What Overstimulation Feels Like in an Open Office

Two people spend nine hours in the same room. Same lighting, same four conversations inside earshot, same dozen interruptions nobody logged. One of them drives home mildly tired. The other cannot decide what to eat, says something sharp to a person who did nothing, and then sits on the edge of the bed for forty minutes doing nothing at all.

The usual explanation for that gap is sensitivity. Sensitivity is real and it is measured and it is part of the answer. It is also the wrong place to start, because it does not move and it does not explain why the same person was fine in the same room six weeks ago.

What explains it is an account with three lines in it. How much arrived. How fast the ceiling came. How much was put back before the next morning. Open-plan offices deserve a whole article here for one reason: most things damage one side of that account, and this one works on both at once.

Overload is a ledger, not a temperament

Three quantities decide whether a given fortnight is survivable, and a single overall score hides which one is doing the work.

The third column is why the two people in that room had different evenings. A loud, interrupted, badly lit day costs very little when it is followed by an hour that is properly empty, and costs a great deal when it is not. Fifteen questions about the last two weeks read the three separately for exactly that reason, because the answer changes what would help.

The column everyone reads wrong

Here is the part worth slowing down for, and it runs against how people rank their own problems.

A high input score is an environment problem. A high recovery score, meaning a high number in the column scored for what is missing, is a structural problem. The second is the more serious of the two.

The reason is about time rather than severity. Input is almost always temporary. A project closes, a quarter ends, a house move finishes, a baby starts sleeping. A recovery window behaves nothing like that. Nobody reinstates the silent commute after the move to a flat ten minutes from the office. Nobody gives back the hour before the house wakes up once something else has been scheduled into it. It goes quietly, it is never the thing anybody apologises for, and afterwards every ordinary week costs more than it used to and nobody can say why.

So a tall input bar means a loud stretch, and loud stretches end. A tall recovery bar sitting next to an unremarkable input bar means the thing to investigate is not the volume of the schedule. It is what used to put something back, and when exactly it stopped.

What an open floor does to both columns at once

On the input side the mechanism is not really loudness. Jahncke and colleagues ran a simulated open-plan office at two noise levels, 39 and 51 decibels, and the twelve-decibel difference was enough to reduce performance on memory tasks and to leave participants more tired and less motivated. Fifty-one decibels is not loud. It is a normal office.

What does the damage is the content. Szalma and Hancock pooled 797 effect sizes from 242 studies of noise and performance, and the pattern that came out was not about intensity. Speech produced a substantially larger decrement than non-speech noise, and intermittent noise produced a larger one than continuous noise. An open office is the worst possible combination of those two findings: speech, arriving at unpredictable intervals, at a volume nobody could reasonably complain about.

The recovery side is less often named and it is the half that matters. An open floor does not merely fail to provide a quiet stretch. It makes withdrawing into one a visible act, performed in front of people whose opinions have consequences, so the cheapest way through the day is to keep absorbing. Kim and de Dear went through 42,764 occupant survey responses from 303 office buildings and found the gains from easier interaction smaller than the combined penalties from noise, sound privacy and visual privacy. In their decomposition the pros came to 0.21 and the cons to 1.07, a net loss of 0.86 satisfaction units for an occupant reporting both.

Bernstein and Turban then looked at what people do rather than what they report. Two corporate headquarters removed walls from a floor, and the researchers tracked face-to-face interaction with wearable sensors and electronic messaging on the company servers. Face-to-face interaction fell by about seventy per cent in both cases. Email sent rose by 56 per cent, instant messages by 67 per cent, words sent over instant message by 75 per cent. Removing the walls did not produce more collaboration. It produced a floor of people who had withdrawn from the room and were talking to each other through screens instead, which is a room where the input never stops and the conversation has moved to a channel that interrupts more cheaply.

Where the evidence actually stops

Writing on this subject is usually confident in one direction, and the honest position is narrower than that.

The broader picture is consistent in direction and weak in design. James and colleagues reviewed 31 studies and found open-plan layouts negatively associated with health, satisfaction and productivity, with no study measuring productivity finding a benefit. Gerlitz and Hülsbeck reviewed 46 articles across several disciplines and reached the same conclusion about single offices outperforming open ones, while noting that journal quality varied considerably across the set. Most of the underlying work is cross-sectional and self-reported, which means it can establish that people in open offices report worse things and cannot establish that the office is what caused them.

