Fixing healthcare incentives

(This post is about my day job, written at my own discretion. All opinions are my own only.)

About twenty years ago, my father had his life saved by a bowel movement and a bad guess. He noticed some blood in the toilet and thought that it might be a sign of something worse, so he made an appointment to see his doctor. On examination, it turned out that the blood came from an ordinary hemorrhoid, the kind most adults get from time to time. Blood in the stool is common enough that absent other risk factors, a doctor will usually just tell the patient to ignore it.

On his way out the door, my father’s primary care doctor mentioned that they could refer him out for a colonoscopy if he really wanted, even though it wasn’t necessary. But my dad went for it, and during the procedure the gastrointestinal (GI) specialist found an unrelated cancerous polyp. Because the cancer was found so early, my father’s treatment was short and successful, instead of the 15% survival rate he would’ve faced had it been caught later. All because of a lucky hemorrhoid.

maybe you can be lucky too!

How can we systematize this early screening and reduce America’s dependence on well-timed hemorrhoids? First, you’d want to gather some statistics and do a cost-benefit analysis based on who is most at risk, with enough years of life remaining to benefit from early intervention.

It turns out that the optimal policy is that adults from 50-75 years of age should get a colonoscopy every ten years. But armed with that knowledge, we find that we’ve only barely scratched the surface of the problem.

1

Now we face the challenge of rolling this policy out to the 93 million Americans who meet the criteria for this screening. They are cared for by roughly 70,000 primary care physicians, who belong to a truly bewildering variety of hospital systems, insurance networks, and independent medical practices. There is no reporting chain that includes all of these doctors, and therefore no one in charge to give the order.

Perhaps the closest person we will find is Seema Verma, the administrator of the Centers for Medicare and Medicaid Services (CMS), confirmed by the US Senate in 2017. None of these doctors report to anyone who reports to her, but she does set the policies and reimbursement plans for Medicare. Medicare is the federal healthcare insurance program covering 65-and-older Americans, which means that Medicare pays the institutions that pay these doctors.

Suppose you have the ear of Ms. Verma. What would you suggest?

CMS could run some PSAs on the benefits of colonoscopies, but the track record for governmental public health campaigns is not encouraging. A scope up the butt every ten years is a lot to push for in a 30-second TV spot. People trust their personal doctors, not talking heads on the airwaves.

Perhaps Ms. Verma should tell doctors to push their patients to get colonoscopies. But remember: none of these docs report to the CMS administrator. Medicare signs the checks that their practices cash, but those checks are written out for specific procedures that the doctors are billing for, not for following top-down orders. Plus, the primary care physicians (PCPs) that have longterm relationships with these patients are not the GI docs who perform and get paid for the colonoscopies. Get a colonoscopy, and your PCP doesn’t see a dime of that reimbursement. Skip the colonoscopy, and it’s the cancer surgeon, anesthesiologist, and hospital night-shift nurses that are getting paid instead when you need life-saving surgery ten years later.

2

Why this emphasis on money, anyways? After all, patient health is the goal and money should be a distant second. I think that is certainly true for the doctors I know, who are intrinsically motivated to keep their patients healthy and well. That’s why many of them became doctors, and most will go above and beyond for their patients without a second thought.

Still, on the margin, the external incentives do matter. A primary care practice is in many regards a small business, with rent and employees to pay and a very limited number of hours in the day. There is always more work to do, and a directive from a far-off federal administrator is not likely to get a lot of attention if it isn’t associated with any concrete up- or downsides. The closer we can align the practice’s financial incentives with their altruistic ones, the better off everyone will be.

One of the biggest misalignments in our system is the “fee for service” model. This means that practices are paid per procedure they perform. An annual care visit is $100. A colonoscopy is $3,000. Colon surgery costs $30,000, plus a few thousand dollars per night in the hospital. (Prices are approximate, split between patient and insurer, and vary wildly – a topic for another time.)

This means that healthcare clinics are paid more when a patient has more procedures. Completely different medical practices would be performing and charging for an annual care visit, a colonoscopy, or a cancer operation. This usually isn’t any kind of direct payment to the doctor, but you can be sure that some hospital administrator is keeping an eye on how these high-reimbursement procedures are growing year-over-year and what is driving their growth.

Various attempts to fix this misalignment have been underway for more than a decade, under the general umbrella of “value-based care”. Value-based care broadly means that a healthcare system should be rewarded for keeping its patients healthy, not for the number and complexity of procedures its patients get. Value-based care is something that Ms. Verma and her predecessors care about a lot. It’s probably the only way to keep Medicare from collapsing under its growing costs (to say nothing of expanding it to the whole population).

