Category Archives: Uncategorized

Elder Care, Now By Robot

I have some friends who are planning to homeschool their kids one day.  When I’ve asked them why, they said their decisions was not a slam against teachers – in fact, one of them was a teacher herself.  Rather, they felt that teachers are put in an impossible position, and education suffers as a result.  On one hand, teachers are charged with teaching every child to the best of their ability.  On the other, they’re given dozens of often rowdy students to manage and are mostly evaluated not by teaching quality but by the mounds of paperwork – lesson plans and so on – that dominate their evenings and weekends.  Each individual teacher might be a paragon of educational talent, but they’re working in a system whose goals are a combination of teaching, warehousing kids so they stay conveniently out of grownups’ way during the daytime, and doing the occasional heartbreaking bit of social work.  And so if you really want to educate your kids, the argument goes, you’re better off teaching them at home, where you can really make teaching the focus.

I’m not sure I totally buy that argument when it comes to schooling, but it definitely applies to nursing homes.  Our Lady of an Undisclosed Location is the hospital of choice for several nursing homes, and so even though I haven’t seen the inside of the local nursing homes myself, I began to build up a picture from the patients that they send our way, like an astronomer inferring the existence of a planet from the debris it drags along its orbit.

I was once taking care of a frail old woman who insisted on getting up to go to the bedside commode by herself.  Usually, we’d have the nurse help pull them to a stand and then walk them over to the commode, but because she insisted, we instead hovered worriedly beside the bed, ready to catch her if she started to stumble.  But I had to ask – why was she so insistent about this, of all things?  And the story she told went something like this.

Imagine you run a nursing home.  You want to take the best possible care of your patients, but the nursing home makes money from each patient they carry.  So of course the natural strategy is to expand your census until you’re just barely able to keep up with the patients you have.  As a caring human being, of course, you might not want to go this route.  You might want to just take good care of a smaller number of patients.  But you know that even if you don’t expand, your competitors will, and soon you’ll be out of business and unable to take care of any patients at all.

Imagine you work for Medicare.  You’re glad to see so many elderly people being taken care of, but you’re well aware of the temptation for nursing homes to cram in so many patients that they can’t take care of them all.  Now, for most normal service businesses, you might just let the market work it out.  Word of mouth would let people know which nursing homes to avoid, just like people learn to avoid bad restaurants.  But the sort of people who need to be in nursing homes are not, as a rule, the sort of people who are actively posting reviews on Yelp, so you decide to take matters into your own hands by creating quality metrics to judge nursing home performance.

Now, one of the worst things that can happen to an elderly person is a fall.  As people age, their bones become more brittle, and their ability to heal falters.  Of patients sent to the hospital after a fall, 40% never regain independent function, and 25% die within a year.  Now, it’s hard to tell someone living at home alone to never fall, but in a nursing home, where you have people supervising the patients at all times, it’s pretty reasonable to expect falls to be extremely rare.  So, you institute a new rule: every time someone falls in a nursing home, you dock the nursing home a good chunk of its pay.  Not quite as good as getting those Yelp reviews maybe, but having some quality metrics is a lot better than letting nursing homes run amok.

Except when we go back to the nursing home, we see the vast yawning gulf between the letter and the spirit of the law.  Oh, the law works – the nursing home is keen to prevent falls and avoid paying the fine.  If anything, the fear gets magnified each step down the chain of command.  If you’re a shift nurse, you really don’t want your patient falling on your shift.  It would mean no end of paperwork, having to explain yourself to your supervisor, and worst of all, getting that “friendly reminder” mass e-mail that doesn’t quite refer to you by name but nevertheless has everyone gossiping the next day.

So you are careful to prevent falls.  Very careful.  So careful that you hover anxiously around whenever patients try to get up out of their beds, and insist on carrying most of their weight when they do.  And so patients have to go through a whole process just to get up to go to the bathroom, and their muscles and coordination get weaker and weaker for lack of practice.

This may be a good point to mention that muscle mass is one of the strongest predictors of mortality in the elderly.  By overzealously preventing falls in the short run, we’re literally making our elders weaker and likely increase their overall mortality.

So, despite everyone acting in perfectly good faith, we have a system that makes nursing home residents less independent, less happy, and sicker in the long run. That’s why this old lady was so eager to do something as simple as try to get up by herself.  As soon as she was outside the watchful eyes of the nursing home, she wanted to regain some of the function that the well-meaning nursing home had drained from her.

But hey, at least the fall statistics look great!

There are other ways, too, that nursing homes managed to systematically fail in taking care of the elderly. For instance, a lot of patients in nursing homes are demented to various extents.  One way that we grade how demented patient is to assess how much they’re alert and oriented, or “A+O” in our shorthand.  If you’re A+O times 3, that means you know your name, where you are, and what year it is.  As people get older, they start to lose these abilities. First they forget about some faraway abstract facts like what year it is or who the president is. Then they lose awareness of where they are and how they got there. Finally, they forget even their own name, although they can still carry on conversations and attend to their own simple needs. And when you’re that demented, when you’re “A+O time zero,” the world becomes this terrifying place – you’re in a strange building, with no memory of how you got there. Strangers are around you all the time doing incomprehensible uncomfortable things to you. Understandably, some of these patients become quite agitated by this turn of events, which hardly wins them many friends among the nursing staff, and the story often ends with the patients sedated by medications with some pretty nasty side effects, just so the staff can take care of them properly. Again, it’s a sad situation, but it’s hard to really blame anyone involved.

Except that one of the simplest interventions to reduce delirium and improve quality of life for the elderly is to give them a predictable environment, preferably with familiar faces around them. It stands to reason that even a fairly confused old lady would still be comforted by being at home in familiar surroundings and seeing her children around taking care of her. But instead, we put them up in these dorm rooms, with antiseptic lighting, noisy roommates, and a new nurse at the end of every shift.  Once again, nursing homes full of caring and attentive staff members manage to be uncaring and inattentive in the most ironic way possible.

To be fair, taking care of old people is really hard. It demands time, attention, and patience, and to a large extent we as a society have decided that we don’t want adult children to have to carry the full burden of all that. I’m as much a free-market man as anyone else, and it makes sense that once we’ve decided to outsource the care of the elderly, we’d have groups of professionals that try to replicate these services, for a fee. This approach works for a lot of everyday life. I don’t want to change my own oil, so I pay my mechanic to do it. I don’t want to build my own bookshelf, so I run down to IKEA. But what nursing homes show us is that there are some services that simply do not survive the transition to the marketplace. A loving daughter taking care of grandma is qualitatively different from three shifts of nurses checking off the boxes from Medicare guidelines.

There is a philosophical puzzle called the Chinese room that goes something like this. Suppose we have a room full of books, each of which has a set of rules pertaining to Chinese characters. A man with no knowledge of Chinese goes into the room, and sits there. Once in a while, a slip of paper with some Chinese characters comes in through the mail slot. The man examines the paper, applies the rules in the books, and scribbles out some characters in response. It turns out that the man on the other side of the door is a Chinese scholar. And he thinks that he is having a very interesting conversation with his mysterious penpal. Now, the question goes, if the man in the room is doing nothing but following a set of rote instructions, can we really say that he “understands” Chinese? One answer to this dilemma is to say that no, the man does not understand Chinese but the room as a whole does. So while the inside of the room looks a little odd, from the outside, the room is a whole does have the emergent property of understanding Chinese.

