A Post About Fight Club and Sample Bias

Imagine if Jack from the movie Fight Club were to lecture you about auto safety.

In the movie Fight Club, Ed Norton (aka Jack) is an actuary who investigates automobile crash scenes in order to estimate the auto manufacturer’s liability.*  Imagine such a person lecturing you about the dangers of driving. All of his most vivid experiences involve real-life car accidents, some of which involve significant carnage. He could probably go on for a very long time stringing together one anecdote after another. He could justify each rhetorical flourish with another example of a family being horribly maimed or killed. You might get skeptical and say something about how you can’t judge the risk of driving just by looking examples of fatal or near-fatal car crashes. That’s a biased sample, to say the least. Obviously you’d want to start with the full sample of all drivers or all car-trips, and estimate the risk of a bad accident as a proportion of this greater total. But by the time you've managed to express this idea, he shuts you down with another vivid anecdote from just the other day in which a family was burned alive in their "safe" automobile. Driving is safe, indeed!

But of course you would be right to ignore his bluster and consult the actuarial tables to quantify the true risk. Generalizing based on the most vivid possible examples is generally a bad idea. If you know that someone has been gathering and assembling the worst 1% or the worst 0.01% of examples of something (or however far out in the right tail you want to go), they will be a biased source of risk information.

Switch the topic to drug legalization. Enter law enforcement and substance abuse treatment personnel and anyone else who deals with society's problems. "You naive drug law reformer, you simply do not know all the horrors I have seen," they might start. And they go on to regale you with example after example, not realizing that the vast majority of the relevant sample is hidden from their view. Of course, people with substance abuse problems are most likely to attract the attention of law enforcement and medical personnel. If you want to know how to deal with specific bad outcomes, these would be good people to consult for their opinions. (Sometimes, sometimes not...I have heard some of these people show very bad judgment even here in their supposed domain of expertise). But if you are trying to determine optimal drug policy, you should have some sense of how the typical potential user response to the various risks and hazards of drugs. Your sample will contain the millions of people who dabble for a while and never develop a habit, or develop some kind of a "habit" but it never becomes a problem. Beware the problem of sample bias. It is lurking everywhere. A skilled demagogue, or even someone who is honest but oblivious, can really mislead you if you aren't careful.

All this is simply to point out the issue of sample bias. Forget for a moment that many of those overdoses, blood-borne pathogens, and other problems are themselves products of prohibition. That fact will also be hidden from the view of someone who simply looks at the screw-ups and tries to extrapolate from there. The non-problem users in his (the cop/E.R medic/social worker's) own world are hidden from his view; the non-problem users in the counterfactual world of rational drug policy are doubly hidden. These law enforcement or medical professionals could at least see some kind of survey evidence for the existence of the hidden population of non-problematic drug users. But they can't "see" how much better the problem-users fair in a counterfactual world where clean needles are freely available, drugs are cheaper and thus don't require property crimes to support a habit, and drugs are of pharmaceutical grade and known purity (thus leading to fewer poisonings). "Seeing" in this way requires the disciplined use of logic and statistics, and somebody who is blinded by vivid anecdotes won't be able to do it. Since I've discussed these issues in other posts, I won't rehash them all here. I just wanted to point out that sample bias isn't the only problem contributing to these folks' misunderstanding of the issue.

Scott Alexander makes a similar point about sample bias in this excellent post. No, you can't just pile a bunch of horrific anecdotes on top of each other. You have to know something about the base rate. You have to know how large a sample you are digging in to dredge up the bad outcomes.
___________________________________________________________________

* He describes his work in the following excerpt: “Take the number of vehicles in the field, (A), and multiply it by the probable rate of failure, (B), then multiply the result by the average out-of-court settlement, (C). A times B times C equals X. If X is less than the cost of a recall, we don't do one.”
At this point in the movie you're supposed to imagine that the manufacturer is a cynical monster, a hulking avatar of capitalist excess and greed. But actually the formula is about the right criterion for issuing a recall. It might be prudent to make the cutoff "X times 1.5" or "X times 3" or something more conservative than simply "X", but there is no way that the value of X is completely irrelevant to the recall decision. Anyway, that's an argument for another post.

