• A. One midterm, one final, no papers, no homework.
• B. Three sources of written information.
  • 1. Posner text.
    • a. Nearest thing available to a survey of law and economics
    • b. Much too much stuff for a one quarter text.
    • c. But useful both as a supplement to the lectures and a source of additional material for those who are interested.
  • 2. Lecture notes: Will be distributed weekly in hardcopy and put on the web page (with links to other relevant material on my web site).
  • 3. Articles: Copycraft. Some are also on my web page, linked to the class page.

• A. Two views of a traffic ticket:
  • 1. A way of transferring money from drivers to governments--a rather peculiar tax.
  • 2. A way of giving people an incentive not to drive too fast.
• B. More generally, think of law as a system of incentives.
• C. Economics is about the implications of individual rationality
  • 1. Do we believe that criminals are rational?
    • a. Muggers pick on little old ladies, not football players, which suggests they are trying to get money at the lowest cost practical.
    • b. Burglars tend to avoid burgling houses with people in them in areas where people frequently own guns.
    • c. Statistical evidence suggests that raising the expected punishment for an offense tends to reduce the number of offenses.
  • 2. How about the rest of us--drivers, potential victims, litigators, ... . Econ analysis of law is about us too.
• D. Hence about how rational individuals will respond to the incentives created by legal rules.

• A. Using economics to deduce the consequences of legal rules.
  • 1. Marginal deterrence example:
    • a. Suppose we raise the penalty for armed robbery to the highest level our legal system permits.
    • b. A legal philosopher would ask "is this just?"
    • c. An economist would ask "do we want armed robbers to routinely kill the witnesses?" since ...
    • d. One result of the change is to eliminate the penalty for murder--if you are already committing an armed robbery.
  • 2. Plea bargaining example:
    • a. A criminal will only accept a plea bargain if he prefers it to the punishment lottery (some probability of some punishment) that he faces if he goes to trial, so ...
    • b. It looks as though plea bargaining reduces average punishment, but ...
    • c. Not if you allow for the fact that a prosecutor who settles many of his cases by plea bargaining has more resources left to prosecute the cases that don't settle, which raises the probability of conviction, which makes people willing to accept harsher plea bargains, so ...
    • d. Permitting plea bargaining might increase average punishment.
  • 3. Freedom of contract example:
    • a. Fixing one term of a contract, for instance requiring landlords to give tenants six months notice before evicting them, looks as though it benefits one party (tenants) at the cost of another (landlords), but ...
    • b. If you allow for the effect that fixing that term has on the other terms (market rents rise in the example), the net effect may well be to make both sides of the contract worse off,
    • c. By forcing them to include inefficient terms in their contract--which the "benefitted" party ends up paying more for than they are worth.
    • d. This suggests a general presumption in favor of the principle of freedom of contract--let the courts enforce contracts as written, rather than deciding what terms the parties ought to have agreed to.
• B. Using economics to predict and understand rules.
  • 1. The Posner Thesis:
    • a. Posner conjectures that common law (judge made) legal rules tend to be economically efficient.
    • b. Testing the conjecture has played a major role in the development of economic analysis of law.
  • 2. Special interest politics is an alternative theory for explaining what laws, especially laws made by legislatures, are like.
• C. Using economics to choose legal rules.
  • 1. Designing efficient rules.
  • 2. Is that worth doing?
  • 3. Part of the attraction of the economic analysis of law is that you get out more than you put in.
    • a. We start by assuming that values to thieves and murderers count just like values to anyone else.
    • b. We conclude that theft and murder ought to be illegal--because even if you count the benefit to the criminal, permitting crimes still turns out to be inefficient.
• D. Part of why it is exciting is that it may provide a theoretical core for law.
  • 1. Traditional legal theory isn't a theory--it is the theory of property law, the theory of tort law, the ..., not any unifying theory that makes sense of the whole thing.
  • 2. Economic analysis of law shows that the same analysis, concepts, problems, ... apply across all of law.

