Especially since my Felicifia username is (rightly) tied to my real name, I want to take extra strides to make sure I am absolutely clear: I'm not advocating theft, nor do I personally steal, nor do I plan on ever stealing in the future. Please don't use this essay as a reason to stop trusting me; instead I offer this as a thought experiment only and an opportunity to discuss a controversial point-of-view.

~

However, the idea of being a Robin Hood -- that is, stealing from the rich and giving that money to the poor -- does seem to generate utility. It is well documented from diminishing marginal returns, the hedonic treadmill, positional goods, and other concepts that $10 billion to a billionaire does not generate nearly as much personal utility than if it were donated to a GiveWell top charity or a EAA top charity.

A relevant thought experiment seems to be that of Heinz, made famous for being used in the work of psychologist Lawrence Kohlberg:

A woman was near death from cancer. One drug might save her, a form of radium that a druggist in the same town had recently discovered. The druggist was charging $2,000.00, ten times what the drug cost him to make. The sick woman's husband, Heinz, went to everyone he knew to borrow the money, but he could only get together about half of what it cost. He told the druggist that his wife was dying and asked him to sell it cheaper or let him pay later. But the druggist said "no." The husband got desperate and broke into the man's store to steal the drug for his wife. Should the husband have done that? .... Why do you think so?

Many people, when faced with this dilemma, seem to argue that the husband did so correctly. In my Intro to Psychology class, an unrepresentative sample, people were split 17-3 in favor. However, as we all know, it's worth pointing out that this thought experiment plays out constantly in every day life.

GiveWell argues that $2000 could save 1.25 lives via malaria bed nets (adding 60 years of human life. Brian Tomasik argues that $2000 could prevent 547 years to 102 millennia of nonhuman animal suffering (PDF) via advertising vegetarianism. Others argue that spending money to mitigate existential risk from artificial intelligence could have even higher expected value.

Thus, if you think stealing something worth $2000 is morally okay if it's done to save a single life, the fact that a $2000 donation could save a life or potentially way more than that should be an incredibly analogous scenario, at least from a consequentialist / utilitarian standpoint.

The situation could be made more realistic. Throughout my life, I find myself in many scenarios where people leave laptops unattended. I'm also easily able to shoplift somewhat valuable items. Occasionally I borrow money from people, and they forget that I owe them that money (and I still pay them back because I keep good records). I assume many people living in affluent communities are able to do the same, and skirt traditional ethics for potential utilitarian gain without getting caught.

So why isn't the proper utilitarian approach to steal things, convert them to money, and donate them to effective charities? Surely the payoff is worth it, at least for consequentialists / utilitarians, right?

Is it just that people are afraid of getting caught? Of having the charity be in the awkward position of receiving illict cash? Of the idea that society would collapse if everyone started engaging in these activites? That people are too personally corruptable? That such an idea clearly fails rule utilitarianism? Just general skepticism that charities actually do what these arguments claim?

Peter Singer once elaborated (I don't remember where and Google won't help me) the toughest moral dilemma I know of: you're paying for a good, and the cashier accidentally gives you back $5 more in change than you were owed based on the transaction. Should a proper utilitarian pocket the $5 and donate it, or should you correct the cashier and hand back the $5?

I personally reject this plan of stealing, and instead do what I can to make sure I'm always on top of my debts. The problem is I can't really articulate why, though I suspect its mostly because of rule utilitarianism. What do you think about this dilemma? How would you handle it?

~

Meta-request: Given that my reputation is at stake, would this be the kind of thing I should post to my blog? Or post to LessWrong to get a more popular response? It seems really interesting, but I don't want people to think that the Peter Hurford guy (my real name) is some guy who is right on the edge of stealing everyone's stuff. Also, does anyone have any ideas how I could make this dilemma more clear and compelling? Thanks!

I remember reading an article on LessWrong about morality. We didn't evolve morality to protect others from ourselves. That's not how evolution works. We evolved it to protect ourselves from ourselves. The reason you feel that stealing is wrong is that people who thought stealing was wrong were less likely to get caught stealing, and were more successful as a result.

People aren't afraid of getting caught. Evolutionary psychology doesn't work that way. Rather than evolving a stronger fear of getting caught, we just evolved to not like making the attempt.

It might be that we have higher benefit and/or lower risk. In the example given, it's highly unlikely that the man would be punished with something worse than the death of his wife. That being said, just because it feels like the benefit is worth the cost doesn't mean it is. A lot of people thought it was, and did not pass on their genes.

