Skip to main content

Architecture and Productivity: Four Theses

I certainly enjoyed this review of a new book, Laboratory Lifestyles, in Nature. This is a look at research buildings and the behavior of scientists in them, and will for many reopen a lot of arguments. Open offices? Open-plan lab space? Where do the break areas go, and the conference rooms? What do they look like? And more broadly, are some of these designs just better than others, both from a practical standpoint (cost, construction, durability) and from a theoretical one (making people more productive and creative)? And if you have opinions on those last topics, and we all do, do you actually have any evidence other than your personal preferences?

I have my own take on those issues, of course. But rather than reiterate those, I’ll put forward some other beliefs of mine, about which I invite comment. Here we go:

  1. Architecture does matter, up to a point. But I think that it’s much more likely that bad design and physical surroundings can be a noticeable influence than good ones can be. Once past certain thresholds, I think that most layouts and designs may be within error bars of each other.
  2. This means that the brave talk of architects about how their research buildings will increase creativity are very likely nonsense. In many cases, said architects have no idea of what the day-to-day work is like, or will be like, in the very buildings they are designing. At the very least, there is little or no empirical evidence for their assertions, and given the number of variables involved, nor is there likely to ever be any.
  3. These variables – chief among them the mixture of researchers, their own personalities, and the problems that they’re working on – are so influential that they will utterly swamp any architectural effects. That is, with lab equipment and budgets being equal, a good mixture of people in a nondescript older building will outperform a bad mixture housed in the latest version of splendor. What you’d want is to take the same mixtures and see how they would have done in both sets of surroundings, but that’s an impossible experiment (see above).
  4. Furthermore, the variance in those researchers and in their personalities (not to mention their actual jobs) is also so high that there is likely no such thing as a single building design which will make them all happy or more productive. When you read these articles about R&D design, they invariably treat the occupants of the building as if they all worked in the same style, and nothing could be further from the truth.

So put me down as a skeptic. Counterarguments are welcome in the comments!

48 comments on “Architecture and Productivity: Four Theses”

  1. RandomWok says:

    Here’s a real-world example to support your observations above: the GSK Stevenage facility. But one must also concede that beyond architectural considerations, any pharma R&D building must have one essential ingredient: Lady Luck smiling on it. Perhaps we need some kind of Feng Shui knowledge for our labs.

    1. Diver Dude says:

      I worked in the GSK Stevenage facility. It wasn’t bad but it certainly wasn’t a game changer.

      My experience has always been that communities of scientists evolve into, and are molded by, their surroundings. The best facilities I have worked in were a bunch of prefabricated WW2 era huts which were freezing in winter and boiling in summer and which immediately fostered a shared feeling of suffering 🙂

      It was the community spirit that was crucial, not the physical architecture.

  2. Mark says:

    Number one issue for me is safety. I want a space that is well ventilated. I could never tell what carcinogens, solvents, mercury vapor, or what-have-you was lurking in the floor boards, ceilings or at the back of fume cupboards or lab shelves. Dust is everywhere and who knows what is adsorbed to the dust and dust balls.

    1. Ken says:

      That was my first thought too. “Open-plan lab space” doesn’t go well with “Things I Won’t Work With”.

      1. Ben says:

        Open plan lab space sounds like a consequence of several “things I won’t work with”.

        Mostly the ones with lots of nitrogen.

    2. Philip says:

      Just imagine having your OO desk space in the lab. I have a rotovap and silica plate cutting station three feet from my desk. Yet EHS says its ok because it’s a separate office space *rolls eyes*

  3. anon3 says:

    I particularly like state of the art research buildings with a lack of bathrooms.

    1. KT says:

      I think this is driven by ADA compliance. New or renovated buildings have one huge wheelchair-accessible bathroom stall where there would have been three regular-sized ones in an older building, or a huge single-user bathroom where there would have been a stall and a urinal before.

  4. Isidore says:

    I find the open space concept for office space quite disagreeable. It is supposed to promote positive interactions and collaboration and not allow people to hide in their cubicles or, behind closed doors, in their offices. It is loud, with multiple conversations going on at once, disruptive with many people coming and going at all times, and distracting to anyone who needs some quiet to review data, compose a memo or manuscript, or read a literature article. It does have the effect (positive, I suppose, from a company management perspective) of conveying to employees that they are all expendable commodities housed in a warehouse-like office area or a stable, where nobody pays much attention if one chair becomes vacant or is occupied by someone new.

