Eight Short Studies On Excuses

The Clumsy Game-Player

You and a partner are playing an Iterated Prisoner's Dilemma. Both of you have publicly pre-committed to the tit-for-tat strategy. By iteration 5, you're going happily along, raking up the bonuses of cooperation, when your partner unexpectedly presses the "defect" button.

"Uh, sorry," says your partner. "My finger slipped."

"I still have to punish you just in case," you say. "I'm going to defect next turn, and we'll see how you like it."

"Well," said your partner, "knowing that, I guess I'll defect next turn too, and we'll both lose out. But hey, it was just a slipped finger. By not trusting me, you're costing us both the benefits of one turn of cooperation."

"True", you respond "but if I don't do it, you'll feel free to defect whenever you feel like it, using the 'finger slipped' excuse."

"How about this?" proposes your partner. "I promise to take extra care that my finger won't slip again. You promise that if my finger does slip again, you will punish me terribly, defecting for a bunch of turns. That way, we trust each other again, and we can still get the benefits of cooperation next turn."

You don't believe that your partner's finger really slipped, not for an instant. But the plan still seems like a good one. You accept the deal, and you continue cooperating until the experimenter ends the game.

After the game, you wonder what went wrong, and whether you could have played better. You decide that there was no better way to deal with your partner's "finger-slip" - after all, the plan you enacted gave you maximum possible utility under the circumstances. But you wish that you'd pre-committed, at the beginning, to saying "and I will punish finger slips equally to deliberate defections, so make sure you're careful."



The Lazy Student

You are a perfectly utilitarian school teacher, who attaches exactly the same weight to others' welfare as to your own. You have to have the reports of all fifty students in your class ready by the time midterm grades go out on January 1st. You don't want to have to work during Christmas vacation, so you set a deadline that all reports must be in by December 15th or you won't grade them and the students will fail the class. Oh, and your class is Economics 101, and as part of a class project all your students have to behave as selfish utility-maximizing agents for the year.

It costs your students 0 utility to turn in the report on time, but they gain +1 utility by turning it in late (they enjoy procrastinating). It costs you 0 utility to grade a report turned in before December 15th, but -30 utility to grade one after December 15th. And students get 0 utility from having their reports graded on time, but get -100 utility from having a report marked incomplete and failing the class.

If you say "There's no penalty for turning in your report after deadline," then the students will procrastinate and turn in their reports late, for a total of +50 utility (1 per student times fifty students). You will have to grade all fifty reports during Christmas break, for a total of - 1500 utility (-30 per report times fifty reports). Total utility is -1450.

So instead you say "If you don't turn in your report on time, I won't grade it." All students calculate the cost of being late, which is +1 utility from procrastinating and -100 from failing the class, and turn in their reports on time. You get all reports graded before Christmas, no students fail the class, and total utility loss is zero. Yay!

Or else - one student comes to you the day after deadline and says "Sorry, I was really tired yesterday, so I really didn't want to come all the way here to hand in my report. I expect you'll grade my report anyway, because I know you to be a perfect utilitarian, and you'd rather take the -30 utility hit to yourself than take the -100 utility hit to me."

You respond "Sorry, but if I let you get away with this, all the other students will turn in their reports late in the summer." She says "Tell you what - our school has procedures for changing a student's previously given grade. If I ever do this again, or if I ever tell anyone else about this, you can change my grade to a fail. Now you know that passing me this one time won't affect anything in the future. It certainly can't affect the past. So you have no reason not to do it." You believe her when she says she'll never tell, but you say "You made this argument because you believed me to be the sort of person who would accept it. In order to prevent other people from making the same argument, I have to be the sort of person who wouldn't accept it. To that end, I'm going to not accept your argument."

The Grieving Student

A second student comes to you and says "Sorry I didn't turn in my report yesterday. My mother died the other day, and I wanted to go to her funeral."

You say "Like all economics professors, I have no soul, and so am unable to sympathize with your loss. Unless you can make an argument that would apply to all rational actors in my position, I can't grant you an extension."