Two of the biggest levers in value-based care today are Medicare Advantage (MA) plans and Accountable Care Organizations (ACOs). A Medicare Advantage plan is an insurance plan offered by a private insurer like United Healthcare or Aetna, run for the benefit of Medicare-eligible patients and paid for by CMS. About a third of Medicare-eligible patients choose MA plans for a variety of different reasons including network coverage, pharmacy plans, or lower costs. CMS compensates these plans a baseline of $10,000 per patient insured per year, more for sicker patients.

This arrangement has some very attractive features from a value-based standpoint and for our nascent pro-colonoscopy campaign. The insurer receives a fixed amount that does not change based on how many medical procedures a patient receives. A patient will hopefully stick with this plan for many years, so preventative care this year means a healthier patient next year and fewer expensive procedures in the future.

An ACO is a similar value-based approach for a group of doctors instead of for an insurance plan. Same deal here: healthier patients and more preventative care equals better business for an ACO, again paid by CMS.

A final important feature of these organizations is that CMS cuts their checks and dictates their terms directly. Did we say $10,000 per patient? We can do a bit better than that, if your MA plan hits a certain target percentage of patients getting those colonoscopies. And if they fall below a certain lower threshold? Well… those were some nice reimbursements that we paid out last year. It’d be a shame if anything happened to them this year.

At last, we’ve found an receptive audience for our colonoscopy messaging. More probes! Less cancer! More preventative care! Fewer surgeries!

3

So here is a corner of the medical system that seems likely to respond to our colonoscopy cheerleading. In the real world, the folks at CMS got here way ahead of us, and are already offering big bonuses to ACOs and MA plans for meeting their targets for colorectal cancer screenings. So what’s left to do?

The devil still lies in the details. Sure, a doctor belongs to an ACO, which theoretically gets paid more by CMS if more patients get screened for colorectal cancer. But that’s assessed once a year, across the whole ACO, which is a collective of hundreds of doctors. So if you do your part but the other docs don’t, no one is seeing any of those extra value-based rewards. And even if everyone hits the threshold, the checks don’t arrive until the next year. Even when the checks arrive, they probably aren’t going to the assistants who are actually calling patients for screenings. Not a very effective feedback loop for motivating behaviorial change.

For a Medicare Advantage plan, the problem is even worse. The relationship between doctors and insurers can be strained sometimes, when claims are rejected and requests for data are ignored. If I’m a doctor, why should I care if your insurance plan is hitting their screening goals for the year? This patient came in with a sore throat and I have to remember to talk to him about getting his butt probed by a different specialist at a different practice? I don’t even get to bill for that!

We’ve finally arrived at the last missing link, the piece that we’ve been working on for the last three years. Recap: CMS is the ultimate bag-holder, the organization charged with keeping all of its patients healthy and the one that gets hit with massive bills if they get sick. CMS is willing to pay extra to risk-sharing ACOs and MA plans if their patients get screenings for early treatment. The last step is to get those incentives to the doctors and medical assistants who can actually talk to patients about the screenings.

So that’s what we are working on now. We have software that identifies eligible patients, and actually pays doctors and medical assistants upfront for making sure that these screenings get done. That software is provided free to the medical providers, and all of it is paid for by the ACOs and MA plans that are eager to spend some money today to make sure their patients are hitting those screening thresholds this year. We do this for colorectal cancer screenings and about a dozen other measures around preventative healthcare and medication adherence. (Our biggest win recently has been helping quarantined patients get their prescription meds through the mail.)

It’s worth drilling into the costs again. For a patient, colorectal cancer surgery is a painful procedure with a long and uncertain recovery. For CMS, it’s $30,000 and a sicker patient to insure in the future. For the ACOs and MA plans, it’s a mark against their orgs and lower reimbursements from CMS.

But if we can convince an overworked medical assistant to stay an extra five minutes to call and convince their patient to go get that early screening, all of that can be avoided. It’s a huge win for everyone if we can pay that assistant an extra $20 to make sure that gets done, especially if we have the software in place to make sure exactly the right patients are getting that call.

Value-based healthcare is all about finding the tiny moments that make the difference. It’s not always glamorous. No CPR on flatlining patients, no flashes of diagnostic genius or heroic last-minute surgeries. It’s routine things like checking glucose levels and making sure patients are getting their prescriptions filled and their screenings on time, keeping them healthy instead of pulling them back from the brink. Our doctors and medical staff do that better when our software makes it easier and more rewarding to do.

My dad had a lucky hemorrhoid that saved his life. We can make sure that no one else has to count on that. If you’d like to work towards that with me, please get in touch!

The self is a leaky abstraction

A few illustrations of the fragmented self.

mr. dictio? i don't feel so good

Most of my enjoyment of a bag of chips comes in the first few bites, and most of the regret comes in the last few. So the obvious thing to do is to open the bag, take a few bites, then close it back up.