• A. The problem: We want a single criterion to define "better" and "worse" for a whole society.
  • 1. How do we judge better and worse for a single individual?
  • 2. How do we aggregate.
  • 3. What is the upper bound--what defines "efficient" as opposed to "more efficient."
• B. Marshall's solution:
  • 1. Individual values are defined by revealed preference--what you would pay for an outcome if you had to.
  • 2. We aggregate values by summing individual values, measured in dollars--how much I would pay plus how much you would pay plus ... .
  • 3. We use a Bureaucrat God as a definition of the upper bound of efficiency
    • a. An imaginary central planner
    • b. Who knows everything anyone in the society knows
    • c. And has unlimited ability to calculate what should be done and make us do it.
    • d. An efficient outcome is one that cannot be improved (i.e. aggregate value cannot be increased) even by a bureaucrat god.
    • e. Note that the term "bureaucrat god" is my invention--don't blame Marshall for it.
• C. Objections:
  • 1. Insulin and heroin--are both of value, judged by people's willingness to pay for them. In other words, we are measuring what people do value, judged by their actions, not what we think they should value.
  • 2. Rich and poor: We are treating a dollar gain or loss to a rich person as just as important as a dollar gain or loss to a poor person, although most of us suspect that a dollar usually represents more happiness for the poorer person.
  • 3. Our upper bound is unattainable, since we have no bureaucrat gods, so why use it to define "efficient?" Isn't the result that all possible institutions are inefficient?
• D. Can you do better?
  • 1. Revealed preference is there to drive your institutions--anything else requires someone else second guessing individual preferences. I may not always know what is in my interest--but I can be trusted to want what is in my interest more than other people can be trusted to want what is in my interest.
  • 2. Ditto for rich and poor, also ...
  • 3. It is hard to use legal institutions to redistribute, and ...
  • 4. Not clear that if you can, you will do it the right way or at reasonable cost.
  • 5. Upper bound is useful as a maximum--and an idealized market achieves it!
• E. It is important to understand the idea of economic efficiency, whether or not you agree with it.
  • 1. Because asking the question "what rules would be efficient" turns out to be an interesting way of understanding the law.
  • 2. So take it as a working hypothesis, feel free to drop it at the end of the quarter.

• A. The simple argument:
  • 1. If the individual making a decision bears all of the costs and receives all of the benefits, then he takes the act which maximizes (private benefit-private cost) which equals (total benefit-total cost), hence he takes the efficient act.
  • 2. This still works as long as net private benefit=net total benefit.
    • a. Suppose A's action harms B by $10 and benefits C by $10--the two effects net out to zero, so he still makes the right choice.
    • b. Example: I become the 1001'st physician. The cost of a medical visit falls by ten cents as a result. That is a loss to the other 1000 physicians--exactly balanced by a gain to their customers.
• B. Consider an ordinary market transaction
  • 1. Buying a good--you pay the price, receive the benefit.
  • 2. Producing a good--you pay the cost of the inputs, receive the benefit in the price.
• C. For this argument to work, however, you have to show that the price I pay really represents the human cost of using the good, the price I receive represents the benefit produced by the good, so that "cost" in the sense I see it (dollars)= "cost" in the sense relevant for economic efficiency (value forgone from other ways of using the input, value of the output).
• D. Price theory does that. See my book, on-line.
• E. But ...

• A. Armies running away.
  • 1. Each individual is behaving rationally
  • 2. Yet all of them are worse off than if they behaved in a different way.
• B. Rational pollution.
  • 1. There are a hundred individuals, each of whom can burn cheap coal or expensive gas to heat his house.
  • 2. The coal pollutes, but ...
  • 3. 99% of the pollution produced by my coal burning is a cost for other people, so I only count 1% of it, so ...
  • 4. I burn coal even if the total cost, including the pollution, is substantially larger than for gas.
  • 5. Note that, on net, we are all worse off. I am externalizing 99% of the cost of my pollution, but paying 1% of the cost of the pollution by each of the other 99 people.
  • 6. But on the margin--the margin of how much I pollute--the pollution cost is almost all external.
• C. More generally, externalities lead to inefficient outcomes.

• A. This can work pretty well for norms such as honesty and courtesy, which involve things that individuals can directly observe and know about, but ...
• B. Not so well for big issues such as global warming or recycling, since the individual who is "enforcing" the norms has no direct information on what actions actually do good or damage and how much:
  • 1. CO2 emissions leading to a greenhouse effect may be a bad thing if you live in (low-lying) Bangladesh, but they make it easier to live in (cold) Canada, and they make agriculture more productive, since CO2 is an input to photosynthesis.
  • 2. Recycling paper may reduce the amount of U.S. forest, since reducing the demand for pulpwood makes it less profitable to use land to grow trees on.
  • 3. For similar reasons, the ivory ban might reduce the number of elephants.