Additionally, I've read that criminals are less well paid than average.

And utilitarians belong to a reference class of highly intelligent, highly motivated individuals who earn an average of perhaps fivefold more than average.

peterhurford wrote:Especially since my Felicifia username is (rightly) tied to my real name, I want to take extra strides to make sure I am absolutely clear: I'm not advocating theft, nor do I personally steal, nor do I plan on ever stealing in the future.

Then I recommend becoming a critic of taxation as well, even if it is of utilitarian value (?). Or does it matter if stealing is legal or not?

Thus, if you think stealing something worth $2000 is morally okay if it's done to save a single life, the fact that a $2000 donation could save a life or potentially way more than that should be an incredibly analogous scenario, at least from a consequentialist / utilitarian standpoint.

Only if your estimates of the charity's utility are really accurate and your theft doesn't cause additional negative externalities. Stealing unattended laptops causes data loss and temporary unexpected deprivation of laptop use. Fencing costs additional time, the police may be involved, which costs money to society, and so on. My car was stolen once just to extract the radio. The total negative externalities of time and damage to me and others were much higher than the thief's gain.

Is it just that people are afraid of getting caught? Of having the charity be in the awkward position of receiving illict cash?

Yup. Both are good reasons, the former is dominating imo. A lot of people would steal if they didn't get caught, which over time leads to a practical equilibrium where stealing usually isn't worth it, all costs and downsides considered. Stealing can only make sense if you see a systematic distortion in that equilibrium. For instance, if there is a systematic failure in addressing free-riders. In these cases, it can make sense to free-ride if the negative externalities aren't too high. It may even create the positive externality of creating pressure to address the free-riding. An example: in a city I used to live, they were too lax in checking tickets for public transportation. I paid for months and my tickets were never checked even once. In addition, the consequences of being caught were miniscule. Consequently, I started free-riding until I moved away a couple of months later. I never got caught.

Of the idea that society would collapse if everyone started engaging in these activites?

This is wrong on two counts: Your behavior doesn't cause the rest of society to behave in the same way, and if many more people started free-riding or stealing, punishments and controls would increase.

Just general skepticism that charities actually do what these arguments claim?

Yes, you should not be overconfident in the altruistic value of your charitable actions, but estimates are the best you've got. If you would donate honestly earned money, and you can profitably free-ride in some context without negative externalities that are higher than your charity's utility, it makes sense. But I think these conditions are less likely than they appear at first glance. The total amount of expected utility you can shift from this are miniscule.

Peter Singer once elaborated (I don't remember where and Google won't help me) the toughest moral dilemma I know of: you're paying for a good, and the cashier accidentally gives you back $5 more in change than you were owed based on the transaction. Should a proper utilitarian pocket the $5 and donate it, or should you correct the cashier and hand back the $5?

It doesn't matter! If your idea of a tough moral dilemma is about the utility equivalent of $5 or $10, you're doing it wrong. This is not a moral dilemma. "Should I quickly kill a child if it is in agony with little hope of improvement, even if I go to prison and the parents want it to live as long as possible?" is a tough moral dilemma. See the difference? Society adapts to thieves, and incompetent cashiers are fired after a while.

Meta-request: Given that my reputation is at stake, would this be the kind of thing I should post to my blog? Or post to LessWrong to get a more popular response? It seems really interesting, but I don't want people to think that the Peter Hurford guy (my real name) is some guy who is right on the edge of stealing everyone's stuff. Also, does anyone have any ideas how I could make this dilemma more clear and compelling? Thanks!

It's probably a mistake to signal that you would steal, when your real name is associated. Then again, if you correctly conclude ordinary thievery is not worth the reputational risk, you can point that conclusion out with your real name next to it, and it's a positive reputational signal.

The most widely applicable problem of this sort would relate to taxes. Although there's the joke that the income tax has made more liars out of Americans than has golf, compliance is relatively high compared to, say, Greece or Russia. The reason for this (as reported in the media) is that Americans think their neighbors don't lie too much, so there's a social norm against egregious cheating. About one precent of Americans are audited in a given tax year. From what I've heard, Americans don't cheat on their taxes as much as a cost-benefit model would suggest.

Trust is an important component of a well-functioning society. More trust means fewer resources wasted on policemen, lawyers, security systems, and bureaucracy. To answer the question of whether utilitarianism would endorse stealing, you'd need to find out how valuable trust is and how much your dishonest actions would reduce it. You'd also need to assess the likelihood of damage to organizations that you're associated with or to the utilitarian movement in general.

