Hobson v. Wilson 737 F.2d 1

HARRY T. EDWARDS, Circuit Judge:

This case presents yet another chapter in the saga of the Federal Bureau of Investigation’s notorious COINTELPRO operation. It is now clear that COINTELPRO has long been abandoned; but, as this case demonstrates, its victims have remained vigilant in seeking redress for past wrongs.

In 1976, several Washington area residents, who had been politically active in the late 1960s and early 1970s, brought suit in District Court claiming that certain of their constitutional rights had been violated. The plaintiffs sought damages and injunc-tive relief against numerous active and retired special agents of the Federal Bureau of Investigation (FBI) and members of the Metropolitan Police Department (MPD), as well as the District of Columbia itself. The amended complaint, filed October 28, 1977, alleged that each defendant had violated plaintiffs’ constitutional rights, individually and through conspiracies, while plaintiffs engaged in lawful political protest against Government policies.1

Following a 17-day trial, over which Judge Oberdorfer ably presided, a jury returned verdicts against most of the defendants and in favor of most of the plaintiffs. The jury also awarded substantial compensatory and punitive damages: awards to the eight prevailing plaintiffs, against the thirteen defendants found liable, totalled $711,937.50.2

On appeal, defendants have raised a number of arguments. We have given every argument thorough consideration and, following a painstaking effort — including examination of the record, verdicts and decisions from the trial court, review of the parties’ briefs and study of the relevant statutory and case law — we have reached the following conclusions: (1) the claims of three of the prevailing plaintiffs against the FBI defendants were barred by the statute of limitations, and the judgments in their favor cannot stand; (2) the evidence was insufficient to support findings of liability against the individual MPD defendants and the District of Columbia, and the judgments against them cannot stand; (3) the evidence was insufficient to support findings that the FBI defendants partici*8pated in a conspiracy with MPD officials, and findings of liability for their participation in a joint FBI-MPD conspiracy are reversed; (4) consistent with our holdings herein, the damages issues are remanded for further consideration; and (5) in all other respects the judgments at trial — including the findings of individual and conspiratorial liability against the FBI defendants — are affirmed, except for our remand for reconsideration of one issue raised on cross-appeal.

Background

I. The Parties

We begin with a brief review of the parties to this action. Because each plaintiff brings an individual claim, as well as conspiracy claims, against each defendant, it is crucial that we set forth clearly the role that each played during the relevant time period.

In the late 1960s and early 1970s, each plaintiff engaged in a variety of activities— such as organizing meetings and demonstrations, and publishing newsletters — to express disagreement with, and rally support against, certain national and local Government policies. Generally, they focused their efforts on three- issues: military involvement in Vietnam, proposals to build a superhighway through the District of Columbia, and equal rights for Black citizens of the District of Columbia. Each individual plaintiff played a leadership role in one or more of these efforts.

Plaintiff Sammie Abbott, a graphic artist, organized and was active in the Emergency Committee on the Transportation Crisis (ECTC), a coalition of Black and White neighborhood associations that opposed freeway construction through the District of Columbia. Plaintiff Abe Bloom, an engineer, was especially active in antiwar organizations, such as the Washington Mobilization Committee (WMC) and the Washington Area Peace Action Coalition (WAPAC), which organized major antiwar demonstrations in the District of Columbia. Plaintiff Reginald Booker, a federal employee, worked with Mr. Abbott on the freeway issue, helped to organize the Black United Front (BUF), and tried to foster local Black opposition to the Vietnam war.

Plaintiff Tina Hobson, a federal employee, participated in the organization of antiwar demonstrations and worked with her husband, the late Julius Hobson, who was active in the Black United Front. Plaintiff Richard Pollock, a college student at the relevant time, was principally involved in antiwar protest, and was active in the Student Mobilization Committee and other antiwar organizations. Plaintiff the Reverend David Eaton, minister of All Souls Church, was one of the founders and leaders of the Black United Front and participated in planning the 1968 Poor People’s Campaign, a major civil rights demonstration.

Plaintiff Arthur Waskow, a Fellow at the Institute for Policy Studies — a research center whose scholars focused on public policy, race relations and foreign policy— organized and spoke at several antiwar demonstrations. Plaintiff Washington Peace Center, an organization affiliated with the Society of Friends, participated in the organization of several antiwar demonstrations during the relevant period.3

During the same period, defendants Brennan, Moore, Jones, Grimaldi and Pang-burn were employed by the FBI. Defendant Charles D. Brennan was chief of the Internal Security Section in FBI headquarters from 1966 to 1970. From 1970 to 1971, Mr. Brennan was assistant director in charge of the Domestic Intelligence Division. In the former position, he was responsible for the FBI’s COINTELPRONew Left activity.4 In the latter job, he had overall responsibility both for COIN-*9TELPRO-New Left and COINTELPROBlack Nationalist.5 Mr. Brennan apparently authored an FBI memorandum announcing the creation of COINTELPRO-New Left in 1968, and another memorandum bearing his name recommended termination of the program in 1971.

Defendant George C. Moore was chief of the Racial Intelligence Section in FBI headquarters from 1967 to 1974. The section supervised field office implementation of COINTELPRO-Black Nationalist. The internal FBI memorandum announcing COINTELPRO-Black Nationalist went out over his name.

Defendant Courtland J. Jones held a liaison position between the Washington Field Office (WFO) of the FBI and FBI headquarters. From 1964 to 1974, he was WFO security coordinator and supervised nine squads, including Squad 5 and Squad 7. Squad 5 initially handled both New Left and racial matters but subsequently was divided into two squads. Squad 5 then handled racial matters and COINTELPRO-Black Nationalist, and Squad 7 was responsible for the New Left and COINTELPRO-New Left.

Defendant Gerald T. Grimaldi was a special agent in the WFO from 1956 to 1971. From the mid-1960s to 1970, he was a member of Squad 7, which investigated persons and individuals affiliated with the New Left; and from 1968 to 1970, he was the designated COINTELPRO-New Left Coordinator. In April 1971, he became supervisor of Squad 5, which handled racial matters and COINTELPRO-Black Nationalist.

Defendant Gerould W. Pangburn, a WFO Special Agent, was a member of Squad 7 in 1968, was then transferred to Squad 5, and was Squad 5 supervisor from 1972 to 1974. In both squads he handled investigations of alleged “racial extremists.”

Plaintiffs also filed claims against MPD officials, as well as against the District of Columbia itself. Defendant Jerry V. Wilson was Chief of Police during the relevant period. Defendant Thomas J. Herlihy supervised the MPD Intelligence Division, and was responsible for its activities, until his retirement in 1973.

Sergeant Jack L. Aeree and Lieutenant George R. Suter were successive supervisors in the MPD Security Information Unit of the Intelligence Division. Sergeant Aeree was in the Unit from 1966 to 1972; Lieutenant Suter was assigned to the Unit from 1968 to 1970 and from 1972 to 1975.

Defendants Christopher Scrapper, Edward Jagen and John W. Mahaney were officers in the MPD Intelligence Division. Officer Jagen served undercover after 1971 and attended meetings of the Washington Area Peace Action Coalition (WAPAC) and the Peoples Coalition for Peace and Justice (PCPJ). Sergeant Scrapper supervised officer Jagen and other informants in the antiwar movement. Officer Mahaney supervised at least two undercover officers, one of whom obtained information on BUF and ECTC, the other of whom was active in two antiwar groups.6

II. The Facts

According to plaintiffs, the FBI defendants conspired with each other, with other FBI agents and with the District defendants 7 to impede plaintiffs’ efforts to associate with others for the purpose of publicly expressing opposition to the Vietnam War, national and local Government race relations policies, and other Government actions. Many of the defendants’ activities from which plaintiffs alleged they were injured were related to COINTELPRO, an FBI program begun in 1967 and discontin*10ued in the early 1970s.8 COINTELPRO had two components; one was COINTELPRO-New Left, which, as implemented, was a “vaguely defined and haphazard”9 operation targeting people who opposed American involvement in the Vietnam War and other related policies of the national Government; the second was COINTELPRO-Black Nationalist, which, as implemented, was directed at people seeking improvement of civil rights for Black people. The goal and strategy of these secret programs are not in doubt: a memorandum prepared and circulated by one of the FBI defendants described the COINTELPRONew Left program as follows:

The purpose of this program is to expose, disrupt and otherwise neutralize the activities of this group and persons connected with it. It is hoped that with this new program their violent and illegal activities may be reduced if not curtailed.10

The lack of any FBI definition of “New Left” apparently resulted in the targeting of almost every antiwar group, including those involved in legitimate, non-violent activities. See Senate Report, note 8, supra, at 88.

The purpose of COINTELPRO-Black Nationalist basically was the same as the New Left program. An airtel dated August 25, 1967 set forth the goals as follows:

The purpose of this new counterintelligence endeavor is to expose, disrupt, misdirect, discredit, or otherwise neutralize the activities of black nationalist, hate-type organizations and groupings, their leadership, spokesmen, membership, and supporters, and to counter their propensity for violence and civil disorder.11

The lack of standards restricting the scope of this program, as with the New Left program, apparently led the FBI to investigate and target persons involved in nonviolent political expression, regardless of their involvement in disorders. See Senate Report, note 8, supra, at 177.

Among the primary goals of COINTELPRO was the prevention of coalitions among Black Nationalist groups and between Black Nationalist groups and the predominantly White New Left. Thus, the COINTELPRO-Black Nationalist program was intended to “[p]revent the coalition of militant black nationalist groups,”12 and to “[p]revent militant black nationalist groups and leaders from gaining respectability, by discrediting them to three separate segments of the community.” 13 Targets of these efforts included the Southern Christian Leadership Conference and the Student Nonviolent Coordinating Committee.14 At least two plaintiffs were identified as Black Nationalists by the FBI and listed in the Agitator list of violence-prone people.15 *11At the same time, the FBI sought to promote political differences within New Left organizations and between New Left and Black Nationalist organizations, and to exploit opportunities to foment animosity.16

Secrecy was of utmost importance to the programs, and memoranda establishing them and requesting specific suggestions from field offices for counterintelligence operations cautioned

that the nature of this new endeavor is such that under no circumstances should the existence of the program be made known outside the Bureau and appropriate within-offiee security should be afforded to sensitive operations and techniques considered under the program.17

FBI memoranda make clear that special security precautions were taken to avoid public exposure, criticism and embarrassment. Leaflets and divisive publications prepared by the FBI were intentionally printed on unwater-marked paper and distributed in unmarked envelopes.18 The Senate Report suggests that the FBI’s recognition of the illegality of the COINTELPRO program merely led to a tightening of security, not termination of the program.19

COINTELPRO invited creative participation by agents in the field offices and spawned an extraordinary range of activities. The memorandum initiating COINTELPRO-Black Nationalist advised agents to whom it was addressed, “You are urged to take an enthusiastic and imaginative approach to this new counterintelligence endeavor and the Bureau will be pleased to entertain any suggestions or techniques you may recommend.”20 The result, as disclosed by the trial record, was at least four categories of activity interfering with plaintiffs’ legitimate protest activities: (1) efforts to create racial animosity between Blacks and Whites; (2) interference with lawful demonstration logistics; (3) efforts to create discord within groups or to portray a group’s motives or goals falsely to the public; and (4) direct efforts to intimidate the plaintiffs.21 We set out below examples of each of these four categories, to provide a sense of the activities of which plaintiffs complain:

—The\WFO prepared and distributed to “friendly media” false press releases calculated to tarnish the reputation of Julius *12Hobson22 and create internal dissension within the ranks of BUF.23

—The FBI wrote a racially-inflammatory leaflet entitled “Give Them Bananas!”, which it sent anonymously to BUF members. The leaflet was the result of FBI efforts to engender animosity between BUF and the New Mobilization Committee (“New Mobe”), an antiwar group, after the New Mobe received from the Black community a demand for payment to support a November 1969 antiwar demonstration. The incident, referred to in the record as the “Head Tax” issue, began with a BUF suggestion that demonstrators each make a contribution; it escalated when the New Mobe received a letter from BUF demanding $25,000 to support the demonstration. Rev. Moore, whose typewritten name appeared at the bottom of the BUF letter, denied writing it. Whether the FBI forged the initial demand letter is in considerable dispute. No dispute exists, however, over the FBI’s responsibility for the “Give Them Bananas!” leaflet, which purports to be the New Mobe’s response. It bears a crude drawing of a black monkey with a banana, and reads in part,

We consider you [Rev. Moore, whose name appeared on the initial demand letter] and your kind as black bandits and the most dangerous of the elements eating away at the movement____ Suck on your bananas, brother, and someday you might learn how to make a fire or build a wheel.24

—The FBI filled in with fictitious names and addresses WMC housing forms used by WMC to identify places to lodge visiting demonstrators; the goal was to have demonstrators at the 1969 Presidential inaugural ceremonies make “useless trips to locate nonexistent addresses” and create confusion.25

—The FBI interfered with a “counterinaugural demonstration” by infiltrating the parade marshals’ walkie-talkie radio communications, countermanding orders and sending marchers outside approved demonstration areas.26

—The FBI encouraged sources to “undertake leadership roles in the various factions and stimulate dissension among them.”27

—The FBI published its own “student” newspaper, The Rational Observer, which attacked the American University student newspaper, The Eagle, as too radical. The FBI publication, which claimed to be published by students, urged students to seek an injunction against publication of The Eagle and to question the motives of those who opposed the war, and warned, “Remember, you will be faced with joining society upon completion of your academic training. Don’t do anything in haste today which could cause you embarrassment tomorrow.” 28

—The FBI engaged in an intensive campaign of harassing interviews to intimidate politically active people and heighten their concern about the consequences of political activity.29

The record does not contain documentation of MPD activities during this period as extensive as that regarding FBI activities. The MPD Intelligence Division had informers in virtually every organization with which plaintiffs were involved — the New Mobe, WMC, BUF, ECTC, Institute for Policy Studies, WAP AC and both plaintiff organizations*1330 — in furtherance of its stated mission to gather information on “persons, groups, and organizations whose activities might be detrimental to the proper functioning of local, state, or national governments.” 31 The record contains testimony about efforts of these informants to disrupt the organizations with which they were involved. For example, evidence at trial indicated that an MPD informant acted as an agent provocateur and attempted to disrupt a peaceful demonstration at the United States Capitol,32 that a vociferous demonstrator who was identified later as an MPD officer urged a crowd to disobey parade instructions given by plaintiff Abbott and instead to march to an area where police awaited,33 and that MPD successfully encouraged informants to take private mailing lists and membership lists, and in one instance to break into an office at night to take a metal strong box.34

III. The Causes of Action

At trial, plaintiffs asserted that they were victims of three conspiracies to violate their civil rights, all actionable under 42 U.S.C. § 1985(3) (Supp. V 1981). They alleged that each of these conspiracies violated their First Amendment rights to assemble for peaceable political protest, to associate with others to engage in political expression, and to speak on public issues free of unreasonable Government interference.35 The jury returned verdicts in favor of most plaintiffs on all three conspiracy claims. One civil conspiracy, the jury found, included the five FBI defendants; a second involved certain MPD defendants and the District of Columbia; a third implicated both FBI and District defendants.

Plaintiffs also filed Bivens36 actions for damages directly under the Constitution, claiming that defendants had individually injured various plaintiffs in the exercise of their First Amendment rights.

IY. The Special Verdicts

The jury returned special verdicts and found most defendants to be liable to most plaintiffs on the basis both of individual and conspiratorial acts. All FBI and MPD defendants, as well as the District of Columbia, were found liable to plaintiffs Bloom, Abbott, Pollock, Waskow and Washington Peace Center. The five FBI defendants, the District, and Wilson and Herlihy of MPD were found liable to plaintiffs Hobson, Eaton and Booker. One plaintiff, Washington Area Women Strike for Peace, was found not to have been injured by any of defendants’ activities.

All defendants were found liable for compensatory damages. In addition, all of the FBI defendants, and Wilson and Herlihy of MPD, were found liable for punitive damages.37 Five plaintiffs recovered $93,750 each, and three recovered $81,062.50 each. Individual defendants’ liability ranged from $75,000 to $9,375.38

Discussion

The FBI defendants, District of Columbia defendants and FBI agent Courtland Jones separately appeal the judgments against them. Each of the defendant *14groups raises issues particular to itself, as well as certain contentions that are common to all. We turn now to consider each argument.

I. Liability Under Section 1985(3)

As a threshold matter, all defendants challenge the applicability of section 1985(3) to this case, although on separate grounds. The D.C. defendants argue that the section does not apply to employees of the District of Columbia or to the municipality.39 The FBI defendants contend that the class-based animus necessary to a section 1985(3) action was not established.40 Finally, FBI defendant Jones argues that the section does not apply to federal officials or to acts occurring within the District of Columbia.41

A. The Statutory Scheme

We begin by considering the language of, and case law relevant to,‘ section 1985(3). The provision reads,

If two or more persons in any State or Territory conspire or go in disguise on the highway or on the premises of another, for the purpose of depriving, either directly or indirectly, any person or class of persons of the equal protection of the laws, or of equal privileges and immunities under the laws ... [and] in any case of conspiracy set forth in this section, if one or more persons engaged therein do, or cause to be done, any act in furtherance of the object of such conspiracy, whereby another is injured in his person or property, or deprived of having and exercising any right or privilege of a citizen of the United States, the party so injured or deprived may have an action for the recovery of damages occasioned by such injury or deprivation, against any one or more of the conspirators.42

By its terms, therefore, the statute requires that a plaintiff must allege and prove four elements: (1) a conspiracy; (2) for the purpose of depriving, either directly or indirectly, any person or class of persons of the equal protection of the laws, or of equal privileges and immunities under the laws; and (3) an act in furtherance of the conspiracy; (4) whereby a person is either injured in his person or property or deprived of any right or privilege of a citizen of the United States.

It is by now well-established that the provision encompasses private conspiracies, and not just actions taken under color of state law. See Griffin v. Breckenridge, 403 U.S. 88, 91 S.Ct. 1790, 29 L.Ed.2d 338 (1971). It is equally clear, in light of Griffin, that the provision does not apply to all conspiratorial tortious interferences with the rights of others, but only to those motivated by some class-based, invidiously discriminatory animus. Id. Thus, the Supreme Court has added a requirement of class-based animus to the list of elements set out above.

In Griffin, the seminal case establishing these principles, the complaint alleged that defendants, private persons, had conspired to carry out an assault “to prevent [the] plaintiffs and other Negro-Americans, through ... force, violence and intimidation, from seeking the equal protection of the laws and from enjoying the equal rights, privileges and immunities of citizens.” Id. at 103, 91 S.Ct. at 1799. Holding that the complaint stated a cause of action at the core of section 1985(3), the Court found it unnecessary to delineate the periphery of such actions. Consequently, because the alleged conspiracy in Griffin was motivated by racial basis, the Court did not have to consider whether non-racial discriminatory animus stated a cause of action under the statute.