Well, in a nursing home we have nurses following instructions from both the administrators of the nursing home and from Medicare. The nurses themselves don’t intrinsically care about the patients in the same way that a son or daughter would – it would simply be impossible to expect that level of emotional labor. Instead they want to go to work, get their work done, and go home, just like anyone else. The administrators know this, and try to write up rules and guidelines that would allow them to treat the patients the way you’d want an elderly parent cared for.  But unlike the Chinese room, even with the best crafted rules, the system demonstrably lacks the emergent property of caring for the elderly.

It would be nice if we could simply revert to the old model of taking care of their parents in multi generational households. But that’s not terribly realistic to ask of everyone. However, there is one area of elder care which I think works reasonably well: the visiting skilled nurse. The notion here is simple. Instead of moving grandma into an institution that takes care of all aspects of her care, the family manages most of the mundane stuff; feeding, conversation, taking walks – and a couple of times a week, a nurse comes by and takes care of the tricky stuff – bathing, making sure the medications are being taken correctly, wound care, and so on.  This, in the best case, is outsourcing done right – with the difficult portions farmed out but done under the supervision of the caring children.

But doing this requires both discipline and courage.  Discipline, because the children need to step up to take on some of the duties of taking care of their parents.  Courage, because in a world where convenience is prized and nursing homes are trusted institutions by family and insurance companies alike, it requires a level of courage to say that no, for some things, care by strangers is simply not the same thing as care by loved ones.  But as my homeschooling friends would probably say, when it comes to the welfare of someone like a child or a parent, the sacrifice is no sacrifice at all.  And so, as strongly as they argue for homeschooling your kids, I would argue in favor of home-care for your parents.

The Paper Chase

Back when Facebook was a new exciting thing, you’d have these outraged stories pop up once in a while, complaining about the way that Facebook kept changing their privacy settings, gave out their data to advertisers, and generally wasn’t that concerned about their privacy. The most cogent response I heard was that these users weren’t paying Facebook money, so they shouldn’t be too surprised that the company had to figure out how to find revenue elsewhere. “If you’re not paying money,” they said, “you are not the consumer. You’re the product.”

Well, when patients are in the hospital, they’re usually not the ones paying their own bills. And in many little ways, the way healthcare is set up really drives home the point that patients are not the consumer. The insurance companies are the consumers.  The patients are the products.

The simplest way to see this is just to look at how doctors spend their time throughout the day. The average day in a hospital looks something like this. I get in around 6:30 AM, get the report from the night intern about what happened with my patients overnight, and have a quick meeting with my resident about the plan for my patients that day. I then take an hour or so to examine all of my patients in person, giving them a little update on their status and our plans but spending no more than 10 minutes per patient. Most of the rest of the morning is then taken up with rounding with the attending (our senior supervising doctor), where we once again go over what we think is going on with each patient and what our plan is. Sometimes during these rounds we would actually go and examine some of the patients in question, but more often, we’d simply crowd around a computer or refer to our notes, and just talk about the patient as an abstract cloud of clinical data.

The entire second half of the day is spent in front of a computer. Some of this computer work is part of patient care – keeping an eye out for a crucial lab result, or pulling up an x-ray that was just taken. But mostly, what we’re doing is documentation. Reams and virtual reams of documentation.

In the old days, documentation meant simply scribbling a couple lines on a clipboard at the foot of the patient’s bed. Maybe something like “Bobby, 30 yo M, here with pneumonia.  Continuing IV penicillin (or whatever they used back in those days), appears to be responding.  Will observe for improvement.”  They’d mention any additional orders for the nurses, and move on.  Then as now, doctors were always in a terrible hurry, which is where we got the stereotype of doctors’ terrible handwriting. But each of them had a dozen patients or more apiece.  No matter how fast they scribbled, all this documentation might take a whole hour out of their day.

Nowadays, of course, technology has advanced, and most of our documentation is done on the computer. And with this great leap forward in technology, we call it a good day if we get our notes done in three hours.

What’s in all this documentation?  Well, it’s not the important clinical data – vital signs, lab results, and so on.  Those are recorded by nurses and lab techs – left in much more careful hands than ours.  Rather, what we spend our time on is the paperwork that ensures that the hospital gets paid.  As before, this includes brief reports like the update on Bobby the pneumonia patient.  But it also involves documentation with much less clinical significance – things like a mandatory review of ten organ systems (“Oh hi there Mr. Heart Attack Guy, just for billing purposes, have you had any skin rashes since last night?  Problems urinating?  Anxiety or depression?”) and exquisitely worded diagnoses corresponding to the official ICD-9 table.

These little details mean big money for the hospital.  If a patient comes in with heart failure, and we notice an unrelated ulcer in his buttocks, that nearly doubles the amount the hospital can bill the insurance company.  And so hospitals naturally throw a ton of resources at making sure that every last box gets checked.

This dynamic is also what’s driving the adoption of hugely expensive electronic medical records.  You see, the most part, EMRs are heavily optimized for billing.  Yes, the idealistic talk about electronic medical records revolves around nice-sounding things like improving hospital workflow, having better access for research, and making information portable between hospitals.  But in reality, DOS-level user interfaces abound, many EMRs are inaccessible for research purposes, and almost none of them let you easily take your information to another hospital.  But the one thing that every EMR worth its salt has done is improve compliance with insurance regulations, with hard-coded forms that ensure that every last insurance requirement gets met.  And for that, hospitals are willing to pay whatever it takes.

Back when I was in medical school, Duke signed a contract to purchase Epic, one of the most advanced EMRs, for $700 million.  Now that’s a lot of money, much more than an impoverished resident is used to visualizing.  To put that in perspective, SpaceX, a private rocket company currently sending supplies to the ISS and hoping to send astronauts soon, was founded with an investment of $100 million.  So for what Duke paid for that EMR, you could go to space – seven times over.  And I have every expectation that that was a totally reasonable expense compared with how much Epic will help the hospital improve its billing in the coming years.

But even after this expense, the game is still not done.  Hospitals also hire roving clinical documentation specialists, who virtually peer over our shoulders like English teachers marking up our papers with red ink.  We’d get emails politely inquiring “That patient with heart failure, could you specify whether it’s systolic or diastolic heart failure?” “That patient with a pressure ulcer, could we give a specific grade of pressure ulcer?” Now, to be fair to the documentation specialists, most of their corrections are relevant, and sometimes they’re even clinically meaningful.  But some of them end up sounding hilariously bizarre to medical ears.  I once received a call about a patient whose kidneys had failed and was on dialysis, asking me to specifically “document the extent of her kidney injury” (“Uh, complete and total kidney failure…which is why she’s on dialysis?”)

And for each bit of bureaucracy the hospital adds, with the goal of eking a bit more money out of the insurance company, the insurance company has an equal and opposite piece of bureaucracy, with the goal of denying payment to doctors and hospitals.  Sometimes these denials are for good reason, such as noting that a procedure was unnecessary or that care was substandard.  But often any excuse would do when it comes to saving money.  Some insurance companies throw up clouds of paperwork that need to be done every time an expensive test is ordered, hoping to discourage doctors from ordering those tests for fear of wasting time on that paperwork.  When there’s conflict over whether a procedure should be reimbursed, hospitals and insurance companies alike hire physicians as consultants to help them argue their case.  Like the trenches of WWI, both sides bring incredible resources to bear, and yet the front never moves.  And as residents we just do our duty and try to keep our heads down.

