Open-Plan Offices vs. Private Offices: What the Research Says About Noise, Interruption, and Cognitive Output

The open-plan office was sold, and continues to be sold, on a set of premises that sound reasonable enough on paper: that physical proximity encourages collaboration, that shared space fosters a sense of common purpose, that the removal of walls between workers increases the spontaneous exchange of ideas. These premises are not entirely wrong. Some of them are supported by some of the evidence some of the time. What the broader research record shows, however, is that the open-plan office as typically implemented imposes cognitive costs on its occupants that its designers and advocates have consistently underweighted — costs that fall most heavily on the kinds of work that most require sustained, focused attention, and that accumulate over time in ways that are not always immediately visible in productivity metrics but show up clearly in physiological and neurological measures of stress and cognitive fatigue.

The private office, meanwhile, has acquired an unfair reputation as a symbol of hierarchy and hoarded privilege — a design choice whose downsides (isolation, reduced spontaneous communication) are treated as obvious while its cognitive upsides are rarely discussed in the terms the research supports. The result is a debate conducted largely on ideological and aesthetic grounds, with the neuroscience frequently cited selectively in support of positions already held and rarely allowed to complicate them.

What follows is an attempt to describe what that neuroscience actually says — about noise, about interruption, about the brain’s capacity for sustained attention, and about the conditions under which different kinds of cognitive work are most effectively performed.

The Noise Problem: What Ambient Sound Does to the Brain

Noise is the most consistently documented problem with open-plan offices, and its effects on cognitive performance are among the best-established findings in environmental psychology. The research record on noise and cognition spans several decades and multiple methodological approaches, from controlled laboratory experiments to naturalistic workplace studies, and its conclusions are unusually consistent: certain types of noise reliably impair certain types of cognitive work, and the office environments most people now work in produce exactly those types of noise at exactly the levels that the research has established are most damaging.

Irrelevant Speech and the Phonological Loop

The most cognitively disruptive form of office noise is not the loudest. It is speech — specifically, intelligible speech occurring in the background while the listener is trying to perform tasks that involve language processing, reading, or writing. This is called the irrelevant speech effect, first documented systematically by Dylan Jones and colleagues in the 1990s, and it operates through a specific neurological mechanism: the phonological loop.

The phonological loop is a component of working memory — the brain’s temporary holding system for information currently being processed — that handles verbal and acoustic information. It has two parts: a phonological store that holds sound-based information for a second or two, and an articulatory rehearsal process that refreshes that information by subvocally repeating it. When you read a sentence and hold it in mind while you think about it, the phonological loop is doing part of that work. When you compose a sentence while writing, the phonological loop is involved in that process too.

Intelligible speech in the background automatically engages the phonological loop — the brain processes speech as speech, obligatorily and without the listener’s consent, because recognizing and processing language is one of the most deeply automatic of all cognitive functions. This means that a conversation happening at a nearby desk competes directly with the cognitive resources the listener needs for their own verbal processing. The interference is not metaphorical; it is mechanistic. The phonological loop cannot simultaneously serve the background conversation and the foreground task, and the foreground task suffers.

A series of studies by Göran Sörqvist and colleagues, published in the 2010s, established that the irrelevant speech effect is modulated by the listener’s working memory capacity: people with higher working memory capacity are somewhat better able to suppress the interference, but no one is immune to it, and the suppression itself consumes cognitive resources that are then unavailable for the primary task. In other words, even the people best equipped to work in noisy environments pay a cost for doing so — they just pay it in the currency of suppression effort rather than raw performance decrement.

The 65 Decibel Finding

A study published in the Journal of Consumer Research in 2012 by Ravi Mehta, Rui Zhu, and Amar Cheema produced findings that have been widely cited in arguments for open-plan offices: moderate ambient noise at around 70 decibels was found to enhance creative performance compared to low noise (50 dB) or high noise (85 dB). The finding is real and replicated in some subsequent work, and it has a plausible mechanism — moderate arousal facilitates the kind of loosely associative thinking that contributes to creative ideation.

What the finding does not support, and what it has frequently been misused to support, is the general claim that open-plan office noise enhances cognitive performance. The tasks on which moderate noise produced benefits in Mehta et al.’s study were specifically designed to measure creative, associative thinking — unusual uses for objects, remote associates tests. The tasks most common in office work — reading for comprehension, writing, analysis, coding, detailed calculations — are precisely the tasks for which the irrelevant speech literature predicts noise-related impairment. Creative thinking and focused analytical work are not the same cognitive processes, they are not supported by the same neural systems, and the noise conditions that benefit the former measurably impair the latter.