Pejtersen and colleagues surveyed 2,403 Danish office workers and found sickness absence rising with the number of people in the room, adjusted for age, sex, socioeconomic status, body mass index, smoking, alcohol and leisure exercise.

The gradient is suggestive and the design is a one-off questionnaire, so it cannot rule out the possibility that the kinds of jobs done in large shared rooms differ from the kinds done behind doors. It is evidence. It is not proof, and presenting it as proof would be the same move the furniture industry made in the other direction.

So the defensible summary is narrow. Open-plan floors reliably raise the amount of intelligible, unpredictable speech reaching a person, reliably reduce acoustic and visual privacy, and reliably shift conversation from the room to the screen. Whether they make a given organisation less productive has a messier answer than either side of the argument admits.

The cost of being interrupted is not the interruption

Most people price an interruption at the length of the interruption. Two minutes for the question, ten for the conversation about the weekend. That accounting is wrong, and the error is in what happens afterwards.

Mark, Gudith and Klocke ran a controlled experiment on interrupted office work and found something that sounds like good news. Interrupted participants finished in no more time than uninterrupted ones, compensating by working faster. The compensation showed up on every other measure: higher workload, more stress, more frustration, more time pressure, more effort. The work got done and the person paid for it.

Mark and colleagues' field observation explains why. Shadowing information workers minute by minute, they found that when an interrupted task was picked up again on the same day, the average wait was over twenty-five minutes, and in between the person had worked inside an average of 2.26 other unrelated activities. The interrupted thing is not resumed. It is queued behind two other things, each of which has to be loaded and unloaded in turn.

The same study gives the cleanest available statement of what sitting with other people does. Among people working in physical proximity to colleagues, 62.9 per cent of work stretches were interrupted before they finished. Among those working apart, 49.3 per cent were. The mechanism is overhearing: a colleague who can hear that the moment is convenient will use it.

What the ceiling looks like from inside

The experience of reaching it is specific enough that people recognise the description before they accept the explanation.

There is irritation with no object attached, arriving before any decision has been made about it. A chair scraping. A question asked twice. Someone chewing. There is the urge to leave the room in the middle of a sentence, which usually gets suppressed and costs something to suppress. There is sharpness directed at whoever is nearest when the ceiling arrives, nearly always someone close, followed by guilt worse than the sharpness was.

Then there is the version that gets misread more than any other.

Sitting completely still, unable to start anything, not enjoying it, not sad about it, just empty.

That is shutdown, and it gets called laziness by the person doing it and depression by the people watching. It can be either. More often it is what a saturated system does when it stops accepting input, and it is the most reliable sign that the ceiling was passed some hours earlier. The tell is what it responds to. An empty evening that follows a loud week and lifts after a genuinely quiet hour is a load problem. A flatness that is there on good days too, that has lasted weeks, that takes the pleasure out of things that used to work, belongs with a doctor rather than with a questionnaire about noise.

In the first twenty minutes, subtract

Most advice at this point reaches for breathing exercises. They are aimed at arousal, and arousal is downstream. At the ceiling the useful move is subtraction: leave the room, kill the overhead light, put something in the ears, and do not speak to anybody for twenty minutes.

Which is where headphones need placing accurately, because they are the most recommended intervention on this subject and they operate on one column only. They lower input, often a lot, and that is a real gain. They do nothing whatsoever for recovery. A day spent inside them is still a day of tracking, deciding, being interrupted and being watched. The failure mode is the reason the distinction is worth labouring: a tool that only reduces input lets a person last longer inside a deficit, so the collapse arrives later, harder, and looking like it came from nowhere.

When the room is not yours to change

Here is the sentence most writing on sensory overload avoids. Open-plan offices, small children, city commuting and shared housing are the common sources, and not one of them yields to a technique. Nearly every article on this subject quietly assumes a reader who can control their own environment, and that reader is in the minority.

If the score is coming mostly from input and the environment is fixed, two things still move.

The first is predictability, which is worth more than quantity. A fixed, defended, unremarkable thirty minutes at a known time beats an unpredictable ninety. The reason is not scheduling discipline. Peters, McEwen and Friston set out the case that uncertainty is itself what the brain spends energy on, and bracing for input that could arrive at any second is a continuous expenditure that happens whether or not the input arrives. Knowing when the quiet is coming lowers the load separately from how much quiet there is, because preparation is a cost in its own right. The practical form is a slot, named, at a time, in a calendar other people can see.