It’s worth emphasizing here that I’m not criticizing doctors or even hospitals for being callous money-grubbing businessmen.  All this effort is absolutely necessary to keep the lights on and be able to continue treating patients who need medical attention.  In fact, it’s a strong endorsement of the ethics of the medical profession that doctors care for patients as well as they do, given the high stakes of the distracting paperwork game.  We still check in on our patients daily, comfort them and address their concerns, and spend considerable time agonizing what treatment would be best for them.  None of this behavior is economically optimal, but we do it because we care.

Still, on a good day I get to spend maybe 20% of my time face to face with patients.  I wouldn’t be surprised if they secretly suspected we all go out golfing in the afternoons.

Dude, where’s my hover-gurney?

Star Trek Into Darkness came out when I was in med school, and so like a good little geek I went with some friends to go see it.  The movie was entertaining enough, but for my fellow med students, the best part had nothing to do with spaceships and lasers.  It was a brief scene near the beginning, set in a futuristic hospital.  The building was a glass and steel tower and the doctors were carrying tablets, but the real gee-whiz gadgetry was the hover-gurneys floating in the background as orderlies carried patients from place to place.  As soon as we got out of the movie, we turned to each other excitedly: “Did you see the hover-gurney?” “Neat, hover-gurneys!”  But to me, the hover-gurney was more than a cool prop – it was, probably by accident, a perfect encapsulation of how healthcare adopts new technology.

Think about it.  Let’s say you had all of this miraculous technology.  You’ve got robots with practically human intelligence, teleportation, let alone whenever advances in genetics and nanotechnology they’ve invented.  Imagine how different the hospital would look if instead of having to do a operation, you could just teleport out a tumor.  Or if you could program a nanobot to sweep out a patient’s coronary arteries.  When it comes to moving patients from place to place, you could teleport patients to their destination, getting them there in perfect comfort.  Or maybe you could have robotic gurneys, eliminating the need to have an orderly whose job is just to push patients around all day.

Instead, what we saw on the screen was a hospital that simply swapped out one technology (wheels) with a slightly superior technology (antigravity) while leaving the rest of the process exactly intact.  I don’t know if the producers had a healthcare consultant on staff, but that scene was a great illustration of how healthcare tends to react to change.

Take a doctor from 1900 and transport him to 1960 and he wouldn’t even know where to begin.  In that interval we developed antibiotics, chemotherapy, x-rays, and the modern medical school curriculum.  Life expectancy in the US increased by over twenty years.  Since then, progress has been much more incremental.  Our chemotherapy is more refined, and we’ve developed a few less invasive surgical techniques.  But take a doctor from 1960 and transport him to modern times, and after little studying and a lot of computer training, he’d be ready to get back into practice.

Why the slowdown?  In part, you might chalk it up to low-hanging fruit.  Going from “no antibiotics” to “antibiotics” is a huge leap, one that’s hard to replicate by, say, developing a next-generation antibiotic that happens to have a couple fewer side effects.  But to a large extent there are also active forces that make it harder to incorporate technology.  Many a promising idea using proven technology has failed to revolutionize the clinic, because it somehow failed to navigate the institutional incentives towards becoming accepted into daily practice.  Let’s take a closer look at four promising technologies.  Two of them have successfully been adopted within healthcare.  Two of them failed to make the grade.

Back when doctors still scribbled on paper charts, there were two great ideas about how the computer might revolutionize medicine.  The first was the electronic medical record: basically taking the same documentation that doctors did and putting it into a computer so it can be searched and accessed anywhere.  The second was the expert systems: designing computer algorithms that could diagnose patients on their own, replicating the thought patterns of clinicians when they first hear about a patient.

Recently IBM’s Watson wowed the world by beating Jeopardy contestants, and there’ve been some exciting news stories about using the same system for medical diagnosis.  But in fact even as early as the 70s, researchers have developed “expert systems,” computer programs that can diagnose as well as fully trained doctors.  INTERNIST-I was a general purpose diagnostic system at the University of Pittsburgh, Stanford’s MYCIN system beat out infectious disease specialists in diagnosing infections and recommending antibiotics.  These proven technologies had a ton of potential – just imagine having expert diagnostic talent deployed in third-world countries, or having a computer system double check your doctor’s work to reduce medical errors.  And yet, while the electronic medical record has taken off, these expert systems were never deployed in clinical practice and remain mothballed in academic labs.

Why did EMRs get adopted while expert systems languish in obscurity?  A big part of the answer is that EMRs fit slid neatly into an already-existing niche in a hospital’s workflow.  Instead of writing a paper chart, doctors and nurses typed the same stuff into a computer.  There was no need to change any other part of the process.  As time went on, EMRs started to be increasingly optimized for billing insurance companies.  They could then promise hospitals not just searchable medical records but also more billing revenue, and so installing EMRs increasingly became a no-brainer.

Expert systems, on the other hand, didn’t fit into a niche.  There’s no way that an algorithm would entirely replace doctors in practice – patients expect to see a comforting person in a white coat, and of course doctors would fight tooth and nail for their jobs.  So how would you integrate this system into the hospital?  The best you could do is use it as a double check on doctor’s judgment.  But busy doctors wouldn’t really have time for this delay, and no insurance company is offering to pay extra for computer diagnosis.  So even though in theory expert systems had a lot of potential to produce better decisions and deliver better patient care, its very disruptiveness prevented it from making any headway.  As Peter Thiel said, disruptive technology may be cool, but disruptive kids get sent to the principal’s office.

Our next set of case studies involves a different sort of computer-aided diagnosis – computer diagnosis of pap smears and lung cancer screening CTs.  Every woman between 20 and about 60 is scheduled for a Pap smear every couple of years.  This test, a screen for early signs of cervical cancer, involves interpreting a large microscope slide full of cells.  Today, almost every single one of the slides is read by a computer, with no human intervention at all.  At the same time, many elderly smokers are recommended to get a high resolution CT scan to look for early signs of lung cancer.  Interpreting this study means scrolling through many images of the lung on a CT, looking for small lung nodules.  But despite being a similar needle-in-a-haystack problem, almost nobody uses computers for this study.

Why is the use of computer aided diagnosis so different in these very similar cases?  Again, a lot of it goes back to history.  By the time computers came on the scene, it wasn’t doctors reading most Pap smears.  Instead, a trained technician screened all the studies – the vast majority were normal – and only passed on the really tough cases to the physician to read.  So the notion of having a pre-reading screening process was already well established.

Not only that, at that point in time, there was an undersupply of these technicians, and new work hour rules were limiting the number of slides that a technician could read in a day.  This became a real bottleneck in doctors’ ability to read lots of slides (and bill for the service.)  So when computer-aided diagnosis company started peddling their wares, doctors were eager to take them up on their offer.  The largest of these companies, Cytyc, was worth $6.2 billion when it was acquired.

For radiology, there’s no such tradition of pre-reading studies.  So adding a computer into the mix would represent a much bigger disruption of the workflow.  In theory, radiologists could adopt the same workflow as the pathologists and let computers pre-read the studies, letting them read faster and bill more.  But this would represent significant behavioral change, and there isn’t as much legal precedent to make them feel comfortable trusting their medical licenses to the computer.  There might still be a fortune to be made here, but the fact that it has not happened yet suggests that the barriers are a good deal greater.