Noise, Cortisol, and the Long-Term Cost

Beyond immediate task performance, chronic exposure to office noise has physiological consequences that accumulate over time. A landmark study by Gary Evans and Dana Johnson, published in the Journal of Applied Psychology in 2000, assigned female clerical workers to either a quiet control condition or a simulated open-office noise environment for three hours. Workers in the noise condition showed significantly elevated urinary epinephrine levels — a marker of physiological stress activation — and made significantly fewer attempts to solve an unsolvable puzzle afterward, a standard measure of motivational persistence. Critically, the workers in the noise condition did not report finding the noise particularly stressful; their subjective experience of stress was not elevated even as their physiology showed clear stress responses.

This dissociation between subjective and physiological stress is important. It means that workers who have adapted to noisy environments — who no longer consciously notice or report the noise as bothersome — may still be experiencing its physiological effects. The cortisol and epinephrine responses to chronic low-level stressors do not habituate as readily as the conscious perception of those stressors does. An employee who has stopped noticing the background noise of an open-plan office may have physiologically habituated to its subjective unpleasantness while still sustaining its hormonal and neurological costs.

Interruption: The Switching Cost the Brain Always Pays

Noise is a diffuse, continuous stressor. Interruption is an acute one, and its cognitive costs operate through different mechanisms — mechanisms that the research has characterized in considerable detail and that paint a picture even less favorable to open-plan office design than the noise literature does.

Task Switching and the Prefrontal Cost

When a worker is interrupted mid-task and switches to responding to the interruption — an overheard conversation that demands a response, a colleague stopping by the desk, a question called across the room — the brain must perform a cognitive operation called task switching: disengaging from the current task’s mental set and engaging the mental set required by the new task. Task switching is not instantaneous. It takes time, it consumes prefrontal resources, and it leaves a residue.

The residue is called the switch cost — a period of reduced performance on the new task immediately after the switch, during which the brain has not fully engaged the new task’s mental set. The switch cost has been consistently measured at between 200 milliseconds and several seconds depending on the task complexity and the degree of mental set conflict between the tasks. But the switch cost after the interruption has ended — the cost of resuming the original task — is where the real damage accumulates. A study by Gloria Mark, Victor Gonzalez, and Justin Harris at the University of California, Irvine, published in 2005, found that after an interruption, workers took an average of 23 minutes to return to the interrupted task at the same level of engagement they had before the interruption. The interruption itself may have lasted seconds. The cognitive recovery time was nearly half an hour.

Interruption Frequency in Open-Plan Offices

Research on actual interruption rates in office environments has produced figures that, set against the 23-minute recovery finding, suggest a working day in which sustained cognitive engagement is structurally almost impossible in open-plan settings. Studies by Mark and colleagues in real workplace environments have found that knowledge workers are interrupted or self-interrupt approximately every three to five minutes in open-plan office settings. If each interruption requires 23 minutes of recovery time to return to full engagement with the original task, and interruptions are occurring every three to five minutes, then the arithmetic of the working day becomes bleak: the brain is essentially never given long enough between interruptions to reach, let alone maintain, the state of deep cognitive engagement in which the most demanding and valuable work is performed.

The term “deep work,” popularized by computer scientist and author Cal Newport, refers to this state — cognitively demanding work performed in conditions of sustained, uninterrupted focus — and Newport has argued that the conditions of modern open-plan offices are structurally hostile to it. The neuroscience supports this framing. The prefrontal cortex, which coordinates the executive functions required for complex analytical and creative work, operates most effectively under conditions of sustained, uninterrupted engagement. The switch costs and recovery times that interruption imposes on prefrontal processing are not trivial inconveniences; they are fundamental constraints on the depth and quality of cognitive output that the environment permits.

The Self-Interruption Problem

A complication in the interruption literature is the finding that a substantial proportion of workplace interruptions are self-generated rather than externally imposed. Workers in open-plan environments do not only interrupt each other; they interrupt themselves — breaking away from demanding tasks to check email, glance at social media, or initiate conversations — at rates that research suggests account for roughly 40 percent of all observed workplace interruptions. This self-interruption rate may itself be partly a product of the open-plan environment: a workspace that offers constant social stimulation and minimal structural support for sustained focus may train workers, over time, toward a style of fragmented, interruption-seeking engagement that persists even when external interruptions are absent.

This behavioral adaptation is neurologically plausible. The brain’s attentional systems, shaped by experience, may calibrate their expectations for interruption to match the environment. Workers who have spent years in high-interruption environments may develop attentional habits that make sustained focus uncomfortable and periodic interruption-seeking rewarding — a learned pattern of neural engagement that is the cognitive fingerprint of the open-plan office, and that does not necessarily disappear when those workers move to quieter settings.

What the Private Office Provides Neurologically

The case for private offices is not merely the absence of the problems described above, though that alone would be substantial. Private and semi-private workspace configurations provide specific positive neurological conditions that go beyond noise reduction and interruption prevention.

