delivered the opinion of the Court.

This case presents a question of statutory construction: whether 42 U. S. C. §1983 (1976 ed., Supp. V) authorizes a convicted person to assert a claim for damages against a police officer for giving perjured testimony at his criminal trial. The Court of Appeals for the Seventh Circuit held that witnesses are absolutely immune from damages liability based on their testimony, and rejected the petitioners’ contention that government officials who testify about the performance of their official duties may be held liable under § 1983 even if other witnesses may not. We agree with that conclusion.

The Court of Appeals heard argument in three separate cases raising the absolute immunity issue and decided them in a single opinion. Two of these cases are before us on a writ of certiorari. Petitioner Briscoe was convicted in state court of burglarizing a house trailer. He then filed a § 1983 complaint against respondent LaHue, a member of the Bloo-mington, Indiana, police force, alleging that LaHue had violated his constitutional right to due process by committing perjury in the criminal proceedings leading to his conviction.1 *327LaHue had testified that in his opinion Briscoe was one of no more than 50 to 100 people in Bloomington whose prints would match a partial thumbprint on a piece of glass found at the scene of the crime. According to Briscoe, the testimony was false because the Federal Bureau of Investigation and the state police considered the partial print too incomplete to be of value, and without the print there was no evidence identifying him as the burglar. He sought $100,000 in damages. The District Court granted LaHue’s motion for summary judgment on four separate grounds: (1) the facts alleged in the complaint did not suggest that LaHue had testified falsely; (2) allegations of perjury alone are insufficient to state a constitutional claim; (3) LaHue had not testified “under color of law”; and (4) Briscoe’s claim was collaterally estopped by his criminal conviction.

Petitioners Vickers and Ballard were jointly tried and convicted of sexual assault in state court. They subsequently brought a civil action under § 1983 against respondent Hun-ley, a member of the Cedar Lake, Indiana, police force, alleging that he had deprived them of their constitutional rights to due process and a fair trial. They alleged that, by giving false testimony suggesting that they had been able to harmonize their stories before making exculpatory statements to police, he had prejudicially diminished the credibility of those statements. Each plaintiff sought $150,000 in compensatory and $50,000 in punitive damages. The Federal Magistrate granted a motion to dismiss the complaint on alternative grounds: (1) Hunley had not testified “under color of law”; (2) he was entitled to absolute witness immunity; and (3) peti- - tioners had failed to state a claim under § 1983 because they did not allege that the prosecutor had knowingly used false testimony. The District Court affirmed the dismissal on the first ground. Both cases were appealed to the United States Court of Appeals for the Seventh Circuit.2

*328Although other issues were argued in the Court of Appeals, its holding in both cases was predicated squarely on the ground that, in litigation brought under 42 U. S. C. §1983 (1976 ed., Supp. V), all witnesses — police officers as well as lay witnesses — are absolutely immune from civil liability based on their testimony in judicial proceedings. 663 F. 2d 713 (1981).3 Because of the importance of the immunity question, which has given rise to divergent conclusions in the Courts of Appeals,4 we granted certiorari. 455 U. S. 1016 (1982).5

*329Before confronting the precise question that this case presents — whether § 1983 creates a damages remedy against police officers for their testimony as witnesses — we begin by considering the potential liability of lay witnesses on the one hand, and of judges and prosecutors who perform integral functions in judicial proceedings on the other hand. The unavailability of a damages remedy against both of these categories sheds considerable light on petitioners’ claim that Congress intended police officer witnesses to be treated differently.

I

There are two reasons why § 1983 does not allow recovery of damages against a private party for testimony in a judicial proceeding. First, § 1983 does not create a remedy for all conduct that may result in violation of “rights, privileges, or immunities secured by the Constitution and laws.” Its reach is limited to actions taken “under color of any statute, ordinance, regulation, custom, or usage, of any State or Territory . . . .”6 It is beyond question that, when a private *330party gives testimony in open court in a criminal trial, that act is not performed “under color of law.”7

Second, since 1951, when this Court decided Tenney v. Brandhove, 341 U. S. 367, it has been settled that the all-encompassing language of § 1983, referring to “[e]very person” who, under color of law, deprives another of federal constitutional or statutory rights, is not to be taken literally.8

“It is by now well settled that the tort liability created by § 1983 cannot be understood in a historical vacuum. . . . One important assumption underlying the Court’s decisions in this area is that members of the 42d Congress were familiar with common-law principles, including defenses previously recognized in ordinary tort litigation, and that they likely intended these common-law principles to obtain, absent specific provisions to the contrary.” City of Newport v. Fact Concerts, Inc., 453 U. S. 247, 258 (1981).

See Pierson v. Ray, 386 U. S. 547, 554 (1967).

The immunity of parties and witnesses from subsequent damages liability for their testimony in judicial proceedings9 *331was well established in English common law. Cutler v. Dixon, 4 Co. Rep. 14b, 76 Eng. Rep. 886 (Q. B. 1585); Anfield v. Feverhill, 2 Bulst. 269, 80 Eng. Rep. 1113 (K. B. 1614); Henderson v. Broomhead, 4 H. & N. 569, 578, 157 Eng. Rep. 964, 968 (Ex. 1859);10 see Dawkins v. Lord Rokeby, 4 F. & F. 806, 833-834, 176 Eng. Rep. 800, 812 (C. P. 1866). Some American decisions required a showing that the witness’ allegedly defamatory statements were relevant to the judicial proceeding, but once this threshold showing had been made, the witness had an absolute privilege.11 The *332plaintiff could not recover even if the witness knew the statements were false and made them with malice.12

