Tort Law

    The economic analysis of tort liability is one of the jewels of law and economics.  As you know, the central economic insight is that exposure to tort liability may induce potential victims and injurers to take actions before an accident occurs that will minimize the social costs of accidents (where those social costs include the precautionary costs of victims and injurers, the accident losses suffered by both parties, and the administrative costs of determining liability).  The following questions, observations, and exercises illustrate some matters beyond those covered in Chapters 8 and 9 in Law and Economics.  

Index to Sections on Tort Law

Liability for Injuries from Guns: A Classroom Exercise

Facts: Manufacturer produced a large-bore pistol and sold it by mail to Buyer, who turned out to be someone with a violent history.  Buyer fell into a rage because Victim would not turn down the volume on his stereo.  Buyer shot Victim, who bled to death.  Victim would have survived a wound in the same place with a small-bore pistol.  Victim's heir sues Manufacturer.

Plaintiff's theories: Plaintiff offers these alternative theories for establishing defendant's liability:

1.  People who engage in ultra-dangerous activities like dynamite blasting or keeping a pet tiger are strictly liable for any harm that results.  Manufacturing and selling guns is an ultra-dangerous activity, so the manufacturer is strictly liable for the harm suffered by the plaintiff. 

2.  Compared to a small-bore gun, a large-bore gun is unreasonably dangerous as measured by the consumer expectations test or the risk-utility test.  The large-bore gun is, consequently, a defective design.  Manufacturers are strictly liable for harm caused by defective designs.  The defect caused the victim's death, so the manufacturer is strictly liable for the harm suffered by the plaintiff. 

3.  The manufacturer did not take any precautions to insure that the buyer was not a dangerous person.  The burden of some precautions by the manufacturer are less than the reduction in expected harm to gunshot victims.  By the Hand Rule, the manufacturer was negligent.  But for the manufacturer's negligence, the injurer might not have obtained a gun.  The manufacturer's negligence caused the victim's death, so the manufacturer is liable under a negligence rule.  

    Have two members of the class argue for "Theory 1" and two argue against "Theory 1."  Likewise for "Theory 2" and "Theory 3."  Each side should use economic analysis to argue about the social consequences of each theory.  

Back to Index of Sections on Tort Law

Joint Care for Accidents

Recall that tort law seeks, among other things, to induce potential victims and injurers to take efficient precaution, so as to minimize the social costs of accidents.  One of the important background assumptions of tort law is that the parties may be unable to bargain so that their precautionary actions must be taken independently and without knowledge of what the other party may be doing.  But there are circumstances in which joint care may be more efficient than independent care by the parties.  

    Consider the following example, from Judge Richard Posner's Economic Analysis of Law 170 (5th ed., 1999):  

“Assume that a railroad wishes to run over an easement though a farmer’s field.  The train throws off sparks, sparks that will ignite the farmer’s crops and cause $150 in damages [each time the train passes through].  The railroad can suppress the sparks entirely, reducing the probability of an accident to zero, by renting a ‘super spark arrester’ for $100 per run.  Moreover, even if the railroad does not employ an arrester, the farmer can eliminate the possibility of accident by leaving a firebreak at a cost of $110.  Finally, the parties can prevent any accident by acting jointly—i.e., the railroad can rent a generic spark arrester for $50 per run, while the farmer leaves a smaller firebreak at a cost of $25.”  

    Can tort law induce the parties to take the joint care option?  

For an interesting attempt to answer that question, see Alan J. Meese, “Injurer Activity and Victim Precaution in a Joint Care Setting: A Correction” (Marshall-Wythe College of Law, College of William & Mary, 1999).   

Back to Index of Sections on Tort Law

Durable and Non-durable Precaution

In an important article, “Why Are People Negligent? Technology, Nondurable Precautions, and the Medical Malpractice Explosion,” 82 Northwestern U. L. Rev. 293 (1988), Mark Grady--now Dean of the George Mason University School of Law--drew a distinction between durable and non-durable precaution.  Consider the farmer-railroad example above.  Suppose that accident costs can be equally minimized by either of two methods of precaution: (1) by digging a firebreak (a durable precaution) or (2) by spraying the at-risk crops with a chemical fire retardant before each passage of a train (a non-durable precaution).  The initial outlay of the durable precaution is much more than that of the non-durable precaution.  However, suppose that the stream of benefits of the durable precaution exceeds that of the non-durable precaution.  How can tort law induce parties to take durable precaution? 

