Defamation Litigation Patterns Across the United States, England, and Australia

Duke University From the SelectedWorks of David Unwin July 18, 2013 Defamation Litigation Patterns Across the United States, England, and Australia David Unwin Available at: https://works.bepress.com/david_unwin/1/

Defamation Litigation Patterns Across the United States, England, and Australia Table of Contents I. Introduction 1 II. Substantive Defamation Law. 3 a. Libel/Defamation Law in the United States.3 b. Libel/Defamation Law in England and Wales.....8 c. Libel/Defamation Law in Australia... 10 d. Comparing the U.S., English, and Australian Defamation Laws.. 13 III. Defamation Litigation Patterns 13 a. U.S. Defamation Litigation Data... 13 b. English Defamation Litigation Data..21 c. Australian Defamation Litigation Data.. 28 d. Comparison of Key Statistics.33 IV. Counter-Arguments. 40 V. Conclusion... 43 I. Introduction A good name is more desirable than great riches; to be esteemed is better than silver or gold. 1 A good reputation is perhaps an individual s most valuable asset. Yet it is capable of being destroyed in an instant. Defamation law recognizes and protects this interest by imposing tort liability. However, the freedom of speech is another bedrock legal principle. Imposing strict liability on any false, defamatory remark would unduly chill free speech. Thus, each national legal system must balance these two broad principles and reach a legal outcome which effectively protects an individual s reputation without unduly burdening the exercise of speech. Section II of this paper will examine three jurisdictions defamation law jurisprudence: the United States, England and Wales, and Australia. It will specifically compare the public figure/private figure and public concern/private concern distinctions 1 Proverbs 22:1 (New International).

in the U.S. and the qualified privilege defenses in England and Australia. The United States has adopted a jurisprudence which heavily favors the freedom of speech, whereas English and Australian courts have elected to give greater protection to an individual s reputation. Section III will examine defamation litigation patterns across these three jurisdictions. It will also address the question: what is the most valuable comparative statistic? Plaintiff win rate statistics are not illustrative as they suffer from severe sample selection bias, and defamation claims per capita statistics suffer from insufficient data. However, defamation claims as a percentage of civil claims statistics are illustrative and are supported by reliable data. This section will specifically compare the Pennsylvania Court of Common Pleas, the Queen s Bench Division in London, and the New South Wales Supreme Court, Common Law Division Civil. These data coupled with additional broader empirical studies which are also discussed in Section III demonstrate that defamation claims comprise a larger percentage of total civil claims in England and Australia than they do in the U.S. Furthermore, English and Australian awards are significantly greater than U.S. awards. Section IV will address possible counter-arguments to my analysis, and Section V will conclude that the diverging substantive jurisprudence of the three nations has led to dramatically contrasting litigation patterns in which potential and actual defamation plaintiffs fare much better in England and Australia than their counterparts do in the United States. 2

II. Substantive Defamation Law A. Libel/Defamation Law in the United States In the U.S., the plaintiff must show: (1) a statement of fact; (2) that is false; (3) and defamatory; (4) of and concerning the plaintiff; (5) that is published to a third party (other than the plaintiff and defendant); (6) not absolutely or conditionally privileged; (7) that causes actual injury...(8) that is the result of fault by the defendant... (9) that causes special (pecuniary) harm in addition to generalized reputational injury... 2 i. New York Times Co. v. Sullivan In New York Times Co. v. Sullivan, the U.S. Supreme Court squarely addressed the competing norms of accurate, truthful reporting and the freedom of speech. There, the New York Times published an advertisement which described the local police quelling a peaceful demonstration by black university students. 3 The plaintiff prevailed at the Alabama Supreme Court. 4 The Supreme Court reversed and adopted a new test which incorporated the First Amendment commitment to free speech. 5 The Supreme Court pointed to the profound national commitment to the principle that debate on public issues should be uninhibited, robust, and wide-open, and that it may well include vehement, caustic, and sometimes unpleasantly sharp attacks on government and public officials. 6 The Court recognized that imposing a strict liability fault standard would unduly chill valuable public speech protected by the First Amendment and thus adopted a new actual malice standard. 7 2 Rodney A. Smolla, Law of Defamation 1:34 (2d ed. 2012). 3 New York Times Co. v. Sullivan, 376 U.S. 254, 257 (1964). 4 Id. at. 256 5 Id. at 279-80. 6 Id. at 270 7 Id. at 279-80. 3

The constitutional guarantees require, we think, a federal rule that prohibits a public official from recovering damages for a defamatory falsehood relating to his official conduct unless he proves that the statement was made with actual malice -that is, with knowledge that it was false or with reckless disregard of whether it was false or not. 8 Thus, the Supreme Court held that, for public officials, the First Amendment required that the plaintiff must show that the defendant published the falsehood with knowledge or reckless disregard as to its falsehood. 9 Moreover, the plaintiff mush make this showing with convincing clarity to prevail. 10 The Court found that the New York Times had published the falsehoods with mere negligence and thus the plaintiff had not satisfied the actual malice test. 11 Therefore, the Sullivan decision has profoundly shaped U.S. libel law by adopting a high degree of fault (recklessness or knowledge) and imposing a weighty evidentiary burden (convincing clarity) on the plaintiff. This rigorous test underscores the U.S. commitment to the freedom of speech as constititutionalized by the First Amendment. ii. Gertz v. Robert Welch, Inc. However, the Sullivan test only applies to public figures. Later, the Supreme Court in Gertz v. Robert Welch, Inc. addressed the important distinction between public figures and private individuals. There, a magazine published falsehoods about a lawyer who had represented a family in a civil action against a police officer. 12 The Court described how public figures can more easily resort to self-help than can private individuals: Public officials and public figures usually enjoy significantly greater access 8 Id. 9 Id. 10 Id. at 285-6. 11 Id. at 287-8. 12 Gertz v. Robert Welch, Inc., 418 U.S. 323, 326 (1974). 4