When I start to think about doing this, something weird happens. I move the bag closer to my face, my hands pluck out the chips more rapidly, and I actually speed up my eating for a few seconds before I’m able to overcome this impulse and close the bag. It’s as though there’s a part of myself that knows the treats are about to disappear, objects vehemently, and begins to fight back.

The abstract of a Nature Neuroscience paper (Soon, 2008):

There has been a long controversy as to whether subjectively 'free' decisions are determined by brain activity ahead of time. We found that the outcome of a decision can be encoded in brain activity of prefrontal and parietal cortex up to 10 seconds before it enters awareness.

St. Paul, much earlier and rather more dramatically:

I have the desire to do what is good, but I cannot carry it out. For I do not do the good I want to do, but the evil I do not want to do—this I keep on doing. Now if I do what I do not want to do, it is no longer I who do it, but it is sin living in me that does it.

And finally, an AskReddit post titled: “If your future self was trying to murder you, would you resist?”

“I'd fight like hell,” says user rhedrum. “I'm constantly sticking it to future self anyway, why would I take any of his shit if he came to kill me. On the other hand, I would take past self out if I had the chance to go back in time. He has done nothing but screw me over and its about time he paid for it.”

Whether you find these examples funny, frightening, or just a bit weird, you can probably see the common thread that runs through them. Your self is not a coherent whole, pursuing a set of goals continuously throughout time. You fight yourself, forget yourself, change without warning.

But what’s the alternative? It seems so useful to think of selves and people and personalities. How do you even make sense of the world if you don’t exist? The Buddhists say that the self is an illusion, but it’s hard to imagine what you ought to see in its place.

So I’d like to propose a different position. The self is not an illusion, but a leaky abstraction.

what is an abstraction?

Imagine a simple four-function calculator. Cheap plastic, about 15 buttons on the front. You press 1 + 3 =. What do you expect to see? How do you make that prediction?

Perhaps you pictured some little black cells on a pale green backdrop, four in total making the shape of a rudimentary 4. To get there, you did some arithmetic. The whole thing took you less than a second.

Someone with a detail-oriented view of the world and an obsession with electronics could take quite a bit longer. They might think, when I press the 1 key that compresses a rubber membrane and causes a bit of copper to fall into place to complete a printed circuit. This changes the voltage running to one port on the processor, which has a series of logic gates and capacitors that work to store inputs and simulate arithmetic. And a similarly complicated process translates the processor signals back into the liquid crystal display.

You could drill down even further. Electrons move through the copper wires in nondeterministic ways, which are described by certain patterns in quantum mechanical equations…

This is the wrong level of abstraction to answer the question. Missing the forest for the trees, as the saying goes. But it depends! If someone asks you the height of the tallest tree, you have to pop back down a level to see the trees again.

When I ask what you see when you input 1 + 3 into a calculator, the right abstraction for the calculator is something like “this thing does arithmetic, then displays the result on a little green screen.”

So an abstraction is a mental shortcut, an ellison of unnecessary details which helps you reason about the world more effectively.

what is a leaky abstraction?

Suppose you input 1 + 3 =, and the screen shows you a 1. Suddenly you realize that there’s some flaw in your idea that this thing does arithmetic and shows you the right answer. How do you figure out what’s going wrong?

Perhaps the + or 3 buttons are broken. Or maybe it’s low on batteries and that’s causing some kind of problem. Maybe part of the display is broken, so the extra cells that are required to display the 4 aren’t showing up, even though it’s doing the math correctly.

A little part of you wonders, what does it even mean for a calculator to do math? How does that work, anyways?

This is called a leaky abstraction. One of the details beneath the surface, something you abstracted away, is popping its little head back up again. Your model was so useful before, but now you have to remember that there were layers underneath it.

the self is a leaky abstraction

At the beginning of this post, I poked some holes in the idea of the self. The self is fragmented, not unified. Mutable, not constant. There’s plenty of holes there, enough for serious edge cases to slip by. Is it fair to punish people for what they do when they are “not themselves”, overcome by rage or drunkenness or mental degeneration? Should we continue to hold a reformed adult responsible for something that a child did decades ago, just because they share the same body and memories?

What good is this abstraction doing, anyways? Isn’t the self an illusion, a construct dividing the islands of our universal consciousness? Just a piece of wool pulled over our eyes by an evil demiurge?

I don’t think so. The self is an abstraction, but a useful one. I have memories of the things that past-self did, both good and bad. I can learn from the actions of my past selves in a much more visceral way than I can learn from the stories of other people. Future rewards and punishments can shape my present behavior. I’m not completely the same every day, but I stay similar enough to be comprehensible to other people and even loved and trusted by some.