Back to Index of Sections on Tort Law

The McDonald's Hot Coffee Case

Those who contend that the U.S. system of justice is out of control frequently begin with the McDonald's hot coffee case.  According to a common understanding of that case, a woman spilled hot coffee on herself while driving.  She had purchased the coffee at the drive-through window of a McDonald's Restaurant.  She sued McDonald's for compensation for her injuries (about $11,000) and was awarded nearly $3 million in compensatory and punitive damages by a New Mexico jury.  

    So, what's wrong with that?  The most fundamental criticism of the case seems to be that the plaintiff had no strong cause of action.  She sought compensation for something that was almost certainly an everyday accident and well within her control.  There does not seem, at first blush, to be anything about the facts above that indicates fault on the part of McDonald's.  Second, most countries do not allow punitive damages, and foreign students are hard pressed to understand why these facts in this case make out an argument, even by U.S. standards, for punitive damages.  Rather, this seems to be an instance of a local jury piling on damages in favor of a local resident against an out-of-state corporation.  

    But a complete account of the facts in the case may present a slightly different story.  Professor Rick Hasen of Loyola Law School of Los Angeles once circulated this account of the case, based on an article in the news magazine, Newsweek:  

    Stella Liebeck, 79, purchased the coffee and while driving her car, placed the coffee cup between her legs and tried to remove the lid.  The cup spilled and coffee ran into her lap.  Wearing a sweatsuit and sitting in a bucket seat, she received second- and third-degree burns across her buttocks, thighs, and labia. 

    After the spill, Liebeck spent seven days in the hospital and three weeks recuperating at home with her daughter in attendance.  This was followed by skin grafts.  During this period, she lost 20 pounds—to 83 pounds—almost 20% of her body weight. 

    Liebeck wrote to McDonald’s and asked them to turn down the coffee temperature, which was set at 170 degrees.  She also asked for her out-of-pocket medical expenses of about $2,000 plus the lost wages of her daughter.  McDonald’s offered $800.  She sued, asking for no less than $100,000 in compensatory damages, including pain and suffering, and triple punitive damages.  Just before trial, she offered to settle for $300,000, but McDonald’s rejected the offer. 

    The case went to trial in August, 1994.  After the trial one juror said, “I was just insulted.  The whole thing sounded ridiculous to me.”  McDonald’s moved for summary dismissal, defending the heat of its coffee and blaming Liebeck for spilling it.  She was, according to McDonald’s, the “proximate cause” of the injury. 

    Photos were shown of Liebeck’s burned skin, and a burn expert, Dr. Charles Baxter (Southwestern Medical School), testified that 170-degree coffee would cause second-degree burns within 3.5 seconds of hitting the skin.  Christopher Appleton, a quality assurance supervisor at McDonald’s headquarters testified that the company had not lowered the heat under the coffee despite receiving 700 burn complaints in 10 years.  Safety consultant Robert Knall said that 700 complaints was about 1 in 24 million cups and "basically trivially different from zero." 

    A juror's response, "Each statistic is somebody badly burned.  That really made me angry."  The juror also was not impressed with the CAUTION:  CONTENTS HOT label on the cup.  She said she needed her glasses to read it. 

    After four hours of deliberation, the jury found for Liebeck.  She was awarded $200,000 for compensatory damages, reduced by 20 percent because Liebeck had contributed to the accident.  They also awarded $2.7 million in punitive damages.  One juror said, “It was our way of saying, ‘Hey, open your eyes.  People are getting burned’.” 

    Trial Judge Robert Scott reduced the award to $640,000, calculating the punitive damages at three times the compensatory damages.  He stated that it “was appropriate to punish and deter” McDonald’s corporate coffee policy.  Scott, a self-described conservative Republican, says the case “was not a runaway.  I was there.” 