It’s Going to Be Okay

The sun will rise tomorrow. Buildings will not fall. You’ll get up, go to work, go back home to hang with your people, then go to bed again. A lot of stuff will happen that you can’t do anything to stop, but then that’s always true. We tend to put way too much thought into what government does and way too little into how to make our own lives better. Which is ridiculous considering that we have so little control over the former and so much control over the latter.

Do you want these next four years to not suck so bad? It’s not that hard. Be nicer to your spouse, kids, parents, and friends. Do more helpful chores around the house to take the load off of other people. Play with your kids more. Leave each day a little bit better than you found it.

Do the same at work. Hustle a little more. Learn a new skill, even a new career. Consider taking some online classes or doing some kind of professional exams to improve your resume a little. There are probably little things you can do that will pay off a little, and there are big things you can do that will pay off big. Find out what they are and do them.

Are you stretched too thin already? Maybe do the opposite of what I’m saying and give yourself a break. I certainly know people who go to both extremes, that of “too much” and of “too little” effort.

None of this is to say that bad government is irrelevant to the quality of your life. Governments can do horrible things that make your life miserable. My point is that it makes sense to allocate your attention and resources to the things that you have control over. There are things you have little or no control over, and sometimes these things make for interesting hobbies. I myself am a voracious reader on all things political, so if it sounds like I'm making "politics" into a vice then it's a vice I indulge to the hilt. I'm an infovore, and I enjoy the intellectual exercise of policy analysis. But I don't let it distract from the ways I can improve my own life. There are many levers to pull and dials to adjust that affect the quality of your life, and most of them are thoroughly under your control. 

How To Think About Government Regulation

Suppose a customer wants some kind of safety feature or quality assurance on a product they are buying. Say it’s a warning tone in an automobile, or a quality inspection on apples, or whatever else you can imagine. Suppose the customer is willing to pay up to $5 extra to add this feature. If the feature costs less than $5 to produce, the company will obviously add it to the product. If I can spend $2 on added quality and charge the customers an extra $5 per unit, that’s an extra $3 per unit I can pocket. (Or, more likely you’ll charge somewhere between $2 and $5 and share some of the surplus with the consumer. Anyway, re-write the example with whatever numbers you prefer.) On the other hand, if it costs more than $5 to produce the feature, the customers don’t really want it. They aren’t willing to pay the cost of production.

If you think this through, you start to realize that government regulation is gratuitous. If a regulation “forces” a company to produce something that the customer is already willing to pay for, that regulation is completely redundant. A free market will already supply that particular “regulation.” If a regulation forces a company to produce something that the customer is not willing to pay for, it’s actually doing harm. What this regulation is really doing is forcing the consumer to bear the cost of something they don’t actually want. If you impose an extra, say, $10 in production costs on something by adding a feature that the customer only values at $5, you’ve destroyed $5 in value per unit. You’ve harmed the consumer by $5 for each unit purchased.

This ignores “externalities,” or costs to third parties. The consumer might not want to purchase a $20 pollution-mitigating mechanism on their car even though the benefit to society is, say, $50, because the consumer only gets a tiny fraction of that benefit. Regulation of such "externalities" can in theory be justified because everyone wants everyone to have the pollution-mitigating device even though everyone is tempted to cheat by not buying his own. But my above discussion certainly applies to regulation that is meant to benefit the consumer directly. A safety inspection on the meat that I purchase directly benefits me; I don’t benefit third parties by assuring the safety of the food that I eat. There is no “pollution” effect on most of these kinds of safety/quality regulations. 

The above argument also applies to various labor regulations. If the cost of a safety practice (or safety equipment, or a comfortable amenity like climate control) is less than the value the worker places on it, then the employer will buy it for his workers. If your workers each value the additional safety at $100 per week and it only costs you $50 per worker-week to provide the safety, you would have to pay your workers an additional $100 to endure risks that you could mitigate for $50. Even if there is "market power" or something else going on here, there is clearly a deal to make and a surplus to share. (Here is a good treatment at Econlib.)