Similarly, the Court identified two constitutional sources of congressional power to reach the private conspiracy alleged in the case before it. First, the Court observed that section 2 of the Thirteenth Amend*15ment enables Congress to impose liability on private persons for conduct “far beyond the actual imposition of slavery or involuntary servitude,” id. at 105, 91 S.Ct. at 1800, and concluded that “Congress was wholly within its powers under § 2 of the Thirteenth Amendment in creating a statutory cause of action for Negro citizens who have been the victims of conspiratorial, racially discriminatory private action aimed at depriving them of the basic rights that the law secures to all free men.” Id.43 Second, the Court noted that the right of interstate travel “is constitutionally protected, does not necessarily rest on the Fourteenth Amendment, and is assertable against private as well as governmental interference,” id. at 105, 91 S.Ct., at 1800, and that plaintiffs had alleged interference with that right. Congress therefore unquestionably had the particular power to protect the right of interstate travel from private interference. The Court made clear that, in identifying these two sources of power, it did not mean to imply the absence of any others. Id. at 107, 91 S.Ct. at 1801.

Quite recently the Supreme Court revisited section 1985(3) and offered some guidance on the outer limits of the provision. In United Brotherhood of Carpenters and Joiners v. Scott, — U.S. —, 103 S.Ct. 3352, 77 L.Ed.2d 1049 (1983), the Court reviewed an action brought by a construction company and two of its employees, principally against a number of unions and their members. The complaint alleged that defendants had conspired to deprive plaintiffs of their legally protected rights, contrary to 42 U.S.C. § 1985(3) (Supp. V 1981), by planning and executing attacks on persons and property at a construction site known to employ non-union workers. The Court held, first, that an alleged conspiracy to infringe First Amendment rights is not a violation of section 1985(3) “unless it is proved that the state is involved in the conspiracy or that the aim of the conspiracy is to influence the activity of the state.” Id. 103 S.Ct. at 3356-57. On this point, the Court observed that the First and Fourteenth Amendments protect the individual against state action, not against wrongs committed by individuals, and it made clear that a section 1985(3) conspiracy to violate either of these constitutional provisions is not made out without some proof of state involvement. At the same time, the Court reaffirmed that the interpretation of the section is not generally limited by the state action constraints of the Fourteenth Amendment; section 1985(3) in fact applies to wholly private conspiracies as long as they are aimed at interfering with rights constitutionally protected against private as well as official encroachment — such as the Thirteenth Amendment and the right to travel. In other words, the rights protected by section 1985(3) exist independently of the section and only to the extent that the Constitution creates them. Thus, when state action is involved, the whole spectrum of rights against state encroachment that the Constitution sets forth comes into play. When no state action is involved, only those constitutional rights that exist against private actors may be challenged under the section. ,______.

The Court in Scott also considered the kind of class-based animus that is encompassed by the section and rejected the notion that the section forbids conspiracies against workers who refuse to join a union. More generally, the Court held that the provision was not intended to reach conspiracies motivated by bias toward others on account of their economic views, status or activities, id. at 3360, or, put another way, motivated by economic or commercial animus.

At the same time, the Court in Scott reaffirmed, on the basis of the applicable legislative history, that section 1985(3) was intended at a minimum to reach animus *16against Blacks “and those who championed their cause, most notably Republicans, ” id. at 3359 (emphasis added). Whether it also reached “wholly non-racial, but politically motivated conspiracies,” id. at 3359, was dubbed a difficult question and was not reached.44

It is against this background that we consider defendants’ arguments that the District Court erroneously permitted plaintiffs to proceed with their section 1985(3) claims.

B. Applicability of Section 1985(3) to the District of Columbia and Its Employees

The District of Columbia argues that section 1985(3) does not apply to the District of Columbia or its employees. For this proposition it offers no analytical support, but rather rests on a passing and conclusory remark in an opinion from this Circuit to the effect that section 1985 “has never been applicable to District employees.” 45 We disagree. Nothing in the statute or in Supreme Court opinions supports the view put forth by defendants. For the following reasons, we hold that the District and its employees may be sued for damages under section 1985(3).46

The District defendants do not offer any rationale for their proposed limitation on the reach of section 1985(3), and we therefore must undertake an independent inquiry into the possible bases for that argument. The proposed limitation has two possible sources: either it is an interpretation of the geographical limit contained in that section — the section by its terms encompasses conspiracies acted upon “in any State or Territory” — which, arguably, was intended to eliminate from the scope of the statute all conspiracies formed in the District of Columbia; or it might be a confused analogy to section 1983 case law, holding that the District of Columbia is not *17a “State or Territory” within the meaning of 42 U.S.C. § 1983.47

The Supreme Court has twice construed the phrase “any [or every] State or Territory” as used in the civil statutes of the 1860s and early 1870s. In Hurd v. Hodge, 334 U.S. 24, 31, 68 S.Ct. 847, 851, 92 L.Ed. 1187 (1948), the Court ruled that the phrase, as used in 42 U.S.C. § 1982 (1976), includes the District of Columbia. That section, which first appeared as section 1 of the Civil Rights Act of 1866, provides:

All citizens of the United States shall have the same right, in every State and Territory, as is enjoyed by white citizens thereof to inherit, purchase, lease, sell, hold, and convey real and personal property.48

In District of Columbia v. Carter, 409 U.S. 418, 93 S.Ct. 602, 34 L.Ed.2d 613 (1973), the Court reached the opposite conclusion when construing 42 U.S.C. § 1983 (current version at 42 U.S.C. § 1983 (Supp. V 1981)). At the time,49 that section provided in pertinent part, “Every person who, under color of any statute ... of any State or Territory____” The Court concluded that the District of Columbia was not a “State or Territory” within the meaning of that language.

To determine the meaning of the phrase “State or Territory” as it is used in section 1985(3), we track the analysis in District of Columbia v. Carter and test for divergence. Initially, however, it is crucial to note that section 1985(3) is facially more akin to section 1982 — which has been construed to encompass the District — than to section 1983. Sections 1982 and 1985(3) both use the “State or Territory” phrase as a geographical limit (unlike section 1983); consequently, to exclude the District of Columbia from their scope is to create an inexplicable safe haven in the District for persons — both private and official — who would be liable elsewhere. In contrast, the “color of law” requirement in section 1983 (which is not found in either sections 1982 or 1985(3)) implicates federalism issues not relevant to our inquiry; therefore the case law construing section 1983 offers a less compelling analogy. In other words, to draw lines on the basis of source of authority is to recognize a coherent distinction, whereas to draw lines on the basis of place of action is to be impractical, as Hurd v. Hodge made clear.

The issue is somewhat complicated, however, by the fact that sections 1982 and 1983 originated in two different acts, each with a different purpose. Section 1985(3) first appeared in the Ku Klux Klan Act of 1871,50 as did section 1983, whereas section 1982 appeared earlier, in the Civil Rights Act of 1866.51 This difference in origin was central to the Carter Court’s analysis of sections 1982 and 1983. Thus, after explaining that section 1982 was enacted to enforce the Thirteenth Amendment’s nationwide prohibition against slavery and involuntary servitude, the Court explained,

The situation is wholly different, however, with respect to § 1983. Unlike § 1982, which derives from the Civil Rights Act of 1866, § 1983 has its roots in § 1 of the Ku Klux Klan Act of 1871, Act of Apr. 20, 1871, § 1, 17 Stat. 13. This distinction has great significance, for unlike the 1866 Act, which was passed as a means to enforce the Thirteenth Amendment, the primary purpose of the 1871 Act was “to enforce the Provisions of the Fourteenth Amendment.” 17 Stat. 13; ... And it has long been recognized that “[different problems of statutory meaning are presented by two enactments deriving from different constitutional sources.”

409 U.S. at 423, 93 S.Ct. at 605. Since section 1985(3) also derives from the Ku *18Klux Klan Act of 1871, arguably the same limits apply to its interpretation as to that of section 1983, in particular the limits imposed by the Fourteenth Amendment. In that case, it might be contended that the District of Columbia would not be within section 1985(3)’s purview, as it was not within that of section 1983, simply because the District is not a State within the meaning of the Fourteenth Amendment, see Bolling v. Sharpe, 347 U.S. 497, 499, 74 S.Ct. 693, 694, 98 L.Ed. 884 (1954); see also Morgan v. District of Columbia, 550 F.Supp. 465, 470 (D.D.C.1982) (reaching this conclusion), aff'd without opinion, 725 F.2d 125 (D.C.Cir.1983). If we were to read into section 1985(3) the limits imposed by Fourteenth Amendment analysis, as the Carter Court read them into section 1983, the result would be to remove from the scope of section 1985(3) all conspiracies acted upon in the District, by whomever planned, not simply or even necessarily conspiracies in which the District or its employees participate. The problem with this analysis is that there is no more reason to read the word “state” as being limited by the scope of the Fourteenth Amendment than there is to read the nature of the conspiracies described in section -1985(3) as limited by that concept — so that only conspiracies under color of state law would be covered. The two would seem to go pari passu. But the latter interpretation has been conclusively rejected by the Supreme Court. See Griffin v. Breckenridge, 403 U.S. 88, 91 S.Ct. 1790, 29 L.Ed.2d 338 (1971).

By interpreting section 1985(3) to encompass private conspiracies, the Court in Griffin necessarily eschewed an extension of the Fourteenth Amendment’s limitations to section 1985(3). The holding in Griffin is wholly inconsistent with the limited commands of the Fourteenth Amendment, which constrain only the States and those acting under color of State authority. Similarly, in Scott the Court explicitly reaffirmed the proposition that section 1985(3) encompasses wholly private conspiracies and unequivocally stated that “the section is not limited by the constraints of the Fourteenth Amendment.”52 Accordingly, while the Carter Court found persuasive section 1983’s Fourteenth Amendment origins, and relied on those origins to read into section 1983 limitations on its reach, Supreme Court decisions construing section 1985(3) point in precisely the opposite direction. With express Supreme Court direction not to apply Fourteenth Amendment analysis to the construction of section 1985(3), we can perceive no reason to interpret the provision as the Court interpreted section 1983, where a contrary rule controlled.

Nor do we find persuasive for our purposes any other argument put forth in Carter to limit section 1983’s reach in the District. In fact, the remaining factors point toward construing section 1985(3) to cover District employees. The Carter Court observed that unlike section 1983, section 1982 was intended to reach private parties, to act as “an ‘absolute’ bar to all such discrimination, private as well as public, federal as well as state.” 409 U.S. at 422, 93 S.Ct. at 605. “With this in mind,” the Court concluded, “it would be anomalous indeed if Congress chose to carve out the District of Columbia as the sole exception to an act of otherwise universal application,” id., particularly when the “dangers of private discrimination ... were, and are, as present in the District of Columbia as in the States.” Id. Precisely the same analysis applies to section 1985(3), which applies to all conspiracies, whether public or private, and the Carter argument quite clearly counsels against “carving out” the District of Columbia. We have no reason to believe private persons in the District are any less capable than persons elsewhere of participating in conspiracies of the kind addressed in section 1985(3).

On the basis of the Court’s analysis in Carter, we conclude that none of the reasons offered for eliminating the District from the scope of section 1983 applies to section 1985(3) or counsels reaching the *19same result. In light of the Supreme Court’s direction not to read Fourteenth Amendment restraints into section 1985(3), we would be remiss to limit the reach of that section and exempt the District and its employees. In the firm belief that Hurd v. Hodge and the analysis contained therein controls this case, we reject defendants’ proposed limitation on section 1985(3).

C. Applicability of Section 1985(3) to Federal Officers

We turn next to consider defendant Jones’ contention that section 1985(3) does not contemplate actions against federal officers. The apparent source of this argument is an antiquated decision of the Second Circuit, whose holding has repeatedly been read out of context by District Courts, and has now effectively been overruled by the Supreme Court. Accordingly, we reject Mr. Jones’ argument and hold that section 1985(3) encompasses actions against federal officers, subject, of course, to considerations of qualified immunity.

Initially, we note that this Circuit has previously permitted actions to be brought under section 1985(3) against federal officers. See Fitzgerald v. Seamans, 553 F.2d 220 (D.C.Cir.1977) (White House official may be liable in section 1985(3) action). As a result, we pause here only to resolve any lingering doubts about the rationale of the law of this Circuit, not to decide what the law should be. Because the law in this area for years was based on conclusory, unsupported statements, and misguided interpretations of an unfortunately cryptic opinion, we want to make absolutely clear the basis of our decision.

The source of confusion is Gregoire v. Biddle, 177 F.2d 579 (2d Cir.1949), cert. denied, 339 U.S. 949, 70 S.Ct. 803, 94 L.Ed. 1363 (1950), a delphic opinion by Judge Learned Hand, which has repeatedly been cited for the proposition that section 1985(3) does not contemplate suits against federal officers. See, e.g., Lofland v. Meyers, 442 F.Supp. 955, 957 (S.D.N.Y.1977); Williams v. Halperin, 360 F.Supp. 554, 556 (S.D.N.Y.1973). In Gregoire, the Second Circuit affirmed the dismissal of a complaint against Francis Biddle, the Attorney General, and other federal officials, to recover for false arrest as an enemy alien. The complaint was grounded, inter alia, on the Civil Rights Act, including section 47(3), now 1985(3). The court began its analysis by summarizing a decision of that Circuit, affirmed by the Supreme Court, which had held United States Attorneys absolutely immune from a civil action for malicious prosecution. The Supreme Court, in affirming, cited to earlier cases establishing absolute immunity for judges for acts done in the exercise of their judicial function.53 Judge Hand concluded that the Court’s affirmance of the circuit court established absolute immunity for officers of the Department of Justice, when engaged in prosecuting private persons. 177 F.2d at 580. Defendants in Gregoire were officers of the Justice Department engaged in prosecuting a private person. Accordingly, Gregoire held only that the common law grants executive officers engaged in prosecution an absolute privilege from civil actions, not that all federal officers are thus sheltered. It concluded that, absent any indication that Congress intended to remove that immunity, the complaint required dismissal.

It was in this context that Gregoire turned to consider the Civil Rights Acts, presumably to determine whether they were meant to waive prosecutorial immunity. Judge Hand rejected plaintiff’s argument that then-section 47(3), currently section 1985(3), created a claim against any two people, including federal officers, who conspire to injure another for spite or improper motives, or, in other words, rejected a reading of the section that would transform it into a general federal tort law. He remarked, “it is apparent that [the words] could not have had such a scope without *20destroying their validity constitutionally.” Id. at 581. Regardless whether Judge Hand was more concerned with the idea of private conspiracies under the statute or federalization of tort law, it is now clear that neither of these factors necessarily poses any obstacle to the section’s constitutionality. In light of Griffin, which both established the constitutionality of the very statutory construction rejected by Judge Hand — that the conspiracy statute applies to wholly private conspiracies — and read into the statute a requirement of class-based discriminatory animus to assure the statute was not a general tort provision, Judge Hand’s construction of the statute on these issues no longer merits consideration.

It is therefore apparent that cases relying on Gregoire to preclude suits against federal officials under section 1985(3) have no present force. First, these cases always have been in error to the extent they read into Gregoire an absolute privilege for persons other than prosecutors. Second, these cases have in any event been superseded in their analysis of section 1985(3), along with Gregoire itself, by Griffin. We have not found either in case law or in the language of the statute any reason to exclude all federal officers from the meaning of the word “persons” in section 1985(3). Pre-Griffin cases therefore are of no preceden-tial weight on this point, and post-Griffin cases that rely only on Gregoire, or on other cases citing only Gregoire, without any mention of Griffin, similarly should not be followed.

Considerable recent case law rejects the proposed limitation on section 1985(3) and supports our conclusion. See Jafree v. Barber, 689 F.2d 640, 643 (7th Cir.1982) (section 1985(3) action against federal officer states cause of action); Gillespie v. Civiletti, 629 F.2d 637, 641 (9th Cir.1980) (same); Dry Creek Lodge, Inc. v. United States, 515 F.2d 926, 931 (10th Cir.1975) (same); Bergman v. United States, 551 F.Supp. 407, 414-15 (W.D.Mich.1982) (same); Peck v. United States, 470 F.Supp. 1003, 1008-12 (S.D.N.Y.1979) (construing section 1985(3) to permit suits against federal officers, rejecting contrary decisions in the district); Founding Church of Scientology v. Director, Federal Bureau of Investigation, 459 F.Supp. 748 (D.D.C.1978) (allowing section 1985(3) action against federal officer). As one Southern District of New York judge explained, in rejecting Gregoire and the cases that followed,

Unless there is a rationale, unknown to the past cases, for holding that federal officers are not “persons” under § 1985(3), there is no longer any reason to exclude from coverage federal officers acting under color of federal law. Since such a rationale is inconceivable, Griffin 's holding that § 1985(3) applies to any person requires that it apply to federal officers.

Moriani v. Hunter, 462 F.Supp. 353, 356 (S.D.N.Y.1978). We agree and hold that plaintiffs did not fail to state a claim upon which relief could be granted by naming either federal officers as defendants or, for reasons stated in section 1(B), for naming as defendants the District of Columbia and its employees.

D. Class-Based Discriminatory Animus

FBI defendants also contend that plaintiffs failed to establish the existence of the “racial, or perhaps otherwise class-based, invidiously discriminatory animus” 54 requisite to recovery under section 1985(3). Their challenge is not directed at the trial judge’s instructions on the matter, which were not challenged below,55 but at *21the sufficiency of the evidence on this element of plaintiffs’ case. We hold that there was sufficient evidence to permit a jury to find that the FBI defendants acted with the requisite animus, and that the claim must therefore fail.

We note at the outset the existence of some debate over the kinds of class-based discriminatory animus that section 1985(3) requires. Scott sought to break down the potential scope of the provision into three categories of animus — economic and commercial, purely political (meaning wholly non-racial), and racial (including the traditional supporters of civil rights) — and held that the first category was not within the reach of the section. Griffin, on the other hand, had squarely placed the third category (i.e., “racial”) within the section’s reach. Both cases left unanswered whether “political activity” simpliciter defines a class of people covered by section 1985(3). At the same time, several circuits have ruled that politics and religion define such a class. See, e.g., Keating v. Carey, 706 F.2d 377, 386-88 (2d Cir.1983) (discrimination on basis of political affiliation constitutes class-based discriminatory animus); Ward v. Connor, 657 F.2d 45, 47-48 (4th Cir.1981) (discrimination against members of Unification Church), cert. denied, 455 U.S. 907, 102 S.Ct. 1253, 71 L.Ed.2d 445 (1982); Hampton v. Hanrahan, 600 F.2d 600, 623 & n. 22 (7th Cir.1979) (discrimination based on political affiliation with racial overtones), modified on other grounds, 446 U.S. 754, 100 S.Ct. 1987, 64 L.Ed.2d 670 (1980); Means v. Wilson, 522 F.2d 833, 839-40 (8th Cir.1975) (discrimination against supporters of insurgent candidate for tribal council presidency), cert. denied, 424 U.S. 958, 96 S.Ct. 1436, 47 L.Ed.2d 364 (1976); Glasson v. City of Louisville, 518 F.2d 899, 911-12 (6th Cir.) (discrimination against critics of the President), cert. denied, 423 U.S. 930, 96 S.Ct. 280, 46 L.Ed.2d 258 (1975); Marlowe v. Fisher Body, 489 F.2d 1057, 1064-65 (6th Cir.1973) (discrimination against Jews); Action v. Gannon, 450 F.2d 1227, 1232 (8th Cir.1971) (en banc) (worshippers at predominantly White parish disrupted by Black civil rights workers); see also Comment, Private Conspiracies to Violate Civil Rights, 90 Harv.L.Rev. 1721, 1728 (1977) (“[T]he legislative history behind section 1985(3) points unmistakably to the conclusion that discrimination [on the basis of political affiliations or beliefs] was intended to be actionable.”); cf. Wilhelm v. Continental Title Co., 720 F.2d 1173, 1176 (10th Cir.1983) (handicapped persons not a class within meaning of section 1985(3)), cert. denied, — U.S.-, 104 S.Ct. 1601, 80 L.Ed.2d 131 (1984); DeSantis v. Pacific Telephone and Telegraph Co., 608 F.2d 327, 333 (9th Cir.1979) (homosexuals not a protected class).

