Chad went an mentioned a buzzword that sets my teeth on edge: deferred maintenance.

It’s not just the economics of academia that suck in this regard. I’ve seen my fair share of it, too. As Chad notes, maintenance is not sexy, and it’s hard to get people not directly involved fired up about it.

Things break, even when Dino and Luigi Vercotti aren’t shaking you down. They break even when you can avoid the overzealous attitude of “If it’s stuck, force it. If it breaks it needed replacing anyway.” And there various approaches you can take to this fact of life —

1. Fix/replace things when they break
2. Maintain the equipment in good working order, so it breaks down less often
3. Try to anticipate when something will break, and replace it just before that happens

If we’re talking about things we really need, the first option is pretty much mandatory. Sure, when the singing fish in the rec room breaks down you can let it slide, but not so much when your lab equipment goes. But the other options entail some prioritizing and risk management. If you can afford to let some apparatus run until it dies, and the downtime while you replace it doesn’t affect you, then your choice is pretty clear, but very often in business and research you can’t let that happen. Broken equipment means some people can’t do their jobs, and other people are probably working overtime getting things working again. And yet, all too often the first response to budget cuts is to cut the maintenance budget.

At first glance, from a beancounter’s bottom-line approach, this looks good, which it’s why it’s so tempting. Money not spent is money saved, at least in the short-term, myopic acasual view, and new equipment does tend to last a while even without doing much to maintain it. So you seem to have saved money. But eventually, the equipment dies before it would have, had it been properly looked after, and now you’re stuck. You hadn’t expected the equipment to die, bureaucratically speaking— there’s no money in the budget for a replacement. You don’t already have a spare widget, because that’s an expenditure, and the goal was to cut the budget.

In business situations, where you deliver a product of some sort, you can at least quantify downtime in terms of lost productivity, and make a business case for keeping things running. But what if that metric isn’t there? There are a lot of situations where lost time isn’t measured (people “steal” budget this way — eliminate a position by shifting work to another department. They look like they’ve saved money, and the extra time burden isn’t accounted for because it’s other departments that are taking up the slack)

Here’s where the really insidious part of the process comes into play. From what I’ve seen (in different bureaucracies), even though run-of-the-mill budget expenses are squeezed, if you can turn it into a crisis you can still get money. Without this widget, your operations screech to a halt, and that’s a calamity, dammit, so you take your case to a higher level, and they search for money to patch the figurative or literal gaping hole that’s appeared. Money is siphoned off from other sources (perhaps, ironically, from other programs cutting maintenance budget), and that saves the day. Until the next emergency.

And nobody has learned anything useful. Systems that run this way have this very destructive feedback loop. The catastrophes get fixed, after a fashion, and few stop to think about how that situation — and the large amount of money spent — could have been averted with less money ultimately spent. And the really frustrating thing is that the people involved in these decisions are smart enough to know that not changing the oil in their car isn’t a viable money-saving tactic, and yet can’t seem to transfer that mindset to the business case.