Cleamtee Garner v. Memphis Police Department

MERRITT, Circuit Judge. The principal question before us concerns the constitutionality of Tennessee’s fleeing felon statute, T.C.A. § 40-808 (1975) under the Fourth, Eighth and Fourteenth Amendments. The Tennessee statute, as interpreted by the District Court and by other federal and state courts, authorizes police officers to use deadly force in order to capture unarmed suspects fleeing from nonviolent felonies. The statute reads: “If .. . the defendant ... either flee or forcibly resist, the officer may use all the necessary means to effect the arrest.” In the present action for wrongful death under 42 U.S.C. § 1983 (1976), a Memphis police officer shot an unarmed boy fleeing from the burglary of an unoccupied house. We hold the Tennessee statute unconstitutional because it authorizes unnecessarily severe and excessive, and therefore unreasonable, methods of seizure of the person under the Fourth and Fourteenth Amendments. I. On the night of October 3, 1974, a fifteen year old, unarmed boy broke a window and entered an unoccupied residence in suburban Memphis to steal money and property. Two police officers, called to the scene by a neighbor, intercepted the youth as he ran from the back of the house to a six foot cyclone fence in the back yard. After shining a flashlight on the boy as he crouched by the fence, the officer identified himself as a policeman and yelled “Halt.” He could see that the fleeing felon was a youth and was apparently unarmed. As the boy jumped to get over the fence, the officer fired at the upper part of the body, using a 38-calibre pistol loaded with hollow point bullets, as he was trained to do by his superiors at the Memphis Police Department. He shot because he believed the boy would elude capture in the dark once he was over the fence. The officer was taught that it was proper under Tennessee law to kill a fleeing felon rather than run the risk of allowing him to escape. The youth died of the gunshot wound. On his person was ten dollars worth of money and jewelry he had taken from the house. The District Court dismissed the suit brought by decedent’s father against the *242 City under 42 U.S.C. § 1983 (1976) to recover damages for wrongful death caused by claimed constitutional violations of the Fourth, Eighth and Fourteenth Amendments. In accordance with Monroe v. Pape, 365 U.S. 167, 81 S.Ct. 473, 5 L.Ed.2d 492 (1961), the District Court held that a city is not a “person” subject to suit under § 1983. Before we heard the first appeal, Monroe was overruled on this point by Monell v. Department of Social Services, 436 U.S. 658, 98 S.Ct. 2018, 56 L.Ed.2d 611 (1978). The District Court also dismissed the case against the officer and his superiors holding, in accordance with our decisions in Beech v. Melancon, 465 F.2d 425 (6th Cir.1972), cer t. denied, 409 U.S. 1114, 93 S.Ct. 927, 34 L.Ed.2d 696 (1973); Qualls v. Parrish, 534 F.2d 690 (6th Cir.1976); and Wiley v. Memphis Police Department, 548 F.2d 1247 (6th Cir.), cert. denied, 434 U.S. 822, 98 S.Ct. 65, 54 L.Ed.2d 78 (1977), that the officers acted in good faith reliance on Tennessee law which allows an officer to kill a fleeing felon rather than run the risk of allowing him to escape apprehension. On appeal, a panel of this Court consisting of Chief Judge Edwards and Judges Lively and Merritt affirmed the District Court’s holding that the individual defendants were protected by the doctrine of qualified immunity because they acted in good faith reliance on T.C.A. § 40-808. Garner v. City of Memphis, 600 F.2d 52 (6th Cir.1972). We reversed and remanded the case against the City of Memphis, however, for reconsideration by the District Court in light of Monell v. Department of Social Services, supra. Because Monell held that a city may be liable in damages under § 1983 for constitutional deprivations that result from a “policy or custom” followed by the city, 436 U.S. at 694 and n. 66, 98 S.Ct. at 2037-2038 and n. 66, we instructed the District Court to consider the following questions: 1. Whether a municipality has qualified immunity or privilege based on good faith under Monell? 2. If not, is a municipality’s use of deadly force under Tennessee law to capture allegedly nondangerous felons fleeing from nonviolent crimes constitutionally permissible under the Fourth, Sixth, Eighth and Fourteenth Amendments? 3. Is the municipality’s use of hollow point bullets constitutionally permissible under these provisions of the Constitution? 4. If the municipal conduct in any of these respects violates the Constitution, did the conduct flow from a “policy or custom” for which the City is liable in damages under Monell? 