I'm confused, and suspect the difficulty is in our ethical assumptions, not your math.
You don't say so explicitly, but I presume in this scenario you get *all* the benefit, while your neighbours pay part of the cost. Perhaps you are a part time thief and a full time worker - only 5% of your gains are stolen. Do we allow you to steal, even encourage you, because you do more real work than stealing? Precious few people would agree with this contrived example.
Does this change in your favor if the harm you do is less direct and targeted than theft? Why?
And in particular, consider anything that makes the rich richer and the poor poorer, even while producing an increase in total wealth. Even ignoring diminishing marginal returns (Adding 1 billion to Bezos' wealth won't create as much happiness or utility as e.g. enabling 500 poor people to retire in comfort), this still seems wrong to me in an ethical sense. Those not gaining from Bezos' new-found wealth should not be paying for it. Full stop. Utility to Bezos (or any other especially rich person) is not of any value to me.
If theft were merely a transfer the standard economic efficiency argument would offer no reason to be against it. The argument for preventing theft is again an incentive argument. The opportunity to steal from people results in people spending time and money doing so — casing houses, developing skills at lock picking, ... . As long as you can steal $20 at a cost of $10 worth of time and effort it is worth other people becoming thieves. In equilibrium, the marginal thief is spending $9.99 to steal $10.00, the cost of stealing having risen and the return declined as the number of thieves increases. So the net effect is a cost of about the amount stolen — less to the extent that some people are particularly talented thieves who can steal $15 at a cost of $5, more if you include the cost to potential victims of precautions. This is a standard example of what economists call rent seeking. If you can make money by spending resources transferring rather than producing the transfer is a wash but the resources are a net cost.
I reinvented the idea in my first book, Anne Kruger later invented the label in her article — I think without knowing about Tullock, whose treatment was not only the first but the broadest, or me.
Your other point is a common objection to the usual economic definition of efficiency. I discuss the issue in Chapter 15 of _Price Theory_.
> the CO2 produced in the process was at least as great as would have been produced by burning gasoline instead. We still have the program because, although it did not reduce CO2 emissions, it did raise the price of corn, and farmers vote.
I have two questions, Dr. Friedman, and this seems to be a good (albeit not perfect) post in which to ask them…
The way in which we deal with externalities is one of the many questions that would have to get answered in an anarchocapitalist society or any other such system of voluntary order. When I express my belief that a condition of voluntary order would be preferable, the usual response I get is as you have no doubt come to expect: "How could such a society work?" "Without government, we'd all be clubbing each other for rat meat within five minutes." Etc.
I believe that the best cases for how such a society could work are made in your "The Machinery of Freedom," Hoppe's "Democracy: The God That Failed," and Rothbard's "A Libertarian Manifesto." However, people tend to get irritated when my answer to their question is to suggest that they read three full books! So, my first question is this: Is there a place in which the answer to this question is effectively summarized? A resource that describes how rights-protection agencies might function, for example, or that addresses common objections? I would be happy with a resource that does so in overview fashion, or anything you might have written in summary of your particular arguments on the subject. Anything to which I could direct people's attention without asking that they read 1500 pages of anarchist theory! David Gordon at Mises offered me one good suggestion, which I am examining now. If you have any others, I would be much obliged.
My second question is this: I have been searching for an audiobook version of "The Machinery of Freedom," but thus far, I have only been able to find your recording on YouTube, and as far as I can tell, that only covers the first half or so of the book. Is there anywhere where I can purchase a complete version in audio format?
> Consider a decision where the person making it receives all of the benefit but pays only 95% of the cost, with the remaining 5% falling on someone else. [...] If benefit is more than 100% of cost he does it and should, since doing it benefits both him and us.
I don't understand the last part - how does it benefit us if he receives all the benefit?
I think he means from a utilitarian perspective. If there's a cost of $50 and a benefit of $100, it's $50 more to the economy overall, even if $2.50 of the cost was paid by someone who got none of the profit.
Correct, except that economic efficiency is only an imperfect proxy for utility maximization, for reasons one of which Dinonerd pointed out in her comment.
One is a system where, when actions have $100 benefits and $50 costs (to borrow your example), but $2.50 of those costs are borne by people other than the one taking the action, that person could eliminate the ethical concern by finding the person or people incurring that $2.50 (let's assume they're easy to find) and pay them that exact amount out of his $100. Assuming no friction, the other people are no better or worse than before, and the first person is up by exactly $50 ($100 minus the $2.50 he paid them, and minus the $47.50 that was the 95% of the cost he originally paid).