    The two sides ultimately settled for an undisclosed amount. 

    According to Andrea Gerlin’s excellent Wall Street Journal article, “A Matter of Degree: How a Jury Decided that a Coffee Spill is Worth $2.9 Million,” Sept. 1, 1994, p. A1, members of the jury learned at the trial that 180-degree coffee like McDonald’s served may produce third-degree burns in about 12 to 15 seconds.  Lowering the temperature by 20 degrees (to 160 degrees Fahrenheit) would increase the time for the coffee to produce such a burn to 20 seconds.  Those extra 5 to 7 seconds in many cases could provide adequate time to remove the coffee from exposed skin, thereby preventing such burns.  Ms. Gerlin also reported that McDonald's reason for serving such hot coffee in its drive-through windows was that, because those who purchased the coffee typically wanted to drive a distance with the coffee, the high initial temperature would keep the coffee hot during the trip. 

Questions:

1.  If you had been on the jury, how would you have decided this case?  

2.  Is it an adequate defense to Ms. Liebeck's harms for McDonald's to say that the benefits of the hot coffee exceeded the costs, where the costs are measured as the probability of an injury times the medical and other expenses of an accident?  

3.  Could McDonald's avoid liability in subsequent cases by increasing its warning about the temperature of its coffee?  How would you advise them to change their warning?  

Note:

    You might compare this case with McMahon v. Bunn-O-Matic, 150 F.3d 65 (1998).  Jack and Angelina McMahon had stopped at a gas station-convenience store to get coffee.  Later, as they were driving, Angelina tried to pour the coffee from the Styrofoam cup provided at the convenience store into a smaller cup that would, she felt, have been easier for her husband, Jack, to hold while driving.  She was burned severely by the hot coffee falling on her lap.  The McMahons sued the manufacturer of the Styrofoam cup and lid and the manufacturer of the commercial coffee maker.  They settled their complaints against the cup and lid manufacturers but sought to proceed to trial against Bunn-O-Matic.  Their theories were that the manufacturer had a duty to warn consumers of the product that the coffee sold to them would be 180 degrees Fahrenheit and that any coffee served at a temperature of greater than 140 degrees Fahrenheit was "unfit for human consumption" and, therefore, a defective product.  Judge Easterbrook held for the Seventh Circuit Court of Appeals that neither of these theories withstood scrutiny.  

Back to Index of Sections on Tort Law

Punitive Damages

We present an economic discussion of punitive damages on pp. 351-55.  There we argue in favor of the imposition of a "punitive multiple" -- the inverse of the enforcement error -- to correct for shortcomings of the civil liability system.  (Be certain to see A. Mitchell Polinsky & Steven Shavell, "Punitive Damages: An Economic Analysis," 111 Harv. L. Rev. 870 (1998), for a more complete analysis but one that is in the same spirit as ours.)  Our argument in favor of the imposition of punitive damages is clearly one that is against the prevailing grain.  Almost every civil law system prohibits the imposition of punitive damages in civil liability, and in this country there has been a rising chorus of eminent voices calling for the curtailing or ending of punitive damages.  In 1996 the U.S. Supreme Court handed down BMW of North America v. Gore, 517 U.S. 559, which is excerpted on our Cases page on this Web site.  That case created some guidelines for the imposition of punitive damages so as to make them within the "excessive fines" clause of the Eighth Amendment to the U.S. Constitution.  (In BMW a jury assessed the company $2 million in punitive damages for touching up a paint defect on a new automobile delivered to an Alabama purchaser but not acknowledging that paint job to the auto's purchaser.  The U.S. Supreme Court held that the ratio of punitive damages to compensatory damages in that case -- 500 to 1 -- was excessive.)   

