Unit 5: Legal Procedure
So far, we have been dealing with substantive aspects of law – what the law is and should be. By contrast, in this unit we will study procedural aspects of law – the way in which the legal system processes claims.
The basic tradeoff inherent in the legal process is that more thorough procedures reduce errors and inefficiencies, but more thorough procedures are also most costly to administer. The optimal legal system balances out the marginal benefit of efficiency improvements from a system that makes better decisions with the marginal cost of administering the system.
We will begin this unit by reviewing the basic stages of a legal claim – complaint initiation, information exchange, pre-trial settlements, trials and appeals. We will then deal with a number of special topics in legal procedure like nuisance lawsuits, class-action suits and the market for legal services.
Complaint Initiation
The number of complaints filed depends on three factors.
1. How many injuries occur – More injuries lead to more complaints. 2. The cost of filing complaints – Higher cost leads to fewer complaints.
3. Expected value of complaints – Higher judgements and higher probability of success lead to an uncertain change in the number of complaints. There is more incentive for victims to file claims, but also more incentive for injurers to avoid injuring people.
An important fact to note right away is that the state subsidizes legal complaints since the filing fee is typically much less than the state’s cost associated with litigating the dispute (e.g. paying judges, building courthouses, etc.). The amount of the subsidy varies widely across countries and jurisdictions, but many countries’ constitutions contain clauses to the effect that access to courts is a universal right that is guaranteed to everyone, which has traditionally been interpreted to mean that filing fees should be low or zero.
Question: Breach of contract cases influence production, whereas property disputes mostly impact distribution (not total production). If this proposition is true, explain the consequences for setting filing fees at efficient levels for contract versus property cases.
Answer: If the proposition is true, it suggests that remedying contract issues (which impact efficiency) is more important that remedying property issues (which just impact distribution). Thus, filing fees should be lower for contract cases than for property cases.
Information Exchange
Discovery is the process by which parties exchange information with each other. In Common Law, the two sides exchange information on their own.1 In civil law, the judge adjudicates information
exchange. The parties do not request information directly of each other.
What is the impact of information exchange on how the process unfolds? Unsurprisingly, it depends on the kind of information being exchanged. Let’s start with two preliminary concepts.
Relative optimism occurs when each side is relatively optimistic about its own chances of winning the case (or the judgment amount in the event of a win). Clearly, relative optimism is bad for the prospect of being able to settle a case in advance. If the plaintiff is sure of winning or is confident of a large award, he will demand a high payout to settle the case in advance. But if the defendant is himself sure of winning, he won’t be willing to pay much to settle the case in advance. Bad news for being able to settle the case in advance.
Relative pessimism occurs when each side is relatively pessimistic about its own chances of winning the case (or the judgment amount in the event of a win). If the plaintiff thinks there is a high chance of losing the case at trial, he will accept even a low settlement. If the defendant thinks the plaintiff has a high chance of winning or that his damages are high, then he will be willing to pay a relatively large settlement to avoid trial. This combination bodes well for being able to settle the case in advance. The maxim is “Bad news is good news for settlements.”
Now, how is this related to information exchange? People tend to voluntarily disclose information that would be bad for the other side. For example, the plaintiff would be happy to reveal private information that his damages are very high or that his chance of winning the case is high. This kind of information disclosure enhances the other side’s pessimism. All in all, voluntary
information disclosure corrects false optimism and encourages settlement.
By contrast, people tend to withhold information that would make the other side more optimistic. For example, a plaintiff probably would not want to inform the defendant that his damages are low
1 In some cases, the law requires one side to answer questions presented by the other side. The rules are complicated,
or that his case isn’t very strong. If he’s forced to reveal this information, it would make the defendant more optimistic about his case, making settlement less likely. Involuntary information exchange enhances the other side’s optimism and reduces the likelihood of a settlement.
To summarize, voluntary pooling of information is unequivocally good. It increases information spread and reduces errors, and it also lowers administrative costs by correcting false optimism and encouraging settlement. Involuntary information exchange is a mixed bag. While it does increase information flow and reduce errors (true of any kind of information), it raises administrative costs because it enhances the other side’s optimism and discourages settlement.
