By James Kwak
Planet Money’s latest podcast features an interview with Matt LeBlanc, an efficiency expert. LeBlanc’s job is to observe various processes and figure out ways to make them more efficient. The idea, is that by increasing efficiency companies can save money, which ends up helping everyone through higher productivity and lower prices, even if some people get laid off along the way.
I am as much of a compulsive efficiency nerd as anyone (well, almost anyone). LeBlanc lays out his toiletries in the morning in a specific order in order to minimize transition time. When I lived in Berkeley, I figured out the fastest way to drive to school. The various possible routes were different paths through a grid that included some stop signs and some street lights; the best route involved slowing down at one intersection, looking to see if what color the light at an intersection was, and making a decision based on that. On one of my previous blogs I wrote a post about the quickest way to get through a security line at an airport. (Tip #1: Don’t unload your bags into the plastic trays until shortly before you reach the X-ray scanner. Your bags were designed to help you carry a lot of stuff with two hands; if you unpack them early, you have to move your unpacked stuff with the same two hands. Tip #2: Put your bags through the scanner before your computer and toiletries bag; that way you can have your bags ready and waiting on the other end so you can pick up the computer and slide it into your bag in one motion.) One of my pet peeves is businesspeople who fly frequently, make faces when standing behind families in the security line, and then slow down the line themselves because they haven’t figured out how to get their stuff onto the conveyor belt immediately after the person in front of them.
But I’ve become very skeptical of the simple argument for efficiency studies. (To be fair, LeBlanc is probably equally skeptical; but the podcast only put forward the simple argument.) The idea is that time has a monetary value (say, the per-hour employment costs of each employee), and if you save time, you save money. One example that LeBlanc mentions is moving printers. It seems to make sense on its face. You spend time walking to and from the printer. Therefore, printers should be located to minimize the total time people spend in transit, which could mean moving the printer closer to the heavy users of printing. Then those people can spend more time at their desks being productive.
But there is a serious fallacy in this argument: the assumption that the constraint on productivity is time at your desk. Let’s leave aside the issue of whether you are productive walking to the printer. The more serious issue is that you aren’t equally productive the whole time you sit at your desk. What if you spend your extra two minutes (in reduced time picking up printouts) at I Can Has Cheezburger?
Well, the efficiency expert may counter, all I need to assume is that a fixed percentage of your desk time is productive. But that’s still a big assumption. Maybe the real constraint on your daily productivity is mental energy, and you only have enough mental energy to do four hours of real work a day. Then your extra two minutes will all go to looking at pictures of cats with ungrammatical captions. Even more likely, maybe the real constraint is your internal sense of what a reasonable day’s work is. Many of us have either left early because we got a lot done or stayed late because we got little done. Maybe the real constraint is how much work your supervisor expects you to do. Maybe the real constraint is how much your colleagues get done, either for process reasons or simply because workplace norms are set by group as a whole. Maybe the real constraint is your motivation level. Maybe the real constraint is customer demand. (Another of LeBlanc’s examples is a cafe where the barista only spends half his time actually making a drink; the most plausible explanation is that you need to staff for potential demand, but actual demand fluctuates and is generally below potential demand.)
All of these possibilities seem much more likely to me than the idea that the limiting constraint is time spent at your desk. And if any of them is true, then moving the printer has gained you nothing.
You might think this is less true of more routine tasks, like moving and unpacking boxes (another LeBlanc example). Maybe, but I’m still skeptical. Let’s say you figure out a way to unload a truck, unpack the boxes, and put the stuff where it should go in half the time as before. Will the same people get twice as much done? Maybe, but I doubt it. Again, this assumes that the binding constraint on productivity is time. What if the work is physically strenuous, and while the old process included a lot of unnecessary pauses, those pauses were necessary to allow people enough time to rest? Then if you try to push them twice as hard, using a more efficient process, their bodies will break down. (If you could actually cut the effort in half, though, you might be onto something.) What if the limiting constraint is boredom: people just can’t work at peak efficiency for eight hours straight, and the old process with its delays gave them time to chat, look around, and relax a bit while working?
In many tasks, there is probably some room for efficiency improvement that can actually result in sustained higher productivity. But the benefits are only a fraction of the theoretical benefits you get by multiplying time savings by the money value of time.
BlackBerry (and its competitors) have made a fortune off the myth of efficiency. The reason BlackBerry is so popular with corporations is the idea that now people can be working while waiting in lines at the airport.* (Judging from the ads, this is the core use case.) That time is now money, at least according to the efficiency theory. But what are people doing? They are clearing out emails. If there is any benefit to anyone, it is that they will spend a little less time in the evening clearing out emails on their computers; but they won’t be doing any other work, because the length of their workday isn’t set by a clock, but by their sense of when they’ve done enough for the day. (For a lot of people, their willingness to knock off at the end of the day is related to the amount of email left in their inboxes.)
In addition, a lot of the supposed BlackBerry benefit is destroyed by four factors. First, working on a BlackBerry is less efficient than working on a computer (it takes more time to get the same stuff done), so some of your benefit (time waiting in line) is wasted in lower productivity. Second, checking your email constantly causes you to respond to emails and deal with issues that you could have simply ignored had you waited until you got home or to your hotel (since questions or issues posed in email often resolve themselves if you simply wait a few hours). Third, having a BlackBerry causes you to spend more time on email than you need to, because you can. Fourth, the quality of work you do on a BlackBerry is lower than on a computer. For example, with a computer, you can answer a question by finding a specific data source and actually finding the answer; with a BlackBerry, you are more likely to give an unhelpful answer like “try looking at source X,” which you may have misidentified, and which is less helpful to the person asking the question. But people lobby their companies to pay for their BlackBerrys because they want them, and companies often agree because they think they’re getting a more efficient workforce.
In case you’re wondering, I have no BlackBerry and no similar device for checking email while waiting in airport lines. (My dumbphone can, in a real pinch, check my email, but it’s so bad at it I rarely use it.)
* Obviously, there are some professions where constant accessibility is an issue–say, IT support personnel. The vast majority of BlackBerry users do not fit into this category.
Update: I should have emphasized that I agree with the basic point that there are vast, vast inefficiencies in the economy that, if eliminated, could have enormous benefits for all of us. In particular, we could make much more use of automation, especially through the expanded use of software (if we could find software companies that make software that works well, that is). Completely eliminating human touches reduces effort and takes all those other binding constraints I mentioned above off the table. Software can also sift through vast numbers of similar cases and determine which ones require skilled human intervention and which ones can be handled automatically according to some set of rules. We have made enormous advances in automation of manufacturing over the last century; I think we can make analogous, though smaller, gains through the automation of many service industries. What I’m more skeptical of are time-and-motion efficiency gains, especially when it comes to knowledge workers.