In the words of one 19th-century court, in damages suits against witnesses, “the claims of the individual must yield to *333the dictates of public policy, which requires that the paths which lead to the ascertainment of truth should be left as free and unobstructed as possible.” Calkins v. Sumner, 13 Wis. 193, 197 (1860). A witness’ apprehension of subsequent damages liability might induce two forms of self-censorship. First, witnesses might be reluctant to come forward to testify. See Henderson v. Broomhead, supra, at 578-579, 157 Eng. Rep., at 968. And once a witness is on the stand, his testimony might be distorted by the fear of subsequent liability. See Barnes v. McCrate, 32 Me. 442, 446-447 (1851). Even within the constraints of the witness’ oath there may be various ways to give an account or to state an opinion. These alternatives may be more or less detailed and may differ in emphasis and certainty. A witness who knows that he might be forced to defend a subsequent lawsuit, and perhaps to pay damages, might be inclined to shade his testimony in favor of the potential plaintiff, to magnify uncertainties, and thus to deprive the finder of fact of candid, objective, and undistorted evidence. See Veeder, Absolute Immunity in Defamation: Judicial Proceedings, 9 Colum. L. Rev. 463, 470 (1909).13 But the truthfinding process is better *334served if the witness’ testimony is submitted to “the crucible of the judicial process so that the factfinder may consider it, after cross-examination, together with the other evidence in the case to determine where the truth lies.” Imbler v. Pachtman, 424 U. S. 409, 440 (1976) (White, J., concurring in judgment).14

At least with respect to private witnesses, it is clear that §1983 did not abrogate the absolute immunity existing at common law, and petitioners do not contend otherwise. Like the immunity for legislators at issue in Tenney v. Brandhove, the common law’s protection for witnesses is “a tradition so well grounded in history and reason” that we cannot believe that Congress impinged on it “by covert inclusion in the general language before us.” 341 U. S., at 376.

II

The Court has already addressed the question whether § 1983 permits damages recoveries from judges, prosecutors, and other persons acting “under color of law” who perform official functions in the judicial process. Again, we have found that, in light of common-law immunity principles, § 1983 did not impose liability on these officials. We have held that state judges are absolutely immune from liability for their judicial acts, Pierson v. Ray, 386 U. S. 547 (1967); Stump v. Sparkman, 435 U. S. 349 (1978), and that state prosecutors have absolute immunity from liability for their actions in initiating prosecutions, Imbler v. Pachtman, supra.

The central focus of our analysis has been the nature of the judicial proceeding itself. Thus, in his opinion concurring in the judgment in Imbler v. Pachtman, supra, Justice White explained that the absolute immunity of public prosecutors was “based on the policy of protecting the judicial process.” *335424 U. S., at 439. He explained that this protection extended equally to other participants, including counsel and witnesses.

“The reasons for this rule are also substantial. It is precisely the function of a judicial proceeding to determine where the truth lies. The ability of courts, under carefully developed procedures, to separate truth from falsity, and the importance of accurately resolving factual disputes in criminal (and civil) cases are such that those involved in judicial proceedings should be ‘given every encouragement to make a full disclosure of all pertinent information within their knowledge.’” Ibid.

The common law’s protection for judges and prosecutors formed part of a “cluster of immunities protecting the various participants in judge-supervised trials,” which stemmed “from the characteristics of the judicial process.” Butz v. Economou, 438 U. S. 478, 512 (1978); cf. King v. Skinner, Lofft 54, 56, 98 Eng. Rep. 529 (K. B. 1772) (“[NJeither party, witness, counsel, jury, or judge can be put to answer, civilly or criminally, for words spoken in office”). The common law recognized that

“controversies sufficiently intense to erupt in litigation are not easily capped by a judicial decree. The loser in one forum will frequently seek another .... Absolute immunity is thus necessary to assure that judges, advocates, and witnesses can perform their respective functions without harassment or intimidation.” Butz, supra, at 512.

In short, the common law provided absolute immunity from subsequent damages liability for all persons — governmental or otherwise — who were integral parts of the judicial process. It is equally clear that § 1983 does not authorize a damages claim against private witnesses on the one hand, or against judges or prosecutors in the performance of their respective duties on the other. When a police officer appears as a witness, he may reasonably be viewed as acting like any *336other witness sworn to tell the truth — in which event he can make a strong claim to witness immunity;16 alternatively, he may be regarded as an official performing a critical role in the judicial process, in which event he may seek the benefit afforded to other governmental participants in the same proceeding. Nothing in the language of the statute suggests that such a witness belongs in a narrow, special category lacking protection against damages suits. We must ask, however, whether anything in the legislative history of § 1983 points to a different conclusion.

I — I HH h-H

Petitioners point to a number of references throughout the debates on the 1871 Act to widespread perjury by Ku Klux Klan witnesses in state criminal trials.15 They urge that, because perjury was one of the specific evils with which Congress was concerned, recognizing an absolute immunity for witnesses would conflict with congressional intent. We find this argument unpersuasive. The Act consisted of several sections establishing different remedies for disorder and violence in the Southern States.17 The legislative history and statutory language indicate that Congress intended perjury *337leading to unjust acquittals of Klan conspirators to be prohibited by §2, the civil and criminal conspiracy section of the statute, now codified in relevant part at 42 U. S. C. § 1985(3) (1976 ed., Supp. V) and 18 U. S. C. §241. But the language of §1 — now codified as §1983 — differs from that of §2 in essential respects, and we find no evidence that Congress intended to abrogate the traditional common-law witness immunity in §1983 actions.

