STEPHEN F. WILLIAMS, Circuit Judge:
In the late ’80s and early ’90s articles appeared in the American press asserting an “October Surprise” — a scheme by members of the 1980 Reagan-Bush campaign team to thwart President Carter’s efforts to negotiate
the release of Iran’s American hostages by inducing the Iranians to delay their agreement. Ultimately (in January 1993) a bipartisan task force of the House of Representatives emphatically rejected these claims. See Joint Report of the Task Force to Investigate Certain Allegations Concerning the Holding of American Hostages by Iran in 1980, H. Rep. No. 102-1102 (102d Cong., 2d Sess.) (“Task Force Report”). In the meantime, however, defendant Craig Unger wrote and defendants Esquire Magazine and Hearst Corporation (collectively “Esquire”) published an article sketching out the conspiracy theory in lurid terms. The magazine’s October 1991 cover asked, “Did the Republicans conspire with Iran and Israel to delay the release of the hostages and steal the 1980 presidential election?” The article’s answer appeared to be a qualified Yes.
In a breathless and kaleidoscopic account rivaling an Oliver Stone movie, Unger writes that plaintiff Robert McFarlane, while an aide to Senator Tower, attended a February 1980 meeting with Iranian officials in Teheran — a meeting that “helped set up later meetings in Madrid, which in turn paved the way for the crucial October rendezvous in Paris.” (In the “October rendezvous” George Bush and William J. Casey are supposed to have closed the deal by which Iranians stalled hostage negotiations in exchange for promises of arms. But see Task Force Report at 173 (finding that the records and testimony “conclusively prove candidate George Bush’s whereabouts in October 1980,” and that he did
not
travel to Paris in the period alleged).) It is not said just how the February meeting “helped set up” later meetings (or how those “paved the way” for the supposed climax in Paris), but in the course of the account the defendants used language effectively calling McFarlane an Israeli spy. McFarlane focuses on this passage, which quotes from Ari Ben-Menashe, a self-professed former Israeli spy and a major source for conspiracy theorists:
In February 1980, Ben-Menashe says, Robert “Bud” McFarlane, then an aide to Senator John Tower, and Earl Brian, a businessman who had been secretary of health in Reagan’s California cabinet, met highly placed Iranian officials in Teheran. a [sic] Richardson on behalf of one of his clients, a computer-software company called Inslaw, Ben-Menashe states that both McFarlane and Brian had a “special relationship” with Israeli intelligence, McFarlane having been
recruited
by Rafi Eitan, a legendary Israeli agent who was the model for a leading character in John LeCarre’s
Little Drummer Girl. “McFar-lane was the famous Mr. X in the Pollard case,”
adds Ben-Menashe, referring to the trial of Jonathan Pollard, an American convicted of spying for Israel. In Pollard’s case there were persistent allegations about
another, unnamed American who secretly worked for the Israelis.
Both McFarlane and Brian have declined comment.
McFarlane and Brian’s visit, Ben-Me-nashe says, helped set up later meetings in Madrid, which in turn paved the way for the crucial October rendezvous in Paris.
(Emphasis added, except for
Little Drummer Girl.)
Unger’s article discloses doubts about Ben-Menashe’s credibility, including quotations from intelligence officials and journalists calling him a “fake” and a “con man.” The article also says that Ben-Menashe took a lie detector test for a news organization and “failed miserably,” but then adds that “it’s almost impossible to dismiss him.”
McFarlane brought suit against both Esquire and Unger in the U.S. District Court for the District of Columbia, alleging that the above passage falsely conveyed to Esquire’s readers that McFarlane was an Israeli spy and a traitor to his country. After discovery the district court granted defendants’ motion for summary judgment. It held that it had no personal jurisdiction over Unger and that no reasonable jury could have found, by clear and convincing evidence, that Esquire had published the piece with “actual malice,” i.e., with knowledge that the statements were false or with reckless disregard of their probable falsity, which McFarlane, as an undisputed public figure, would have to prove in order to win his case. See
New York Times Co. v. Sullivan,
376 U.S. 264, 279-80, 84 S.Ct.
710, 725-26, 11 L.Ed.2d 686 (1964);
Liberty Lobby, Inc. v. Dow Jones & Co., Inc.,
838 F.2d 1287, 1293 (D.C.Cir.1988). We affirm.
I.