to the channels of effective communication and hence have a more realistic opportunity to counteract false statements then private individuals normally enjoy. 13 Moreover, public figures usually have assumed the risk of defamatory statements whereas private individuals have not. The Court allows the media to assume that public officials and public figures have voluntarily exposed themselves to increased risk of injury from defamatory falsehood concerning them. No such assumption is justified with respect to a [pr]ivate individual. 14 Thus, the Court allowed the states to decide the applicable level of mental fault to defamation actions brought by private individuals, so long as the states did not impose a strict liability standard: We hold that, so long as they do not impose liability without fault, the States may define for themselves the appropriate standard of liability for a publisher or broadcaster of defamatory falsehood injurious to a private individual. 15 However, the states could only allow for punitive damages if the private individual plaintiff had proven actual malice: we hold that the States may not permit recovery of presumed or punitive damages, at least when liability is not based on a showing of knowledge of falsity or reckless disregard for the truth. 16 Thus, the states may adopt a negligence standard with regard to private individuals but cannot allow for punitive damages unless the plaintiff has proven recklessness or knowledge. The Court held that Gertz was a private individual and not a public figure. A person can be a public figure for all aspects of her life only if there is clear evidence of her fame or notoriety and her pervasive involvement in community affairs: Absent clear 13 Id. at 344. 14 Id. at 345. 15 Id. at 347. 16 Id. at 349. 5

evidence of general fame or notoriety in the community, and pervasive involvement in the affairs of society, an individual should not be deemed a public personality for all aspects of his life. 17 Gertz had not achieved this fame or notoriety. 18 Moreover, the Court described that a person can be a public figure as to a certain issue. Although Gertz represented the family whose son the police officer had shot, he had not publicly engaged the issue: He plainly did not thrust himself into the vortex of this public issue, nor did he engage the public's attention in an attempt to influence its outcome. 19 The Court plainly rejected the notion that lawyers, as officers of the court, are presumptively public figures. 20 Therefore, an individual must be generally famous or have thrust himself into the vortex of a public issue in order to be a public figure for purposes of libel law. iii. Dun & Bradstreet, Inc. v. Greenmoss Builders, Inc. In Dun & Bradstreet, Inc. v. Greenmoss Builders, Inc., the Supreme Court adopted an additional distinction between matters of public concern and private concern. There, a credit reporting agency had inaccurately reported that the plaintiff s business was in bankruptcy. 21 The Court held that this falsehood was not a matter of public concern but a matter of private concern and stressed the fact that this report was only issued to five subscribers who could not further disseminate the information. 22 It also 17 Id. at 352. 18 Id. 19 Id. 20 Id. at 351. 21 Dun & Bradstreet, Inc. v. Greenmoss Builders, Inc., 472 U.S. 749, 751 (1985). 22 Id. at 762. 6

pointed to the commercial nature of the information, asserting that this speech is hardy and unlikely to be unduly chilled by libel law. 23 The Supreme Court applied a new test for this speech which was of private concern. It allowed the states to impose punitive damages even absent a showing of actual malice: In light of the reduced constitutional value of speech involving no matters of public concern, we hold that the state interest adequately supports awards of presumed and punitive damages-even absent a showing of actual malice. 24 Thus, for speech involving matters of private concern, the states can allow punitive damages even if the plaintiff has merely shown negligence. The following table summarizes the results of these three cases: Table 1: Mens Rea in U.S. Libel Law (Constitutional Floor) Mens Rea (for Any Damages) Mens Rea (for Punitive Damages) Public Figure Recklessness or Knowledge Recklessness or Knowledge Private Figure Negligence Recklessness or Knowledge Private Concern Negligence Negligence This table only shows the constitutional floor which the First Amendment requires. Of course, states can choose to adopt higher levels of fault given that libel is a state tort action. 23 Id. 24 Id. at 761. 7

B. Libel/Defamation Law in England and Wales In England and Wales, the plaintiff merely needs to show that a published statement has adversely affected his reputation. 25 The plaintiff does not even need to show that statement was false or that he suffered damages. 26 However, like in the U.S., truth is a complete defense. 27 However, in cases where the statement is false, the defendant may still prevail under a Reynolds qualified privilege, a kind of responsible journalism defense. In Reynolds, the plaintiff was the former Taoiseach (Prime Minister) of Ireland who had recently resigned during a political crisis. 28 He sued a newspaper who had alleged that he had deliberately suppressed vital information from D il (the House of Representatives) of the Irish Parliament and his coalition cabinet colleagues. 29 Lord Nicholls of the House of Lords articulated a new test of ten non-exhaustive criteria which would establish a defense of qualified privilege: 1) The seriousness of the allegation; 2) The nature of the information, and to the extent to which the subject matter is a matter of public concern; 3) The source of the information; 4) The steps taken to verify the information; 5) The status of the information; 6) The urgency of the matter; 7) Whether comment was sought from the plaintiff; 8) Whether the article contained the gist of the plaintiff s side of the story; 9) The tone of the article; and 10) The circumstances of the publication, including the timing. 30 25 Reynolds v. Times Newspapers Ltd., [2001] 2 A.C. 127 (H.L.) 192 (appeal taken from Eng.). 26 Id. 27 Id. 28 Id. at 191 29 Id. 30 Id. at 205. 8