She says "If you did grant this extension, it wouldn't encourage other students to turn in their reports late. The other students would just say 'She got an extension because her mother died'. They know they won't get extensions unless they kill their own mothers, and even economics students aren't that evil. Further, if you don't grant the extension, it won't help you get more reports in on time. Any student would rather attend her mother's funeral than pass a course, so you won't be successfully motivating anyone else to turn in their reports early."

You think for a while, decide she's right, and grant her an extension on her report.

The Sports Fan

A third student comes to you and says "Sorry I didn't turn in my report yesterday. The Bears' big game was on, and as I've told you before, I'm a huge Bears fan. But don't worry! It's very rare that there's a game on this important, and not many students here are sports fans anyway. You'll probably never see a student with this exact excuse again. So in a way, it's not that different from the student here just before me, the one whose mother died."

You respond "It may be true that very few people will be able to say both that they're huge Bears fans, and that there's a big Bears game on the day before the report comes due. But by accepting your excuse, I establish a precedent of accepting excuses that are approximately this good. And there are many other excuses approximately as good as yours. Maybe someone's a big soap opera fan, and the season finale is on the night before the deadline. Maybe someone loves rock music, and there's a big rock concert on. Maybe someone's brother is in town that week. Practically anyone can come up with an excuse as good as yours, so if I accept your late report, I have to accept everyone's.

"The student who was here before you, that's different. We, as a society, already have an ordering in which a family member's funeral is one of the most important things around. By accepting her excuse, I'm establishing a precedent of accepting any excuse approximately that good, but almost no one will ever have an excuse that good. Maybe a few people who are really sick, someone struggling with a divorce or a breakup, that kind of thing. Not the hordes of people who will be coming to me if I give you your exemption."

The Murderous Husband

You are the husband of a wonderful and beautiful lady whom you love very much - and whom you just found in bed with another man. In a rage, you take your hardcover copy of Introduction To Game Theory and knock him over the head with it, killing him instantly (it's a pretty big book).

At the murder trial, you plead to the judge to let you go free. "Society needs to lock up murderers, as a general rule. After all, they are dangerous people who cannot be allowed to walk free. However, I only killed that man because he was having an affair with my wife. In my place, anyone would have done the same. So the crime has no bearing on how likely I am to murder someone else. I'm not a risk to anyone who isn't having an affair with my wife, and after this incident I plan to divorce and live the rest of my days a bachelor. Therefore, you have no need to deter me from future murders, and can safely let me go free."

The judge responds: "You make a convincing argument, and I believe that you will never kill anyone else in the future. However, other people will one day be in the position you were in, where they walk in on their wives having an affair. Society needs to have a credible pre-commitment to punishing them if they succumb to their rage, in order to deter them from murder."

"No," you say, "I understand your reasoning, but it won't work. If you've never walked in on your wife having an affair, you can't possibly understand the rage. No matter how bad the deterrent was, you'd still kill the guy."

"Hm," says the judge. "I'm afraid I just can't believe anyone could ever be quite that irrational. But I see where you're coming from. I'll give you a lighter sentence."

The Bellicose Dictator

You are the dictator of East Examplestan, a banana republic subsisting off its main import, high quality hypothetical scenarios. You've always had it in for your ancestral enemy, West Examplestan, but the UN has made it clear that any country in your region that aggressively invades a neighbor will be severely punished with sanctions and possible enforced "regime change." So you decide to leave the West alone for the time being.

One day, a few West Examplestanis unintentionally wander over your unmarked border while prospecting for new scenario mines. You immediately declare it a "hostile incursion" by "West Examplestani spies", declare war, and take the Western capital in a sneak attack.

The next day, Ban Ki-moon is on the phone, and he sounds angry. "I thought we at the UN had made it perfectly clear that countries can't just invade each other anymore!"

"But didn't you read our propaganda mouthpi...ahem, official newspaper? We didn't just invade. We were responding to Western aggression!"

"Balderdash!" says the Secretary-General. "Those were a couple of lost prospectors, and you know it!"

"Well," you say. "Let's consider your options. The UN needs to make a credible pre-commitment to punish aggressive countries, or everyone will invade their weaker neighbors. And you've got to follow through on your threats, or else the pre-commitment won't be credible anymore. But you don't actually like following through on your threats. Invading rogue states will kill a lot of people on both sides and be politically unpopular, and sanctions will hurt your economy and lead to heart-rending images of children starving. What you'd really like to do is let us off, but in a way that doesn't make other countries think they'll get off too.