Question: A plaintiff sues a defendant for causing him physical injury. The defendant agrees that he is liable, but there is a dispute about the level of damages. Rule 35 of US Federal Court Procedures allows the defendant to force the plaintiff to undergo a medical examination. A lawyer makes the following remark: “The medical exam could reveal that the plaintiff’s injuries are more serious or less serious than what the defendant thought. Thus, on average, Rule 35 has no impact on the defendant’s settlement offer or on the likelihood of resolving the issue before trial.” Explain the error in the lawyer’s reasoning. Does Rule 35 lead to more settlements or fewer settlements?
Answer: If the information uncovered in the medical exam were favorable to the plaintiff’s case, he would
have disclosed it voluntarily. Thus, if the defendant has to force a medical exam, it is more likely to reveal
something that is unfavorable for the plaintiff’s case. This makes the defendant less willing to settle the case – involuntary information disclosure discourages settlement.
Settlements
In Common Law countries, 95% of cases are settled before trial. Only 5% of disputes ever reach the trial stage. Nevertheless, what would have happened at the trial is of key importance because it sets the threat values for pre-trial bargaining – a trial will happen if a bargain is not reached.
If all parties agreed about the probability of winning, awards and costs, then the Coase Theorem would apply. All cases would be settled before trial, leaving both parties off due to the surplus created by avoiding legal fees at trial. Here is an example.
Question: Farrah has a $300 legal claim against Noah. It would cost each side $50 to hire a lawyer to go to trial. Both sides agree that Farrah has a 40% chance of winning her case. What settlement will Farrah and Noah reach in the reasonable solution?
Answer: Let’s compute the expected payoff to each side from going to trial.
Farrah has a 40% chance of winning $300, a 60% chance of winning nothing and has to pay $50 in fees.
𝐸𝐸Π𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹ℎ= 0.4($300) + 0.6($0)−$50 = $70
Noah has a 40% chance of losing $300, a 60% chance of losing nothing and also has to pay $50 in fees.
𝐸𝐸Π𝑁𝑁𝑁𝑁𝐹𝐹ℎ = 0.4(−$300) + 0.6($0)−$50 =−$170
The reasonable solution is that Noah pays Farrah a settlement of $120, and Farrah drops her case. Both are better off. Farrah ends up with $120 instead of her expected value of $70. And Noah ends up only paying out $120 instead of his expected loss of $170.
The surplus is simple to see in this example – it’s the $100 in legal fees that are avoided by settling before trial. In the reasonable solution, the surplus is split and both Farrah and Noah are $50 better off than they are if they cannot reach a settlement. This is just an application of the Coase Theorem. The settlement replicates the expected outcome of the case, without the legal fees.
A problem arises if Farrah and Noah have different expectations about what is going to happen at trial, in particular if each is relatively optimistic about winning the case. Settlement is still possible with relative optimism as long as it’s not too severe.
Question: In the example above, suppose that Farrah thinks that she has a 60% chance of winning the case, but that Noah thinks that Farrah’s chance of winning is only 30%. Is settlement possible in the reasonable solution? What settlement will Farrah and Noah reach?
Answer: Let’s compute the expected payoff to each side from going to trial, based on each side’s subjective perceptions.
𝐸𝐸Π𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹ℎ = 0.6($300) + 0.6($0)−$50 = $130
𝐸𝐸Π𝑁𝑁𝑁𝑁𝐹𝐹ℎ = 0.3(−$300) + 0.7($0)−$50 =−$140
Farrah expects to gain $130 by going to trial, while Noah expects to lose $140. Thus, in the reasonable solution, Farrah and Noah settle the case in advance for $135.
Even though the actual surplus from avoiding trial is $100, the perceived surplus based on Farrah and Noah’s subjective probabilities is only $10. However, there is still room for a settlement. But this is not the case if relative optimism is too severe.
Question: In the example above, suppose that Farrah thinks that she has a 90% chance of winning the case, but that Noah thinks that Farrah’s chance of winning is only 30%. Is settlement possible in the reasonable solution?
