Tuesday, March 31, 2020

Work refusals and COVID-19

A question COVID-19 raises is when can a worker successfully refuse work as unsafe due to the virus? The short answer (at least in Alberta) is likely to be only when the work is unusually and egregiously dangerous. Or when the consequences for the employer of denying the refusal would be profound.

In Alberta, you can always refuse unsafe work if you have reasonable grounds to believe that the work is unsafe. During a refusal, a worker must paid and the worker can be directed to perform alternative work. A worker cannot be retaliated against for a refusal (or, at least, such retaliation would be illegal…).

A refusal requires the employer to investigate and either (1) control the hazard or (2) tell you the work is safe. If the worker continues to have reasonable grounds to believe the work is unsafe, the worker may file a complaint with OHS and an officer will investigate. The investigation can result in compliance orders (to make the work safe) or a determination that the work is not hazardous.

The real question is in which circumstances can a worker successfully refuse unsafe work? That is to say, in what circumstances can a reluctant employer be forced to make work safe(r)?

(For the purposes of this post, I’m going to set aside cases where workers are providing medical services. These workers have working conditions and professional obligations that muddy the analysis. This article speaks a bit about the challenges for medical workers.)

For example, consider this case in Calgary, where a COVID-positive individual showed up to file legal documents, and thereby exposed the court clerks and security personal. Could the rest of the clerks (who are still on the job) refuse to work the public counter because of the risk of exposure?

A more every-day example might be grocery store cashiers, who must interact with anyone who comes in (undoubtedly included presently asymptomatic carriers of COVID-19) in close quarters. Could they refuse to check out groceries because the work is unsafe?

Alberta’s publications about work refusals assert various versions of “Health and safety hazards that are not normal for the job, or normal hazards that are not properly controlled, are considered a dangerous conditions (sic) that could trigger a work refusal.” (p.5).

I was unable to find a legislative anchor for this “normal for the job” assertion. I expect the thinking is that hazards that are normal for the job should already be controlled to the degree reasonably practicable. And the government wouldn’t want, for example, firefighters to refuse to fight fires (assuming whatever controls are reasonably practicable are in place). So it is only unusual hazards or normal hazards that someone aren’t controlled (e.g., due to changing circumstances) that can give rise to legitimate work refusals.

Assuming this “normal for the job” rule is in play, could grocery cashiers successfully refuse unsafe work on the basis that working in a pandemic is unsafe? An employer might well argue that exposure to infectious disease is a normal hazard of the job. Yet, workers might argue that a highly virulent pandemic that can result in potentially devastating health effects is either not a normal hazard of their job or is a normal hazard that is not adequately controlled.

There have been a number of COVID-related work refusals in Canada. An early one was March 14 among casino workers in Ontario. The workers noted that workers were exposed to COVID-19 during work despite the employer’s cleaning protocols, that casino chips were never cleaned, and the employer was not following its own cleaning protocols (e.g., too few staff cleaning, the cleaning was being done with the wrong chemicals, no hot water for hand washing). The Ministry of Labour did a seemingly superficial investigation and found the work was safe.

The case law on refusals during outbreaks of infectious diseases appears to find work with an inherent risk of exposure (transportation in the link above) and where the employer takes reasonable steps to comply with public-health advice will be deemed safe. Obviously context matters: workers at greater risk of ill-health may face a lower bar in establishing the work is unsafe.

So coming back to the grocery cashier example, an employer might successfully argue that regular cleaning of the cashier’s workspace, the provision of PPE (e.g., gloves), and administrative rules (e.g., don’t touch your face, wash your hands) constitutes reasonably practicable controls. Thus workers might not be able to require the employer to take further steps (e.g., provide masks or barriers).