The Ku Klux Act, 17 Stat. 13, was enacted on April 20, 1871, less than a month after President Grant sent a dramatic message to Congress describing the breakdown of law and order in the Southern States. Cong. Globe, 42d Cong., 1st Sess., 236, 244 (1871). During the debates, supporters of the bill repeatedly described the reign of terror imposed by the Klan upon black citizens and their white sympathizers in the Southern States. Hours of oratory were devoted to the details of Klan outrages — arson, robbery, whippings, shootings, murders, and other forms of violence and intimidation— often committed in disguise and under cover of night. These acts of lawlessness went unpunished, legislators asserted, because Klan members and sympathizers controlled or influenced the administration of state criminal justice. In particular, it was alleged that Klan members were obligated, by virtue of membership in the organization, to protect fellow members who were charged with criminal activity. They had a duty to offer themselves for service on grand and petit juries, and to violate their jurors’ oaths by refusing to indict or to convict regardless of the strength of the evidence. They also were bound to appear as witnesses, and again to violate their oaths by committing perjury, if necessary, to exculpate their Klan colleagues.18 Perjury was thus one of the *338means by which the Klan prevented state courts from gaining convictions of Klan members for crimes against blacks and Republicans.

It is clear from the legislative debates that, in the view of the Act’s sponsors, the victims of Klan outrages were deprived of “equal protection of the laws” if the perpetrators systematically went unpunished.19 Proponents of the measure repeatedly argued that, given the ineffectiveness of state law enforcement and the individual’s federal right to “equal protection of the laws,” an independent federal remedy was necessary and Congress had the power to provide it.20 See Monroe v. Pape, 365 U. S. 167, 174 (1961).

Section 2 was designed specifically to provide criminal and civil remedies in federal court for the conspiratorial activities of the Klan. Indeed the provision singles out those who “go in disguise upon the public highway.” Earlier versions of the section enumerated precisely the activities that had been attributed to the Klan — murder, manslaughter, mayhem, robbery, assault and battery, perjury, subornation of perjury, criminal obstruction of legal process or resistance of of-*339fleers in discharge of official duty, arson, or larceny. Cong. Globe, supra, at 317. The more general language in the final version of § 2 was also intended to apply to the abuses that had been described repeatedly in congressional debate.21 Part of the provision is particularly well tailored to reach conspiracies to commit perjury in order to prevent punishment of fellow Klansmen. It provides penalties whenever two or more persons shall

“conspire together ... for the purpose of preventing or hindering the constituted authorities of any State from giving or securing to all persons within such State the equal protection of the laws, or shall conspire together for the purpose of in any manner impeding, hindering, obstructing, or defeating the due course of justice in any State or Territory, with intent to deny to any citizen of the United States the due and equal protection of the laws . . . ,”22

This evidence does not, however, tend to show that Congress intended to abrogate witness immunity in civil actions under § 1, which applied to wrongs committed “under color of . . . law.” The bill’s proponents were exclusively concerned with perjury resulting in unjust acquittals — perjury likely to be committed by private parties acting in furtherance of a conspiracy — and not with perjury committed “under color of *340law” that might lead to unjust convictions. In hundreds of pages of debate there is no reference to the type of alleged constitutional deprivation at issue in this case: perjury by a government official leading to an unjust conviction. Indeed, the legislative history is virtually silent even with regard to perjury by private persons leading to convictions of innocent defendants.23 There is a simple enough reason for this lacuna: the Klan had other, more direct, means of dealing with its victims. A “reign of terrorism and bloodshed” did not require the formal processes of law; at most, drumhead tribunals were convened at dead of night.24 Even when the organization’s intended victims had been taken into custody and charged with crimes, the evidence before Congress suggested that the Klan resorted to vigilante justice rather than courtroom perjury.25

In summary, the legislative history supports criminal punishment under §2 for a witness who conspired to give perjured testimony favorable to a defendant, with the effect of preventing effective enforcement of the laws, and liability in a civil suit against the perjured witness by the defendant’s victim. But these are not the issues before us today. We are asked to extrapolate from pro-defendant perjury to pro-prosecution perjury, and if willing to make that step, we are further invited to apply legislative history relating to § 2 — a section specifically directed toward private conspiracies — to § 1 — a section designed to provide remedies for abuses under *341color of law. We decline the invitation. The debates of the 42d Congress do not support petitioners’ contention that Congress intended to provide a § 1 damages remedy against police officers or any other witnesses.26

I — I C

Petitioners, finally, urge that we should carve out an exception to the general rule of immunity in cases of alleged perjury by police officer witnesses.27 They assert that the reasons supporting common-law immunity — the need to *342avoid intimidation and self-censorship — apply with diminished force to police officers. Policemen often have a duty to testify about the products of their investigations, and they have a professional interest in obtaining convictions which would assertedly counterbalance any tendency to shade testimony in favor of potentially vindictive defendants. In addition, they are subject to § 1983 lawsuits for the performance of their other duties, as to which they have only qualified immunity, and their defense is generally undertaken by their governmental employers. Further, petitioners urge that perjured testimony by police officers is likely to be more damaging to constitutional rights than such testimony by ordinary citizens, because the policeman in uniform carries special credibility in the eyes of jurors. And, in the case of police officers, who cooperate regularly with prosecutors in the enforcement of criminal law, prosecution for perjury is alleged to be so unlikely that it is not an effective substitute for civil damages.

These contentions have some force. But our cases clearly indicate that immunity analysis rests on functional categories, not on the status of the defendant.28 A police officer on the witness stand performs the same functions as any other witness; he is subject to compulsory process, takes an oath, responds to questions on direct examination and cross-examination, and may be prosecuted subsequently for perjury.

