Editor’s note: Our friends at Harvard’s Digital Media Law Project have taken a look at the legal implications of the sort of erroneous reporting — in both traditional and in social media — we saw in the aftermath of the Boston Marathon bombings. Here, project director Jeff Hermes explores the case law to put the issues into context.
There has been outstanding coverage and analysis of the Boston Marathon bombings and subsequent manhunt by both mainstream and independent media, but one particular aspect which has stood out for many commentators was the role played by social media. For many in the Boston area, platforms such as Twitter and Reddit became a key way for the community to share its experiences and reactions, anger, fear, and prayers. In the Watertown area, social media took on another aspect as the way in which many of us who sat locked in our houses reached out to one another for news and support.
However, there was also controversy around social media, in particular concerns about “witch hunts” among social media users while the suspects were still unknown. While these problems were not limited to social media platforms, there has been significant commentary about how social media either helped or hindered law enforcement efforts and public understanding in a crisis situation.
As we all try to gain perspective on the events of last month, it is helpful to remember that this far from the first time that reporting on acts of terror has generated mistakes and misidentifications. To the contrary, the natural impulse to identify the perpetrators of horrific acts as quickly as possible has often led to reporting of law enforcement efforts that swept up individuals later cleared of any wrongdoing. Examination of these situations, and the legal cases that resulted, may reveal whether there are unique issues that can be laid at the feet of social media or if these issues appear generally in reporting after terrorist attacks. Here are three case studies.
On December 21, 1988, an explosive device detonated on Pan Am Flight 103 over Lockerbie, Scotland, resulting in the deaths of all passengers and crew, as well as others on the ground being killed by falling wreckage. On January 31, 2001, a Libyan intelligence officer was convicted and imprisoned for the bombing. During the intervening twelve years, there was widespread discussion and speculation about who might be responsible for the bombing, with the government of Libya being at the center of the leading theory, but with many alternative theories advanced.
One such alternative theory was published in a cover story in Time magazine in April 1992, suggesting that the bombing had been the work of a Palestinian group seeking to eliminate U.S. counterterrorism agents on the flight. According to the article, the Palestinian group identified the flight carrying the U.S. agents with the assistance of a U.S./Iranian double agent named David Lovejoy. Time also ran a photograph purporting to be of Lovejoy, which it obtained from an affidavit filed by an attorney for Pan Am in a civil lawsuit related to the bombing. The affidavit claimed that an unnamed source had identified Lovejoy as the man in the photograph.
In fact, the picture was of another man, Michael Schafer. As the U.S. Court of Appeals for the Eleventh Circuit would later describe the mistake, “Time’s article, therefore, erroneously identified Schafer, then working in his family’s janitorial business in Austell, Georgia, both as a traitor to the United States government and a player in the bombing of Pan Am 103.” Schafer demanded a retraction from Time, which the magazine published more than a month later. He also sued Time, Inc., for defamation in federal district court in Georgia.
The jury returned a verdict for Time in less than an hour. Schafer appealed the ruling, and the Eleventh Circuit reversed, granting a new trial because it found that the jury instructions issued by the district court were confusing. Specifically, the instructions were vague as to the question of whether Schafer needed merely to prove that Time was negligent in checking its facts (the proper standard for liability), as opposed to some level of intent to injure or constitutional “actual malice.” However, the Eleventh Circuit held that Time would be allowed to argue in the next trial that it was not negligent in relying on the attorney affidavit.
Time settled the case (Schafer v. Time, Inc.) with Schafer before retrial.
On July 27, 1996, a bomb exploded in Centennial Olympic Park in Atlanta, killing one person and injuring more than 100 others. For three days thereafter, media coverage identified Olympic security guard Richard Jewell as a hero of the event, based upon his reporting of a suspicious unattended package and his reduction in casualties from the event through his efforts to evacuate people from the area around the package. On July 30, however, the Atlanta Journal-Constitution identified Jewell as the “focus” of the FBI’s investigation of the event, which initiated pervasive and detailed analysis in the press of Jewell, his past, and his potential connection to the bombing.