Different pieces of me want different things, but they can work together towards mutual satisfaction. The point of putting away the bag of chips is not to deprive my inner glutton, but to give it more enjoyment given the same costs, and to balance the needs of the glutton-fragment with the other self-shards that want to look good and feel good.

We should try to see the abstraction of self for what it is: a useful tool. Sometimes I need to look beneath the surface, to see why I am at war with myself. At those times I think it’s better to see myself as an abstraction on top of disagreeing fragments, not as a soul possessed by unreasoning sin. Sometimes I need to abstract above myself – not just a self, but part of a group that cares about all of its members.

The self can be an abstraction that helps us guide the pieces that make up the whole. Fragmented, but not shattered. Leaky, but still holding together.

Faithfulness

Doubt is a cherished state of mind for most intellectuals. “Thinking critically” is the first thing we’re taught in school, and the ability to question and update our assumptions is what elevates us above the less enlightened. We think that unexamined belief is easy and doubting things is worthwhile and hard. When we think of faith, if we think of it all, we think of outdated religious claims that haven’t been subjected to enough skepticism.

When in doubt, you might say, add more doubt.

But is this the right model? There’s an alternative meaning of faith that is quite useful, that survives in phrases like “being faithful to your spouse”. Roughly speaking, this means keeping the promises that you make, even after a long time has passed, even when you don’t feel like it.

'til death do us part, or we get bored of each other, whichever one comes first.

I think this is more than a historical coincidence. While it’s true that we don’t often seek out new evidence to change our minds, we do change our minds and our behavior all the time, regardless. Often this happens without any evidence or reasoning process at all. As each day passes, we just feel a little less conviction about our commitments. Is going to the gym today really that important? Will another cigarette really make any difference? Perhaps what we need is just a bit more critical thinking about your original premises?

No! Nothing relevant has changed since you first made that commitment. Except that the strength of your conviction fades day by day with no external input at all, just as it does for everyone else. So faith is the ability to fight that constant, unreasoning doubt that otherwise nibbles everything away. It is the practice of keeping your commitments, both to the belief and to the action.

1

Here are some examples of propositions where it’s useful and difficult to “keep the faith”.

  • Regular exercise will improve my health, mood, and appearance.
  • This startup will definitely succeed, as long as we work hard and don’t give up.
  • My co-conspirator and I commited to cooperating, we will not defect against each other.
  • If I keep talking to girls, some of them will like me.
  • If I hang onto my investments through this downturn, stocks will recover and tend to appreciate over time.
  • If I keep on writing, my writing will improve.

Of course, any of them may or may not be true, depending on the situation. But suppose you’ve convinced yourself (even provisionally) that such a proposition is true, and committed yourself to act accordingly.

One thing that these things all have in common is that we need to make a decision before we can see the outcome. Most of the time, we will have to stay faithful to that decision and continue to act on it for some time before we see any results. At some point we will find ourselves asking whether the process is really working.

It’s a fine question to ask. But often it’s just a standin for the statement “I don’t feel like doing this anymore.” In Steven Pressfield’s book The War of Art, he calls this feeling “resistance,” and it’s the feeling of reluctance or frustration that you’ll get whenever you try to do something creative and difficult. For most of us, we’ll have to push past that feeling before we can ever achieve something we can be proud of.

On the other hand, sometimes a process really isn’t working. How long should you keep at something without seeing any results, before you give up?

One answer is that you should set that limit before you get started. Hopefully this is the time during which you will be the least biased, either by wanting to give in to doubt and cutting the time too short, or by being too stubborn about sunk costs and hanging on for too long. Separate out the planning and execution phases. Planning is when you indulge and explore your doubts. During execution, you faithfully hold to the commitments you made during planning.

In some ways though, this begs the question. How do you set the limit during planning? It depends almost entirely on the situation, but here is one piece of advice that’s been useful to me: do things for just one iteration longer than a reasonable person would. The tendency is to give up too early rather than too late.

(I remember this advice attributed to Paul Graham, but I can’t find any corroboration of that.)

2

I described faithfulness earlier as a kind of corrective to the human tendency to drift and doubt. There are a few other useful hints that fall out of that framework.

When I was running a startup, I noticed that my doubts and anxieties were worst in the morning just after I woke up. We had already decided on a course of action and set a timeframe for how long we would commit to it without expecting to see any results. So this doubt and anxiety wasn’t doing me any good, it was just keeping me in bed for longer.

In a reason-based framework, I might try to stay in bed for longer while I worked out the doubts. But in the context of being faithful to my promises, I just had to remember that I had already considered those doubts during our planning and decided to commit to this course of action anyways. So I made a rule for myself: no thinking about work in the morning at all, until I had gotten on the subway. I noticed that after I had woken up a bit and started taking action, the doubts and anxieties seemed to quiet down.