The point isn't to say that "no regulation is good" or "all regulation is pointless." That is close to my own position, but I won't try to sell that here. The point is rather that you should think more critically about some proposed piece of at-first-blush reasonable regulation. What is the cost? Why aren't consumers already demanding it ("demanding" implicitly meaning "being willing to pay the production costs")? Is there an obvious externality or market failure here? Is there really asymmetric information (in which a sophisticated company bamboozles low-information customers), or are customers actually indifferent to the information and thus ignore readily-obtainable information? Think about it. You might not get all the way to "There should be no government regulation at all." But plainly the stereotype of greedy corporations looking to slash safety and quality costs to earn a quick buck is a mistake. Slashing quality costs a corporation more than the cost savings if you think for a moment about the customer's willingness to pay. (I'm assuming here that actual fraud, selling something other than what you say you are selling, is illegal under any regulatory framework. Just as theft and murder are illegal under any regulatory framework. There are common law doctrines that exist basically everywhere and aren't a function of the regulatory state.)

Reacting to "Outrage Porn"

Emotionally reacting to individual news stories is kind of silly. Some news is useful. But the latest sob story about some horrific crime is not useful and it’s barely even information.

I like the term “outrage porn”, typically in reference to these outrageous news stories about individual crime cases (or sometimes outrageous non-crimes). I imagine this dialogue:

Person 1: Did you realize that a really horrible thing happened this week?

Person 2: Well, I guess I assumed that *something* horrible happened this week, considering that horrible things happen every single day. I don’t see why knowing about a specific instance would meaningfully add to my knowledge.

Person 1: Yeah, but did you see…THIS! (Horrible, outrageous headline.)

Person 2: WHAT! This is an outrage!

Person 1 + Person 2: Rabble rabble rabble!

Sometimes the news meaningfully informs you about your world. As far as I can tell these outrage stories do not. I think some people string together collections of these news stories, each one alone being merely an anecdote, and fool themselves into thinking they are incisively analyzing a real social trend. No, you need some statistics to do that, along with the competence to interpret those statistics. Take off the "moral outrage" hat for a moment and put on the "skeptical analyst" hat. It's rewarding in its own right, and it will help you communicate more meaningfully with people who hold opposing worldviews.

Declining to “Respect Your Opinion”

When is it reasonable to decline to “respect” somebody else’s opinion? I’ve done this before. I have two specific examples in mind, and in both cases I could not bring myself to say or type the words, even though I felt like some sort of olive branch was being extended to me. This is true even though I respect most of the people I know as people. You can be a good person, a loving parent, a hard worker, and I deeply respect those qualities in my friends and acquaintances. I just might not particularly respect their political commentary on topics they might be ignorant or dogmatic about.

I don’t think I’m being completely unreasonable or prickly here. If someone has at least a plausible argument in their defense, I usually respect their opinion. But there is a bar to clear. If someone adopts a position that I find morally repugnant, I don’t respect their opinion (unless it is particularly well argued or somehow interesting, and even then it is the approach and not the conclusion that I respect). If someone utterly fails to bring any argument to bear on the discussion, I probably don’t respect their opinion. If someone just asserts “The earth is the closest planet to the sun” and has no response when presented with contrary information, I feel no obligation to respect that. Some political arguments are literally this terrible. Likewise, if the arguments presented are just bad, and the presenter of these arguments refuses to budge after their total dissection, I don’t feel obligated to respect their opinion. If any of this sounds really snobbish, try imagining yourself saying “I respect your opinion” to a neo Nazi, or an intransigent child making a factually incorrect claim, or a person adamantly claiming “2+2=5, because my hair is a bird.” Everyone has a bar to clear. In fact, depending on how repugnant or how poorly argued your interlocutor’s position is, you may well say, “I don’t respect your opinion, and I don’t particularly respect you either.” 

Learning the Front Handspring

I thought I’d do a post about a physical skill I’ve been learning: the front handspring. I always wanted to learn those cool gymnastic moves, and now at the age of 35 I’m finally doing it. Sort of. It’s been a long, exhausting, frustrating journey but I finally have a serviceable front handspring.

This won’t be a tutorial; there are plenty of good ones out there. Watch a few Youtube tutorials (here is my favorite, but I've watched many others) and read this Crossfit document and this series of three blog post by a gymnastics coach. I had to watch every tutorial I could find about fifty times and re-read the blog posts and Crossfit document about a dozen times to catch all the mistakes I was making. So have your library of tutorial videos and reading materials ready to go.