Answer: Let’s compute the expected payoff to each side from going to trial, based on each side’s subjective perceptions.
𝐸𝐸Π𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹ℎ = 0.9($300) + 0.6($0)−$50 = $220
𝐸𝐸Π𝑁𝑁𝑁𝑁𝐹𝐹ℎ = 0.3(−$300) + 0.7($0)−$50 =−$140
Farrah expects to gain $220 by going to trial, while Noah expects to lose only $140. Based on their perceptions of the probabilities at trial, no settlement is possible.
consistency and clarity about court rulings is inefficient because it drags people into disputes. Clear expectations about what will happen at trial makes cases easier to resolve in advance.
Question: Assume that delay is more costly for the plaintiff than it is for the defendant. How does the availability of an appeals system (which delays resolution of the case) impact the bargaining between the two parties to settle out of court?
Answer: The defendant can threaten to appeal a trial verdict if he loses, which reduces the expected payoff of going to trial for the plaintiff. This tilts the bargaining towards the defendant and makes the plaintiff willing to accept a lower settlement to resolve the case before trial.
Trials
Let’s start with a quick comparison of how trials operate in Common Law and in civil law.
Common Law Civil Law
Adversarial process – Judges are neutral referees over process. Each side fashions its own arguments
Inquisitorial process – Judges actively develop questioning and gather evidence.
Courts have general jurisdiction over
many matters Judges have specialized jurisdiction Lawyers prepare cases with witnesses
before trial
Limited contact between lawyers and witnesses before trial
Strict rules of admissible evidence More latitude in what evidence is allowed
In either system, trials are expensive. Legal fees are high, and the trial process often features long delays and uncertainty about the outcome. As a result, as we saw in the last section, it is advantageous and parties will often make efforts to settle their cases and avoid trial completely.
Appeals
Different courts use different rules on whether appeals courts must consider all cases that are filed or whether they can choose which cases to review. An important distinction is that Common Law appeals courts typically review only matters of law and procedure. They do not re-litigate issues of fact. In other words, you can’t appeal in a Common Law court simply because you don’t agree with the decision. You have to cite something specific that was done incorrectly at the first trial. Appeals courts in civil law jurisdictions tend to review cases de novo – from the beginning.
To the first point, appeals courts are obviously more likely to overturn errors than correct decisions. Thus, litigants themselves screen out which cases they are going to bother to appeal. If a litigant knows that a decision is in egregious error, then the litigant is more likely to pursue an appeal because he knows that the initial decision is likely to be overturned. But if the litigant knows that his chances are low, he’s not going to pay the cost of appealing. Thus, appeals courts in effect draw on private information from litigants themselves about which cases involve the most egregious errors. Only certain cases are “screened into” appeals courts by litigants themselves.
Secondly, appeals courts are more likely to screen out inefficient rulings. For example, if a court allocates property to Anna, who values it less then Sara, then the ruling is inefficient and Sara has more incentive to appeal than if the ruling had been efficient in the first place. Inefficient allocations provoke more appeals and provoke greater expenditure on appeals that are filed.
Because inefficient laws are contested more frequently and with greater resources, they are overturned more often than efficient laws are. Thus, the legal system creates an “invisible hand” effect that over time sifts out inefficient laws.
The problem in practice is that the incentive to appeal inefficient laws is not as large as it should be because there are both public and private benefits from the appeal – all society benefits when bad precedent is reversed on appeal. A litigant may not net enough private benefit to justify going through an appeal of an inefficient ruling, but it might have been efficient from a social perspective for him to do so if the public benefits are large enough.
Now, to fix this problem, the state does subsidize the legal system. Lower courts mostly apply law to specific disputes. The cases affect only litigants themselves, and there is no precedent set. Thus, there is no large public goods spillover. By contrast, appeals courts set precedents for interpreting law that apply to all lower courts. So there is a significant public goods effect to appeals court rulings because they set precedent and impact everyone.