Personal jurisdiction over Unger
McFarlane claims that the court had personal jurisdiction over Unger by virtue of subsection (3) or (4) of the District’s statute relating to personal jurisdiction based on conduct linked to the District:
(a) A District of Columbia court may exercise personal jurisdiction over a person, who acts directly or by an agent, as to a claim for relief arising from the person’s—
(1) transacting any business in the District of Columbia;
(3) causing tortious injury in the District of Columbia by an act or omission in the District of Columbia;
(4) causing tortious injury in the District of Columbia by an act or omission outside the District of Columbia if he [1] regularly does or solicits business, [2] engages in any other persistent course of conduct, or [3] derives substantial revenue from goods used or consumed, or services rendered, in the District of Columbia.
D.C.Code 1981 § 13-423(a). Although the D.C. Court of Appeals reads subsection (1) to extend to the “limits of due process,” see
Environmental Research Int’l v. Lockwood Greene Eng.,
355 A.2d 808 (D.C.1976), McFarlane does not assert it. See also
Colder v. Jones,
465 U.S. 783, 104 S.Ct. 1482, 79 L.Ed.2d 804 (1984) (exploring limits of due process as applied to personal jurisdiction over out-of-state writer of an article defaming a resident of the forum state). And the subsections he does assert are unavailing. The first, § 13-423(a)(3), requires that both tortious injury and an act predicate to it take place within the District. But Unger’s acts were not in the District; it is undisputed that he wrote the article in New York and delivered it to Esquire in New York. The case is therefore like
Moncrief v. Lexington Herald-Leader Co.,
807 F.2d 217 (D.C.Cir.1986), where the defendant in Kentucky published and mailed a newspaper that allegedly defamed the plaintiff in the District, and we found no basis for jurisdiction under subsection (3).
McFarlane argues that in a libel action the injury is part of the tort, so that, in law, the defendant has committed an act within the District. We rejected that claim in
Moncrief,
on the ground that to accept it would obliterate subsection (3)’s careful distinction between “injury” and “act.”
Id.
at 220-21. McFarlane appears to concede that circuit precedent excludes Unger from the purview of § 13-423(a)(3), and invites us to overrule
Moncrief.
But we have no power to do so, even if we thought it desirable. One panel of the court does not have authority to overrule another. See, e.g.,
United States v. Caldwell,
543 F.2d 1333, 1369 n. 19 (D.C.Cir.1974).
Section 13-423(a)(4) is of no more help to McFarlane. It contemplates jurisdiction when there is tortious injury within the District accompanied by any of three specified kinds of additional contacts between the District and the defendant, not necessarily related to the contested act or injury.
Crane v. Carr,
814 F.2d 758, 763 (D.C.Cir.1987). McFarlane has not shown that Unger had any of the specified types of contacts. His appeal makes no claim at all on the third type—deriving substantial revenue from goods used or consumed, or services rendered, in the District. Invoking the first and second types (regular business or persistent course of conduct), he points to the fact of Unger’s having written for six national publications whose circulations include the District and, on two occasions, writing for District-based publications (the
Washington Post
and the
New Republic).
But
writing
an article for a publication that is circulated throughout the nation, including the District, hardly constitutes doing or soliciting business, or engaging in a persistent course of conduct,
within
the District. The writer is not the publisher; Unger’s contacts must be assessed separately. See
Keeton v. Hustler Magazine,
465 U.S. 770, 781 n. 13, 104 S.Ct. 1473, 1482 n. 13, 79 L.Ed.2d 790 (1984).
Thus we are left with two articles appearing in Washington-based publications, one in
the
Washington Post
(which was written after MeFarlane’s complaint was filed and is therefore no basis for personal jurisdiction, see
Minghoffer v. S.N.C. Achille Lauro,
937 F.2d 44, 52 (2d Cir.1991);
Asarco, Inc. v. Glenara, Ltd.,
912 F.2d 784, 787 n. 1 (5th Cir.1990)) and one in the
New Republic.
If “regularly” and “persistent” are to have any meaning, sale of two articles to District-based publications over a career in journalism cannot amount to “regularly” doing business or to a “persistent” course of conduct. Thus, without regard to the circuit’s “news-gathering exception” to jurisdiction, see
Moncrief,
807 F.2d at 222-25, we agree with the district court that McFarlane has failed to show contacts between Unger and the District satisfying § 13-423(a)(4). We affirm its finding of a lack of personal jurisdiction over Unger.
We also affirm the district court’s decision to dismiss instead of transfer the case against Unger. The plaintiff asked for a transfer under 28 U.S.C. § 1406(a), which permits transfer “in the interest of justice” of a “ease laying venue in the wrong division or district.” Where personal jurisdiction is absent and venue is proper (or at least unchallenged, as here), courts have disagreed as to whether a request for transfer should be analyzed under § 1406(a) or instead under 28 U.S.C. § 1404(a). See generally 15 Charles A. Wright, Arthur R. Miller & Edward H. Cooper, Federal Practice and Procedure § 3827, at 263-67 (1986). In this particular case it appears to make no difference. Our standard of review is the same—abuse of discretion—under both sections.