In October 1996, Jewell was cleared of any responsibility for the bombing, with the FBI issuing an unusual official statement that he was no longer a suspect. (In 2003, Eric Robert Rudolph was convicted of the bombing and sentenced to life imprisonment.) Jewell then filed defamation lawsuits against a number of media outlets, including NBC, CNN, the New York Post, and Cox Enterprises d/b/a the Atlanta Journal-Constitution. The latter two cases generated significant court decisions.
In October 1998, the district court granted the Post’s motion in part and denied it in part. The district court held that statements that Jewell was the “prime” or “main” suspect in the FBI’s investigation were substantially true, based upon Jewell’s own admissions. Jewell acknowledged that he was a suspect, but denied that he was the “main” suspect; the court found that to be a distinction without a substantial difference in terms of the impact of the Post’s report.
With respect to defamatory meaning, many of the statements about which Jewell complained in the Post articles were facially innocuous, such as claims that Jewell “was a straight arrow who overdid everything” and was “desperate to stand out as a hero.” However, the district court found that, read in context, these characteristics were presented in support of speculation by the Post and others that Jewell was responsible for the bombing; as a result, the court held that these statements carried a defamatory meaning. Similarly, the court held that statements that Jewell “fit the profile of the bomber” were not innocuous, just because innocent people as well as the guilty party might fit a profile; in context, the statement indicated a belief that Jewell was in fact the guilty party.
Nevertheless, the court held that the context of the New York Post articles was such that, for at least some of the statements at issue, a reasonable reader would understand them to be mere speculation or opinion:
It does not strain the concept of judicial notice … to note that everyone hoped that the individual(s) responsible for the crime would quickly be brought to justice…Given this quick succession of events and publications, a reasonable reader would have understood the information concerning Jewell’s involvement in the bombing to have been preliminary in nature…The preliminary nature of reported information is a contextual factor which supports, but by no means dispositively so, a finding that the statements are ones of opinion.
But the court also recognized countervailing considerations:
[T]he simple fact of the matter is that these statements were published in a newspaper…A newspaper column is the product of some deliberation, not the heat of a moment. Prior to publication, it passes through the hands of professional editors and it thus carries with it the cloak of credibility and authority of the particular newspaper and the profession…These undoubtedly are circumstances encouraging the reasonable reader to be less skeptical and more willing to conclude that the report is stating or implying facts garnered by a professional news gatherer and reporter.
The district court found that while many statements in the articles were phrased as speculation or in loose and rhetorical language, others were stated as fact. Moreover, the court found that many of the statements of opinion — such as beliefs and speculation that Jewell fit the profile of the bomber — might be found to create false implications of fact. As such, defamation claims on most of the statements were allowed to proceed.
The case against the Post was dismissed in March 1999 following a settlement in an undisclosed amount.
In contrast to the New York Post, the Atlanta Journal-Constitution did not settle with Jewell, continuing to fight the case to conclusion after Jewell’s death in 2007. Eventually, the Georgia trial court granted summary judgment to the defendants with respect to all statements at issue, which ruling was affirmed by the Georgia Court of Appeals in 2011. The Court of Appeals affirmed that the statements at issue were either substantially true or opinions that could not be proven false.
Notably, the Court of Appeals considered the context of the statements at issue in a manner similar to the federal district court in the New York Post case, but reached a significantly different result:
We cannot conclude that the statements contained within the August 1 and August 4 articles, construed in the context of the entirety of those articles and given their reasonable and natural meaning, amounted to an accusation by the Media Defendants that Jewell planted the bomb. Rather, a reasonable reader would have understood the information to be preliminary in nature and published during the very early stages of an ongoing investigation. Moreover, both articles reported not only the suspicion of Jewell’s involvement, but also evidence tending to belie that suspicion. And finally, the record definitively establishes that at the time of the publications, investigators did, in fact, suspect that Jewell may have planted the bomb and were actively investigating that theory…
Although the July 31 article repeats the opinion of investigators who reportedly believed that Jewell may have placed the 911 call, it includes within its text the factual premise of that reported opinion. For example, the article sets forth what some described as Jewell’s fervent approach to his prior law-enforcement duties; expressions of concern made by Jewell’s former employer; Jewell’s prior arrest for impersonating a police officer; and Jewell’s reported ownership of a similar knapsack. Nothing in the article suggests to the reader any defamatory facts other than those disclosed within its text which, in context, is obviously a report on the very early stages of an intense and ongoing investigation.