Given the facts of this case, it is unnecessary to decide whether purely political or other activity without any racial overtones falls within section 1985(3). The FBI conspiracy allegedly targeted plaintiffs in significant part because of their involvement in and support of civil rights. At a minimum, according to Griffin, section 1985(3) reaches conspiracies motivated by animus against Blacks and those who support them. As the Court in Scott explained,

The central theme of the bill’s proponents was that the [Ku Klux] Klan and others were forcibly resisting efforts to emancipate Negroes and give them equal access to political power. The predominate purpose of § 1985(3) was to combat the prevalent animus against Negroes and their supporters. The latter included Republicans generally, as well as others, such as Northerners who came South with sympathetic views towards the Negro.56

Thus, in Griffin, the Court found an alleged conspiracy against Blacks (and against a non-party who was mistaken to *22be a civil rights worker), designed to prevent them from seeking equal rights under the laws, to satisfy the requisite animus requirement. Similarly, we find that the uncontroverted evidence plaintiffs offered to prove the requisite animus discloses that defendants’ actions were undoubtedly aimed at plaintiffs at least in part on account of their racial politics—that is, their participation in and support of civil rights protest, so as to take the conspiracy out of the purely political, wholly non-racial category. The evidence strongly supports the conclusion that the COINTELPRO conspiracy was “impelled by a commingling of racial and political motives,”57 and is far removed from the kind of purely political disputes that seemingly have troubled the Supreme Court in its reading of section 1985(3).58 The blatant racial overtone of the FBI program, coupled with plaintiffs’ various organizational efforts, make clear the entanglement of race and politics that characterized the implementation of the COINTELPRO conspiracy.

First, the plaintiffs clearly constitute a class of Black and White civil rights workers and White supporters of their efforts and, thus, fall at the core of section 1985(3)’s concern. Plaintiffs Abbott and Booker were prominent in the Emergency Committee on the Transportation Crisis, a coalition of Black and White neighborhood groups opposed to construction of superhighways through the District’s residential areas, because of the consequent displacement of many poor Blacks; plaintiffs Booker and Eaton were involved in organization of the Black United Front, a civil rights group. Plaintiff Hobson, an active member of the peace movement, also worked for and supported the civil rights efforts of her husband, Julius Hobson. Many of these plaintiffs also participated in protests against the Vietnam War, in part “to heighten the awareness of the public on particularly the racial miseries of the war ... [meaning] that minority youths were disproportionately being drafted and fighting for the war.” 59 At the same time, the plaintiffs who were primarily involved with antiwar groups, such as the Washington Mobilization Committee and the New Mobilization Committee, worked with Black civil rights groups on peace marches60 and otherwise, both to support civil rights generally and to increase Black awareness of the high casualty figures among Black soldiers *23and the perceived need, as a result, to participate in antiwar protest. Given the undisputed evidence that the antiwar plaintiffs and civil rights plaintiffs joined together to seek common goals of concern to each, we find they represent, at a minimum, a class of civil rights workers and their supporters at the core of section 1985(3)’s concerns.

*22[W]e find difficult the question whether § 1985(3) provided a remedy for every concerted effort by one political group to nullify the influence of or do other injury to a competing group by use of otherwise unlawful means. To accede to that view would go far toward making the federal courts, by virtue of § 1985(3), the monitors of campaign tactics in both state and federal elections, a role that the courts should not be quick to assume.

*23It is equally clear that the FBI defendants’ conspiracy was directed against plaintiffs because of their participation in these very activities. Considerable testimony described efforts by the FBI to drive a wedge into this alliance between civil rights groups and peace groups.

The FBI’s COINTELPRO program sought not only to “neutralize” and “disrupt” Black groups and antiwar groups separately, but also to exploit any dissention between them in an effort to deter formation of an alliance. One of the most shocking examples of this effort is found in an FBI document, written in August 1969, relating to the BUF head tax demand on the New Mobilization Committee. It discloses that an aim of COINTELPRO-New Left, to which it is referenced, was to split the antiwar and civil rights groups:

The Washington Field Office has recommended and the Bureau concurs in that recommendation that this is an ideal situation to exploit through the Counterintelligence Program.

Recipient offices are to furnish recommendations for such action to the Bureau by 8/29/69 without fail. Consideration is to be given to utilizing informants in both racial and nonracial protest groups in this matter. It is noted that the nonracial protest groups, particularly the NMC, can be accused of racism in refusing to go along with this demand. At the same time, such groups could be split further by some individuals calling the Front’s demands extortion while other individuals in the group support the demands.

The NMC has called for a massive demonstration at Washington, D.C., on 11/15/69. This demonstration may result in confrontation with authorities. The BUF is demanding that it be paid to let this demonstration go on in Washington, D.C. Committee members are upset over these demands and this situation offers an opportunity to divide and to embarrass them. This airtel requests recommendations from the field.61

An October 1969 memorandum from the Special Agent in Charge (SAC) in New York to the FBI Director, also referenced to COINTELPRO-New Left, followed:

Enclosed are two copies of a suggested leaflet entitled “Give Them Bananas!”, designed to widen the rift between the New Mobilization Committee To End The War in Vietnam (NMC) and the Black United Front (BUF) over the BUF demand that the NMC “pay” for their planned March on Washington 11/15/69.

Enclosed leaflet has been written in the jargon of the New Left, necessitating the use of a certain amount of profanity.

It is realized this material is racial in tone but it is believed this is the one point of vulnerability in the NMC-BUF combination.

Bureau authority is requested to prepare and mail anonymously the enclosed leaflet to selected individuals active in the NMC and BUF both in the New York and Washington areas.62

Subsequent documents reveal that Bureau authority was granted, with the caution, “Take all necessary steps to protect the identity of the Bureau as the source of these leaflets,” 63 and that copies of the leaflet were forwarded to WFO for distribution.*2464 Rev. Moore, purported author of the BUF head tax demand, testified that he in fact received the FBI’s leaflet.65 As distributed, it was titled, “Give Them Bananas!”, was “decorated” with a drawing of a monkey, and contained a string of profane, racist and degrading remarks that our senses of propriety and revulsion prevent us from reprinting in full.66

We believe the foregoing amply discloses that the FBI actions were sufficiently related to matters of race to place the FBI conspiracy solidly within even the narrowest reading of section 1985(3).67 Section 1985(3) was “intended, perhaps more than anything else, to provide redress for victims of conspiracies impelled by a commingling of racial and political motives.”68 We would be hard pressed to imagine a case with evidence stronger than we find here of precisely such commingled motives. Accordingly, we reject the FBI defendants’ contention that plaintiffs’ evidence on this element of their claim did not support the verdict.

II. Harlow v. Fitzgerald and Defendants’ Qualified Immunity

After trial and entry of judgment in this case, the Supreme Court issued its opinion in Harlow v. Fitzgerald, 457 U.S. 800, 102 S.Ct. 2727, 73 L.Ed.2d 396 (1982), which altered the law of qualified immunity.69 All defendants claim that the rule announced in Harlow applies to this case and requires judgment in their favor. While we agree that the principle announced in Harlow controls this case, we find that it does not require entry of judgment in favor of defendants.

A. Qualified Immunity and the Harlow Standard

Prior to Harlow, qualified immunity had both subjective and objective elements. The objective element involved a presumptive knowledge of and respect for “basic, unquestioned constitutional rights.” Wood v. Strickland, 420 U.S. 308, 322, 95 S.Ct. 992, 1001, 43 L.Ed.2d 214 (1975). The subjective element addressed the “permissible intentions,” id., or good faith of the Government actor. In other words, an official would not receive qualified immunity if he

knew or reasonably should have known that the action he took within his sphere of official responsibility would violate the constitutional rights of the [plaintiff] or if [the official] took the action with the malicious intention to cause a deprivation of constitutional rights or other injury.

Wood, 420 U.S. at 322, 95 S.Ct. at 1001.

The Supreme Court’s decision to grant state70 and federal71 officers qualified, instead of absolute immunity “reflected an attempt to balance competing values: not only the importance of a damages remedy to protect the rights of citizens ... but also ‘the need to protect officials who are required to exercise their discretion and the related public interest in encouraging the *25vigorous exercise of official authority. Harlow, 457 U.S. at 807, 102 S.Ct. at 2733 (quoting Butz v. Economou, 438 U.S. 478, 506, 98 S.Ct. 2894, 2910, 57 L.Ed.2d 895 (1978)). By striking the balance in this manner, the Court determined that the need for damage remedies required rejection of an absolute immunity for federal officers. Yet, the fears remained that Government officials might be chary in the exercise of their most important public functions if constantly subject to discovery and trial over their discretionary decisions. To shield Government officers from undue interference, the Court admonished lower courts applying the qualified immunity standard to engage in “firm application of the Federal Rules of Civil Procedure [to] ensure that federal officials are not harassed by frivolous lawsuits,”72 and endorsed the idea of resolving the immunity issue of damage suits at the summary judgment stage. 73

Experience with the qualified immunity defense proved it to be often ineffective in resolving suits on summary judgment. If any factual dispute existed as to whether the official acted with malicious intent, or with a belief that a clear standard prohibited his conduct, that dispute required a subjective determination necessitating a trial. “The need for such determinations thus frustrated the goal of terminating insubstantial lawsuits on summary judgment.” National Black Police Association, Inc. v. Velde, 712 F.2d 569, 574 (D.C.Cir.1983), cert. denied, — U.S.—, 104 S.Ct. 2180, 80 L.Ed.2d 562 (1984).

Harlow redefined the qualified immunity defense to eliminate the subjec-. tive element. Under the new standard, “government officials performing discretionary functions generally are shielded from liability for civil damages insofar as their conduct does not violate clearly established statutory or constitutional rights of which a reasonable person would have known.” Harlow, 457 U.S. at 818, 102 S.Ct. at 2738. The Court further explained,

If the law was clearly established, the immunity defense ordinarily should fail, since a reasonably competent public official should know the law governing his conduct. Nevertheless, if the official pleading the defense claims extraordinary circumstances and can prove that he neither knew nor should have known of the relevant legal standard, the defense should be sustained. But again, the defense would turn primarily on objective factors.

Id. at 818-19, 102 S.Ct. at 2739. In other words, if the trial judge determines that the law was not clearly established at the time the conduct occurred, the inquiry ceases and the official is entitled to summary judgment. If the law was clearly established, the Government actor is presumed to have known about it; unless he can bring forward undisputed facts establishing that because of extraordinary circumstances he neither knew nor should have known of the unlawfulness, summary judgment in his favor must be denied.

B. Application of the Harlow Standard

In applying the foregoing principles to this case, we conclude that the acts defendants were alleged to have committed violated fundamental and well-established constitutional rights, and that defendants are entitled neither to immunity nor, consequently, to reversal of the verdicts against them on this ground.

Our task, in applying Harlow, is to measure the constitutionality of the acts alleged in this action by reference to clearly established rights at the time the acts occurred. At first blush, the task as so stated appears simple and straightforward. A moment of reflection, however, reveals various uncertainties in the Harlow decision and exposes a test of some imprecision. It is not clear, for example, how a court should determine well-established rights: should our reference point be the opinions of the Supreme Court, the Courts of Ap*26peals, District Courts, the state courts, or all of the foregoing?74 Furthermore, it is unclear whether we must assure ourselves only of the existence of a constitutional right, such as the constitutional right of freedom of speech, or religion, at the relevant time. Alternatively, it might be claimed that a court must inquire more deeply into the constitutionality of the particular conduct of the Government actor, such as the legality of domestic national security warrantless wiretaps,75 at a given time. Put another way, how broadly or narrowly defined is the right that must be well-established?

At the extremes, the answers are clear. Supreme Court precedent “establishes” the law; to the extent that the Court’s opinions give guidance we obviously do not doubt that the law is well-established. It is equally clear that the right at issue can be defined neither so broadly as to parrot the language in the Bill of Rights, nor so narrowly as to require that there be no distinguishing facts between the instant case and existing precedent. The former reading of Harlow would, of course, undermine the premise of qualified immunity that the Government actors reasonably should know that their conduct is problematic. The latter reading, on the other hand, would unquestionably turn qualified into absolute immunity by requiring immunity in any new fact situation. In future cases, courts will of course work through the area between these extremes and answer these and other questions. On the facts of this case, however, these broad parameters suffice, because the illegality of the conduct alleged was well-established “by any reasonable definition of the phrase.”76

In an effort to overcome the obvious, defendants focus on the trial evidence and argue that each individual act that they were shown to have committed was lawful, and that they consequently are immune. This argument seriously misconstrues the nature of the qualified immunity defense, and in particular the separate questions of fact and law. We pause here to address the proper course of pleadings and proof on this issue, to make clear why we find defendants’ argument irrelevant to this element of our review.

The difficulty, we presume, stems from the bipartite nature of Harlow’s inquiry. First, a district court must determine whether the right alleged to have been violated was well-established; it then must determine whether the defendant reasonably should have known of its existence. While Harlow asserted that a “reasonably competent public official should know the, law governing his conduct,” 457 U.S. at 819, 102 S.Ct. at 2739, in some extraordinary instances “what a reasonable person in like circumstances should have known”77 has been held to create a question of fact. Step-by-step, the inquiry proceeds as follows: assume plaintiff alleges that defendant violated Right X, and defendant moves for summary judgment on grounds of qualified immunity, claiming Right X was not well-established when the alleged acts were committed and, alternatively, that extraordinary circumstances prevented him both from knowing and having reason to know the relevant legal standard. If the district court judge determines as a matter of law that Right X was well-established, a question of fact might still remain as to whether defendant reasonably neither knew nor should have known of it. The first issue is purely legal and should be susceptible to initial determination on the complaint and summary judgment papers. The latter is potentially an issue for trial, depending *27upon the unique circumstances of each case.78

Applying these principles, we must consider in this action only whether the right that plaintiffs alleged to have been violated was well-established at the time the alleged acts occurred. We consider irrelevant to this inquiry defendants’ assertions that the evidence does not support those allegations; such evidence is properly considered as an element of our inquiry into the sufficiency of the evidence, not qualified immunity. Because no defendant argues that exceptional circumstances exist to demonstrate that he reasonably neither knew nor should have known of the relevant legal standard, our inquiry is finished following our determination whether the applicable law was well-established.79

Plaintiffs alleged in their complaint80 that from 1967 or 1968 to 1974 the District and FBI defendants engaged in a variety of legal and illegal activities in a specific effort “to disrupt and interfere with the plaintiffs’ political activities, [,including] urging violent or unlawful actions, and supplying the public and/or news media with false information about the plaintiffs and their plans.”81 Plaintiffs also alleged that “[s]ome or all of the activities of defendants” 82 were conducted pursuant to a plan, known as COINTELPRO-New Left, which, they contended, was “designed to conduct surveillance upon and to cause disruption of the activities of what the defendants regarded as the ‘New Left.’ ” 83 Plaintiffs claimed that because these actions were undertaken expressly to disrupt their lawful activities, they violated, inter alia, plaintiffs’ First and Fifth Amendment rights.84

The extraordinary nature of these charges makes this an easy case. Whatever authority the Government may have to interfere with a group engaged in unlawful activity, and however it may be permitted to impede or deter rights of lawful association as a by-product of legitimate Government actions, it is never permissible to impede or deter lawful civil rights/political organization, expression or protest with no other direct purpose and no other immediate objective than to counter the influence of the target associations.

As of 1967, the existence of a First Amendment right of association for lawful purposes was beyond dispute and its broad contours were quite clear.85 A line of Su*28preme Court cases had expressly established first, that Government cannot constitutionally punish membership in or association with any organization, absent clear proof that a person specifically intends to accomplish the illegal aims of the organization, see Scales v. United States, 367 U.S. 203, 228-30, 81 S.Ct. 1469, 1485-86, 6 L.Ed.2d 782 (1961); Noto v. United States, 367 U.S. 290, 299-30, 81 S.Ct. 1517, 1521-22, 6 L.Ed.2d 836 (1961); second, that Government cannot constitutionally punish individual or group advocacy of any position, unless it amounts to incitement to lawless action, see Bond v. Floyd, 385 U.S. 116, 87 S.Ct. 339, 17 L.Ed.2d 235 (1966); Yates v. United States, 354 U.S. 298, 77 S.Ct. 1064, 1 L.Ed.2d 1356 (1957); and third, that lawful associations and their members have the right to be protected from facially legitimate Government actions that would deter membership or otherwise thwart their efforts to associate and petition the Government for redress of their grievances. The first two principles at a minimum rendered absolutely unconstitutional any direct Government interference with persons because they participated in organizations, if those organizations did not advocate violence or other lawless action, or because they held certain views, if those views were not accompanied by incitement to illegal action or a specific intent to accomplish illegal ends by force and violence. The third principle was not absolute, but made unconstitutional Government action taken for legitimate purposes if it significantly interfered with protected rights of association, unless the Government could demonstrate a substantial, NAACP v. Button, 371 U.S. 415, 444, 83 S.Ct. 328, 343, 9 L.Ed.2d 405 (1963), or compelling, Bates v. Little Rock, 361 U.S. 516, 524, 80 S.Ct. 412, 417, 4 L.Ed.2d 480 (1960), interest to justify the infringement, and that the interest could not more narrowly be accommodated. See Shelton v. *29Tucker, 364 U.S. 479, 488, 81 S.Ct. 247, 252, 5 L.Ed.2d 231 (1960) (“In a series of decisions this Court has held that, even though the governmental purpose be legitimate and substantial, that purpose cannot be pursued by means that broadly stifle fundamental personal liberties when the end can be more narrowly achieved.”).

These principles leave no doubt that Government action, taken with the intent to disrupt or destroy lawful organizations, or to deter membership in those groups, is absolutely unconstitutional. The Government could not constitutionally make such participation unlawful; consequently, it may not surreptitiously undertake to do what it cannot do publicly. Nor can we fathom any conceivably legitimate governmental interest in such an undertaking. If the targets act unlawfully, criminal laws are available. Nor do we accept any argument that proper law enforcement requires systematic disruption and neutralization of lawful organizations.

It is therefore absolutely clear that the actions defendants were alleged to have taken violated well-established rights. The defendants allegedly directed and participated in a program designed to obstruct plaintiffs’ efforts to work peaceably for political change and civil rights, and to splinter and neutralize plaintiffs’ organizations. The constitutional right of association of the kind in which plaintiffs were engaged was well known, as was the degree of protection from direct interference that such lawful association was to be accorded. Additionally, while the associational right was not absolute, permissible limits of Government intrusion in various contexts were sufficiently defined. In a case such as this one, in which the pleadings and proof disclose a program that at its tamest violated the narrowly defined associational rights expressly discussed in these cases, and which in fact extended beyond violations previously contemplated, the law was undoubtedly “well-established.” In such circumstances, to require a prior Supreme Court holding on the particular facts of this case would not only immunize but actually reward the Government for inventing and pursuing ever more egregious conduct. Indeed, there never could be such a ruling from the Court, because Harlow would always immunize the Government actors. Simply put, where it is apparent that less direct, and facially legitimate intrusions on plaintiffs’ rights violate the Constitution, it is beyond question that sweeping, intentional intrusions do so as.well.