Moreover, to the extent that traditional reasons for witness immunity are less applicable to governmental witnesses, *343other considerations of public policy support absolute immunity more emphatically for such persons than for ordinary witnesses. Subjecting government officials, such as police officers, to damages liability under § 1983 for their testimony might undermine not only their contribution to the judicial process but also the effective performance of their other public duties.

Section 1983 lawsuits against police officer witnesses, like lawsuits against prosecutors, “could be expected with some frequency.” Cf. Imbler v. Pachtman, 424 U. S., at 425. Police officers testify in scores of cases every year, and defendants often will transform resentment at being convicted into allegations of perjury by the State’s official witnesses. As the files in this case show, even the processing of a complaint that is dismissed before trial consumes a considerable amount of time and resources.29

This category of § 1983 litigation might well impose significant burdens on the judicial system and on law enforcement resources. As this Court noted when it recognized absolute immunity for prosecutors in Imbler, if the defendant official “could be made to answer in court each time [a disgruntled defendant] charged him with wrongdoing, his energy and at*344tention would be diverted from the pressing duty of enforcing the criminal law.” 424 U. S., at 425. To some degree the individual’s burden might be alleviated by the government’s provision of counsel, but a case that goes to trial always imposes significant emotional and other costs on every party litigant.

It is not sufficient to assert that the burdens on defendants and the courts could be alleviated by limiting the cause of action to those former criminal defendants who have already vindicated themselves in another forum, either on appeal or by collateral attack. We rejected a similar contention in Imbler. Petitioner contended that “his suit should be allowed, even if others would not be, because the District Court’s issuance of the writ of habeas corpus shows that his suit has substance.” Id., at 428, n. 27. We declined to carve out such an exception to prosecutorial immunity, noting that petitioner’s success in a collateral proceeding did not necessarily establish the merits of his civil rights action. Moreover, we noted that “using the habeas proceeding as a ‘door-opener’ for a subsequent civil rights action would create the risk of injecting extraneous concerns into that proceeding.” Ibid. We emphasized that, in determining whether to grant postconviction relief, the tribunal should focus solely on whether there was a fair trial under law. “This focus should not be blurred by even the subconscious knowledge that a post-trial decision in favor of the accused might result in the prosecutor’s being called upon to respond in damages for his error or mistaken judgment.” Id., at 427. The same danger exists in the case of potential liability for police officer witnesses.30

*345There is, of course, the possibility that, despite the truth-finding safeguards of the judicial process, some defendants might indeed be unjustly convicted on the basis of knowingly false testimony by police officers.31 The absolute immunity for prosecutors recognized in Imbler bars one possible avenue of redress for such defendants. Similarly, in this case, the absolute witness immunity bars another possible path to recovery for these defendants. But we have recognized, again and again, that in some situations, the alternative of limiting the official’s immunity would disserve the broader public interest. As Judge Learned Hand wrote years ago:

“As is so often the case, the answer must be found in a balance between the evils inevitable in either alternative. In this instance it has been thought in the end better to leave unredressed the wrongs done by dishonest officers than to subject those who try to do their duty to the constant dread of retaliation.” Gregoire v. Biddle, 177 F. 2d 579, 581 (CA2 1949), cert. denied, 339 U. S. 949 (1950).32

In short, the rationale of our prior absolute immunity cases governs the disposition of this case. In 1871, common-law immunity for witnesses was well settled. The principles set forth in Pierson v. Ray to protect judges and in Imbler v. Pachtman to protect prosecutors also apply to witnesses, who perform a somewhat different function in the trial process but whose participation in bringing the litigation to a *346just — or possibly unjust — conclusion is equally indispensable. The decision of the Court of Appeals is affirmed.

It is so ordered.

with whom Justice Blackmun joins, except as to Part I, dissenting.

I cannot agree that police officers are absolutely immune from civil liability under 42 U. S. C. § 1983 (1976 ed., Supp. V) for testimony given in criminal proceedings. The extension of absolute immunity conflicts fundamentally with the language and purpose of the statute. I would therefore be reluctant in any case to conclude that §1983 incorporates common-law tort immunities that may have existed when Congress enacted the statute in 1871. But in this case the conclusion is especially unjustified. First, absolute immunity for witnesses was by no means a settled legal proposition in 1871. Most notably, in 1845 this Court had cast serious doubt on the existence of absolute immunity for testimony given in judicial proceedings. Second, the origins and history of § 1983 strongly suggest that Congress meant to abrogate any absolute immunity for government officials involved *347in the judicial process, including police officers. Finally, considerations of public policy deemed necessary to justify absolute immunity in our past cases do not support an absolute immunity for officer-witnesses.

The majority opinion correctly states that this case presents a question of statutory construction. Ante, at 326. Yet it departs from generally accepted principles for interpreting laws.

In all other matters of statutory construction, this Court begins by focusing on the language of the statute itself.1 “Absent a clearly expressed legislative intention to the contrary, that language must ordinarily be regarded as conclusive.” Consumer Product Safety Comm’n v. GTE Sylvania, Inc., 447 U. S. 102, 108 (1980). The language of § 1983 provides unambiguous guidance in this case. A witness is most assuredly a “person,” the word Congress employed to describe those whose conduct § 1983 encompasses.2 The ma*348jority turns the conventional approach to statutory interpretation on its head. It assumes that common-law tort immunities provide an exemption from the plain language of the statute unless petitioners demonstrate that Congress meant to override the immunity. See ante, at 336. Thus, in the absence of a clearly expressed legislative intent to the contrary, the Court simply presumes that Congress did not mean what it said.

