cowichan District Hospital Construction Site,” December 2, 2022, https://cowichantribes.com/application/files/1216/7002/6373/2022_12_02_CDH_Statement.pdf. This accommodation was not granted, and BCIB refused permission for Khowutzun and its member companies to continue working on the project. 40 40 R. Barron, “Work Blockade at New Cowichan District Hospital Site Over, for Now,” Cowichan Valley Citizen, December 13, 2022, https://www.cowichanvalleycitizen.com/news/work-blockade-at-new-cowichan-district-hospital-site-over-for-now-841031.

In response, Jon Coleman and other members of Khowutzun began a protest that halted construction work on December 2. 41 41 R. Barron, “Updated: Picket Line Stops Work at New Duncan Hospital in Dispute Over Union Requirements,” Cowichan Valley Citizen, December 2, 2022, https://www.cowichanvalleycitizen.com/news/picket-line-stops-work-at-new-duncan-hospital-in-dispute-over-union-requirements-840666. The protest lasted eleven days, with construction work resuming on December 13, despite the situation not having been resolved. 42 42 R. Barron, “Work Blockade at New Cowichan District Hospital Site Over, for Now.” Finally, on February 8, 2023, Minister of Health Adrian Dix announced in the BC Legislature that Khowutzun would be able to work on the project “without a change to their workforce,” implying that employees of Khowutzun would no longer be required to belong to one of the Agreement’s affiliated unions. 43 43 Legislative Assembly of British Columbia, Fourth Session, 42nd Parliament, Office Report of Debates (February 8, 2023), at p. 8755, https://www.leg.bc.ca/hansard-content/Debates/42nd4th/20230208pm-Hansard-n260.pdf.

The Coleman affair highlights a tension that exists at the heart of British Columbia’s CBA. On the one hand, the Agreement sought to increase the diversity of the construction workforce in BC, including by increasing the representation of Indigenous workers. On the other hand, by requiring all workers to belong to an affiliated union, the Agreement led to an Indigenous contractor being denied permission to work on the project. Thus, a conflict can exist between the requirement to unionize and the desire for a more diverse workforce. In this case, the requirement to unionize dominated, until the Minister of Health acquiesced to an exception for Khowutzun.

The same tension can be seen in the stated goal of employing local residents. Jon-co Contracting was a local business, yet the requirement for workers to join an affiliated union led to its being removed from the project.

This tension points to a fundamental flaw in British Columbia’s CBA. BCIB claims that its role is to “help expand the province’s labour pool.” 44 44 BCIB, About Us, https://bcib.ca/about-us/. However, by elevating the requirement to join an affiliated union as a condition of employment for the vast majority of workers on Agreement projects, the Agreement can have the perverse effect of shrinking the labour pool. As the Coleman affair shows, it can shut out a local contractor whose workers were qualified to do the work and in fact were already doing it. It effectively excludes all workers who wish to affiliate with other unions or who do not wish to unionize at all. In this respect, it cannot be said to expand the labour pool.

This paper has highlighted a number of ways that British Columbia’s CBA does not align with the defining characteristics of successful CBAs. Each does not tell the full story on its own, but together they paint a picture of a regime that poses some problematic outcomes. These are as follows:

• The Agreement resembles a PLA more than it does a traditional CBA. This matters because these are two separate concepts that should be distinguished. At best, the Agreement can be seen as a mixture of the two.
• Making BCIB the employer of all employees on Agreement projects oversteps the government’s role. Governments should be there to help and support employers, not to supplant them.
• Most fundamentally, the restriction of its workforce to those who join one of its nineteen affiliated unions results in a reduction of both supplier diversity and workforce diversity. The Coleman affair highlights how the Agreement can have a negative impact on both. A reduction of supplier diversity means fewer firms competing for government infrastructure projects, which means higher prices for taxpayers. The cost overruns, which the government blamed in part on labour shortages, may be evidence of this. A reduction of workforce diversity means that the government’s own aims in establishing the Agreement can be undermined.
• As an example, the Cowichan District Hospital Replacement Project received only two bids from construction firms. It has seen cost overruns of at least $559 million, which the BC Minister of Finance attributed at least partly to labour shortages. Allowing access to a wider pool of labour could have lessened the impact of these adverse conditions.

This is not to say that British Columbia’s CBA has no positive effect at all. The statistics on local residents and some underrepresented groups participating in the Cowichan District Hospital Replacement Project suggest that it may have had some positive effect in promoting employment among these groups, although there are some mixed results in this respect as well.

