Angry callers, complex calls: How collective bargaining protects call centre workers

Have you had this experience? You have a problem with a product, so you visit the company website. The FAQs don’t answer your question, so you try the website chatbot, but none of the AI-generated responses resolve the issue. Some time later, you’re connected with a human agent, but by this time you are extremely frustrated. Does the call go well?

Sean O’Brady is an assistant professor in the Department of Human Resources and Management at the DeGroote School of Business. His work investigates how AI and algorithmic management technologies impact working conditions for low-wage workers, and how unions can mitigate harmful uses of those technologies. O’Brady’s latest publications investigate union power and job quality in call centres as well as spatial control and union resistance in digital remote work.

We asked O’Brady a few questions about these papers and how collective action can improve outcomes for workers and companies.

Can you provide a high-level overview of your work on union activity in North American call centres?

Broadly speaking, I look at what unions do to reduce the harmful effects of technology in call centre environments. Because we don’t have the right model for the 21st century, union power has declined. Despite this, we still see evidence of positive union impact in the industry.

Companies don’t want to invest a lot in call centres. They invest just enough to make it work; they’re always trying to cut costs, and through the last 20 years and the rise of different technologies, it’s a lot easier to outsource call centre work. Call centre workers were already losing jobs, and now they’re heavily performance managed. With the rise of AI, every detail of their work is scrutinized and monitored.

Can you explain the term “spatial control” as it applies to remote work environments, as well as why and how workers are resisting it?

A spatial control regime is where employers can use the space where you work as leverage to force changes in employment contracts and working conditions. It gives employers a new source of leverage that they didn’t previously have. In unionized contexts it becomes problematic because this is an area where unions and employers negotiated less on in the past and hence was largely unregulated. Then work-from-home arrangements were introduced very quickly at the onset of the Covid-19 pandemic.

A few scenarios that have resulted include employers passing on costs that come from working at home, like limiting refunds for required equipment or making employees absorb the cost of disruptions such as internet or power outages. Grievances have decreased because workers are worried about losing access to work-from-home. They often have less voice and less solidarity since they no longer see colleagues and union representatives in person.

In the Canadian context, we don’t have rights to work-from-home, and even in unionized environments, it’s difficult to negotiate better rights to work-from-home. Part of the issue with the spatial control regime is that workers like working from home.

It becomes a carrot, and on the one hand you want to reduce worker exploitation, but because workers like working from home, they often don’t want to rock the boat, and they aren’t willing to strike or mobilize in that environment because of this tension.

But unions still negotiate important protections, like limits on digital monitoring and protections from layoffs or arbitrary termination. Some have even negotiated provisions that prevent managers from arbitrarily refusing requests to work from home.

How are automated algorithms and AI impacting job quality?

One of the big issues workers face in call centres is that they’re required to use the company’s tools, they’re constantly being monitored, and this data is filtered into the AI systems themselves. Another issue is that companies are making it more difficult for customers to connect directly with live service representatives. They are encouraging them to use AI systems like company chatbots, even if there are problems with them, because organizations want to cut labour and improve their tools. Customers are also entangled in this, and this amplifies the tensions they have with the workers, where the workers keep telling them to use the systems.

So, we do see a high number of these workers on medication, and an increase in the number of people having to take stress leave.

What about job security?

In a collective bargaining scenario, you are required to have just cause to fire somebody. But what we see is that the use of these technologies, especially the constant tracking, is putting stress on these protections.

It used to be [that if] you have an employee who did a bad call, the manager would bring them up and say, “Maybe you’re not qualified.” Then the union would get involved and say, “Maybe you didn’t train them well. It was a bad week for this person, but they have good statistics for the last year.”

Maybe it’s someone who complained about working conditions, or who took a lot of leave; they try to justify termination. But even top performers make mistakes. This is evidence that can be produced for anyone, and arbitrators look at the evidence and often don’t see it as targeted. Unions are still working hard to protect these workers, but some say it’s getting tougher.

What are the costs for employers who treat their workers like this?

First of all, when you’re not investing in workers and the call centre environment, not giving workers rights, and people are leaving, it comes at a significant cost to customer service. It doesn’t produce high-quality customer service, and it doesn’t produce gains for both parties.

We have survey evidence that shows customer service outcomes are ameliorated when workers have autonomy to use their experience and are not scared of punitive management.

In a non-union environment where you don’t have protections and there’s a culture of fear, they won’t get that information about the negatives of production. Also, experiments with AI can be costly for organizations. AI-related stress and burnout can increase expensive forms of turnover and absenteeism. By promoting more sustainable uses of these technologies, unions are lowering these costs for employers.

How can unions protect workers from labour replacement and work intensification?

Unions can protect people from losing their jobs. For example, many collective agreements have “bumping rights”: if you’re a call centre worker and you’re going to lose your job, this gives you rights to another position in the company — you can move around.

I think unions need more rights, and employers fear expanding those rights because it’s not part of our culture. In countries where workers have stronger legislated labour rights and data protections, better practices can become normalized. For example, our team examined cases in Germany and Italy where you can only monitor teams, not individuals. Managers often agree with it because it creates a kind of professionalism, and there are fewer costs from unhealthy work. You can have an environment where quality service is recognized and voices are heard, but this works best in contexts with strong legislated rights and where unions are well organized in the workplace.