Absolute immunity for witnesses conflicts not only with the language of § 1983 but also with its purpose. In enacting § 1983, Congress sought to create a damages action for victims of violations of federal rights; absolute immunity nullifies “pro tanto the very remedy it appears Congress sought to create.” Imbler v. Pachtman, 424 U. S. 409, 434 (1976) (White, J., concurring in judgment). The words of a statute should always be interpreted to carry out its purpose.3 Moreover, Members of the 42d Congress explicitly stated that § 1983 should be read so as to further its broad remedial goals. As the sponsor of the 1871 Act, Representative Shellabarger, declared:

“This act is remedial, and in aid of the preservation of human liberty and human rights. All statutes and constitutional provisions authorizing such statutes are liberally and beneficently construed. It would be most strange and, in civilized law, monstrous were this not the rule of interpretation. As has been again and again decided by your own Supreme Court of the United States, and everywhere else where there is wise judicial *349interpretation, the largest latitude consistent with the words employed is uniformly given in construing such statutes and constitutional provisions as are meant to protect and defend and give remedies for their wrongs to all the people.” Cong. Globe, 42d Cong., 1st Sess., App. 68 (1871).4

It might be appropriate to import common-law defenses and immunities into the statute if, in enacting § 1983, Congress had merely sought to federalize state tort law. But Congress “intended to give a broad remedy for violations of federally protected civil rights.” Monell v. New York City Dept. of Social Services, 436 U. S. 658, 685 (1978) (emphasis added). Different considerations surely apply when a suit is based on a federally guaranteed right — in this case, the constitutional right to due process of law — rather than the common law.5 The Congress that enacted § 1983 had concluded that “a deprivation of a constitutional right is significantly different from and more serious than a violation of a state right and therefore deserves a different remedy even though *350the same act may constitute both a state tort and the deprivation of a constitutional right.” Monroe v. Pape, 365 U. S. 167, 196 (1961) (Harlan, J., concurring). Therefore, immunities that arose in the context of tort actions against private parties provide little guidance for actions against state officials for constitutional violations. “It would indeed be the purest coincidence if the state remedies for violations of common-law rights by private citizens were fully appropriate to redress those injuries which only a state official can cause and against which the Constitution provides protection.” Id., at 196, n. 5.

Given the language and purpose of § 1983, I have serious doubts about any further extension of absolute immunity to state officials in actions under § 1983. At a minimum, I do not believe the Court should extend absolute immunity to state officials “in the absence of the most convincing showing that the immunity is necessary.” Imbler v. Pachtman, supra, at 434 (White, J., concurring in judgment). For the reasons elaborated below, I believe that the case for absolute witness immunity is far from convincing.

HH hH

The majority’s decision is predicated on its conclusion that “[i]n 1871, common-law immunity for witnesses was well settled.” Ante, at 345. I disagree with this view of the law as it stood when Congress enacted § 1983.

To begin with, some of petitioners’ allegations would clearly not have been barred by doctrines of immunity at common law. The majority discusses only the immunities associated with actions for defamation at common law. Ante, at 330-331, n. 9. However, petitioner Briscoe did not allege solely that Officer LaHue had testified falsely at his trial, a claim resembling one for defamation. He also alleged that Officer LaHue had made knowingly false charges at two probable-cause hearings, one of which resulted in Briscoe’s arrest.6 At common law, such an allegation would have *351formed the basis of an action on the case for malicious prosecution,7 or the related action known by its Latin name, crimen feloniae imposuit (imputing the crime of felony).8 Both English and American courts routinely permitted plaintiffs to bring actions alleging that the defendant had made a false and malicious accusation of a felony to a magistrate or other judicial officer.9 No immunity barred these suits. Indeed, an absolute immunity would have been illogical, for it would have allowed a defendant to avoid the related common-law action for false imprisonment by the simple expedient of proffering false charges before a magistrate and thereby securing an arrest warrant.10

*352Even with respect to the common-law action for defamation which the majority discusses, I cannot agree that an absolute immunity for witnesses was well-settled law in 1871. In 1845, this Court had rejected both the rule of absolute im*353munity and its logical underpinnings, proposing instead that a plaintiff allege and prove malice in the case of privileged communications.

In White v. Nicholls, 3 How. 266 (1845), Justice Daniel wrote for a unanimous Court in dicta a veritable treatise on the law of defamation and privileged communications.11 The Court began by noting the existence of various exceptions “which, in the elementary treatises, and in the decisions upon libel and slander, have been denominated privileged communications or publications.” Id., at 286. One of these “exceptions” was for “[w]ords used in the course of a legal or judicial proceeding, however hard they may bear upon the party of whom they are used.” Id., at 287. The Court then stated:

“But the term ‘exceptions,’ as applied to cases like those just enumerated, could never be interpreted to mean that there is a class of actors or transactions placed above the cognisance of the law, absolved from the commands of justice. It is difficult to conceive how, in society where rights and duties are relative and mutual, there can be tolerated those who are privileged to do injury legibus soluti; and still more difficult to imagine, how such a privilege could be instituted or tolerated upon the principles of social good. The privilege spoken of in the books should, in our opinion, be taken with strong and well-defined qualifications. It properly signifies this, and nothing more. That the excepted instances shall so far change the ordinary rule with respect to slanderous or libellous matter, as to remove the regular and usual presumption of malice, and to make it incumbent on the party complaining to show malice, either by the construction of the spoken or written matter, or by the facts and circumstances connected with that matter, or with the situation of the parties, adequate to authorize the conclusion.” Ibid, (emphasis added).

