Sources of Stress at Work: Understanding What Really Causes Workplace Stress

Your stress and resilience programmes might just be wrong

If your stress management strategy begins and ends with teaching people to breathe more slowly, you have not addressed stress. You have addressed symptoms while the building stays on fire. Or, as I saw it put in the FT many years ago, you’re breaking someones leg and teaching them how to make a splint.

Work-related stress is one of those topics where everyone agrees it matters and almost nobody agrees on what to do about it. Organisations commission wellbeing surveys, launch mindfulness programmes, distribute fruit baskets, and wonder why nothing changes. The answer, we would argue, is that most stress interventions are aimed at the wrong level. They focus on helping individuals cope with stressful conditions rather than changing the conditions themselves.

This is not entirely the fault of the organisations involved; the psychology of stress has historically emphasised individual appraisal and coping, and there is a well-funded industry devoted to selling personal resilience as the solution to structural problems. But the research tells a different story: workplace stress is overwhelmingly a product of how work is organised, managed, and experienced, and the most effective interventions target those organisational factors directly.

This article examines the major sources of stress at work, the theoretical models that explain how those sources produce harm, and the practical implications for organisations that want to do more than gesture at the problem.

The HSE Management Standards: Six Areas That Matter

The UK Health and Safety Executive’s Management Standards represent one of the most practically influential attempts to specify what, exactly, makes work stressful. Rather than offering a grand theory, the Management Standards identify six key areas of work design that, if not properly managed, are associated with poor health, lower productivity, and increased absence. They are:

• Demands (workload, work patterns, and the work environment),
• Control (how much say people have in the way they do their work),
• Support (the encouragement, sponsorship, and resources provided by the organisation, line management, and colleagues),
• Relationships (promoting positive working and dealing with unacceptable behaviour),
• Role (whether people understand their role and whether the organisation ensures they do not have conflicting roles), and
• Change (how organisational change is managed and communicated).

What makes the Management Standards valuable is their specificity. They do not simply say “reduce stress.” They point to concrete, measurable aspects of the working environment that can be assessed, discussed, and improved.

The main limitation of the Management Standards, in our view, is that they are descriptive rather than explanatory. They tell you what to look at, but they do not explain the psychological mechanisms through which these factors produce harm. For that, we need theory.

One illuminating lens here that we like (and view many things through) is the predictive processing framework, which models the brain as a prediction machine wired to minimise surprise. On this account, working conditions characterised by low control, ambiguity, and poorly managed change generate sustained prediction error — a state the brain experiences as threat. This helps explain why predictability and autonomy are not merely pleasant features of a good job but structural conditions for a brain that is, at a fundamental level, trying to maintain coherent models of the world. Our article on Predictive Coding: The Brain as a Prediction Machine brings this further to life.

Karasek’s Demand-Control Model: The Grandfather of Work Stress Theory

Robert Karasek’s (1979) demand-control model is one of the most influential theoretical framework in the history of occupational stress research. Its central proposition is elegant: it is not demands alone that produce strain, but the combination of high demands and low control. A surgeon performing a complex operation is under enormous demand, but she also has substantial control over how she works, what decisions she makes, and how she manages the situation. A call centre worker facing relentless targets has equally high demands but virtually no control over the pace, content, or methods of their work. According to Karasek, the surgeon is in a high-strain situation but also a potentially active and stimulating one, while the call centre worker is in the most toxic quadrant of the model: high demand, low control.

The model generates four combinations:

• Low strain (low demands, high control),
• Passive (low demands, low control),
• Active (high demands, high control), and
• High strain (high demands, low control).

The high-strain combination is associated with the greatest risk of cardiovascular disease, psychological distress, and burnout. The active combination, interestingly, is associated with learning and personal development, which suggests that demands are not inherently damaging when people have the autonomy to meet them on their own terms.

