Looking Back Over the Semester

Econ 522 – Lecture 23 (December 4 2008)

One thing I forgot last lecture. I mentioned the debate over whether rich and poor should face different sentences for similar crimes. In the U.S., this is not generally done – some crimes are punished by a choice of a fine or jail time, but fines and jail sentences, at least in theory, are not differentiated according to the criminal’s wealth.

In Europe, fines are more common for minor criminal offenses, and they are higher for the rich than for the poor. Specifically, day fines are commonly used – that is, rather than a fine being for a certain amount of money, it is specified as a certain number of days’ salary – so that rich people do face a greater penalty.

Looking back over the semester…

We started out by introducing the notion of efficiency, or Kaldor-Hicks efficiency, or wealth maximization, and argued that it’s a pretty good goal for the legal system to aim for

With property law, the Coase conjecture suggested that if there are no transaction costs, efficiency is easy to achieve – just define property rights, choose any initial allocation of rights, and allow people to bargain with each other
(When there are transaction costs, we’re left with a choice: design the law to minimize transaction costs, or design the law to allocate rights efficiently as often as possible.)
But setting up and enforcing a system of property rights costs something, so efficiency demands only doing this when the gain outweighs the cost – we saw this with Demsetz

With contract law, we looked for rules that would yield efficient breach, and efficient reliance, and efficient disclosure of information, among other things

With tort law, we looked for rules that would create incentives for efficient precaution to prevent accidents, and efficient levels of risky activities

And in the last few lectures, with both civil and criminal law, we discussed designing the legal system more generally to minimize the combined cost of implementing the system and living with its errors – which is the same as maximizing the resources left to society

So with each branch of the law, the question we’ve asked is, what would an efficient system look like, and how do the actual rules in place measure up to the standard of efficiency?

Now that we’ve made it through all that, it seems like a good idea to look again at efficiency, and return to two key questions:

is efficiency the right goal for the law?

and should we expect the common law system to naturally tend toward efficiency?

To answer the first question, we started with Posner

Posner points out that the appeal of a Pareto-improvement is that everyone would consent to it
He tries to make an analogous argument for Kaldor-Hicks improvements, or changes that increase overall efficiency

Obviously, you can’t do this after the fact: once I’ve hit you with my car, I’d prefer a negligence system (or a system of no liability), and you’d prefer a strict liability system
But Posner introduces the idea of ex-ante compensation
He points out: “if you buy a lottery ticket and lose the lottery, then… you have consented to the loss”
And similarly: “Suppose some entrepreneur loses money because a competitor develops a superior product. Since the return to entrepreneurial activity will include a premium to cover the risk of losses due to competition, the entrepreneur is compensated for those losses ex ante.”

Suppose we all agree that a negligence rule is most efficient than a strict liability rule
And suppose for simplicity that everybody in society is the same – everyone is the same age, is equally good drivers, drives the same amount, and so on
If I just hit you with my car, you’d prefer a strict liability rule
But if we all got together before any accidents happened and chose a system, we’d both agree to the negligence rule
If you wanted to be compensated if you ever got hit by a car, you could buy accident insurance – which, if a negligence system really is more efficient, would be cheaper than your expected liability would be under a different system

So Posner’s idea: we were all compensated ex-ante for the choice of the more efficient system. (If the system is negligence rather than strict liability, then liability insurance is cheaper.)
And therefore, it’s as if we all consented to it

Obviously, if people are different – some people don’t drive, but still get hit by cars – the argument is harder, but he tries to make it more general
Suppose we introduced a law that was inefficient, but favored tenants over landlords
We might think that tenants would like it and landlords would not
But since rental rates are determined competitively, it would probably just result in higher rents, to compensate landlords for the change in the law
So Posner’s general argument: if we could all meet up to choose a rule ahead of time, we’d all choose the efficient one
So assume that we all consented to it, and go with it
(He points out this is like the principle in contract law, where to deal with an unforeseen contingency, you try to impute the terms the parties would have chosen, if they had addressed that scenario in the contract.)