The Court of Appeals also held that the comparison of Jewell in one column to a notorious serial killer was “loose, figurative language” and “[n]onliteral commentary that cannot reasonably be interpreted as stating actual facts about an individual.”
The Georgia Supreme Court denied review of the Court of Appeals decision on January 9, 2012, ending the case almost exactly 15 years after it started.
Over the course of several weeks following the September 11, 2001, terrorist attacks, an unknown individual mailed a series of envelopes containing anthrax powder to various media outlets and the offices of two U.S. Senators, resulting in the deaths of five people and the infection of seventeen others. Throughout 2002 in the immediate wake of the attacks, Nicholas Kristof, a columnist for The New York Times, wrote about the FBI’s handling of information related to a particular suspect he called “Mr. Z.”
Much of Kristof’s writing was devoted to collating information about Mr. Z that indicated he was responsible for the attacks. This culminated in an August 13, 2002 article in which Kristof identified Mr. Z as Dr. Steven Hatfill, a research scientist employed by the Department of Defense, and stated that the FBI should “end this unseemly limbo by either exculpating Dr. Hatfill or arresting him.”
Hatfill sued The New York Times for defamation in federal district court in Virginia (Hatfill v. The New York Times Co.; Hatfill v. The New York Times Co.). The district court initially dismissed Hatfill’s complaint, finding that Kristof was merely reporting on an ongoing investigation actually focused on Hatfill at the time and was “careful to disavow any conclusion of Hatfill’s guilt.” The U.S. Court of Appeals for the Fourth Circuit reversed in 2005, and remanded to allow Hatfill’s case to proceed. The Fourth Circuit specifically rejected the Times’ argument that the columns were agnostic as to Hatfill’s guilt:
The columns did not describe any other actual or potential target of investigation, and they recounted detailed information pertaining to Hatfill alone. Once Kristof named Hatfill as Mr. Z (and perhaps even before that time), a reasonable reader of his columns could believe that Hatfill had the motive, means, and opportunity to prepare and send the anthrax letters in the fall of 2001; that he had particular expertise with powder forms of anthrax, the type used in the mailings; that his own anthrax vaccinations were current; that he was the prime suspect of the biodefense community as well as federal investigators; that he had failed numerous polygraph examinations; that specially trained bloodhounds had “responded strongly” to Hatfill, his apartment, and his girlfriend’s apartment but not to anyone else or any other location; and that Hatfill was probably involved in similar anthrax episodes in recent years. Based on these assertions, a reasonable reader of Kristof’s columns likely would conclude that Hatfill was responsible for the anthrax mailings in 2001.
Importantly, the Fourth Circuit did not address whether the statements in the columns might be either true or protected as statements of Kristof’s opinion — only whether the statements were capable of damaging Hatfill’s reputation.
Hatfill’s claims ultimately failed after he was found to be a limited-purpose public figure in connection with the larger controversy over government readiness for bioterrorist attacks. As a result, he was required to prove that Kristof published the statements at issue with constitutional “actual malice,” i.e., subjective knowledge of their falsity or a high degree of awareness of probable falsity. And, as the Fourth Circuit held, the record in the case contained “substantial evidence to support The New York Times’ contention that Kristof actually believed that Dr. Hatfill was the prime suspect,” including a long list of facts with respect to the FBI’s investigation whose truth Hatfill did not dispute. As a result, the court ruled that “no reasonable jury could find that Kristof had a high degree of awareness that Dr. Hatfill was not the anthrax mailer.”
The Fourth Circuit therefore affirmed summary judgment in the Times’ favor, and the U.S. Supreme Court denied certiorari, ending the case.
The FBI’s investigation closed without an arrest after a new primary suspect, Bruce Edwards Ivins, committed suicide in 2007; in 2008, the FBI declared that Ivins had been solely responsible for the attacks.
The applicability of the final legal conclusions in the cases discussed above is likely to be limited in any lawsuits that arise out of last week’s events. Although these cases invoke general principles of defamation law, precise interpretations of the law will vary from state to state, and decisions on issues such as truth, falsity and opinion will turn on the facts of particular cases. Nevertheless, the cases above remain interesting for the aspects of news reporting in a post-attack environment that the courts considered relevant to questions of liability and damage.