Another is a system where there are an abundance of such actions that may be taken, all yielding $100 and costing $47.50 to the principal and $2.50 to others, and everyone just takes those actions as often as they please. (Contrived example: suppose there's an effectively unlimited supply of fish in a nearby stream, each fish can be sold for $4 (and never changes - let's suppose demand is inelastic) and requires $2 of effort, 5% of it being making some noise that disturbs people in earshot.) No one bothers to compensate others for the inconvenience, but everyone is better off. Anyone bothered by the continual disruption can quickly fix that by taking the action themselves, and if they don't, it's presumably because they're doing something with an even greater return than that.
The latter system has the upside of not requiring the work of tracking down whoever's incurring that 5% of the cost, although it also relies on that action being available to all. That, or some similar action at least. This seems plausible in the real world, where there exist large numbers of opportunities, many of which come with externalities to unknown parties but are still positive utility.
Ethically speaking, it's still tempting to try to track the exact amounts, just in case someone's chasing down all the "easy" activity with low costs to self relative to cost to others. However, in many cases, these externalities might not be worth the cost of tracking them, as the OP points out. The more obvious the externality, the less expense we might expect to incur in order to know about it.
Would you package externalites in terms of rights? For instance, Business B is harming Business A by competing with it (externality). Buts its not Business A's right to not be harmed from competition. Or take minor inconveniences (loud noise, emitting odors, etc.) imposed by others, where the costs are just expected to be lumped. People have the right to impose minor inconveniences.
And rights would be subject to numerous considerations (who was there first, who is better able to deal with the costs, what does the social calculus look like, etc.). Rather than make everyone pay for all their costs, we'd figure out the rights arrangement first and then determine damages.
Not all externalities should be taxed, only those that the imposing party doesn't have a right to impose.
You are describing two issues, neither of which need to be described in terms of right.
Competition is a pecuniary externality, meaning that it imposes a cost on the competitor but provides a matching benefit to the competitor's customers, so no net externality.
Small externalities are still externalities, but the cost of dealing with them is larger than the cost of leaving them undealt with.
There are a bunch of other complications coming from the ideas in "The Problem of Social Cost," one of the two articles that earned Ronald Coase a Nobel prize. If transaction costs are low enough, the inefficiency can be eliminated by bargaining. If the value to you of playing loud music late at night is less than the value to me of your not doing so, then if you have the right to do it I can bribe you not to. If I have the right to forbid it and the value is higher to you, you can pay me to let you. So one way of looking at the whole subject is that it is a transaction cost problem and you get the right answer however rights are defined, as long as transaction costs are low enough.
I go into that in some detail in Chapter 4 of my _Law's Order_.
I’m a fan of the book. My point was just that if we are to treat externalities as social torts, we should look to whether a duty was violated before going onto damages.
Legal duties are not identical to moral duties, although there is a lot of overlap. Law does not prohibit everything that is immoral, nor should it, and similarly it requires some things that are not based on moral claims beyond that fact that they have become law (e.g. drive on left or right).
But having said that, I think there is a lot to what you said, and something fishy about the economic analysis. I suspect that the topic of externalities becomes a yawn unless we assume, as our host does, that utility is comparable between persons. If I prove that the market underproduces some good, that doesn’t necessarily determine the size of the subsidy that might improve things. Is a penny enough, or should producers receive half of global resources?
If the existing legal/institutional structure produces one thing and not another, we can look at this two ways. Economists and politicians can try to discover the optimum and subsidize or impose it. Or a market or something like it can allow people to experiment with solutions and seek the one they prefer. Sometimes the later is difficult, perhaps impossible. But the former seems certain to generate error, bias, and corruption.
Would a hybrid be better? Can we imagine government providing or encouraging solutions that do not require one size to fit all, allowing different people to do things differently? It sounds good, but I wonder.
Legal duties depend on moral duties for their legitimacy. Even coordination problems rest on moral duties (the decision itself isn’t a moral one, yet obeying it is a moral duty). Sure, not everything moral must be legally enforced, but everything legal must have moral backing for legitimacy. See my substack which is dedicated to this argument.
Markets move resources to where they are most valued, given fixed property rights, as the Coase theorem demonstrates. And property rights determinations are fundamentally moral/legal issues. So the system is fairly hybrid.
I'm confused, and suspect the difficulty is in our ethical assumptions, not your math.
You don't say so explicitly, but I presume in this scenario you get *all* the benefit, while your neighbours pay part of the cost. Perhaps you are a part time thief and a full time worker - only 5% of your gains are stolen. Do we allow you to steal, even encourage you, because you do more real work than stealing? Precious few people would agree with this contrived example.
Does this change in your favor if the harm you do is less direct and targeted than theft? Why?