• A. EPA calculates the cost imposed by a factory polluting, the cost of preventing it, permits it if banning it produces net cost. but ...
• B. How can they have the information necessary to do that?
  • 1. In particular, how can they know the cost to polluters of controlling, by ...
    • a. Pollution control
    • b. Alternative production methods
    • c. Not producing the good at all, or producing less.
  • 2. When the person with the best information about cost of controlling pollution is the producer, and it is in the producer's interest to argue that there are no cost justified ways of controlling the pollution, since then he will be allowed to continue polluting.
• C. Even if they did the calculation correctly for the firm itself, if it is allowed to pollute (because doing so is efficient), the price of its output will not include all of the cost imposed, so everyone else will be using an inefficiently large quantity. So they really have to do the calculation for the whole economy, not just a single firm--and then tell the firms that are allowed to pollute how much output (the efficient quantity) they are allowed to produce.
• D. And what makes it in EPA's interest to act that way, even if they can?
  • 1. Economics is supposed to apply rationality to everyone.
  • 2. So we really need a theory of the political marketplace; for some pieces of one, see my book.

• A. Impose an effluent fee equal to the damage done by the pollution.
• B. The firm is thus forced to internalize the externality--its costs, including the effluent fee, now equal the total cost imposed by its action.
  • 1. So it chooses the efficient alternative.
  • 2. Controls pollution if doing so is worth the cost
  • 3. Switches to alternative production methods if that is cheaper
  • 4. Closes down if cost is now greater than what the customers will pay.
  • 5. Note that the fine might go to the state, as in this example, or to the victim, as in tort law. We will discuss later why it matters which it is.
• C. Furthermore, the pollution cost is now included in the price for the firm's output, so everyone downstream has the right incentives.
• D. Problem solved--until 1960, when Coase published "The Problem of Social Cost."

• A. There is a small private lake; half the shore belongs to a chemical factory, half to a fishing resort, nobody else uses it.The factory's pollution kills most of the fish, reducing the resort's revenue from $50,000/year to $10,000/year.
  • 1. So we charge the factory $40,000/yr effluent fee.
  • 2. Stopping the pollution only costs $30,000/yr, so the factory stops polluting, but ...
  • 3. The resort could switch to a riding resort, with no fishermen, and make $40,000/year.
  • 4. So "No pollution/fishing" solution costs $30,000/year to factory
  • 5. "pollution/riding" costs $10,000/yr lost revenue to the resort
  • 6. We have gotten the inefficient outcome.
• B. The general argument is that an externality is not a cost A imposes on B, but a cost jointly due to decisions by A and B.
  • 1. Air pollution in Los Angeles wouldn't be a problem if nobody lived there. If that seems unreasonable, consider ...
  • 2. Dropping bombs next to people camping on bombing ranges imposes externalities on the campers, but it seems likely that moving the campers is the cheaper solution.
  • 3. Simple case of two roommates: My playing loud music imposes a cost on you when you want to sleep. But if the rule is "I can't play when you want to sleep," then your sleeping imposes a cost on me--I can't play loud music when you're sleeping.
    • a. If I can play the music even if you want to sleep, I will sometimes due so inefficiently (when the cost imposed on you is more than the benefit to me)
    • b. If I can't play it when you are sleeping, you will sometimes sleep inefficiently (when ... ).
• C. Coase offers some interesting real world examples:
  • 1. The physician and the candy factory.
  • 2. Chimneys not drawing after the building next door was rebuilt--higher.
  • 3. Stopping up the vent by which brewery fumes escaped --via an old mining tunnel--from a basement brewery.
  • 4. A new building shading a hotel swimming pool.
• D. The implication of this argument is that nothing works.
  • 1. In setting general rules, such as "polluters must pay a fine equal to damage done," the legal system does not know which party can adjust at lower cost.
  • 2. If it makes the wrong party liable, that party may adjust, so the other party has no incentive to.
  • 3. And the optimal solution might involve adjustments by both parties.
• E. Think of the question as "how do we bundle property rights."
  • 1. Does my ownership of part of the lakeshore imply the right to pollute, or the right not to have the lake polluted?
  • 2. Does my ownership of the hotel pool include ownership of the sunlight falling on it?
  • 3. Who owns the right to control the movement of currents of air--the air carrying the smells from the brewery into someone else's courtyard, or carrying (or not carrying) the smoke from my chimney away?
  • 4. Frat house and recording studio--does a landlowner have a right to make noise or be free from noise?
  • 5. How should we decide such questions? That is where the Coaseian argument is leading.