Re the cashier, the stores where I live often advertise sales and update the signs that show the prices, but somehow forget to update the price in their computer systems. They probably end up overcharging people by thousands of dollars a year. I'd gladly swindle them out of some money. But in general, honesty is the best policy, which doesn't quite rhyme, does it.

Consequently, I started free-riding until I moved away a couple of months later.

A genuine case of free-riding! I'd never before fully contemplated the origins of that phrase.

P.S. What happened to all of HedonicTreader's posts?

Yeah, it probably isn't the best thing to discuss these things publicly. Academics can get away with publishing outrageous, pernicious articles, but that's because no one reads them, and of those who do, no one would actually change take such things to heart.

Re taxes, minimize what you owe to the extent allowable by law. You shouldn't feel obligated to report income no one else does, such as income from the paper route or mowing people's lawns. Re stealing, this seems like a bad idea in most circumstances. How much money could you get for a stolen laptop? Probably not a whole lot, when you consider all the transaction costs, risks, and stress.

What about the case of stealing from a billionaire? Well, if utilitarians get a reputation for being the sort of people who would steal money if we had the chance, an "out" utilitarian probably wouldn't be given the opportunity to manage large sums of money. For this reason, it's probably not good to put pro-stealing or pro-cheating arguments online. It may not even be a good idea to believe them in your heart of hearts. It might be better for utilitarians to be perceived as artless and ingenuous rather than conniving and Janus-faced. And the best way to be perceived that way is to be that way.

And utilitarians belong to a reference class of highly intelligent, highly motivated individuals who earn an average of perhaps fivefold more than average.

I have a lot of work to do, then! The median American wage is more than $40,000, so I should be clearing close to a quarter of a million a year. Yikes.

Pat wrote:P.S. What happened to all of HedonicTreader's posts?

I made a new account because of lost password and changed email address.

HedonicTreader wrote:

Pat wrote:P.S. What happened to all of HedonicTreader's posts?

I made a new account because of lost password and changed email address.

Using my admin god powers, I changed the password of the old "HedonicTreader" account and PM'd you the password. I assume you'd like it back?

peterhurford wrote:Using my admin god powers, I changed the password of the old "HedonicTreader" account and PM'd you the password. I assume you'd like it back?

Thanks!

DanielLC wrote:The reason you feel that stealing is wrong is that people who thought stealing was wrong were less likely to get caught stealing, and were more successful as a result.

I recognize my evolutionarily driven intuitions might label stealing as wrong, especially given its the prevailing social view, but I feel like this is largely irrelevant to whether this kind of "Robin Hood" would actually maximize utility.

~

RyanCarey wrote:Additionally, I've read that criminals are less well paid than average.

I don't think that's specifically *because* they are criminals, right? Sounds like a third variable effect, to me. The type of "Robin Hood" I was talking about wasn't being a full-time thief, either.

~

HedonicTreader wrote:I recommend becoming a critic of taxation as well, even if it is of utilitarian value (?). Or does it matter if stealing is legal or not?

I think the legality of it matters. Taxation proves that some sort of systematic, institutionalized, predictable, and in many cases consensual theft can be good for utility.

~

HedonicTreader wrote:Only if your estimates of the charity's utility are really accurate and your theft doesn't cause additional negative externalities. Stealing unattended laptops causes data loss and temporary unexpected deprivation of laptop use. Fencing costs additional time, the police may be involved, which costs money to society, and so on. My car was stolen once just to extract the radio. The total negative externalities of time and damage to me and others were much higher than the thief's gain.

These are all good points. I'd suspect that there are very few, if any, scenarios where I could steal and not get bogged down in negative externalities and personal costs that outweigh the benefit from stealing.

~

HedonicTreader wrote:

peterhurford wrote:Is it just that people are afraid of getting caught? Of having the charity be in the awkward position of receiving illicit cash?

Yup. Both are good reasons, the former is dominating imo

You don't necessarily have to give the charity illicit cash, however. You could use the stolen goods to supplement your personal income, allowing you to give more of your genuine salary to non-profits.

~

HedonicTreader wrote:Stealing can only make sense if you see a systematic distortion in that equilibrium. For instance, if there is a systematic failure in addressing free-riders.