    1. Been there and still there says:

      And I suspect much cheaper to build… and sell or rent. It’s “flexible”!

    2. DrOcto says:

      I often think the main purpose of open-plan offices is to reduce non-productive internet surfing. Much less likely when anyone can just walk past your computer at any time.
      If you’re in a one or two man office, you can increase your own productivity simply by having your back to the door. Or maybe I am just projecting and you’re all perfect little angels.

      1. bluefoot says:

        The funny thing is that I do a lot *less* non-productive internet surfing when not in open-plan because I can actually focus. With all the noise, etc in the open plan office space, my thought process is often, “I am not getting anything worthwhile done anyway, so I’ll check the news/In the Pipeline/the stock market, etc.” When I had an actual office (or in my current situation can grab a “huddle room”), I could get a lot more done in a day. Plus when I had an office, I didn’t have to wander around for 10 minutes ahead of every call or 2-3 person meeting to find a space to have the call/meeting.

      2. O'Brien says:

        Why have big brother watch when it can be outsourced to your fellow workers?

  5. Ted says:

    I love the fact that almost half of the people (kids…) I see in the lab are wearing bluetooth headphones, while another 40% have a more discreet set of earbuds. It’s the last 10% that were apparently nominated to the combination social/safety committee…


  6. loupgarous says:

    Eli Lilly’s Endocrine Group’s office space was a building formerly used for packaging and warehousing birth control pills (appropriately, I guess). It had the advantage (to the architects) of no partition walls, just a huge open area which could be broken up into cubicles and conference rooms, with the steady aggregate murmurs of a hundred analysts, statisticians, and management folk.

    The team within Endocrine Group to which I was assigned (insulin analogs) moved not long after I joined it to a smaller space but essentially also a large open space, thankfully with just a couple dozen of us murmuring. Still cubicle-bound, with our very own conference room. The move made no real difference, and cubicle structure just made it easier for some of us to tattle to our team leader about how the newspaper cartoons we stuck on our cubicle walls were disturbing them. The conversation from the Lilly company hire about my cartoons to our team leader about this was audible above the murmur.

    If architects really want to help productivity, they should bring back fully enclosed offices for everyone.

  7. Jonathan Starr says:

    “It does have the effect … of conveying to employees that they are all expendable commodities housed in a warehouse-like office area or a stable, where nobody pays much attention if one chair becomes vacant or is occupied by someone new.”

    Yes, yes it does.

    I’ve always thought the open plan vs other debate was just an expression of extrovert personalities. The type of people that do stuff like conning investors out of large-enough piles-of-cash to buy buildings unfortunately don’t seem to understand that the numerate class who will be stuck in said corporate hell are more often introverts that feel exposed when in the open and don’t want to look at other “people” let alone talk to them.

    who knows…

    1. Hap says:

      Either it’s an expression of dominance (“We’re winners and you’re not.”) or of irrelevance (that the people that work there don’t matter). I’m not sure whether the people that work in open areas don’t matter to management or are held in contempt. Either way, it doesn’t help.

      1. fajensen says:

        …. the people that work in open areas don’t matter to management or are held in contempt. ….
        We can have Both, you know: The weak asks “Why?”, the strong asks “Why Not?”

        There is a Copenhagen-based architectural company – who are not actually, Architects because to be that one has to be a member of “Arkitektforeningen (Danish Association of Architects)” and they have requirements and stuff … basically they are bunch of consultants.

        The service provided by this company is the design of Collaborating Working Environments – open plan offices, with task-specific work stations, “collaboration areas” a.k.a. where the coffee machine is, and the Killer Feature, the Key Sales Point, is that they will always provision this “Collaborative Innovation Space” for 20% fewer employees that there is.

        And soon this will be because the smart people doesn’t care for the kind of “Collaboration” there is in a monkey-cage, where one has to either get in early, stay late, or queue for a “workstation” while watching the dum-dum’s happily talking all day while waiting to work.

        When that lot shows up, it is hight time to polish that CV and resurrect the LinkedIn profile!