C. Pleading Unconstitutional Motive

Harlow focused on the need to enable courts to dismiss “insubstantial” lawsuits before discovery and trial, and it adjusted the qualified immunity defense to facilitate that goal. The kind of case we confront today, involving allegations of unconstitutional motive, offers to litigants a possible means to circumvent the new rule, simply by pleading that any act was performed with an intent to violate clearly established constitutional rights and thereby surmounting the threshold test set out in Harlow. We recognize that in some instances, plaintiffs might allege facts demonstrating that defendants have acted lawfully, append a claim that they did so with an unconstitutional motive, and as a consequence usher defendants into discovery, and perhaps trial, with no hope of success on the merits. The result would be precisely the burden Harlow sought to prevent. Accordingly, in cases involving a claim that defendants acted with an unconstitutional motive, we will require that nonconclusory allegations of evidence of such intent must be present in a complaint for litigants to proceed to discovery on the claim. The allegations on this issue need not be extensive, but they will have to be sufficiently precise to put defendants on notice of the nature of the claim and enable them to prepare a response and, where appropriate, a summary judgment motion on qualified immunity grounds.

As a general proposition, the problem we address is not a new one. In Butz v. Economou the Supreme Court apparently recognized that general conclusory allegations that Government actors had breached *30constitutional rights could unduly enmesh them in burdensome litigation, and it called for a firm application of the Federal Rules of Civil Procedure and for dismissal of insubstantial claims on summary judgment.86 Similarly, every other circuit has articulated a requirement of particularity in pleading for civil rights complaints. In one formulation, the Second Circuit adheres to the following rule:

[Cjomplaints containing only “concluso-ry,” “vague,” or “general allegations” of a conspiracy to deprive a person of constitutional rights will be dismissed____ Diffuse and expansive allegations are in-sufficient____ In this case, appellants’ unsupported allegations, which fail to specify in detail the factual basis necessary to enable appellees intelligently to prepare their defense, will not suffice to sustain a claim of governmental conspiracy to deprive appellants of their constitutional rights.

Ostrer v. Aronwald, 567 F.2d 551, 553 (2d Cir.1977) (citations omitted); see also Contemporary Mission, Inc. v. United States Postal Service, 648 F.2d 97, 106-07 (2d Cir.1981) (“Where a plaintiff fails to produce any specific facts whatsoever to support a conspiracy allegation, a district court may, in its discretion, refuse to permit discovery and grant summary judgment.... Courts must be particularly cautious to protect public officials from protracted litigation involving specious claims.”). Moreover, every other circuit follows suit in requiring that civil rights complaints be pleaded with at least a minimum of specificity.87 Nor is the principle unknown in this Circuit. See Lombard v. United States, 690 F.2d 215, 227 (D.C.Cir.1982) (applying the Second Circuit’s Ostrer rule to claims under the Federal Tort Claims Act), cert. denied, — U.S.-, 103 S.Ct. 3086, 77 L.Ed.2d 1347 (1983).

The test articulated by the Second Circuit is more than adequate to address the Harlow concerns to which we have referred. We simply remind our trial courts that some factual allegations must support claims of unconstitutional motive. Plaintiffs who fail to allege any specific facts to support a claim of unconstitutional motive cannot expect to involve Government actors in protracted discovery and trial. On receipt of such a complaint, Government defendants might move for dismissal or, alternatively, for summary judgment. Then plaintiffs must produce some factual support for their claim to avert dismissal.

In so holding we do not forget that in some circumstances plaintiffs are able to *31paint only with a very broad and speculative brush at the pre-discovery stage, and that overly rigid application of the rule we articulate could lead to dismissal of meritorious claims.88 Thus, while we hasten to add that district court judges must act cautiously in this regard, and freely give leave to amend an inadequate complaint, we- conclude that Harlow requires that mérely conclusory allegations of unconstitutional motive, devoid of factual support, must be found lacking and be dismissed.

Applying these principles, we find plaintiffs’ complaint89 was drafted with sufficient specificity to withstand dismissal in this case. Plaintiffs not only alleged specific kinds' of disruptive activity that could have had no other purpose than to disrupt, but also alleged facts describing the COINTELPRO program designed to conduct surveillance and cause disruption of plaintiffs’ activities. These allegations clearly were neither general, conclusory, nor devoid of factual support and would have sufficed to survive a motion for dismissal under the standards we articulate today.

D. Municipal Liability

The District of Columbia argues that it is entitled to a new trial because the jury was instructed incorrectly on municipal liability. It asserts that the District Court’s proposed instruction was correct, but that the instruction actually given was erroneous.90 Whatever objection might have been raised, we find that the District did not raise this argument after the instruction was given and, consequently, it has waived any right to pursue this matter on appeal.

Rule 51 of the Federal Rules of Civil Procedure states with unmistakable clarity that “[n]o party may assign as error the giving or the failure to give an instruction .unless he objects thereto before the jury retires to consider its verdict, stating distinctly the matter to which he objects and the grounds of his objection.” Case law in this Circuit91 and elsewhere92 leaves no doubt that parties who have an opportunity to object to instructions before the jury retires and do not avail themselves of it cannot claim on appeal that the instructions were erroneous. Yet, that is precisely what the District of Columbia seeks to do.

Indeed, this case attests to the wisdom of the rule. The purpose of Rule 51 is to give the trial judge an opportunity to correct any inadvertent omission, ambiguity or error, reconsider any ruling, or make any other necessary changes or clarifications before the jury retires, and in that way avoid the delay and expense of a retrial.93 The rule obliges counsel to follow the charge to the jury, and to point out alleged errors in the instructions as given. Thus, “[t]he necessity of a retrial is avoided when, by design or through sheer neglect, the losing party fails to make objection at the proper time.”94

Here, the District argues that Judge Ob-erdorfer mistakenly deviated from his own proposed instruction, making this precisely the kind of case to which Rule 51 is addressed. Had the District simply read *32along as the instructions were given and called the alleged deviation to the Judge’s attention, we presume that the judge would have corrected any error. It appears that plaintiffs’ counsel followed that course with respect to other matters and received at least one clarifying instruction on another separate issue.95 Meanwhile, when given the opportunity, the District counsel raised no independent objections and requested no clarifications. Under such circumstances, we decline at this late date to entertain the District’s claim.96

III. Statute of Limitations

All defendants argue that the District Court erred in denying their motions for directed verdicts on statute of limitations grounds. Having reviewed the applicable legal standard and the relevant evidence, we conclude that the District Court should have granted the motions of the FBI defendants as to plaintiffs. Abbott, Bloom and Booker. The remaining motions — as to all plaintiffs by the District defendants, and as to the other plaintiffs by the FBI defendants — were properly denied.97

When no federal statute of limitations governs the period of repose for actions brought under the Civil Rights Act, or under the rationale of Bivens v. Six Unknown Named Agents, 403 U.S. 388, 91 S.Ct. 1999, 29 L.Ed.2d 619 (1971), a federal court must look to the limitations period applicable to the most nearly analogous state cause of action.98 On that basis, it is clear that the three-year limitations periods provided in D.C.Code Ann. § 12-301(8) (1981) controls this case.99

Plaintiffs’ complaint, filed July 16, 1976, seeks damages for actions occurring from 1967 or 1968 to 1974. In other words, this suit was initiated after the applicable three-year limitations period. Plaintiffs argue, however, that in this case the défendants’ fraudulent concealment tolled the statute of limitations. It is to this aged doctrine of *33fraudulent concealment that we now turn our attention.

A. Fraudulent Concealment: Case Law

Initially, all defendants agree that federal law provides the applicable tolling doctrine. This assumption is consistent with the law in this Circuit100 and, since we reach the same result under local law as well,101 we take it to be true for purposes of this discussion.

The keystone of federal fraudulent concealment doctrine is Bailey v. Glover, 88 U.S. (21 Wall.) 342, 22 L.Ed. 636 (1874), in which the Court wrote that where a party injured by another’s fraudulent conduct

remains in ignorance ... without any fault or want of diligence or care on his part, the bar of the statute does not begin to run until the fraud is discovered, though there be no special circumstances or efforts on the part of the party committing the fraud to conceal it from the knowledge of the other party.

Id. 88 U.S. (21 Wall.) at 348. The Court also held that, absent laches or negligence on plaintiff’s part, the limitations period does not begin to run until plaintiff discovers his cause of action when “the fraud has been concealed or is of such character as to conceal itself.” Id. 88 U.S. (21 Wall) at 349-50 (emphasis added). Shortly thereafter, in Wood v. Carpenter, 101 U.S. 135, 25 L.Ed. 807 (1879), the Court gave some indication of the meaning of the phrase “of such character as to conceal itself,” when it wrote, “Concealment by mere silence is not enough. There must be some trick or contrivance intended to exclude suspicion and prevent inquiry.” Id. 101 U.S. at 143. Read together, Wood and Bailey establish first, that equitable tolling generally has two elements, (successful) concealment by defendant and diligence by plaintiff, and second, that a defendant who contrives to commit a wrong in such a manner as to conceal the very existence of a cause of action, and who misleads plaintiff in the course of committing the wrong, may be found to have concealed the wrong. This second principle distinguishes between acts that are self-concealing (such as frauds) and acts where, absent a subsequent act of concealment, only the perpetrator, but not the fact that a cause of action might exist, would be unknown (such as a burglary). In the former case, concealment is established by the nature of the act; in the latter case, additional acts of concealment are required to trigger the tolling doctrine.102

*34Although the foregoing principles generally have been accepted by the courts, the case law reflects a variety of formulations to apply the concealment doctrine.103 In each instance, however, before a defendant’s “exposure to liability is given a potentially infinite duration, there [is] some minimum of culpability — if not affirmative concealment, then at least the construction of a scheme which is by its nature unknowable.” 104

Applying these principles to the facts of this ease, we are able to put to one side actions involving wrongs that are by their nature “knowable” and taper our analysis to focus on the “self-concealing” wrongs that warrant different treatment. Bearing in mind Wood’s requirement of “some trick or contrivance intended to exclude suspicion,” 101 U.S. at 143, we conclude that defendants must engage in some misleading, deceptive or otherwise contrived action or scheme, in the course of committing the wrong, that is designed to mask the existence of a cause of action.105 The deception may be as simple as a single *35lie or as complex as that which we confront here, so long as the defendants conceal “not only their involvement, but the very conduct itself.” Richards v. Mileski, 662 F.2d 65, 70 (D.C.Cir.1981).

This Circuit has recently refined its approach to cases involving self-concealing wrongs and placed on the defendant the burden of proving that the plaintiff did not exercise due diligence. Thus, in Richards v. Mileski, the court held “[w]hen tolling is proper because the defendants have concealed the very cause of action, ... they have the burden of coming forward with any facts showing that the plaintiff could have discovered ... the cause of action if he had exercised due diligence.” 662 F.2d at 71.106 The jury in this case was so instructed. In finding for plaintiffs on the statute of limitations issue, the jury found that defendants had not proved by a preponderance of the evidence that plaintiffs failed to exercise reasonable diligence in pursuing their claims. Defendants do not seriously challenge that finding.

Before turning to apply these principles to defendants’ principal assertions about the sufficiency of the evidence on fraudulent concealment, we pause to note an obvious, albeit often overlooked, proposition. The doctrine of fraudulent concealment does not come into play, whatever the lengths to which a defendant has gone to conceal the wrongs, if a plaintiff is on notice of a potential claim. A key aspect of a plaintiff’s case alleging fraudulent concealment is therefore proof that the plaintiff was not previously on notice of the claim he now brings. By “notice,” we refer to an awareness of sufficient facts to identify a particular cause of action, be it a tort, a constitutional violation or a claim of fraud. We do not mean the kind of notice — based on hints, suspicions, hunches or rumors — that requires a plaintiff to make inquiries in the exercise of due diligence, but not to file suit.107

While the precise standard for notice that amounts to constructive discovery and triggers the statute of limitations is not precisely delineated, for our purposes the contours are sufficiently clear. For one, the plaintiff must know facts giving notice of the particular cause of action at issue, not of just any cause of action. Richards provides an example. Plaintiff Richards was a career employee with the United States Information Agency until 1955, when he resigned under the duress of false charges of homosexual activity. At the time, he knew the charges were false. Years later he brought suit against six former federal officials alleging a variety of tort injuries. The suit was dismissed as untimely. On appeal this court reversed and held that the complaint stated sufficient allegations of fraudulent concealment to toll the statute of limitations. It found persuasive Richards’ argument that, while he was aware of the falsity of the charges in 1955, only in 1978 had he become aware that his superiors had knowingly filed false reports. On this issue; the court held that claims Richards might have brought in 1955 (such as wrongful discharge) and those he filed in 1978 were different — that it “was no mere ‘detail’ in 1955 that the false charges against Richards had been fabricated as part of a deliberate conspiracy against him, or that his own superiors rather than an unknown informant were *36the source of his misery”108 — and that mere knowledge of falsity did not constitute such notice as to bar tolling the statute.

Second, plaintiffs knowledge of the grounds for a suit must generally extend to an awareness of the persons responsible for plaintiffs injury. We by no means imply that a plaintiff may postpone suit until he knows every defendant by name and title. However, simply because a person knows he has been injured by one person cannot reasonably mean he should be held to know of every other participant.109 This Circuit’s decision in Fitzgerald v. Seamans, 553 F.2d 220 (D.C.Cir. 1977), illustrates the point. Plaintiff Fitzgerald sued officials of the Air Force and the Executive Branch charging a conspiracy to deprive him of his civilian position with the Air Force and of his constitutional rights in retaliation for testimony he gave before Congress. We held there that Fitzgerald’s prior administrative appeals disclosed that he had reason to know he was being eased out of his job in retaliation for his testimony in time to file his suit against the. Air Force officials within the given period of repose. However, we also held that he had no reason to know of the involvement of a White House official until significantly later. Because the White House official was “a person of influence in a different center of power,” id. at 229, the court declined to hold that Fitzgerald was on notice of his participation. We therefore must probe a plaintiff’s knowledge to determine whether he was on notice of all possible defendants, and not just a subgroup, as well as the particular cause of action.

B. The Tolling Doctrine Applied

On the basis of the foregoing, our analysis is quite straightforward. We find that the record contains overwhelming and uncontradicted evidence of defendants’ efforts to construct a scheme that would remain secret and that the wrong fits neatly under the “self-concealing” or “unknowable” label. At the same time we find that three plaintiffs were on notice of their claims against the FBI, though not against the District defendants, more than three years before they filed suit, and that the statute of limitations therefore has run on those particular claims.

1. The Self-Concealing Wrong

The FBI took numerous steps to assure that no aspect of its COINTELPRO program would come to the attention of the public, or, indeed, of anyone outside the FBI. Steps were also taken to limit knowledge of COINTELPRO operations even among FBI personnel. Thus, in any early FBI memorandum regarding COINTEL-PRO-Black Nationalist, it was written,

You are also cautioned that the nature of this new endeavor is such that under no circumstances should the existence of the program be made known outside the Bureau and appropriate within-office security should be afforded to sensitive operations, and techniques considered under the program.

IV J.A. 1728. A later airtel from FBI headquarters again cautioned the Field Office agents that COINTELPRO proposals must be designed to avoid embarrassing the Bureau. IV J.A. 1735. The record discloses numerous ways in which the FBI sought to hide its activities. Two examples suffice. First, the Bureau published The Rational Observer on unwatermarked paper “[f]or the sake of security,” IV J.A. 1853 (memorandum from SAC, WFO to Director, FBI), and falsely attributed authorship of the articles therein to “a small *37group of students,” IV J.A. 1854. Second, the Bureau mailed out the “Give Them Bananas!” leaflet110 anonymously, in “unmarked white envelopes,” taking “all necessary steps to protect the identity of the Bureau as the source of these leaflets.” 111 These affirmative efforts at concealment leave no doubt that the FBI defendants engaged in a “self-concealing” wrong. There also was ample evidence to find that the District defendants on their own engaged in activities that they deliberately concealed. They sent informants into the New Mobe, WMC, BUF, ECTC, Institute for Policy Studies, WAP AC, and into the two plaintiff organizations, who misrepresented their identities.112 Moreover, the MPD Intelligence Division shredded all its political surveillance files in 1974. III J.A. 1317 (testimony of Jerry Wilson). Whether the MPD did so for housekeeping or for concealment purposes was a question of fact for the jury.

The record in this case is replete with examples of efforts to prevent public awareness of the conduct at issue in this case. Our review leaves no doubt of the existence of evidence that defendants deliberately constructed schemes of such a kind that plaintiffs would not even suspect that any outsider was meddling in their lawful activities, much less that their constitutional rights were being violated. Accordingly, this case properly went to the jury under the framework established by Richards for adjudicating claims of fraudulent concealment emanating from self-concealing wrongs, and we decline to hold that insufficient evidence supports the jury’s verdict on this issue.113

*382. Notice to Trigger the Statute of Limitations

Up to now, we have determined, first, that the evidence supports the conclusion that defendants concealed their activities— by contriving schemes that were by their nature unknowable — and that defendants failed to offer evidence of facts plaintiffs might nevertheless have uncovered with due diligence. Thus, unless we determine as a matter of law that plaintiffs nonetheless had notice of the cause now sued upon, plaintiffs will have proved “fraudulent concealment.” We turn now to this last inquiry.

Initially, we must make clear what we are and are not seeking. Plaintiffs recovered in the District Court on a claim that their First Amendment rights were infringed. We therefore look only for evidence of knowledge of facts about that claim. Knowledge of possible other wrongs, such as illegal wiretaps or burglaries, or of proper law enforcement activities aimed at plaintiffs, such as surveillance, surely do not create a cause of action under the First Amendment by themselves and will not meet our standard of notice.

The FBI defendants suggest a variety of incidents whose occurrence they believe placed plaintiffs on timely notice of their claims against the FBI. First, they point to articles by former FBI Special Agent Robert Wall, published in the New York Review of Books on January 27, 1972, and in the Potomac Magazine, a Sunday supplement to the Washington Post, on March 5, 1972.114 The articles by and about Wall purported to describe secret FBI activities, including the existence of COINTELPRONew Left and its intent to disrupt New Left and antiwar groups, FBI infiltration of New Left groups, the FBI’s effort to exploit dissension between the New Mobilization Committee and the Black United Front arising out of the “head-tax” demand, the FBI’s investigation of the Poor People’s Campaign, and the FBI’s investigation of the Institute for Policy Studies and persons associated with it. Defendants remind us that several plaintiffs admitted to reading these articles more than three years prior to filing suit. Second, other information was publicly available regarding FBI activities, according to defendants. They point to testimony in a separate civil action regarding FBI investigations of the New Mobilization Committee, and to disclosures (in unidentified publications) following burglary of an FBI office.115 The FBI defendants concede, however, that no plaintiff admitted to being aware of such information. Third, the FBI defendants detail specific instances in which each plaintiff knew or suspected he or she was the subject of an FBI investigation or surveillance.