This analysis of how the right to refuse plays out reflects a couple of underlying dynamics:
  1. Workers and employers often want different levels of safety. Employers typically want the least expensive form of safety necessary to keep their business operating. Workers might prefer a more costly (and likely more effective) set of hazard controls. This conflict is rooted in a basic struggle to capture the surplus value of labour. And employers are often able to use their greater labour market and legal power to limit injury-prevention efforts to only those instances when injury prevention is cost-effective.
  2. OHS is about maintaining production and safe workplaces is a bi-product, although not a necessary one. The point of OHS laws is to prevent class conflict over the injury and death of workers from disrupting production and imperiling social stability. It does this in a couple of ways. OHS laws channel conflict into manageable processes (e.g., a work refusal versus a wildcat strike) that also frame the government as a neutral actor. But the government can (and usually does) write and interpret the law in ways the minimally impact production (e.g., setting a low bar around what is considered safe). OHS rules also make workplaces somewhat safer, but not necessarily very safe.
So given the likely limited utility of the right to refuse in the face of COVID-19, what could the grocery workers do to make their workplaces safer. Here are some strategies:
  • Social pressure: Employers are (sometimes) subject to social pressure. A group of workers asking for reasonable safety precautions may be able to get what they want because an employer may feel badly about refusing their demands. This is especially the case when the target of this pressure is within the workers’ social circle (e.g., a supervisor or the owner of a small business). In this case, the demand might be a re-organization of work processes or the provision of additional protective equipment. For example, an independent grocery store in Edmonton has (voluntarily, I should add) hung clear plastic between cashiers and customers at the checkout to create a barrier.
  • Public pressure: Shit-talking your employer on social media is cause for termination in most workplaces. But it can also be an effective pressure tactic, especially if your employer is acting in an egregiously unsafe manner during a pandemic. The internet provides a variety of tools that can allow workers to try and mask their identities.
  • Insubordination: No one can force a worker to perform work, they can only attach sanctions for not doing the work. A work refusal (e.g., not coming in, not performing a particular piece of work, performing work in non-sanctioned ways) outside of the OHS Act is insubordination and may attract discipline (e.g., termination). But, if everyone does it, it can be a powerful tool for pushing employers take more action than they might otherwise be prepared to do. The risk, of course, is getting fired. The more people who cooperate with the insubordination, the harder it is for the employer to discipline.
  • Quit: While quitting in a terrible job market is not a viable option for everyone, it is always an option. And, as financial support for the unemployed is loosened during the pandemic, it becomes a somewhat more accessible option (which tells you a lot about why EI eligibility rules have been tightened over time). The article linked above about home care nurses being asked to perform their jobs without adequate protections is evidence that workers will sometimes choose their health and safety (or that of their loved ones) over a paycheque.
It will be interesting to see how the many refusals over COVID-19 going on right now play out over time. An interesting case is a number of nurses in Alberta who refused to perform COVID-19 swab tests without N95 masks.

Alberta Health Services (AHS) maintained N95 masks are not required for swab tests. The nurses argued that the method of transmission is not clear so the precautionary principle suggests use is warranted. Eventually, an agreement was reached that staff could have access to the masks.

AHS could have likely held firm and even chosen to discipline these nurses if they continue to refuse (assuming the OHS investigation concluded the masks were not required). But context matters and the politics of discipline are bad.

Nurses are a very sympathetic group of workers who are placing themselves in harm’s way during a pandemic. And all they were asking for were slightly more effective masks. And, if AHS choose to discipline them, AHS ran the risk of escalating job action (i.e, a wildcat strike) in healthcare .

Precipitating a wildcat strike (by disciplining the nurses) at the moment workers’ labour power is the greatest (during a pandemic and when the public is scared) would have been a bad outcome. So the employer accommodated the nurses' demands for greater safety.

-- Bob Barnetson

Tuesday, March 24, 2020

Workers' compensation and COVID-19

One of the outcomes of COVID-19 is that some workers will contract the virus at work. Of those, some will become ill and will miss work. Some will require medical treatment. Some will likely die. In theory, these workers are eligible to file a claim through the workers’ compensation system. These claims have the potential to impose huge costs on workers’ compensation systems.

