It's the same phenomenon that gets people to think that working long hours, or working in "open plan" offices for that matter, is super productive.
You watch TV for a bunch of hours and there's a constant hum of energy which is mostly just regurgitating the same information over and over again in different ways but there is a little kernel of new information coming out in dribs and drabs. And at the end of X hours it feels like those hours were spent productively because you have been exposed to new information along the way. Typically though you'd be better off just reading a summary later.
In an office environment when you work, say, 12 hours a day you'll find that you do productive work all throughout that period, even in the parts after "normal" hours. And that suckers you into thinking that the whole stretch was valuable, even though typically there are lots of periods of downtime and reduced productivity in those 12 hours and usually you'd be better off just working 8 hours or less and having better time management. Similarly, when you happen to overhear co-workers talking about something important and you go join the discussion and it turns out to be very fruitful you have that same post hoc rationalization that how things happened is the best and/or only possible way for productive work to happen.
I've been against open plan office environments since long before it became cool. There are some things that I just don't need to see and hear. It could be anything from a coworker blowing their nose, a conversation, or body language. It's distracting. If a company won't give a programmer a proper office with a door, it tells you a lot about the company, and what it tells me isn't good for productivity.
In that case it would appear to be unavoidable. However, when comparing the costs between offices I think it is important to take lost productivity in to account.
When looking at $cost_of_office_with_private_space versus $cost_of_open_plan_office it might make sense to go for the open plan office.
But $cost_of_office_with_private_space versus ($cost_of_open_plan_office + $productivity_cost) might lead to a different conclusion.
So how do you measure $productivity_cost accurately? Do you account for cases like where employee A, B and C come in the shared office with a new kind of idea going to rocket your business revenues to sky level?
I've got no problem with "ramen profitable" startups doing it out of necessity— after all, there's a lot of non-optimal things new companies do. What I object to are companies that are flush with cash, profitable— maybe they've even IPO'd— and can afford to foster a productive work environment, but believe open floor plans are the way to go because Facebook does it.
You watch TV for a bunch of hours and there's a constant hum of energy which is mostly just regurgitating the same information over and over again in different ways but there is a little kernel of new information coming out in dribs and drabs. And at the end of X hours it feels like those hours were spent productively because you have been exposed to new information along the way. Typically though you'd be better off just reading a summary later.
In an office environment when you work, say, 12 hours a day you'll find that you do productive work all throughout that period, even in the parts after "normal" hours. And that suckers you into thinking that the whole stretch was valuable, even though typically there are lots of periods of downtime and reduced productivity in those 12 hours and usually you'd be better off just working 8 hours or less and having better time management. Similarly, when you happen to overhear co-workers talking about something important and you go join the discussion and it turns out to be very fruitful you have that same post hoc rationalization that how things happened is the best and/or only possible way for productive work to happen.