600 F.2d 52, at 54-55. On remand, Judge Wellford ordered memoranda and oral argument on the issue of whether the trial should be reopened. By order dated February 29,1980, he denied further hearings and dismissed the case on the merits, holding that the constitutional claims had already been fully adjudicated. Because there had been no constitutional violation, the holding of Monell that cities could be liable for violations occurring pursuant to a policy or custom of the city did not require a different result. Plaintiff’s motion for reconsideration was granted and he was allowed to submit further briefs and make an offer of proof. The Judge considered the offer of proof and once again ruled against plaintiffs in a written opinion dated July 8, 1981. He held that the wisdom of a statute permitting the use of deadly force against all fleeing felons was a matter of policy for the legislature rather than the judiciary, and that the Tennessee statute was not unconstitutional on its face, nor as applied by the police officer in this case. Addressing the question of the City’s good faith immunity, the District Court held that Owen v. City of Independence, 445 U.S. 622, 100 S.Ct. 1398, 63 L.Ed.2d 673 (1980), prevented the city from claiming immunity from liability based on the good faith of its agent. Nevertheless, it found that it was still an open question whether the City might claim immunity if the City itself was relying in good faith on the Tennessee law as interpreted by the federal and *243 state courts. Judge Wellford did not believe it necessary to address the constitutionality of the use of hollow point bullets, because he found that there was no causal connection between the use of hollow point bullets and Garner’s death. ■ II. We consider the Fourth Amendment question first because, unlike the other more general constitutional provisions raised, the Fourth Amendment is specifically directed to methods of arrest and seizure of the person. The question under the Fourth Amendment is one of first impression in this Circuit. The narrow question presented is whether a state law authorizing the killing of an unarmed, nonviolent fleeing felon by police in order to prevent escape, constitutes an unreasonable seizure of the person. The Fourth Amendment provides for the “right of the people to be secure in their persons ... against unreasonable .. . seizures.” The Amendment also provides that where a warrant is necessary it must describe “the person to be seized.” When an officer “accosts an individual and restrains his freedom to walk away,” the Fourth Amendment comes into play. Terry v. Ohio, 392 U.S. 1, 16, 88 S.Ct. 1868, 1877, 20 L.Ed.2d 889 (1968). “[A] person is ‘seized’ ... when, by means of physical force or a show of authority, his freedom of movement is restrained.” United States v. Mendenhall, 446 U.S. 544, 553, 100 S.Ct. 1870, 1877, 64 L.Ed.2d 497 (1980). Killing the individual is the most decisive way to make sure that he does not “walk away,” a method “unique in its severity and irrevocability.” Gregg v. Georgia, 428 U.S. 153, 187, 96 S.Ct. 2909, 2931, 49 L.Ed.2d 859 (1976). It is plainly a “seizure” of the “person.” The question therefore becomes whether this method of capturing suspects is “reasonable” under the Fourth Amendment. Tennessee courts have interpreted their statute regarding the capture of fleeing felons to create a jury question on the issue of the “reasonableness” and the “necessity” of using deadly force. But the “reasonableness” and “necessity” of the officer’s action must be judged solely on the basis of whether the officer could have arrested the suspect without shooting him. Purporting to follow the rule developed in England at common law allowing the use of deadly force against suspects fleeing from any felony, Tennessee courts have interpreted their statute to mean that once it is determined that the officer probably could not have captured the person without firing, the jury should find the police action reasonable under the statute. Scarbrough v. State, 168 Tenn. 106, 110, 76 S.W.2d 106 (1934) (officer may kill automobile thief “as a last resort” to prevent escape and the question of “necessity of killing” is one for jury); see also to the same effect Love v. Bass, 145 Tenn. 522, 238 S.W. 94 (1921) and State v. Boles, 598 S.W.2d 821 (Tenn.App. 1980) and the cases cited in those opinions. It makes no difference that the felony was nonviolent or that the felon was unarmed and not dangerous to the physical safety of others. It is true that the common law permitted the killing of a felon who resists arrest without regard to the nature of the felony. But it did so at a time when all of the small number of felonies then in existence were capital crimes. Since any felon at large would be hanged or otherwise executed if taken and tried, he was an “outlaw” who was automatically dangerous and posed an imminent threat to the physical safety of others. The common law, however, prohibited the use of deadly force against a fleeing