Would you think it's utility maximizing to free-ride in analogous scenarios? How sure would you have to be about there being no negative externalities?

~

HedonicTreader wrote:

peterhurford wrote:Of the idea that society would collapse if everyone started engaging in these activites?

This is wrong on two counts: Your behavior doesn't cause the rest of society to behave in the same way, and if many more people started free-riding or stealing, punishments and controls would increase.

These are good points. I've never really liked much the "what if everyone did that?" kind of arguments anyway...

~

HedonicTreader wrote:It doesn't matter! If your idea of a tough moral dilemma is about the utility equivalent of $5 or $10, you're doing it wrong. This is not a moral dilemma. "Should I quickly kill a child if it is in agony with little hope of improvement, even if I go to prison and the parents want it to live as long as possible?" is a tough moral dilemma. See the difference? Society adapts to thieves, and incompetent cashiers are fired after a while.

I realize that $5 is not large enough to be significant, but I think the general principle still underlies this entire dilemma. You're dramatic retelling, however, does seem more compelling along the same lines, however, and I wouldn't particularly know how to immediately answer.

~

HedonicTreader wrote:It's probably a mistake to signal that you would steal, when your real name is associated. Then again, if you correctly conclude ordinary thievery is not worth the reputational risk, you can point that conclusion out with your real name next to it, and it's a positive reputational signal.

Now that I've correctly concluded I won't steal, I might end up doing publishing that.

~

Pat wrote:Trust is an important component of a well-functioning society. More trust means fewer resources wasted on policemen, lawyers, security systems, and bureaucracy. To answer the question of whether utilitarianism would endorse stealing, you'd need to find out how valuable trust is and how much your dishonest actions would reduce it. You'd also need to assess the likelihood of damage to organizations that you're associated with or to the utilitarian movement in general.

This is a good point along the negative externalities line. I also think people become utilitarians in part because they see the world has been fair to them, but not fair to others, and want to reconcile that fact. Thus, open thievery would decrease the number of utilitarians entirely apart from the repetitional problem.

~

Pat wrote:What about the case of stealing from a billionaire? Well, if utilitarians get a reputation for being the sort of people who would steal money if we had the chance, an "out" utilitarian probably wouldn't be given the opportunity to manage large sums of money.

Also a great point.

~

Pat wrote:For this reason, it's probably not good to put pro-stealing or pro-cheating arguments online.

I can delete this entire thread if you'll think it wise. However, I'm inclined to keep it as lessons learned, though not without concern.

peterhurford wrote:Would you think it's utility maximizing to free-ride in analogous scenarios?

In theory. Otherwise you'd be implying that not free-riding is the most efficient form of charity. Let's say you can pirate a software product with no downsides to yourself, and pay the difference to the best charity (in addition to what you would otherwise have donated). Let's assume your pick of good charities is reasonably good. Should you pirate? I'd say yes, unless you would also do the following: Imagine a different scenario in which all else is equal, but someone else has pirated the product and you have paid the full price. In this case, would you buy another copy as a form of optimal charity? Probably not. If so, you should pirate the copy instead of buying it (unless it is somehow inferior or more effort or risky to pirate).

In practice, I think it's unrealistic to assume most free-riders to donate all the extra money to efficient charities. And the amounts are never high - if they were, society would either find ways to deter the defectors, adapt to different business models, or maybe the negative externalities are so high that not defecting is a more optimal form of charity than giving money to the best charity you know.

It appears my following assumption was wrong:

A lot of people would steal if they didn't get caught, which over time leads to a practical equilibrium where stealing usually isn't worth it, all costs and downsides considered.

In this Blogging Heads discussion, Dan Ariely argues most humans defect less than a pure cost-benefit analysis would predict.

Let's say you can pirate a software product with no downsides to yourself, and pay the difference to the best charity (in addition to what you would otherwise have donated). Let's assume your pick of good charities is reasonably good. Should you pirate? I'd say yes, unless you would also do the following: Imagine a different scenario in which all else is equal, but someone else has pirated the product and you have paid the full price. In this case, would you buy another copy as a form of optimal charity? Probably not. If so, you should pirate the copy instead of buying it (unless it is somehow inferior or more effort or risky to pirate).

Very clever. I think the argument can be generalized to show that stealing is justified in many other cases.