• A. Back to the factory and the resort--plus bargaining.
• B. The Coase Theorem: Any initial assignment of rights will lead to an efficient outcome, provided that transaction costs are zero.
  • 1. Not always the same outcome--does the candy factory or the physician have to pay the other to get his way?
  • 2. Which matters for the distribution of income, if we take nominal property holdings (who owns this piece of land) as fixed, because we are changing actual property holdings (who owns which rights with regard to the land).
  • 3. It matters for the use of resources only if the owners are also the consumers.
    • a. A hermit who neither has nor wants much money lives on the shore of the lake and loves having an unpolluted lake to look at. If the factory has a right to pollute, the hermit will not offer the factory enough to make it in their interest to stop polluting. But if the hermit has the right not to have the lake polluted, the factory will not offer the hermit enough to get his permission to let them pollute.
    • b.In such a situation, we have different uses of resources according to who starts with the rights--but both outcomes are efficient!
    • c. After all, how your desires get weighted in determing the efficient outcome depends in part on how much money you have, and different initial allocations correspond to different distributions of wealth.
    • d. If the hermit owns the right to decide that the lake won't be polluted, he is richer than if he doesn't, and the efficient outcome therefore gives more weight to his preferences.
    • e. Just as, in a world where I happen to own downtown San Jose, somewhat more of goods I very much like will be produced.

• A. Factory has a property right--it is free to pollute and not liable for the results of its pollution.
  • 1. Without transactions, the factory pollutes whether or not doing so is efficient.
  • 2. With transactions,
    • a. resorts face a public good problem in raising the money to pay the factory not to pollute.
    • b. Because if the factory stops polluting, all resorts benefit, whether or not they helped pay.
• B. Factory pays a fine equal to the damage done by its pollution.
  • 1. No transactions are possible.
    • a. If damage done by the pollution is <cost of controlling it, factory continues to pollute, fishing resorts switch to riding resorts, efficient outcome.
    • b. If damage done>cost of controlling>cost of switching to riding, the factory stops polluting, the resorts don't switch, inefficient outcome--or ...
    • c. The factory keeps polluting for a while, the resorts switch, the fine falls, efficient outcome via strategic behavior by factory.
    • d. If damage done>cost of switching to riding >cost of controlling>, the factory stops polluting, the resorts don't switch, efficient outcome--or ...
    • e. The factory keeps polluting for a while, the resorts switch, the fine falls, inefficient outcome via strategic behavior by factory.
  • 2. Transactions are possible.
    • a. If damage done by the pollution is <cost of controlling it, factory continues to pollute, fishing resorts can't make any offer to control pollution that the factory would accept, efficient outcome.
    • b. If damage done>cost of controlling>cost of switching to riding, the factory pays the resorts to switch to riding, efficient outcome--or ...
    • c. The factory keeps polluting for a while, the resorts switch, the fine falls, efficient outcome via strategic behavior by factory.
    • d. If damage done>cost of switching to riding >cost of controlling, the factory stops polluting, the resorts don't switch (there is no price the factory would offer that they would accept), efficient outcome--or ...
    • e. The factory keeps polluting for a while, the resorts switch, the fine falls, inefficient outcome via strategic behavior by factory. Or the resorts pay the factory to stop polluting--but again face the public good problem.
• C. Factory pays a damage payment to the resorts for the damage done by its pollution.
  • 1. This is like B above, except that
  • 2. Strategic behavior by the factory doesn't work, because the resorts are getting paid for the damage done, which also implies
  • 3. That if the factory continues to pollute, the resorts have no incentive to switch to riding if there are no transactions, but ...
  • 4. The factory can pay them to do so, and will if it is efficient (if the cost of switching is less than damage done and less than cost of pollution control). No public good problem--the ones that don't agree will continue to collect damages until they do.
• D. Resorts have a property right--any resort can enjoin the factory from polluting.
  •  
  • 1. If it is efficient for the factory to pollute, with no transactions they won't be able to, and ...
  • 2. With transactions, there is a holdout problem in getting permission from the resorts. Any single resort can block the deal by refusing its permission, so some may try to get a big share of the gain by threatening to block, so the deal may fall through.
• E. Resorts can enjoin--but must pay damages equal to the cost that doing so imposes on the factory. Implications are left as an exercise for the reader--homework not to be turned in.
• F. Damage payment vs fine:
  • 1. Damage payment eliminates the opportunity for strategic behavior by the factory. If the factory wants to keep polluting, the resorts will be happy to keep operating as fishing resorts--and suing the factory for their losses.
    • a. This solves a problem that arises because of the one vs many nature of the situation.
    • b. In 1 vs 1, resort is in as good a position to be strategic as factory.
    • c. In many vs many, each factory knows that its decision has little effect on the resorts, so it doesn't pay to try to behave strategically.
    • d. Consider driving dangerously for strategic reasons--so that other people will drive very carefully, or leave the road to you. One driver is such a small part of the environment faced by other drivers that it doesn't pay him to modify his behavior to try to affect theirs.
    • e. But ... a friend of mine followed a policy of not taking dents out of her car, on the theory that other drivers would conclude that it was prudent for them to be careful near her, since she didn't seem to care. So it may become a 1 vs 1 interaction at the last moment.
  • 2. Damage payments also eliminate the incentive of the "victim" who is being compensated to eliminate the problem himself if there are no transactions.
  • 3. And a fine leads to a "double incentive" problem if there are transactions.
    • a. Go back to the resort/factory problem, assume a fine, small number of resorts.
    • b. Pollution damage = $1,000,000/year, control costs $1,500,000/year, riding resorts not an option.
    • c. Without transaction, the factory keeps on polluting, but ...
    • d. Resorts offer to pay the factory another $600,000 to stop polluting.
    • e. The factory stops.
    • f. Coase + Pigou is too much incentive.