A second example:
I have noticed that I tend to forget nice things quickly after they happen. At first I thought that this would just drive me to have more nice experiences, instead of resting on my laurels. But this tendency became a problem when my wife started working long shifts for her medical residency. Sometimes I would get the feeling that she never did anything nice for me at all. Did she even appreciate or value me? If she did, why was I feeling this way?

When I really focused my attention on the question, I could remember that of course we’d had many great experiences together, and that in fact she was as warm and attentive as anyone could ask when we had time together. The problem was just that we’d sometimes go weeks without more than an hour or two of real quality time together when her schedule got busy, and in the meantime I would forget all about the great times that we’d had, unless I tried really hard to draw them back into my attention.

So I started to keep a log of nice things that we’d done for each other, and I found that it was much easier to just consult that log when I was feeling down instead of trying to recall those memories unaided. This was incredibly helpful for staying faithful, not just to the narrow sense of sexual monogamy, but also to those other promises that we make: to love and to cherish, in good times and in bad. And I think the key there was just to remember that the tendency is to doubt and forget, and not to place too much stock in those feelings.

A third tip is to surround yourself with people who are trying to stay faithful to the same kinds of things as you. If your friends are all building the same mental habits as you, things go much easier than if they are cynical and doubtful about the things you’re trying to hang on to.

3

Finally, of course, it helps to have faith in the right things. I think we’ll leave that outside the scope of this post though.

Faith is not a process for choosing what to believe in or what to commit to. It is a tool for helping with what comes next. Creeping doubt and regression to status quo is natural, even in the absence of any new evidence. Faith is the virtue of counteracting that regression, through repeated internalization of those truths and commitments.

I hope you can find a few things that are worthy of your faith, and that having found them, you will be faithful.

Group Coordination at Raves

Think about the last time you tried to get eight people to do the same thing. Now imagine that four of them are on drugs, two have to use the bathroom, you’re all in a crowded arena filled with a hundred thousand (100,000) other people, and none of your cell phones work.

Welcome to group coordination at raves.

Can you find your friends in this picture? Me either.

1

I exaggerated slightly. It’s far too loud to hold a phone conversation, but you can infrequently send and receive text messages. However, the network is overwhelmed by the hundred thousand other people who are also texting their friends, so your messages will be delayed by ten to thirty minutes and might fail outright.

Here are some common scenarios you might find unexpectedly difficult to solve.

  • You and some friends have a great view of the stage, but you can’t find the rest of your group. You’re willing to abandon your spot but only if you’re confident you can meet up with them.

  • You find yourself unexpectedly alone. You’re willing to do whatever it takes to find your friends, but you aren’t sure how to coordinate without reliable text service.

  • You are waiting for a friend at a pre-arranged meeting spot but they aren’t showing up. You were 5 minutes late for your agreed-upon time, so you aren’t sure whether they showed up and left already, or they are still on their way, or they decided not to show up at all.

A normal person in one of these situations might send a text like, “I’m near the Circuit stage, want to meet up?”

That message would be a complete failure. One of your friends might see the message 30 minutes later, and at that point they will have no idea if you are still there or even precisely when you sent the text. They could text back to confirm, but then you will face the same issues when you get their message.

Here is a better message: “Can we meet under the giant daisy at 2AM? Currently 1:15, I’m with Jess and Sam. Will wait there for 10 min.”

The important components here are a timestamp, a rendezvous time far enough in the future that your message will likely be read, and a plan that you can unilaterally commit to.

When your friend gets this message, they will have much more information to work with. If it’s past 2:10 when they see it, they will know not to bother. If they get the message in time, they can begin heading to the spot without waiting for more round-trip texts. Also, they will know how much they want to join your group. For some people, that answer will depend on who else is with you, and also on how much fun they are having at that exact moment.

2

Let’s take another look at scenario 3. You’re sitting under a large daisy sculpture, wondering if your lost friend is going to show up at all or whether you already missed them. You’re getting cold and feeling impatient to get back to the lights and music. How long should you wait there? How long should you have asked your friend to wait?

Of course the immediate answer is that you should wait as long as you committed to waiting. A surprising number of people won’t do this, and it’s always disappointing to see. But how long should you commit to waiting in the first place?

When I first started going to raves, I often committed to 10 minutes and waited for up to 20. It seemed like a classic (iterated) prisoner’s dilemma, and I’d always want to cooperate with my friends and seek the greater good of the group. I thought that people who weren’t making the effort were being selfish and short-sighted. There’s really nothing worse than being alone at one of these events if you don’t know anyone else – something about the whole experience brings out this deep need for connecting with other humans at the same time that it makes it very difficult to even stay in proximity with them.