Piper Aircraft Co. v. Reyno,
454 U.S. 235, 257, 102 S.Ct. 252, 266-67, 70 L.Ed.2d 419 (1981) (adopting abuse of discretion standard, noting need to balance multiple factors, in transfer under § 1404(a));
Naartex Consulting Corp. v. Watt,
722 F.2d 779, 789 (D.C.Cir. 1983) (applying abuse of discretion, without discussion, in transfer under § 1406(a)). And, although § 1404(a) calls on the court to consider “the convenience of parties and witnesses” (in addition to “the interest of justice,” which both sections specify), neither party addresses arguments uniquely pertinent to convenience. Accordingly, we need not try to resolve the “nearly hopeless muddle” of conflicting reasoning and precedent as to which statute properly applies.
Ellis v. Great Southwestern Corp.,
646 F.2d 1099, 1106-07 (5th Cir.1981).
The district court’s denial of a transfer matters because the statute of limitations for defamation has now apparently expired in New York, the place where McFarlane could most readily secure jurisdiction over Unger. The district judge explained that while denying such a transfer thus exacts a “heavy penalty,” the fault for the problem rests with McFarlane, who was put on notice, by Un-ger’s answer to the complaint, of Unger’s intention to rely on a defense of lack of personal jurisdiction, and nonetheless failed to file a protective suit in New York. The district court further reasoned that McFar-lane would not be prejudiced by denial of transfer, as Esquire had adequate insurance. That strikes us as something of an oversimplification, as McFarlane's quest for vindication would more easily be satisfied (as we shall see) in a suit against Unger than against Esquire, because of Unger’s far greater awareness of the reasons to doubt the truth of the article’s claims. Nonetheless, because of McFarlane’s notice of the defense and his counsel’s inexplicable failure to file a protective suit, we think the district court was within its discretion in denying a transfer.
II.
Evidence of Esquire’s actual malice
McFarlane concedes that he is a public figure, having held various high-level positions in the Reagan Administration, including those of special envoy to the Middle East and National Security Advisor. As such, McFar-lane must show that the evidence in the record could support a reasonable jury finding by clear and convincing evidence that the allegedly defamatory statements were made with actual malice,
Anderson v. Liberty Lobby, Inc.,
477 U.S. 242, 255-56, 106 S.Ct. 2505, 2513-14, 91 L.Ed.2d 202 (1986), i.e., with “knowledge that [they were] false or with reckless disregard of whether [they were] false or not,”
New York Times Co. v. Sullivan,
376 U.S. at 279-80, 84 S.Ct. at 725-26. See also
Harte-Hanks Communications,
Inc. v. Connaughton,
491 U.S. 657, 692, 109 S.Ct. 2678, 2698, 105 L.Ed.2d 562 (1989) (actual malice shown if there is a “purposeful avoidance of the truth”);
Tavoulareas v. Piro,
817 F.2d 762, 775-76 (D.C.Cir.1987) (citing cases). We first consider whether Unger’s state of mind may be attributed to Esquire for these purposes, and then, having concluded that it may not, examine the evidence as to the state of mind of Esquire’s employees.
A.
Attribution of Unger’s state of mind to Esquire
If Unger had been an employee of Esquire, his state of mind could undoubtedly be attributed to his employer. See
Cantrell v. Forest City Publishing Co.,
419 U.S. 245, 253-54, 95 S.Ct. 465, 470-71, 42 L.Ed.2d 419 (1974). As McFarlane effectively acknowledges, however, Unger was working as an independent contractor, not as an employee. McFarlane notes that Unger did use the prestige of Esquire by representing himself to interviewees as working for Esquire, so we assume that Esquire may have established some sort of agency relationship with Unger. Thus we consider the question whether the malice of a non-employee agent can be imputed to the principal under
New York Times v. Sullivan.
New York Times
itself is a good place to start. There the Court threw out the case against four individuals who (the Court assumed) had authorized use of their names as makers of the defamatory statement, explaining that “there was no evidence whatever that they were aware of any erroneous statements or were in any way reckless in that regard.” 376 U.S. at 286, 84 S.Ct. at 729. Thus the Court refused to impute to the individuals as principals any information in the minds of persons they authorized to act as their agents in the matter.