        Management does not care a lot about *how* the numbers are made as long as they move the needle on their KPI’s.

    2. eub says:

      I actually am something of an extrovert in my professional personality — chatty, at least: I’m most productive if I can say “hey, look at this thing” many times a day and get the same from the people with me — but open plan doesn’t work for me. You can’t talk in an open-plan office, that’s the strong social convention, because too many other people don’t want the distraction.

      Put the chatty people in offices with half a dozen other chatty people. Put quiet people in offices with other quiet people. This is more important that what specific topic they’re working on. There, do I win a prize for increasing global productivity?

    3. albegadeep says:

      I have a desk job in manufacturing (you’re attracting a fair number of non-chemists to your blog, Derek!), and am a considerable introvert. I was moved from a tiny 4-person room at the far end of the plant to a full-height cube in the main admin area. It’s been driving me nuts. Conversations all around, people walking by constantly, constant distractions. Just the thought of an open-plan office gives me the willies. Put me in a 6’x6′ broom closet first!

      1. Derek Lowe says:

        Readers completely out of this field make me very happy!

  8. Hap says:

    Does constructing new shiny buildings or renovations correlate to how management views the people that work for it or the work they do? It doesn’t seem so (open offices, labs designed to look pretty and enhance reputation but not actually do work in), and as long as that’s the case, the architecture of lab buildings is unlikely to correlate to anything useful.

  9. dearieme says:

    In the New Museums Site in Cambridge there’s a plaque to Crick and Watson. Apparently the building it’s affixed to was chosen because it was less embarrassing than the old bicycle shed they actually worked in.

    This is too good a story to check.

    1. Diver Dude says:

      The accepted story in Cambridge is that they did the most creative parts of their work just around the corner in the “snug” of the The Eagle.

      Also too good to check but the beer is *very* good.

      1. dearieme says:

        I have checked though. Yep, the old bicycle shed had earlier been their hideout; before their spell it had been a hut used for some other science-ish purpose. Anyway there are photos of its being destroyed. So if the Powers That Be had put the plaque in the right place originally they’d just have had to rescue it and fix it somewhere else. And no doubt some loonie would have said that they shouldn’t demolish the old shed because of its extraordinary historical value as demonstrated by its plaque.

        I think the truly cautious approach would have been to affix the plaque to King’s College Chapel.

        The alternative analysis is that people might care to disguise the fact that earth-shaking science can be done in rackety old, cold huts if you combine the right individuals with the right scientific culture.

  10. DPM says:

    I recall a talk by Mark Fishman, given at the Broad Institute ~6 or 7 years ago, where he spoke somewhat on the effect of open offices in biotech. (I assume he gave this talk multiple times elsewhere, so maybe someone else may remember the details.) I remember he advocated for libraries and other quiet, private spaces for scientists to periodically hide in (i.e. escape the distracting, noisy open lab spaces), and I think part of his rationale was that many (most?) scientists lean toward introversion. My own experience in the Broad’s open labs were mixed. I certainly loved the community, ready access to collaborators, etc., but it was damn hard to read or write a paper/grant/etc.!

    1. ex-London Chemist says:

      We have open plan office, because that’s what the management group wanted

      The management group are about 95% “E” on the Myers Brigg Scale

      If it gets too noisy (oh, it does!) we’re banned from using any form of headphone/earplug to filter out the sound…..

      Who said “Hell is other people?”

      1. Hap says:

        Maybe it’s just your management that is Hell (or belongs there).

      2. loupgarous says:

        Jean-Paul Sartre, in his short story “No Exit”, which described three damned souls – a South American journalist who died after covering the wrong story and a French lesbian couple who suicided – whose punishment was to remain in each others’ company in an apartment with no doors.

        1. loupgarous says:

          I forgot to mention I was answering ex-London Chemist’s question

          “Who said “Hell is other people?”

    2. Skeptic says:

      In my experience, open design labs are fine (unless you’re working with something hazardous) as long as everyone has a desk somewhere quieter.

  11. myma says:

    I had the genuine thrill of interning at IBM TJ Watson research center (Yorktown Heights). Beautiful building, crescent shaped, designed by Eero Saarinen. The interior colors at one point were painted a full color wheel, which diminished in effect one by one after a policy to repaint places that needed repainting could only choose blue or grey. And it had real offices, and real labs, yet people would run into each other naturally because the corridor design was simple = inner edge radius, outer edge radius, spokes on a wheel between the edges.

