[fmap]

This whole situation is ridiculous. I am working in a university and I have my own office. Turns out that employees are by far the most expensive part about doing research in computer science, and the cost of giving everyone their own office is easily offset by the (noticable) increase in productivity.

Everyone I know who went on to an industry position is working in an open plan office... and it is not unusual to hear stories about people putting in a few days of home office to get some actual work done.

This is obviously anecdotal, but based on other responses in this thread and elsewhere there is more than enough evidence to justify doing a real study. It is not difficult to measure changes in productivity over time accross a sufficiently large population and the amount of money being spent on engineers means that even a marginal increase in productivity will be worth the cost.