• A. The central planning approach:
  • 1. Decide whether it is efficient for the factory to pollute.
  • 2. If the answer is yes, dismiss the resorts' lawsuit against the factory.
  • 3. If the answer is no, either make the factory liable or grant an injunction forbidding it from polluting
  • 4. But this requires the court to know enough to figure out whether it is efficient for the factory to pollute.
  • 5. And the right answer might be neither "pollute" nor "don't pollute" but some reduction in pollution, along with some adjustment by the resorts.
• B. The Pigouvian approach:
  • 1. Decide which party is the lower cost avoider of the problem (which does not require you to know whether it is worth avoiding the problem).
  • 2. Make that party liable for the cost of the problem.
  • 3. And leave it to him to decide whether to avoid the problem or pay damages.
• C. The Coaseian approach:
  • 1. Consider the different ways the initial rights might be defined.
  • 2. And the different outcomes that might be efficient.
  • 3. And how easily, given transaction cost problems, one could get from each initial definition of rights to each outcome.
  • 4. And how likely it is that each outcome will be the efficient one
  • 5. And choose your starting point accordingly--the one from which it is relatively easy to get to the outcomes that are likely to be efficient.
  • 6. More precisely, consider that the cost of the problem is the transactional cost of solving it if it is solved or the inefficiency resulting from not solving it if it is not solved, and choose the starting point that minimizes the expected cost.

• A. If I want to
  • 1. use your car to drive somewhere, I borrow, buy, or rent it.
  • 2. If I want to "use" your car by running into it, I do so and you sue me.
• B. Think of these as two different ways of handling the externalities/incompatible uses problem.
  • 1. The property approach: The owner gets to decide how his property is used, and if it is valuable to someone else to use it, the someone else gets the owner's permission.
  • 2. The liability approach. Anyone can use the property--and owes the owner damage payments for any cost imposed on him.
• C. Note that property rules don't just enforce themselves, so another way of putting it is that
  • 1. under liability rules, the purpose of damages is to produce the efficient level of use directly (i.e. Pigouvian approach);
  • 2. under a property rule, the purpose of damages is to force the transaction onto the market, and have the efficient level produced by market transactions.
• D. In our recent discussion of the factory and the resorts, property rules were
  • 1. absolute right (the factory can pollute if he wants to, so if the resorts object, they must persuade him not to) or
  • 2. injunctive right (the factory cannot pollute without permission from the resort)
  • 3. The difference being that in the former case the right is held by the person with physical control over the action, so no legal procedure is needed in order to implement it. In the latter case, the right is held by someone else, so he must go to court to forbid the person with physical control from acting.
• E. Why would we choose one rule or the other?
  • 1. A property rule makes sense if transaction costs for reallocating the right are low--the case on ordinary markets, where people can freely buy and sell their property.
  • 2. A liability rule makes sense if transaction costs are high:
    • a. auto accidents, since you can't contract in advance with every other driver for permission to impose a risk on them.
    • b. A lost hunter who finds a locked cabin and breaks in--high transaction costs because the owner can't be found in time.
    • c. Pollution with many small victims
    • d. What about costs imposed on one neighbor? There is a bilateral monopoly bargaining problem if the difference between cost imposed and cost of controlling is likely to be very large.
  • 3. Another relevant factor is how easily a court can measure damages. If it is very hard, then the property solution might be better even with substantial transaction costs. That is how we handle many bilateral monopoly cases--the purchase of a tract of land that is very valuable to only one purchaser, for example.
  • 4. A third possible rule is the central planning solution: The court (or the government) transfers the right to the person they think values it most, typically the government--the military draft, for example, or jury duty.
    • a. Note that this requires even more court information than the liability rule--and creates a greater incentive to spend resources litigating.
    • b. Imagine that the way I got a house was to persuade a court that it was worth more to me than to anyone else--at which point it was mine for free.
    • c. As opposed to the liability method--I seize the house, and the court makes me pay its value to the previous owner, or ...
    • d. The property method--which is the one we actually use.
    • e. One reason we don't use either the property or the liability approach for (private) slavery is that we are pretty sure the efficient bundling is for me to own myself.
• F. All of this is really an application of the Coaseian approach:
  • 1. The legal system uses property rules when that minimizes the problem of getting goods to their highest valued use.
  • 2. It uses liability rules when that minimizes ...