So subsidize appeals heavily, right? Well, there’s an obvious problem. If we subsidize appeals too
heavily, people might bring frivolous appeals. We want to provide incentives for appeals specifically for appeals that will result in correcting a mistake or some previously inefficient precedent. One intriguing policy option is to charge a high fee for appeals, but to refund it if the litigant is successful. This provides an incentive for litigants to pursue appeals for judgements that they know to be in serious error.
Role of Judges
present evidence, ask questions and fashion arguments. In civil law courts, by contrast, judges ask questions directly of lawyers and witnesses and try to ferret out testimony on their own. In essence, the Common Law system requires high effort by lawyers but requires less of judges, vice versa for civil law courts.
Which is better? As usual, there are two sides. Lawyers operate in a competitive market and thus have an incentive to pursue the best interest of their clients. Judges are independent – either appointed to long terms or shielded by many layers of bureaucracy.
The benefit of having lawyers present evidence and fashion arguments is that lawyers have a lot to gain by putting high effort into case, while judges do not. On the other hand, lawyers have no incentive to present an unbiased version of the truth since their interest is in protecting their clients. That’s the basic tradeoff. The Common Law system induces more investment in the search for facts and arguments because lawyers make their living in a competitive market. But the civil law system creates more independence in fact-finding because judges do not have a vested interest in a particular outcome.
Market for Legal Services – Supply and Demand
From basic microeconomics, we know that an increase in the supply of something reduces its value. But a casual observation about lawyers is that an increase in the number of lawyers creates more litigation and creates more need for lawyers.2 The resolution to this puzzle is that an increase in the supply of lawyers reduces the price of legal services. As a result, the number of claims rises,
but they are lower-valued claims on the margin. In other words, increasing the supply of lawyers brings cases into the market that wouldn’t have been worth fighting otherwise.
An important point is that the supply of lawyers is severely restricted. In the US, the American Bar Association licenses lawyers, who cannot practice without a license. The degree is very expensive and the exam is difficult.3 In Germany, the law sets a minimum price for legal services, reducing competition on the basis of price.
Market for Legal Services – Moral Hazard
The practice of law is a textbook example of a market that features asymmetric information. The lawyer knows more about the law than you do and can make a much more informed guess about the expected outcome of a case than you can. Furthermore, the outcome is random to some extent, making it difficult for you to determine as a client whether your lawyer did a good job; your lawyer
2 A old joke – There was a lawyer who was all alone in town, and he had no business. Then another lawyer moved
into town and he had all the cases he could handle.
3 Many economists criticize professional licensure that is organized by the profession itself because of the incentive
may lose the case even if he puts in high effort. These substantial information asymmetries can create inefficiencies in the operation of the market for legal services.
Our first question – Do lawyers have an incentive to provide the efficient level of service to their clients? Under the three standard compensation schemes, unfortunately the answer is no.
1. If lawyers are paid by the hour, they have an incentive to spend too much time on each case and invest less effort in finding new cases.
2. If lawyers are paid a fixed fee for particular services, they have an incentive to perform many services and minimize the amount of time spent on each one.
3. If lawyers are paid a contingency fee (a percentage of the judgment in the event of a win), then lawyers underinvest in each case. Why? They only receive a fraction of the rewards created by their additional effort. It might be efficient for the lawyer to put in some more effort to increase the expected judgment in a case by $10,000, but the lawyer may not find it worth it if he can only keep a fraction of this amount.
Our second question – Do lawyers have an incentive to provide honest information to their clients about the expected outcome of the case? Again, under any compensation scheme, the answer is no.
1. If lawyers are paid by the hour or with a fixed fee, they have an incentive to exaggerate the expected value of potential clients’ claims in order to get more cases.
2. If lawyers are paid a contingency fee, the lawyer might not take a case even if it would be efficient to do so because he only keep a fraction of the benefits.
The classic resolution to these moral hazard / incentive issues is to hire lawyers with a reputation and with an interest in maintaining their reputation. If you were a lawyer’s last client and last case, he might not have an incentive to provide the efficient level of service for you. But if the lawyer is interested in keeping his clients, getting good recommendations, or attached to a firm with a valuable name brand, then he has an interest in providing you good service in order to maintain his reputation.