And this is the overall sense I get when it comes to incorporating technology into healthcare.  If a new gadget neatly replaces an existing gadget, it’s adopted relatively quickly.  If it directly impacts the bottom line, such as EMRs offering hospitals better billing revenues, so much the better.  But true disruption – something that fundamentally changes the way medicine is done – is extremely difficult to sell, even if the technology works and the benefits could be enormous.  That’s why doctors are happy to carry around iPads but only a few are willing to Skype with patients.  That’s why radiologists were happy to swap keyboards for Dictaphones, but screening patients’ genomes is a niche pursuit.

In some ways, this conservatism is a necessity.  There aren’t many cowboy doctors left; the trend today is towards highly specialized medicine and enormous hospitals.  A solo practitioner might be able to tinker with new technology, but for something as complex as a large hospital, any disruption has huge ripple effects that administrators are eager to avoid.  And yet, there is an unfortunate side too.  A lot of the great advances in medicine – sterile surgery, radiation therapy, even the residency training process – were innately disruptive, with some doctors losing turf while others gained, and with everybody having to figure out how to deliver medical care in fundamentally different ways.  It seems no coincidence that the death of the cowboy doctor and the rise of bureaucratized medicine has come with a slowdown in medical innovation.  But savvy medical entrepreneurs still find ways to produce some innovations that can make patients’ lives better.  And one day, I hope, they’ll at least give us that hover-gurney, because that thing looked pretty awesome.

The City as Drill Sergeant

I recently got back from a road trip through New England with a friend.  The trip itself was fun – scenic drives, picturesque towns, and some beautiful remote hikes.  But one of the things that made the strongest impression on me happened the day before we left.

My friend had some business to take care of before we headed out, so I spent the afternoon strolling around Cambridge and Boston.  It’s a good town for walking, with handsome campuses interspersed with humbler neighborhoods of houses and shops.   That day, it was clear and unseasonably warm out, and the whole town turned out to enjoy the weather after a record-breaking winter.  Everywhere I went I’d see young, thin, well-educated people – a pretty different demographic than I’m used to in West Philadelphia.

On the subway I’d see students with their noses buried in books or Economist articles on their iPhones.  On the banks of the Charles I saw young couples strolling hand in hand, or lounging on the grass watching the antics of a group of boys on paddleboards in the river, who were whooping as they got splashed with frigid water.  In cafes I’d see small groups of grad students with heads together, talking animatedly over their beers.  And everywhere, there were the ubiquitous joggers, running from nowhere to nowhere in easy loping paces.

And as the afternoon dragged on, I noticed a little voice in the back of my head start saying things like “Hmm, I wonder what that gaggle of students is working on?  What’s that guy learning about right now?  How did that couple meet?  Hmm, maybe I should swing by the gym myself,” and so on.  Nothing terribly obtrusive, but just a tiny low-level paranoia that I was missing out on some important improving experience that these other people around me were taking advantage of.  It was a funny feeling that I hadn’t really encountered except for back when I was in college.

But when I sheepishly mentioned it to my friend, he immediately recognized it and said that he felt it all the time, walking around the campus of business school and seeing other highly competent people going about their business.  And really, it was no bad thing.  This brief experience did, in fact, inspire me to hit the gym more seriously once I got back, and to bang these musings into essay form on the flight back home from Boston.  Certainly, I could imagine it growing into oppressive insecurity if it were stronger, but on the whole it resembles exercise – a hormetic stressor that makes you stronger.

Paul Graham wrote an essay about what messages cities send to ambitious people.  I suspect that something like this is the actual medium by which these messages get sent.  Nothing obvious, just a constant reminder that other high-level people are out there improving themselves, and that you’d better keep doing the same or else fall behind.  And this is what helps to produce both talented individuals and talented ecosystems.

Everything else is an epiphenomenon.  All the ads on the subway exhorting people to work for a company trying to cure cancer, or to go back to school to get on the cutting edge of robotics, the indie bookstores and small museums, even the tech companies that get started, fundamentally derive from the existence of smart people whetting one another’s appetites.  I didn’t notice its absence in Philadelphia, but I certainly noticed its presence in Cambridge.  I suspect that this feeling of “fire in your belly” is underrated as an ingredient in the good life.

Elasticity to innovation

I’ve been doing a book club with a few friends on The Idea Factory, a history of Bell Labs. I’ve previously written about this book, but in thousand ideas around it came across the concept that I hadn’t thought of before.

The obvious question that comes up when thinking about Bell Labs is “what made Bell Labs awesome, and what are the analogous institutions today?” Some of these factors are obvious. Bell Labs was most creative during the period when innovation information technology was one of the most dynamic fields of science; they were lucky because they picked the field that had a lot of low hanging fruit and which was not, in the end, regulated to death, unlike medicine and aerospace. They also had the benefit of recruiting a lot of top talent during a period when bright Midwestern kids were being recruited to places like Caltech MIT for the first time. In a previous article, I suggested that one hidden advantage was that Bell Labs had a very specific planning horizon that let it look farther than the next quarter but didn’t let it have its heads entirely in the clouds.

Another idea that came up, though, is the notion that Bell Labs also had a high “elasticity to innovation.” That is, the company was positioned in a way so that every innovation ended up having a large effect on the company’s bottom line.

There are two reasons why this is the case. One, which we’ve mentioned before, is that the company was a monopoly. There weren’t any phone companies of comparable size, and so that meant that there was no worry that any innovation they adopted would be quickly adopted by its competitors as well. If Dell adopted a new way to make computers, they’d only have a short window before HP did the same, but if you’re a monopoly, that by definition means you get dibs on the innovation forever.

In the second and more interesting reason is that this period of history is one where the telephone network was just beginning to really take off. Less than 10% of homes had a telephone, and this meant that the company knew that if they could improve the telephone experience, they were looking at a 10x increase in phone adoption and a 100x increase in call volume.  Innovation meant dollars, lots of them. And rather than taking innovation to be some magical, highly contingent process, we could think of it as something that responds to incentives, and appears only when it’s really needed.

This helps shed some light on why some monopolies are innovative and others aren’t. This is a question that I had struggled with earlier, and which Peter Thiel also touched on in his book Zero to One. Clearly, some of the really innovative companies have been monopolies, but lots of monopolies didn’t innovate much at all. I think that this notion of elasticity to innovation helps to explain the difference. If you are a diamond mine or the only bridge across the river, there’s no way for you to increase your revenue by 10x by adopting some new process. Your main asset is your location and monopoly status. Rationally, then, these companies would invest mostly in maintaining their monopoly status and exploiting their pricing power, giving relatively little thought to innovation. Only in its monopolies where you are creating the market as you go along, such as Bell Telephone or early airplane manufacturers, do you have a strong incentive to innovate. And indeed, we see that monopolies in these areas tend to be the ones that we think of as innovators.

This theory helps us make some predictions about which modern monopolies we might expect to be unusually productive. Companies like Microsoft for enterprise software or Google for search we wouldn’t expect to be extraordinarily productive, despite the monopoly status. However, the same companies may end up being innovative in different areas. Googles self-driving car, for example, is a case where the market, and by extension the company’s revenues, will only exist in as much as they were able to create meaningful innovation. Disruptive companies are innovative for the same reason; their market only exists in as much as they can wrench it from the control of incumbents. But there’s no special DNA in disruptive companies that will prevent them from being stagnant once they have become the kings of the hill.

