In 2016, when Amazon workers began organizing at a warehouse in Chester, Virginia, the company tracked where employees gathered, posted anti-union messaging in bathroom stalls, and held town halls to discourage them. The National Labor Relations Board forced Amazon to admit, in writing, that it had illegally surveilled and threatened workers. The union drive still failed.

That sequence — anticipate, disrupt, escalate — isn’t random, according to new research from Timothy Werner, professor of business, government, and society and Wade T. and Bettye C. Nowlin Centennial Professor of Business Administration at the McCombs School of Business at The University of Texas at Austin.

Rather, it’s a systematic playbook that a wide variety of companies deploy far beyond the warehouse floor, Werner says. He calls it “organizational repression” — a term he borrowed from research on how governments suppress political dissent.

He applies the expression to the ways companies manage collective pressure from stakeholders: non-shareholder groups with an interest in a company, such as employees, activists, and communities. The term covers companies’ actions from union-busting to greenwashing under one strategic umbrella.

“We were trying to find a more encompassing term that would capture all these different ways in which organizations, as opposed to states, could engage in this behavior,” Werner says. “These things are more alike than scholars have previously recognized.”

Strategy Against Stakeholders

Past research has assumed that companies are largely respectful toward stakeholder activism, Werner says. They might resist stakeholder pressure, but they might also collaborate or simply ignore it.

“What we wanted to show with this paper was that there are actually ways in which firms can undermine that ability to organize in the first place,” he says.

With Natalie Holzaepfel and Olga Hawn, both of The University of North Carolina at Chapel Hill, Werner built his framework around three phases that stakeholder movements typically move through.

• Emergence, in which individuals privately notice a grievance.
• Coalescence, in which people start organizing.
• Formalization, in which the group becomes a structured movement with allies.

For each phase, the researchers identify matching corporate strategies that can prevent or discourage it.

Emergence: Stop it before it starts. Companies work to convince people there’s nothing worth mobilizing over. Exxon Mobil, for example, began funding research downplaying climate change as early as the 1970s — years before it became a target of activist campaigns.

Another strategy, Werner says, is to cultivate a reputation as being unreceptive to activism, making mobilization feel pointless.

Coalescence: Make joining costly. Once a movement begins to coalesce, companies target the people most likely to join. When Delta Air Lines faced a 2024 unionization push among flight attendants, it offered a carrot: a 5% pay raise, but only to nonunion workers.

Other companies have taken an opposite approach: Brandish sticks such as demotions, changing schedule to conflict with meetings, or implicitly threatening to fire known organizers.

Formalization: Divide and isolate. If a movement fully organizes — recruiting members and forming alliances with outside groups — the most effective corporate response shifts toward fracturing the coalition itself, Werner says.

The pipeline company Energy Transfer, facing protests over its Dakota Access Pipeline, allegedly hired private security companies to disrupt activist networks. It also filed lawsuits against protest groups, eventually winning more than $600 million from Greenpeace.

Not Risk-Free

The researchers don’t pass any ethical judgments on organizational repression, Werner emphasizes. They simply propose the theory that it’s a systematic and underexplored set of tactics that warrants further study.

“We take no stance as to whether repression is good or bad,” Werner says.

It also isn’t guaranteed to work, he adds. A company that moves too aggressively may risk a backlash that strengthens the very movement it’s trying to stop.

The researchers’ next step is to test the theory empirically, using data such as whistleblower reports, lawsuits, and leaked corporate documents. Says Werner, “We want to see how often — and how effectively — companies actually deploy these tactics in practice.”

“Organizational Repression of Stakeholder Collective Action” is published in the Academy of Management Review.

Open-plan offices entail a clearly increased risk of workplace bullying compared with employees having their own office or sharing with just a few colleagues. This is shown in research from Linköping University, Sweden. 

“Increased bullying is a tangible negative consequence of how you choose to organise the workplace. It’s important to highlight this, as it hasn’t previously been examined,” says Michael Rosander, professor at the Division of Psychology at Linköping University.

Open-plan offices, where many employees share the same space, have become increasingly common. Employers often justify this development as a way to use premises more efficiently and to encourage creative interactions between employees. However, research has shown that open-plan offices do not promote health, job satisfaction or productivity.  

Until now, it has been unclear whether open-plan offices also affect the risk of bullying and employees’ motivation to look for another job. Through surveys of more than 3,300 randomly selected individuals in employment in Sweden, Michael Rosander has now provided an answer. The results are published in the journal Occupational Health Science. 

Thirty per cent of those with some form of office-based work reported that they worked in a traditional open-plan office with no access to private space. Thirteen per cent worked in so-called activity-based offices, where employees spend part of their time in an open-plan environment but also have access to designated rooms for tasks requiring peace and quiet. The remainder had their own office or shared one with only a few colleagues.