So that’s Posner on why the law should aim for efficiency
Cooter and Ulen take on the question as well, and offer another reason why the law should aim for efficiency
They acknowledge that society may have goals other than efficiency – in particular, a society may care about the distribution of wealth, not just maximizing the total amount of it
But, they argue, if society wants achieve redistribution, it is better to design the law to be efficient, and use the tax system to control distribution
They give several reasons for why the tax system is a better way to redistribute wealth than the legal system:
The tax system can directly target people with high or low incomes, rather than relying on rough proxies (like “consumers” and “investors”, or “doctors” and “patients”)
Effects of changes in the legal system are harder to predict than changes in tax rates
Lawyers are more expensive than accountants (or really, transaction costs are higher when redistribution is through the legal system)
A law aimed at redistribution would function like a tax on a particular activity; but the more narrow a tax is, the more distortion it causes, and therefore the more deadweight loss it creates

So that gives us two views for why we should aim for a legal system that is efficient, or maximizes wealth:

Posner: it’s what we’d all agree to ex-ante
Cooter and Ulen: if we want to redistribute, it works better to do it through taxes

This brings us to the second question: can we expect the common law to naturally tend toward being efficient?

Posner gives one reason why it would

He points out that many people believe the government, and the courts, respond to politically powerful interest groups
But the same idea of ex-ante consent suggests that even these groups would probably want efficient laws
If landlords, as a group, were more politically connected and influential than tenants, we might expect them to influence the laws
But like we said before, landlords wouldn’t get that excited about inefficient pro-landlord laws, because they’d probably just lead to lower rents to compensate
So well-connected groups wouldn’t resist efficient rules

Cooter and Ulen go further, and offer three specific mechanisms that could lead the common law toward being efficient:

First, the common law often implements social norms or existing industry practices. These norms and practices likely developed, and lasted, because they were better than the alternatives, so they are likely efficient
We saw this at the beginning of the course, with whaling
The whaling industry in different places developed norms and practices which were well-suited to each environment and therefore efficient
And common-law judges chose to respect these norms, which made the common law efficient as well

Second, some judges may deliberately try to push the law in the direction of efficiency
Under the common law, judges generally follow precedents
But judges do occasionally reverse precedents and “make new law”
If judges are more willing to reverse inefficient precedents and replace them with efficient ones, then this will push the common law in the direction of efficiency
We saw a dramatic example of this with the Hand Rule, where the judge explicitly incorporated efficiency into the legal standard of care for a negligence rule
We also saw this with the ruling in Boomer v Atlantic Cement, where even though precedent was to issue an injunction in nuisance cases – this was a cement factory that created dust and noise – the court felt that an injunction would be inefficient, and chose to abandon precedent and order only damages.

Cooter and Ulen also consider a third way the common law could naturally evolve toward efficiency: inefficient laws might lead to more litigation than efficient ones
Inefficient laws have higher social costs than efficient ones
So they must have higher costs for at least one party
So there should be a greater incentive to challenge them or violate them, leading to more litigation
Even if we imagine that courts do not consciously favor efficiency, but just randomly make new law a certain fraction of the time, more litigation around inefficient rules leads to a greater chance that inefficient rules are the ones that are overturned in favor of new rules
If the new rule is inefficient, it will still lead to a large amount of litigation
But if it’s efficient, it will lead to less, and is therefore more likely to survive
And so the common law would evolve, over time, in the direction of efficiency

(Cooter and Ulen concede that while this is an appealing argument, it may be a fairly weak effect
This is because the private gain from challenging a “bad” law is different from the social gain
If I go to court and get an inefficient law overturned, there is a positive externality on the rest of society
Because I bear most of the cost of doing this, but don’t get all of the benefit, the incentives to challenge inefficient laws are not that strong, and therefore it won’t happen as much as it “should”.)

However, there’s an opposing view that litigation will not always lead to efficient laws

A paper by Gillian Hadfield, “Bias in the Evolution of Legal Rules”
Hadfield imagines that every case is a little bit different – so the “ideal” rule might vary from case to case
But it’s impossible to have legal rules that are too complicated/ambiguous, so the best the court can hope to do is to be right “on average”
The court will form an opinion of what rules are efficient on average based on the cases it sees
But the cases the court sees will not be a random sample of all possible cases
That is, since there will be a bias in which cases go to trial, the court will have a biased view of what rules are efficient.