"Luckily, we can make a credible story that we were following international law. Sure, it may have been stupid of us to mistake a few prospectors for an invasion, but there's no international law against being stupid. If you dismiss us as simply misled, you don't have to go through the trouble of punishing us, and other countries won't think they can get away with anything.

"Nor do you need to live in fear of us doing something like this again. We've already demonstrated that we won't go to war without a casus belli. If other countries can refrain from giving us one, they have nothing to fear."

Ban Ki-moon doesn't believe your story, but the countries that would bear the economic brunt of the sanctions and regime change decide they believe it just enough to stay uninvolved.

The Peyote-Popping Native

You are the governor of a state with a large Native American population. You have banned all mind-altering drugs, with the honorable exceptions of alcohol, tobacco, caffeine, and several others, because you are a red-blooded American who believes that they would drive teenagers to commit crimes.

A representative of the state Native population comes to you and says: "Our people have used peyote religiously for hundreds of years. During this time, we haven't become addicted or committed any crimes. Please grant us a religious exemption under the First Amendment to continue practicing our ancient rituals." You agree.

A leader of your state's atheist community breaks into your office via the ventilation systems (because seriously, how else is an atheist leader going to get access to a state governor?) and says: "As an atheist, I am offended that you grant exemptions to your anti-peyote law for religious reasons, but not for, say, recreational reasons. This is unfair discrimination in favor of religion. The same is true of laws that say Sikhs can wear turbans in school to show support for God, but my son can't wear a baseball cap in school to show support for the Yankees. Or laws that say Muslims can get time off state jobs to pray five times a day, but I can't get time off my state job for a cigarette break. Or laws that say state functions will include special kosher meals for Jews, but not special pasta meals for people who really like pasta."

You respond "Although my policies may seem to be saying religion is more important than other potential reasons for breaking a rule, one can make a non-religious case justifying them. One important feature of major world religions is that their rituals have been fixed for hundreds of years. Allowing people to break laws for religious reasons makes religious people very happy, but does not weaken the laws. After all, we all know the few areas in which the laws of the major US religions as they are currently practiced conflict with secular law, and none of them are big deals. So the general principle 'I will allow people to break laws if it is necessary to established and well-known religious rituals" is relatively low-risk and makes people happy without threatening the concept of law in general. But the general principle 'I will allow people to break laws for recreational reasons' is very high risk, because it's sufficient justification for almost anyone breaking any law."

"I would love to be able to serve everyone the exact meal they most wanted at state dinners. But if I took your request for pasta because you liked pasta, I would have to follow the general principle of giving everyone the meal they most like, which would be prohibitively expensive. By giving Jews kosher meals, I can satisfy a certain particularly strong preference without being forced to satisfy anyone else's."

The Well-Disguised Atheist

The next day, the atheist leader comes in again. This time, he is wearing a false mustache and sombrero. "I represent the Church of Driving 50 In A 30 Mile Per Hour Zone," he says. "For our members, going at least twenty miles per hour over the speed limit is considered a sacrament. Please grant us a religious exemption to traffic laws."

You decide to play along. "How long has your religion existed, and how many people do you have?" you ask.

"Not very long, and not very many people," he responds.

"I see," you say. "In that case, you're a cult, and not a religion at all. Sorry, we don't deal with cults."

"What, exactly, is the difference between a cult and a religion?"

"The difference is that cults have been formed recently enough, and are small enough, that we are suspicious of them existing for the purpose of taking advantage of the special place we give religion. Granting an exemption for your cult would challenge the credibility of our pre-commitment to punish people who break the law, because it would mean anyone who wants to break a law could just found a cult dedicated to it."

"How can my cult become a real religion that deserves legal benefits?"

"You'd have to become old enough and respectable enough that it becomes implausible that it was created for the purpose of taking advantage of the law."

"That sounds like a lot of work."

"Alternatively, you could try writing awful science fiction novels and hiring a ton of lawyers. I hear that also works these days."