• A. Why we discuss them here: Because issues of allocating risk show up in legal rules all the time.
  • 1. When there is an auto accident, who compensates whom?
  • 2. When a bottle of coke explodes and injures you, is Coca Cola liable?
  • 3. When a contract breaks down because of unanticipated problems, which party bears the cost?
• B. Why insure?
  • 1. Risk aversion
  • 2. aka declining marginal utility of income.
  • 3. Suppose I face a 50% chance of an income of $10,000 for next year, and a 50% chance of $110,000.
    • a. I expect that dollars will buy me much more important goods if I have only $10,000 of them than if I have $110,000, so I would like to "transfer" money from the one state of the world to the other.
    • b. Which I do by betting someone $50,000 that I am going to have an income of only $10,000 next year.
    • c. Now if I win my bet I have $10,000+$50,000=$$60,000 and if I lose I have $110,000-$50,000=$60,000; I prefer that lottery to the one I started with.
    • d. In other words, I have insured against a low income.
• C. Why not insure--part I: Moral hazard
  • 1. You will take inefficiently low precautions if you have insured your factory against fire.
  • 2. More generally--we are creating an externality by transferring the cost of my poor precautions to the insurance company.
  • 3. We can reduce the problem by coinsurance--the insurance company only pays part of the loss.
    • a. This solves most of the problem if most of the precautions worth taking are cheap ones--still worth taking when I only bear part of the cost of a fire.
    • b. It doesn't do so well if the important precaustions are expensive--only worth taking if I get most of the benefit.
    • c. We can also increase the probability of fire a lot by insuring for more than 100%
  • 4. But this might sometime be a feature rather than a bug! Suppose the insurance company knows more than I do about preventing fire, and requires me to take its advice as a condition of insurance.
    • a. I once read an article that claimed there was an extensive literature on safety engineering for many dangerous activities, such as chemcial companies--driven by insurance companies
    • b. But no such literature for nuclear reactors--because they are insured primarily by the government
    • c. With the result that reactor control rooms are designed to impress local government officials whose pe support is needed to build the reactor, not professional safety engineers working for insurance companies--and end up looking like a set for a science fiction movie.
    • d. I don't know if this is true--but it could be.
  • 6. We are back with Coase--the loss is jointly caused, and we want to give both parties the right incentives.
  • 7. Consider the coke bottle which exploded because I was emphasizing a rhetorical point on a hot summer day by shaking my hand--with the bottle in it.
• D. Why not insure--part II: Adverse selection.
  • 1. Market for lemons story.
    • a. $8000 cream puffs, $4000 lemons, sellers know which are which, buyers don't, half the cars whose owners would like to sell them if they can get a reasonable price are each.
    • b. Do they all sell for $6000? Then owners of lemons would unload even if they don't really want to sell them, but very few creampuffs would sell, so ...
    • c. Average car sold at $6000 is almost certainly a lemon, so not worth $6000, so price will be lower, so ...
    • d. An equilibrium where price is a little about $4000, reflecting the existence of a few creampuff owners who badly want to get rid of their cars, and almost all the cars sold are lemons.
    • e. This is an example of market failure since lots of owners of creampuffs value their cars at less than the buyer does, but can't sell.
    • f. Solution--guarantee the car. But ...
    • g. That produces a moral hazard problem on the other side, since it reduces the buyer's incentive to take precautions.
    • h. The two objectives are in tension, since from a moral hazard standpoint the risk should be born by the buyer, from an adverse selection standpoint by the seller.
    • i. My friend Ami Glazer exploited a different version of the same principle, by finding a used car dealer who was willing, for an additional payment, to guarantee the car he was selling--then buying the car without the guarantee.
  • 2. Insurance story:
    • a. "I want life insurance right now." Don't sell it to him.
    • b. More generally, good risks end up underinsured, bad risks end up overinsured.
    • c. Tension between who has the assymetric information (buyer) and who can best risk spread (seller)
• D. The difficulty of solving two or three problems at once.
  • 1. We will be seeing this problem again in the context of contract law.
    • a. Buyer and seller of coke bottles both affect the chance of an explosion
    • b. Each starts with some specialized knowledge.
    • c. Seller can risk spread best.
    • d. Note that one solution is to figure out who was "at fault"--but that isn't always easy.
  • 2. Sometimes you can construct an ingenious solution:
    • a. Robin Hanson's health care scheme: You decide how much your life is worth. Buy that much life insurance, from a company which will then provide you medical care at its expense. Now resell the excess life insurance to a second company, so as not to end up with insurance you don't want. The first company has an incentive to figure out what medical care is worth having, given the value of your life to you (and it). Make sure the two companies can't communicate, and the second doesn't know who you are--since it is now in their interest for you to die.
    • b. We are creating one market to solve the moral hazard problem, another to solve the risk spreading problem.
  • 3. Another example of this problem we will come back to is the difference, in tort law, between optimal damages to deter behavior and optimal damages to insure the victim.