This also lets us make some predictions in medicine. In this age of hospital consolidation, there are several companies, such as Kaiser and UPMC, which have local monopolies. By virtue of their location in consolidation, they basically have a lock on the medical care of patients in their geographic region. However, crucially, they can’t actually make a great deal more money by being better hospitals. Even if they came up with a new kind of robotic surgery, or reduce hospital infections in half, there’s only a limited extent to which patients would come specifically to them. And so, while I think it’s possible to see some process innovations from economies of scale, we wouldn’t expect explosive medical innovation to take root in these monopolies. On the other hand, there are some companies that are creating the market as they go along. This includes concierge physicians, which are trying to disruptively grab market share from established healthcare systems, as well as services like 23andme and Foundation Medicine, which are premised on creating entirely new markets for genomics and cancer diagnostics. Of course, we can’t predict whether or not these companies will succeed, but this model does predict that they have the incentives to make a valiant effort.

Personal finance for new residents

As a new resident, I was recently put in the funny position of going from negative income to positive income for the first time.  And so, just before the start of residency, I took some time to set up a personal finance system from the ground up.  Since starting residency, I’ve helped a few fellow interns sort out bits and pieces of their financial infrastructure, so I thought I’d write up my system, in case it’s helpful to future interns.

As a note, this is not a fantastic path to riches, just a simple system that works.  My philosophy in setting all this up is to set it and forget it – finding the optimal tool for each function, and then leaving it alone to work its magic.  After all, as busy residents, we don’t have time to keep tweaking and adjusting.  So it makes sense to do the homework upfront, set it up right, and set it on autopilot.

  1. Bank

The purpose of a bank is not a holding area for savings (that comes later).  The purpose is for it to be a simple funnel between paychecks coming in and payments (for bills, credit cards, and savings) going out.

For that purpose, online banks are significantly better than brick and mortar banks.  They let you write and deposit checks and have ATMs, all the functionality that I actually use, and they’re equally FDIC insured, so your money is equally safe.  But because they don’t have to pay for all the physical locations and staff, they have lower costs and can pass along some of these savings to you.  This means two things: slightly higher interest rates, and more importantly, no hidden fees.  Many other banks have minimum deposits or hidden fees; online banks tend not to, which makes them less of a headache to deal with.

I use Capital One 360 as main banking account; there are other good options as well, such as Ally and GE Capital.  I direct deposit my pay there, and set up auto bill pay for everything possible, including rent, utilities, and credit cards.  It’s possible to supplement this with a local bank or credit union account if you want some of the old-school banking services; in practice I haven’t found that necessary.

  1. Credit cards

As long as you pay off your balance every month, credit cards can be a useful convenience.  They also provide a tiny slice of rewards back, which is a nice bonus.  Since we’re not carrying a balance, the main thing to look at is the rewards.  I ended up using the Sallie Mae Mastercard, which offers by far the best rewards among non-fee cards – effectively 5% back for groceries, gas, and Amazon purchases, albeit with a cap on each category.  There are cards with annual fees that offer a greater percent back, and if you spend a lot more than I do (you can do the math yourself; it generally runs around $15-20k/year) it may make sense to get a fee credit card with even better rewards.  It’s possible to get super nerdy with getting multiple different credit cards for different spending, but I haven’t gone to those lengths, keeping in mind that the difference will likely be less than $100/yr.

  1. Investments

If you have money left over, it’s worth thinking about where to put it.  Depending on your situation, it may be worth it to pay off loans quickly.  But it’s also well worth thinking about investing some of the surplus.

Before you even think about stocks versus bonds, think about taxes.  “Retirement savings” may not sound exciting, but there are two categories of retirement savings that are tax-exempt, which is a pretty big deal.  At a resident’s salary, that means an instant 25% bonus on your savings, which is well above what even the smartest Wall Street jockey can get you.  Briefly, in normal taxable accounts, you pay taxes when you receive the income, and you pay capital gains taxes when you withdraw it in the future.  With a Roth IRA, you pay taxes on your income, but you don’t pay taxes when you withdraw it.  With a 401k, you don’t pay taxes now, but you pay taxes when you withdraw it in the future.  Many hospitals offer a match to your 401k contributions; this is free money you’d be a fool to turn down.  So at the minimum you should set your contribution to your 401k such that you get the full employee match.  (For people working at nonprofits, which includes most hospitals, you have a 403b rather than a 401k; they work the same way.)

As residents, we’re in a unique position where we’re making fairly small amounts of money now, but have a solid expectation of making an attending’s salary in the future, which also means paying an attending’s tax rate.  So it makes sense to prioritize the Roth IRA, and pay the taxes now, while we’re poor and the taxes are lower, rather than in the future.  But both tax-deferred savings accounts are pretty good, and overall I’m maxing out my contribution limits in this order: Roth IRA ($5,500 limit) > 401k ($18,000 limit) > ordinary savings, after accumulating a small rainy-day buffer in my ordinary savings account.

It’s important to note that for the Roth and 401k, there are severe penalties for withdrawing funds early.  These savings are for retirement, not a buffer for immediate spending.  It’s possible to use the 401k money as backing for a loan for a mortgage, so you still get some value from the money stored in there, but it’s not meant to be touched until retirement.

Money in the retirement accounts can be invested, just like ordinary savings.  I won’t get too much into portfolio theory; I generally believe the efficient market hypothesis that it’s very difficult to consistently beat the market.  So, I’m a fan of low cost, broad-based index funds, such as those offered by Vanguard.  (The “low-cost” aspect is worth emphasizing.  Some funds have expense ratios of 1% per year or more, which can add up to enormous savings lost over a lifetime, even if the market is otherwise doing well.)  For early savings, a Vanguard target date retirement fund is a good option – it automatically rebalances a broad portfolio of stocks and bonds and shifts from aggressive to conservative as you get older.  Furthermore, and this is important for first-timers, it has a low minimum contribution; $1,000 compared to $3,000 or more for other funds.   From there, it’s worth exploring further and diversifying into things like international markets and REITs (real estate), but this is a good way to get started.

You’ll have noticed that I haven’t mentioned a savings account in this section.  That’s because in practice, it makes more sense to keep your rainy day fund in a taxable investment account, where it’s working for you at a much higher rate than your bank can offer.  You can still sell off some stocks in the account to cover your expenses if an emergency arises.

  1. Mint

Mint is a great piece of software.  You can link it to your credit card and bank accounts, and it passively reads your activity and compiles a snapshot of all the money you spend each month.  This lets you really see where all your money is going, and what you’ve been spending on.  Over time, you can develop a sense of where it makes sense to cut back, and where it’s okay to spend more money.

  1. Profit!

And that’s pretty much it.  With all this set up, your monthly pay is deposited into your (online) checking account, and your bills and credit cards are automatically paid off.  Your investment savings are also automatically funneled off, and the rest is yours to spend as you like.  With that, you can have your finances mostly set on autopilot, and just worry about getting through residency.