*354The Court invoked these principles in discussing the specific exception for words used in a judicial proceeding, relying on the views of one English judge who had rejected absolute immunity.12

“With respect to words used in a course of judicial proceeding, it has been ruled that they are protected by the occasion, and cannot form the foundation of an action of slander without proof of express malice; ... in the case of Hodgson v. Scarlett, 1 Barn. & Aid. 247, it is said by Holroyd, J., speaking of the words of counsel in the argument of a cause, Tf they be fair comments upon the evidence, and relevant to the matter in issue, then unless malice be shown, the occasion justifies them. If, however, it be proved that they were not spoken bona fide, or express malice be shown, then they may be actionable.’” Id., at 288 (emphasis added).13

If Congress in 1871 actually examined the subject of common-law witness immunity, it could hardly have overlooked White v. Nicholls since that case was the sole pronouncement on the subject from the highest Court in the land. Congress might well have concluded — as did the Tennessee Supreme Court in 1871 — that the principles enunciated in White were “settled law.” Saunders v. Baxter, 53 Tenn. 369, 383. In an age when federal common law prevailed, see Swift v. Tyson, 16 Pet. 1 (1842), a Supreme Court decision would have been the natural focus for a Congress establishing a federal remedy which was accompanied by a new grant of federal jurisdiction.14 In short, the most salient feature in the *355landscape of the common law at the time Congress acted was an opinion rebuffing absolute immunity in favor of a qualified immunity based on the absence of malice.15

*356The majority’s decision is also predicated on its conclusion that there is “no evidence that Congress intended to abrogate the traditional common-law witness immunity in § 1983 actions.” Ante, at 337. In fact, there is considerable evidence in the legislative history that Congress did intend to abrogate the immunity of participants in state judicial proceedings.

A

At petitioners’ urging,16 the Court has extensively examined the legislative history of § 2 of the 1871 Ku Klux Klan Act, 17 Stat. 13, now codified as 42 U. S. C. § 1985(3) (1976 ed., Supp. V). However, the forerunner of § 1983 was § 1 of the 1871 Act, not §2. As the majority points out, ante, at 337, 340-341, the two sections differ significantly in their language and purpose. It is thus hardly surprising that the debates over § 2 shed little light on § 1. In my view the inquiry should focus on the history of § 1. Only by examining the *357genesis of that provision can it be determined whether Congress intended to abrogate certain common-law immunities.

The origin of § 1 is not open to serious question. The language and concept of the provision were derived in large part from §2 of the Civil Rights Act of 1866, 14 Stat. 27.17 The author of § 1 clearly stated the relationship between the two Acts in introducing the 1871 measure:

“My first inquiry is as to the warrant which we have for enacting such a section as this [§ 1 of the 1871 Act]. The model for it will be found in the second section of the act of April 9,1866, known as the ‘civil rights act.’ That section provides a criminal proceeding in identically the same case as this one provides a civil remedy for, except that the deprivation under color of State law must, under the civil rights act, have been on account of race, color, or former slavery. This section of the bill, on the same state of facts, not only provides a civil remedy for *358persons whose former condition may have been that of slaves, but also to all people where, under color of State law, they or any of them may be deprived of rights to which they are entitled under the Constitution by reason and virtue of their national citizenship.” Cong. Globe, 42d Cong., 1st Sess., App. 68 (1871) (emphasis added).

Because the two provisions are so intimately connected, a full examination of the history of § 1 of the 1871 Act must begin with § 2 of the 1866 Act.

B

The Civil Rights Act of 1866 was the first federal statute to provide broad protection in the field of civil rights. Its primary purpose was to guarantee the newly emancipated Negro equality with whites before the law. Section 2 of the Act provided criminal liability for any person who, acting under color of law, deprived another of his rights because of race. This provision was extensively debated. Controversy centered in large part over its intended application to state officials integral to the judicial process.

The liability of state judicial officials and all official participants in state judicial proceedings under §2 was explicitly and repeatedly affirmed.18 The notion of immunity for such officials was thoroughly discredited. The Senate sponsor of *359the Act deemed the idea “akin to the maxim of the English law that the King can do no wrong. It places officials above the law. It is the very doctrine out of which the rebellion [the Civil War] was hatched.” Cong. Globe, 39th Cong., 1st Sess., 1758 (1866) (Sen. Trumbull). Thus, §2 was “aimed directly at the State judiciary.” Id., at 1155 (Rep. Eldridge). See also id., at 1778 (Sen. Johnson, member of the Senate Judiciary Committee) (§ 2 of the 1866 Act “strikes at the judicial department of the governments of the States”).

Two unsuccessful efforts were made to amend § 2. First, Representative Miller introduced an amendment to exempt state judges from criminal liability. Id., at 1156. Second, and of particular significance, Representative Bingham introduced an amendment to substitute a civil action for the criminal sanctions contained in the proposal. Id., at 1266, 1271-1272. The sponsor of the 1866 Act, Representative Wilson, opposed the amendment largely on the ground that it would place the financial burden of protecting civil rights on poor individuals instead of on the government. Id., at 1295. At the same time he stressed that there was “no difference in the principle involved” between a civil remedy and a criminal sanction. Ibid.

After the 1866 bill passed the Senate and House, President Andrew Johnson vetoed it. His opposition was based in part on the fact that § 2 of the bill “invades the judicial power of the State.” Veto Message, in id., at 1680. The President warned that “judges of the State courts . . . [and] marshals and sheriffs, who should, as ministerial officers, execute processes, sanctioned by State laws and issued by State judges, in execution of their judgments, could be brought before other tribunals and there subjected to fine and imprisonment for the performance of the duties which such state laws might impose.” Ibid. Within two weeks, both the Senate and the House overrode the veto. Various Congressmen responded to the President’s criticisms and freely admitted that § 2 of the legislation was aimed at state judicial systems. As a member of the House Judiciary Committee, Representative *360Lawrence, declared: “I answer it is better to invade the judicial power of the State than permit it to invade, strike down, and destroy the civil rights of citizens. A judicial power perverted to such uses should be speedily invaded. The grievance would be insignificant.” Id., at 1837. See also id., at 1758 (response of Sen. Trumbull to President’s veto message); id., at 1838 (statement of Rep. Clarke). The bill became law on April 9, 1866.