Nonetheless, a different rule might apply to a corporate defendant, which can act only through agents of one sort or another — employees or non-employee agents. Yet the courts have not so found. Commonly they have interpreted
Cantrell
as barring liability on any theory other than
respondeat superior
(which is limited to employees). See, e.g.,
Price v. Viking Penguin, Inc.,
881 F.2d 1426, 1446 (8th Cir.1989);
Secord v. Cockbum,
747 F.Supp. 779, 787 (D.D.C.1990). But
Cantrell
doesn’t say that. The writer in question was an employee of the corporate defendant, and, although the trial court had given an instruction somewhat muddling the categories of employee and agent, no one had objected.
Cantrell,
419 U.S. at 253-54 & n. 6, 95 S.Ct. at 470-71 & n. 6. So
Cantrell
presented no occasion for the Court to address the issue of when the mental state of non-employee agents may be imputed to the principal. See also
Masson v. New Yorker Magazine, Inc.,
832 F.Supp. 1350, 1371 (N.D.Cal.1993) (discussing Cantrell).
McFarlane invokes
Gertz v. Robert Welch, Inc.,
680 F.2d 527, 539 n. 19 (7th Cir.1982), but in that case the Supreme Court had already determined that the plaintiff there was not a public figure, so that states were free to impose liability on whatever ground they chose, “so long as they do not impose liability without fault.”
Gertz v. Robert Welch, Inc.,
418 U.S. 323, 347, 94 S.Ct. 2997, 3010-11, 41 L.Ed.2d 789 (1974). On remand the Seventh Circuit applied an actual malice standard anyway, because Illinois law had adopted that standard for statements derived from public proceedings (some of which were at issue), and because (it ultimately concluded) the plaintiff had satisfied even that heightened standard with respect to yet other statements not covered by that privilege.
Gertz,
680 F.2d at 537. As
Masson
explains, 832 F.Supp. at 1872-73, in so doing the court was willing to attribute the knowledge of the independent-contractor writer to the defendant under Illinois agency law for purposes of Illinois defamation law.
Gertz,
680 F.2d at 539 n. 19.
Gertz
neither asserts nor cites authority for the proposition that vicarious liability can be the basis for finding actual malice under
New York Times,
apart from
respondeat superior.
In the next section of its opinion, to be sure, the
Gertz
court affirmed the imposition of punitive damages against the principal,
Gertz,
680 F.2d at 540, which under federal constitutional law requires a showing of actual malice, see
Gertz,
418 U.S. at 349, 94 S.Ct. at 3011-12;
Dun & Bradstreet, Inc. v. Greenmoss Builders, Inc.,
472 U.S. 749, 756-57, 105 S.Ct. 2939, 2943-44, 86 L.Ed.2d 593 (1985), but with no discussion or analysis of the issue. In practice, then,
Gertz
has little significance as a ruling on actual malice under the
New York Times
standard, and our own analysis leads us to conclude that under
New York Times
actual malice may not be attributed outside
respon-deat superior.
Liability for the acts of a non-employee agent normally flows from the principal’s authorization of the agent’s acts, and establishment of the principal-agent relationship as a threshold matter is based largely upon control of one party by the other. See Restatement of Agency 2d §§ 140,14. The question here is what
kind of
control (if any) might suffice to tie Esquire to Unger’s knowledge for these purposes. In
Hunt v. Liberty Lobby,
720
F.2d
631 (11th Cir.1983), the court addressed the possibility that a principal might be liable for the malice of a non-employee agent on the grounds of its control over the agent’s activities. Cf. Restatement of Agency 2d § 254 (discussing imputation of agent’s defamation to principal, without reference to malice). But the court in
Hunt
seemed to assume that the positions of independent contractor and of agent were mutually exclusive,
Hunt,
720 F.2d at 649, which they are not, Restatement of Agency 2d § 14 N cmt. a, and therefore conflated the enterprise of showing someone to be a de facto employee with that of showing agency. In any case, the court found no control by the publisher, as the writer had simply sold the publisher a finished product.
Hunt,
720 F.2d at 648-49 & n. 30. This overlooks the fact of the publisher’s ability to edit and, ultimately, to veto the publication—the court does not say that the article was sold on a take-it-or-leave-it basis. Similarly, in
Price v. Viking Penguin,
881 F.2d 1426, the court examined the editor’s relation to the author. Finding that it was limited to matters of structure, priorities and clarity, the court declared that the author “clearly was not an employee of Viking such that liability could be imputed, were we to find any,”
id.
at 1446. Despite the court’s reliance on the finding, that the writer was not an employee, its examination of the editing process suggests that it may have viewed some forms of editorial control as a possible basis for vicarious liability for the state of mind of a non-employee agent, classification of the writer as an employee being clearly out of the question.