Weaponized reductionism

One of the fun parts of Neal Stephenson’s Anathem is the way the book forces the reader into a sort of reverse culture shock, understanding the full strangeness of a post-industrial society through the eyes of his cloistered scholars.  He describes familiar things in this world with such care and attention to detail that they become strange again:

“Before leaving town we stopped, or rather slowed down, at a place where we could get food without spending a lot of time.  I remembered this kind of restaurant from my childhood but it was new to the Hundreders.  I couldn’t help seeing it as they did: the ambiguous conversation with the unseen serving-wench, the bags of hot-grease-scented food hurtling in through the window, condiments in packets, attempting to eat while lurching down a highway, volumes of messy litter that seemed to fill all the empty space in the mobe, a smell that outstayed its welcome.”

I’ve been reading Marcus Aurelius’s Meditations recently, and one of the things that really jumped out at me was the extent to which this Roman emperor deliberately used the same reductionist technique to not to let the pomp and luxury of his office make him proud or put on airs.  He wrote:

“[Cultivate] disgust at what things are made of… marble as hardened dirt, gold and silver as residues, clothes as hair, purple dye as shellfish blood.”

He applies this sort of reductionism throughout his book, and uses it to try to see through the social conventions of his day, and arrive an objective view of what in his life was worth valuing.  Instead of living in a resplendent palace, emperor of the known world, he forced himself to see it as living in an artificial cave, ornamented with bits of shiny stuff dug out of the earth, and instead of lording it over senators he simply saw himself as dealing with other humans who happened to wear hairs dyed in purple shellfish extract.  “Latch onto things and pierce through them, so we see what they truly are.  That’s what we need to do all the time…to lay them bare and see how pointless they are, to strip away the legend encrusts them.  Pride is a master of deception.”  Without saying a single tendentious or untrue thing, he manages to transform these luxuries and totems into the mundane and frankly bizarre.

Elsewhere, he performs the exact opposite trick, turning a mundanity into a transcendent purpose:

“What’s left for us to do?  I think it’s this: to do (and not do) what we were designed for.  That’s the goal of all trades, all arts, and what each of them aims at: that the thing they create should do what it was designed to do.  The nurseryman who cares for the vines, the horse trainer, the dog breeder – this is what they aim at…I do what is mine to do; the rest doesn’t disturb me.”

And what was this telos that we were designed for?

“Revere the gods; watch over human beings.  Our lives are short.  The only rewards of our existence here are an unstained character and unselfish acts.”

So he uses reductionism to tear down temptations, while telling grand, non-reductionistic narratives to encourage himself to virtues.  And his book is really powerful in changing perspective because of this selective use of reductionism.

Now, you could imagine the opposite: a situation where people explain away virtue as just old-fashioned twaddle – or worse, nothing more than the deceptions of an entrenched power structure or an emergent property of Darwinian selfishness.  At the same time, they might tell really compelling, teleological stories about the importance of success at short-term things, whether being a business titan, winning a political debate, or dating someone really attractive.  Of course this would be pretty crazy and I don’t know why anyone would do this.

Learning Chinese as an ABC

This past summer, I took advantage of the two month gap between the end of medical school and the start of residency to do some serious vacationing, visiting family in Taiwan and sightseeing through Singapore and several cities in southern China. Leading up to this adventure, I spent some time beforehand – starting in the last year of med school – starting to learn Chinese.

I’m in a fairly common position for an American-born Chinese guy.  Growing up, we spoke Chinese at home, so I picked up a fair amount by osmosis – but only the sorts of things that parents would talk about with their kids in casual conversation – no curse words, no analysis of Warring-States era philosophical schools.  So I started out with a decent grasp of grammar and vocabulary, and the challenge was to go the next step and learn to read and write.  My informal goal was to be able to read at a grade-school level – the point where I could walk down a street and read signs, menus, and sitcom subtitles.

The vacation was a lot of fun, involving lots of sightseeing, catching up with relatives, and good food.  I also got plenty of opportunity to continue studying the language, and I found to my surprise that I had largely achieved my goals; I could get around town, order from menus, and even read simple books – and I’d gotten there with a minimal investment – studying Chinese about 15 minutes a day before my trip.  Along the way, I learned some valuable lessons and made some mistakes, and I wanted to share this experience for anyone else looking to pick up Chinese reading skills.

Use spaced repetition.

About a year before my trip, I sat down and tried to figure out what kind of study plan would make sense to try to accomplish my goal.  I realized that what my problem really boiled down to was matching characters to pronunciations and definitions – often, words that I already knew how to use in conversation.

I then realized that spaced repetition flashcards would be a great tool for this job.  I’ve been a fan of spaced repetition for a long time now, and flashcard programs like Anki really shine when it comes to vocabulary learning.  The electronic flashcards would let me test myself for both reading (“here’s the character, what’s the definition?”) and writing (“here’s the definition and pronunciation; what’s the character?”), and the spacing algorithm could monitor my progress and ensure that I spent time on just the words I most needed to review.

I found a freely available list of the 3,000 most commonly used Chinese characters and converted them into an Anki deck.  For each word, this deck contains a card showing the character and asking the definition, as well as a card showing the definition and asking how to write the character.  Paired with the touchscreen of a phone or tablet, this let me conveniently practice writing and recognizing characters.  I worked my way down the list, learning the characters for all the words I already knew how to use in conversation.  Starting several months in advance with a very leisurely schedule of 5-10 new words per day, I had almost a thousand characters under my belt before heading out.

(If you’re interested in a copy of the deck, contact me and I’d be happy to send it your way!)

Quick and dirty

In Chinese classes, they usually teach characters by having you learn a character, then write it out a dozen or so times.  This is a classic massed practice technique, and it makes sense in the classroom setting.  Having intensive training lets you lock in the words, and since your teacher can ensure that you’ll start seeing those words in your vocabulary lessons immediately, you have the spaced practice component as well.

Real life, however, isn’t as well organized as a vocabulary workbook.  You’ll be seeing street signs, magazines, TV subtitles, and more, and your goal is comprehension, not scoring 100 on a test.  There are statistics floating around about how if you learn the most common 200 characters you’ll be at 50% comprehension, but that’s a pretty misleading notion.  Yes, you’ll be able to recognize 50% of the characters in a sentence, but that doesn’t mean you’ll have a 50-50 chance of figuring out what it means.  Just imagine reading that previous sentence without knowing what “recognize,” “characters,” or “figuring out” mean – even small holes in a sentence – usually around the more difficult or specific vocabulary words – can kill your comprehension.  Getting over that hump – from understanding a few words but losing the gist, to understanding the gist but missing a character or two – is very important to making the most of immersion.

That means the breadth of your vocabulary is key.  It’s much better to have a general sense of the meanings of 1,000 characters than a superb grasp of 100.  So be aggressive about adding new words to your vocabulary.  Don’t worry if you have shaky or vague sense of what they mean – that’s what Anki and everyday immersion are for.

For similar reasons, speed is more important in the real world than it was in school.   In the real world, you don’t have more than a few seconds to read a sign or catch a subtitle, so when you’re out and about, it’s important to deliberately train the skill of reading fast, even at the expense of some precision.

Simplified or traditional: pick one

As you probably know, there are two Chinese writing systems.  Traditional Chinese characters are used in all pre-1949 writing and are still used in Taiwan and Hong Kong.  Simplified Chinese characters are used in mainland China and Singapore.  There are complicated linguistic arguments in favor of both systems, but in practice, just pick one system and stick to it.  If you’re going to a country where traditional characters are used, learn that; if you’ll be in a place where simplified characters are more common, go with simplified.