Lord Nicholls added that this list is not exhaustive and the assigned weights to the factors will vary from case to case. 31 The jury will decide the primary facts whereas the judge will perform the balancing test and make the ultimate determination as to whether the qualified privilege has been established. 32 Lord Nicholls also mentioned recent developments in the case law of the European Court of Human Rights. 33 A year earlier, the UK Parliament had passed the Human Rights Act of 1998 which incorporated the European Convention of Human Rights and Fundamental Freedoms. Article 10(1) provided that [e]veryone has the right to freedom of expression. This right shall include freedom to hold opions and to receive and impart information and ideas without interference by public authority and regardless of frontiers 34 Article 10(2) qualifies that the exercise of this freedom may be subject to such formalities, conditions, restrictions or penalties [which] are necessary... for the protection of the reputation or the rights of others. 35 Thus, the Reynolds qualified privilege defense was designed to track this new supranational enactment which emphasized the freedom of expression but qualified it as being subject to penalties which are necessary to protect the reputation of others. 36 Unlike U.S. law, English law does not focus on the plaintiff but rather examines the propriety of the journalists behavior. Moreover, the Reynolds factors have not led to a clear, predictable jurisprudence but have allowed courts to second-guess journalistic 31 Id. 32 Id. 33 Id. at 203-4. 34 Human Rights Act of 1998, 10(1) (Eng.). 35 10(2). 36 Reynolds at 207. 9

standards and practices. 37 The factors have not been compartmentalized in the jurisprudence. 38 Moreover, the fact that the list is non-exhaustive allows judges to add new factors, further muddying the jurisprudence. 39 Therefore, English law provides a kind of responsible journalism defense. However, this defense is not robust. Before 2010, there had been only seven successful invocations of this defense. 40 Moreover, the plaintiff does not need to prove any level of mens rea. 41 For these reasons, the substantive English law is much more favorable to libel plaintiffs than is substantive U.S. law. C. Libel/Defamation Law in Australia In Australia, the plaintiff must show that the statement or imputation was likely to injure his reputation, to injure him in his profession or trade, or that it was likely to cause others to shun or avoid or ridicule or despise him. 42 Like in the U.S. and England, truth is an absolute defense in Australia. 43 Moreover, a defendant may use either the common law qualified privilege defense under Lange or a statutory qualified privilege defense under the Uniform Defamation Acts. In Lange v. Australian Broad. Corp., the former Prime Minister of New Zealand, David Lange, sued the Australian Broadcasting Corporation, alleging that its television 37 Russell L. Weaver et al., The Right to Speak Ill: Defamation, Reputation, and Free Speech 109 (2006). 38 Id. 39 Id. at 110. 40 David Price et al., Defamation Law, Procedure and Practice 117 (4th ed. 2010). 41 See Reynolds at 192. 42 Mirror Newspapers Ltd. v. World Hosts Proprietary Ltd. (1979) 141 CLR 632, 640 (Austl.). 43 See e.g., Defamation Act (New South Wales) s 25; Defamation Act (Queensland) s 25. However, truth was not a complete defense in New South Wales before the Uniform Defamation Acts of 2005. See Defamation Act (1974) (New South Wales) s 15, 16. 10

broadcast conveyed that he was guilty of abuse of public office and was unfit to hold public office. 44 Unlike the U.S. Supreme Court, the High Court of Australia has the power to develop the common law of the states. 45 In light of the Australian Constitution s implied freedom of communication, the High Court established a common law qualified privilege which applied only to political and government matters. 46 However, the defendant must establish that his conduct was reasonable. 47 The High Court listed a set of factors in making this determination: But, as a general rule, a defendant's conduct in publishing material giving rise to a defamatory imputation will not be reasonable unless the defendant had reasonable grounds for believing that the imputation was true, took proper steps, so far as they were reasonably open, to verify the accuracy of the material and did not believe the imputation to be untrue. Furthermore, the defendant's conduct will not be reasonable unless the defendant has sought a response from the person defamed and published the response made (if any) except in cases where the seeking or publication of a response was not practicable or it was unnecessary to give the plaintiff an opportunity to respond. 48 Moreover, this defense is limited solely to matters of political and government matters. 49 Indeed, the defense may only include political information which is required to help electors make voting decisions. 50 Statements about corporate entities likely fall outside the scope of the defense. 51 However, statements about public administration may come 44 Lange v. Australian Broad. Corp. (1997) 189 CLR 520, 521 (Austl.). 45 Weaver at 82-3. 46 Lange at 576. 47 Id. at 575. 48 Id. 49 Id. 50 Weaver at 84. 51 See Orion Pet Products v. RSPCA (Vic.) (2002) F.C.R. 191 (Austl.) 11

within the defense. 52 In summary, this defense has a much narrower scope than the Reynolds qualified privilege defense and does not cover matters of wider public interest. 53 In 2005, the Australian states passed identical Uniform Defamation Laws which included a qualified privilege defense. 54 Under the statute, the defendant s conduct in publishing the defamatory statement must be reasonable under the following factors: 1) the extent to which the matter published is of public interest; 2) the extent to which the matter published relates to the performance of the public functions or activities of the person; 3) the seriousness of any defamatory imputation carried by the matter published; 4) the extent to which the matter published distinguishes between suspicions, allegations and proven facts; 5) whether it was in the public interest in the circumstances for the matter published to be published expeditiously; 6) the nature of the business environment in which the defendant operates; 7) the sources of the information in the matter published and the integrity of those sources; 8) whether the matter published contained the substance of the person s side of the story and, if not, whether a reasonable attempt was made by the defendant to obtain and publish a response from the person; 9) any other steps taken to verify the information in the matter published; and 10) any other circumstances that the court considers relevant. 55 These factors closely resemble the ten Reynolds criteria. 56 This defense has a much wider scope than the Lange common law defense as it potentially applies to any statement of public interest. However, both defenses adopt highly rigorous reasonableness tests. 52 See Chapman v. Conservation Council of South Australia (2002) 82 S.A.S.R. 449 (Austl.) 53 Weaver at 84. 54 See e.g., Defamation Act (New South Wales) s 30; Defamation Act (Queensland) s 30. 55 Id. 56 See Rhonda Breit, Uniform Defamation Laws in Australia: Have they struck a better balance, 10 (Apr. 15, 2013 8:14 PM) http://www.networkinsight.org/verve/_resources/breitr.pdf 12

