Punitive Damages in Financial Injury Verdicts
Punitive damages lie at the heart of the controversy about the civil justice system and have been the focus of reform efforts at both state and local levels of government. Many argue that punitive damages have become an unpredictable feature of the legal landscape, imposing burdens on business that are out of proportion to the alleged wrong-doing. Others counter that punitive damages are necessary to punish and deter egregious behavior and that amounts awarded are reasonably related to corporate behavior.
Previous research conducted by the Institute for Civil Justice (ICJ), based on jury verdicts from 1985 to 1994 in 15 jurisdictions across the country, documented that, in general, punitive damages are rare events. They are awarded in less than 4 percent of all verdicts; however, the average punitive award amount was more than $1 million in many of the jurisdictions examined (Moller, 1996).
This earlier research also established that within certain case types, punitive damages occur relatively more often. In particular, almost half of all punitive awards were made in cases in which the damages were financial, rather than personal in nature. These verdicts, which we call financial injury verdicts to distinguish them from personal injury verdicts, comprise disputes arising from contractual or commercial relationships including, for example, disputes arising from insurance or employment contracts or from unfair business practices.
Little detailed information has been available about financial injury cases. To provide an empirical basis for the ongoing debate about punitive damages, we draw on the ICJ's jury verdict database to describe the number of punitive damage awards in financial injury cases in selected jurisdictions during the period 1985 to 1994 and to examine patterns in these awards. We provide a separate analysis of jury verdicts reached from 1992 to 1997 from all of Alabama. Because caps on punitive damages are a prominent feature of many reform proposals, we also estimate what percentage of the financial injury punitive awards in our database would have been affected by caps of various sizes and how the caps would have affected the total amount of punitive damages awarded in such cases.
Most civil disputes are resolved by the parties without recourse to a lawsuit, and most of the civil cases in which a lawsuit is filed never reach a jury. However, even though they are rare, jury verdicts are important. Juries decide cases totaling billions of dollars annually. But the dollars that are awarded by juries are only one aspect of the influence of jury verdicts. Although the award is important to the parties in a case, the signals that awards send to the rest of society may have a far greater influence on our nation's social and economic well being.
Jury verdicts have both a deterrent and a shadow effect. The jury's decision in any particular case indicates the potential costs of engaging in behaviors similar to the defendant's. Jury verdicts thus set standards that can influence behavior and, by deterring unduly risky behavior, affect the financial and personal risks we impose on each other. At the same time, the jury's decision in any particular case indicates the potential outcome of any similar dispute. Jury verdicts thus cast shadows that influence claiming and settlement behavior, ultimately affecting the mix of cases that future juries will see and, coming full circle, the outcomes and precedents that will guide system participants in the future.
Legal theorists have observed that awards of compensatory damages have both deterrent and shadow effects. In principle, individuals and organizations involved in activities that can result in harm can learn from their own experiences and from those of other similarly situated parties both the likelihood that they will be found liable for someone's losses and the likely size of those losses. They can adjust their behavior with a view toward the expected cost should someone claim that they are liable for harm. And, if a claim is brought, those expectations provide the parties a basis for negotiating a reasonable settlement.
But the deterrent and shadow effects of punitive damage awards may be far stronger and, thus, more significant than the corresponding effects of compensatory awards. Punitive damages are designed to punish a defendant for grossly inappropriate actions and, in so doing, to deter future such actions by signaling that their consequences can be severe. Because punitive damages are awarded in a fraction of all verdicts, they are less frequent, and thus less predictable, than compensatory awards. And, because punitive damages can be many times the compensatory award (although some states have imposed limits on punitive damages in some types of cases), their size is less predictable. Individuals and organizations may find it more difficult to develop expectations as to both the kinds of behaviors that will result in a punitive award and the amount of any such award.
The effects of punitive awards on individuals' and organizations' behaviors will depend on how they weigh the uncertainties of a punitive award. Critics of the current system, for example, argue that the risk of a very large punitive award sometimes drives defendants to settle cases in which they believe the claim is not meritorious, or to settle meritorious claims for far too much.
How are business decisionmakers and litigants likely to respond to the risks of punitive damage awards? The literature on risk perception and management in business decisionmaking suggests that in assessing risks, most business decisionmakers focus on worst-case scenarios and will go to great lengths to avoid exposing their companies to very large financial losses or potential bankruptcy (March and Shapira, 1987)
Previous ICJ work (Garber, 1993) emphasizes that the nature of the worst-case scenario differs across industries. For example, for medical products, mass torts are very salient, and companies could be liable for punitive damages in many different cases. In contrast, in the automobile industry, the worst-case scenario might involve adverse publicity triggered by punitive damage awards, consequent loss of sales and company reputation, and responses of safety regulators. Punitive awards are important from this perspective because the presence of a punitive award increases the likelihood that newspapers and other media will report on a verdict and trigger such adverse publicity (Garber, work in progress).
