I justified it by saying, "We pay this individual over $100,000 a year, they are key to the following efforts, and the chair will last them for their entire career with the company. Not to mention it costs about 1% what we paid the recruiter who brought this person to us."
I made a similar argument at IBM (which got me into some trouble) which was that revenue per employee was up significantly from the period when everyone had an office, so why not have an office with some offices, some cubes, and some open space and let people work in the space where they are most productive? Total cost per employee was still going to be a lot less relative to their revenue generation than it had been in the past.
Too often people focus too closely on the one time costs and not the overall expense.
Admitting that the workers might know anything better than the management, even about their own preferences, is disruptive to this concept.
My previous supervisor (current coworker) was/is excellent about this. Yeah, of course the actual data will be front and center in the decisionmaking process, but he always makes a point to actually validate that data (through hands-on / direct observation or through his direct reports), knowing full well that there's no such thing as perfect data (there's always a missing metric or an unreliable source somewhere).
More importantly, he's willing to put himself behind his decisions, and always encouraged the rest of us to do the same for our own direct reports (and while I don't and didn't have any, I still took his lessons to heart): if you need to rely on / appeal to external sources of authority (be they higher-ups, data, "corporate policy", etc.) to assert your own authority, then your own authority is illusory, at best. By all means explain your reasoning, of course ("transparency is a dependency of trust" applies to any kind of computational system, including the kind that sits between our ears), but it should be clear that it's your reasoning, not someone (or something) else's.
http://slatestarcodex.com/2018/08/29/bullst-jobs-part-1-of-%...