The District defendants reiterate many of the same pieces of evidence that the FBI defendants offer. They also point to a Washington Post article, published in June 1973, in which a former MPD officer described MPD efforts to infiltrate the Institute for Policy Studies and the Student Mobilization Committee, and his own (apparently unauthorized) efforts to coordinate electronic surveillance of demonstrators with federal law enforcement groups.116

We are not persuaded that any of these examples, standing alone, suffices to have put plaintiffs on notice of their claims. A person’s suspicion that he is a target of lawful law enforcement activity, such as *39surveillance, or even that his organization has been infiltrated by informants, cannot conceivably constitute notice of possible constitutional violations, without creating the anomalous situation of requiring persons to file suit on a hunch, only to be dismissed for failure to state a claim.117 A court simply cannot “require that an aggrieved party have proceeded from the outset as if he were dealing with thieves.” Hudak v. Economic Research Analysts, Inc., 499 F.2d 996, 1002 (5th Cir.1974), cert. denied, 419 U.S. 1122, 95 S.Ct. 805, 42 L.Ed.2d 821 (1975). To require potential plaintiffs to spin a tale of unconstitutional activity from a mere awareness of lawful investigation would impose just such a burden of omnipresent suspicion.

Nor does the mere fact that a plaintiff might have read a newspaper article detailing an FBI scheme to disrupt certain organizations 118 cast upon him a duty to presume that he was a target and to file suit. In a different setting, we declined to infer notice of an accounting firm’s fraudulent practices on the basis of one newspaper article criticizing certain accounting practices, and held, “As a matter of law, we believe that one article challenging the accounting procedures of a reputable firm is insufficient to impute knowledge of fraud to appellants.” Wachovia Bank & Trust Co. v. National Student Marketing Corp., 650 F.2d 342, 349 (D.C.Cir.1980), cert. denied, 452 U.S. 954, 101 S.Ct. 3098, 69 L.Ed.2d 965 (1981). While such an article might trigger a duty to inquire, we do not believe that such seemingly outrageous allegations as appeared, unverified, in the Wall articles on their own constitute as a matter of law notice of constitutional violations to all unidentified victims. This is particularly so in a case such as this one, in which plaintiffs’ injury, at least initially, was difficult to identify and, consequently, would not necessarily have provided verification that plaintiffs were targets of the FBI’s covert activities. We do not, of course, imply that an article can never constitute notice, but only that it was within the range of the jury’s authority to resolve the issue.

Nonetheless, as to some plaintiffs, not one but several of the factors to which defendants point converged; as a matter of law, we hold that the confluence of factors placed three plaintiffs on notice of their claims. Those plaintiffs who not only read the Wall articles more than three years before filing suit, but also knew that they were subjects of FBI investigation, had enough timely information to claim that they were victims of unconstitutional FBI activities. The FBI defendants’ directed verdict motions as to those plaintiffs were therefore improperly denied. These plaintiffs did not, however, have any reason to know of District involvement; therefore, their claims against the District defendants are not time-barred.

According to the FBI defendants, four plaintiffs, Abbott, Bloom, Booker and Was-kow, were aware in early 1972 of the Wall allegations, and also of other “red flags” signaling a cause of action. We consider each of these plaintiffs in turn.

Mr. Booker: Mr. Booker testified that he read an article by Special Agent Wall in the Washington Post, evidently at the time of its publication. II J.A. 829-30 (testimony of Reginald Booker). The article specifically noted the existence of COINTELPRO and reported that the project was designed, “to thwart and undermine the activities of any organization that fell into the category of ‘New Left.’ ” IV J.A. 2183, 2188 (Wall article). More specifically, the article noted that the FBI tried to “create dissent among *40the various groups involved in the New Left to prevent them from working together,” id., and it specifically gave as an example the FBI’s forgery of a letter to the NMC demanding payment for Black support of NMC peace efforts. Id. Mr. Booker had been present at BUF meetings in the late 1960s focusing on whether BUF should propose that a head tax be paid by demonstrators coming into Washington, see II J.A. 811-12 (testimony of Reginald Booker), and believed no such policy ever was adopted. Id. Moreover, he had had meetings with FBI agents in which they had tried to get him to be an informant. Id. at 826. Thus, Mr. Booker was on notice both of the injury he had suffered and of the unlawful intent on the part of the FBI that made the injury actionable. The ease is easily distinguishable from the situation in Richards, in which the plaintiff knew of the former (injury) but not the latter (unlawful motive), and requires a contrary result.119

Mr. Abbott: Mr. Abbott is in a position similar to that of Mr. Booker. He recalls that in 1972 he “read ... very avidly” the Wall articles in the New York Review of Books and the Washington Post. II J.A. 934-35 (testimony of Sammie Abbott). At the same time, he had personal knowledge of the head tax issue discussed in the articles, id. at 919-20, knew the FBI was interested in using him as an informant, id. at 924, and well before 1972 he was aware that his telephone was tapped. Id. at 931-32. As with Mr. Booker, this dual awareness of injury and motive sufficéd to put Mr. Abbott on notice of his constitutional claim against the FBI.

Mr. Bloom: Mr. Bloom is in a position like that of Messrs. Abbott and Booker. He, too, testified that he read the Wall article in the Washington Post when it was published and heard about the New York Review of Books piece. II J.A. 702-03 (testimony of Abe Bloom). He was involved in the head tax controversy discussed in the article, id. at 678-80, and he also suspected he was under surveillance. Id. at 663. Most pointedly, Mr. Bloom testified that after reading the Wall article, he came to believe that the Rev. Moore demand letter was written by an FBI agent. Id. at 702-03. With such strong evidence that Mr. Bloom knew of his injury, and of the role of the FBI in causing that injury, as early as 1972, we do not believe his claims against the FBI should have gone to the jury.

Mr. Waskow: The facts regarding Mr. Waskow are not so clear, and we believe his claims properly went to the jury for resolution of the fraudulent concealment issue. First, Mr. Waskow admitted that he had read articles about Special Agent Wall, but did not testify as to when he read them; nor did he describe them in sufficient detail to enable us to determine which of many articles on COINTELPRO he might have read, or whether they might have put him on notice of the intent of the FBI’s COINTELPRO program. See II J.A. 902-03 (testimony of Arthur Waskow). We cannot conclude on this basis that Mr. Was-kow read any article by or about Special Agent Wall more than three years before filing suit. While there was evidence that Mr. Waskow suspected that he and the Institute for Policy Studies were targets of an FBI investigation, see id. at 884, 897-99, and that he knew about the head tax issue, id. at 868, on their own such suspicions certainly do not amount to knowledge of constitutional violations. The jury could reasonably have concluded that Mr. Was-*41kow did not know the contours of his claim until mid-1973.

In sum, we hold that the motions for directed verdicts on timeliness grounds should have been granted in favor of the FBI defendants only, as to plaintiffs Booker, Abbott and Bloom. As a matter of law, each of the three was on notice of his claim more than three years before filing suit. None of the other plaintiffs testified to sufficient awareness of his injury and the motive with which it was inflicted to require the issue to be taken from the jury. Moreover, we find from the record that none of the plaintiffs was cognizant of facts regarding the District defendants that should have led them, as a matter of law, to file constitutional claims against the District or its employees. The District and MPD represent a wholly separate organization from the FBI and, on the basis of Fitzgerald v. Seamans, we decline to hold plaintiffs to a constructive knowledge of the possible involvement of organizations outside the FBI.

C. Remaining Objections to the Fraudulent Concealment Instructions

The FBI defendants challenge the trial court’s failure to instruct the jury that “mere silence by a defendant is not fraudulent or deliberate concealment.” Brief of FBI Appellants, p. 67. They also claim as error the court’s instruction that concealment for law enforcement purposes may be deemed to be fraudulent or deliberate concealment; they assert that this instruction conflicts with alleged principles that concealment by a third party may not be attributed to named defendants, and that concealment must have been wrongful. Id. According to defendants, these “erroneous” instructions entitle them to a new trial. We disagree.

First, the unrefuted evidence presented at trial made absolutely clear that the FBI conspirators here had not merely been silent, but had made affirmative misrepresentations, and had assumed false identities as protestors, students, and parade coordinators, to conceal their official status. Much as in Richards, defendants here admitted that they misrepresented their actual position and then remained silent, knowing their representations were false. On those facts, an instruction that “mere silence” does not constitute concealment could only have been confusing. Regardless whether a proposed instruction correctly states a legal abstraction, when the instruction is not applicable to the facts, it is properly denied.120

Second, we reject both defendants’ characterization of the District Court’s instruction on concealment for law enforcement purposes and its challenge to that instruction. The court did not instruct that concealment for law enforcement purposes “may be deemed to be fraudulent or deliberate concealment,” as though all law enforcement secrecy amounted to fraudulent concealment. Rather, the court instructed that if there was deliberate concealment, “it doesn’t make any difference whether the defendant or conspiracy kept the material [information] from the plaintiff because the agency employing the defendant or the conspirators believed that law enforcement considerations required concealment.” 121

Neither of defendants’ challenges to this instruction warrants a new trial. For one, the instruction was not contrary to defendants’ “principle” that concealment by a third party may not be attributed to the named defendants — assuming, arguendo, that we accept that “principle.” First, the instructions generally required that con*42cealment be attributable to the defendant or, with the conspiracy claims, to the overall conspiracy;122 and second, the particular instruction at issue addressed concealment of information by defendants because defendants’ employers believed it necessary, not concealment by the employers. Nor was the instruction contrary to defendants’ “principle” — which, again, we only accept arguendo —that concealment must be wrongful. The District Court properly noted that, as in Smith v. Nixon, 606 F.2d 1183 (D.C.Cir.1979), cert. denied, 453 U.S. 912, 101 S.Ct. 3147, 69 L.Ed.2d 997 (1981), and Fitzgerald v. Seamans —cases in which we found claims of Government secrecy to constitute adequate allegations of concealment — the “wrongfulness” in this case “consisted of defendants’ efforts to conceal COINTELPRO and the MPD activities from the public despite the unlawfulness of the FBI and MPD programs.” 123 In other words, to the extent that “wrongful concealment” is required, it may be “wrongful” to devise a scheme to conceal information even if done on the belief that one’s superiors believe law enforcement purposes to compel secrecy. The instruction as given did not conflict with this principle. Moreover, we suspect that this argument is but a masked effort to challenge the District Court’s failure to charge that concealment must be wrongful. Since defendants did not raise this objection below, we decline to speculate further on their perceived inconsistencies between the instruction and the asserted “principle.” We find none and reject the claim.

IV. Defendant Courtland Jones

Defendant Jones raises several arguments unique to his appeal that result from the protracted procedural history of plaintiffs’ case against him. Our review of the details of that history convinces us that Mr. Jones’ assertions are without merit.124

We begin with a summary of the procedural history of the case against Mr. Jones, the details of which are critical to our decision. Mr. Jones was not named as a defendant in the first complaint filed by plaintiffs in July 1976. “By 1977, however, once discovery had begun, plaintiffs learned of Jones’ role in the ‘internal security’ activities of the WFO, and apparently believed him to have injured them in connection with COINTELPRO effort at the WFO.” Hobson, 556 F.Supp. at 1184. Plaintiffs therefore included Mr. Jones as one of the FBI defendants named in their amended complaint filed October 28, 1977.125 Plaintiffs made several attempts thereafter to serve Mr. Jones, first at the WFO, later at his residence. Mr. Jones has admitted that on December 19, 1978, he was aware that the papers were served at his residence but, because the papers were left at his residence, and not served on him personally (or left with a person of suitable age), he has claimed he was not served.126 In December 1978, pursuant to 28 C.F.R. § 50.15 (1977), Mr. Jones requested that the Department of Justice appoint counsel for him. Hobson, 556 F.Supp. at 1184. The Department assigned to him the lawyer who was then and throughout trial *43continuously representing all the FBI defendants.

In April 1979, Messrs. Jones and Schlar-man, another defendant, moved for dismissal of the claims against them, on the ground that “service of process was not effected upon a person of ‘suitable age and discretion’ at their homes in accordance with Rule 4(d)(1) of the Federal Rules of Civil Procedure.” Hobson, 556 F.Supp. at 1184 (quoting a November 9, 1979, memorandum opinion by Judge Pratt). The District Court judge granted the motion “without prejudice to fresh attempts to serve process.” Id.

In November 1980, after the case was reassigned to Judge Oberdorfer, before whom it was tried, the court directed the parties to “ ‘propose a practical solution for the problem posed by the failure to effect service on a number of defendants,’ including defendant Jones.” Id. at 1185 (quoting Hobson v. Wilson, No. 76-1326 (D.D.C. Nov. 14, 1980)). According to Judge Ober-dorfer,

At status calls held in late 1980 and in January 1981, the Court discussed the failure of plaintiffs to serve Jones and other defendants; however, counsel for the “served” FBI defendants] declined to take any action, claiming not to represent those unserved defendants and because those defendants “are not under the personal jurisdiction of the Court.”

Id. (quoting Letter of David M. White (dated Nov. 25, 1980), filed as Attachment to Order of Dec. 8, 1980).

On June 21, 1981, plaintiffs perfected service of process on Mr. Jones. Pretrial orders then in effect scheduled discovery to terminate at the end of July 1981. Mr. Jones did not seek an extension of the discovery period,127 but in August 1981 moved for dismissal of the claims against him, inter alia, for failure to prosecute under Rule 41 of the Federal Rules of Civil Procedure. He argued principally that he would be unfairly disadvantaged by trial in the autumn of 1981. The District Court rejected his motion:

[H]e will suffer no prejudice if plaintiffs are precluded from raising any claim against him not stemming from acts already involved in the litigation against other defendants who are represented by Jones’ counsel, and who were with Jones in the Federal Bureau of Investigation at the time of his allegedly unlawful conduct. Counsel for Jones is invited to submit with his pretrial brief an appropriate order concerning claims to be precluded.

Id. at 1185 (quoting Hobson v. Wilson, No. 76-1326, slip op. at 5 (D.D.C. Oct. 29, 1981)). Mr. Jones’ FBI counsel did not submit the proposed order to the court.128

After trial, Mr. Jones renewed his argument that dismissal was required under Rule 41 for lack of prosecution. The court again disagreed.

On appeal, Mr. Jones raises several arguments. First, he contends that even if the statute of limitations was properly tolled as to all defendants named in the July 1976 complaint, it was not tolled as to himself,, because he was not named in a complaint until the following year. In addition, he argues, the date of the amended complaint does not relate back to the original complaint because it does not meet the requirements of Rule 15(c) of the Federal Rules of Civil Procedure. Second, he argues' that *44the statute of limitations was not tolled until service of process was perfected in 1981, and that plaintiffs surely cannot argue that they did not know of the claim until three years before 1981. Third, he again renews his assertion that the action should have been dismissed under Rule 41 for failure to prosecute. We consider each argument in turn.

First, as even Mr. Jones appears to concede,129 the first official verification of the COINTELPRO program occurred in November 1974, when Attorney General Saxbe held a widely publicized press conference. That press conference led the Fifth Circuit to conclude, in United Klans of America v. McGovern, 621 F.2d 152, 153-54 (5th Cir.1980), that the statute of limitations on plaintiffs’ claims challenging COINTELPRO began running in November 1974, because “[wjhere events receive such widespread publicity, plaintiffs may be charged with knowledge of their occurrence.” Id. at 154. In United Klans, the court observed that Attorney General Saxbe had identified plaintiffs’ group as one of the COINTELPRO targets. We need not consider here whether that press conference as a matter of law put all plaintiffs on notice of their claims against the FBI. We have absolutely no reason to believe that the plaintiffs whose actions against the FBI remain in light of our statute of limitations discussion had notice before that press conference. Mr. Jones has not pointed us to evidence to that effect. Thus, assuming arguendo that the conference made public sufficient information to put all plaintiffs on notice, they still filed their claim against Jones in October 1977, within three years of that press conference.130 We therefore conclude that plaintiffs amended their complaint, and filed the amended complaint, within three years of the earliest date on which they might have had notice of their claims on the basis of public knowledge. We reject Mr. Jones’ argument that the action against him was not timely filed and find it unnecessary to consider whether the filing date of the amended complaint relates back to the original complaint.

Mr. Jones next argues that even if the amended complaint was filed within the period of repose, the statute was not tolled until service was complete. We find that this argument is based on an incorrect view of the role of service of process in a case in this Circuit and, accordingly, reject it.

It is now well-established that in diversity actions, in which a federal court applies a state statute of limitations, the court must also follow the state’s rule as to whether the filing of a complaint, the service of process, or some other procedural hurdle commences the action and tolls the statute of limitations. Walker v. Armco Steel Corp., 446 U.S. 740, 100 S.Ct. 1978, 64 L.Ed.2d 659 (1980). While it is also quite clear that a federal court confronted with a federal cause of action, but borrowing a state statute of limitation in the absence of a federally prescribed limitation period, must follow federal law as to the procedural steps that commence an action for statute of limitation purposes, the Supreme Court’s decision in Walker v. Armco Steel Cory, has led some litigants to argue that State law controls.131 On this *45point, we note that Justice Marshall, writing in Walker, expressly stated that the holding did not extend to federal causes of action;132 however, we need not reconsider that issue here because, under decisions both of this court in federal actions and of the District of Columbia Court of Appeals, it is unquestionably the case that the filing of a complaint commences an action for statute of limitation purposes. See Jordan v. United States, 694 F.2d 833, 837 & n. 7 (D.C.Cir.1982) (appellants need only timely filing, not proper service, to toll the statute of limitations in federal tort claims action); Varela v. Hi-Lo Powered Stirrups, Inc., 424 A.2d 61, 62 (D.C.1980) (en banc) (holding that D.C. law requires “only the filing of a complaint to commence an action and thereby toll the statute of limitations,” and that questions about plaintiffs diligence in obtaining service of process should be addressed by means of a motion for failure to prosecute). There can be no doubt that the statute tolled as to claims against Mr. Jones at the latest when the amended complaint was filed. Delay in service of process is an issue more properly addressed as an element of a motion to dismiss for failure to prosecute. We now turn to consider Judge Oberdorfer’s ruling on that issue.

In reviewing a ruling of the trial court on a motion to dismiss under Rule 41(b), our task is extremely limited. Absent a clear abuse of discretion, we must affirm the ruling of the trial court. Cherry v. Brown-Frazier-Whitney, 548 F.2d 965, 969-70 (D.C.Cir.1976); Sheaffer v. Warehouse Employees Union, 408 F.2d 204, 206 (D.C.Cir.), cert. denied, 395 U.S. 934, 89 S.Ct. 1996, 23 L.Ed.2d 449 (1969). Our inquiry is especially circumscribed when we review a refusal to dismiss; it is generally accepted that cases in that posture warrant a particularly narrow standard of review, see Finley v. Parvin/Dohrmann Co., 520 F.2d 386, 390 (2d Cir. 1975),133 under which discretion

is abused when the judicial action is arbitrary, fanciful or unreasonable, which is another way of saying that discretion is abused only where no reasonable man would take the view adopted by the trial court. If reasonable men could differ as to the propriety of the action taken by the trial court, then it cannot be said that the trial court abused its discretion.