This conclusion could be reached, I believe, also based on considerations of a different sort. It would be quite surprising if social norms and legal institutions, neither of which are the result of a utilitarian optimization process, were such as to make stealing always or almost always unjustified from a utilitarian perspective. Admittedly, one might come up with sophisticated arguments to defend this antecedently implausible harmony between common-sense morality and utilitarianism (as Henry Sidgwick and R. M. Hare, for instance, have done). Such arguments, however, smack of bias, and may be resisted by applying a "powerful debiasing heuristic" proposed by LessWrong user Vladimir M. According to this heuristic, "when a seemingly rational discussion of some deeply problematic and controversial topic sounds pleasant and reassuring, there's probably something fishy going on."

It would be quite surprising if social norms and legal institutions, neither of which are the result of a utilitarian optimization process, were such as to make stealing always or almost always unjustified from a utilitarian perspective.

Maybe you're right. But utilitarian values are in sync with commonsense ones on a lot of issues, and where they're out of sync, it is usually better to try to change things through socially acceptable methods. You could say, isn't it oddly convenient that prominent utilitarians in the animal-welfare movement are opposed to using violence to bring about change. Why aren't they assassinating poultry farmers, or whatever. One reason is that violence is usually counterproductive, and there are socially acceptable ways of bringing about change that are effective.

So it's necessary to say what it is that's fishy. The utilitarian case against stealing is based on

• costs to the thief (opportunity cost [i.e., you could make money legally instead of committing a crime], stress, fines, reduced earnings at future jobs, jail time)
• costs to society (stealing inconveniences people and increases the amount that society has to spend on unproductive activities)
• bad PR for the cause you were trying to advance. For example, the Animal Liberation Front probably has done more harm than good for the animal-welfare movement.
• decreasing people's trust of utilitarians

It seems that some of the time, some people will find themselves in situations where the chances of being caught sufficiently low that it would make sense from an self-interested economic perspective to steal. The question then is how much weight we should give to the remaining concerns.

If you think that economic growth is good, since it produces technologies that could be used to populate the universe with well-off beings (or whatever), you might be concerned about the societal cost of theft. Even if you are, you might think that resources are so inefficiently allocated that stealing could be justified. And some people aren't sure that economic growth is a good thing. So the societal-cost argument probably won't convince everybody.

Then there are the costs to whatever cause you're trying to advance. It seems that if it were to emerge that you had been funneling stolen money into a given charity, that charity's reputation would be damaged. This might not be a big deal if the charity was really small. In this case, there might not be any such thing as bad publicity. But you'd have to consider whether the taint would extend more broadly (e.g., to the animal-welfare movement). This isn't clear to me.

Should we worry that people would distrust utilitarians some of them decided to steal on utilitarian grounds? It seems a bit theoretical. Most people don't even know what utilitarianism is. What's the chance that a potential employer would (a) know that you're a utilitarian, (b) be aware that utilitarians aren't sure that stealing is always wrong, and (c) decide not to hire you because of this. Right now, it's probably close to zero. I'd guess that, empirically, utilitarians are good, trustworthy workers. But if utilitarians were to start stealing whenever they have the chance, lack of trust could become a big problem.

The downsides to software piracy are much diminished. It is a form of theft, but it's somewhat socially acceptable. I think stealing laptops is an entirely different matter.

Interesting piece of fictional evidence: The pilot episode of the crime investigation show The Mentalist featured a utilitarian "cheating for charity gone wrong" murder.

It seems that some of the time, some people will find themselves in situations where the chances of being caught sufficiently low that it would make sense from an self-interested economic perspective to steal. The question then is how much weight we should give to the remaining concerns.

Pat, consider this scenario. You are sitting at a cafe with a laptop that you intend to sell, and use the proceeds to support some cost-effective charity. (This laptop was given to you as a birthday present, and you don't really need it.) Suddenly, you notice a thief running away with the laptop of a person who was ordering a drink at the counter. You recently had an accident, so you can't catch the thief. But you can leave your own laptop at this person's table. Because both laptops are brand new and the same brand and model, the owner will not notice the difference. And since no one noticed the theft either, if you go ahead and leave him your laptop, it will be as if the theft never happened. Should you do it? If the answer is No, as seems eminently plausible, why do folks here still insist that stealing a laptop would be unjustified because of the negative externalities it will create? Why is it impermissible for you to create these externalities, but permissible for you not to prevent these externalities from being created?

Why do folks here still insist that stealing a laptop would be unjustified because of the negative externalities it will create?