• A. Auto example: speed limit or liability for accidents.
• B. More generally: Punish by outcome or punish by observing inputs to the production of bad outcomes, and charging for them.
• C. Advantage of ex post:
  • 1. Utilizes private information of actor re what he is doing and its consequences. The most dangerous thing I do while driving is to think--about things other than the driving.
  • 2. Only have to litigate when an accident happens.
  • 3. At which point the damage is there to be measured.
• D. Advantage of ex ante:
  • 1. Utilizes the legislator's knowledge--but why can't he transmit that directly by telling drivers what speed is dangerous (or whatever)?
    • a. Maybe they won't believe him.
    • b. But this seems to assume predictable irrationality--which is a bad idea in economics.
  • 2. Don't have to measure damage each time--just estimate expected value.
  • 3. Permits much smaller punishments. Less litigation cost per case, plus ...
  • 4. More efficient punishment--a lot of $100 fines instead of occasionally hanging someone.
• E. Things that are not advantages but might seem to be.
  • 1. "Deter vs punish." But both methods deter.
  • 2. Utilize private information about how important it is to drive fast (or whatever)--both methods do that.
    • a. With ex ante, I decide whether it is worth it to me to risk the fine.
    • b. With ex post, I decide whether it is worth it to me to risk the accident, and the resulting costs.

     

• A. Insurance and the multiple margin problem:
  • 1. Typically insoluble with a single variable--optimal product liability.
  • 2. Sometimes soluble by adding a second market--tort plus insurance.
  • 3. Or a second mechanism: Product liability with well informed consumers.
  • 4. But ... lots of problems left:
    • a. The fine is both the punishment for a tort and the incentive to sue.
    • b. A damage payment affects the incentives of both "tortfeasor" and "victim."
    • c. Solving b gets back to a--criminal punishment eliminates victim's incentive to report the accident.
• B. Ex post/Ex ante punishment:
  • 1. Punishment cost, inefficiency.
  • 2. The basic argument. Implies that we will almost always use ex post for small offenses.
  • 3. Why punish attempts? Ex post for ex ante for ex post. Because of punishment costs.
  • 4. General point--in a world of costless enforcement, punishment is merely a pricing mechanism.
    • a. In a world of costly enforcement, the arguments become more complicated.
    • b. There is a balance of pricing and preventing
    • c. both because there are some efficient offenses, and because whether it is worth the cost of deterring an offense depends on how much it costs to do so and how much damage the offense does.