The model was later extended to include social support as a third dimension, creating the demand-control-support model. The addition was important because it captured something the original formulation missed: that isolation compounds the effects of high demand and low control, while social support buffers them. The most dangerous working conditions, according to this extended model, are those combining high demands, low control, and low social support, a combination sometimes called “iso-strain.”

The demand-control model has accumulated decades of epidemiological evidence linking its predictions to cardiovascular outcomes, mental health problems, and musculoskeletal complaints. But it has also attracted significant criticism for its simplicity. Real jobs do not divide neatly into four quadrants. Control is not a single dimension; it encompasses decision authority, skill discretion, and temporal autonomy, among other things. And the model says nothing about reward, recognition, fairness, or the meaning people derive from their work, all of which turn out to matter a great deal.

Siegrist’s Effort-Reward Imbalance: When Work Doesn’t Pay Back

Johannes Siegrist’s (1996) effort-reward imbalance (ERI) model addresses a gap that Karasek’s framework leaves open: the question of reciprocity. Siegrist’s central argument is that work is built on an implicit social contract: people invest effort and, in return, expect adequate rewards (see Psychological Contracts too). When the effort invested consistently exceeds the rewards received, the resulting imbalance produces sustained stress, with measurable consequences for cardiovascular health, mental health, and wellbeing.

Crucially, Siegrist defined rewards broadly. They include financial compensation, certainly, but also esteem and recognition, career opportunities (including job security and promotion prospects), and status consistency. The model therefore captures something that many people intuitively recognise but struggle to articulate: you can be well paid and still deeply stressed if your work is unrecognised, your career prospects are stagnant, and your professional status feels precarious. The effort-reward imbalance is not about money alone; it is about the fundamental sense that your investment in work is being reciprocated.

The ERI model has been validated across a wide range of occupational groups and countries, with particularly strong evidence linking effort-reward imbalance to coronary heart disease, depression, and poor self-rated health. Its practical implication is clear: organisations that expect sustained high effort without providing commensurate recognition, development opportunities, and fair compensation are generating stress at a structural level, regardless of how many resilience workshops they offer.

The JD-R Model: A More Flexible Architecture

The Job Demands-Resources model, developed by Bakker and Demerouti (2007) and covered in detail in our separate article on the topic, emerged partly from a frustration with the rigidity of earlier frameworks. Where Karasek specified demands and control, and Siegrist specified effort and reward, the JD-R model proposes that every job has its own unique set of demands and resources, and that the balance between them determines whether people experience strain, engagement, or some combination of both.

The model’s contribution to understanding stress is its dual-process structure: demands drive a health impairment process leading to exhaustion and burnout, while resources drive a motivational process leading to engagement and performance. Resources can also buffer the impact of demands, meaning that high demands are less damaging when people have adequate autonomy, support, feedback, and development opportunities to draw on.

From a stress perspective, the JD-R model’s flexibility is both its strength and its challenge. It accommodates virtually any source of stress (any demand that requires sustained effort qualifies), which makes it highly applicable but also, as critics have noted, potentially unfalsifiable. Nevertheless, we very much like this model. Its diagnostic utility is considerable: it provides a structured way to map the specific demands and resources in any given role or team, which is precisely what is needed for targeted intervention.

Challenge Stressors vs. Hindrance Stressors: Not All Demands Are Created Equal

One of the most important refinements to our understanding of workplace stress came from LePine, Podsakoff, and LePine (2005), who demonstrated that the relationship between demands and outcomes depends fundamentally on the type of demand. Drawing on earlier work by Cavanaugh and colleagues, they distinguished between challenge stressors and hindrance stressors.

• Challenge stressors: are demands that, while effortful, are experienced as opportunities for growth, mastery, and achievement: workload, time pressure, job complexity, and responsibility.
• Hindrance stressors: are demands experienced as obstacles that thwart progress and offer no compensating sense of development: role ambiguity, role conflict, organisational politics, red tape, and job insecurity.