Conclusion

In all these stories, the first party wants to credibly pre-commit to a rule, but also has incentives to forgive other people's deviations from the rule. The second party breaks the rules, but comes up with an excuse for why its infraction should be forgiven.

The first party's response is based not only on whether the person's excuse is believable, not even on whether the person's excuse is morally valid, but on whether the excuse can be accepted without straining the credibility of their previous pre-commitment.

The general principle is that by accepting an excuse, a rule-maker is also committing themselves to accepting all equally good excuses in the future. There are some exceptions - accepting an excuse in private but making sure no one else ever knows, accepting an excuse once with the express condition that you will never accept any other excuses - but to some degree these are devil's bargains, as anyone who can predict you will do this can take advantage of you.

These stories give an idea of excuses different from the one our society likes to think it uses, namely that it accepts only excuses that are true and that reflect well upon the character of the person giving the excuse. I'm not saying that the common idea of excuses doesn't have value - but I think the game theory view also has some truth to it. I also think the game theoretic view can be useful in cases where the common view fails. It can inform cases in law, international diplomacy, and politics where a tool somewhat stronger than the easily-muddled common view is helpful.

Comments

sorted by
magical algorithm
Highlighting new comments since Today at 12:34 PM
Select new highlight date
All comments loaded

What Thomas Schelling would do. Partly tongue-in-cheek.

The Clumsy Game-Player: agree to the deal, then perform an identical "finger slip" several turns later.

The Lazy Student, The Grieving Student, The Sports Fan: make the deadline for reports a curve instead of a cliff. Each day of delay costs some percentage of the grade.

The Murderous Husband: if you really don't want these things to happen, make the wife partially responsible for the murder in such cases, by law. (Or the lover, if the husband chooses to murder the wife.)

The Bellicose Dictator: publicly threaten sanctions unless the invading army withdraws immediately. Do this before any negotiations.

The Peyote-Popping Native, The Well-Disguised Atheist: when the native first comes to you, offer to balance out the permission to smoke peyote with some sanction against the Native American church. Then the atheists won't bother asking for a free lunch.

make the deadline for reports a curve instead of a cliff. Each day of delay costs some percentage of the grade.

We had this system for my second year physics project at university. I hadn't started it when the deadline arrived and decided the penalty rate was too steep to bother starting when the deadline passed. Several weeks later I was summoned to explain why I hadn't handed the project in and I explained that it hadn't seemed worth starting given how little it would be worth by the time I finished it (by this point the penalty had long since reduced the potential grade to ~0). They told me if I completed it before the end of term they wouldn't apply the penalty

I precommit to acting as if I had made any precommittment I find myself wishing I had made. If I make this clear before iterated prisoners' dilemma, a rational partner would not try the "finger slipped" excuse against me, because I would wish that I had precommitted to punishing defection due to finger slipping.

I would still allow the grieving student to turn in the paper late, because in that situation, I do not wish I had precommitted to rejecting that excuse.

Hopefully, of course, everybody involved will intuitively understand what sorts of things you are likely to "wish [you] had precommitted to".

Sounds like a weak precomittment. Schelling includes the theory of excuses in his work, and they are a key part of bargaining, since pre-commitments that can be averted without appearing to weaken the bargainer's position will be.

IOW, once a breach has been made it will be in both parties' interests not to have the threat carried out, and any "wiggle room" in the precommitment will be exploited. Because of this, bargainers are well-advised to make the circumstances that will trigger their threat as unambiguous and externally verifiable as possible.

I don't see any way to do this with your model precommitment, unless the agent(s) you're bargaining with and any third parties observing have access to your source code.

IOW....

If typing an abbreviation saves you less than 10 keystrokes, but increases the time taken to parse your post by at least 30 seconds for at least one reader, it almost certainly isn't socially optimal to use it (although I did get the pleasure of an 'aha' moment when I finally figured out what 'IOW' was supposed to mean).