Although
Hunt
and
Price
seem to require an employment relationship, they might also be understood as supposing that some kinds of intense editorial involvement by a publisher’s employees might entangle them in the independent writer’s thought process enough to serve as a basis for holding the publisher vicariously liable. But why should editorial controls that take the employee editors themselves to a point
short
of actual malice be deemed to establish the publisher’s actual malice? One answer might be that not to do so creates a perfect escape hatch for deep-pocket publications. But the answer is not completely satisfactory, as any writer who knowingly ventures into legally risky waters, and who is reluctant to experience personal bankruptcy, would presumably demand an indemnity agreement, putting the publisher on the line. Further, actual malice is a First Amendment protection predicated on a subjective state of mind,
Hutchinson v. Proxmire,
443 U.S. 111, 120 n. 9, 99 S.Ct. 2675, 2680 n. 9, 61 L.Ed.2d 411 (1979) (citing
New York
Times), which surely cuts against any extension of vicarious liability beyond respondeat superior. Because we doubt that actual malice can be imputed except under
respondeat superior,
and because in any case McFarlane presents no evidence showing Esquire’s supervision, of the
process
by which Unger turned raw data into finished article (as distinct from control over his final product), cf. Restatement of Agency 2d § 14 N cmt. b, we conclude that McFarlane may show Esquire’s malice only through evidence of the information available to, and conduct of, its employees.
B.
Evidence that Esquire editors had actual malice
Because both parties in their original briefs made little effort to distinguish between the possible malice of the two defendants, we ordered an additional round of briefing directed to evidence of actual malice on the part of Esquire’s editors. We as
sume, in favor of McFarlane, that information in the mind of each individual Esquire editor may be aggregated with information in the mind of every other editor, but we do not decide the point. Our review persuades us that McFarlane failed to submit evidence from which a jury could find such malice.
McFarlane’s supplemental brief on the issue is obscure. In a series of paragraphs the brief asserts that “Esquire editors knew ...” various facts. The facts are largely ones that appear on computer disks containing Unger’s transcriptions of his notes of interviews with various sources, which Unger sent over to Esquire. But with one exception the Esquire editors questioned on the matter said (and McFarlane offers no reason why a jury should disbelieve them) that they did not examine the transcripts; the single exception, research editor Mark Warren, said that he had spot checked some of them—so far as appears, only those of Ben-Menashe and Richardson. But McFarlane does not claim material conflicts between the Ben-Menashe transcripts and the published article’s accounts of his claims. Thus, as we shall see, McFarlane is driven largely to rely on Esquire’s alleged distortions of what Richardson had to say.
McFarlane’s contentions that Esquire had actual malice comprise four basic points. First, he suggests that Esquire editors not only had reasons to doubt Ben-Menashe’s truthfulness but also in fact doubted it—so much so that their decision to publish without further corroboration showed malice. Second, McFarlane claims that Esquire either knowingly, or with reckless disregard, participated in Unger’s (alleged) fabrication of a “pedigree” for Ben-Menashe in the form of an endorsement by Elliot Richardson. Third, McFarlane suggests that general aspects of Esquire’s presentation of the article—segment headings, promotional blurbs elsewhere in the magazine, and the yellow highlighting of certain passages—add to the “defamatory sting” of the specific passage on McFarlane. Finally, McFarlane says that Esquire editors fabricated the article’s statement that he had denied comment. We review these points in turn, recognizing that McFarlane is entitled to an aggregate consideration of all of these claims—with the evidence construed most favorably to him—to see if he has met his burden.
Tavoulareas,
817 F.2d at 794 n. 43.
1.
Reasons to doubt the credibility of Benr-Menashe
Ben-Menashe is the source for many of the details in “October Surprise,” including the ones in the passage under attack in this case. Several of Unger’s sources made clear to him their belief that Ben-Menashe was a liar, and Esquire does not deny its awareness of these views; it couldn’t, as the article passed a substantial chunk of them through to the readers. The article directly quotes a former CIA officer and a Washington Post journalist as calling him, respectively, a “liar” and a “con man.” It notes that when he took a he detector test he “failed miserably,” and it quotes an ABC News producer as saying that in the he detector test Ben-Menashe “goes way off the chart on ah relevant questions. My theory is that a lot of what he says is true, but that Ari exaggerates his own role and muddies the water.” But full (or pretty full) publication of the grounds for doubting a source tends to rebut a claim of malice, not to estabhsh one. See
