I got into a short-lived debate on Twitter on the reasons firms like the one I work for have open office layouts. I know that they used to be sold on their merits, and then people grew a bit soured on them. Meanwhile, offices with a door have their proponents too. Me, I tend to believe that they save greatly on the cost of real estate, and that’s why tech companies use them. My opponent pointed out that Facebook, the recent owner of the world’s biggest open office floor plan, is hardly pressed for savings, so surely at least some of the people who push them must be thinking that they have some other merit than their cost advantages.

Economists tend to think of stuff as goods and bads: goods are stuff you will pay for; bads are stuff you will pay to get rid of. Beer is a good (work with me); garbage or illness are bads. Sometimes goods and bads come as a bundle: a banana is a good; a banana peel is a bad. You can’t buy a banana without the peel, so the cost of getting rid of the peel figures in your calculation of whether to buy a banana. This cost is usually small enough that you don’t think about it, but it does influence your decision all the same. If you think that it doesn’t, imagine a world in which the smallest amount of bananas available for sale were 1 ton, at $1 per ton. In such a world, about 1999 pounds of those bananas would quickly turn into a bad you’d need to get rid of somehow. Is $1 still a bargain?

The office layout, whether open or closed, is like a banana: the good comes bundled with the bad. Open is good for collaboration, proximity to the problems, efficient communication; it’s bad for quiet, concentration and flow. Closed is good for flow, quiet and concentration; it’s not so great for communication, or seamless collaboration. You need both kinds of goods and you need to avoid both kinds of bads but they cannot be separated cleanly. What makes the office layout different from a banana is that the good and the bad do not come in fixed proportions, as in one peel for each banana. Instead, you can pick the mix that works best for you. Closed is more expensive than open, so we buy it when flow and concentration yield productivity gains that justify the difference. It is possible to buy too much of it though: when people can hide in the office to read Ars Technica, or when the door gets in the way of quick feedback, productivity gains fall, and closed may no longer pay for itself. Open has the opposite problem: if the chatter and noise get in the way of flow, productivity suffers to the point that the savings from open are erased. So what’s typical? Mostly open, or mostly closed?

This reminds me of the argument about cities: are big, anonymous cities good or bad for your health? If you’re part of an oppressed minority, they can be oases of freedom; when the risk that the village elders might decide to have you stoned is great enough, more freedom is definitely good for your health. When you can trade, you don’t need the protection of the village elders, or the safety of a large clan; if either demands in exchange that you keep your head down and forego whatever pleasures this arrangement forbids, the strictly business ethos of the city is a better deal to you than the communal sharing of village life. But anonymity breeds loneliness, and that comes with all kinds of bad health effects of its own. What’s the net effect? Your own mileage will vary, but if you want to know what works best for most people, you need only look at which is growing. The world’s population is voting with its feet – for the big cities.

Cities do not draw the bulk of their growth from people fleeing brutal oppression; it would be a horrible world if they did. Most migrants’ reasons are more mundane, but they do add up. Caring and sharing in the village is an expensive way to live. After a day of toil doing subsistence farming, that’s all you have: subsistence. You get to sleep and do it again tomorrow. Cities, where trade partners are many and close together, offer better value for their inhabitants’ time, however small the gains for most people who choose them.

It seems to me that people judge the office layout by the same measure – what works for the most, for the most mundane of reasons – and open wins. Flow and concentration are great, but most people don’t use them well enough to justify the extra rent.