And in particular, consider anything that makes the rich richer and the poor poorer, even while producing an increase in total wealth. Even ignoring diminishing marginal returns (Adding 1 billion to Bezos' wealth won't create as much happiness or utility as e.g. enabling 500 poor people to retire in comfort), this still seems wrong to me in an ethical sense. Those not gaining from Bezos' new-found wealth should not be paying for it. Full stop. Utility to Bezos (or any other especially rich person) is not of any value to me.
If theft were merely a transfer the standard economic efficiency argument would offer no reason to be against it. The argument for preventing theft is again an incentive argument. The opportunity to steal from people results in people spending time and money doing so — casing houses, developing skills at lock picking, ... . As long as you can steal $20 at a cost of $10 worth of time and effort it is worth other people becoming thieves. In equilibrium, the marginal thief is spending $9.99 to steal $10.00, the cost of stealing having risen and the return declined as the number of thieves increases. So the net effect is a cost of about the amount stolen — less to the extent that some people are particularly talented thieves who can steal $15 at a cost of $5, more if you include the cost to potential victims of precautions. This is a standard example of what economists call rent seeking. If you can make money by spending resources transferring rather than producing the transfer is a wash but the resources are a net cost.
I discuss this in Chapter 20 of my Price Theory.
http://www.daviddfriedman.com/Academic/Price_Theory/PThy_Chapter_20/PThy_Chapter_20.html
Starting at PART 2 -- THE COST OF CRIME.
The original article on the subject was by Gordon Tullock: "The Welfare Cost of Tariffs, Monopoly, and Theft":
https://micros21.classes.ryansafner.com/readings/Tullock-1967.pdf
I reinvented the idea in my first book, Anne Kruger later invented the label in her article — I think without knowing about Tullock, whose treatment was not only the first but the broadest, or me.
Your other point is a common objection to the usual economic definition of efficiency. I discuss the issue in Chapter 15 of _Price Theory_.
http://www.daviddfriedman.com/Academic/Price_Theory/PThy_Chapter_15/PThy_Chap_15.html
The relevant part, beyond the framework at the very beginning, starts at PARETIAN AND MARSHALLIAN EFFICIENCY.
> the CO2 produced in the process was at least as great as would have been produced by burning gasoline instead. We still have the program because, although it did not reduce CO2 emissions, it did raise the price of corn, and farmers vote.
Ain’t that the truth. You gotta laugh really.
I have two questions, Dr. Friedman, and this seems to be a good (albeit not perfect) post in which to ask them…
The way in which we deal with externalities is one of the many questions that would have to get answered in an anarchocapitalist society or any other such system of voluntary order. When I express my belief that a condition of voluntary order would be preferable, the usual response I get is as you have no doubt come to expect: "How could such a society work?" "Without government, we'd all be clubbing each other for rat meat within five minutes." Etc.
I believe that the best cases for how such a society could work are made in your "The Machinery of Freedom," Hoppe's "Democracy: The God That Failed," and Rothbard's "A Libertarian Manifesto." However, people tend to get irritated when my answer to their question is to suggest that they read three full books! So, my first question is this: Is there a place in which the answer to this question is effectively summarized? A resource that describes how rights-protection agencies might function, for example, or that addresses common objections? I would be happy with a resource that does so in overview fashion, or anything you might have written in summary of your particular arguments on the subject. Anything to which I could direct people's attention without asking that they read 1500 pages of anarchist theory! David Gordon at Mises offered me one good suggestion, which I am examining now. If you have any others, I would be much obliged.
My second question is this: I have been searching for an audiobook version of "The Machinery of Freedom," but thus far, I have only been able to find your recording on YouTube, and as far as I can tell, that only covers the first half or so of the book. Is there anywhere where I can purchase a complete version in audio format?
Thanks for all you do.
> Consider a decision where the person making it receives all of the benefit but pays only 95% of the cost, with the remaining 5% falling on someone else. [...] If benefit is more than 100% of cost he does it and should, since doing it benefits both him and us.
I don't understand the last part - how does it benefit us if he receives all the benefit?
"Us" includes him. All of this is summing over everyone affected, including the person taking the action.
I think he means from a utilitarian perspective. If there's a cost of $50 and a benefit of $100, it's $50 more to the economy overall, even if $2.50 of the cost was paid by someone who got none of the profit.
Correct, except that economic efficiency is only an imperfect proxy for utility maximization, for reasons one of which Dinonerd pointed out in her comment.