One of the challenges of dealing with infectious disease claims is determining whether the disease was caused by a workplace exposure (rather than a community exposure). This challenge is particularly difficult during a pandemic, when a large percentage of workers (possibly 60-70%) will contract the illness and they might well contract the illness through non-work contact.

Alberta’s Workers’ Compensation Board (WCB) has issued some preliminary advice about COVID-19 claims [NOTE: See update below]. The crux is that the WCB will be accepting claims where COVID-19 is a “direct result of their employment.”

The WCB’s Policy 03-01 (Occupational Diseases) addresses the question of work-relatedness of infectious diseases by applying a three-part test. Infectious diseases are eligible for compensation if:

a) the nature of employment involves sufficient exposure to the source of infection, and
b) the nature of employment is shown to be the cause of the condition, or
c) the nature of employment creates an increased risk of exposure for the worker.

I think the sensible way to read these requirements is a claim must meet (a) and either (b) or (c). The case-law I quickly looked at wasn’t perfectly clear. The WCB’s fact sheet on infectious diseases suggests all three tests must be met, which can't be right given the “or” between (b) and (c).

Assuming the (a) + (b) or (c) reading of the policy is correct, I wonder a bit about the examples the WCB provides around COVID-1 claims. (Note that the WCB does flag that each claim will assessed on its merits.)

The COVID-19 Fact Sheet goes on to assert:
… if your worker is at greater risk than the general public of contracting the virus while at work, and they lose time from work after contracting the virus, report the claim to WCB.

However, if your staff is not at greater risk than the general public of contracting the infection, do not report. Examples include people who believe they caught the COVID-19 virus from a co-worker or client.
Accepting claims from health-care workers who treat COVID patients seems sensible. I’m less convinced by the analysis presented for cafeteria and grocery workers. Given the virulence of COVID-19 and the fact that grocery and cafeteria employment require workers to extensively interact with the public (who will include carriers), I would think these workers might well meet the (a) + (c) test. Basically, the workers’ employment puts them at greater risk than the general public.

It may be that the WCB is relying upon the “but-for” test here. The “but-for” test basically asks whether the work exposure was necessary for the injury to have occurred. That is to say, if not for the work, would the injury have occurred? If yes, then the work did not cause the injury.

In a pandemic, all cases of COVID-19 (including those experienced by health-care workers) could well occur without a workplace exposure. So, if the WCB applies the “but for” test to health-care workers who treat COVID-19 patients, shouldn’t they also be excluded from coverage?

It is important to recognize that the WCB is trying to provide broad advice and must adjudicate each claim on its merits. Consequently, actual claim acceptance and denial may play out differently than the examples suggest.

That said, the examples provided may cause some workers and employers not to file claims. The narrow reading of the policy evident in the examples may also shape how WCB staff apply the policy when faced with claims.

An interesting question to ask is who benefits from a narrow reading of Policy 03-01? One answer is that the WCB benefits. By restricting access to compensation, the WCB will protect its accident fund from a potentially huge number of claims (including a relatively large number of fatalities). The cost of these claims would hit at the same time as the WCB’s accident fund is likely taking a beating in markets. Such a draw on the fund might also trigger premium increases, thereby creating employer (and likely government) backlash.

I don’t mean to suggest that the WCB staff are acting like Bond villains (they are likely simply trying to adjudicate claims in circumstances of considerable ambiguity). But I do think it is worthwhile considering how organizational pressures and logics can influence how organizational actors interpret policy. And this can lead to outcomes (e.g., widespread claims denials) which sit uneasily with the purpose of workers’ compensation, which is to compensate workers for work-related injury and death.

-- Bob Barnetson

Update: So today (March 26) the WCB issued a new advice sheet to employers that removes the examples fro the March 9 sheet and adds this paragraph:
A claim is likely to be accepted if a worker contracts the illness and is performing what the province deems to be an “essential service” that puts them in regular contact with the general public. A worker will also likely be covered in the event of a widespread outbreak at their place of work.
This seems to broaden the group of workers who would likely have claims accepted, although who is an essential service has not yet been established. There are also other changes too complex to elaborate.