Because there are differing perspectives about which features of punitive damage awards are most likely to influence decisionmakers and litigants, we report multiple measures of punitive damage awards. The mean award may be important because it is, in a statistical sense, the amount a defendant could expect to pay if a jury made a punitive award. The median award is important because the odds are exactly equal that a given punitive award will be greater than or less than the median punitive award. The 90th percentile award captures the worst-case scenario.
The ICJ's jury verdict studies are supported by a unique jury verdict database constructed by the Institute over the past 15 years. This study focuses on data collected from 1985 to 1994 in jurisdictions constituting a diverse sample of geographical locations, population, growth, and income. The jurisdictions include all state trial courts of general jurisdiction in the states of California and New York; Cook County, Illinois (Chicago); the St. Louis, Missouri, metropolitan area; and Harris County, Texas (Houston). The states in which these jurisdictions lie differ on important legal standards relevant to punitive damage awards, including limits on punitive awards, specification of the type of behavior that can warrant a punitive award, and different burdens of proof that must be met for a punitive award to be made. Variation in laws between states may explain some of the variation in jury verdict outcomes that we observe.
For all financial injury verdicts in which punitive damages were awarded, we identified the particular type of dispute that led to the punitive damage award and added information about the parties, their relationship to each other, and the industrial sector in which the dispute arose.
We supplement data from these jurisdictions with information obtained from the Administrative Office of the Alabama Courts for verdicts reached in that state's trial courts of general jurisdiction during the period 1992 to 1997. Because these data differ in important ways from the data in the ICJ database, we report the Alabama data separately.
Overall, about one-quarter of the U.S. population (1995) lives in the jurisdictions included in this study.
It is important to keep the following facts in mind when interpreting jury verdict data:
- Cases tried to verdict may not be representative of all claims filed. Many claims are settled before reaching trial.
- The pattern of civil jury outcomes in any year or jurisdiction reflects the mix of cases tried to verdict in that year or jurisdiction, as well as jury decisions. The mix of cases may reflect changes in court jurisdiction, legal rules, and system user behavior.
- A substantial fraction of jury awards are reduced after verdict by trial or appellate court action or by settlement.
- We cannot assume that the patterns observed in one jurisdiction will be replicated in any other specific jurisdiction.
To put our discussion of financial injury verdicts in perspective, we first analyzed punitive awards in all verdicts in California; Cook County, Illinois; Harris County, Texas; metropolitan St. Louis; and New York. We found patterns of punitive damages similar to those that emerged in our prior studies. Of the approximately 1,300 punitive damage verdicts in the database, about 50 percent occurred in financial injury verdicts.
To describe trends in financial injury punitive damage awards, we used the following descriptive measures:
- The number of punitive damage awards, the percentage of all verdicts (including defense and plaintiff victories) in which punitive damages were awarded, and the proportion of the total amount of damages awarded in all cases that was punitive.
- The punitive award amount: The mean award (arithmetic average) and the median award (typical award because half the award amounts lie above and half below it) to provide measures of central tendency, and the 90th percentile (10 percent of the awards lie above this amount) to provide information regarding extremely high award amounts.
- The relationship between amount of compensatory damages and amount of punitive damages because the policy debate often focuses on this relationship. We calculate the ratio of punitive damages to compensatory damages for each case. A ratio of 1.0 means that punitive and compensatory damages were equal for that case. A ratio of greater than 1.0 means that punitive damages were higher than compensatory damages for that case. A ratio less than 1.0 means that punitive damages were lower than compensatory damages for that case. We focus on the median of the distribution of all these ratios.
To link our analysis more closely to the policy debate over punitive damages, we categorized the financial injury cases in our database to reflect case types frequently mentioned in that debate. The categories we used in our analysis are shown in Table 1.
Types of Disputes in Our Sample
|Types of Dispute ||Definition |
|Insurance ||Disputes involving the existence, interpretation, or performance of an insurance contract. |
|Disputes arising out of an employee-employer relationship. |
|Disputes arising out of the existence of a security instrument, including stocks, bonds, and other instruments of finance or ownership. Includes shareholder derivative suits. |
|Disputes arising out of the sale, lease, or improvement of real property. |
|An aggregate of many types of contractual disputes other than those identified above. |
|Financial injury cases arising out of noncontractual relationships between the parties. Largely antitrust and unfair business practice. |
As Table 2 shows, in the jurisdictions we analyzed, punitive damages were awarded in 14 percent of all financial injury verdicts. Most of the punitive awards occurred in insurance, employment, and real property disputes.