The Music Fan

A month before the paper was assigned a student was talking with their professor and mentioned that the punk band "The Exemplars" was their favorite band and that they would do anything to see them live but that they almost never perform. The day after the paper was due the student comes in and says "I'm sorry I didn't get the paper in on time, professor, but you won't believe what happened. The Exemplars, that band I told you about held a one-time-only surprise concert downtown yesterday and I just couldn't pass up the opportunity to attend. It is true that my excuse isn't traditionally considered a good one like that of the grieving student, but months ago I signaled to you quite clearly that an Exemplars concert would be the kind of thing that would lead me to turn a paper in late. You excused the grieving student because it is assumed in our culture that a death of a family member will give someone reason not to write a paper. But though we don't assume most people will not write a paper because of a concert you could have predicted an Exemplars concert would cause me not to write the paper. This excuse won't be used often by other students since to be successful one has to indicate a stance that would cause them to not write the paper under certain circumstances and then hope those circumstances arise."

If I were a teacher, I'd ideally allow this (and the sports fan too, for that matter, if I knew he was a big enough fan).

There's no problem in the ideal form, but I would anticipate a lot of problems in real life - other students not understanding exactly how much this person loved the Exemplars and misinterpreting it as "You get an exemption for going to a rock concert", other students trying to convince you they like [thing X] just as much as this guy liked the Exemplars and you having to judge lots of difficult cases to see if their fandom is truly as great as this guy's is and inevitably getting some of them wrong.

If I really knew what to do, this post would have been titled "Eight Short Studies On Excuses, Plus The Answers To Them"

You know, most of these problems can be avoided if you accept assignments late and have most of the course grade depend on one or more large individual projects that must be turned in "sometime before the end of the semester". The simplifying assumption is that a student who doesn't try to learn anything deserves to remain ignorant -- it's nice how neatly this works out.

I give my students a recommended due date, but allow them to turn in any assignment at any time before Finals Week without penalty, for any reason. It works better than you might think.

(I wish that I was allowed to insist that everybody do an individual project. There is so much cheating it's a bit ridiculous, and the worst part is that the students who cheat are just good enough at covering their tracks that I can't find absolutely definitive evidence. This is especially common among Chinese students, for some reason. Again, I figure that if people harm their own learning by cheating on assignments, then their actions are self-punishing.)

In "Predictably Irrational", they mention a study (don't have the book with me, can't be more specific) where a teacher assigns three projects, due at various points throughout the year. The students do relatively well.

In another class the same teacher assigns three projects which can be turned in whenever the students want. The students in this class do quite a bit worse than the first students because they're procrastinating and do all three projects the last week.

In yet another condition, students are allowed to turn in projects at any time during the year, but they're also allowed to voluntarily pre-commit to a due date at the beginning of the year (they fail the project if they don't have it in by their own due date). In this condition, the students who pre-committed did as well as the students in the first class, and the students who didn't pre-commit did as poorly as the students in the second class.

Based on that study, I predict your students would do better if you gave them assigned due dates.

I do give them assigned due dates, actually; I just make it clear to everyone that I'll accept arbitrarily late work without penalty. I make a show of collecting papers on the date they're due. About 80--90% of the students turn in their work on the due date.

Adding to the incentive scheme here is the fact that the classwork builds on earlier work, so if they just procrastinate for a few weeks, they'll be in a pretty dire situation. If they have to go see a funeral or a concert, or they've been really busy some week with exams or a big project for another class, or whatever, then my method gives them enough flexibility for that, but the great majority of people do not procrastinate the way you might predict.

I guess you could call this system "obviously fictitious due dates". Everybody knows that the due dates are made of handwaving and rainbows, but that doesn't make them much less effective. And this way I don't have to listen to any damn excuses, and my students don't have to think of any; I just smile and recommend catching up before next week.

The murderous husband would traditionally have received a reduced sentence in common law jurisdictions, since such a killing would not have been considered murder but rather manslaughter committed in the sudden heat of passion as the result of adequate provocation. A husband’s killing his wife’s lover upon discovering them together is one of the paradigm examples of this sort of manslaughter. (If the husband does not immediately kill the lover, but instead leaves and then kills him at another time, the killing would then be murder because it was not committed in the sudden heat of passion; or if the husband only finds out about the adultery, but does not actually witness it, it would not be considered adequate provocation.)