    There is some controversy about whether the BMW decision has had or will have any discernible impact on the penchant of U.S. juries to award punitive damages and of lower courts to oversee some of those awards as excessive. An article in the September 13, 1999, Wall Street Journal quoted many practitioners as perceiving that the BMW case had constrained the award of punitive damages.  As evidence they cite a California state court jury's assessing the General Motors Corporation $4.9 billion in punitive damages on behalf of six people burned when the gas tank on their Chevrolet exploded in a collision and the trial court judge's later reducing that amount to $1.2 billion.  Some courts seem to be finding that a ratio of punitive damages to compensatory damages in the range of 5-to-1 or 4-to-1 is acceptable and that anything above that will be reduced.  

    However, in a recent article in The Wall Street Journal ("No End in Sight as Punitive Damages Go Up, Up, Up," March 13, 2000, A47), former U.S. Attorney General Dick Thornburgh argued that BMW has had no discernible ameliorative effect.  A California court recently assessed a punitive award of $4.8 billion dollars against General Motors when a drunk driver struck a GM auto, killing four passengers.  (The amount was later reduced to $1.1 billion.)  In a South Carolina case, a court assessed a $260 million punitive damages award against DaimlerChrysler after an accident in which a mother, driving a DaimlerChrysler product, ran a stoplight and was struck by another car, killing the mother's unbuckled child.  (That award is now on appeal to the Fourth Circuit Court of Appeals in Richmond, Virginia.)  The article goes on to report on other large punitive damages awards, principally, but not exclusively, against automobile manufacturers, that have occurred since 1996  

    In early June, 2000, the Department of Justice announced that later in the Summer of 2000, they would issue a new study of punitive damages.  The study is a joint product of the Bureau of Justice Statistics of the Department of Justice and the nonprofit National Center for State Courts.  The study looked at over 10,000 tort trials held in 1996 in the 75 largest counties in the United States.  Of those 10,000 trials, slightly less than half—about 4,700—were won by plaintiffs, and in only 3.3 percent of those cases won by plaintiffs were punitive damages awarded. 

    One of the most striking findings of the study is that “judges are more than three times as likely as juries to award punitive damages in the cases they hear.”  Bench trials resulted in punitive damages in 7.9 percent of cases; trials to a jury resulted in punitive damages in only 2.5 percent of cases.  The median punitive amount awarded by a judge was $75,000; that awarded by juries was $27,000. 

    Other interesting conclusions of the study: about 85 percent of the cases were tried to a jury; automobile accident claims amounted to almost half of the 10,000 cases, and in those cases, plaintiffs were more likely to prevail (57.5 per cent success) than in any other kind of tort claim; medical malpractice was the third most common type of case tried but the category in which plaintiffs were least likely to succeed (winning 23.4 percent of the time). 

Question:

    What is the most efficient policy toward punitive damages?  Should they be prohibited?  Allowed only on a showing by "clear and convincing" evidence or some higher standard of proof?  Limited to a fixed multiple -- say, two or three -- of the compensatory damages?  Consider your answer from the ex ante point of view of potential injurers, potential victims, actual victims, and actual injurers.  In light of the fact (at least as argued by former Attorney General Thornburgh) that the BMW guidelines have had no discernible impact, how would that affect your argument for or against punitive damages?  

Back to Index of Sections on Tort Law

Split-Award Statutes and Punitive Damages

In an interesting and important new manuscript, Professor Andrew Daughety and Jennifer Reinganum of Vanderbilt University -- "Found Money?: Split-Award Statutes and Settlement of Punitive Damages Cases" -- consider the impact on a variety of litigation and settlement issues of state statutes that require plaintiffs to give a portion of any punitive damages award to the state government.  The (economic) theory that might be invoked to justify this sharing is that punitive damages are appropriate in circumstances in which there is a social harm beyond the harm that has been inflicted on identifiable persons and property and that, therefore, the state should receive some "compensation" for that social harm in the form of punitive damages.  Presumably, the plaintiff ought to be allowed to keep a portion of the punitives in order to create an incentive for plaintiffs in civil cases to undertake to ask for punitives where they are appropriate.  