Over the years, I’ve changed my mind on this. Different people want different things – some just want the lights and music, some want to be part of something bigger, and some have one or two friends that they really care about and are indifferent to the larger group. If you make them commit 20 minutes of their night to the cohesiveness of a group they just don’t care that much about, you’re setting yourself up for disappointment.

There are also these little perfect moments where the lights are just right, you’re feeling the beat in your chest and in your feet and in the eyes of the strangers around you and the DJ on the stage, and these fleeting moments are impossible to predict. If I knew that one of my friends was having that perfect moment, I wouldn’t want to drag them away from it. I value people who keep their commitments, but I also want to build in the flexibility to enjoy the magic when it happens.

My updated answer is that you should pick one or two buddies that you absolutely commit to sticking with. If they go to the bathroom, you go with them or else you stay rooted to the spot where they left you for an hour if you have to. For the rest of the group, set some meetup times at the beginning of the night, give it five minutes or ten at the most, and make it explicit that people can bail if they want to.

Be honest with yourself about whether anyone in the group wants to be that buddy for you. If not, you’ll have to make some compromises or be ready to go it alone. I’ve sometimes paid for friends’ tickets if they were on the fence, just to have someone reliable on hand.

3

Coordination is hard at the best of times, and breakdown is inevitable at events like these. I once took a taxi 40 minutes out into the desert at 6 AM to rescue friends who had skipped out on their planned ride back and gotten stranded at the racetrack, only to discover that they also skipped out on me. (Still love you guys!)

Everyone is out there seeking their own bliss, that elusive perfect trick of light and sound and drugs too if we’re being honest, and no two people ever want or get exactly the same night. Good intentions and promises often aren’t enough when the lasers flash, the beat drops, the roll peaks. Sometimes it’s all too enticing and it’s just too easy to get separated in the night and the massive pressing crowd.

But there are other perfect moments that I remember too, when people do follow through on their promises and come through for their friends. They miss out on their favorite DJs to take care of people who need them, wrestle themselves down from serotonin highs long enough to reunite for the one set we promised we’d do together. What makes those moments possible? Sometimes it’s the strength of friendships, sometimes it’s dumb luck. Sometimes, I think, it’s the grace and synchronicity and unity of all things, the hidden oneness that pulses underneath and alongside, always.

And sometimes, it’s because you sent the right text.

Struggling in Dance Class

I recently started going to a walk-in Beginner House Dance class on Monday nights. It’s been a struggle!

The first thing to know about walk-in beginner classes is that the other attendees are never actually beginners. The average student in the class, learning the choreography for the first time, looks substantially better than the second-best dancer on a club dance floor. The first walk-in I ever attended was a beginner hip-hop class with my friend Cyndie, who swore up and down that she’d never taken one before. It turned out that she had instead choreographed one in college.

As for myself, I’m not a great dancer but I like to try. I once got a high-five from a group of cool-looking black teens while dancing at Gov Ball, probably more for enthusiasm than technique. I have a basic two-step and I’m comfortable dancing while sober, which I think puts me above average in the general pop and firmly at the bottom of every walk-in class I’ve been to.

The bottom of a class is an uncomfortable place to be, especially in a setting where the other students can watch your every misstep. During each class our teacher tries to run us through a couple of basic moves, then adds on variations. She shows one move slowly and repeats it four times, asking us to follow. By the third repetition I can just about make out what her feet are doing. By the fourth I start to reason about which leg goes where and which foot supports my weight, while the other students are doing pretty close copies. Then: double-time! I have no chance, of course. Four more reps and it’s on to double-time, and also turn in circles!

I’ve always been a good student in school so it’s a novel feeling to see the pure confusion in our teacher’s eyes as she observes me. For the fourth time, it’s left-left-right! Is he spacing out or slightly brain-damaged? I imagine the other students side-eyeing my clumsy steps, wondering why the class can’t collectively move on from this basic routine yet. I think of each class as a battle between my frustration and competence levels. When the competence meter fills, I learn a move that’s eluded me before and gain confidence to continue. When frustration overflows, I shut down, make more mistakes, and struggle to take on feedback.

At the lowest points I start to worry that this will hinder the enjoyment and fluency of dancing for me outside of class too. This seems to be a general pattern with consciously practicing to improve tacit skills. There’s a period of time where I regress in performance because I’m consciously directing a behavior which is normally managed subconsciously. Then this degrades self-confidence, which further reduces performance.

So why haven’t I quit yet?

First, I’ve been trying to get more comfortable doing things that I’m bad at. Being bad at something is the first step to getting good at it. Plus, I’ll miss out on a lot of fun things in life if I only do things that I’m innately good at. Growing up I avoided most sports and I never put much effort into music because it didn’t feel like I was good at those things. But my friends who put in the time and struggled through being bad now have more hobbies that they can enjoy.