Who should pay legal fees?
In the US, each side pays its own litigation expenses. In the UK, the loser is required to pay his own legal expenses and the legal expenses of the winner. There are fewer disputes in the UK than in the US. Is this the reason?
Theoretically, it turns out that the answer is not quite so simple. On one hand, the UK system discourages plaintiffs from filing frivolous cases that they think they have a low probability of winning, because they would be responsible for the defendant’s legal bills.
On the other hand, what if a case features relative optimism by both sides? Here, the “loser pays” UK system might actually result in more lawsuits because each side expects to be able to shift its legal bills to the other side, and both sides are optimistic about winning.
Overall, despite appealing intuition, it is not at all clear that “loser pays” necessarily results in fewer lawsuits.4
Nuisance Suits
Suppose Anna sues Bob, but Bob’s litigation costs will be significantly higher than Anna’s. The difference tilts the reasonable bargaining solution in favor of Anna because she knows that Bob will be willing to pay a lot to avoid trial.
In such a setting, a plaintiff might even file a claim with an expected value of zero (a guaranteed loser) hoping that the defendant will pay a settlement just to avoid the costs associated with litigation. This is known as a nuisance suit. A real-life example is that a plaintiff might file a case against a construction company – even a case that she expects to lose – because she knows that the construction company has to stop the project while the case is being litigated, and thus has an incentive to settle the case quickly to avoid the costly litigation period.
To illustrate how nuisance suits work, first suppose that both sides have the same litigation costs. Anna and Bob will both lose $1000 if the case goes to court. The case is a guaranteed loser, with no chance that Anna will win any money from Bob at trial. In this case, the threat values if the case ends up in court are 𝐸𝐸Π𝐴𝐴𝐴𝐴𝐴𝐴𝐹𝐹 = −1000 and 𝐸𝐸Π𝐵𝐵𝑁𝑁𝐵𝐵= −1000. In the reasonable solution, Bob need not pay Anna anything to resolve the case – both are $1000 better off if the case is not filed.
But suppose instead that Bob’s litigation costs would be $5000 while Anna’s are only $1000. The threat values if the case ends up in court are 𝐸𝐸Π𝐴𝐴𝐴𝐴𝐴𝐴𝐹𝐹 =−1000 and 𝐸𝐸Π𝐵𝐵𝑁𝑁𝐵𝐵 = −5000. In the
4 One final point – Some people argue that there are serious equity concerns with the “loser pays” system. Not all
losing cases are frivolous cases. Under a loser pays system, a plaintiff with limited economic means might be reluctant
to file a legitimate case against a wealthy defendant if she thinks there is even a chance that she will lose the case, out
“reasonable solution” Bob pays Anna $2000 to drop her case even though both know that Anna’s case is a guaranteed loser in court! It might seem absurd, but Anna and Bob are both better off by reaching a settlement. Anna gains $2000 instead of losing $1000, and Bob loses $2000 instead of losing $5000. In essence, Anna can take advantage of Bob’s high litigation costs and use it as leverage to demand a settlement up front.
However, this is one problem with this tactic. Suppose that Bob is stubborn and outright refuses to pay a settlement. Is Anna really going to follow through on her threat to take Bob to court, knowing full well that she will lose $1000 if she does so? At the same time, is Bob really going to follow through on his threat to not pay a $2000 settlement knowing full well that he will lose $5000 if he does so? It’s a brinkmanship game, and it all depends on who backs down first. Reputations are important. Bob might have an incentive to fight anyway, to scare off future litigants who might try to file nuisance suits. But Anna might have an incentive to go to a losing trial too, in order to convince future defendants to pay settlements or else.
Note that the “loser pays” system would solve this problem, but what about in the US system? Overall, the problem of nuisance suits is really hard to deal with in the US system unless the defendant can somehow screen out who is serious and who isn’t.
Class-Action Lawsuits
A class-action suit occurs when a large group of people with basically identical claims collectively bring their cases to court. For example, suppose that a bank illegally overcharges every customer. One customer might hire a lawyer and attempt to recover on behalf of everyone with the same claim. In Common Law countries, the judge must certify the plaintiffs as a class.