C

This Court has from time to time read § 1983 against the “background” of common-law tort liability.19 Far more pertinent to this case, however, is the background provided by the 1866 Civil Rights Act. Representative Bingham, who had introduced the amendment to substitute civil liability for criminal liability in the 1866 Act, had become chairman of the House Judiciary Committee by the time of the 42d Congress. Senator Trumbull, the Senate sponsor of the 1866 Act, was chairman of the Senate Judiciary Committee in 1871. Representative Shellabarger, who had participated in the debates on the 1866 legislation,20 drafted the 1871 Act.

Congress was well aware that the “model” for § 1 of the 1871 law could be found in the 1866 Civil Rights Act. Cong. Globe, 42d Cong., 1st Sess., App. 68 (1871) (Rep. Shella-barger). The manager of the bill in the Senate, George Ed-munds, stressed that § 1 was merely “carrying out the principles of the civil rights bill” that had been passed in 1866. Id., at 568. Representative Coburn stated that § 1 “gives a civil remedy parallel to the penal provision” in the Civil Rights Act. “If this penal section is valid, and no one dares controvert it, the civil remedy is legal and unquestionable.” Id., at 461. See also id., at 429 (Rep. McHenry in opposi*361tion) (“The first section of the bill is intended as an amendment of the civil rights act”); id., at 365 (Rep. Arthur in opposition) (§ 1 is “cumulative, as far as it goes, with certain provisions in the civil rights bill”).

The fact that § 2 of the Civil Rights Act was the model for § 1 of the 1871 Act explains why the debates in the 42d Congress on § 1 were so perfunctory.21 Of all the measures in the Ku Klux Klan Act, § 1 generated the least controversy since it merely provided a civil counterpart to the far more controversial criminal provision in the 1866 Act. See id., at 568 (Sen. Edmunds) (“The first section is one that I believe nobody objects to”); id., at App. 313 (Rep. Burchard) (“To the first section, giving an injured party redress by suit at law in the United States courts in the cases enumerated, I can see no objections”); Monell v. New York City Dept. of Social Services, 436 U. S., at 665 (debate on § 1 was limited and the section passed without amendment); Developments in the Law — Section 1983 and Federalism, 90 Harv. L. Rev. 1133, 1155 (1977).

Opponents of § 1 of the 1871 Act repeated the same arguments that had been made against § 2 of the 1866 Act. They warned of the liability for judicial officers that would result from enactment of § 1.22 Indeed, in portraying the inevitable consequences of the 1871 Act, Senator Thurman pointed to criminal prosecutions of state judicial officers that had already taken place under the 1866 Act.23 These statements can hardly be dismissed as exaggerated rhetoric from opponents of the 1871 Act. Instead, they simply reflect the fact that the battle over liability for those integral to the judicial process had already been fought in 1866 when Congress *362adopted the far more serious criminal sanction aimed at state judicial systems. Section 1, in contrast, provided for “the mild remedy of a civil action.” Cong. Globe, 42d Cong., 1st Sess., 482 (1871) (Rep. Wilson, member of the House Judiciary Committee). So it was not surprising that the arguments of the opponents to the 1871 Act would fall on deaf ears. It is also noteworthy that Representative Shella-barger, who was hardly reluctant to interrupt speakers who were misconstruing his proposal,24 never disputed the opponents’ characterizations with regard to the liability of state judicial officers.25

To assume that Congress, which had enacted a criminal sanction directed against state judicial officials,26 intended sub silentio to exempt those same officials from the civil counterpart approaches the incredible.27 Sheriffs and marshals, while performing a quintessential^ judicial function such as serving process, were clearly liable under the 1866 Act, notwithstanding President Johnson’s objections. Be*363cause, as Representative Shellabarger stated, § 1 of the 1871 Act provided a civil remedy “in identically the same case” or “on the same state of facts” as § 2 of the 1866 Act, it obviously overrode whatever immunity may have existed at common law for these participants in the judicial process in 1871.

D

The lack of historical support for witness immunity sharply contrasts with the substantial historical support for legislative immunity which this Court recognized in Tenney v. Brandhove, 341 U. S. 367 (1951), a case on which the majority relies. Ante, at 330, 334. Legislative immunity enjoyed a unique historical position since it had been conceived in the Parliamentary struggles of the 17th century and enshrined in the Speech and Debate Clause of the Constitution. The vast majority of States had adopted constitutional provisions providing a parallel protection against civil and criminal liability. See 341 U. S., at 372-375.

Moreover, the history of § 1 supports incorporation of legislative immunity. For example, when the specter of holding state legislators liable under § 2 of the 1866 Act was raised by President Johnson’s veto message,28 the Senate sponsor of the Act was quick to disavow any such intention. Senator Trumbull argued at some length that legislators did not fall within the scope of the Act because they “enact” laws rather than act “under color of” state law.29 Whatever the validity of this distinction, it no doubt reflected the reluctance of Congress to impinge on the immunity of state legislators. But while the Radical Republican Congress was a “staunch advocate of legislative freedom,” 341 U. S., at 376, it displayed no solicitude for state courts.30 The debates over the 1871 Act are replete with hostile comments directed at state judicial *364systems.31 It is entirely reasonable to conclude that Congress intended to make state legislators immune from civil liability under § 1 of the 1871 Act. No similar evidence exists to support an immunity for police officers testifying as witnesses.32

IV

The majority also bases its decision on considerations of public policy, which purportedly mandate absolute immunity for police officers sued under § 1983 for their testimony as witnesses. Ante, at 341-345. This Court has recognized absolute immunity only in “exceptional situations” where public policy makes it “essential.” Butz v. Economou, 438 U. S. 478, 507 (1978).33 In my view, the case for official-witness immunity falls far short of this standard.