suspect whose crime did not require execution and who, therefore, was not likely to become a dangerous outlaw. Pollock and Maitland describe as follows the felony at common law and the method by which a felon could be taken: But the very ease with which the king’s peace spread itself until it had become an all-embracing atmosphere prevented a mere breach of that peace from being permanently conceived as a crime of the highest order. ... It was otherwise with felony. This becomes and remains a name for the worst, the bootless crimes. *244 The specific effect of the ‘words of felony’ when they were first uttered by appellors, who were bringing charges of homicide, robbery, rape and so forth, was to provide that, whatever other punishment the appellees might undergo, they should in all events lose their land... . At all events this word, expressive to the common ear of all that was post hateful to God and man, was soon in England and Normandy a general name for the worst, the utterly ‘bootless’ crimes.... The felon's lands go to his lord or to the King and his chattels are confiscated. The felon forfeits life or member. If a man accused of felony flies, he can be outlawed. We have now to speak of the various processes which the law employs in order to compel men to come before its courts. They vary in stringency from the polite summons to the decree of outlawry.... When a felony is committed the hue and cry should be raised.... The neighbors should turn out with the bows, arrows, knives, that they are bound to keep and, besides much shouting, there will be hornblowing; the ‘hue’ will be ‘horned’ from vale to ville. Now if a man is overtaken by hue and cry while he has still about him the signs of his crime, he will have short shrift. Should he make any resistance, he will be cut down. There is hardly room for doubt that this process had its origin in days when the criminal taken in the act was ipso facto an outlaw. He is not entitled to any ‘law,’ not even to that sort of ‘law’ which we allow to noble beasts of the chase. Even when the process is being brought within some legal control, this old idea survives. If there must be talk of proof, what has to be proved is not that this man is guilty of a murder, but that he was taken red-handed by hue and cry. II Pollock and Maitland, History of English Law, 464-66, 578-80 (2d ed. 1959). It is this common law rule allowing all fleeing felons to be killed, a rule based on the ancient concept of outlawry, that Tennessee courts have adopted in interpreting their statute. These killings were acceptable at common law because only violent crimes were classified as felonies, and all were punishable by death and subject to outlawry. The killing of a fleeing felon merely accelerated the time of punishment. The rule of outlawry permitting the killing of the fleeing felon did not apply to misdemeanors and lesser crimes. Lesser criminals who took flight from their crimes could not be killed to prevent their escape. See Comment, Deadly Force to Arrest: Triggering Constitutional Review, 11 Harv. C.R.-C.L. L. Rev. 361, 364-65 (1976). It is inconsistent with the rationale of the common law to permit the killing of a fleeing suspect who has not committed a life endangering or other capital offense and who we cannot say is likely to become a danger to the community if he eludes immediate capture. Those states like Tennessee that cite the common law in defense of their rule permitting the killing of any fleeing felony suspect exalt the form of the common law rule over its substance and purpose. Tennessee law authorizing the use of deadly force against all fleeing felons is at odds with the purpose and function of the common law principle because there are now hundreds of state and federal felonies that range all the way from violations of tax, securities and antitrust laws and the possession of stolen or fraudulently obtained property to murder and crimes of terror. A state statute or rule that makes no distinctions based on the type of offense or the risk of danger to the community is inherently suspect because it permits an unnecessarily severe and excessive police response that is out of proportion to the danger to the community. This line of reasoning concerning the origin, development and current status of the common law rule is similar to the reasoning of the Eighth Circuit in its en banc decision *245 in Mattis v. Schnarr, 547 F.2d 1007 (8th Cir.1976), vacated as moot per curiam sub nom Ashcroft v. Mattis, 431 U.S. 171, 97 S.Ct. 1739, 52 L.Ed.2d 219 (1977). There the court held a similar state statute in Missouri unconstitutional under the Fourteenth Amendment as a matter of substantive due process. After tracing some of the history of the fleeing felon doctrine and cataloguing in comprehensive fashion the state statutes on the question, as well as federal