• A. Homesteading story.
  • 1. If I get ownership of 160 acres of government land by being the first person to farm it for X years, there is an incentive to start farming prematurely--losing money on the farming in exchange for obtaining ownership of valuable land. If I wait until the land is worth farming, someone else will already be there.
  • 2. So the land gets settled prematurely--early enough so that the losses during the early years just balance the value of ending up with the land.
  • 3. Implying that the homesteading act wiped out the land value of much of the U.S.
  • 4. The analysis becomes a little more complicated if you take account of the fact that different potential homesteaders have different costs for homesteading, so that it is only the marginal homesteader who gets no net gain from homesteading.
  • 5. And the actual history of thepublic lands prior to the homesteading act suggests that there are serious problems with the obvious alternative of auctioning the land off.
    • a. When the land is opened for homesteading there are already settlers living on it illegally.
    • b. They petition congress to led them buy "their land" at the minimal auction price.
    • c. And if congress doesn't agree, they use local pressure to rig the auction, so as to make sure that each settler gets the 160 acres containing the five or ten acres he is actually farming for the minimal auction price.
• B. General principle: If I can spend resources on making sure that I get some valuable right instead of someone else getting it, I will. If there are lots of potential claimants, they will compete for the right until, on average, the cost of getting it equals its value--at which point their rent seeking expenditures have canceled out the benefit of the right--the land value of much of the U.S. in the homesteading case.
• C. Litigation implication.
  • 1. When I spend money litigating, the object is to get a damage payment at someone else's expense; and his object is to keep me from doing so.
  • 2. So both of us are engaging in a form of rent seeking--although not with free entry into the business.
  • 3. So there is no reason to expect an efficient level of expenditure on litigation.
• D. Legislation implication:
  • 1. If you have a system that transfers money to members of a particular group, that gives people an incentive to join that group, even if doing so is costly, which reduces the net benefit received by those who collect the money.
    • a. Obvious example--turning down a not very attractive job in order to collect welfare benefits.
    • b. Less obvious example: A law paying $10,000/year to anyone who is legally blind. Nobody is likely to become blind in order to collect--but someone who is almost blind might decide to declare himself legally blind--at the cost of not being able to get a job that his actual eyesight is sufficient to do.
  • 2. In addition, groups have an incentive to spend resources lobbying to get and to prevent such transfers.
• E. Crime implication:
  • 1. What is wrong with theft anyway--economically speaking? Isn't it just a transfer?
  • 2. No. The opportunity to steal gives thieves an incentive to spend resources on stealing.
    • a. The marginal thief ends up leaving a job that pays $20,000 in order to spend his time stealing $20,001.
    • b. The thief gains $1, the victims lose $20,000
    • c. And what is stolen is often worth more to the victim than to the thief (goods rather than money)
    • d. And there is the additional cost of precautions to prevent theft.
    • e. So on net theft reduces the total size of the pie.
• F. Another way of putting this is that rent seeking involves inefficiencies due to negative externalities--one consequence of my homesteading the land is that it isn't there for you to homestead.

• A. Bilateral monopoly.
  • 1. Economic example: I have the only apple, worth $1 to me. You are the only customer, and value it at $2. If we can agree on a price, there is a net gain of $1 to be divided between us, with the division implied by the price.
  • 2. My six-year-old threatens to throw a tantrum if she doesn't get her way. Doing so imposes net costs, so there is a gain to finding some mutually acceptable outcome. Don't assume you can simply be firm and always win. You may have thought out the logic of bilateral monopoly bargaining better than she has, but she has a hundred million years of evolution on her side--during which offspring who succeeded in getting a larger share of parental resources were more likely to survive to reproduce.
  • 3. Doomsday machine.
    • a. The idea. Lots of very dirty bombs buried under the Rockies. If the Russians attack, the bombs go off and the fallout kills everyone on earth. So the Russians won't attack, and we don't need to pay for nuclear arms.
    • b. The reality: Our (and their) nuclear systems were doomsday machines, with human triggers. They worked, and therefore were not used--fortunately.
  • 4. Bully: If you train yourself to punch out anyone who gets in your way, and people know it, they will stay out of your way, and you don't have to fulfill the commitment--until you run into someone else following the same strategy, and one of you ends up dead. A doomsday machine on the individual level.
  • 5. In general, the outcome of such games depends largely on issues of commitment and reputation, which are hard to include in our analysis.
• B. For more on game theory, see the game theory chapter in Price Theory.

Table of Contents