So, what do you need to solve your company’s problems? Right, money, I know. But how about something else, maybe a new building? How about a whole new site? Sound appealing? Ditch those old labs and offices, move into something that doesn’t look like people have been paid to beat on it with garden trowels for the last 10 years, clear out all the junk and start fresh.

When I think about it, every company I’ve ever worked at has succumbed to some form of this urge. Sometimes it’s been understandable: when your employees are scattered around so widely that everyone has to get in a van when someone calls a meeting, that’s not a great use of everyone’s time. One research site I worked at that had been in use ever since the Eisenhower administration, or maybe Truman’s. Something had to be changed, since there was no longer any room to put anything, or anyone. And as much as I had gotten used to the scuffed linoleum and the shiny washed-out green mental-institution tiles on the walls, I did think the atmosphere was capable of being improved.

Getting to the new land was a struggle, though. We had to clear out the old labs, which meant unearthing stuff that hadn’t seen sunlight — well, fluorescent light, even — for many, many years. The first layers were things that we’d ordered. The second layers were things we recognized. Behind those, well . . . when I pulled out a dropping mercury electrode setup, I knew that things were getting out of hand. (It’s still the only one I’ve ever seen in person). Dumpster after dumpster of antique lab junk was hauled off into the wilds of New Jersey. In retrospect, the big moving project was probably the only way that most of that stuff was ever going to get thrown away.

But the new building, surely that was an improvement, right? Well, on one level, it had to be. There was finally room to turn around in, and room to put everything (especially after the 1960s stuff had been vaporized). And at first, things really were much better, even though we had the usual moving pains and odd annoyances. (Whoever thought of putting locks on every solvent cabinet under every hood, for example had a suspicious view of human nature.) But in time, I noticed that something was missing.

Even though the architects had promised a “high-interaction facility,” there were people I used to see all the time whom I hardly ever ran into any more. The cause, I think, was the move to a multiple-floor building. In the old one, I’d had to walk past most of the chemistry labs to get anywhere. They were all on the same level, connected through several ingrown buildings, and you ran into people whether you wanted to or not. Having everyone scattered across several different floors, though, meant that you had to make a special trip to see some of the groups. Borrowing reagents became the main mode of cross-floor encounter: better than nothing, but still not exactly what the designers had in mind, I’ll bet.

A more recent fad to compel people to interact with each other should be mentioned: the open office, with no real partitions and no permanent desks. I’ve never been sure what the value of hearing everyone else talk on the phone might be, and combining this plan with a vast, impressive, and echo-filled atrium seems ever stranger to me. But that seems to be the fashion. You wonder if the architects would like to work there themselves.

We didn’t have that layout to contend with, anyway. But the other problem I had with the new building took a while to work its way into my consciousness. It was, at least on the first floor, a little too nice. The lobby was clearly built for public consumption, with white statues leading to the impressive wood-paneled lecture hall. The old lab building, by contrast, had no lobby at all — you walked in, and there you were, pretty much in the stairwell, with nothing to do but go to work. The old lecture hall, for its part, made you want to be somewhere else (back in the lab, for example), and if anyone had installed one of those statues, people would have used it as a coat rack. It’s not like I missed the 1950s green tile, exactly, but replacing it with cherrywood paneling was maybe a bit too far in the other direction. It didn’t remind us that we were trying to get somewhere; it made it seem as if we’d already arrived.

And that wasn’t the case, if by “arrived” one meant, “finally able to discover drugs when they needed to be discovered.” The company went through some rough patches in later years because not enough new things were coming through the pipeline. It’s not that I blame the building, because plenty of other companies have gone through the same thing while operating out of everything from lawn sheds on up. But it was incongruous if you knew how things were really going.

This all goes through my mind when I see new research labs going up, because the same problems as before are going to be moved in there with the boxes and crates. Of all the factors affecting drug R&D productivity, I’m not sure that the newness and layout of the buildings even makes the top 10, or the top 20. But these are things that you can do something about, and they are impressive, and they do (at least superficially) make a big change. And since most of the really important problems don’t fall into any of those categories, this response might seem especially tempting.

It might seem especially so at a certain point in an organization’s lifespan. At least a new lab building has a reasonable chance of being functional. Giving people more hood space and room for equipment can’t be, in the end, a complete waste of effort. But what are we to make of it when a company decides that what they really need is a new building for the senior management? You can bet that in those cases, the nice wooden paneling will not be confined to the lecture hall; you’ll be hard-pressed to keep it out of the toilet stalls. (Maybe a white statue dispensing paper towels?)

C. Northcote Parkinson, the man who coined Parkinson’s Law (“Work expands to fill the time available”), had something to say about that. Using the example of the building up of New Delhi as the capital of British India, he pointed out that too often, the urge to construct a huge new headquarters strikes after things have peaked. Examples are numerous. By the time St. Peter’s was finished, the papacy had lost a good part of its power, and in the 20th century, the League of Nations completed its Palace of Nations just in time for the whole organization to shrivel up and disappear. “Perfection of planning is a symptom of decay,” he wrote. “During a period of exciting discovery or progress there is no time to plan the perfect headquarters. The time for that comes later, when all the important work has been done.”

So if your company is moving into new research space, well, enjoy, and try to get as much out of the experience as you can. But if it’s putting up a palace, watch out. Things went downhill for the Old Kingdom of Egypt after the Great Pyramid went up, so you’d better make sure that you’re not building another one.