This post will be more of a motivational and trouble-shooting post, perhaps filling in some holes or emphasizing different problems than the tutorial videos.

First of all, get a yoga ball and maybe some gym mats. If you have a very soft carpeted floor (what I started on), that might work, but it’ll be murder on your knees, hips, and back to do handsprings repeatedly on a hard surface. You can also try it in the grass outside. Position your yoga ball somewhere on the floor, take a hurdle step toward the yoga ball, kick up into the tallest handstand you can manage with your hands placed just in front of the ball, and fall forward over the yoga ball. The ball will serve as your spotter and bounce you up onto your feet. Trust me, you won’t get bounced off in some odd direction. You’ll bounce up in pretty much the direction you were headed. Take a deep breath and try. Then try a few more. You’ll get used to going over yourself, which is an important confidence builder. You'll probably start off flopping onto the ball at the small of your back and bouncing off with most of your weight. This means your arms aren't straight enough. As you progress, your arms will straighten until you're barely grazing the ball on the way over, probably grazing with your shoulder instead of your hips. 

Now record yourself doing a few. Review the tape, then review the tutorials on Youtube. On an iPhone you can drag yourself through the video at any speed you like by sliding your finger at whatever speed and freeze on the frames you need to. Notice everything you’re doing wrong. You’re probably bending your arms too much when you’re upside down with your weight on your hands. This is wrong; your arms are supposed to be straight. You’re probably reaching down to the ground with your arms. This is wrong; your back leg is supposed to kick up and drive your hands to the ground. Your arms and kicking leg are supposed to be in a straight line, as if you were one of those drinking bird toys dipping down for a drink. Your forward leg is supposed to start bent, then spring to give you some lift just as your hands touch the floor. Think about this as your lunge leg is helping to spring you up into a really tall handstand, it’s just that you’re going to go over yourself rather than staying up. You’re probably tucking forward, trying to look for the ground in front of you. This is wrong. You’re supposed to curl your legs under you and arch your back powerfully, like doing a back-bridge in midair. Lie on the ground for a moment. Try tucking your chin forward and doing a back bridge, then try bridging with your head flexed back. It’s a lot harder to arch your back with your chin tucked forward, but in order to land properly you need this arch in your spine. When you go over, you should be looking backwards and down at your hands, not trying to look forward to spot your landing. Then, there’s the “block” with the hands, where you push yourself up off the floor. Make sure you’re doing this at the right time.

I went through this about a thousand times, looking at my terrible front handsprings in slow-motion on my iPhone video and trying to spot my errors. I noticed that I was bending my left arm while I was upside down in a handstand. To correct this error, I had to make sure I was jumping off my lunge leg, which gives you upward momentum. The force of my back-kicking leg was driving me down, so I was forced to bend my arm to absorb the impact. I was indeed tucking my chin on the way up, rather than looking back so that my back could properly arch. As a result, I was landing very hard on my heels with my knees bent and my weight far behind my feet, rather than landing more softly on the balls of my feet. Doing it the wrong way is murder on your joints, so if you have a problem here fix it fast. I repeatedly missed one or another of these steps and had to record myself to see what I was missing. I repeatedly drilled the one part of the movement that was missing, and then promptly forgot another piece. It’s frustrating as hell, because I could do them fine one day then forget how to do it the next. After several days of messing up badly, I went back to doing them over the yoga ball, which would briefly fix my problem before some other problem arose. But I think I’m finally at a point where I’m reasonably proficient in this move. It took about three months of practice, but I got there.

If possible, get a good coach and schedule a few practice sessions under their tutelage. They’ll be able to correct things that you don’t see yourself doing, suggest useful drills that address your deficiencies, and tell you if you’re doing the move safely or not. I really wish I had had a few sessions with a good gymnastics coach, but it was fun trying to learn it on my own. 

Private Drug Prohibition is Adequate. Government Enforced Prohibition Unnecessary.