Tavoulareas,
817 F.2d at 788 n. 35.
We are not, of course, saying that one may altogether shield a defamation simply by including the source’s reputation as a liar. Here Esquire supphed an answer of sorts to the question of why, knowing Ben-Menashe’s flaws, they still saw fit to pass his accusations on to its readers. Editor William Blythe approved addition of the phrase, “Yet it’s almost impossible to dismiss him,” directly after the recitation of Ben-Menashe’s vulnerabilities. Explaining the decision, he testified, ‘We wouldn’t have used him as a source unless we thought he had some knowledgeability_ We ... knew that he ... was the [sic; “a”?] source of the Iran-Contra story, and certainly that had cheeked out, and also that Congress was investigating Ari Ben-Menashe’s charges and using him as a witness.” Esquire’s editor-in-chief, in the course of elaborating on his denial that he “knew that what Ben-Menashe was saying was as likely to be false as it was to be true,”
also pointed to Ben-Menashe’s apparent vindication in Iran-Contra. And the editors relied on Unger’s record; he had worked with Esquire editor David Hirshey on two earlier pieces, and Hirshey had found him “exemplary.” Reliance on a reporter’s reputation can indeed show a lack of actual malice by a publisher. See
Speer v. Cutaway Newspapers, Inc.,
828 F.2d 475, 478 (8th Cir.1987);
McManus v. Doubleday & Co.,
513 F.Supp. 1383, 1390 (S.D.N.Y.1981). Cf.
Washington Post Co. v. Keogh,
365 F.2d 965, 971-72 (D.C.Cir.1966) (exploring difficulties in using writer’s questionable reputation against the publisher). McFarlane does not dispute the validity of these bases for Esquire’s going forward (though he does, of course, assert Esquire’s allegedly inadequate attention to countervailing negatives).
As to Ben-Menashe’s being interviewed by congressional staffers, we have in the past given weight to the fact that a source told congressional investigators the same story that he told the defendant, under circumstances where lying could have serious consequences. See
Tavoulareas,
817 F.2d at 791 (source gave information to congressional investigators in formal setting which, though without oath, could possibly lead to criminal liability for lying). But no one should suppose that such colloquies add up to some sort of congressional endorsement. That it may make sense for congressional staffers to give the time of day to a wildly improbable source, on the off chance that he will yield something useful, tells virtually nothing about the source’s credibility. Further, Esquire does not even claim to have had information that Ben-Menashe’s statements to the investigators were the same as to Unger, and it offered no evidence that the circumstances of the interviews in any way would have alerted Ben-Menashe to the penalties for lying.
McFarlane says that Esquire “reviewed” a report by the PBS television documentary “Frontline,” showing that an arms dealer, Houshang Lavi,
not
Ben-Menashe, attended a meeting at the L’Enfant Plaza with some Reagan campaign foreign policy advisers.
(This
meeting actually occurred. See Task Force Report at 109-18.) From the fact that the final article did
not
mention the L’Enfant Plaza meeting, whereas Unger’s original draft did, expressing doubt about Ben-Menashe’s claim to have been present there, McFarlane argues that the jury could infer a willful intent to suppress an instance where Ben-Menashe was proven flat-out false. It is not clear to us that the pronouncements of an obscure international arms dealer are so self-evidently true that they could be said to establish the falsity of Ben-Menashe’s claims, though of course the contradiction is not trivial. (In fact, the strongest evidence presented to the Task Force on the identity of the foreign interlocutor, a contemporaneous memo by participant Richard Allen, identifies him as one A.A. Mohammed, a Malaysian and so far as appears in no way connected to Ben-Me-nashe or Lavi. See
id.
at 115.) Moreover, McFarlane has not directed our attention to any depositions indicating what real contact Esquire editors may have had with the Frontline report or with raw data as to the L’Enfant Plaza meeting generally.
In sum, given Ben-Menashe’s supposed role as a source in Iran-Contra, Unger’s reputation with Esquire, and the inherent difficulties in verifying or refuting a claim that someone is the agent of a foreign power, the proofs do not add up to the possibility of a reasonable jury finding of clear and convincing evidence of reckless awareness of probable falsity, and in no way show an actual belief in falsity.
2.