Transitioning between one system and another is doable but very tricky, especially for a beginner.  As with most languages, the transition from traditional to simplified characters applied a small number of definite rules, plus a lot of idiosyncratic decisions.  As a result, you can use the rules and mostly understand a sentence from the other writing system means, but it takes a fair bit of effort and there will be gaps in your knowledge.  My uncle, who lives in Taiwan, related that he once needed to consult a rare engineering textbook that was only available in simplified Chinese.  He managed to get through it, but it involved quite a bit of effort – like reading Chaucer.

My experience bore this out.  I’d focused on learning simplified Chinese in my studies, but part of my trip took me to Taiwan, where traditional Chinese ruled.  Here, I was much less adept; often I’d be puzzled by a character and look it up, only to find that it was simply the traditional version of a word I already knew.  This confusion effectively shrunk my vocabulary, which reduced my comprehension and hindered my progress.


There are a lot of apps that help in learning Chinese, but Pleco was the best I found.  It has an extensive dictionary that allows you to look up characters by writing them on a screen, as well as a cool OCR feature that lets you take a picture of something you want to read – a billboard or a printed page – and can identify the characters for you.  Either way, you can look up new words on the spot, and save them for later review – every word you see, even when you’re out and about, is a potential addition to your vocabulary.  This is a powerful way to build your vocabulary quickly.

“Respect your elders!”

One thing that I definitely got out of this experience was a new appreciation for my parents, and for immigrants in general.  Even with some reading ability, when you’re using a different language, every little task – from navigating the subway to ordering at a restaurant – is just a little bit trickier, and this burden definitely adds up over time.  And yet, at my age, my parents were able to make that transition and read and speak far better English than I speak Chinese.  On my desk I have a small stack of kid’s books that I’m slowly working through; at my age my father was reading grad school textbooks.  I may never get to that level of fluency, but if I get no other use out of my language-learning adventure, I can now really appreciate what a big hurdle that was.

The Subversiveness of Professionalism

This essay was the winner of the Andrew Puckett Essay Award at Duke University in 2014.

“I don’t know why everyone’s playing these silly games,” the man said, even before my preceptor and I walked into the room.  “I went to my surgeon and he said he wouldn’t operate unless I quit smoking.  What kind of treatment is that?”   Halfway through a busy surgery clinic, a patient like this was the last thing I wanted to see.  As he fidgeted and leaned forward in the chair, I could tell that he was exactly that sort of talkative patient that I still couldn’t figure out how to disengage from – at least, before I got a pocket lecture on the history of his illness, his parents’ health, and the nice old lady who comes by and looks after his dogs while he’s away at the hospital.

I was surprised to see my preceptor – who’d been all about efficiency earlier in the day – pull up a chair and level with the patient.  He explained clearly why his peripheral artery disease was uniquely sensitive to his smoking habit, and how surgery without any lifestyle change would have been useless to him.  In the end, the patient left not only satisfied but with a new resolve to quit smoking now that he understood its link with his disease.  When we got out, the preceptor remarked that we’d set ourselves back a good half hour on the clinic schedule.  “But in these situations, you have to do the right thing.”  I knew what he meant.  As a result of his decision, both he and the clinic staff would have to stay later.  But he turned around a patient who would otherwise inevitably lose control of his disease.  In the long run, this conversation with the patient was one of the most powerful prescriptions he could wield.  With a half-hour talk, he did more good for this patient than he could have in a whole afternoon in the OR.

A few weeks later, we had a lecture on professionalism in medicine.  Just as we were settling down into the gray seats in the auditorium, the speaker asked us for our own definitions of professionalism.  The answers were all over the place, from “being a team player” to “working hard,” – and the anodyne lecture did little to dispel our confusion.  As I existed I distinctly overheard someone guessing that professionalism simply meant whatever people wanted it to mean.  Yet as unsatisfying as I found the lecture’s exploration of the topic, I couldn’t agree with this cynical take.  I had seen, in that surgery clinic, a small example of what I admired most in medicine, and there had to be a foundation behind my preceptor’s stance.

So what was professionalism, really?  One way to approach a definition is to see what the opposite of professionalism is.  The obvious answer – the answer the lecturer would give – is that it’s sloppiness and indolence – various ways of not getting your work done.  But I don’t think that’s quite true; after all, those would be vices in any situation, not just in a job or a “profession.”  Instead, the true opposite of professionalism is being a company man.

A company man is someone who plays by the rules, who internalizes the values of his employer or the rewards system of the market and aspires to nothing greater than to play that game well.  It’s the kind of person – peaceable, flexible, capable – who can fit into any institution and make it grow.  Company men were what made it possible to build General Motors in the mid-1900s, and they’re what makes it possible to build large university health systems today.  Modern capitalism needs company men, and to be a company man is no bad thing.

But it’s a very different thing from being a professional.  A professional is someone who holds to values – whether the physician’s oath of Hippocrates or the lawyer’s reverence for the law – that transcend any other claims the world might have on him.  Professionals understand the system that surrounds their work, and are able to work with it, but do not let the needs of the institution displace their professional creed as their guiding value system.  And a profession is nothing more than a group of people who hold to a common creed.

It was a small act of professionalism that allowed my preceptor to spend the extra time it took to help that patient.  He may not have even consciously realized that he was making that choice – it may have been instinct by this point in his training.  But in that act was a small echo of the professionalism of Ignaz Semmelweis, who spent decades going above and beyond the call of duty to drastically reduce death in childbirth.  And of Barry Marshall, whose courageous self-experimentation overturned a century of dogma on the cause of stomach ulcers.  In the long journey from ignorance and folk medicine into the scientific medicine we practice today, every step was made possible only by an act of professionalism – of someone making the choice to actually help people when the easy answer was to keep doing things the way they’d been done before.

And that courage is still needed.  Scientific, evidence-based medicine may be enthroned today, but there are new challenges to providing patient care.  Distorted payment systems force doctors to spend less and less time with each patient.  The rigors of dealing with billing encourage doctors to give up independent practice and throw in their lot with ever-larger healthcare organizations.  None of these changes are bad in themselves – indeed they’re necessary defensive measures given the environment we practice in.  Yet at the same time, these changes are a recipe for turning professionals into company men.  It puts ordinary physicians under pressure to change their practice to suit the payment system or the organization’s needs, leaving them less flexibility to let their practice be guided by the dictates of their professional creed.

Perhaps the most ominous sign that professionalism is in danger is that it’s seen as a harmless bit of virtue, featured in inspirational posters and preached to medical students.  For professionalism, properly understood, is not a peaceful creed.  It’s a force that says that you must do what is right by the patient, no matter what hospital policy or your billing agent may say about it.  It says that fitting in matters much less than doing what’s right.  It is a force that topples orthodoxies, disrupts routines, and upsets accounting managers to no end.  If we had a truly vital ethos of professionalism, it would not be the kind of idea you put up on inspirational posters.  It would be the kind of ideology you furtively whisper to your close friends, and public agitation for professionalism would be a fireable offense.