D. Comparing the U.S., English, and Australian Defamation Laws The U.S. has adopted a much more defendant-friendly defamation jurisprudence than England and Australia. Public figure plaintiffs must establish actual malice, i.e. recklessness or knowledge, in order to prevail. Conversely, in England and Australia, the burden quickly shifts to the defendant to show that his conduct was reasonable under a non-exhaustive list of ten criteria. However, in recent years, the legal regimes have moved closer together as the U.S. qualified its Sullivan jurisprudence with Gertz and Dun & Bradstreet, England adopted the Reynolds criteria, and Australia passed the Uniform Defamation Acts. Nevertheless, the defendant still has a much greater burden and must show greater care in England and Australia. Two possible reasons for this disparity include the First Amendment to the U.S. Constitution and the deeply held commitment to the freedom of speech found in the U.S. Because neither England nor Australia have a comparable seminal legal text or national legacy, they weigh the reputation vs. speech balancing differently and adopt a more reputation-protecting, more speech-inhibiting policy. III. Defamation Litigation Patterns A. U.S. Defamation Litigation Data In 2005, there were 187 slander/libel trials in U.S. state courts. See Table A1. As a tort, defamation law is state law. The Bureau of Justice Statistics (Department of Justice) provides helpful summary data of all U.S. state court litigation in 2005: 13

Table A1: Summary Data of U.S. State Court Litigation in 2005 Number of Civil Claims 7.4 million Number of Civil Trials 26,948 % of Civil Claims Which Reached Trial 0.36% Number of Tort Trials 16,397 Number of Slander/Libel Trials 187 Slander/Libel Trials as Percentage of Civil Trials 0.7% Number of State Civil Claims Per 100,000 People 2,500 * Number of Slander/Libel Trials Per 100,000 People 0.063176 * * Assuming U.S. population in 2005 was 296 million. Source: Bureau of Justice Statistics Special Reports, Civil Bench and Jury Trials in State Courts, 2005, Table 1. 57 These data clearly show that the U.S. is a very litigious country. There were 2,500 state civil claims per 100,000 people in 2005. Over 99% of these claims were settled or dismissed before trial. Even so, there were 26,948 civil trials, 16,397 of which, were tort trials. However, there were only 187 slander/libel trials in total. Slander/libel trials were only 0.7% of civil trials. With some assistance from the Court Statistics Project, the Administrative Office of the Pennsylvania Courts recently conducted an empirical study of all civil case types brought before their general jurisdiction Court of Common Pleas. 58 During June December 2010, 766 slander claims were filed in the Pennsylvania Court of Common Pleas. See Table A2. Multiplying this number by 12/7 (12 months/ 7 months) yields an 57 This report is available here: http://bjs.gov/content/pub/pdf/cbjtsc05.pdf 58 See Implementing a Civil Cover Sheet: The Pennsylvania Experience, Court Statistics Project, available at: http://www.courtstatistics.org/otherpages/publications/~/media/microsites/files/csp/highlights/csph_vol18_n2.ashx 14

estimate of 1,313 total slander suits in 2010. Thus, there were roughly 10 slander suits per 100,000 people in Pennsylvania in 2010. Table A2: Slander Suits Per 100,000 in Pennsylvania Number of Slander Claims (Jun Dec 2010) 766 Estimated Number of Slander Claims (2010) 1,313 Estimated Slander Suits Per 100,000 People 10.34* Pennsylvania Court of Common Pleas *Assuming the population of Pennsylvania in 2010 was 12.7 million. 59 Source: Implementing a Civil Cover Sheet: The Pennsylvania Experience, Court Statistics Project There were 127,445 total civil claims brought to the Pennsylvania Court of Common Pleas from June December 2010. Thus, only a very small percentage (0.6%) of state civil claims in Pennsylvania were slander claims. See Table A3. Table A3: Slander Suits as Percentage of Total Civil Claims Number of Slander Claims (Jun Dec 2010) 766 Number of Civil Claims (Jun Dec 2010) 127,445 Slander Suits as Percentage of Civil Claims 0.601% Pennsylvania Court of Common Pleas Source: Implementing a Civil Cover Sheet: The Pennsylvania Experience, Court Statistics Project 59 See United States Census Bureau website at: http://quickfacts.census.gov/qfd/states/42000.html. 15

Returning to the nationwide study by the Bureau of Justice Statistics, one can observe plaintiffs in slander/libel trials prevailed roughly 40% of the time. See Table A4. Table A4: Plaintiff Win Rate at Trial Plaintiff Win Rate in Torts 51.6% Plaintiff Win Rate in Slander/Libel 39.4% Source: Bureau of Justice Statistics Special Reports, Civil Bench and Jury Trials in State Courts, 2005, Table 5. Plaintiffs won roughly 40% of slander/libel trials whereas plaintiffs in tort trials overall won roughly 50% of cases. Thus, in 2005, there were few defamation plaintiffs who reached the trial stage, and most of these plaintiffs lost at trial. However, the median trial award amount for successful slander/libel claims was roughly equivalent to the median trial award amount for successful tort claims. See Table A5. 16

Table A5: Trial Award Amount Torts Slander/Libel Median Trial Award Amount $24,000 $24,000 <$10,000 32.7% 22.5% $10,001 - $50,000 30.6% 40.0% $50,001 - $250,000 21.3% 2.5% $250,001 - $1 million 10.4% 22.5% > $1 million 5.0% 12.5% Source: Bureau of Justice Statistics Special Reports, Civil Bench and Jury Trials in State Courts, 2005, Table 6. The median trial award amount for both slander/libel plaintiffs and overall tort plaintiffs was equal to $24,000. About one third of tort awards were less than $10,000 whereas 40% of slander/libel awards were between $10,000 and $50,000 and over a third of slander/libel awards were over $250,000. Thus, the mean slander/libel award was almost certainly significantly greater than the mean tort award. Moreover, a greater percentage of plaintiffs in slander/libel trial sought punitive damages than plaintiffs in overall tort trials. See Table A6. Over a fifth of slander/libel plaintiffs sought punitive damages. Among these plaintiffs, almost two thirds successfully recovered punitive damages. However, the median amount of punitive damages in slander/libel trials ($13,000) was much lower than the median amount of punitive damages in overall tort trials ($55,000). Nevertheless, one fourth of the slander/libel punitive damage awards exceeded $1 million whereas only 17% of tort punitive damage awards exceeded $1 million. 17