    But BUT also after many years of homesteading and lab group expansion and contraction, the group labs were scattered all higgeldy piggeldy. I had to carry materials from one lab on one floor, to the far other side, other floor often enough that I would time it, and the inner radius would be several min faster than the outer radius because of length.

    1. tim Rowledge says:

      Yep; I got to work there briefly when I was an IBM research fellow in the early 80’s. Back when we were developing th 801 processor, one of the very first RISC machines.
      Lovely building in lovely surroundings, great colleagues. Shame about the vile weather….

    2. eub says:

      I didn’t know that about the history of the color scheme! Long before my time.

      I liked the effect that when you were walking down a spoke hallway, the glass door at the far end was angled to precisely reflect the hallway next to you. You couldn’t see yourself but you might see somebody else. Occasionally you would see somebody in the other hall walking about the same as you, and unconsciously take it as your reflection — and then your apparent reflection would stop and turn into a room, it was the damnedest thing to see.

  12. Ed says:

    One major failing, which I had the misfortune of dealing with all the way through grad school, is combining desks and lab benches in the same room. Particularly when lab safety strictly prohibits food or drink anywhere in the room, and the only “lunch room” was whichever conference room was occasionally empty. How the hell can anyone do any reading or writing without a cup of coffee? And people pretty much only ate lunch with a few close labmates, so there really was a lack of spontaneous and fruitful interaction.

    Of course, the de facto solution was to keep that cup of coffee hidden behind a pile of papers, and listen carefully for the sound of dress shoes entering the lab (90% chance of sales rep, 9% chance of lab safety personnel, 1% chance of deanlets or funders on tour.)

  13. CMCguy says:

    Probably an imaging correlation but I have seen/been part of numerous companies build then move to fancy new facilities then end up announcing downsizing or shut downs within 6 to 12 months. The focus behind construction often was more on giving impressions of grandeur over optimal functionality and wise budget management. I guess most Banker and Wall Street types need to see visible signs in creation of a showplace over investment in less obvious strengthening the personnel and tools to support project progress. And akin to layoffs experiencing a move to new labs, even if modernized and well designed, productivity usually will suffer for months or longer that it takes to resettle and relearn work flows (e.g. where to find everything) thus seems start with a negative balance that must be over come to reach or improve prior efforts.

    1. Scott says:

      When I transferred from the USS Georgia to the USS Kentucky (ships of the same class, and so almost identical in construction), I still had to re-learn where everything was. Things were stored in different places, and even the fire hoses were numbered differently.

    1. RandomWok says:

      Was waiting for someone to bring this up!

  14. Li says:

    “little or no empirical evidence”. (see EE Smiths (sci)fictional Arisians). It is a fundamentally flawed statement, imho, given that the light-cones of everything of practical import collide within the previous second or so. Is such a statement useful? When you use the word “empirical”, do you mean peer reviewed? replicable? Does the vast literature of testimonial evidence “count” in your world? Anyway, it’s a pet peeve of mine when anyone (usually someone with less research experience than DL) makes claims about the lack of “evidence” for this or that. (Prior to 1900 or 1950 was there any “evidence” for plate tectonics?) “No evidence” is just plain wrong. “No evidence to prefer hypothesis X over hypothesis Y” is imho the only acceptable context to make such vague claims. Anyway, that said, until we have AI`s sufficiently advanced to accurately model human cognition (and so make replicable experiments testing such alternatives ) we shouldn’t (again, imho) make claims that (currently) impossible evidence hasn’t been collected when it just simply CAN’T be observed. Creativity is a problem. We don’t know what it is nor where it “comes from”. Would the General Theory of Relativity ever have been created without Einstein? I really don’t know, do you? (I’d be willing to bet that Special Relativity would exist today, without him.) Would GR exist if Einstein hadn’t married his first wife? No way to know. That’s the problem with history, it’s historical (not a thermodynamic state). BTW, afaik there is no debate that open offices (OO) are significantly cheaper to build and maintain than traditional office space, that doesn’t deserve, imho, any discussion here. I suspect we can agree that OO`s will provide for the invention of x patents and closed offices for y patents and that x & y have a non-null intersection and that their intersection is large compared to either. Do we measure x:y by number? by 10th year sales volume? by ROI? by number of citations? There is, it seems to me, three important issues: 1. Is R&D a (sunk) cost or a profit center? (my ans: mostly a cost, especially if not well (and closely) managed), 2. Do researchers themselves know what their individual optimally creative work environment is?, (Ans: no) and 3. Is their work environment where we expect their creativity to occur? (think about your travel budget if this were the case).(Ans. creativity is a mental process which is extended in time (mental work) and can be increased with appropriate exposure to diverse ideas. Some of this will happen at work, and some of it will happen elsewhere.) I personally abhor OO layouts, but then again I’ve not won a Nobel, so …