    Eight states -- Georgia, Illinois, Indiana, Iowa, Kansas, Missouri, Oregon, and Utah -- currently require that a portion of the punitive damages be given to some department of the state government.  All of these states but Missouri require "clear and convincing" evidence of gross, willful, or wanton behavior before the assessment of punitive damages.  The statutes differ in other interesting ways.  For example, half of the eight make the allocation to the state "net of (plaintiff's) attorney fees, and half do not.  Three of the states receive 50 percent of the punitive damages award; one receives 60 percent; three receive 75 percent; and in Illinois the split is at the "judge's discretion."  Several of the states put the amount into general revenues, but most designate a particular organ of state government as the recipient -- for instance, the Department of Human Services (Illinois), the Violent Crime Victims' Compensation Fund (Indiana), the Civil Reparations Trust Fund (Iowa), the Health Care Stabilization Fund (Kansas), and the Tort Victims' Compensation Fund (Missouri).  

    Professors Daughety and Reinganum use a sophisticated model of litigation versus settlement decisions to find that "the amount of settlement is increasing in the likelihood of the defendant being found liable, in the size of both the compensatory and punitive damages, and in the share of the punitive damages award that the plaintiff may keep.  ...  [S]plit-award statutes simultaneously lower settlement amounts and the likelihood of trial, as both [plaintiff and defendant] seek to cut out the State (since the State's share only arises post-trial.  We also find that (for all states but one) the revenue-maximizing sharing rate is 50 percent (for the remaining state, the revenue-maximizing share should be approximately [two-thirds]) [and that these] results are robust to variations in economics parameters and to whether the State's share is gross or net of the plaintiff's attorney's fee."  They also use the model to draw inferences about the motivations of the states in enacting split-award statutes, finding that some appear to have been motivated by deterrence reduction and others by revenue enhancement.  

For a copy of the paper, please contact Professors Daughety and Reinganum at the Department of Economics, Vanderbilt University, Nashville, TN 37235.  

Back to Index of Sections on Tort Law

Coyote v. Acme

One of the most delightful criticisms of the U.S. products liability system is a short story by Ian Frazier entitled "Coyote v. Acme."  It appears in Frazier's collection Coyote v. Acme (1996).  We wish that we could include the entire story here, but to do so would clearly violate the copyright laws.  The story concerns an action brought by the hapless Wile E. Coyote.  As any viewer of the famous Roadrunner cartoons knows, Mr. Coyote has been trying to capture the Roadrunner through a series of ever-more-complicated stratagems.  None of them works, and they frequently backfire on Mr. Coyote, causing him harm.  Frazier's story is the account of the plaintiff's pleadings in Wile E. Coyote v. Acme Company.  Here are the first two paragraphs and Mr. Coyote's prayer for relief in the final paragraph:  

"Opening Statement of Mr. Harold Schoff, attorney for Mr. Coyote: My client, Mr. Wile E. Coyote, a resident of Arizona and contiguous states, does hereby bring suit for damages against the Acme Company, manufacturer and retail distributor of assorted merchandise, incorporated in Delaware and doing business in every state, district, and territory.  Mr. Coyote seeks compensation for personal injuries, loss of business income, and mental suffering caused as a direct result of the actions and/or gross negligence of said company, under Title 15 of the United States Code, Chapter 47, section 2072, subsection (a), relating to product liability.  

    Mr. Coyote states that on eighty-five separate occasions he has purchased of the Acme Company (hereinafter, "Defendant"), through that company's mail-order department, certain products which did cause him bodily injury due to defects in manufacture or improper cautionary labeling.  Sales slips made out to Mr. Coyote as proof of purchase are at present in the possession of the Court, marked Exhibit A.  Such injuries sustained by Mr. Coyote have temporarily restricted his ability to make a living in his profession of predator.  Mr. Coyote is self-employed and thus not eligible for Workmen's Compensation. ...  

[A hilarious description follows of Mr. Coyote's misadventures with Acme's Rocket Sled, "Little Giant" Firecracker, Self-Guided Aerial Bomb, and Acme Spring-Powered Shoes.]  