Second, I enjoy dancing and I go to a number of events where that’s the thing to do: weddings, concerts, raves, clubs, etc. I get bored of doing things unless I see some path to getting better, and I didn’t feel like I was really improving as a dancer despite the hours that I’ve spent dancing. The timing on these classes is a little unfortunate, since I expect that the majority of my dancing is behind me – economically speaking, it would have been better to invest into the skills earlier so that they would have more time to pay off later. But even though the area under the curve will be small, I still hope to enjoy the process of improving.

Against the Pressure Theory of Motivation

When we were growing up, one of my dad’s favorite phrases was: “if they’re hungry enough, they will eat.” This was meant literally, meaning that my mom needn’t worry so much about making every meal delicious – we might refuse to eat it at first, but as long as we didn’t eat anything else, eventually we would be forced by our hunger to eat. My mom never listened to this advice, and I grew up eating delicious, fussed-over food.

He had a similar theory of motivation, roughly “if you aren’t acting towards X, it’s because you don’t want it enough.”

I think of this as a pressure differential theory of motivation. Imagine two rooms A and B separated by a barrier, where room A is initially filled with a colored gas and B is not. If the barrier is very leaky, gas will naturally diffuse from A to B. If the barrier is sturdier, the gas might stay contained in room A. However, if you continually pump A full of more and more gas, eventually it will leak into B somehow, either by forcing itself through small leaks, or eventually by blowing out the wall between them.

Therefore, if you want to move the gas into room B, it’s not necessary to understand the strength or build of the wall between them. All you need to do is increase the pressure enough.

(For this metaphor, the containing walls around the system are infinitely strong – although you could come up with allegorical extensions for ‘blowing out the outside wall’ etc.)

2

I have not found this advice to be very practical in my life. The most common cause of distraction seems to be some low-level anxiety. One common thing that will happen is that I’ll find myself looping between Reddit, Discord, Twitter, checking each website three times in five minutes. Obviously I know that not enough time has passed for anything to have changed. What’s really happening is that I have a bit of anxiety about the actual task at hand, so I try to self-soothe by procrastinating. But procrastinating is actually making the anxiety worse, so I then try to self-soothe with something slightly different, and so on.

The pressure theory of motivation does not provide any solutions for this loop. It’s tempting sometimes to up the pressure with negative self-talk. Stuff like “you’ll never succeed if you keep on like this”, “this is why you never get anywhere”, etc. But I am already too anxious, so more pressure only amps up the anxiety, encourages more maladaptive procrastination, and deepens the loop.

If we want to extend the pressure metaphor, we can think of a special valve that only operates within a certain range of pressure difference – too high and it shuts itself off, like a Chinese finger trap snapping shut when you pull your fingers apart too fast.

There’s a different failure mode of genuinely low motivation that probably can be fixed with some self-applied pressure, but that hasn’t usually been my problem. If anything, just trying to avoid boredom is enough to kick me out of that mode.

The point is, you can’t just blindly increase the pressure to get more progress. The pressure has to be in just the right zone, not too high and not too low.

It could be that this problem specific to negative self-talk. There are other ways to ramp up pressure that may be more adaptive. For example, you could make an agreement with a friend that if you don’t accomplish a goal by a certain time, you must donate $100 to a charity, perhaps even one that is run by people who you don’t like very much. I haven’t explored these avenues very much though, because I expect them to have similar problems.

3

The final problem with the pressure model of motivation is that it doesn’t distinguish between short-term motivation and long-term goals. Personally I’ve found that long-term goals have relatively little immediate effect on getting me to do something in the present. (Yeah, I do want to have a popular blog, but I also want to hit the snooze button. Which one do you think wins?)

Focusing purely on short-term motivation, I’ve found that there are three components that almost always play a major role. Actual aversion to the task, the displeasure of starting costs, and the pleasure of flow state. Each of these has an associated trick.

To get over actual aversion to the task, I try to generate a sense of compassion towards my future self. This task will be done eventually, and I might as well eat the disutility with my present self, instead of inflicting it on a future self who might be less able to afford it.

To get over the displeasure of starting costs, I just try doing the task for five minutes. This might ease me into the flow state. Once I get into a flow state, it becomes naturally pleasurable to continue working on the task.

To remain in flow for as long as possible, I separate out the brainstorming, planning, and execution. Each of these has their own distinct flow, and moving from one to another breaks flow.

4

Overall, I’ve found it more helpful to focus at a detailed level on how to get things flowing on a moment-by-moment level. This is more difficult because each task has a slightly different shape in my mind and requires different tactics to get myself moving on it.

In contrast, the pressure theory of motivation is one-size-fits-all – if you aren’t working hard enough, just yell at yourself more! But the more detailed and flow-based way of getting things done is the way that usually works for me.