*365Police officers and other government officials differ significantly from private citizens, around whom common-law doctrines of witness immunity developed. A police officer comes to the witness stand clothed with the authority of the State. His official status gives him credibility and creates a far greater potential for harm than exists when the average citizen testifies.34 The situation is aggravated when the official draws on special expertise. A policeman testifying about a fingerprint identification or a medical examiner testifying as to the cause of a death can have a critical impact on a defendant’s trial.35 At the same time, the threat of a criminal perjury prosecution, which serves as an important constraint on the average witness’ testimony, is virtually nonexistent in the police-witness context. Despite the apparent prevalence of police perjury,36 prosecutors exhibit extreme *366reluctance in charging police officials with criminal conduct because of their need to maintain close working relationships with law enforcement agencies.37 The majority thus forecloses a civil sanction in precisely those situations where the need is most pressing.

Moreover, the danger that official witnesses would be inhibited in testifying by the fear of a damages action is much more remote than would be the case with private witnesses. Policemen normally have a duty to testify about matters involving their official conduct. The notion that officials with a professional interest in securing criminal convictions would shade their testimony in favor of a defendant to avoid the risk of a civil suit can only be viewed with skepticism. In addition, police officials are usually insulated from any economic hardship associated with lawsuits based on conduct within the scope of their authority.38 In any event, if the Court truly desires to give police officers “ ‘every encouragement to make a full disclosure of all pertinent information within their knowledge,’” ante, at 335 (quoting Imbler v. Pachtman, 424 U. S., at 439 (White, J., concurring in judgment)), then at the very least it should permit § 1983 suits which allege that officials withheld key information while testifying.39

*367The majority’s primary concern appears to be that §1983 suits against police witnesses would impose “significant burdens on the judicial system and on law enforcement resources.” Ante, at 343. As an empirical matter, this contention is unfounded. Both the Sixth Circuit and the District of Columbia Circuit have permitted such suits for over five years, see ante, at 328-329, n. 4, but there is no perceptible drain on legal resources in those Circuits compared to other Circuits that bar such lawsuits. Moreover, a comprehensive study of § 1983 suits filed in the Central District of California, which includes Los Angeles, indicates that only about 30 actions for false arrest were filed annually in that District.40 Police officers arrest much more frequently than they testify, and an arrest will undoubtedly make many individuals disgruntled. Yet, lawsuits based on such allegations constituted only 0.5% of all cases filed in the Central District,41 or an average of only one for every 243 full-time police *368officers in the city of Los Angeles.42 This does not appear to be a “significant burden.”43 The simple fact is that practical obstacles alone are enough to deter most individuals from suing the police for official misconduct.44

In considering the competing interests at stake in this area, the majority strikes a very one-sided balance. It eschews any qualified immunity in favor of an absolute one. Thus, the mere inquiry into good faith is deemed so undesirable that we must simply acquiesce in the possibility that government officials will maliciously deprive citizens of their rights.45 For my part, I cannot conceive in this case how patent violations of individual rights can be tolerated in the name of the public good. “The very essence of civil liberty certainly consists in the right of every individual to claim the protections of the laws, whenever he receives an injury.” Marbury v. Madison, 1 Cranch 137, 163 (1803).

*369V

For all of the above reasons, I believe that the majority has failed to sustain the heavy burden required to justify an immunity so plainly at odds with the language and purpose of § 1983. I therefore respectfully dissent.

dissenting.

Justice Marshall’s dissenting opinion, post, presents an eloquent argument that Congress, in enacting § 1983, did not intend to create any absolute immunity from civil liability for “government officials involved in the judicial process . . . .” Post, at this page and 347. Whatever the correctness of his' historical argument, I fear that the Court has already crossed that bridge in Pierson v. Ray, 386 U. S. 547 (1967), and Imbler v. Pachtman, 424 U. S. 409 (1976).

I entirely agree with Justice Marshall, however, that the policies of § 1983 and of common-law witness immunity, as they apply to witnesses who are police officers, do not justify any absolute immunity for perjurious testimony. I therefore dissent for the reasons stated in Part IV of Justice Marshall’s opinion.

dissenting.

I join all of Justice Marshall’s dissenting opinion except Part I. I cannot join its Part I, for I adhere to the views I expressed for the Court in City of Newport v. Fact Concerts, Inc., 453 U. S. 247, 258-259 (1981), regarding the role played by history and policy in determining whether § 1983 incorporates a particular common-law immunity. It is proper to assume — indeed, the Court in the past has assumed — “that members of the 42d Congress were familiar with common-law principles . . . and that they likely intended these common-law principles to obtain, absent specific provisions to the contrary.” Id., at 258. If an immunity was well established in the common law in 1871, careful analysis of the policies supporting it, and those supporting § 1983, governs the determination whether that immunity was retained.

In my view, Justice Marshall’s dissent convincingly demonstrates that the Court finds little support for its decision in the present case either in the language of the statute, the history of the common law, the relevant legislative history, or policy considerations.

I therefore dissent.