    Mr. Coyote respectfully requests that the Court regard these larger economic implications and assess punitive damages in the amount of seventeen million dollars.  In addition, Mr. Coyote seeks actual damages (missed meals, medical expenses, days lost from professional occupation) of one million dollars; general damages (mental suffering, injury to reputation) of twenty million dollars; and attorney's fees of seven hundred and fifty thousand dollars.  Total damages: thirty-eight million seven hundred and fifty thousand dollars.  By awarding Mr. Coyote the full amount, this Court will censure Defendant, its directors, officers, shareholders, successors, and assigns, in the only language they understand, and reaffirm the right of the individual predator to equal protection under the law."  

    Don't miss this delightful piece.  

Back to Index of Sections on Tort Law

A Game Theoretic Analysis of Tort Liability Rules

One of the discernible trends in law and economics -- and in Law and Economics -- is the greater use of game theory.  For example, our discussion of contract law in Chapter 6 centers on the use of legal rules to induce players in an agency game to adopt socially optimal, dominant strategies in concluding a consensual agreement.  At the time we wrote the third edition there were a few game-theoretic treatments of tort liability issues, but they were mostly in the professional literature and not readily accessible to students and others just learning  law and economics.  (See, for example, Daniel Orr, "The Superiority of Comparative Negligence: Another Vote," 20 J. Legal Stud.119 (1991), and Tai-Yeong Chung, "Efficiency of Comparative Negligence: A Game Theoretic Analysis," 22 J. Legal Stud. 395 (1993).)   

    In a recent, excellent book -- Game Theory and the Law (1995) -- Professors Douglas Baird, Robert Gertner, and Randal Picker, all of the University of Chicago, introduce some of the basic concepts of game theory through the device of considering a tort liability situation: motorist-pedestrian accidents.   We'll use their discuss to introduce the game theory of tort liability rules.  We assume that you are at least minimally familiar with game theory.  

    Baird, Gertner, and Picker (BGP) -- pp. 9 - 23 -- assume that an accident might occur between a motorist and a pedestrian and that, if it does, the pedestrian will suffer $100 in losses.  The accident is certain to happen if neither party takes precaution.  But even if they both take care, there is still a 0.10 probability of the accident's occurring.  Each party can take care for an expense of $10.  Under these assumptions, let us see how each party will behave under different legal regimes.  

    First, consider how the parties will behave in a legal regime in which the motorist will never be held liable for the the accident losses, as represented in the following table:    

Figure 1.1  Regime of No Liability.  Payoffs: Pedestrian, Motorist

Source: Baird, Gertner, & Picker, Game Theory and the Law 10 (1995).

    Notice that the Motorist has a dominant strategy -- namely, taking no care, under which he will always have lower expenses than under taking due care.  The Pedestrian has an iterated dominant strategy -- that is, a strategy that is dominant, given that the other player has a dominant strategy.  He knows that the Motorist's dominant strategy is to take no care.  Given that the Motorist will take no care, the Pedestrian's best strategy is also to take no care.  Therefore, the combination in the upper-left hand corner of the payoff matrix is the one that will result.  There will be an accident; it will inflict $100 in losses on the Pedestrian; neither the Motorist nor the Pedestrian will take care.  

    Now imagine a different legal regime -- one in which the Motorist is liable for the Pedestrian's losses if there is an accident.  BGP refer to this regime as "pure strict liability."  The payoffs to the parties under this regime are given in BGP's Figure 1.2:  

Figure 1.2  Pure strict liability.  Payoffs: Pedestrian, Motorist.  

Source: Baird, Gertner, & Picker, Game Theory and the Law 15 (1995).

    Note that the payoffs in the upper-left hand corner of this figure are exactly the opposite of those in the previous figure.  The reason is that under pure strict liability if there is an accident and the Pedestrian suffers $100 in losses, the Motorist is liable for those losses.  Note further that the Pedestrian now has a dominant strategy: never take care.  Given that fact, the Motorist has an iterated dominant strategy: take no care.  

    Now, consider a third legal regime -- negligence with contributory negligence.  Under that liability standard, as we know from Chapter 8, the Motorist is liable if an accident occurs and he failed to take care and the Pedestrian did take care.  The anticipated payoffs to the parties under this regime are given in the next payoff matrix:  

Figure 1.3 Negligence with contributory negligence.  Payoffs: Pedestrian, Motorist.  