How does this work?
A given legal rule will have different effects on different individuals/firms
Some might find it not that costly to follow the rule: Hadfield calls these compliers
Others will find it too costly to comply with the rule, and will choose to be violators
And finally, some might both these options too costly, and drop out of the market entirely.

Her point is that being right on average requires doing what is efficient, given the true mix of compliers and violators
But the court mostly sees cases involving violators, so it is likely to choose the rule which would be efficient if everyone in the market looked like the violators
If compliers and violators are different enough, this rule will not be efficient overall.

I think this is simplest to understand by an example
We made the point earlier that successive application of the Hand Rule could be used to clarify the standard for negligence
Suppose product liability in a certain industry is covered by a simple negligence rule, and the Hand Rule is used to determine the legal standard of care.
First, suppose all the firms in the industry face the same cost-benefit tradeoff from taking precaution.

Firms initially take a low level of precaution, so there are some accidents
Some lead to lawsuits, the court finds that more precaution would have been cost-justified, firms respond by taking a higher level of precaution
Eventually, some lawsuits fail because more precaution was not cost-justified, and we’ve stumbled upon the efficient level of precaution, which is now the legal standard for avoiding negligence

But now suppose all the firms in the industry aren’t the same
Suppose that for whatever reason, some firms find it fairly cheap to take precaution, while some firms find it expensive
The court can’t distinguish between these firms, so it can’t mandate different levels of care for the two types of firms
(Or even if it could, maybe it shouldn’t.)

A legal standard for negligence would lead (in theory) to every firm taking the same level of precaution
The level that would minimize total social costs would be based on the average costs of all the firms in the industry. (DRAW IT)

However, think about what would actually happen as firms applied the Hand Rule
Suppose there’s some uncertainty about the exact legal standard
Firms who can take precaution cheaply don’t have any need to run the risk – they can set a high level of precaution, leading to very few accidents and very few lawsuits, and be fairly sure they’ll avoid liability when accidents do occur.

On the other hand, the firms who find it very costly to take precaution will take less
Even if they want to meet the legal standard of care, they have a much greater incentive to “cut it close” – to just barely meet the legal standard, rather than exceed it by a lot
Some firms may even find precaution so costly that they choose to ignore the legal standard and accept liability.

But now think about the cases the court will see: they will consist almost entirely of cases involving firms with high precaution costs
The firm can only use the information it has; so as it applies the Hand Rule, it will constantly be assessing what precautions are cost-justified for these higher-cost firms
So by applying the Hand Rule, courts will settle on a legal standard based on what is cost-justified for high-cost firms; this rule would be efficient if all the firms in the market were high-cost
But applying this rule to all the firms in the market – both the low- and the high-cost firms – is inefficient. (In this case, it sets the level of precaution below the socially optimal level, leading to too many accidents.)

Numerical example of Hadfield

Four possible levels of precaution: none, low, medium, high
Accidents cost $1,000
Chance of an accident: 20%, 10%, 5%, 1%
Two types of firms: high-cost and low-cost
High-cost firms: $0, $60, $120, $180
Low-cost firms: $0, $30, $60, $90
Equal number of high- and low-cost firms
Efficient rule: high-cost firms should take low precaution, low-cost firms should take high precaution
Single rule for every firm: most efficient would be for everyone to take medium level
But if the initial rule is unclear, low-cost firms will take more precaution than high-cost firms, so it will mostly be …

Hadfield’s claim is that this will happen not just in setting negligence standards, but in all situations where the court adjusts legal rules based on what appears efficient from the cases it sees
Whatever the existing rule, it will lead to some “compliers” and some “violators”
The court will mostly see cases involving violators, so it will gravitate toward rules that would be efficient if everyone was violators
As long as compliers and violators are different enough, these rules will not be efficient.

Hadfield doesn’t use a formal model, but argues the point fairly well
She argues that as long as there is enough heterogeneity among potential cases – that is, as long as different cases vary enough to merit different optimal rules – and as long as the court only learns about cases that go to trial, this bias will lead to inefficient rules
She also discusses the assumption that courts only learn about the cases they see, and why this is.

So that gives us a few reasons why the common law will naturally be efficient, and one reason why it might not

There’s one other thing I thought I’d talk about today