For traditional open-plan offices, the survey responses showed a clearly increased risk of bullying compared with those who had their own office or shared an office with only a few colleagues. The difference remained regardless of factors such as personality traits and the extent of remote working. This suggests that the problems are indeed caused by the work environment in the office.  

The researchers’ explanation is that in traditional open-plan offices it is easier to notice colleagues’ shortcomings and become irritated by them. If someone gets frustrated and takes it upon themselves to “do something about” a colleague’s behaviour, and there are no clear guidelines for handling such situations, there is a risk that it may escalate into bullying. Those who are subjected to bullying lack access to a private space for retreat. 

Activity-based open-plan offices, by contrast, showed no increased risk of bullying, likely due to the availability of private spaces. However, in both types of open-plan office, employees were more likely to consider changing jobs. One possible explanation is that activity-based offices also involve more distractions, according to Michael Rosander.

For employers who have introduced, or are planning to introduce, open-plan offices, there are some lessons to be learned. One is to be prepared to deal with irritation and conflicts before they escalate. Another is the importance of providing rooms where employees can work undisturbed. Placing individuals with similar needs and tasks near one another may also reduce the risk of disruption.

“Traditional open-plan offices are in themselves negative for the individual, for productivity, and make people more likely to leave their job. Social interaction also suffers. So it’s worth considering how to handle it,” says Michael Rosander.

To get people to pay attention, you have to make it engaging. But what makes content engaging often comes at the cost of detail – shaping what people learn and what they think they’ve learned. The result: People can come away with the wrong idea, even when what they read isn’t factually wrong.

That tension sits at the core of research from Marta Serra-Garcia, a behavioral economist at the University of California San Diego’s Rady School of Management. The study, published in the American Economic Review, examines how incentives in the online attention economy shape the way scientific information is communicated – and what readers ultimately take away from it.

A trade-off in the attention economy

You don’t need bad actors for people to get the wrong idea. Incomplete information can be enough.

Crucially, the research finds that attention-grabbing summaries are not more likely to be factually inaccurate. Instead, they tend to include less information – especially key details about how studies were conducted.

“This is not a simple story that clickbait is bad,” said Serra-Garcia, associate professor of economics and strategy and Phyllis and Daniel Epstein Chancellor’s Endowed Faculty Fellow at UC San Diego’s Rady School. “You need to get people’s attention in order for them to learn something, and it’s good to encourage curiosity. Yet there’s a trade-off: Material designed to engage can also unintentionally contribute to the kinds of misunderstandings that can fuel misinformation.”

The finding comes from a large, multi-stage experimental study in which freelance writers produced nearly 600 summaries of actual scientific research, and more than 3,700 participants were then tested on what they learned from them.

Why “in mice” matters

In one study used in the experiment, a compound in broccoli reduced cancer cell growth – in mice. Leave out those last two words, and the finding can sound far more directly relevant to human health than it actually is.

“Why can’t we say ‘in mice’?” Serra-Garcia said. “It’s not very hard to add. It’s two words. But once you say ‘in mice,’ maybe fewer people will click.”

Study results were consistent. Summaries written to attract attention were shorter, easier to read and more engaging – but included less detailed information, especially about sample sizes and methods.

Given the option to seek out more information, most readers did not. That mirrors real-world behavior: Studies of social media use suggest most content is shared without users ever clicking through to read more.

Among those who relied on summaries alone in Serra-Garcia’s study, knowledge dropped by about 6-7 percentage points. Readers were also more likely to draw incorrect conclusions – such as assuming findings applied to humans or reflected firm medical guidance.

Inside the experiments

To isolate these effects, Serra-Garcia conducted a multi-stage experimental study. In the first stage, 149 freelance writers produced nearly 600 summaries of the same set of studies – covering topics such as cancer, sleep, vaccines and climate – under different instructions: to inform readers accurately, or to attract attention by encouraging clicks or shares. 

In the second stage, more than 3,700 participants read those summaries under different conditions, including whether they could click through for more information.

The results held across experiments: Attention-driven summaries increased engagement and prompted some readers to learn more – but left many others with less complete understanding.

AI and the attention economy

The same pattern emerged when a human wasn’t doing the writing. In additional tests, when a large language model was prompted to attract attention, it also produced less detailed summaries – suggesting the effect is driven less by who creates the content than by the objective it’s optimized for.

For Serra-Garcia, the findings point to an ongoing challenge for researchers, journalists and institutions alike.

“How do you make science engaging and important to readers,” she said, “without missing the essentials that convey the full picture?” 

The research was funded in part by National Science Foundation grant no. 2343858. 

Read the full study: “The Attention – Information Trade-off.”