Both types of stressor produce strain, but only challenge stressors are positively associated with motivation, engagement, and performance. Hindrance stressors are negatively associated with all three.

The practical implication is significant: the goal of stress management is not simply to reduce all demands, because doing so would also eliminate the challenge demands that give work its meaning and energy. The goal is to reduce hindrance demands while ensuring that challenge demands are accompanied by adequate resources.

Role Ambiguity and Role Conflict: The Quiet Destroyers

Role ambiguity (uncertainty about what is expected of you) and role conflict (facing incompatible expectations from different sources) are among the most consistently documented sources of workplace stress, and among the most frequently ignored. They lack the dramatic quality of workload or interpersonal conflict, but their cumulative impact is substantial. Meta-analyses have repeatedly shown that both are associated with anxiety, depression, reduced job satisfaction, lower performance, and increased turnover intention.

Role ambiguity is particularly insidious because it undermines the basic psychological need for competence. If you do not know what success looks like, you cannot know whether you are achieving it. In our work with organisations, we frequently encounter people who have never received a clear description of their role’s priorities, whose responsibilities have expanded incrementally without any formal acknowledgement, or who receive contradictory signals from different managers about what matters. The resulting anxiety is not a personal failing; it is a design failure.

Role conflict, meanwhile, is endemic in matrix organisations, project-based work, and any structure where people report to multiple stakeholders with different priorities. Being asked to prioritise time, cost and quality simultaneously, to be both innovative and compliant, or to serve a client’s needs while also meeting internal targets, are all forms of role conflict. The stress arises not from any single demand but from the impossibility of fully satisfying competing ones..

Technostress: The Demands We Didn’t See Coming

The digitalisation of work has introduced a category of stressors that the classic models did not anticipate. Technostress, a term that has been in circulation since the 1980s but has gained urgency in the era of smartphones, constant connectivity, and remote work, refers to the stress experienced as a result of the use and omnipresence of information and communication technologies.

Researchers have identified several dimensions of technostress: techno-overload (technology forcing people to work faster and longer), techno-invasion (the blurring of work and personal life through always-on connectivity), techno-complexity (the cognitive demands of learning and maintaining proficiency with multiple platforms), techno-insecurity (fear of being replaced by technology or by people who are more technologically competent), and techno-uncertainty (the stress of constant changes and upgrades to technological systems).

Presenteeism and Leaveism: Stress Made Invisible

Two related phenomena deserve mention because they represent ways in which stress becomes invisible to the organisations generating it. Presenteeism, attending work while unwell, has been shown to cost organisations more than absenteeism in terms of lost productivity, and it is driven largely by workload pressure, job insecurity, and organisational cultures that stigmatise absence. People come to work sick not because they are heroically committed but because they fear the consequences of not doing so, or because the work will simply pile up until they return.

Leaveism, a more recently coined term, describes the practice of using allocated leave (annual leave, flexi-time) to catch up on work that could not be completed during normal hours, or working while officially on leave. It is, in effect, a way of hiding overwork within the statistics of leave-taking, making workload problems invisible at an organisational level. If everyone uses their holiday to clear their email backlog, the organisation’s leave data looks healthy while its people are anything but.

Individual vs. Systemic Sources: The Attribution Error Organisations Make

There is a persistent and damaging tendency in organisational life to treat stress as an individual problem. When someone is struggling, the instinct is often to ask what is wrong with them rather than what is wrong with their working conditions. This mirrors the fundamental attribution error in social psychology: we attribute others’ behaviour to their character rather than to their situation. And it is enormously convenient for organisations, because if stress is an individual problem, the solution is individual too: resilience training, mindfulness apps, employee assistance programmes, and advice to practise better self-care. None of these things are harmful in themselves, and some are genuinely helpful as part of a broader strategy. But when they constitute the entire strategy, they function as a way of offloading organisational responsibility onto individual employees.