decisions, administrative rules and scholarly commentary, the Eighth Circuit observed that “the historical basis for permitting the use of deadly force by law enforcement officers against nonviolent fleeing felons has been substantially eroded,” 547 F.2d at 1016. At common law “since all felonies ... were punishable by death, the use of deadly force was seen as merely accelerating the penal process... . ” 547 F.2d at 1011 n. 7. Likewise, in Jones v. Marshall, 528 F.2d 132 (2d Cir.1975), the Second Circuit in a scholarly opinion by Judge Oakes observed that a rule which permits the use of deadly force against nonviolent fleeing felons is not consistent with the purpose and function of the common law rule. Although the Jones case, like our earlier opinion in this case, insulates the officer from federal liability when, in reliance on a similar state statute, he shoots a nonviolent fleeing felon, the court commented: [T]he common law rule evolved when only a few crimes were felonies, and all of them involved force or violence ... and were punishable by death or forfeiture of lands and goods. See ALI, Model Penal Code § 3.07, Comment 3 at 56 (Tent. Draft No. 8, 1958). (“Such rational justification for the common law rule as can be adduced rests largely on the fact that virtually all felonies in the common law period were punishable by death”).... As the scope of “felony” crimes has expanded wholly away from the concept of violence which underlay its common law origin, the use of the felony label to justify especially severe police behavior has become increasingly strained. As stated by Judge McCree in his concurring opinion in Beech v. Melancon, 465 F.2d 425, 426-27 (6th Cir.1972), cert. denied, 409 U.S. 1114, 93 S.Ct. 927, 34 L.Ed.2d 696 (1973): “. .. I would find it difficult to uphold as constitutional a statute that allowed police officers to shoot, after an unheeded warning to halt, a fleeing income tax evader, antitrust law violator, selective service delinquent, or other person whose arrest might be sought for the commission of any one of a variety of other felonies of a type not normally involving danger of death or serious bodily harm.” We have thoroughly explored the digests and the electronic case retrieval systems, and our research discloses only one appellate decision discussing Fourth Amendment limitations on the use of deadly force to capture a fleeing suspect. In Jenkins v. Averett, 424 F.2d 1228 (4th Cir.1970), a black youth took flight at night. The police officer cornered the boy and shot him. The District Court dismissed the federal constitutional claim. Applying a Fourth Amendment analysis, the Fourth Circuit in an opinion by Judge Sobeloff reversed. Holding that the Fourth Amendment “shield covers the individual’s physical integrity,” the Court found a constitutional violation because “our plaintiff was subject to the reckless use of excessive force.” 424 F.2d at 1232. The only other discussion of the reasonableness of the use of deadly force by police in a Fourth Amendment context is that of Chief Justice Burger in his dissenting opinion in Bivens v. Six Unknown Federal Narcotic Agents, 403 U.S. 388, 91 S.Ct. 1999, 29 L.Ed.2d 619 (1971). In Bivens the Court held that the Fourth Amendment creates a direct constitutional tort claim for violation of a citizen’s right to be free of illegal searches of the home and seizures of the person. Although Bivens was not a fleeing felon case, Chief Justice Burger, in the course of his Fourth Amendment analysis in dissent, observed: I wonder what would be the judicial response to a police order authorizing ‘shoot to kill’ with respect to every fugitive. It *246 is easy to predict our collective wrath and outrage. We, in common with all rational minds, would say that the police response must relate to the gravity and need; that a ‘shoot’ order might conceivably be tolerable to prevent the escape of a convicted killer but surely not for car thieves, pickpockets or a shoplifter. Bivens v. Six Unknown Agents, 403 U.S. 388, 419, 91 S.Ct. 1999, 2016, 29 L.Ed.2d 619 (1971) (Burger, C.J., dissenting) (emphasis added). The Sixth Circuit long ago in United States v. Clark, 31 F. 710 (6th Cir.1887), expressed similar doubts about the validity of a rule allowing deadly force against all fleeing felony suspects: Suppose, for example, a person were arrested for petit larceny, which is a felony at the common law, might an officer under any circumstances be justified in killing him? I think not. The punishment is altogether too disproportionate to the magnitude of the offense. Id. at 713. The Tennessee statute in question here is invalid because it does not put sufficient limits on the use of deadly force. It is “too disproportionate.” It does not make distinctions based on “gravity and need” nor on “the magnitude of the offense.” Before taking the drastic