Source: Baird, Gertner, & Picker, Game Theory and the Law 17 (1995).

    The Pedestrian has a dominant strategy.  Regardless of the action taken by the Motorist, the Pedestrian should take due care.  Given that fact, the Motorist has an iterated dominant strategy of also taking due care.  

    Now consider a fourth legal regime -- strict liability with a defense of contributory negligence.  Under this regime if there is an accident, the Motor is liable unless the Pedestrian failed to take due care.  The payoffs are given in this table:   

Figure 1.4  Strict Liability with contributory negligence.  Payoffs: Pedestrian, Motorist.  

Source: Baird, Gertner, & Picker, Game Theory and the Law 18 (1995).

    Again the Pedestrian has a dominant strategy: take due care.  Given that fact, the Motorist, too, has a dominant strategy: take due care.  In contrast to the outcome under the "negligence with contributory negligence" standard, however, here the Motorist can anticipate paying for the Pedestrian's losses even when both of them take due care. 

    One of the most interesting aspects of Baird, Gertner, and Picker's game theory analysis of tort liability is there consideration of comparative negligence.  Recall that under that tort liability standard, the Motorist will be liable if there is an accident and he fails to take due care and the Pedestrian does take due care.  If the Motorist takes due care and the Pedestrian also takes care and there is an accident, then the Pedestrian bears the $100 losses from the accident.  If an accident occurs and neither party has taken due care, then the Pedestrian's losses are apportioned between the Motorist and the Pedestrian.    

    To consider the incentives for care created by this system, Baird, Gertner, and Picker add a third option to the two that have figured heretofore in their analysis.  In addition to taking no care or due care, the Pedestrian and the Motorist can take "some care."  They assume that due care costs each of them $3 and that some care costs each of them $1.  They also assume that when both parties exercise due care, the probability of an accident's occurring drops to one in fifty or 0.02.  Thus, the expected cost of the accident to the Pedestrian, when both parties exercise due care is $2 ($100 x .02).  

    Now we come to the loss-allocation rules.  If there is an accident and the Pedestrian suffers $100 in losses and neither party has taken any care, then the Pedestrian's losses are split 50-50: each party bears $50 in the loss.  (This is the entry in the upper left-hand corner of the matrix below.)  If there is an accident and each party has take some care, the losses are again split 50-50.  Because some care costs $1, each party will incur a $50 loss plus the $1 loss of purchasing some care, for a total loss of $51.  (This is the entry in the middle of the matrix below.)  Suppose that the Motorist exercises some care; the Pedestrian exercises no care; and there is an accident.  In that case BGP assume that the Motorist will pay $1 of the Pedestrian's losses and also incurs the $1 cost of paying for some care, for a total loss of $2.  The Pedestrian spent nothing on care and bears $99 of his $100 in losses.  (This is shown in the middle entry of the top row of the matrix.  The opposite case -- where the Motorist takes no care, and the Pedestrian takes some care -- has opposite entries in the matrix.)  Finally, if both the Motorist and the Pedestrian exercise due care, then each incurs the $3 cost of due care, and the Pedestrian additionally incurs the expected losses ($2) of an accident.  BGP summarize all of this in the matrix below:  

Figure 1.5  Comparative negligence (skewed sharing rule).  Payoffs: Pedestrian, Motorist.  

Source: Baird, Gertner, & Picker, Game Theory and the Law 21 (1995).

    How will the Pedestrian and the Motorist behave when faced with this game?  BGP begin with the observation that the incentives facing the parties in this legal regime are not as clear-cut as in the previous regimes.  Neither party has a dominant strategy.  Notice, for example, that some care is usually worse than due care but that is not always the case.  There are circumstances when some care is better than due care -- specifically, when the other party plays no care.  But it is a close call.  