Table A6: Punitive Damages Torts Slander/Libel Number of Plaintiffs Who Sought Punitive Damages 822 38 % of Plaintiffs Who Sought Punitive Damages 5.01% 20.32% Number of Plaintiffs Awarded Punitive Damages 254 24 % of Plaintiffs Seeking Punitive Damages Who Were 30.90% 63.16% Awarded Punitive Damages Median Amount $55,000 $13,000 Number of Punitive Damage Awards over $250,000 59 9 Number of Punitive Damage Awards Over $1 million 43 6 % of Punitive Damage Awards Exceeding $1 million 16.93% 25.00% Source: Bureau of Justice Statistics Special Reports, Civil Bench and Jury Trials in State Courts, 2005, Table 7. This disparity in punitive damages explains the previous statistics reported in Table A5. Although the median slander/libel trial award is comparable to the median tort trial award, the mean slander/libel trial award is much greater due to the possibility of large punitive damages. Therefore, in 2005, there were few slander/libel plaintiffs who reached the trial stage, even fewer of whom prevailed at trial. However, of those who prevailed, there were 6 who recovered over a $1 million in punitive damages. These data depict a generally arduous road for the majority of slander/libel plaintiffs but a potentially large payout at trial for a select few. Adopting a broader time horizon may provide another valuable perspective. In Libel Law in the Trenches, David Logan examines the Libel Resource Defense Center s 18

study of libel suits against media defendants from the 1980s and 1990s. 60 However, this study focuses only on suits against media defendants and thus examines a slightly different question. 61 Table A7: Plaintiff Win Rates in the 1980s and 1990s Plaintiff Win Rate in Summary Judgment Motions between 1980-1996 23.3% Plaintiff Win Rate at Trial between 1980-1999 63.1% Source: Logan at 510, 513. Logan s statistics suggest a particularly difficult pre-trial stage for plaintiffs against media defendants, with less than a quarter of plaintiffs surviving the summary judgment stage. However, over 60% of plaintiffs prevailed at the trial stage. However, the picture appears slightly less rosy for plaintiffs when focusing the analysis on more recent data: Table A8: Plaintiff Win Rates in the 1990s Plaintiff Win Rate in Summary Judgment Motions between 1995-1996 17.6% Plaintiff Win Rate at Trial in the 1990s 60.9% Source: Logan at 510, 513. Plaintiffs were less successful at the summary judgment and trial stage in the 1990s. Extrapolating this downward trend could explain why we saw that the plaintiff win rate at trial was 39.4% in 2005. However, Logan s statistics only measure plaintiff 60 David A. Logan, Libel Law In The Trenches: Reflections On Current Data On Libel Litigation, 87 Va. L. Rev. 503, 509 (2001). 61 Id. 19

win rates against media defendants. Moreover, plaintiff win rates have very low probative value as they suffer from severe sample selection bias. 62 See later discussion in subsection D. The trial awards between 1980-1999 were fairly high. Almost one half were greater than $250,000 and over one quarter exceeded $1 million. See Table A9. Table A9: Range of Trial Awards Between 1980-1999 Total $0 - $250,000 $250,001 - $999,999 Over $1 million 273 140 (51.2%) 61 (22.3%) 72 (26.4%) Source: Logan at 514. However, the affirmance rate of trial awards is fairly low. Appellate judges frequently reduced the award or reversed entirely. Conversely, appellate judges rarely disturbed a verdict for the defendant. See Table A10. Table A10: Final Awards Between 1980-1999 Affirmance Rate of Trial Award 27.8% Affirmance Rate of Verdict for Defendant 81% Mean Payout $281,000 Median Payout $72,500 Source: Logan at 516, n. 68, 517. 62 See George L. Priest and Benjamin Klein, The Selection of Disputes for Litigation, 13 J. Legal Stud. 1, 5 (1984). 20

Less than 30% of trial awards for the plaintiff stayed in tact on appeal, whereas 80% of verdicts for the defendant survived on appeal. Over the two decades, the final mean payout was $281,000 and the final median payout was $72,500. Therefore, the data suggest that while a successful libel plaintiff during the 1980s and 1990s may receive a large trial award, this award was frequently reduced on appeal. However, this figure of $72,500 is still much greater than the 2005 median trial award amount of $24,000 reported by the Bureau of Justice Statistics. This disparity suggests that plaintiffs generally recover larger awards against media defendants. B. English Defamation Litigation Data There are few readily available English libel statistics. Indeed, Judith Townend laments the dearth of data about defamation cases in a recent blog post. 63 However, after piecing data together from a number of different sources, one can roughly estimate certain statistics. The UK Ministry of Justice regularly publishes court statistics on its website, both as reports and spreadsheets. To compare with the U.S. summary data, I selected the year 2005. 63 Judith Townend, A dearth of data about defamation cases in England & Wales (Apr. 15, 2013 8:39 PM), http://meejalaw.com/2012/09/19/a-dearth-of-data-about-defamationcases-in-england-wales/ 21