Similar aspects, including reports of preliminary investigations by law enforcement, reliance on photographs and other information of uncertain provenance or import, and efforts to engage in armchair sleuthing to deduce the responsible party, can be seen in the aftermath of last week’s events in both social media and professional reporting. Examples include:
There are, however, many significant differences from prior events. The most basic is the time frame, which was greatly accelerated from prior events. Articles indicating suspicion of Richard Jewell circulated three days after the Olympics bombing, and he was exonerated over two months later. The Time article inadvertently implicating Michael Schafer was published over three years after the Lockerbie bombing, and a correction ran more than a month later. The New York Times columns about Steven Hatfill began to run in 2002, several months following the anthrax mailings, and Hatfill was not unequivocally exonerated by the government until 2008. In the case of the Marathon bombings, the entire process of investigation, suspicion, misidentification and exoneration took less than a week: The identification of the Tsarnaev brothers as suspects was reported nationally by 8:00 a.m. on Friday, April 19.
This is not to suggest that last week’s misidentifications have not been devastating to the individuals involved and their families. Even brief identification as a terrorism suspect can cause tremendous distress, as shown by the desperation of a teenager depicted in the April 18 New York Post story to clear his name, and the increased suffering of Sunil Tripathi’s family, already in torment over his disappearance. That said, the pace of events was such that correct information was released within days, if not hours, so that at least some aspects of the ordeal were not prolonged.
Another important difference between last week’s behavior — particularly on Reddit — and prior incidents was the degree to which the ethics of attempting to identify suspects remained at the forefront of the discussion. An extended thread on Reddit (note: profanity beyond the link) discussed the propriety of crowdsourced efforts to identify suspects, including presumptions of innocence, lessons from the Jewell case, and the possibility of ruining an innocent person’s life through rushes to judgment. Users also maintained a list of cleared individuals (link to cached copy on Google) to limit the impact of the crowdsourced search. The architecture of Reddit itself, with its voting mechanism, functions to move discredited information out of public view. Reddit’s staff both publicly and privately apologized to the family of Sunil Tripathi and others affected, despite the fact that Reddit, under Section 230 of the Communications Decency Act, likely could not be held liable for user content.
I will leave to others the questions of the general utility of crowdsourcing in these situations, and whether this evidence of transparency and self-consciousness on Reddit offsets the consequences of users’ activity. For interesting recent perspectives on those questions, see: Alexis Madrigal’s piece in The Atlantic here and his subsequent exchange on Twitter with Dan Sinker of the Knight-Mozilla OpenNews Project; this article by James Surowiecki in The New Yorker; and this article here at Nieman Lab by Mike Ananny.
But however these questions are answered, the constant criticism and challenging of conclusions being drawn in coverage of last week’s events is significant. In past events, readers had no effective way to publicly question the information they were receiving from the media. Although the individuals wrongfully identified could file defamation claims in the courts to make a strong public denial of accusations in the press, this tactic is not without its hazards. Lawsuits by their nature cause media organizations to dig in their heels in defense of their reporting, leading to protracted litigation that a plaintiff would have to see through or abandon at the cost of appearing to admit the truth of the accusation. Social media provides another channel of feedback, allowing the public to act as a check on sloppy journalism and unfounded speculation; as Erik Wemple commented for The Washington Post, “When tragedy strikes America, Twitter remembers bad reporting.”
In the wake of confusing and senseless tragedy, there is a fundamental human impulse to try to exert control over an often impenetrably chaotic situation. Where a tragedy is caused by human action, that attempt to exert control often manifests as an attempt to identify those responsible. This is not new to last week’s events; the reporting and public discussion in the aftermath of the Marathon bombing bear striking similarities to reporting on past terrorist attacks where the attacker was initially unknown.
But despite these similarities, social media platforms allow the public to engage with the fragmentary information available after this kind of event in a much deeper way, as part of the community. While it is debatable whether this form of communal activity is on the whole beneficial or detrimental, social media also provide forums in which this debate itself can take place. This may help to ameliorate some of the more damaging consequences of past events, where misinformation was perpetuated over an extended period before the error was publicly revealed.
Jeff Hermes is director of the Digital Media Law Project. This is article is based on a piece originally published at the DMLP’s website.