Class-action suits present a tradeoff. On one hand, the benefit of class action is that if many people are involved and the facts for each case are basically identical, there is a large savings in litigation costs by litigating the case only one time. The cost is that class action creates a much larger opening for nuisance suits. Even if the probability of success in any individual case is low, a defendant facing a class action might be pressured to settle it in order to avoid even a small chance of a huge, cataclysmic loss. Plaintiffs’ lawyers would not have this leverage if there were no class action.
A related issue is filtering. Suppose that many people are harmed by a consumer product, with varying levels of harm. The firm offers $1000 to anyone who will settle immediately. If it costs $200 to go to court, then anyone whose damages are less than $1200 will just take the settlement rather than suing. In this way, the settlement is essentially a filter for the less serious claims.
the other hand, a higher settlement offer “overpays” people whose claims would have been worth less than the amount offered. The defendant must balance out this marginal benefit with the marginal cost in choosing the optimal settlement offer.
Does the defendant prefer class action or individual cases in this scenario? Clearly, separating out the cases involves higher litigation costs. On the other hand, the problem with class action is that the defendant loses his ability to filter out frivolous or fraudulent cases. Under class action, the defendant is forced to settle with everyone in order to resolve the case. If cases are filed individually, the defendant can use settlement offers to screen out which plaintiffs are serious.
Joint and Several Liability
Does joint and several liability make settlements easier or more difficult? The intuitive answer is that joint and several liability promotes settling cases before trial. The plaintiff has a lot of leverage against each defendant because he can threaten to sue any one of them for the full amount. These fears make defendants more likely to settle their cases.
But it’s not this simple. Another way of thinking about it suggests that joint and several liability might make settlements more difficult. Consider a plaintiff who is unsure about winning his case. Joint and several liability gives him “insurance” against losing his first case because he could follow it up with a case for the whole amount against the second injurer if he loses the first case. This “insurance” raises the plaintiff’s expected value associated with filing cases and pushes him to demand larger settlements, which makes a successful settlement less likely.
Question: Alisha suffered a $50,000 injury in an accident for which Henry and Matthew are jointly responsible. All parties agree that Alisha has a 60% chance of winning any case she files, and each party incurs $2000 of legal fees for any case that goes to trial.
a. Suppose that Alisha is required to sue Henry and Matthew separately, each for $25,000.
i. What is the smallest settlement Alisha will accept to drop her lawsuit against Henry?
ii. What is the largest settlement that Henry is willing to pay?
iii. Can the case be settled before trial in the reasonable solution? What is the settlement?
b. Now suppose that courts use joint and several liability. Alisha decides to first sue Henry for the
full $50,000. If she loses that case, she will sue Matthew for the whole amount.
i. If Alisha loses her case against Henry, what is the expected payoff from her subsequent
lawsuit against Matthew?
ii. What is the smallest settlement Alisha will accept to drop her lawsuit against Henry?
iii. What is the largest settlement that Henry is willing to pay?
iv. Can the case be settled before trial in the reasonable solution? What is the settlement?
Answer:
a. Several liability
i. Alisha’s expected payoff from going to trial is $13,000, so this is the minimum
settlement that she would accept.
𝐸𝐸Π𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴ℎ𝐹𝐹= 0.6(25,000) + 0.4(0)−2000 = 13,000.
ii. Henry expects to lose $17,000 if they end up at trial, so this is the maximum settlement
that he would pay.
𝐸𝐸Π𝐻𝐻𝐻𝐻𝐴𝐴𝐹𝐹𝐻𝐻= 0.6(−25,000) + 0.4(0)−2000 =−17,000.
iii. In the reasonable solution, Henry pays Alisha a $15,000 settlement to avoid a trial.
b. Joint and several liability
i. 𝐸𝐸Π𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴ℎ𝐹𝐹= 0.6(50,000) + 0.4(0)−2000 = $28,000
ii. Alisha’s expected payoff from her initial lawsuit against Henry is $39,200, which is the
lowest settlement she will accept. The important idea is that, if she loses against Henry, her expected value from that point forward is $28,000, because she would follow up with her suit against Matthew, which provides her with an expected payoff of $28,000.