Fabrication of the Elliot Richardson endorsement of Ben-Menashe’s credibility
McFarlane claims that “October Surprise” artificially boosted Ben-Menashe’s credibility, fabricating a “pedigree” for him with quotations from Elliot Richardson — referred to in the article as the “moral hero” of Watergate — purporting to show that Richardson took Ben-Menashe seriously. First, recall that the disputed passage includes the following sentence:
In a sworn affidavit submitted by Elliott [sic] Richardson on behalf of one of his clients, a computer-software company called Inslaw, Ben-Menashe states that both McFarlane and Brian had a “special
relationship” with Israeli intelligence, MeFarlane having been recruited by Rafi Eitan, a legendary Israeli agent who was the model for a leading character in John LeCarre’s
Little Drummer Girl.
The statement is technically false in one narrow respect—Richardson evidently did not appear as counsel of record in the Inslaw case. But he later said (and it is not disputed) that he was on the legal team for Inslaw, and he evidently so represented himself to Unger, who was Esquire’s only claimed source of information about Ben-Menashe’s Inslaw affidavit.
The sentence has a far more serious flaw. Its latter part, beginning with “MeFarlane having been
recruited
by Rafi Eitan,” is not in the affidavit. Nor indeed is there anything in the affidavit remotely supporting the statement in the article’s following sentences, making MeFarlane out to be an equivalent of Pollard—a spy who had pled guilty to, in effect, selling masses of top secret U.S. material to the Israelis. While MeFarlane does not dispute that Ben-Menashe made these assertions to Unger, the text of the affidavit supports only the idea of a “special relationship.” Thus, says MeFarlane, Esquire stretched the Richardson imprimatur from the relatively innocent “special relationship” all the way to the charge of espionage.
Esquire offers two answers. First, it now argues that there is “no material difference” between the “recruited by” and “special relationship” phrases. This is fanciful. “Special relationship” seems infinitely elastic, while “recruited,” in context, suggests a switch of allegiance to a foreign power. Second, Esquire’s editors testified that they recognized that the passage might appear to suggest that the whole accusation was in the affidavit, but that they ordered changes—their recollection was the addition of the comma—to indicate the contrary. Their brief also notes that the
affidavit
surely would not have cited
Little Drummer Girl,
alerting the reader to the fact that the spy charge was not advanced in the affidavit. Although we find the comma theory pretty thin, this imprimatur stretch seems as consistent with linguistic muddle as with reckless disregard, and in context not enough, even in conjunction with other evidence, to show actual malice by Esquire editors.
MeFarlane puts more stress on Esquire’s use of a truncated quote from Richardson in a mention of the Ben-Menashe affidavit in the course of a generic canvassing of Ben-Menashe’s spotty credibility:
And former attorney general Elliott [sic] Richardson ... has submitted sworn affidavits by Ben-Menashe on behalf of a client. A standard legal gambit, perhaps, but Richardson finds Ari Ben-Menashe a compelling witness. “I take him seriously as being who he says he is,” says Richardson.
Richardson’s actual comment, as reflected in Unger’s earlier draft and substantially corroborated in his notes, was, “Quite apart from what he knows, I take him seriously as being who he says he is.” Esquire’s editors removed the first clause from Unger’s draft, in McParlane’s view greatly increasing the intensity of the endorsement.
Esquire argues that any problem here is solved by an affidavit of Richardson, filed in this litigation, saying that “[t]he statements and quotations attributed to me accurately reflect statements I made to Mr. Unger, and accurately reflect my views of Mr. Ben-Me-nashe.” But while the statement may aid Esquire in a defense of truth (on a rather peripheral matter,
not
the defamatory material itself), it does not dispose of the question of actual malice, which turns on Esquire’s subjective beliefs and purposes at the time of publication. Esquire saw Unger’s notes, and they speak for themselves.
Nonetheless, we think the explanation by Esquire—that they deleted the clause because of space considerations and because of its ambiguity—altogether plausible. “Quite apart from what he knows” could be taken to mean that Richardson believed Ben-Menashe’s claimed identity but not his statements, or it could mean that Richardson wanted to emphasize his belief in Ben-Me-nashe’s identity, not simply Ben-Menashe’s apparent knowledge. The clause
is
ambiguous. And with or without it this Richardson endorsement seems to add up to very little.
We are more troubled about a different discrepancy between Unger’s notes and the final article. The notes, quoting Richardson’s statements to Unger about Ben-Me-nashe, say (correcting obvious typographical errors):
[0]ne thing that is true of people like him is that they live in a world of such constant deception [ ] that they are used to moving without misstep between truth and fabrication so he doesn’t help much with the October surprise story to verify the truth.