Table B1: England and Wales Summary Data in 2005 Number of Civil, Non-Family Claims 1,969,069 Number of Money Claims 1,576,733 Number of Hearings or Trials 63,380 % Cases Heard or Tried 3.2% Number of Civil, Non-Family Claims Per 3,715 100,000 People* *Assumes the population of England and Wales in 2005 was 53 million. Source: UK Ministry of Justice, Table 1.1, 1.2 of Court Statistics Quarterly April June 2012, All Tables appendix. 64 These data report all civil, non-family claims filed throughout the county courts of England and Wales. Only about 3% of civil, non-family claims reached trial in 2005. This number is far greater than the equivalent U.S. statistic: only 0.36% of civil claims in the U.S. reached trial. See Table A1. Furthermore, there were over 3,700 civil, nonfamily claims per 100,000 people in England and Wales. This number is surprisingly significantly higher than the equivalent U.S. statistic of 2,500 state civil claims per 100,000 people. However, the U.S. statistic does not include federal civil claims, and the English statistic does not include family claims. Nevertheless, it is apparent that both countries have similar levels of litigiousness. The Ministry of Justice also reports on specific divisions within the High Court. The Queen s Bench Division handles defamation claims across England and Wales. 65 64 This source is available here: https://www.gov.uk/government/publications/courtstatistics-quarterly-earlier-editions-in-the-series 65 The Queen s Bench Guide, Ministry of Justice at 7-8, available at: http://www.justice.gov.uk/downloads/courts/queens-bench/queen-bench-guide.pdf 22

However, the relevant data reports only the claims which are handled at the Royal Courts of Justice in London. Table B2: Queen s Bench Division Claims in 2011 Slander/Libel Total Slander/Libel as % 15,000 50,000 28 939 2.98 Over 50,000 61 2,129 2.87% Unspecified Value 76 1,658 4.58% Total Claims 165 4,726 3.49% Source: UK Ministry of Justice, Table 6.2 of Queen s Bench Division spreadsheet. 66 There were 165 defamation claims brought in 2011 at the Queen s Bench Division in London. Most of these claims were for over 50,000 or an unspecified value. Roughly 3.5% of all claims brought to the Queen s Bench Division in London were defamation claims. Table B3 examines this statistic from 2005-2011. 66 This source is available here: https://www.gov.uk/government/publications/judicialand-court-statistics-annual 23

Table B3: Percentage of Defamation Claims Year Claims Issued in London at the Queen s Bench Defamation Claims Issued in London at the Queen s Bench 2011 4,726 165 3.49% 2010 4,864 158 3.24% 2009 5,694 298 5.23% 2008 5,173 259 5.00% 2007 4,794 233 4.86% 2006 4,246 213 5.02% 2005 3,841 252 6.56% Defamation Claims as % of All London QB Claims Source: Judicial Statistics 2011: Number of defamation claims remains low, The International Forum for Responsible Media Blog. 67 Tracing this statistic from 2005-2011 reveals that defamation claims have been decreasing somewhat both absolutely and as a percentage of all claims issued in London at the Queen s Bench Division. However, this percentage has ranged from 3.24 6.56%, a much larger percentage than the percentage of U.S. defamation claims to U.S. civil claims which was 0.6%. See Table A3. This distinction suggests that defamation claims comprise a larger percentage of civil claims in England than in the U.S. Table B4: Number of Defamation Claims in 2011 Number of Defamation Claims in London 165 at the Queen s Bench Division Number of Defamation Claims Per 100,000 2.06* in London (QB) *Assuming London s population in 2011 was 8 million. 68 67 This source can be found here: http://inforrm.wordpress.com/2012/07/15/judicialstatistics-2011-number-of-defamation-claims-remains-low/ 24

In 2011, there were 165 defamation claims brought to the Queen s Bench Division in London. Thus, there were roughly 2 defamation claims at the Queen s Bench per 100,000 London residents. See Table B4. Table B5: Notable Defamation Trials from 1990-2009 Plaintiff Win Rate 65.38% Mean Trial Award or Settlement (among 72,670 Plaintiff Winners) Median Trial Award or Settlement 42,500 Source: Calculated from Appendix 3 of Price et al. s Defamation Law, Procedure and Practice. Appendix 3 of Price et al. s Defamation Law, Procedure and Practice contains a schedule of notable defamation cases from 1990 2009. From this schedule, I calculated the plaintiff win rate, mean trial award or settlement, and median trial award or settlement. See Table B5. The plaintiff win rate of 65.38% is much higher than the 39.4% U.S. plaintiff win rate in 2005 but is comparable to Logan s statistic of a 60.9% U.S. plaintiff win rate in the 1990s. See Tables A4, A8. The median trial award of 42,500 is significantly greater than the U.S. median trial award amount in 2005 which was $24,000. See Table A5. However, it is comparable to Logan s estimate of $72,500 in Table A10. In 2009, Sir Rupert Jackson issued a preliminary report on civil litigation costs. 69 68 See Census 2011: London s population booms to Eight Million, London Evening Standard, available at: http://www.standard.co.uk/news/london/census-2011-londonspopulation-booms-to-eight-million-7946191.html 69 Lord Justice Jackson publishes preliminary findings of his civil litigation cost review, 25

In Appendix 17, he included a schedule of libel claims against the media resolved by settlement or judgment in 2008. From these data, I made the following calculations reported in Table B6. Table B6: Defamation Trials and Settlements in 2008 Number of Cases in which Plaintiff Won at 140 Trial or Obtained Favorable Settlement Mean Award or Settlement 88,937.20 Median Award or Settlement 11,556.50 Source: Appendix 17 of Sir Rupert Jackson s preliminary report. 70 There were 140 cases in which the plaintiff won at trial or obtained a favorable settlement. The mean award or settlement was almost 89,000, a larger figure than the mean award or settlement from the 1990 2009 data in Table B5. However, the median award or settlement was around 11,500, a much smaller figure than the corresponding figure in Table B5. Perhaps this figure has been somewhat distorted by the large number of small settlements reported in Appendix 17. Judicial Communications Office news release, http://www.judiciary.gov.uk/media/mediareleases/2009/news-release-1309 70 This source can be found here: http://www.judiciary.gov.uk/publications-andreports/review-of-civil-litigation-costs/reports/civil-litigation-costs-review-preliminaryreport 26