Tuesday, March 17, 2020

How Alberta’s Bill 1 Constrains Workers’ Rights to Protest

This post originally appeared on the Canadian Law of Work Forum.

On February 25, Alberta’s government introduced Bill 1 (the Critical Infrastructure Protection Act) in the Legislature. If passed, Bill 1 will allow police to arrest without warrant anyone who is present at any location defined as essential infrastructure or who interferes with the operation of the infrastructure “without lawful right, justification or excuse”.

The legislation is putatively designed to reduce the likelihood of economic disruption caused by civil disobedience, such as the recent railway blockades in support of the Wet-suwet’en First Nations. This Act will, however, significantly constrain the ability of workers to exert pressure on employers in Alberta.

Bill 1: An Overview
Bill 1 prohibits entering, damaging, destroying or interfering with the operation of essential infrastructure “without lawful right, justification, or excuse.” It also prohibits aiding, counselling, or directing someone else to do so.

In addition to arrest without warrant, those found in violation of the act face a minimum fine of $1000 for a first offense, with a maximum fine of $10,000. Second and subsequent offences see the maximum fine rise to $25,000. Each offence can also result in a jail term of not more than 6 months. Each day that an offense continues is considered a separate offence under the Act.

An essential infrastructure includes pipelines, utilities, mines and oil production sites, highways, railways, powerplants, agricultural operations, and dams. Bill 1 also allows cabinet to extend the list of essential infrastructure through regulation.

Of particular note is that a highway, as defined in Section 1(1)(p) the Traffic Safety Act is a verybroad term. It includes any publicly- or privately-owned thoroughfare, street, road, trail, avenue, parkway, driveway, lane, alley, square, bridge, or causeway, including adjacent sidewalks and boulevards.


Alberta’s United Conservative government asserts that Bill 1 is necessary to protect the province’s economy from disruption. Numerous commentators have noted that the behaviours prohibited by Bill 1 are already illegal under other statutes. In this way, Bill 1 is unnecessary duplication and may be simply a political nod to government supporters, including employers.

An alternative analysis is that Bill 1 is designed to heighten the cost associated with effective pressure tactics that workers might exert. Precluding individuals from being present on sidewalks or boulevards “without lawful right, justification or excuse” would, for example, dramatically raise the risk and cost associated with information and solidarity pickets.

For example, during the recent lock-out of Co-op refinery workers in Regina, Unifor locals and allies have shut down fuel distribution depots and conducted information pickets of gas stations in order to exert economic and reputation pressure on the refinery. These actions occurred on sidewalks, boulevards, and driveways and could constitute entering essential infrastructure without lawful reason. Bill 1 would allow the police to immediately arrest such protestors and require a minimum $1000 fine for a first offence. The spectre of these consequences will undermine the viability of such actions in the future, thereby benefitting employers.

Of particular concern to Alberta labour activists is the impact of Bill 1 on public-sector labour relations. Virtually every public-sector collective agreement in Alberta is up for negotiation this year. The United Conservative government has signalled its intention to reduce public-sector compensation. To help achieve wage rollbacks, the government delayed arbitrations last summer. It then gave itself the power to issue binding and secret bargaining mandates to public-sector employers. In light of this, unions expect the government is also considering back-to-work legislation, a strategy recommended by Alberta’s recent Blue-Ribbon panel on provincial finances.

Consequently, Alberta’s public-sector unions are taking unprecedented steps to prepare their members for job action. This member mobilization—including information picketing outside public facilities—is publicly framed as preparing for legal strikes. At mobilization events, however, wildcat strikes are being explicitly discussed. Many activists suggest that wildcat strikes may be much more effective at causing the government to shift its position at the bargaining table. Although it has a conservative reputation, Alberta has a recent history of wildcat strikes in health care (2000, 2012), construction (2007), and the prison system (2013).