As Esquire put most of its eggs in the Ben-Menashe basket, both for October Surprise generally and for the accusation against McFarlane, this seems a sharp renunciation of its star witness by his putative champion. The statement is in Unger’s notes of his talks with Richardson, and it is undisputed that Mark Warren browsed in precisely those notes. But at no time, so far as we can tell, did McFarlane’s counsel ever ask Warren whether he’d spotted this passage and, if so, what he made of it. Accordingly, its presence, though puzzling, cannot much help McFarlane.
We are also troubled by the fact of Unger’s uncontradicted testimony that he sent to Esquire a draft including a quote about Ben-Menashe from Richardson, “I can’t tell if the son of a bitch is telling the truth.” The editors apparently removed it, with the purpose, according to Unger, of “maintaining the integrity of what Mr. Richardson said.” This appears to be some sort of jargon for suppressing material inconsistent with a broad effort to build up Ben-Menashe’s credibility. Still, in view of the article’s inclusion of solid material damning Ben-Menashe, the removal gets only limited weight.
3.
Embellishments
McFarlane points to aspects of the article’s presentation that generally hype its broad theme of Republican skullduggery—the yellow highlighting of certain sentences in the article that push the theory forward (but not ones that detract from it), the inclusion of a cover sticker [“OCTOBER SURPRISE DID THE Republicans conspire with Iran and Israel to DELAY the RELEASE of the hostages
and
steal
the
1980 PRESIDENTIAL election?”] (typography in original), the placement of color photographs, and the use of bold section titles. But while these displays of bias may be the sort that deprive Esquire of any “neutral reporting” privilege,
In re UPI,
16 Media L. Rep. (BNA) 2401, 2408 (D.D.C.1989), but see
White v. Fraternal Order of Police,
909 F.2d 512, 514, 528 (D.C.Cir.1990) (leaving open the scope of the privilege under District law), none of the displays relates directly to McFarlane or goes directly to Esquire’s subjective intent to tell a mistruth or speak with willful disregard of truth. “The fact that a commentary is one sided and sets forth categorical accusations has no tendency to prove that the publisher believed it to be false.” ’
Westmoreland v. CBS, Inc.,
601 F.Supp. 66, 68 (S.D.N.Y.1984).
4.
The report of McFarlane’s refusal to comment
In the last stages of editing, Esquire changed the article’s treatment of McFar-lane’s response to the charges. “McFarlane ... denied the charges” became “McFarlane ... refused comment.” Esquire’s David Hir-shey testified that they had been holding space for McFarlane’s response, but were told by Unger that McFarlane had refused to agree to any kind of interview. Indeed, McFarlane wrote that he thought he would be “of little help” to Unger, essentially declining to be interviewed.
McFarlane notes that Unger’s letter requesting help from McFarlane was phrased very blandly, saying that he was about to do an article on the October Surprise and asking for an interview. There was not a clue as to the depths of the charges (“By the way, I plan to accuse you of being an Israeli spy.”).
And Mark Warren acknowledged having seen a copy of the letter to MeFarlane, and the latter’s reply, in the editing process. In a later telephone call, according to MeFar-lane’s executive assistant, Unger mentioned October Surprise and Inslaw, but, again, not a word about espionage. Thus Warren and Esquire were on some notice of what had led to McFarlane’s position. In retrospect, this looks at least very careless, but not enough to entangle Esquire in Unger’s apparent misleading, which would be necessary for Esquire’s behavior to help show actual malice on its part. “[P]laintiff must prove more than an extreme departure from professional standards.”
Harte-Hanks Communications, Inc. v. Connaughton,
491 U.S. 657, 665, 109 S.Ct. 2678, 2685, 105 L.Ed.2d 562 (1989).
The standard of actual malice is a daunting one. MeFarlane has been accused of espionage and, because the trial court rightly believed that want of adequate evidence of malice was the simplest way of addressing the claim against Esquire, has as yet had no opportunity to secure vindication. Compare Arnold A. Lubasch, “Time Cleared of Libelling Sharon But Jurors Criticize its Reporting,” New York Times, Jan. 25, 1985, at A1 (reporting jury finding that Time’s story about Israeli general Ariel Sharon was false and defamatory but not published with actual malice); cf.
Herbert v. Lando,
603 F.Supp. 983, 988 (S.D.N.Y.1985) (noting submission of special verdict to jury in
Sharon v. Time, Inc.,
83 Civ. 4660CADS), 1985 WL 186675 (S.D.N.Y.1985)). Nonetheless, as the dismissal was correct, we must affirm.
Because of the correctness of the finding on malice and of the trial court’s dismissal of the claim against Unger for want of personal jurisdiction, the judgment is
Affirmed.