Table B7: Additional Mean and Median Award Figures June 1997 January 2009 January 2004 January 2009 Mean Award just short of 48,000 under 38,000 Median Award 25,000 just over 20,000 Source: Alistair Mullis and Andrew Scott, Something Rotten in the State of English Libel Law?: A Rejoinder to the Clamour for Reform of Defamation, para. 41 (2010). 71 In their 2010 article, Professor Alistair Mullis and Dr. Andrew Scott estimated that the mean award was between 1997 2009 was just short of 48,000 whereas the median award was 25,000. However, these statistics decreased to under 38,000 and just over 20,000 respectively during the 2004 2009 period. The awards during this period are much more comparable to the $24,000 U.S. median trial award figure in Table A5. However, it appears that Mullis and Scott s figures were calculated following appeals whereas the U.S. figure is calculated before appeals. Mullis and Scott suggest that the Court of Appeal has in recent years sought to dampen the escalation of awards made by juries. 72 This description is starkly contrasted to the story depicted by Prosser in his 1992 article The English Libel Crisis: A Sullivan Appellate Review Standard Is Needed. 73 He describes how from 1964 to 1992, no jury libel award had been reduced by the Court of Appeal. 74 However, the Court of Appeal may have adopted a more rigorous standard of appellate review in recent years as Mullis and Scott have suggested. 71 This source can be found here: http://www.lse.ac.uk/collections/law/news/libel.pdf 72 Mullis and Scott at para. 41. 73 Sean Thomas Prosser, The English Libel Crisis: A Sullivan Appellate Review Standard Is Needed, 13 N.Y.L. Sch. J. Int l & Comp. L. 337 (1992). 74 Id. at 353-4. 27

C. Australian Defamation Litigation Data Australian civil litigation data was even more difficult to find than English civil litigation data. However, I did find some summary data of New South Wales, the Australian state in which Sydney is located. New South Wales is likely a good representative sample of Australia in total. Marfording and Eyland reported on civil litigation in New South Wales. See Table C1. Table C1: New South Wales Summary Data in 2003 Civil Claims Filed 20,009 Civil Claims Filed Per 100,000 People 299 Civil Claims Resolved by Judgment 9.28% Source: Annette Marfording and Ann Eyland, Civil Litigation in New South Wales: Empirical and Analytical Comparisons with Germany, University of New South Wales Faculty of Law Research Series 2010, Working Paper 28 (2010), p. 110, Table 4.3. Over 9% of civil claims were resolved by judgment, a much greater figure than the 0.36% U.S. figure and the 3.2% English figure. See Tables A1 and B1. However, the Australian figure includes judgment by default, judgment by acknowledgment, and judgment by renunciation. Thus, it is not a precise comparable statistic. However, the overall litigiousness in New South Wales seems much lower than in the U.S. or in England. There were only 299 civil claims per 100,000 people. In the U.S., there were over eight times as many state civil law claims per 100,000 people, and in England and Wales, there were over twelve times as many civil, non-family claims. Thus, the U.S. and English populations appear to be much more litigious than the Australian population. 28

Table C2: Defamation Claims as Percentage of Civil Claims New South Wales Supreme Court, Common Law Division--Civil Number of Defamation Claims Filed (2009) 6,313 Number of Civil Claims Filed (2009) 73 Defamation Claims as Percentage of Civil 1.16% Claims Source: Supreme Court of New South Wales, Annual Review 2009, p. 56-7 75 Table C2 shows that 73 defamation claims were filed in its Common Law Division, Civil Section. This division excludes equity claims as the NSW Supreme Court has a separate Equity Division. 76 Over 1% of claims filed in the NSW Supreme Court s Common Law Division Civil were defamation claims. This figure is nearly twice as much as the comparable U.S. percentage (0.6%) but less than half of the comparable English percentage (3.49%). See Tables A3 and B3. Michael Newcity conducted a through comparative empirical study of Australian and U.S. defamation law. 77 He examined 435 cases brought to the New South Wales Supreme Court from February 1979 June 1981. 78 He examined the NSW Supreme Court because Sydney has been characterized as the defamation capital of Australia. 79 He derived the U.S. figures from reports by the Iowa Libel Research Project. 80 75 This report is available here: http://www.supremecourt.lawlink.nsw.gov.au/agdbasev7wr/supremecourt/documents/pdf /supremecourtannualreview09.pdf 76 See About the Supreme Court, Supreme Court of New South Wales website at: http://www.supremecourt.lawlink.nsw.gov.au/supremecourt/sco2_aboutus.html 77 Michael Newcity, The Sociology of Defamation in Australia and the United States, 26 Tex. Int l L.J. 1 (1991). 78 Id. at 11-12 79 Id. 80 Id. at 53. 29

Table C3: Australian and U.S. Defamation Percentages Australia (1979 1981) U.S. (1974 1984) % Claims Which Go To Trial 20.9% 24% Plaintiff Win Rate 62.6% 61% Source: Newcity at 52, 53, 55, n. 186. It appears that, during this time period, Australia and the U.S. have comparable figures for the percentage of claims which reach trial and for the plaintiff win rate. Roughly 20% of claims reached trial at the NSW Supreme Court and the plaintiff prevailed in over 60% of the trials. Table C4: NSW Supreme Court Trial Awards <A$1,000 4 18.2% A$1,001 - A$10,000 4 18.2% A$10,001- A$25,000 3 13.6% A$25,001 A$100,000 9 40.1% >A$100,000 2 9.1% Total 22 100% Source: Newcity at 57. Of the 22 cases in which the plaintiff received money damages, nine of those cases resulted in an award in the range of A$25,001- A$100,000. The mean award was A$44,355. These figures are significantly greater than the U.S. figure of $24,000 in Table A5. 30