Variation in Punitive Awards in Financial Injury Case Types (1985 to 1994)
|Punitive Damage Awards ($1992)|
|Type of Dispute ||Number of Punitive Awards ||Punitive Awards as % of Number of Verdicts ||Mean||Median||90th Percentile|
|Insurance ||134 ||13 ||7,933,676 ||652,000 ||13,572,000 |
|Employment ||125 ||17 ||2,689,033 ||194,180 ||2,060,200 |
|Securities ||6 ||21 ||30,269,389 ||1,229,080 ||174,342,000 |
|Real property ||113 ||12 ||2,110,888 ||94,700 ||2,048,000 |
|Other contracts ||258 ||15 ||6,283,804 ||277,875 ||8,423,360 |
|Other commercial ||11 ||36 ||1,654,966 ||956,470 ||3,370,840 |
|Overall ||647 ||14 ||5,344,876 ||250,000 ||6,223,400 |
Because there are so few cases in the securities and other commercial categories, we do not believe statistics for these case types necessarily reflect cases not in our database.
Punitive damage award amounts were often high in these cases. The mean punitive awards varied from $2.1 million to $7.9 million, with an overall mean of $5.2 million. (We exclude securities and other commercial cases because we have so few data points.) The overall 90th percentile award amount was $6.2 million.
Given these large award amounts, it is not surprising that punitive damages represent a large portion of the total amount of damages awarded in these case types. As Figure 1 shows, punitive damages represent more than half of all the damages awarded overall, including those cases in which there was no punitive award and more than 60 percent in insurance and securities cases.
Figure 2 shows the median of this ratio for each type. The highest median ratio is found in insurance verdicts, where punitive awards are almost four times compensatory awards. The overall median ratio is 1.4.
Figure 2--Median Ratio of Punitive Award to Compensatory Award
VARIATION IN PUNITIVE AWARDS BETWEEN JURISDICTIONS
As Table 3 shows, punitive damages are awarded more often in California and Harris County than in the other jurisdictions in our database. Punitive damages are awarded in about 20 percent of all financial injury verdicts in California and in 14 percent of all financial injury verdicts in Harris County.
|Punitive Damage Awards ($1992)|
|Number of Punitive Awards ||Punitive Awards as % of Number of Verdicts||Mean||Median||90th Percentile |
|California ||429 ||21 ||5,844,685 ||355,800 ||7,210,700 |
|Cook County ||17 ||4 ||3,233,558 ||250,000 ||3,000,000 |
|Harris County ||130 ||14 ||6,669,497 ||206,314 ||8,335,000 |
|St. Louis Metro. area || |
|New York ||16 ||4 ||567,579 ||55,127 ||3,755,500 |
|Overall ||647 ||14 ||5,344,876 ||250,000 ||6,223,400 |
Award amounts vary across jurisdictions. The mean punitive damage award is considerably higher in California, Cook County, or Harris County than in New York or the St. Louis metropolitan area. And in the former three jurisdictions, punitive damages represent more than half of total damages awarded.
Although the ratio of punitive award to compensatory award also varies across jurisdictions, the pattern of variance is different (Fig. 3). The St. Louis metropolitan area, which has the lowest median, mean, and 90th percentile punitive awards, has a higher median ratio of punitive award to compensatory award than either California or Harris County.
Figure 3--Median Ratio of Punitive Award to Compensatory Award
in Different Jurisdictions
We examined the entire population of financial injury verdicts in our database for each of the five-year periods 1985-1989 and 1990-1994. The number of punitive awards has decreased between these periods, both in absolute numbers and as a percentage of the overall number of verdicts. Punitive damages were awarded in about 16 percent of all financial injury verdicts in the 1985-1989 period and in about 13 percent of all financial injury verdicts in the 1990-1994 period. This change reflects the facts that plaintiffs are winning at a slightly lower rate, and given that they have won the case, plaintiffs are also being awarded punitive damages at a slightly lower rate as well.
However, the mean award amount increased from $3.4 million to $7.6 million between these two periods. In addition, punitive damages represent a larger portion of all damages awarded, rising from about 44 percent of all damages awarded in the 1985-1989 period to slightly less than 60 percent of all damages awarded in the 1990-1994 period.
In contrast, the median ratio of punitive damages to compensatory damages fell over the two periods from 1.5 to 1.2, indicating that there is an increase in the number of punitive damage awards in which the amount of punitive damages awarded is small relative to the compensatory damages awarded.
In the technical volume that is a companion to this executive summary, we provide more detailed analyses of punitive awards in financial injury cases. In particular, we describe trends in punitive awards in the four types of financial injury cases in which these awards occur most frequently--insurance, employment, real property, and other contracts. We examine patterns in punitive awards according to the legal theory employed in these disputes. And we compare differences in awards depending on whether the plaintiff was an individual or a government, business, or other entity.
