Dec 12 2018
“[…]In two intervention-based field studies of corporate headquarters transitioning to more open office spaces, we empirically examined—using digital data from advanced wearable devices and from electronic communication servers—the effect of open office architectures on employees’ face-to-face, email and instant messaging (IM) interaction patterns. Contrary to common belief, the volume of face-to-face interaction decreased significantly (approx. 70%) in both cases, with an associated increase in electronic interaction. In short, rather than prompting increasingly vibrant face-to-face collaboration, open architecture appeared to trigger a natural human response to socially withdraw from officemates and interact instead over email and IM.[…]”
Michel Baudin‘s comments: I got curious after reading multiple blanket statements on LinkedIn to the effect that open workspaces decrease office productivity. The authors all refer to the same “Harvard study” without giving any details. Is the Harvard label sufficient to quell any doubts? As the notorious Reinhart-Rogoff paper on austerity shows, it is nothing of the kind.
On closer scrutiny, the Bernstein-Turban’s study is serious but limited in scope. The readers of readers of readers of their paper draw increasingly cosmic conclusions that the study does not support. To locate it, you must thread your way through multiple layers of papers. Each one simplifies and amplifies the results of the previous ones. In the process, they forget any of the nuances and restrictions of the original authors.
A Small, Self-Selected Sample And Limited Conclusions
Bernstein-Turban originally studied the communications used of a few dozen headquarters employees of two multinational corporations, before and after cubicle walls were removed. It’s more than the 5 women on which the theory of the Hawthorne effect rests but still a minuscule sample from one type of office. There is no discussion of what the employees actually do, and conclusions about corporate headquarters may not apply to other office environments.
In addition, the participants were volunteers. The authors know about self-selection bias and handle it by asserting that “a comparison of HR data for participants and nonparticipants provided no evidence of nonresponse bias.” All this says, however, is that the people who declined to participate are similar to the participants in terms of their HR data. We don’t know, for example, their feelings about the transition to an open workspace.
Other authors use this study to bash open workspaces as a concept. In doing so, they are guilty of confirmation bias. They don’t like open workspaces and are happy to use a study with “Harvard” on it for support. Open workspaces may be ineffective in general but this study only shows one special case.
No One-Size-Fits-All For Offices
You can organize offices in many different ways for different tasks. Cubicles are not the only alternative to open workspaces. For tasks that require confidential exchanges of information or quiet concentration for extended periods, you need private offices. On the other hand, you can run a call center in an open workspace. It’s too noisy for engineers but the electronics shield both callers and responders from the noise.