Table C5: Mean and Median Trial Award Mean Trial Award Median Trial Award A$44,355 A$25,001 A$100,000 Source: Newcity at 57. The appeal rate in Australia was much lower than the appeal rate in the U.S. during this period. See Table C6. Less than 30% of the trial verdicts were appealed whereas in the U.S., 90% of defamation trial verdicts were appealed. Table C6: Comparative Appeals Rates Australia (1979 1981) U.S. (1974 1984) Appeal Rate 27.5% 90% Source: Newcity at 60. Even more strikingly, the plaintiff would more frequently appeal than the defendant and often would win more frequently on appeal. See Table C7. Thus, plaintiffs win over 60% of the time at trial and over 55% of the time on appeal! Table C7: Appeals Rates Defendant Appeal Rate 22.8% Plaintiff Appeal Rate 33.3% Plaintiff Win Rate on Appeal (combining plaintiff and defendant appeals) Source: Newcity at 60. 57.1% However, these data must be taken with a grain of salt. Because of the small sample size, these data are almost certainly not statistically significant. In Newcity s 31

sample, 91 cases reached trial verdicts and only 25 were appealed. 81 Nevertheless, these data suggest profoundly plaintiff-friendly litigation outcomes. Newcity also examined the average number of defamation suits per year at the NSW Supreme Court between 1986-1987. See Table C8. Table C8: Defamation Suits Per Capita Average Defamation Suits Per Year at the 231 NSW Supreme Court (1986-1987) Defamation Suits Per 100,000 People in 4.08* New South Wales (1987) Average Defamation Suits Per Year in the 110 U.S. (1974-1984) Defamation Suits Per 100,000 People in the 0.049 U.S. (1974 1984) *Assuming the population of New South Wales in 1987 was 5,660,500. Source: Newcity at 63, n. 220. There was an average of 231 defamation suits brought to the NSW Supreme Court alone between 1986-1987. There were roughly 4 defamation suits per 100,000 people in New South Wales which is roughly double the comparable 2011 London figure of 2.06 defamation suits per 100,000. This figure is over eighty times greater than Newcity s U.S. figure of 0.049 defamation suits per 100,000 people. This disparity suggests that defamation suits are far more prevalent per capita in New South Wales than in the U.S. However, Newcity admits that his U.S. figure is based on elaborate and highly speculative mathematic calculations. 82 Thus, I will not use this U.S. figure in the following comparison section. 81 Id. at 60. 82 Id. at 63. 32

D: Comparison of Key Statistics Drawing from the above data, this section will compare key statistics across the three jurisdictions. Table D1 shows that the U.S. and England are far more litigious than Australia. In the U.S., there were over eight times as many civil law claims per capita than in Australia, and in England and Wales, there were over twelve times as many civil claims per capita than in Australia. Table D1: Overall Litigiousness United States (2005) 2500* England and Wales (2005) 3,715** New South Wales (2003) 299 *State Civil Claims **Civil, Non-Family Claims Source: Tables A1, B1, and C1 Civil Claims Per 100,000 People Table D2 yields some surprising results. It appears that there are five times more defamation suits per capita in Pennsylvania than defamation suits per capita in London and over twice as many defamation suits per capita than in New South Wales. However, this comparison is somewhat misleading. The Pennsylvania figure does not take into account the number of federal defamation claims filed. Moreover, the London figure does not take into account the number of defamation claims filed outside the Queen s Bench Division of the High Court. For instance, the county courts in England and Wales 33

can also hear tort claims. 83 Furthermore, the New South Wales Supreme Court is not the only Australian court in New South Wales to hear defamation cases. For instance, the District Court of New South Wales also hears defamation cases. 84 Thus, these statistics are merely court-specific and do not accurately convey the number of defamation suits per capita. Table D2: Number of Defamation Suits Number of Defamation Suits Per 100,000 People Pennsylvania States (2010) 10.34 London (2011) 2.06 New South Wales (1987) 4.08 Source: Tables A2, B4, and C8. However, Table D3 provides much more meaningful statistics. It compares the number of defamation claims as a percentage of civil claims in the Pennsylvania Court of Common Pleas, the Queen s Bench Division in London, and the NSW Supreme Court, Common Law Division Civil. 83 See County Court, Judiciary of England and Wales website, available at: http://www.judiciary.gov.uk/you-and-the-judiciary/going-to-court/county-court/countycourt#headinganchor1 84 See Civil Cases, District of New South Wales website, available at: http://www.districtcourt.lawlink.nsw.gov.au/districtcourt/case_types/civil_cases.html 34

Table D3: Defamation Suits as a Percentage of Civil Claims Pennsylvania Court of Common Pleas (June Dec 2010) Queen s Bench, Royal Courts of Justice in London (2011) NSW Supreme Court, Common Law Division Civil (2009) Number of Defamation Suits Source: Tables A3, B2 and C2. Number of Civil Claims 766 127,445 0.60% 165 4,726 3.49% 73 6,313 1.16% Defamation Suits as a Percentage of Civil Claims This comparison is far more meaningful. All three courts are courts of first instance and regularly handle defamation claims. Moreover, these statistics are all very recent. The number of defamation claims as a percentage of civil claims in the Queen s Bench Division (3.49%) is roughly three times greater than the number of defamation claims as a percentage of civil claims in the NSW Supreme Court Common Law Division Civil (1.16%). It is roughly five times greater than the number of slander claims as a percentage of civil claims in the Pennsylvania Court of Common Pleas (0.60%) and the number of slander/libel trials as a percentage of overall state civil trials (0.7%). See Table A1. This comparison indicates that defamation claims in England and Australia comprise a much larger percentage of civil claims than defamation claims do in the U.S. In fact, defamation cases have their own docket within the Common Law Division. 85 85 NSW Supreme Court Annual Review at 16. 35