Finley, 520 F.2d at 390 (quoting Delno v. Market Street Railway, 124 F.2d 965, 967 (9th Cir.1942)). This narrow standard is attributed to the procedural posture that a refusal to dismiss will have when it reaches the appellate court. “There is a natural reluctance to reverse a plaintiff’s judgment on the merits because of a pretrial order refusing to dismiss for want of prosecution.” Finley, 520 F.2d at 391. The reason is simple: a trial court’s discretionary denial of a defendant’s motion to dismiss is not nearly so harsh as would be an appellate court’s decision that a judgment for plaintiffs must be reversed and remanded for a dismissal with prejudice.

In this case, while we share Judge Oberdorfer’s view that plaintiffs displayed “unusual indifference” to the requirement of service,134 we also find entirely reasonable his conclusion that defendant Jones was not prejudiced by the delay. Absent *46prejudice, we cannot find that the motion to dismiss was improperly denied, particularly since plaintiffs obviously were trying to prosecute their claim when the motion was filed. Moreover, we are persuaded that the alternative to dismissal that was imposed— limiting plaintiffs’ claims against Mr. Jones to those stemming from acts already involved in the litigation against other defendants — adequately protected Mr. Jones’ interests.

This case is immediately distinguishable from those in which delay in service causes defense counsel to rush in their preparations. See, e.g., Anderson v. Air West, Inc., 542 F.2d 522, 525 (9th Cir.1976). Here, Mr. Jones shared counsel with the other defendants and, because of the court-imposed limitation on claims against Mr. Jones, he received the full benefit of those preparations. Mr. Jones’ failure both to request an extension in discovery, and to submit suggestions of claims to be precluded, cannot now be turned against the plaintiffs. Dismissal is a severe sanction, and alternative sanctions are not only permissible but often preferred. See 9 C. Wright & A. Miller, Federal Practice and Procedure § 2370, at 203 (1971) (“appellate courts do not look favorably on dismissal with prejudice if there are lesser sanctions that could vindicate the purpose of the rules”). It simply is not for a defendant to snub the court’s alternative of limiting the claims against him and then argue prejudice.

Moreover, this case differs from one in which a defendant on whom tardy service is perfected is unfairly robbed of an expectation that he will not have to defend a claim. Here the trial judge made clear to FBI counsel that plaintiffs sought to serve Mr. Jones and that the court wanted the parties to work out a procedure. It was undoubtedly reasonable for Judge Oberdor-fer to conclude that Mr. Jones’ appointed counsel relayed this information.135

On the basis of the foregoing, we conclude that the District Court did not abuse its discretion in denying Mr. Jones’ motion to dismiss under Rule 41(b).

V. Juror Contact

After the close of evidence but before closing arguments in this action, plaintiffs’ counsel contacted, and had a brief conversation with, an excused juror. All defendants argue that the conversation deprived them of a fair trial and ask for a new trial. While we fully agree with the District Court’s conclusion that the juror contact was improper, we nonetheless agree with the trial judge’s determination that no prejudice flowed from this misconduct so as to warrant a new trial.

On December 14, 1981, following the close of evidence, but prior to closing arguments, one of the plaintiffs’ attorneys contacted a person who had sat on the jury early in the trial but had been excused on December 3, 1981. In a telephone conversation, the attorney discussed with the dismissed juror her view of the case and the evidence. Four days later, on December 18, the court examined the juror on the record136 and invited questions from counsel. The same day, the court examined the lawyer who had made the contact.137 Thereafter, counsel for all defendants moved for a mistrial, which was denied.

According to the testimony, the conversation with the juror was extremely brief. The dismissed juror freely offered her opinions regarding the evidence she had heard, while denying that the jurors had discussed the merits of the case among themselves.138 According to the attorney, who testified from her notes of the conversation, the juror comented that the quantity of evidence was overwhelming and con*47fusing, that the plaintiffs appeared to be able to accomplish the goals they had set out despite defendants’ activities, that she did not understand the significance of The Rational Observer, that she thought it was a “major point that the FBI was involved in sending false signals to marshals during certain demonstrations,” that she was not convinced the defendants managed to incite the police against “the peace people,” and that the statute of limitations questions seemed very difficult.139 The juror’s recollection of the conversation, while less precise, was substantially in accord.

The District Court focused on the possible impact of the conversation on the fairness of the trial and expressly found that “there was no prejudicial impact on the jury’s verdicts stemming from the incident.” Hobson, 556 F.Supp. at 1188.140 In taking this approach, the court adhered to the Supreme Court’s decision in Mattox v. United States, 146 U.S. 140, 13 S.Ct. 50, 36 L.Ed. 917 (1892), which establishes a presumption that contact with a sitting juror should invalidate a verdict unless the harmlessness of the contact is shown. While we agree generally with the trial court’s analytical process, we believe that this case is not directly controlled by Mattox but is rather a logical extension of it, and should be recognized as a variation on the circumstance of Mattox. Indeed, no case to which we have been pointed, nor that we have found, addresses the standard of review of attorney contact with an excused juror. Moreover, confusion even about the permissibility of such contact apparently gave rise to the incident we confront.141 Accordingly, we will try here to make absolutely clear the conduct we will expect of officers of the court, and the consequence of a failure to adhere to such principles.

That the Mattox rule is firmly entrenched in our jurisprudence is beyond dispute. The case leaves no doubt that once an improper communication has been made to a juror, a presumption of prejudice arises that is rebuttable only by a strong contrary showing. As the Mattox Court held: “Private communications, possibly prejudicial, between jurors and third persons, or witnesses, or the officer in charge, are absolutely forbidden, and invalidate the verdict, at least unless their harmlessness is made to appear.” Mattox, 146 U.S. at 150, 13 S.Ct. at 53; see also Remmer v. United States, 347 U.S. 227, 229, 74 S.Ct. 450, 451, 98 L.Ed. 654 (1954). While Mat-tox was a criminal case, the same principle *48has been applied to civil actions. See Rinker v. County of Napa, 724 F.2d 1352 (9th Cir.1983) (section 1983 action); Krause v. Rhodes, 570 F.2d 563 (6th Cir.1977) (section 1983 action), cert. denied, 435 U.S. 924, 98 S.Ct. 1488, 55 L.Ed.2d 517 (1978); United States v. Harry Barfield Co., 359 F.2d 120 (5th Cir.1966) (taxpayer refund suit); Washington Gas Light Co. v. Connolly, 214 F.2d 254 (D.C.Cir.1954) (tort action); Port Terminal & Warehousing Co. v. John S. James Co., 92 F.R.D. 100 (S.D.Ga. 1981) (antitrust action).

Both Mattox and the Supreme Court’s subsequent decision in Remmer v. United States considered the possible prejudicial effect of third-party contacts with sitting jurors and determined that the likelihood of prejudice warrants imposition of a rebuttable presumption of prejudice. Following the same analysis, we hold that attorney contact with an excused juror, during trial, is potentially harmful and similarly creates a rebuttable presumption of prejudice. We believe that concerns related to those implicated in Mattox and Remmer exist in this case and warrant such a presumption to safeguard against prejudice. However, the two scenarios also differ significantly. These differences should come into play when the judge determines whether the contact has in fact been prejudicial. We would anticipate, for example, that in the case of an excused juror, rebuttal will be more easily accomplished, and the trial judge will find áctual prejudice with less frequency.

Behind the Mattox presumption of prejudice lies the most fundamental premise of our jury system: that each juror enters a case impartial, and throughout trial remains uncontaminated by outside pressures. Trial by jury presupposes that the jury works in a controlled environment, untouched by any influence other than that properly permitted by the trial judge. “This protection and safeguard must remain inviolate if trial by jury is to remain a viable aspect of our system of jurisprudence.” United States v. Harry Barfield Co., 359 F.2d at 124. The assumption of impartiality achieves two ends: it insures that defendants receive a fair trial and it maintains the faith of litigants in the probity of verdicts and the integrity of the system. See Krause v. Rhodes, 570 F.2d at 568 (quoting United States v. Ferguson, 486 F.2d 968, 971 (6th Cir.1973)).

Extraneous influence on a sitting juror might have one or more of several results. It creates a passageway through which information may flow both into and out of the jury room. Contact might intimidate, either obliquely or directly; it might subtly create empathy; it might disclose facts of which the jury ought not to be aware.142 Contact also might result in disclosure by a juror of confidential information about the jury deliberations, or the impressions of individual jurors. The existence of any communication link between the jury and interested parties thus potentially contaminates the controlled environment irreparably. These concerns give rise to the presumption of prejudice. Of course, the resulting degree of actual prejudice varies, depending, for example, on whether the contact is directly with a party, relates to the merits of the case, involves threats, or is communicated to other jurors. These factors no doubt are persuasive when the trial judge determines whether the presumption of prejudice is rebutted.

Contact with an excused juror, before the verdict is rendered, could also influence the jury. First, indirectly, such a conversation makes possible precisely the same kind of influence on jury deliberations as direct contact with a sitting juror. An excused juror could serve as a conduit of information into the jury room. Indeed, in the single case we have found involving a defendant’s communication with an excused juror, the excused juror telephoned a sitting juror to recount the conversation, who then repeated what was said to other jur*49ors. See United States v. Boscia, 573 F.2d 827 (3d Cir.), cert. denied, 436 U.S. 911, 98 S.Ct. 2248, 56 L.Ed.2d 411. (1978). If this were the only thing to be feared, we might have some reservations about a rule that any discussion between an excused juror and any third party, about the merits of the case, necessarily creates a presumption of prejudice (primarily because the likelihood of prejudice dissipates drastically); a cautionary instruction to the excused juror not to discuss the case, coupled with a judge’s inherent discretion to declare a mistrial, might produce an adequate safeguard.

However, there is in fact much more to be feared. Contact between a party, his attorney, or an agent of either, and an excused juror, additionally offers parties the opportunity to learn information to which they ought not be privy. Perhaps most importantly, an excused juror might relay to parties or their counsel his or other jurors’ impressions of the credibility of witnesses and the strength of evidence, thereby permitting counsel to address specific concerns with additional witnesses and different evidence. If such contacts were permitted, counsel undoubtedly would undertake detailed interrogations of excused jurors. This would enable counsel to benefit unfairly and unequally from access to information that the process quite properly denies, and thereby make a mockery of the underlying principles on which trial procedure is based. Additionally, such contact can provide a mini-voir dire, without benefit of notice, participation or a record. There is absolutely no guarantee, as there is in a proper voir dire setting, that all parties will be present, or be treated to the same information. There is no safeguard against revelation of personal attributes or concerns of jurors to which a knowledgable party might then address arguments. In other words, the integrity of the jury process does not stop at the door to the jury box or jury room. It extends equally to the controlled environment of the voir dire. We simply will not permit the parties or their counsel to circumvent the voir dire process and grill excused jurors about their knowledge of the prejudices or biases of remaining jurors.

Accordingly, we find that contact between an excused juror and a party, his counsel, or an agent of either, implicates the two concerns underlying Mattox — the fairness of the trial and the overall integrity of the jury system — and similarly has the potential to prejudice the verdict. We therefore conclude that a presumption of prejudice arises from a showing that such contact has occurred. In this case, the trial judge found that the presumption was rebutted and that the contact had no prejudicial impact. We find Judge Oberdorfer’s reasoning convincing and conclude that his denial of the mistrial motion constitutes a proper exercise of discretion.

Our review of the trial court’s ruling on the mistrial motion is limited to assuring that the judge did not abuse his discretion.143 In this regard, his appraisal of the prejudice resulting from contact with an excused juror is entitled to great weight, and ought not be disturbed unless it is “manifestly unreasonable,” Miller v. United States, 403 F.2d 77, 84 n. 11 (2d Cir.1968), or unless new facts have been discovered since the ruling.

Several factors support Judge Oberdor-fer’s ruling of no prejudice. First, there is absolutely no evidence that any sitting jur- or was affected by the conversation. The excused juror did not mention the conversation to any sitting juror, and no sitting juror was contacted directly. Thus, half of our concern — for information flowing into the jury’s controlled environment — is relaxed. Second, the timing of the contact minimized the usefulness of any information that might have been received. Had the excused juror expressed concern about the honesty of any witness, or the probity of any evidence, plaintiffs’ counsel could not have shifted strategy or shored up testimony. After each side had completed its case, it simply was too late to adjust the *50evidence. Third, the nature of the information relayed was not of a kind to enable counsel to address the private biases or concerns of individual jurors in closing argument. At bottom, the conversation simply informed counsel that complex topics are complex, and that overwhelming amounts of information are overwhelming. To be sure, the excused juror focused on the difficulty of the statute of limitations issue — a proposition with which we fully agree — perhaps signaling to counsel to focus in closing argument on that issue. Yet we find it completely reasonable to conclude that counsel would otherwise have addressed in closing argument a major issue in the trial, and that defense counsel must have been prepared, and had full opportunity, to respond. In addition, since we have resolved much of the statute of limitations issue as a matter of law, we would be especially reluctant to override the trial judge and find that the conversation prejudiced the verdict as a matter of law.

In light of all of the foregoing, we affirm the ruling below on the jury contact question.

VI. Sufficiency of the Evidence

All defendants argue that the evidence at trial supports neither a finding of conspiratorial nor of individual liability against any of them.144 They concede that the jury was properly instructed to consider whether each defendant had, by individual conduct, injured each plaintiff, and that it found as much on the special verdict form. They also do not challenge the court’s instructions on the elements of a conspiracy or conspiratorial liability. Instead, they contend that the evidence did not support sending the case to the jury.

Defendants unsuccessfully raised this argument before the District Court, in motions for a judgment notwithstanding the verdict. Our task at this stage is limited. In reviewing the ruling of the District Court, we must view all evidence, and inferences to be drawn therefrom, in the light most favorable to plaintiffs. See Alden v. Providence Hospital, 382 F.2d 163, 165 (D.C.Cir.1967), quoted in Doe v. District of Columbia, 701 F.2d 948, 965 (D.C.Cir.1983) (Edwards, J.) (separate concurring statement). Moreover, unless we “determine that ‘no reasonable man could reach a verdict in favor of’ the prevailing party, the jury’s verdict must be allowed to stand.” Doe, 701 F.2d at 965 (quoting Muldrow v. Daly, 329 F.2d 886, 888 (D.C.Cir.1964)) (separate concurring statement). If reasonable persons could even disagree about the sufficiency of the evidence, we must uphold the judgment.

Stripped to its essentials, plaintiffs’ claim is that defendants undertook an array of activities for the unlawful purpose of disrupting and impeding plaintiffs’ lawful and peaceful efforts to protest Government policies. Defendants argue that individually each activity was a proper exercise of law enforcement authority. The gravamen of plaintiffs’ response is that it is immaterial here that certain of defendants’ activities were for .law enforcement' purposes because, simultaneously with these allegedly *51legitimate activities, defendants undertook a massive scheme for the express purpose of disrupting and defeating plaintiffs’ exercise of protected rights. In short, plaintiffs forcefully contend that it is a sham for defendants to plead that their activities were merely innocent efforts to obtain information for law enforcement authorities. As a general proposition, we agree with plaintiffs’ assertion that defendants have, in some important respects, badly mischar-acterized the evidence submitted to the jury.

Proof of the defendants’ unlawful purposes was offered in a variety of forms, including documents expressly articulating the COINTELPRO design, and testimony about incidents that could not have been designed simply to gather information or otherwise engage in lawful actions. Thus, in some instances activities that might appear proper — such as interviews of demonstrators — are alleged to have been improper because they were performed with an intent to intimidate, not to gather information. In reviewing the evidence, we fully recognize that we must be chary not to impute unconstitutional motive where none is to be found, and, in particular, not to permit any defendant to be penalized for engaging in legitimate law enforcement activity. At the same time, we must recognize that, in the presence of competent evidence, it was for the jury to decide the motive or intent with which actions were taken.

The foregoing principles do not, of course, require us to uphold a jury’s decisions in the absence of any evidence that might reasonably support its findings. In this case, our review of the trial transcripts and the evidentiary submissions, coupled with the parties’ extremely helpful briefs on these points, lead us to conclude that the evidence cannot support all aspects of the jury’s verdict. In particular, we find insufficient evidence that either the named, individual MPD defendants or the District of Columbia participated in a conspiracy to violate plaintiffs’ constitutional rights either within MPD or between persons in MPD and the FBI, and we reverse judgments against those defendants arising from those two alleged conspiracies. We also find insufficient evidence supporting the individual liability of the District or any MPD officer to any plaintiff and reverse the individual liability judgments against the District and each MPD defendant. Similarly, we find insufficient evidence that the FBI defendants participated in any conspiracy with persons in MPD and reverse those findings. The District Judge’s ruling on all other findings by the jury — as to the individual and conspiratorial liability of the FBI officers — is affirmed.

A. The Conspiracies

The jury found each defendant to have participated in one or more civil conspiracies designed to deprive plaintiffs of their First Amendment rights by disrupting their organizations and impeding their associations with other persons. One conspiracy was alleged to exist within MPD, a second was alleged between MPD and the FBI, and the third allegedly existed within the FBI. We will first consider the two conspiracies involving the MPD.

A civil conspiracy is defined as an agreement between two or more people to participate in an unlawful act or a lawful act in an unlawful manner. See Halberstam v. Welch, 705 F.2d 472, 477 (D.C.Cir. 1983); Hampton v. Hanrahan, 600 F.2d 600, 620-21 (7th Cir.1979), modified on other grounds, 446 U.S. 754, 100 S.Ct. 1987, 64 L.Ed.2d 670 (1980). An express agreement among all conspirators is not necessary. “A plaintiff .... need not prove that each participant in a conspiracy knew the ‘exact limits of the illegal plan or the identity of all participants therein.’ ” Hampton, 600 F.2d at 621 (quoting Hoffman-LaRoche, Inc. v. Greenberg, 447 F.2d 872, 875 (7th Cir.1971)). The conspirators “must share the general conspiratorial objective, but they need not know all the details of the plan ... or possess the same motives.” Id. Thus, to “demonstrate the existence of a conspiratorial agreement, it simply must be shown that there was ‘a *52single plan, the essential nature and general scope of which [were] known to each person who is to be held responsible for its consequences.” Id. (quoting Hoffman-LaRoche, Inc. v. Greenberg, 447 F.2d 872, 875 (7th Cir.1971)). To make the conspiracy actionable, there must also be an overt act in furtherance of the object of the conspiracy that injures plaintiff in his person or property or, in a section 1985(3) action, which deprives him of having or exercising any right or privilege of a citizen of the United States.

The District defendants argue,".. first, that they engaged only in lawful intelligence-gathering and therefore did not agree among themselves to do an unlawful act or a lawful act by unlawful means.145 Second, they argue, the evidence failed to demonstrate the existence of an MPD-FBI joint conspiracy.146 Without concluding that no such conspiracies existed, we agree with defendants that there was not enough evidence for the jury to conclude that the individual MPD defendants shared the “general conspiratorial objective” of intentionally disrupting plaintiffs’ First Amendment rights, either among themselves or with the FBI. No evidence linked to any individual MPD defendant warrants even an inference of unlawful motive in carrying out routine law enforcement activity.