  15. Anonymous says:

    Lipinski has written (can’t find the essay now) that having an office was crucial to his formulation of Lipinski’s Rules. (Some would use that an argument AGAINST open offices or FOR open offices. 🙂 ) It was where he could go to think about results and data and work on his ideas without distraction. Many others have similar, if less famous, stories to tell.

    There is a story that AC Cope’s office at Columbia (before he moved to MIT) was a converted closet underneath a stairwell. (Shades of Harry Potter, before Harry Potter!) (Personally, I have my doubts about that story, but even a converted broom closet is better than open office bedlam.)

    There is at least one fairly well known CEO of a mid-sized tech company (sorry, I can’t remember the specifics at the moment) who makes a big deal about open offices for everyone, including himself. You don’t even get a cubicle or desk; there is no reserved space. You show up and grab whatever space is available that day. He and his assistant never have a problem getting the space they need in the center of the room.

    I recently visited some open office space for engineering grad students at a local university. There was constant traffic, bumping into and squeezing by occupied chairs, a lot of distracting noise, and zero privacy, even for a phone call. I asked about the Prof: he was safely ensconced in his rear office behind the secretary’s outer office.

    HOW MANY OTHERS in management and admin and professing would give up their offices to join the Great Unwashed in the open office space?

    Put down my vote for private office space.

  16. You know, as I read this piece I mentally exchanged the architectural terms with genome/genetic terms and it still pretty much scanned.

  17. Sartre was onto something... says:

    I’m in a hot-desking situation in an open office, and assume it’s 100% about the ratio of people to space. Of course, that doesn’t sound warm or fuzzy so the design consultants talk about Silicon Valley and use lots of keywords to sex it up. But it’s simply a design that accommodates more people per square meter. Does anyone really believe all that other stuff?

    1. Hap says:

      It’s supposed to make it easier to clear everyone out and resell the property without much work when they lay everyone off. Don’t forget about that.

      I wonder when the C-team gets to hot-desk. That would be fun.

      1. Isidore says:

        I think it’s more than just being able to clear the property quickly, after all cubicle partitions are not permanent, I have seen them removed or reconfigured very quickly. I think it is more about power, conveying to employees that they are expendable commodities who can be readily replaced.

        1. Hap says:

          I guess this works as long as there’s a large supply of cheaper researchers elsewhere, at least until they realize that the value added by their “management” is negative and decide they can do it better, or decide that management can be insourced to wherever they are with significant cost savings.

          On the other hand, I keep waiting for logical consequences, and they doesn’t seem to happen, so somewhere there’s a hole in the logic. There seems to be an awful lot of “consequences don’t apply to us” going around.

  18. Argon says:

    “Open office” design is driven almost entirely by reducing cost. Nobody seriously thinks of it as an optimal office environment (except when HR has to sell it). It kills concentration. I’d also say it doesn’t inhibit web surfing as many people just put privacy shields on their monitors. (I don’t think web surfing is bad, FWIW – Work performance is what matters at the end of the day)

    “Open labs” are another matter. Not great for synthetic chemistry but perfectly fine for most biology as long as good housekeeping practices are maintained. I’m not too happy about working in a lab full of mass specs with their pumps, but otherwise, it’s pretty workable.

    One thing I’ve found is that very often, the purpose or content of the lab changes before the area is rebuilt. If you’re dealing with architects before the building even goes up, just make sure they don’t lock in designs that are hard to change after construction, because odd are, they won’t be filled with the originally intended occupants.

Comments are closed.