That is, the law has already pre-committed to the reduced sentence in this peculiar class of cases. At least this is true traditionally; I believe that in at least some jurisdictions, this principle may have eroded somewhat or been changed by statute.

"Mahmud Karim, of the Dadolzai lineage, was enraged to learn that a member of a "brother" lineage, the Kamil Hanzai, had carried off the palm trunks he'd prepared to roof his temporary mud-brick dwelling during the seasonal date harvest. Karim quickly mobilized a war party from among his Dadolzai lineage mates (including a few allies from closely related lineages) to retrieve the trunks--by fighting the Kamil Hanzai, if necessary.

But why hadn't Karim first simply walked over to the Kamil Hanzai and tried to clear the matter up? Indeed, it was later discovered that the palm trunks had been taken by mistake. Why risk a battle without first making a reasonable effort to talk the problem out? That sort of question is liable to be posed by someone living where a state monopolizes the legitimate use of force, and police and courts can therefore be relied upon to keep the peace. In a nonstate setting, where anarchy is kept under control only by the threat or use of force, it often makes sense to send a war party first and ask questions later.

A lone emissary from the Dadolzai making an inquiry or offering to negotiate a settlement would have conveyed an impression of weakness. Only by making publicly known their capacity to swiftly unify and fight to preserve their interests would the Dadolzai prevent future abuse in the lawless desert environment, whatever the intentions of the Kamil Hanzai had been in this particular case. The Dadolzai meant to fight only if blocked from retrieving the palm trunks, yet it was crucial that they be seen as willing to do battle. Had the Kamil Hanzai in fact seized the palm trunks with hostile intent, a lone Dadolzai emissary would have been in serious danger. Only after retrieving the palm trunks unopposed did the Dadolzai send an emissary--not a Dadolzai, but a member of a neutral lineage who would not be at personal risk--to inquire after the Kamil Hanzai's intent. And only then was it discovered that the apparent theft of the palm trunks had been an innocent mistake."

http://www.weeklystandard.com/Content/Public/Articles/000/000/014/947kigpp.asp?nopager=1

Luckily, we can make a credible story that we were following international law...

Credible, to whom? Implausible deniability is an interesting phenomenon. Everyone knows it's a lie, but we all say the lie to each other, and pretend it is not.

It seems more true (lesswrong?) that laws are for those without power, while those with power act according to interest, making up implausible denials when they want to pretend that laws are for everyone. The need for implausible denials places some small pressure on the powerful to act consistently, but not much.

If the powerful don't feel it is in their interests to protect West Examplestan, they won't. East Examplestan will tailor it's action to allow some implausible excuse, those with power will take it, and the lie of law lives on.

For the Peyote example, the powerful don't really care if a limited number of people can take drugs, while the religious certainly do care to protect religious exemptions. If there were only 10 religious people in the country, probably Peyote gets banned and we're done with it.

"Existing a long time" is just an excuse, one of many possible, and again, one that literally privileges vested interests - those who have had the interest for a long time. But if 100 million people are willing to bomb and kill for a week old religion that wanted to use Peyote as a sacrament, "existing a long time" would no longer be a part of the magic formula granting exemption from drug laws.

Follow the money. Follow the guns. Follow the violence.

I find a lot of game theory suspect. The right action will always depend on your prior for the behavior of the other players, and for any prior P that recommends X over Y, you can find a prior P' that recommends Y over X. Treating Game Theory as if it provides "the right" answer, without explicitly identifying the priors on the other players is hugely dangerous. I didn't see any priors established in this article. I saw rationalizations that would hold for some unidentified priors, and not hold for others.

Typos:

These paragraphs

You say "Like all economics professors [....] I can't grant you an extension.

A third student comes to you [....] the one whose mother died.

should end in quotation marks.

In this sentence

"The student who was here before me, that's different.

"me" should be "you".

This paragraph

You respond "Although my policies [....] sufficient justification for almost anyone breaking any law.

should end in a (double) quotation mark and most of the quotation marks in it are nested so should be single.

(Upvoted, by the way.)

You missed the "they enjoys" one. You're fired as official Less Wrong proofreader. (just kidding, thanks for the corrections)