The second is staggering. Put the work that requires real concentration somewhere other than the hours with the most arriving, rather than attempting both in one window and losing on both counts.

Neither changes the environment. Both change what it costs, which is the only lever on offer when the walls are not negotiable.

Sensitivity sets the slope, this sets the position

One distinction prevents most of the confusion in this area.

Sensory-processing sensitivity is a trait: a consistent, cross-situational, largely lifelong level of reactivity, with its own measurement literature and its own arguments, covered in the entry on the highly sensitive person. Overload is a state: the relationship, over one particular stretch of weeks, between what is arriving and what is being returned.

The trait sets the slope. The state sets the current position on it. A highly sensitive person with a real recovery window can go months without being overloaded at all. A person who is not especially sensitive, working on an open floor and going home to two small children, will be overloaded regardless of temperament. A high score on one is not an answer to the other.

When this is not overload at all

One pattern is worth separating from everything above, because a questionnaire cannot do it.

Overload that has appeared suddenly in someone who never had it, particularly alongside headache, ringing in the ears or new sensitivity to light, may not be overload. Sensitivity to light and sound is a documented part of the premonitory phase of migraine, arriving hours to days before any headache does. Noise sensitivity is common after concussion and resolves slowly: in a longitudinal study of 341 adults after mild traumatic brain injury, 45 per cent reported it in the first week and 28 per cent still reported it a year later. The aftermath of an infection and several medications produce the same picture. None of that is distinguishable from a loud fortnight by any set of questions, which makes a new and sudden onset a reason to see a doctor rather than to try harder at managing the office.

For everyone else the arithmetic is unglamorous and it holds. The room is one line in the account. The hour afterwards is the line that decides what the room costs.

Sources

• Bernstein, E. S., Turban, S. (2018). The impact of the 'open' workspace on human collaboration. Philosophical Transactions of the Royal Society B, 373(1753), 20170239.
• Kim, J., de Dear, R. (2013). Workspace satisfaction: the privacy-communication trade-off in open-plan offices. Journal of Environmental Psychology, 36, 18-26.
• Jahncke, H., Hygge, S., Halin, N., Green, A. M., Dimberg, K. (2011). Open-plan office noise: cognitive performance and restoration. Journal of Environmental Psychology, 31(4), 373-382.
• Szalma, J. L., Hancock, P. A. (2011). Noise effects on human performance: a meta-analytic synthesis. Psychological Bulletin, 137(4), 682-707.
• Pejtersen, J. H., Feveile, H., Christensen, K. B., Burr, H. (2011). Sickness absence associated with shared and open-plan offices: a national cross sectional questionnaire survey. Scandinavian Journal of Work, Environment & Health, 37(5), 376-382.
• James, O., Delfabbro, P., King, D. L. (2021). A comparison of psychological and work outcomes in open-plan and cellular office designs: a systematic review. SAGE Open, 11(1).
• Gerlitz, A., Hülsbeck, M. (2023). The productivity tax of new office concepts: a comparative review of open-plan offices, activity-based working, and single-office concepts. Management Review Quarterly, 74(2), 745-775.
• Mark, G., Gudith, D., Klocke, U. (2008). The cost of interrupted work: more speed and stress. Proceedings of the 2008 CHI Conference on Human Factors in Computing Systems, 107-110.
• Mark, G., González, V. M., Harris, J. (2005). No task left behind? Examining the nature of fragmented work. Proceedings of CHI 2005, 321-330.
• Peters, A., McEwen, B. S., Friston, K. (2017). Uncertainty and stress: why it causes diseases and how it is mastered by the brain. Progress in Neurobiology, 156, 164-188.
• Karsan, N., Goadsby, P. J. (2018). Biological insights from the premonitory symptoms of migraine. Nature Reviews Neurology, 14(12), 699-710.
• Shepherd, D., Heinonen-Guzejev, M., Heikkilä, K., Landon, J., Theadom, A. (2021). Sensitivity to noise following a mild traumatic brain injury: a longitudinal study. Journal of Head Trauma Rehabilitation, 36(5), E289-E301.