The Office That Never Made Sense

The open office plan arrived in the 1990s with promises: collaboration would increase, hierarchies would flatten, creativity would flourish in the free flow of human contact. Two decades later, the evidence assembled to test these promises did not support them. Noise was up, focused work was down, unplanned collaboration had actually decreased, and sick days had increased. The design that was supposed to optimize human performance had optimized against it. This was not a surprise to anyone who understood how human social architecture actually works. The open office was not designed around human social needs. It was designed around real estate costs, flattened into a mandate by an ideology of connection that confused proximity with community and visibility with trust.

What the Human Social System Actually Runs On

Human social groups evolved in conditions of controlled visibility. The campfire at night illuminated the inner circle and left the surrounding dark opaque. The village arranged dwellings with semi-private spaces between public gathering areas and private sleeping quarters. Social contact was modulated — you moved through gradients of publicness and privacy depending on the nature of what you were doing and with whom. This modulation matters for a specific reason: the human nervous system does not behave the same way in conditions of constant observation that it does in conditions of partial privacy. Under continuous visibility, physiological stress markers increase, self-monitoring increases, and the kind of relaxed, exploratory cognition that produces creative work decreases. You cannot brainstorm authentically while performing. You cannot take intellectual risks while managing an audience. Research from the University of Exeter's environmental psychology group studying worker experience in open-plan offices found that the loss of privacy and the inability to control interruption were the strongest predictors of dissatisfaction and reduced productivity, outweighing factors like lighting, temperature, and desk ergonomics. Workers were not struggling with the physical environment. They were struggling with the social one.

The Dunbar Problem Applied

There is a number problem embedded in open office design. Human social groups function well at the scales for which the social nervous system was calibrated: the small team of five to seven people who work closely together, nested inside a larger organizational unit of perhaps 15 to 20 who share goals and context. Beyond these thresholds, the social complexity of tracking everyone's needs, status, and communication simultaneously exceeds what individuals can manage without significant cognitive cost. Open offices routinely place 50 to 300 people in a single shared space. At that scale, you are not in a community. You are in a crowd. The social processing demands of a crowd — monitoring for threat, managing impression, tracking multiple simultaneous interactions — are exactly the wrong conditions for concentrated work. The irony is that the plan that was sold on collaboration increased the social noise to the point where the deep work that collaboration is supposed to enable became structurally unavailable for most of the day.

The Tangent: Why the Myth Persisted

Part of the reason open offices persisted despite the evidence is that the evidence was inconvenient for multiple stakeholders simultaneously. Real estate costs were real and measurable. The costs in cognitive performance and employee wellbeing were diffuse and harder to attribute directly. Executives who could afford private offices retained them and did not experience the conditions they had mandated for others. And the ideology of transparency and flatness provided a narrative that made the plan sound progressive rather than cost-driven. This pattern — where the costs of a design decision fall on those with the least power to reverse it, while the benefits accrue to those who made the decision — is worth recognizing as a feature, not an accident.

What Actually Works

Harvard Business School researchers Ethan Bernstein and Stephen Turban, studying the behavioral consequences of transitioning organizations from traditional to open floor plans, found using electronic tracking data that face-to-face interaction decreased by approximately 70 percent following the transition, while digital communication increased. The open plan had not produced collaboration. It had produced overstimulation and retreat into headphones and messaging apps — a worse version of the cubicle design it replaced, but noisier. What produces genuine collaboration is not shared space. It is shared purpose, genuine relationship, and enough privacy that people can think before they speak. These can be facilitated by thoughtful design — spaces that mix enclosed rooms for focused work with informal gathering areas, that give teams identifiable territory without isolating them entirely, that allow people to choose their level of visibility depending on what they are doing. The principles are not complicated. They are just in tension with the economics that drove the design that did not work. And until those economics change, the design tends to persist.