I myself don't think that negative externalities are a compelling reason not to steal. But I'll try. Your example is interesting, but it has some problems. First, there are benefits to people's bearing the costs of their actions. If laptop owner returned to his table and saw that his laptop was missing, he probably wouldn't leave valuable items unattended in similar circumstances in the future. What's more, he would probably say, "SOMEBODY STOLE MY LAPTOP!!!" and everybody in the café would realize that they shouldn't leave their laptops unattended. And some of the customers might tell other people about the episode, causing useful information to spread. The owner of the store might decide to install security cameras. It would become harder to steal laptops in the future.

My explanation for why you aren't obligated to give away your laptop is that no one would expect that of you. It's often a good idea to follow social rules, but it's pointless to do way more inefficient good than what's required.

Hi Pat. I agree that there are benefits to people's bearing the costs of their actions. But I don't think this affects my argument. To the degree that such benefits provide me with a reason not to donate my own laptop to the person at the café, they also provide me with a reason to steal unattended laptops at cafes. Similar remarks apply to the other reasons you list: increased social awareness, adoption of security measures, etc. As far as I can tell, the two cases are symmetrical in the relevant respects.

Security measures have a cost and are beneficial only insofar as they are correctly calibrated with the actual risk. A thief doesn't do a business a favor by forcing it to install security cameras. If nobody ever stole anything, the businesses wouldn't need security cameras.

You could come up with a scenario in which the costs of a theft to society (selling a $2,000 laptop for $400, security, etc.) are sufficient to outweigh the utility gained by directing the $400 to a charity. But this would be contingent on the charity's advantage in producing utility relative to the laptop owner and all the others who incur costs because of the theft. I don't believe this assumption holds in every or nearly every case.

Hi Pat,

I'm sorry, but I don't quite understand your reply.

Do you agree that in the case I described one wouldn't be required to donate one's laptop to the person whose laptop was stolen?

Thanks!

Hi Pablo,

Yes, I agree that you wouldn't be required to donate your laptop.

What I've written is confusing because I shifted my position a few posts ago. I realized that the negative externalities of theft could be outweighed by giving to a sufficiently effective charity. So I shifted my reasons for opposing theft to ones that are harder to falsify, namely

• bad PR for the cause you were trying to advance. For example, the Animal Liberation Front probably has done more harm than good for the animal-welfare movement.
• decreasing people's trust of utilitarians

I believe this is referred to as "motivated reasoning." I further muddied the waters by continuing to argue for the negative-externalities view. As I wrote, "I myself don't think that negative externalities are a compelling reason not to steal. But I'll try [to come up with an argument to this effect]."

Sorry for being so confusing. I hope this helps!

A bit off topic, but I must say that I am impressed by seeing such wariness of being biased, together with quick and honest updating.

I think there's a relevant difference between stealing a laptop to sell for donations and refraining from replacing a laptop instead of donating -- in the case of stealing, you are taking on all the externalities of getting caught, having the reputation of utilitarianism tarnished, of getting charity "tainted money", of not being able to find a buyer, etc. In the case of replacing, none of these costs are factored in. So I think you ought refrain from both stealing and replacing. (Not to mention that there are things you could never replace, like the files on the laptop.)

Taking a different approach, I'm brought back to reflecting upon a question asked of me about utilitarianism -- doesn't utilitarianism say we should never have operas or art, because that kind of spending could be better allocated toward alleviating suffering? Generally the answer is yes (and I do think way too much money is spent on art), however, there are billions of dollars spent on things that should be on the chopping block of re-allocation first, such that I think we would end up still being able to spend on art.

Likewise, while we could profit from laptops, there are many significantly easier ways to profit, I think, such that we would end up not stealing stuff.

Pat,

Thanks for the clarification!

Peter,

My argument was not intended to show that stealing a laptop is justified in all circumstances, but only to cast doubt on the claim that we shouldn't steal because of the costs to society that stealing imposes. Hedonic Treader, for instance, wrote that

Stealing unattended laptops causes data loss and temporary unexpected deprivation of laptop use. Fencing costs additional time, the police may be involved, which costs money to society, and so on. My car was stolen once just to extract the radio. The total negative externalities of time and damage to me and others were much higher than the thief's gain.

If my argument works, then this line of reasoning won't succeed in establishing a duty not to steal laptops, since otherwise it would also succeed in establishing a duty to replace stolen laptops (which, ex hypothesi, we don't have).

I agree, however, that there are many other considerations against stealing, and that these are often sufficient to ground a utilitarian duty not to steal.