According to the District defendants, the record reveals only that they engaged in the collection of intelligence data for law enforcement purposes, such as planning for security and crowd control at future demonstrations, by means of physical surveillance and informants. This process, they explain, not only is lawful, but also was recommended by the Kerner Commission as a means to anticipate and prepare for civil disorder.147 In addition, they concede, they occasionally obtained intelligence by electronic surveillance. We do not agree, and MPD does not seriously argue,148 that the record is devoid of evidence that any District officers engaged in activities disruptive of the plaintiffs’ activities protected by the First Amendment. We put aside such evidence for the moment, however, because we agree with the District that no evidence indicates any individual MPD defendant initiated or approved these actions. We turn our attention instead to evidence of the activities in which the individuals engaged.

Our review of the record leads us to the conclusion that the individual activities of the MPD defendants were in all relevant respects 149 consistent with the recommendations of the Kerner Commission and other law enforcement task forces, and did not support an inference of their participation in an unlawful conspiracy. The task forces either suggested or endorsed the framework by which the MPD defendants operated — from the establishment of the Intelligence Division as it existed in the relevant time period, to the information-gathering procedures it utilized, to the exchange of information with other law enforcement agencies. For example, the Kerner Commission reported on the “absence of accurate information both before and during a disorder” and recommended that police departments develop means to obtain adequate intelligence for planning purposes:

An intelligence unit staffed with full-time personnel should be established to gather, evaluate, analyze, and dissemi*53nate information on potential as well as actual civil disorders. It should provide police administrators and commanders with reliable information essential for assessment and decisionmaking. It should use undercover police personnel and informants, but it should also draw on community leaders, agencies, and organizations in the ghetto.

Kerner Commission Report, note 147, supra, at 269. Plaintiffs’ response is to detail evidence of incidents in which informants acted as agents provocateurs, but without linking the enumerated disruptive activities to any individual defendant. Nor is plaintiffs’ response surprising: the record simply lacks any evidence that the MPD defendants participated in, or knew about any conspiracy to disrupt, as opposed to a conspiracy to learn about plaintiffs’ protest activities.150 We decline to hold that the individual defendants’ positions in the MPD hierarchy alone permit an inference that each named defendant knew about and set in motion the disruptive activities of other MPD officers.

We are equally at a loss to find any evidence that the MPD defendants engaged in a conspiracy with the FBI to infringe plaintiffs’ rights. To be sure, as plaintiffs argue, “[ejvidence of a pattern of mutually supportive activity over a period of time provides a reasonable basis for inferring that the parties are engaged in a common pursuit.” 151 However, when the only evidéiice of such “mutually supportive activity” is conversation and the exchange of intelligence information — a form of cooperation expressly endorsed by presidential task forces at the time152 — it simply is not reasonable to infer unlawful activity. Absent some evidence that the individual MPD defendants knew about COINTELPRO,153 or engaged in disruptive activities *54on their own to complement those of the FBI, or in any other way violative of First Amendment rights went beyond the limits of proper intelligence-gathering, we cannot permit an inference of unlawful motive.154 We therefore conclude that individual MPD defendants could not reasonably have been found to know of, or participate in, an unlawful scheme along with the FBI.

Evidence of disruptive activity by persons other than the individual defendants is present, however, and requires us to consider whether a jury reasonably might infer that the District participated either in an internal MPD conspiracy or in one with the FBI. As to the latter, we must decide whether the flow of information back and forth, coupled with evidence of efforts to impede plaintiffs’ rights, permits an inference that the District and unnamed defendants worked with the FBI toward a common goal, even if the communication on its own does not suffice. Evidence of such disruptive practices includes the following:

—MPD undercover informer Ann Kolego Markovich infiltrated the Antiwar Union, an antiwar group that worked with the Peoples Coalition for Peace and Justice, another antiwar group with which Mr. Pollock was affiliated. According to the evidence, Kolego Markovich acted as an agent provocateur and urged others to commit violent acts during meetings and demonstrations planned by the Antiwar Union and the PCPJ. See II J.A. 789-94 (testimony of Carol Cullum); II J.A. 752-53 (testimony of Richard Pollock). At trial, Kolego Markovich denied this activity; the jury found that she was liable on the merits in some instances but that claims against her were barred by the statute of limitations.

—At an ECTC-demonstration planned by plaintiffs Abbott and Booker, a demonstrator urged the crowd to disregard instructions given by Mr. Abbott — whom the demonstrator called a “sell-out artist, phony, coward” — and march to an off-limits area where police were waiting. About 200 people did so, resulting in numerous arrests. Mr. Abbott later found out that the demonstrator was Jan Francis, an MPD officer. See II J.A. 804-06 (testimony of Reginald Booker); II J.A. 910-14 (testimony of Sammie Abbott).

—Immediately before a major demonstration, MPD informers were told to smash the duplicating equipment of peace groups to render it inoperable. See III J.A. 1393-94 (transcript of MPD hearings).

—The day before another rally, a New Mobe office worker known as Steve Wilcox wrote two bad checks to the telephone company; thereafter he disappeared. As a result, Mr. Bloom and his colleagues had to scramble to raise enough cash to cover the phone bills and avoid disruption of telephone service. See II J.A. 721-29 (testimony of Helen Anne Gurewitz); II J.A. 696-98 (testimony of Abe Bloom). All defendants denied that Mr. Wilcox worked for the FBI or MPD.

These and other like incidents conceivably could have sufficed to permit a jury to infer that an agreement existed among certain persons within MPD to disrupt plaintiffs’ activities — although not that any defendant knew about or participated in it. When different people shown to be in contact with one another act at different times in a similar and facially illegitimate manner, an inference of prior agreement might be quite plausible. See Halberstam, 705 F.2d at 481 (“[M]utally supportive activity *55by parties in contact with one another over a long period suggests a common plan.”). However, these scattered events surely do not permit an inference that an official policy, custom, ordinance, regulation or decision of the municipality caused such incidents, as is required to hold a municipality liable under the Supreme Court’s decision in Monell v. Department of Social Services, 436 U.S. 658, 98 S.Ct. 2018, 56 L.Ed.2d 611 (1978). There is no indication that the municipality impliedly or tacitly approved, authorized or encouraged such conduct by its employees, see Turpin v. Mailet, 619 F.2d 196, 201 (2d Cir.), cert. denied, 449 U.S. 1016, 101 S.Ct. 577, 66 L.Ed.2d 475 (1980), or that any person of sufficient seniority that his or her “edicts or acts may fairly be said to represent official policy,” Monell, 436 U.S. at 694, 98 S.Ct. at 2037, ordered such activities. Plaintiffs have failed to tie the incidents described above either to an official District policy, to an unofficial governmental “custom,” or to any official with policy-making authority, see Rookard v. Health and Hospitals Corp., 710 F.2d 41, 45 (2d Cir.1983). Indeed, we have found those supervisory MPD officers whom plaintiffs have identified not to have been shown to have participated in any conspiracy. We thus are left with a handful of scattered incidents, occurring over a period of years, aimed at a wide array of persons, which are allegedly attributable to low-level MPD officers and informants. From such scant evidence, neither an inference of official policy nor unofficial custom is warranted.

Similarly, exchanges of information and constant communication between particular persons within MPD and the FBI, coupled with patterns of similar activity on their part, might have sufficed to permit an inference of a joint conspiracy — although, again, not that any defendant knew about or participated in it. After considering several civil conspiracy cases, we recently observed that “since in most cases the court will have to infer a conspiracy from indirect evidence, it must initially look to see if the alleged joint tortfeasors are pursuing the same goal — although performing different functions — and are in contact with one another.” Halberstam, 705 F.2d at 481 (construing Davidson v. Simmons, 203 Neb. 804, 280 N.W.2d 645 (1979), and Peterson v. Cruickshank, 144 Cal.App.2d 148, 300 P.2d 915 (1956)). But, again, even if such activities occurred among some District employees, that fact does not permit an inference that they were taken pursuant to a District policy. We therefore conclude that the evidence simply did not suffice to permit the jury to conclude that the District participated in any conspiracy that might have existed within the MPD, or between certain persons in MPD and the FBI.

The FBI conspiracy need not detain us. There was evidence that each FBI defendant knew about, and actively participated in unlawful COINTELPRO activities. See Discussion section VI(B), infra. The essence of a civil conspiracy is participation in' a common and unlawful plan whose goals are known to all members — even if all parties are not privy to each individual act taken in furtherance of the scheme. The evidence here undoubtedly sufficed to permit the jury to conclude that such a conspiracy existed within the FBI, and that the FBI defendants were participants in it. At the same time, we find insufficient evidence that these defendants participated in any conspiracy with persons in MPD and reverse judgments against them arising out of that alleged joint MPD-FBI scheme.

B. Individual Liability

The foregoing discussion leads inexorably to the conclusion that the individual MPD defendants and the District were entitled to a directed verdict on plaintiffs’ Bivens claims. The dearth of evidence that any one of them took any act with the intent to impede any plaintiff’s exercise of rights protected by the First Amendment precludes a finding of individual liability. There simply is no evidence from which a jury reasonably could infer that any one of them was responsible for deprivation of a plaintiff’s First Amendment rights. In the *56same way, the evidence of disruptive practices in which various MPD informants engaged does not allow an inference that the District, through its policies, acted to disrupt plaintiffs’ rights.

The verdicts against the individual FBI defendants are supported by the evidence. In so saying, we recognize that the record reflects extraordinary variety in the level of participation of each defendant: whereas some defendants actively participated in formulating and carrying out some of the most blatantly wrongful activities, such as the “Give Them Bananas!” leaflet, others are implicated only by inference. While we have no doubt that circumstantial evidence and inference can support a jury verdict, we caution that damage awards must reflect disparities in responsibility among the defendants.

Plaintiffs’ least substantial case against an FBI agent is that against Mr. Pangburn. We have carefully scanned the record and conclude that it supports a finding that he acted to injure plaintiffs. Throughout the relevant period, Mr. Pangburn was involved in WFO investigations of racial extremists. He knew about COINTELPRO, and in particular about COINTELPRO-Black Nationalist, and was well aware that its aim was to devise procedures to make extremist groups less effective. More specifically, defendant Pangburn admitted that he investigated the BUF, see Tr. Trans. (Dec. 10, 1981), Vol. 5, at 425, and Julius Hobson, id. at 435, and that he received information about reaction to a BUF demand for $25,000 from the New Mobe. Id. at 437. Finally, when questioned about an FBI document indicating that WFO was told to instruct its agents in BUF to take opposing positions on the issue of the head tax, he acknowledged both that he probably had informants in BUF at the relevant time and that he would have been involved in instructing them, if they were to be instructed. Id. at 444-45 (testimony regarding Plaintiffs’ Ex. 19, IV J.A. 1830-31). On the basis of the foregoing, we conclude that a jury reasonably might conclude that defendant Pangburn knew about FBI efforts to disrupt plaintiffs and took steps to implement the agency's plan to injure plaintiffs. At a minimum, the jury reasonably could have found that defendant Pangburn acted to exacerbate the head tax issue through his BUF informants. Moreover, the evidence permitted an inference that he knew about other COINTELPRO programs directed against BUF and other Black groups and was positioned to help effectuate them. As to his liability to plaintiffs who were not directly involved in BUF, or other “racial extremist” groups with which Mr. Pangburn was concerned, we agree with Judge Oberdorfer that those verdicts were supported by evidence.155

Each of the other FBI defendants was well positioned to create and effectuate individual COINTELPRO activities, and the evidence supports the conclusion that each took steps to disrupt these plaintiffs’ rights. For example, defendant Jones acknowledged that he was in the position to approve or disapprove suggestions from the WFO, that he approved preparation of The Rational Observer156 and other COINTELPRO activities, and that it was part of his job to make sure that Bureau directives on implementation of both COINTELPROs were carried out. III J.A. 1153-55, 1467-68, 1470. Defendant Grimaldi prepared a memorandum to headquarters seeking permission to acquire WMC housing forms to fill in with fictitious addresses, see IV J.A. *571818-22, reported success in interfering with counterinaugural demonstration marshals’ walkie-talkie radio communications, thereby countermanding the marshals’ orders and disrupting the NMC, see IV J.A. 1823-26, and conceived of and published The Rational Observer. See also II J.A. 968-70, 974-75 (testimony of Mr. Grimaldi). Defendants Brennan and Moore of course knew about the COINTELPROs — an intra-agency memorandum announcing the New Left activities went out over Mr. Brennan’s name, IV J.A. 1742, and a memorandum announcing the Black Nationalist activities went out over Mr. Moore’s name, IV J.A. 1729 — and they remained involved throughout the programs’ tenure. Their positions in headquarters clearly permitted an inference that directives from headquarters either were formulated or approved by them, and supervised by them, and that activities that went forward were not disapproved by them. Moreover, their signatures, initials or names on numerous documents confirm that they knew of, sometimes expressly endorsed, and did not put a halt to activities relevant to plaintiffs. See, e.g., IV J.A. 1848 (memorandum from FBI headquarters giving permission to prepare the “Give Them Bananas!” leaflet, with copy to Mr. Moore, handwritten initials of Mr. Brennan); III J.A. 1198-1200 (testimony of Mr. Brennan regarding The Rational Observer); IV J.A. 1810-17 (memorandum issued over Mr. Moore’s name suggesting covert distribution of press releases to cooperative news sources to curtail fundraising efforts for planned civil rights demonstration); II J.A. 1086-98 (testimony of Mr. Moore on same subject).

The foregoing is hardly an exhaustive list of each defendants’ COINTELPRO activities. We offer this brief sketch simply to demonstrate that the record does not lack evidence from which the jury reasonably could have concluded as it did regarding each FBI defendant. We therefore affirm the remaining judgments against them.

VII. Damages

Our disposition of the liability issues poses an obstacle to the immediate resolution of all damage .awards and requires that we remand this case for further proceedings. In so doing, we observe that this case serves as a pristine example of the extraordinary usefulness of special verdicts, particularly in complex cases. Here the jury responded to particular questions about MPD, MPD-FBI and FBI conspiratorial liability, as well as individual liability, and its specific decisions regarding the conspiratorial and individual liability of the FBI defendants are in no way affected by our conclusion concerning the liability of the District defendants. In marked contrast, the special verdicts requested only generalized findings of damages of each defendant as to each plaintiff.157 We therefore are unable to determine what portion of the damages assessed against the FBI defendants is properly attributable to their individual acts, or their participation in the FBI conspiracy, and what part was levied as a result of their alleged participation in the FBI-MPD conspiracy, of which we have found insufficient evidence to sustain a judgment. It is worthwhile to note that, had the same particularized special verdicts been returned both as to damages and liability, both our task, and that of the parties and the District Court on remand, would have been greatly facilitated. In that in*58stance, we might have been able to limit our review to questions concerning the ex-cessiveness of the damage awards.158 Had we found them to be excessive, we might have remanded for a proposal of remittitur or, if plaintiffs did not consent, for a new trial on damages.159 Without the guidance of special verdicts, our options at this stage are necessarily more limited, and as to all but one plaintiff on one claim, we are able only to remand for a new trial on damages.

The single exception is plaintiff David Eaton, in his claim against Gerald Grimaldi. The jury found that Mr. Grimaldi participated in an FBI conspiracy against Mr. Eaton, and individually injured Mr. Eaton, but that he did not participate in a joint FBI-MPD conspiracy against Mr. Eaton. See I J.A. 379. It concluded that Mr. Gri-maldi was liable to Mr. Eaton for $4,687.50, of which two-thirds, or $3,125.00, represented compensatory damages, and the remainder, $1,562.50, punitive damages. I J.A. 382. As to this one award, we therefore consider the FBI defendants’ claims that the damage judgments were excessive in light of the evidence.

The FBI defendants’ challenge is two-fold: they assert that plaintiffs suffered no actual injury warranting compensation, and that the awards are excessive in any event. Judge Oberdorfer rejected the first argument in his opinion denying defendants’ motion for a new trial but, in apparent agreement with the second argument, wrote that he believed the compensatory damage liability of each defendant should be ten to twenty percent of the amount the jury awarded. Hobson, 556 F.Supp. at 1192. Nevertheless, he concluded that the monetary award was not so unreasonably high as to permit a new trial or a proposal of a remittitur.160 We agree with Judge Oberdorfer that the single compensatory award we review is .excessive *59and conclude, contrary to his view, that it is sufficiently disproportionate to the evidence to warrant reassessment. For one, a verdict that is potentially ten times a reasonable award simply is not supportable. Moreover, the amount is out of proportion to harm that actually has been suffered.161 In consequence, we remand this issue to the District Court for a proposal of a remittitur, see note 159, supra, or, if Mr. Eaton does not consent, for a new trial on damages. The remittitur should reflect the highest amount a jury could reasonably have set, not necessarily the District Court's view of the correct award, and must be based on findings of actual injury, in accordance with principles set out in this opinion. Finally, Judge Oberdorfer found that the record supported the jury’s finding that an award of punitive damages was merited in this case. We find his analysis persuasive as to the award in favor of Mr. Eaton and against Mr. Grimaldi, and uphold the award.

As to the remaining claims, we hasten to add that while we may only remand for a new trial, the parties are not so limited. As we have observed once before,

The parties should consider whether, in the interest of efficiency and justice, First Amendment damages could be set by the trial judge on the record thus far adduced at trial.

Dellums v. Powell, 566 F.2d 167, 196 n. 87 (D.C.Cir.1977). We of course can not require the parties to agree to this resolution. We only wish to point out that much of the considerable amount of evidence and testimony offered at trial to demonstrate liability will no doubt be offered again to demonstrate that both compensatory and punitive damages are appropriate. The result will be the expenditure of considerable time and money simply to recreate a record that is largely before the court already.

In remanding, we are aware of the existence of considerable confusion surrounding the interests that may be compensated in constitutional tort actions. The dearth of recent case law from this Circuit concerning damages for First Amendment violations well might leave the District Court uncertain of the proper principles against which to test a jury’s verdict, to set a remittitur or (if the parties agree as suggested above) to fix damages on its own. In particular, absent clarification of the specific interests that may properly be compensated, it is impossible to determine what the level of compensation should be. To provide some guidance for disposition of this case on remand, and for future cases, we set out the basic analytical framework below.

In Carey v. Piphus, 435 U.S. 247, 98 S.Ct. 1042, 55 L.Ed.2d 252 (1978), the Supreme Court held that substantial compensatory damages for violations of procedural due process may be recovered only for actual injury suffered. The Court rejected the argument that plaintiffs should recover simply for the deprivation of the constitutional right, absent any injury, or that deprivations may be presumed to cause injury. Rather, the Court established that plaintiffs must demonstrate injury to secure more than a nominal recovery, and that “damage awards [must] be limited to sums necessary to compensate plaintiffs for actual harm.”162 At the same time, the Court made clear that plaintiffs may pre*60vail on the merits and recover nominal damages without proof of actual injury.163

Carey involved a due process claim under section 1983, but its rationale — that compensation for injury underlies damage recovery — as well as its result have been extended to other constitutional rights, and to claims not based on section 1983. Thus, in Doe v. District of Columbia, 697 F.2d 1115 (D.C.Cir.1983), in which we applied Carey to a Bivens action challenging prison conditions on the basis of the Eighth Amendment, we observed that the Court “meant to extend the basic ‘compensation principle’ to all constitutional rights, substantive as well as procedural.” Id. at 1123. Doe therefore mandates that we apply the principles set out in Carey to plaintiffs’ Bivens claims for First Amendment violations. Section 1985(3) similarly must be read in light of the principle that damages are designed to compensate persons for actual injuries. Carey expressly cited section 1985(3) as incorporating this principle, 435 U.S. at 256 n. 10, 98 S.Ct. at 1048 n. 10, and it declined to construe section 1983 differently. The language of section 1985(3) indeed supports this principle: it permits a person to recover damages occasioned by the deprivation of constitutional rights.164

Having determined that we must apply the principle of compensation to the case before us, and limit any recovery beyond nominal damages to compensation for actual injury, we must now identify the kinds of injury that might give rise to recovery. Carey instructs us to look first to common law tort rules of recovery for pecuniary and non-pecuniary loss. However, the common law rules serve only as a starting point. Where the common law offers no protection to an interest analogous to that protected by a constitutional right,165 we must adapt those rules to assure adequate compensation:

It is not clear, however, that common-law tort rules of damages will provide a complete solution to the damages issue in every § 1983 case. In some cases, the interests protected by a particular branch of the common law of torts may parallel closely the interests protected by a particular constitutional right. In such cases, it may be appropriate to apply the tort rules of damages directly to the § 1983 action____ In other cases, the interests protected by a particular constitutional right may not also be protected by an analogous branch of the common law of torts____ In those cases, the task will be the more difficult one of adapting common-law rules of damages to provide fair compensation for injuries caused by the deprivation of a constitutional right.