    Let's try to simply things.  Suppose the Pedestrian plays no care.  The Motorist can play some care and have a loss of $2 or play due care and have a loss of $3.  Although strictly speaking the Motorist should play some care, let us assume that if the Pedestrian plays no care, the Motorist plays due care.  The same will be true, by assumption, if the Motorist plays no care; then, the Pedestrian will play due care.  This simplifying assumption makes the game a 2 x 2 matrix instead of a 3 x 3 matrix.  The parties' choices are really between some care and due care.  And under that assumption due care dominates some care for both parties.  

    Baird, Gertner, and Picker go on to explore some more complex liability apportionment rules in comparative negligence.  They make several points that are well worth noting.  First, one of the problems with the set up in Figure 1.5 above is that neither the Motorist nor the Pedestrian is systematically rewarded for making additional investments in precaution.  In an ideal liability regime, an additional dollar invested in precaution should always reduce expected liability by more than a dollar up to the point of socially optimal precaution.  At that point an additional dollar invested in precaution will reduce expected liability by exactly one dollar.  Can you show why the set up in Figure 1.5 violates this principle?  

    Second, Baird, Gertner, and Picker specify a loss-apportionment rule in comparative negligence that comports with the central principle of the last paragraph.  Here is their rule:  

A party who fails to exercise due care should bear the liability in proportion to the amount that party failed to spend on due care relative to the amount both parties fell short of exercising due care. (p. 29)  

    To see how this would work, return to Figure 1.5.  Suppose the Pedestrian took some care; the Motorist took no care; and there is an accident, causing $100 in losses to the Pedestrian.  The Pedestrian spent $1 on care when he should have spent $3.  The Motorist spent nothing when he should have spent $3.  Together the parties should have spent $5 more than they, in fact, spent.  The Pedestrian should have spent 2/5 of that total shortfall, and the Motorist should have spent 3/5.  Therefore, the Pedestrian should bear $40 of his own losses, and the Motorist should pay for the remaining $60.  Note that the rule does not allocate liability when both parties exercise due care.  Can you show that this allocation rule creates a dominant strategy for each player to exercise due care when the other party is exercising less than due care?  

    We highly commend this excellent book -- Baird, Gertner, and Picker, Game Theory in the Law (1995) -- in its entirety.  

Back to Index of Sections on Tort Law

Products Liability Heats Up Again

    The New York Times recently reported that after several years of relative quiet products-liability awards have increased significantly in the last six years.  (See Greg Winter, "Jury Awards Soar as Lawsuits Decline on Defective Goods," New York Times, A1, col. 4 (Jan. 30, 2001).)  

    The median award (exclusive of any punitive damages) for a successful plaintiff in a products-liability action has "more than tripled since 1993" -- rising from $500,300 in 1993 to over $1.8 million in 1999.  These data result from a study of 2,751 products-liability verdicts by LRP Publications of Horsham, Pennsylvania.  That group reports further that the greater part of the rise occurred in the period 1996-1999 and that the rate of increase in those three years was the highest in two decades.  

    Why has the increase occurred?  The Times speculates that the principal reason is that the quality of the products-liability actions that are being brought has increased.  This has occurred because a "series of Supreme Court rulings from 1993 to 1999, embraced by defense lawyers as a way of knocking out lawsuits against companies, have raised the bar on who can testify as expert witnesses, the backbone of any product-liability trial."  As a result, the costs of mounting a defective-product claim have risen; plaintiffs' lawyers are, therefore, proceeding to trial only with the strongest cases in which the stakes are so large that they justify the higher costs of products-liability litigation.  Some lawyers told The Times that the minimum cost of bringing a defective-product action is now $100,000.  

    Although most plaintiffs bring product-injury cases in state courts, there is some evidence to support the higher cost-higher quality hypothesis from looking at the number of products-liability cases filed in federal courts.  "[T]he number of product-liability cases filed in federal court has dropped by more than half the last four years, from 32,856 in 1997 to 14,428 in 2000, according to the Administrative Office of the United States Courts."  Further evidence that the quality of those cases that are brought has increased comes from LRP's finding that the success rate of plaintiffs in products-liability cases has risen from 39 percent in 1993 to 46 percent in 1999.  

Back to Index of Sections on Tort Law

© Copyright 2000, Robert D. Cooter & Thomas S. Ulen.  All rights reserved.