Psychological Safety and the Default Mode Network

Working in a space that feels private and secure — where the worker is not being observed, is not performing for an audience, and is not managing the social demands of shared space — activates what neuroscientists call the default mode network (DMN) more freely. The DMN is a set of brain regions — including the medial prefrontal cortex, the posterior cingulate cortex, and the angular gyrus — that is active during rest, mind-wandering, autobiographical reflection, and the kind of loosely associative thinking that underlies creative insight and complex problem formulation. The DMN is suppressed by social observation and the attentional demands of managing a shared environment. Private space reduces those demands and allows the DMN to contribute to cognitive work in ways that constant social exposure prevents.

Personalization, Ownership, and Cognitive Comfort

Research on workspace personalization — the degree to which workers can customize their physical environment — consistently finds that personalization is associated with higher job satisfaction, lower stress, and better cognitive performance. Workers who can arrange their physical environment to suit their preferences, display personal objects, and establish a sense of territorial ownership over their space show measurably lower cortisol levels and better performance on attention and memory tasks than those in depersonalized environments. The private office, as typically configured, offers personalization opportunities that open-plan desks and hot-desking arrangements do not — a point returned to in the discussion of hot-desking elsewhere in this series.

The Collaboration Claim: What the Evidence Actually Shows

The central justification for open-plan offices has always been the claim that they enhance collaboration — that proximity and visibility increase the frequency and quality of interactions between workers, producing better collective outcomes. The research on this claim is more complicated than its advocates typically acknowledge.

The Ethan Bernstein and Stephen Turban Study

The most directly relevant study is a 2018 paper by Ethan Bernstein and Stephen Turban, published in Philosophical Transactions of the Royal Society B, which used electronic monitoring devices to track actual face-to-face interaction in two companies before and after transitions from traditional to open-plan office configurations. The findings directly contradicted the collaboration hypothesis: face-to-face interaction decreased by approximately 70 percent following the transition to open-plan design, while electronic communication (email and instant messaging) increased correspondingly. Employees who found themselves in open, visible, acoustically porous environments responded by reducing spontaneous in-person interaction and retreating to the mediated privacy of screens.

This finding makes psychological sense. Open-plan environments make conversation visible and audible to many people simultaneously, which increases the social stakes of any given interaction and makes workers more reluctant to initiate informal exchanges. The casual conversation that the open plan was supposed to encourage is inhibited, not facilitated, by the awareness that it is being observed by and potentially disruptive to everyone within earshot.

What Actually Predicts Collaborative Quality

The research on what actually improves collaborative outcomes at work points less toward continuous physical proximity and more toward the availability of appropriate shared spaces for deliberate interaction — meeting rooms, collaboration zones, and social areas — combined with the private or semi-private spaces that allow individuals to do the focused preparatory and integrative work that makes collaborative exchanges valuable. Collaboration and focused individual work are not served by the same spatial conditions. The office design that optimizes for one tends to impair the other, and the error of the open-plan movement has been to treat them as if they were.

Hybrid Configurations and the Evidence for Acoustic Zoning

The practical implication of the research is not that every worker needs a private office — a configuration that is expensive and not always possible — but that office design needs to provide genuine acoustic and visual separation between spaces designated for focused individual work and spaces designated for communication and collaboration. This principle, sometimes called acoustic zoning, has been supported by studies showing that even relatively modest physical separations — partial-height screens, enclosed focus rooms, deliberate spatial organization — produce measurable improvements in worker-reported cognitive performance and stress levels compared to fully open configurations.

A 2021 study by Kim and de Dear, building on their earlier 2013 research into the largest dataset of workplace satisfaction ever assembled at the time, confirmed that enclosed private offices consistently produced the highest satisfaction scores on measures of noise, interruption, privacy, and the ability to do focused work — followed by enclosed shared offices, then cubicles with high partitions, then open-plan arrangements. The ranking was consistent across industries, job types, and demographic groups, suggesting that the cognitive preferences it reflects are not idiosyncratic but general.

The open-plan office has been the dominant workplace design paradigm for long enough that its problems have accumulated a substantial evidence base. That evidence base does not argue for a universal return to closed offices — it argues for an honest accounting of what different spatial configurations cost cognitively, and for design decisions made on the basis of that accounting rather than on the basis of aesthetics, cost, or ideology. The brain’s requirements for sustained cognitive work are not flexible in the ways that office design trends have assumed. Noise impairs verbal processing whether or not the management believes in open communication. Interruption fragments attention whether or not the culture values spontaneous collaboration. The prefrontal cortex is not susceptible to being told that the environment is good for it.

Your Workplace and Your Brain: Full Series

• Open-Plan Offices vs. Private Offices: What the Research Says About Noise, Interruption, and Cognitive Output — You are here
• The Neuroscience of Commuting — What Daily Transit Stress Does to the Brain Over Years
• Artificial Lighting, Color Temperature, and Cognitive Performance
• Working From Home vs. Office: How the Brain Adapts to Each Environment
• The Cognitive Cost of Hot-Desking and Not Having a Permanent Workspace
• How Plants and Greenery in the Workspace Measurably Improve Cognitive Performance
• Standing Desks and Cognitive Function — What the Evidence Actually Shows