𝐸𝐸Π𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴ℎ𝐹𝐹= 0.6(50,000) + 0.4(28,000)−2000 = 39,200.
iii. Henry expects to lose $32,000 if they end up at trial, so this is the maximum settlement
that he would pay.
𝐸𝐸Π𝐻𝐻𝐻𝐻𝐴𝐴𝐹𝐹𝐻𝐻= 0.6(−50,000) + 0.4(0)−2000 =−32,000.
iv. No settlement between Henry and Alisha is possible, so a trial is inevitable. Alisha
expects to gain $39,200 by proceeding to trial, while Henry expects to lose only $32,000.
c. In this example, the rule in (a) is more efficient because of the savings in legal costs. In (b),
Alisha’s lawsuit against Henry cannot be settled before trial.
The point of the example is that joint and several liability essentially gives plaintiffs “two bites at the apple” for the full amount of their lawsuit. While the defendant is willing to pay a larger settlement to avoid being sued for the full amount, the plaintiff in turn will demand a larger settlement. Thus, joint and several liability can result in fewer settlements, which increases litigation costs and reduces efficiency.
Burden of Proof
Very few cases are absolutely clear-cut. The objective of courts is to determine the probability that the defendant is guilty / liable based on observable evidence. But do courts apply rules of probability in a consistent way? There are a number of pitfalls.
First is courts’ unwillingness – in total defiance of logic – to accept statistical evidence. Here is an example. Suppose 400 people hold a ticket for a concert but 600 people break in. The policeman arrests an attendee at random and tells a judge there is a 60% chance that he broke in. This is called
sense, both pieces of evidence contain exactly the same information for assessing the defendant’s guilt. Nevertheless, Common Law courts do not allow evidence based on naked statistical reasoning.
The second pitfall is that people not trained in statistics have a great deal of trouble dealing with probabilities of joint events. Here are two examples to illustrate.
Question: A plaintiff alleges that a doctor committed two different errors. The doctor is liable if he committed either error. The probability that the doctor committed error 1 is 0.4 and the probability that he committed error 2 is 0.4. The two errors are independent of each other. Is the doctor liable using a preponderance of the evidence standard?
Answer: Let us consider the four possibilities.
• Pr(Error 1 and Error 2) = 0.4⋅0.4 = 0.16 • Pr(Error 1 but not Error 2) = 0.4⋅0.6 = 0.24 • Pr(Error 2 bur not Error 1) = 0.4⋅0.6 = 0.24 • Pr(Neither error) = 0.4⋅0.4 = 0.36
Thus, the probability that the doctor made at least one error is 0.16 + 0.24 + 0.24 = 0.64 and so he is
liable under a preponderance of the evidence standard (>50%).
Despite the unequivocal mathematical answer, many people would be reluctant to find the doctor liable because the probability for each error is only 0.4. Psychologically, people place more value on one slam-dunk piece of evidence than in a string of less reliable evidence – even if, taken together, the string of many pieces of evidence, is more convincing.
Here is an example to illustrate the converse of this problem.
Question: A plaintiff claims that a doctor committed a series of two errors. The doctor is liable in this case only if he committed both errors. The probability that the doctor committed either error is 0.6, and again the errors are independent of each other. Is the doctor liable using a preponderance of the evidence standard?
Answer: The probability that the doctor committed both errors is 0.6⋅0.6 = 0.36, so the doctor is not liable under a preponderance standard.
Again, jurors or judges without a good understanding of probability might nevertheless be inclined to find the doctor liable since each piece of evidence was separately established with a probability in excess of 50%.
Efficiency of Common Law
Well-known legal scholar Richard Posner, among others, has argued that Common Law is inherently inclined to greater efficiency than civil law.