Although this task of adaptation will be one of some delicacy — as this case demonstrates — it must be undertaken. The purpose of § 1983 would be defeated if injuries caused by the deprivation of constitutional rights went uncompensated simply because the common law does not recognize an analogous cause of action.

Id. at 258, 98 S.Ct. at 1049 (citations omitted). Accordingly, the Court admonishes us to tailor rules governing compensation for injuries caused by constitutional depri*61vation to the interests protected by the right in question, in the absence of adequate common law analogs. In so doing, our inquiry of course must not wander from the quantum of injury to protected interests into a consideration of the “inherent value” of the constitutional right.166

We have previously identified several interests protected by common law tort rules that might be relevant to constitutional deprivations. Persons deprived of their rights “are entitled to compensation for any physical injuries, pain and suffering, emotional distress, and impairment of their prospects for future employment proximately caused by the defendants’ unconstitutional conduct.” Doe, 697 F.2d at 1124.167 Injury to reputation also is a compensable interest.168 Without intimating that this list is exclusive, we believe the record before us includes some evidence of injury to three of these interests — emotional distress, reputation and earning capacity. The latter two require little elaboration.169 The former is more complex, however, and merits some discussion.

The conclusory term “emotional distress” defines a jumble of intangible injuries to a person, as is well-evidenced by the facts of this case. Various plaintiffs testified that they suffered some form of “disagreeable emotion" 170 as a result of defendants’ activities. Mr. Pollock, for example, testified171 about feelings of intimidation as a result of FBI interviews and surveillance.172 Furthermore, numerous plaintiffs were placed in stressful situations as a result of the aggravated head tax demand. It is the task of the jury, or the District Court, on remand to assess the testimony on these points and sort out whether each plaintiff testified to some form of distress,173 whether the conduct directed *62against each plaintiff conceivably justified such distress, and whether the evidence permits a conclusion that each defendant in a Bivens action caused such distress. In other words, the factfinder may measure plaintiffs testimony in light of the surrounding circumstances, and in proper circumstances award damages on the basis of plaintiffs testimony.174 In reaching its conclusion, the court or jury may consider, as elements of compensable injury for emotional distress, humiliation and personal indignity,175 emotional pain,176 embarrassment, fear, anxiety and anguish.177

In phase two of its analysis, the District Court must identify the interests protected by the First Amendment and determine whether injuries caused by the constitutional deprivation at issue will go uncompensated on the basis of traditional common law principles. Carey’s admonition that the common law should be modified to provide “fair compensation” suggests that intangible interests must be compensated if they can be conceptualized and if harm can be shown with sufficient certainty to avoid damages based either on pure speculation or the so-called inherent value of the rights violated. These requirements of certainty in valuation, and of actual harm, address Carey’s concern for jury guesswork in valuing priceless constitutional rights, while leaving open the possibility of recovery for genuine injury resulting from deprivation of those rights.

Several courts have identified First Amendment compensable rights. In Del-lums v. Powell, 566 F.2d 167 (D.C.Cir. 1977), we held that the inability to finish speaking in a demonstration merits compensation, because the loss of an opportunity to demonstrate “constitutes loss of First Amendment rights in their most pristine and classic form.” Id. at 195. Similarly, in Kincaid v. Rusk, 670 F.2d 737 (7th Cir. 1982), the Seventh Circuit held that the restriction of an inmate’s access to books would be compensable if the inmate proved that “jail conditions were ‘less desirable’ because; of the restricted access to reading material ... or that he unsuccessfully attempted to gain access to forbidden reading materials.” Id. at 746 (citations omitted). In the same way, we believe competent proof that a plaintiff was deterred from attending a meeting or participating in a demonstration, as Mr. Pollock testified he was, or that associational privacy was violated, would merit “fair compensation.”

Judge Oberdorfer analyzed in depth the evidence to support a damage award for such injury and concluded that it sufficed to demonstrate that plaintiffs were diverted from their protest activities by defendants’ activities. Hobson, 556 F.Supp. at 1190. He pointed to the delay and racial distrust that affected the peace movement following the head tax episode, and to other efforts to exploit differences between the antiwar and civil rights movements. Id. We add to this example the evidence of confusion that followed the FBI’s creation of fictitious housing forms prior to the counterinaugural demonstration in 1969, and of the FBI’s efforts to disrupt a.demonstration by relaying misinformation under the guise of parade marshals. To our minds, such evidence suffices to demonstrate injury to a protected interest — i.e., involuntary diversion from a protected activity. However, we agree with Judge Ob-erdorfer that quantification of the damage arising from such activity is especially difficult; on remand, the court must bear in mind that injury that is not reasonably quantifiable is to be compensated with nominal damages.

We note, however, that many plaintiffs suggested that their rights of association were violated because the FBI’s actions either deterred others from joining their *63movement or made the movement less effective. Proving these harms is extremely difficult, as is assessing the value of the loss. We find in the record only sparse competent evidence that other persons were dissuaded from participation in plaintiffs’ activities. The only concrete example we have found is in an FBI document, from WFO to the FBI Director, outlining the “tangible results” of FBI publication of The Rational Observer. According to the document, “[a]s a result of this publication, sources advised that students on that campus became aware of the nature of SMC and its affiliation with the YSA and partially as a result of this, refused to participate in an SMC planned demonstration on that campus.” 178 Plaintiff Pollock participated in the SMC on campus and conceivably was harmed by this refusal of others to participate. We are skeptical, however, of the ability of a jury, or any person, to place a value on such a loss.179 We need not reach the issue whether a loss of participants is ever compensable. It suffices to conclude that, in the present case, absent additional support, the impact on the SMC merits only nominal damages. Moreover, absent other competent evidence of loss of participants, plaintiffs other than Mr. Pollock may not recover even nominal damages with respect to this injury.

With the foregoing principles in mind, the District Court might find that one or more plaintiffs were entitled as a matter of law only to nominal compensatory damages, because they showed no element of compensatory injury, or because no value could reasonably be placed on the particular injury demonstrated. In that case, Carey instructs that the court may nonetheless award nominal damages for mere deprivation of constitutional rights. In so saying, we intimate no opinion on the proper disposition of this case on remand.

Finally, Carey clearly contemplates that a Court might award punitive damages in a proper case to punish violations of constitutional rights.180 Such an award surely is permissible in either a Bivens or section 1985(3) action when defendants are found to have acted with a malicious intent to deprive plaintiffs of their rights or do them injury, see Carey, 435 U.S. at 257 n. 11, 98 S.Ct. at 1049 n. 11, or with reckless or callous indifference to the federally protected rights. See Smith v. Wade, 461 U.S. 30, 103 S.Ct. 1625, 1640, 75 L.Ed.2d 632 (1983). Such an award may accompany either a nominal compensatory or a substantial compensatory award.

VIII. Arguments on Cross-Appeal

A. Expungement of FBI Records181

Plaintiffs petitioned below for an order requiring destruction of files about them that are in the custody of the FBI. Judge Oberdorfer denied the motion, principally on the ground that an outstanding order of the District Court, in a separate case, precluded the FBI from destroying the *64records. See Hobson v. Wilson, No. 76-1326, slip op. at 6 (D.D.C. July 22, 1982) (citing American Friends Service Committee v. Webster, No. 79-1655 (D.D.C. July 1, 1981) (H. Greene, J.)), reprinted in I J.A. 549. Plaintiffs renew their request on appeal, in response to which the FBI makes four arguments. Because we find none of these arguments persuasive, and believe the case to which Judge Oberdorfer refers does not preclude such an order, we remand to the District Court for reconsideration of plaintiffs’ petition.

Plaintiffs argue, in essence, that the FBI’s continued possession of files pertaining to them is unlawful. They point to the Privacy Act, 5 U.S.C. § 552a(e)(7) (1982), which provides that no agency shall maintain records describing how an individual exercises rights guaranteed by the First Amendment, and to 5 U.S.C. § 552a(e)(l) (1982), which provides that only such information as is relevant and necessary to accomplish a purpose of the agency shall be maintained. Records retained in violation of such prohibitions must be expunged, they assert.182 In addition, they argue that a federal court may order expungement of records when necessary to vindicate First Amendment rights, and that such relief is proper here. The FBI challenges both of these arguments.

First, the FBI argues that the files are necessary for purposes of this litigation. Surely that need does not stand in the way of destruction after all issues are finally resolved. The FBI admits as much,183 and we cannot imagine any court would order expungement, if appropriate, prior to resolution of the case.

Second, the FBI cites to chapters 31 (“Records Management by Federal Agencies”) and 33 (“Disposal of Records”) of Title 44 of the United States Code for the proposition that it may only destroy records pursuant to the scheme set out in Title 44. The FBI asserts that under section 3101 the FBI must maintain records of its investigative activities,184 and that destruction of records thus maintained may occur only under the terms prescribed in chapter 33 of Title 44. Section 3314 provides that procedures set out in chapter 33 are exclusive, “and records of the United States Government may not be alienated or destroyed except under this chapter.” Chapter 33 establishes detailed procedures for promulgation of regulations for compiling lists of records proposed for destruction, for submission of such lists to the General Services Administration by agency heads, and for authorization from the Administrator prior to destruction. Accordingly, the FBI apparently concludes, no court may order the FBI to destroy records except pursuant to this procedure.

The flaw in the FBI’s argument is that this Circuit has expressly held that chapters 31 and 33 must yield to statutory or constitutional rights elsewhere guaranteed:

Were it necessary to protect important statutory or constitutional rights of ap-pellee, expungement in this case would not be prevented, as the Government has argued, by the command of 44 U.S.C. § 3314 (1970) that Government records “may not be alienated or destroyed except under this chapter”. Since it clearly effects no repeal of other provisions, this general statutory command must be reconciled with other statutory requirements, and must bow to them when they are more specific, as of course it must bow to the Constitution.

Chastain v. Kelley, 510 F.2d 1232, 1236 n. 4 (D.C.Cir.1975). Indeed, it is now well-established that an order for expungement of records is, in proper circumstances, a permissible remedy for an agency’s violation of the Privacy Act. Two cases have expressly held this to be true when an agency has violated the Act’s prohibition on maintenance of records describing an individual’s exercise of rights guaranteed by the First Amendment, contained in 5 U.S.C. *65§§ 552a(e)(7) (1982). See Clarkson v. Internal Revenue Service, 678 F.2d 1368, 1376-77 (11th Cir.1982); Albright v. United States, 631 F.2d 915, 921 (D.C.Cir.1980). It is equally well-established that expungement of records is, in proper circumstances, a proper remedy in an action brought directly under the Constitution. See Paton v. La Prade, 524 F.2d 862 (3d Cir.1975); Chastain v. Kelley, 510 F.2d at 1235; cf. Metadure Corp. v. United States, 490 F.Supp. 1368 (S.D.N.Y.1980) (plaintiff may assert both Privacy Act and constitutional claims concerning agency record maintenance). Accordingly, we must reject the defendants’ argument regarding the applicability of Title 44.

Third, the FBI argues that the District Court has enjoined it from destroying or otherwise disposing of its files, and that the few exemptions to the ban against destruction “are not pertinent to the circumstances of this case.” 185 We disagree. In American Friends Service Committee v. Webster, No. 79-1655 (D.D.C. July 1, 1981), aff'd in relevant respects, 720 F.2d 29 (D.C.Cir.1983), Judge Harold Greene issued an order halting the FBI destruction of records on the ground that the FBI and the National Archives and Records Service had failed to carry out their statutory responsibility to develop records disposal schedules, in contravention of various chapters of Title 44. Judge Greene’s order expressly exempted several categories of records, including Privacy Act expungement material, from the injunction. See American Friends Service Committee v. Webster, No. 79-1655, slip op. at 3 n. 3, 11-12 (D.D.C. June 9, 1981), reprinted in I J.A. 511, 513, 521-22. Thus, insofar as plaintiffs’ request for expungement results from a violation of the Privacy Act, the injunction would not be relevant. Moreover, particularly in light of the rationale for the injunction, we certainly do not read Judge Greene’s order as barring courts from ordering the expungement of records for statutory or constitutional violations. For the very same reasons that certain record destruction is permissible outside the scope of chapters 31 and 33 of Title 44 — with which the American Friends injunction is principally concerned — certain court-ordered expungements may be carried out during the pendency of the injunction. It would simply be inconsistent with case law in this Circuit to construe the injunction, which is narrowly directed at FBI compliance with relevant chapters of Title 44, to preclude courts from ordering expungement in circumstances that we have previously identified as not controlled by the scheme set out in Title 44.

Fourth, the FBI argues that to turn over the original files to plaintiffs would nullify claims of privilege asserted in this action. This argument exemplifies a certain confusion, apparent in both parties’ briefs, concerning the remedy plaintiffs seek. To the extent that plaintiffs seek complete ex-pungement from FBI files to prevent use of the records against them at some future date, plaintiffs need not receive copies of the complete records. To the extent they wish to receive copies of the records, the issues concerning the FBI’s ability to destroy them are irrelevant. We believe the latter issue — i.e., plaintiffs’ access to the records — has been resolved with our af-firmance of the District Court’s holding on privilege and need not be reconsidered. The former argument, on which we focus, is in no way affected by the claims of privilege.

In light of the foregoing, we find no merit to defendants’ arguments in support of the District Court’s ruling on expungement, or to the reasons offered by the District Court in its denial. We therefore remand this issue for a hearing on the merits. As we understand the posture of this case, plaintiffs seek expungement of their records under the Constitution. In such a case, “[djetermination of the propriety of an order directing expungement involves a balancing of interests; the harm caused to an individual by the existence of any records must be weighed against the utility to the Government of their maintenance.” Paton v. La Prade, 524 F.2d at 868; see Chastain v. Kelley, 510 F.2d at 1236-37 (District Court should consider “extent to which the information in the Bureau’s files violates appellee’s rights without serving any legitimate needs of the Bureau.”). In conducting such an inquiry, a court must bear in mind the “limited relevance” of cases involving expungement of criminal records, where “the potential prejudicial effect ... far exceed[s] that of the information here at issue.” Chastain, 510 F.2d at 1237. While such an inquiry is well within the equitable power of the court, where necessary to vindicate rights *66secured by the Constitution, it may be that the court can resolve the issue on statutory grounds without reaching the constitutional question. Once before, when confronted with a petition for expungement based on the Constitution, we based our decision ordering expungement on statutory grounds. In Menard v. Saxbe, 498 F.2d 1017 (D.C.Cir.1974), we granted a request for ex-pungement based on arguments that record maintenance violated several of plaintiffs constitutional rights, on the ground that maintenance of the record was beyond the statutory authority of the agency in question. Similarly, in this case it is conceivable that the limitations of the Privacy Act, regarding maintenance of files concerning the exercise of First Amendment rights, see 5 U.S.C. § 552a(e)(7) (1982), will remove the need to decide “whether or to what extent the Constitution forbids the Government from contributing to the harm to the individual that may attend maintenance” of a file such as those at issue here. Menard v. Saxbe, 498 F.2d at 1029. Accordingly, we remand for consideration of whether the FBI’s maintenance of records on plaintiffs violates the Privacy Act, and, if so, whether the violation warrants expungement in whole or part.186 If the Privacy Act does not require the relief sought, then plaintiffs’ claim based on the Constitution must be addressed.

Conclusion

To summarize, our judgments on this appeal are as follows:

(1) we reverse, on statute of limitations grounds, the trial judgments in favor of plaintiffs Booker, Abbott, and Bloom against the FBI defendants;

(2) we reverse, on sufficiency of the evidence grounds, the trial judgments against the individual MPD defendants and the District of Columbia;

(3) we affirm all other findings of liability against the FBI defendants (except those related to the alleged conspiracy between officials at MPD and the FBI);

(4) we remand for reconsideration of the proper amounts due as damages— specifically, in the case of Mr. Grimaldi’s compensatory liability to Mr. Eaton, we remand to the District Court for a proposal of a remittitur or, if Mr. Eaton does not consent, for a new trial limited to the damages issue, and on all other claims we remand for a new trial limited to the damages issue;187

(5) we affirm the award of punitive damages, in favor of Mr. Eaton and against Mr. Grimaldi; and

(6) we remand for reconsideration of plaintiffs’ cross-appeal regarding the expungement of the FBI records.

It is so ordered.

On Petition for Rehearing

On consideration of the petition for rehearing of Appellants Brennan, Moore, Pangburn and Grimaldi, filed July 23, 1984, it is

ORDERED, by the Court, that the petition is denied.

A statement of the Court is attached.

PER CURIAM.

We deny the petition for rehearing in this case. Two points raised by the FBI defendants in their petition, however, merit comment. The FBI defendants’ principal argument is that it is unjust to find mid- and low-level government officials liable for conspiracy when they merely participated in an illegal program approved by higher authorities. In this case, however, the defendants found liable by the jury (and whose liability has been sustained) could only have been found to have participated in the conspiracy, if at all, in a supervisory capacity, exercising substantial responsibility for developing and implementing a policy knowingly designed to thwart plaintiffs’ exercise of their First Amendment rights. We are simply not dealing with federal agents who were acting without knowledge of or responsibility for the illegal objective of a government program, *67and we need not and do not hold that such agents may be held liable under a conspiracy theory.

Second, the FBI defendants argue for the first time that compliance with an agency’s approved policy should be deemed one of the “extraordinary circumstances” (as yet undefined) giving rise to immunity under Harlow v. Fitzgerald, 457 U.S. 800, 819, 102 S.Ct. 2727, 73 L.Ed.2d 396 (1982). In its most extreme form, this argument amounts to the contention that obedience to higher authority should excuse disobedience to law, no matter how central the law is to the preservation of citizens’ rights. We have no hesitation in rejecting this new argument.

In this ease, we had no occasion to consider and we therefore did not address, the contention that in some circumstances compliance with approved policy, such as compliance only after protesting the policy at issue, could excuse liability. Appellants have never raised and, indeed, the record could not support any such contention. We thus leave open the question whether the “extraordinary circumstances” prong of Harlow is satisfied under circumstances not presented by this case.