Many states have already approved caps on punitive damage awards, and similar measures are being considered in other states and at the federal level. To provide some context for the policy debate, we estimated what the effects would have been of imposing caps on the existing financial injury punitive awards in our database from California, Cook County, Harris County, metropolitan St. Louis, and New York.
The caps we analyzed are different multiples of the compensatory damages awarded in the case. We chose multiples of one through five and ten--an array of proposals that spans legislative efforts in many states.
If punitive damages had been capped at the amount of compensatory damages in each case, 60 percent of all punitive awards would have been affected, and the total amount of punitive damages awarded in these cases would have been reduced by roughly 65 percent. If caps had been imposed at higher levels, fewer awards and a smaller percentage of the damages awarded would have been affected. For example, a cap of three times compensatory damages would have affected the punitive damage award in about one-third of all the financial injury punitive awards and decreased the total amount of punitive damages awarded by 40 percent.
Table 4 displays our estimates of the effects of caps on the punitive damage awards in our database.
Effect of Caps on Punitive Damages in Financial Injury Verdicts
|Level of Limit |
(multiples of compensatory damages)
|Number of Punitive Awards Affected ||Percent of Punitive Awards Affected || |
Decrease of Aggregate Total Award (percent)
|Decrease of Aggregate Punitive Award|
|1 ||386 ||60 ||43 ||66 |
|2 ||280 ||43 ||34 ||51 |
|3 ||219 ||34 ||27 ||40 |
|4 ||184 ||28 ||22 ||33 |
|5 ||168 ||26 ||19 ||27 |
|10 ||102 ||16 ||10 ||12 |
Were such caps to be imposed, the future experience in the states in our database would not necessarily reflect these estimates. Legislation imposing caps would also affect claiming and settlement behavior. In addition, if juries were aware of caps, they might take limits on punitive damages into their calculations of compensatory damages.
The incidence and size of punitive damage awards in Alabama has figured prominently in the national debate over punitive damages, but heretofore little systematic information has been available about such awards in that state. We analyzed data describing verdicts reached in Alabama's trial courts of general jurisdiction from 1992 to 1997.
We estimate that the percentage of all financial injury verdicts in which punitive damages were awarded in Alabama is between 17 and 32 percent during the period 1992 to 1997. To put this range in perspective, in the other jurisdictions we found a low of 4 percent (in New York) and a high of 21 percent (in California). In Alabama, punitive damages represented between 80 and 86 percent of all damages awarded in all financial injury verdicts.
As with the other states in our database, punitive damage awards in
Alabama can be quite high. The mean punitive damage award is between $540,000 and $945,000; the 90th percentile award is between $947,000 and $1.9 million.
The median ratio of punitive damages to compensatory damages in
Alabama is somewhat over 5; this compares to 0.5 to 2.5 in the other jurisdictions studied. These data suggest that in Alabama punitive damages are awarded more often and are higher in any given case relative to compensatory damages than in the other jurisdictions in our database.
We also estimated the effects of a range of caps on punitive damage awards in the Alabama data. Because we cannot determine how this statute affects litigant and jury behavior, our estimates must be read with caution.
We estimate that a cap at the level of compensatory damages would have affected approximately 80 percent of the punitive awards in Alabama and would have reduced the total amount of punitive damages awarded by about 90 percent. A cap at three times compensatory damages would have affected the punitive damages awarded in 60 percent of the punitive damage awards in financial injury cases in Alabama and reduced the total amount of punitive damages awarded in these cases by 82 percent. In our estimates, we assumed that all general awards were compensatory. Had we assumed that they were in part or entirely punitive in nature, the effect of caps would have been larger.
 We omitted one very large financial injury verdict from Harris County--an award of more than $12 billion dollars, more than $2 billion of which was punitive--because it would have strongly affected many descriptions of the data. The next largest verdict in our database was less than one-fortieth the size of this verdict.
 To control for inflation, we converted all award amounts to 1992 dollars.
 The question of whether these changes reflect changed jury behavior or changes in plaintiffs' and defendants' litigation strategies, which, in turn, changed the mix of cases going to verdict, is beyond the scope of this analysis.
 Since 1987, Harris County has capped punitive damages at the greater of four times compensatory damages or $200,000 except in cases of malice or intentional tort. Consequently, for Harris County our estimate is the effect of imposing caps beyond the cap already in place.
 We present our results for Alabama as a range because some "general awards" are reported as a lump sum, without distinguishing what portion, if any, is a punitive award. The lower number in our estimated range of 17 to 32 percent assumes that the awards are entirely compensatory; the upper number assumes that they are entirely punitive.
 Alabama law limits punitive damages to $250,000, unless the defendant has exhibited a pattern or practice of intentional wrongful conduct involving actual malice or libel, slander, or defamation.