The basic reason is that Common Law is based on prevailing norms, and economists (and others) have argued for a long time that these norms should move in the direction of greater efficiency. In a model of society evolving over time, people will observe which social norms produce good outcomes for society. These norms will be preserved in societies where they arise and will be replicated and emulated by other societies. Look at how quickly female access to the labor market and vaccination spread once information could flow freely – it was so obvious that these norms produced good outcomes for society that they replicated themselves in many places.
Turning to the legal system specifically, it would be impossible for any judge or scholar to amass all of the information that one would need to write the law in an efficient way that would cover every scenario. In this way, proponents of the superiority of Common Law argue that civil law is inherently flawed. Decentralized decision-making can capitalize on information from everyone in society in the natural evolution of its norms and values, and thus Common Law – which responds to these norms and values – has the potential to be more efficient.
In brief, Common Law is responsive to social norms, which are themselves moving in the direction of greater efficiency in response to changing conditions in society. This is more efficient than civil law, which is centralized and static.
Empirical Evidence
To close, let’s briefly review some empirical evidence on the legal system in the US.
Several million legal claims are filed every year in the US. Until about 20 years ago, breach of contract was the most common kind of case. Tort cases are more numerous now. Cases involving civil rights, intellectual property and labor disputes are also becoming more common. With respect to criminal cases, there has been a sharp increase in the number of cases filed, but a decline in the number of criminal trials. In other words, more and more criminal cases are settled by plea bargain. This pattern is common across many types of cases – more claims, but fewer trials. Why the decline in the number of trials?
• The cost of trials has increased significantly. Controversies are more complex, requiring expensive lawyers and experts.
• Arbitration and mediation is more common.
• Specifically with respect to criminal law, the landscape increasingly features huge sentence reductions for defendants who plead guilty rather than going to trial.
Fewer than 2% of cases ever go to a trial. The vast majority are either settled or are resolved through pretrial motions like a summary dismissal. There are plenty of lawyers around to deal with all this litigation. The US has one lawyer for every 275 people.5 What can we say about the welfare
effects of this abundance of attorneys on society? Lawyers theoretically facilitate transactions, protect property rights and lead to a more efficient allocation of resources, but the practice of law is saturated with market power and moral hazard, which compromises some of these efficiency gains. The market for lawyers in the US is getting more competitive, though.
Litigation can be costly. Filing fees in the US are a few hundred dollars on average, but can rise into the thousands depending on how long the case goes on. Discovery is bimodal – very cheap in most cases, but very expensive in a few cases. However, lawyer fees dwarf all of this and are by far the most expensive component of litigation costs.
A case settled outside of court requires 9 hours of attorney time on average. A trial, by contrast, requires 9 days of attorney time. If we include the value of the attorneys, judges and juries, the total cost of running a trial is about $4000 a day, or $36,000 for the average trial. This is not even including the opportunity cost of time, the cost of maintaining buildings, etc… In this spirit, deterring wrongdoing and making laws more efficient is important because of the substantial social benefit associated with reducing litigation.
As for the outcomes of trials, cases that are very tilted towards the plaintiff or to the defendant tend to be settled out of court. If it is obvious what will happen at the trial, it is easier to settle out of court. Cases that go to trial should be those cases that have closer to even odds, where there is uncertainty about who will win. Precisely, the Priest-Klein prediction is that plaintiffs should win about 50% of trials. This is not to say that 50% of cases are legitimate, because the Priest-Klein prediction is only about cases that go to trial, which are definitely not representative of all claims that are initiated.6 How does the Priest-Klein prediction hold up in practice? For automobile accidents, products liability and intellectual property it seems pretty robust. By contrast, defendants routinely win 60-70% of antitrust and medical liability cases.
Finally, turning to appeals, evidence shows that they are expensive and rare. Filing fees are high and attorneys are typically specialist experts. Because filing an appeal in a Common Law court requires the litigant to prove that there was an error in the lower court, the burden of proof is quite high. Overall, fewer than 0.2% of cases are appealed.
5 In Germany it’s one lawyer for every 622 people and in Japan it’s one lawyer for every 5800 people.
6 Two caveats. First, the Priest-Klein prediction does not hold if the dispute is over the amount of damages rather than