 I’m not opposed to all forms of drug prohibition. There are some forms that I like and that I believe to be necessary. But in those situations where prohibition is called for, it will virtually always be enforced privately. I’m not talking about some shadowy private security force doing armed SWAT raids on civilian residences looking for drug stashes (think Black Water, the private military company, or Lone Star Security from the game Shadowrun). I’m talking about boring everyday life making drug use an impossibility for most normal people.

Most people spend around half of their day in a restrictive institution of some sort. Most adults go to work. Most children go to school. Even outside of these environments, you may go to a store to do some shopping or go to a restaurant to eat. Many of these environments make drug-induced intoxication unfeasible. You could get drunk and go to work, but you’d most likely be found out and severely punished or, if it’s a repeat offense, almost surely fired. Likewise, if you go to a grocery store rambling incoherently and bothering the other customers, you will most likely get ushered out. Work and school are going to be far less forgiving than the restaurant or the store, but virtually everywhere you go there will be something constraining you from full-on drug intoxication. Even if we assume that workplaces and schools aren’t very good at spotting and punishing drug use, there are intrinsic penalties for these behaviors. Your work will suffer until you eventually get fired or kicked out of school, or at the very least you will not prosper as much as you otherwise could. Even if your boss is oblivious to your high-functioning alcoholism or your methamphetamine habit, if your work suffers your boss will eventually notice that.

Supposing your drug use doesn’t affect your work, there’s no cause for concern anyway as your habit isn’t causing a problem. If someone is very privately using intoxicants in a way that doesn’t affect their work or school, it’s hardly any of their concern. In this sense it’s unreasonable to do drug testing; your boss or school can observe your work product directly and decide whether it’s adequate. With the caveat that some drug use might have no observable effect on day-to-day work output but slightly raise the risk of a catastrophe (think an airline pilot using cocaine), directly observing someone’s work is the sensible solution. Testing for something that isn’t causing an obvious problem is a waste of time.

I can feel your objection welling up. “But…people do drugs anyway and ruin their lives. Clearly private deterrence is inadequate, right?” And to that I say, You’re missing the point. Yes, some people feel the bad consequences of their destructive drug habits. That is how incentives work. The people who behave irresponsibly are hurt by their irresponsibility. Observe that most people don’t have self-destructive drug habits. They see ahead of time that there is an enforcement mechanism in place and decide not to, say, smoke weed before reporting to work, or take heroin before sitting though a class lecture. Perhaps the negative consequences are so obvious that the idea of starting a drug habit doesn’t even occur to most people. It’s not a conscious decision not to smoke a bowl in the morning; the need to get your kids to school and get yourself to work rules that completely out of the question. For most people. And the guy who does try it goes to work smelling of their obvious habit, unable to focus, and gets punished or fired.

Perhaps I haven’t answered the objection fully. The argument is that government enforced prohibition enhances the already-existing (if sometimes tacit) private enforcement of drug prohibition. Does it make sense to go after those remaining screw-ups for whom the implicit penalties for drug use are inadequate? Not really.  Not at all .  It simply does not make sense to try to penalize people out of harming themselves, particularly when the demand for the drug is inelastic (as it must be for any self-destructive drug habit). The intuition here is that I have to harm you more than the drug harms you to get you to stop. Even admitting that a few potential drug users are deterred, the harm to the remaining users is so exacerbated that it doesn’t justify the costs. This is an outcome of very straightforward economic reasoning, requiring no actual grounding in economics or assumption of weird econ theorems or anything. A bit of logic alone will get you to this answer. Unless you posit some very strange, stilted assumption about how drug users respond to legal penalties much more strongly than implicit penalties (of comparable magnitude), you basically have to accept this conclusion: drug prohibition does more harm than good. It’s completely unnecessary. But such deterrence as everyday life naturally provides keeps most of us clean most of the time.

If you’re going to bring up harm to third parties (intoxicated motorists, neglected children, alienated friends, etc.), I answer that point here. In short, the externalities are 1) grossly exaggerated and 2) they are already internalized to the extent that the drug-induced misbehaviors are already illegal. I wish the prohibitionists would take their own arguments a little more seriously. My impression, having thought this through, is that they don’t really have a leg to stand on but they don’t have the patience to think the arguments through to their conclusions.