For the government, Bill 1 would be a useful supplement to existing laws prohibiting job action other than legal strikes. Public-sector institutions could be easily named as essential infrastructure through regulation and picketing prohibited. The threat of immediate arrest would likely reduce the willingness of workers to picket during a wildcat. This, in turn, degrades workers’ ability to solicit public support and/or disrupt operations in order to pressure the government.

While there are no publicly available analyses of the constitutionality of Bill 1, it appears to sit uncomfortably with our Charter freedoms of expression and of peaceful assembly. Both of these rights permit peaceful protests in public spaces, such as sidewalks or lands associated with public buildings. While the use of Bill 1 to suppress picketing and other worker actions would undoubtedly be challenged, such challenges are slow. In the meantime, the law continues to operate.

If this analysis of Bill 1’s effect on labour relations is correct, Bill 1 represents another step in Alberta’s rolling back of workers' rights. Alberta did away with card-check certification, rendered the public-sector replacement worker ban ineffective, and opened up holes in over-time pay provisions last summer. It then interfered with contractually required arbitrations and allowed itself the power to issue secret and binding bargaining mandates. Now it appears poised to restrict workers’ freedom of expression and peaceful assembly.

-- Bob Barnetson

Tuesday, March 10, 2020

Judge strikes down a portion of Canada's sex work laws

In December, AU opened a new course (LBST 415: Sex work and sex workers). One focus of the course is how governments regulate sex work and sex workers, with an eye to which approach yields the best results for sex workers.

In 2014, Canada amended the Criminal Code to decriminalize the sale of sexual services (in most instances). The purchase of sexual services and any acts designed to facilitate the sale of sexual services remain criminalized. 

This approach is often called the Nordic model and is intended to extinguish demand for sex work (although it has not) while making it safer for sex workers to seek police assistance (which it also has not). The law was supposed to have been reviewed by the government in 2019, but it was not.

An interesting development last month was that an Ontario judge struck down portions of Canada’s law on sex work that criminalized procuring, advertising, or materially benefitted from the sale of someone’s sexual services. The case involved a couple who ran an escort agency.

The judge’s rationale was prohibiting advertising violates freedom of expression while the laws against procuring and materially benefitting violate the Charter guarantee of security of person. The crux of the rationale is that the prohibitions make it difficult for sex workers to screen clients, work cooperatively, and to purchase certain services, all of which make sex work safer.

Those opposed to the decision frame this ruling as protecting pimps who traffic in exploited women and girls. While rhetorically powerful, this analysis ignores that there is significant nuance in the “management services” that sex workers may purchase as well as that human trafficking remains illegal.

This line of critique also contributes to the conflation of sex work and human trafficking. While there is overlap (some trafficked women and girls are involved in sex work, and some of this involvement in sex work is against their will), most sex work appears to be consensual activity.

That is not to say there isn’t an element of exploitation involved in sex work. But it is important to identify that sex workers have agency. A part of recognizing this agency is providing sex workers the opportunity to engage in sex work in ways of their choosing. Analysis of New Zealand (where sex work has been legalized) suggests this model yields the best outcomes for sex workers.

-- Bob Barnetson

Tuesday, March 3, 2020

Research: New study on injury under-reporting

The American journal of industrial medicine has recently published an article entitled “Examining the impact of occupational health and safety vulnerability on injury clam reporting in three Canadian provinces.” The survey found that 63.5% of the 326 respondents with physical workplace injuries that required time off or professional medical aid did not report them to the WCB. This finding is broadly consistent with other Canadian studies.

Reporting was higher among non-vulnerable workers who were exposed to hazards. Women, part-timers, those with postgraduate educations, and those working in education, health care and public administration as well as non-unionized workers were less likely to report injuries (although the difference was not statistically significant).

This study adds to the growing literature that suggests WCB claims data is a poor proxy for the true level of occupational injury, even when the injury is serious. Put differently, WCB claims data can provide a misleading picture of the extent of serious injury in the workplace. This has implications for public policy, which is often developed and evaluated based upon claims data.

-- Bob Barnetson