It helped me to think of it in two possible ways:
One is a system where, when actions have $100 benefits and $50 costs (to borrow your example), but $2.50 of those costs are borne by people other than the one taking the action, that person could eliminate the ethical concern by finding the person or people incurring that $2.50 (let's assume they're easy to find) and pay them that exact amount out of his $100. Assuming no friction, the other people are no better or worse than before, and the first person is up by exactly $50 ($100 minus the $2.50 he paid them, and minus the $47.50 that was the 95% of the cost he originally paid).
Another is a system where there are an abundance of such actions that may be taken, all yielding $100 and costing $47.50 to the principal and $2.50 to others, and everyone just takes those actions as often as they please. (Contrived example: suppose there's an effectively unlimited supply of fish in a nearby stream, each fish can be sold for $4 (and never changes - let's suppose demand is inelastic) and requires $2 of effort, 5% of it being making some noise that disturbs people in earshot.) No one bothers to compensate others for the inconvenience, but everyone is better off. Anyone bothered by the continual disruption can quickly fix that by taking the action themselves, and if they don't, it's presumably because they're doing something with an even greater return than that.
The latter system has the upside of not requiring the work of tracking down whoever's incurring that 5% of the cost, although it also relies on that action being available to all. That, or some similar action at least. This seems plausible in the real world, where there exist large numbers of opportunities, many of which come with externalities to unknown parties but are still positive utility.
Ethically speaking, it's still tempting to try to track the exact amounts, just in case someone's chasing down all the "easy" activity with low costs to self relative to cost to others. However, in many cases, these externalities might not be worth the cost of tracking them, as the OP points out. The more obvious the externality, the less expense we might expect to incur in order to know about it.
Would you package externalites in terms of rights? For instance, Business B is harming Business A by competing with it (externality). Buts its not Business A's right to not be harmed from competition. Or take minor inconveniences (loud noise, emitting odors, etc.) imposed by others, where the costs are just expected to be lumped. People have the right to impose minor inconveniences.
And rights would be subject to numerous considerations (who was there first, who is better able to deal with the costs, what does the social calculus look like, etc.). Rather than make everyone pay for all their costs, we'd figure out the rights arrangement first and then determine damages.
Not all externalities should be taxed, only those that the imposing party doesn't have a right to impose.
You are describing two issues, neither of which need to be described in terms of right.
Competition is a pecuniary externality, meaning that it imposes a cost on the competitor but provides a matching benefit to the competitor's customers, so no net externality.
Small externalities are still externalities, but the cost of dealing with them is larger than the cost of leaving them undealt with.
There are a bunch of other complications coming from the ideas in "The Problem of Social Cost," one of the two articles that earned Ronald Coase a Nobel prize. If transaction costs are low enough, the inefficiency can be eliminated by bargaining. If the value to you of playing loud music late at night is less than the value to me of your not doing so, then if you have the right to do it I can bribe you not to. If I have the right to forbid it and the value is higher to you, you can pay me to let you. So one way of looking at the whole subject is that it is a transaction cost problem and you get the right answer however rights are defined, as long as transaction costs are low enough.
I go into that in some detail in Chapter 4 of my _Law's Order_.
http://www.daviddfriedman.com/Laws_Order_draft/laws_order_ToC.htm
I’m a fan of the book. My point was just that if we are to treat externalities as social torts, we should look to whether a duty was violated before going onto damages.
Legal duties are not identical to moral duties, although there is a lot of overlap. Law does not prohibit everything that is immoral, nor should it, and similarly it requires some things that are not based on moral claims beyond that fact that they have become law (e.g. drive on left or right).
But having said that, I think there is a lot to what you said, and something fishy about the economic analysis. I suspect that the topic of externalities becomes a yawn unless we assume, as our host does, that utility is comparable between persons. If I prove that the market underproduces some good, that doesn’t necessarily determine the size of the subsidy that might improve things. Is a penny enough, or should producers receive half of global resources?
If the existing legal/institutional structure produces one thing and not another, we can look at this two ways. Economists and politicians can try to discover the optimum and subsidize or impose it. Or a market or something like it can allow people to experiment with solutions and seek the one they prefer. Sometimes the later is difficult, perhaps impossible. But the former seems certain to generate error, bias, and corruption.
Would a hybrid be better? Can we imagine government providing or encouraging solutions that do not require one size to fit all, allowing different people to do things differently? It sounds good, but I wonder.
Legal duties depend on moral duties for their legitimacy. Even coordination problems rest on moral duties (the decision itself isn’t a moral one, yet obeying it is a moral duty). Sure, not everything moral must be legally enforced, but everything legal must have moral backing for legitimacy. See my substack which is dedicated to this argument.
Markets move resources to where they are most valued, given fixed property rights, as the Coase theorem demonstrates. And property rights determinations are fundamentally moral/legal issues. So the system is fairly hybrid.