The issue, rather, is whether increasing the demographic diversity of the workforce could be accomplished without requiring workers to join a specified set of affiliated unions. As we have seen, this requirement is itself at odds with the principle of workforce diversity by limiting the kinds of workers that are allowed to work on a project based on their union status.

The following high-level recommendations outline ways that the government can achieve the legitimate goals of CBAs without the perverse outcomes and cost overruns that have been associated with the Agreement.

Respecting this right will, at minimum, involve repealing article 8.1, which states that workers on CBA projects must be or become members of an affiliated union, and amending the clauses relating to the hiring process.

Limiting workers to membership with one of the affiliated unions unnecessarily restricts the labour pool for CBA projects. The current rule discourages firms from competing for government contracts if their workers are members of other unions or are not unionized. This decreases potential competition, which drives up prices.

More fundamentally, requiring workers to join an affiliated union violates their freedom of association. They should be free to join or form a union of their choice, or not to, if that is their preference. Allowing workers the right to affiliate as they wish would not only get better value for taxpayers; it would also respect their fundamental freedoms.

A good CBA requires cooperation between governments and contractors in identifying pools of underrepresented workers. It does not require that government supplant the employer, as BCIB does.

Employers play an important role in labour markets. They are in the best position, together with their workers’ unions if they are unionized, to manage their workforces. Governments should not take on this role. After all, the whole purpose of tendering is to have an external contractor manage the project and the workforce more efficiently than government can.

If the government wishes to increase the representation of certain groups in the construction industry, then its role should be to assist employers in seeking out new pools of labour among those groups. This does not require the government to replace employers.

The government should therefore re-orient the relationship between itself and contractors such that the latter once again become employers of their workers. This will likely render BCIB unnecessary.

The government must not lose sight of the goal of increasing opportunities for underrepresented groups. As we have seen, rigid hiring structures can inadvertently lead to a reduction of workforce and supplier diversity. This is clearly the case with the complicated rules and methods used in British Columbia’s CBA.

Instead of supplanting the proper role of employers and mandating that all workers be or become members of an affiliated union, the government’s role in hiring targets should be characterized more by help and support in connecting firms with social agencies and sources of labour. This can be achieved in the context of a traditional CBA. The government should re-focus on these priorities and support, not supplant, employers in realizing them.

These recommendations amount to an overhaul, if not a complete replacement, of the Agreement. The BC government should not be deterred by this task, however. The exception for Khowutzun that the Minister of Health announced on February 8, 2023, was effectively an admission of a fundamental flaw of British Columbia’s CBA. The exception took away the requirement to join a union—a requirement that the Agreement had placed above other goals, including the engagement of local residents and Indigenous contractors that were members of Khowutzun. The problem is that the exception addressed this flaw in a narrow way and did not fix the Agreement itself.

Going forward, the Government of British Columbia should be more mindful of the distinction between CBAs and PLAs. As discussed above, the Agreement has more of the characteristics of a PLA than of a true CBA. A replacement of the Agreement should follow the principles outlined above, especially those of supplier diversity and workforce diversity.

This paper began on the premise that CBAs can be successfully implemented in Canada. The first part examined CBAs in general and summarized previous Cardus research that provided a framework that governments could use to introduce them in their respective jurisdictions.

The problems outlined in this paper with respect to British Columbia’s CBA do not necessarily pertain to CBAs in general. Rather, they arise because of specific issues within the Agreement itself, especially the requirement that all workers on Agreement-related projects join one of a specific set of labour unions. In making this a condition of employment, the Agreement undermines goals that are more inherent to CBAs in general, as evidenced by the Coleman affair in the construction of the Cowichan District Hospital Replacement Project.

The solution for the Government of British Columbia is to design a CBA that aligns with the best practices, outlined in this paper and in previous Cardus research. This starts with respecting the rights of workers to affiliate as they wish, re-establishing firms as the employers of workers on infrastructure projects, and re-focusing on expanding employment opportunities for underrepresented groups. These are among the features of successful CBAs, and they should guide future collaborations between governments and the construction sector on matters relating to the diversity of the workforce.

The checklist below was originally published in the 2022 Cardus report “A Framework for Implementing Community Benefits Agreements.”* It is a summary of the principles of good CBAs, in a format that is useful for governments.