The evidence is clear that the most potent sources of workplace stress are organisational and systemic: excessive workload, lack of control, inadequate support, poor relationships, unclear roles, badly managed change, effort-reward imbalance, and the absence of resources to meet demands. These are not problems that can be solved by teaching people to meditate. They require changes to work design, management practice, organisational culture, and resource allocation.

This is not to say that individual differences do not matter. They clearly do. Some people are more vulnerable to stress than others, and factors such as personality, coping style, and personal circumstances all influence how a given set of working conditions is experienced. But the research consistently shows that organisational-level factors explain more variance in stress outcomes than individual factors, and that interventions targeting the work environment are more effective than those targeting individual coping.

The Business Case: Why This Is Not Just a Moral Argument

For those who require a financial justification for treating people decently, the business case for addressing workplace stress is substantial. Stress-related absence, presenteeism, and turnover are enormously costly. CIPD consistently reports stress as one of the top causes of long-term absence in the UK. Beyond direct costs, chronic workplace stress erodes engagement, creativity, discretionary effort, and organisational commitment, all of which have measurable impacts on performance and productivity. This is a societal level issue.

Organisational-Level Interventions: What Actually Works

If the primary sources of stress are organisational, it follows that the primary interventions should be too. The evidence supports several approaches.

First, job redesign that increases autonomy, reduces unnecessary demands (particularly hindrance demands), and ensures that roles are clearly defined with manageable workloads. This is not about making work easy; it is about making work possible without chronic self-sacrifice. Participative approaches, where employees are involved in identifying problems and co-designing solutions, tend to be more effective than top-down prescriptions, partly because they address the control dimension of stress directly. Job-crafting can be a really valid part of this.

Second, management development that equips line managers to have honest conversations about workload, to recognise early signs of strain, and to create team environments characterised by psychological safety and mutual support. Line managers are, for better or worse, the primary mediators of the employment relationship, and research consistently identifies poor management as one of the most significant sources of workplace stress. Investing in management capability is therefore one of the highest-leverage stress interventions available, though it requires a commitment to genuine development rather than one-day workshops that nobody applies.

Third, organisational culture change that challenges norms of overwork, performative busyness, and the stigmatisation of vulnerability. This is harder to prescribe and slower to achieve, but cultures that normalise sustainable working practices, that reward outcomes rather than hours, and that treat help-seeking as a sign of strength rather than weakness create fundamentally different conditions for their people. Leaders play a disproportionate role here, because culture is, in large part, what leaders model and tolerate.

Fourth, structural changes to workload and resourcing. Sometimes the honest answer to “why are people stressed?” is “because there are not enough of them to do the work.” No amount of redesign or cultural change can compensate for chronic understaffing, unrealistic targets, or resource allocation that treats human capacity as infinitely elastic. This is an uncomfortable truth for organisations under financial pressure, but it remains true regardless.

Learning More

To go deeper, explore these resources on people-shift.com:

• Albrecht’s Four Types of Stress
• Resilience in the World of Work
• The Job Demands-Resources Model
• Wellbeing: What It Is and How to Improve It
• The Relationship Between Performance and Pressure
• Predictive Coding: The Brain as a Prediction Machine

The People Shift View

At PeopleShift, we have spent enough time working with stressed organisations to know that the gap between what the research says and what organisations actually do is wide, persistent, and frustrating. The research says: look at the work, look at the management, look at the culture. Most organisations say: here is a mindfulness app and a fruit bowl.

We do not think this is malicious; we think it is a combination of convenience, inertia, and the entirely human tendency to reach for the easiest solution rather than the most effective one. Individual-level interventions are easier to commission, easier to measure, and easier to point to in a board report. Organisational-level interventions require leaders to examine their own practices, to confront uncomfortable truths about workload and resourcing, and to accept that some of the stress in their organisation is a direct consequence of decisions they have made. That is harder, and so it happens less often.