Review of 'Tribe' by Sebastian Junger

I recently read the book “Tribe” by Sebastian Junger. Here’s the blurb:

We have a strong instinct to belong to small groups defined by clear purpose and understanding – “tribes.” This tribal connection has been largely lost in modern society, but regaining it may be the key to our psychological survival.

Readers will probably be familiar with the overarching narrative – modern society is too atomized, individuals don’t have a shared sense of meaning or reciprocal bonds of dependency. Junger’s main contribution is to look at some societies that seem to provide more of these tribal bonds, especially with regards to their war veterans.

Junger opens the book with an interesting claim: in the colonial era, European settlers often ran away to join Native American tribes, but the reverse almost never happened. This happened often enough that people like Ben Franklin wrote worried letters about it. Why?

Junger claims that Native American society was more attractive because it has a more egalitarian setup. The higher-status members of the tribe only consumed slightly more resources than lower-status members. Most primates have much more egalitarian societies than we do, which suggests we are not evolved to thrive with modern levels of inequality. There was less top-down authority and more self-determination. Coming-of-age initiation rites were common for all males and provided a shared sense of meaning.

Some other interesting, not-super-relevant nuggets: infant care was a shared responsibility in Native American tribal society. There was almost always an adult at hand to pick up and soothe a crying baby. War leaders were chosen separately than peacetime leaders, and it was the peacetime leaders that had the responsibility of negotiating terms of victory or surrender.

2

So Native American tribes provided more of a sense of belonging. What about modern society? Can we ever replicate that sense of tribal community?

Good news: yes, we can. Bad news: we only seem to be capable of doing that in response to a crisis.

Junger describes a few examples including London during the Blitz and Bosnia during the Bosnian War. During the Blitz, Britain had prepared for a big surge in people seeking psychological help and for unrest in the bomb shelters. But they found the opposite – suicides and reported psychiatric disorders actually decreased! People cooperated in maintaining order in the shelters and stepped forward to share their resources. Inequality tends to decrease during times of crisis, and the crisis creates a shared “us vs. them” narrative for the community.

Similar effects were reported in German civilians during the Allied firebombing campaign, and during civil wars in Spain, Algeria, Lebanon, and Northern Ireland. The cities with the highest morale were the ones that experienced the most of the fighting; the untouched ones seemed to suffer from a sense of helplessness.

In fact, this effect is so strong that people report missing the crisis after it’s over. “It was better when it was really bad,” reads one piece of graffiti in Bosnia.

A third part of the book talks about the effects of PTSD, differentiating an acute trauma response from a chronic response. Junger claims that some rate of acute trauma response is expected and constant across societies, but our atomized, less tribal modern societies show much higher long-term PTSD rates when compared to others. I was less personally interested in this portion, so I won’t go further into it.

Actually, quite a lot of the book was focused on the experience of being a soldier in a war, both the positives and the negatives. I only noticed afterwards that the word “Tribe” on the cover has a camouflage background. Someone with more military connection might get more out of it.

3

Overall I was a bit disappointed with this book. I’ve been floundering about for awhile trying to figure out what exactly it is that I want from a community and how to build it, and by first appearances I had been hoping that this book would provide more of that.

The book implies two paths for proactively building a tighter-knit community, although it doesn’t explore either of them. These are egalitarianism and shared crisis. All of the crises brought up in the book are wars or large-scale natural disasters; only a true psychopath would consider triggering something like that to build a sense of tribal community, and anyways, the community in that case only seems to last slightly longer than the crisis.

Egalitarianism as an ideal may be more achievable, although Junger doesn’t talk much about modern attempts to replicate it. It’s important to note that here he means equality not just of resources but also of power. What separates the colonists from the Native Americans was not so much the distribution of wealth but rather the structure of society. One white woman who joined a native tribe is quoted as saying “Here, I have no master, I am the equal of all the women in the tribe. I do what I please without anyone’s saying anything about it.”

There’s an interesting tension here between this and the idea that armies in war also form tribal bonds. Isn’t an army about the most hierarchical structure you could imagine? Unfortunately this isn’t explored at all in the book.

Another tension that occurs to me now: one problem that communitarian societies face is the existence of freeloaders. It becomes important to police the boundaries of your community, since you don’t want outside freeloaders to approach in times of need and then walk out. But a structure to enforce this kind of boundary seems like it could become hierarchical in itself.

Anyways, Tribe was an interesting read, just not what I had hoped for. I suspect it also suffers from the standard pop-sociology flaw of cherry-picked data, but I don’t know the field well enough to point to the omissions. It is very short though, at just 192 pages, and I read it in just two sittings. I’m very interested in books in this vein, so I’m happy to have read this one, and eagerly look forward to any related recommendations any of you might have.