Yet even if nobody’s pushing this robust, warlike professionalism, a humble, quiet version of it still lives on.  I caught a glimpse when I saw my preceptor take the extra time with that patient.  And once I started noticing it, I began to see it more and more.  I saw it in an administrator pushing for a costly quality improvement project; I saw it in the surgeon who took up precious OR time to teach the residents.  I even saw it in my classmates who chose to study important facts they will use in their careers, rather than cramming the minutiae that show up on tests.  Each of these, I knew, would pay a price – the administrator may have a harder time making budget; the students may score lower on their tests.  But it’s that willingness that made the virtue so striking.  And as long as physicians are willing to go that extra mile, I have great hope for the future of the profession.

Bell Labs and the Culture of Innovation

Bell Labs was one of these crazy institutions that did so much, in such a short span of time, that it hardly seems like a real place.  From the transistor to the laser, radio telescopes to programming languages, the amount of innovation that came out of the place puts its modern peer institutions – from other corporate R+D departments to entire universities – to shame.  Even Silicon Valley, our paragon of innovation, is not producing the same volume of truly foundational, far-reaching innovations.  It’s worth wondering why.

I recently read The Idea Factory, a history of Bell Labs, in part to glean a notion of how this place might have achieved such outsized stature.  Certainly Bell Labs was a major employer of bright minds.  But there were no shortage of bright minds at the Harvard physics department either.  Was there some secret to its culture that explained how it employed bright people in such a productive way?  And in the course of my reading, one of the major theories I came up with was that Bell Labs provided its workers with a very rare set of incentives: to be long-term ambitious.

Bell Labs was the research arm of AT+T, a monopoly.  As such, they were in a uniquely lucrative position – and a politically tricky one.  On one hand, the company was able to reinvest its monopoly profits into research in a way that companies in competitive markets simply could not afford.  On the other hand, this privilege rested on its ability to continually persuade regulators that it was operating in the public interest.  One argument they made was that the company was very generous with the free and easy licensing of its inventions, which then benefited the country as a whole.  The example of the transistor is representative: the company licensed it freely to companies for a nominal fee, which led to its widespread adoption and incidentally led to the development of the first true Silicon Valley company, Fairchild Semiconductor.  These sorts of stories came in handy whenever Bell executives had to make the case to Congress that the monopoly should be maintained.

The mandate of Bell Labs, then, was twofold.  On the business side, it was to produce technologies that would increase AT+T’s long-run profit, and on the political side it also had to produce cool technologies that could, in the long run, justify the company’s continued existence.  What this amounted to was a mandate to produce significant results in the long term: you don’t have to hit quarterly targets, but in the long run, you should produce something ambitious.

At the same time, the Bell Labs scientists weren’t completely free to go nuts and do whatever they like, as tenured professors are in theory.  Bell Labs was still a for-profit company, and they were tasked with producing significant, interesting inventions.  If nothing came out of Bell Labs in a few months, that might be okay, but if nothing came out of in five years, it would become pretty obvious that something was going wrong, and heads would start to roll.

That very specific set of incentives – strongly caring about producing marketable inventions not by next quarter or next year, but within the decade – is likely the best possible set of goals for applied science.  On one hand, you’re free to dream big and swing for the fences.  No need to study the trendy topics that will get you grants, or worry about playing games like splitting your results into multiple papers to pad your CV.  On the other hand, you aren’t given quite enough rope to hang yourself.  You’re still expected to be pursuing something commercially valuable, not just wander off into intellectually fun blind alleys or garner prestige for prestige’s sake.  And time and time again, the scientists at Bell Labs used this unique perspective to work on ambitious projects tangentially related to communication, spend years perfecting an invention, and tossing another masterpiece into the world.

One of the best-known talks in science is “You and Your Research” by Richard Hamming, a prominent computer scientist who spent much of his career at Bell Labs.  As well as being valuable advice, the talk illuminates both the scientific culture that animated the place, as well as some of the unspoken assumptions that allowed that culture to exist.  In one telling anecdote, Hamming was persuaded by senior scientists to devote 10% of his time to “Great Thoughts Time”:

“Along those lines at some urging from John Tukey and others, I finally adopted what I called “Great Thoughts Time.” When I went to lunch Friday noon, I would only discuss great thoughts after that. By great thoughts I mean ones like: “What will be the role of computers in all of AT&T?”, “How will computers change science?”… I thought hard about where was my field going, where were the opportunities, and what were the important things to do. Let me go there so there is a chance I can do important things.”

What’s striking here is the assumption – perhaps true in academia and Google but not elsewhere – that unilaterally taking time off to do some private brainstorming that doesn’t check off any to-dos is not only feasible but in fact somewhat supported.  He expected, and faced no pushback, even in the mildest form of feeling an uneasy sense of swimming against the institutional current.  Instead, he’s actively encouraged to do so by his senior scientists!  It’s worth noting, of course, that this is “10% time,” not 20 or even 100% time – the rest of his time was still needed to do his day job.  And he does dedicate a fair amount of his talk to how to deal with resource constraints and how to persuade your boss to let you work on things you prefer.  But just as much of his talk is about how to make the most of an abundance of time – how to train your public speaking skills by giving outside talks, to work with the door open to allow interruptions and serendipity into your work, how not to get distracted playing amusing games with bureaucracy.

It’s instructive to compare the “long-term ambitious” orientation of Bell Labs to the innovative institutions we have today.  Modern research and development largely takes place in non-monopoly companies and are necessarily focused towards filling the pipeline with products that can produce results on a short time-scale.  Entrepreneurship gets a lot of headlines for being innovative, but even there, the amount of runway startups have before investors start demanding measurable traction is fairly short, on the order of one to two years – which precludes a lot of ambitious, long-term projects.  It’s notable that many of the giant technology companies of today, Facebook and Google among them, initially didn’t look like a business at all.  Indeed, they each spent several years as glorified side projects, resisting turning into a typical revenue-driven business.  Amazon, too, is arguably still in deeply in investment-and-building mode rather than trying to cash in and become a mature business.

Academia, the remaining hub of innovation, is arguably both too short-term focused and too unanchored for its own good.  On one hand, the race to getting grant funding and tenured positions is brutal on grad students and junior faculty.  Publish or perish incentivizes working on short-term projects where you can show palpable progress and papers quickly.  On the other hand, in the long run, academia is given enough rope to hang itself.  The judges of science, grant committees and referees, are drawn from scientists themselves; they’re not tethered to an external yardstick the way Bell Labs was, however loosely, tethered to the market.  And it’s therefore entirely possible, despite their best intentions, for entire disciplines, and these powerful committees in particular, to fall into fads and dogmas and become untethered to scientific realities or economic needs, and rank-and-file researchers are forced to follow suit.

If having a well-tuned planning horizon is what made Bell Labs great, then it’s dispiriting that we don’t seem to have institutions with similar planning horizons today.  The vast majority are too short-term focused, and those that are not lack the market discipline that the Bell System brought to bear, however weakly, on the Labs.  Is the answer to return to monopolistic companies?  One could certainly argue that the Bell monopoly more than paid for itself, despite any excess profits it might have made overcharging customers.  However, it is probably for good reason that “monopoly” remains a dirty word today.  In addition to the standard economic arguments against monopolies, most monopolies were not nearly as intellectually productive as Bell Labs.  More accurately, a monopoly seems to set up conditions that are necessary but not sufficient for intellectual productivity.  The most important thing to take away from this is that “long term ambitious” is a very valuable set of incentives, and if we want to replicate the successes of Bell Labs, it’s a culture we’ll have to replicate.