measure of using deadly force as a last resort against a fleeing suspect, officers should have probable cause to believe not simply that the suspect has committed some felony. They should have probable cause also to believe that the suspect poses a threat to the safety of the officers or a danger to the community if left at large. The officers may be justified in using deadly force if the suspect has committed a violent crime or if they have probable cause to believe that he is armed or that he will endanger the physical safety of others if not captured. A statute which allows officers to kill any unarmed fleeing felon does not meet this standard and is therefore invalid. After oral argument in this case, upon motion, the Court permitted the state of Tennessee, through its Attorney General, William M. Leach, Jr., to intervene as a party under 28 U.S.C. § 2403(b) for the purpose of defending the constitutionality of T.C.A. § 40-7-108. The State has filed an able brief. It concedes that Tennessee courts and law enforcement agencies interpret the statute to permit the use of deadly force against any fleeing felon, whatever the felony, “when no lesser means of apprehension reasonably appears available.” (Brief, p. 5.) The State's brief argues, however, that we should not reach the issue of whether Tennessee’s rule may be constitutionally applied to a nondangerous felon fleeing from a non-violent felony because here the officer “could not be certain whether there was an accomplice in the burglarized house, or in the area, and whether the accomplice might be armed.” (Brief, p. 6.) This argument almost always permits the officer to shoot to kill. The officer will seldom be absolutely certain of the situation. The Fourth Amendment resolves this problem, however. It requires probable cause — an objective, reasonable basis in fact to believe that the felon is dangerous or has committed a violent crime. There is no evidence to support such a finding in this case, although as the state argues, and as the District Court found, the officer was not certain on this point. The officer knew only that he was dealing with a youth who had committed a non-violent felony and was apparently unarmed. We do not have to hold the District Judge’s findings clearly erroneous in order to reach this result, because the facts, as found, did not justify the use of deadly force under the Fourth Amendment. An analysis of the facts of this case under the Due Process Clause of the Fourteenth Amendment leads us to a similar result. That clause prohibits any State from depriving “any person of life, liberty, or property, without due process of law.” U.S. Const. Amend. XIV. The right to life, expressly protected by the Constitution, has been recognized repeatedly by the Supreme Court as fundamental in the due process and equal protection contexts. Yick Wo v. *247 Hopkins, 118 U.S. 356, 370, 6 S.Ct. 1064, 1071, 30 L.Ed. 220 (1886) (the fundamental rights “to life, liberty and the pursuit of happiness”); Johnson v. Zerbst, 304 U.S. 458, 462, 58 S.Ct. 1019, 1022, 82 L.Ed. 1461 (1938) (“the fundamental human rights of life and liberty”); Roe v. Wade, 410 U.S. 113, 93 S.Ct. 705, 35 L.Ed.2d 147 (1973) (right to life protected by Fourteenth Amendment when fetus becomes viable). When a fundamental right is involved, due process requires a state to justify any action affecting that right by demonstrating a compelling state interest. Roe v. Wade, supra; Mattis v. Schnarr, 547 F.2d 1007, 1019 (8th Cir.1976) (en banc). Laws which infringe on fundamental rights must be “narrowly drawn to express only the legitimate state interests at stake.” Roe v. Wade, supra. The law challenged here is not so narrowly drawn. Certainly there are state interests in law enforcement served by this law which allow police to shoot all fleeing felons. Those interests are compelling when the fleeing felon poses a danger to the safety of others. We do not consider these interests sufficiently compelling to justify the use of deadly force to protect only property rights. As the Eighth Circuit pointed out in striking down a similar law: We find nothing in this record ... to support the contention of the state that statutes as broad as these deter crime, insure public safety or protect life. Felonies are infinite in their complexity, ranging from the violent to the victimless. The police officer cannot be constitutionally vested with the power and authority to kill any and all escaping felons, including the thief who steals an ear of corn, as well as one who kills and ravishes at will. Mattis v. Schnarr, supra 547 F.2d at 1019-20 (footnote omitted). Where, as here, human life is the right at stake, a statute that sweeps as broadly as this one violates due process of law and must be struck down. The principles and distinctions we have enunciated here have been cast in the form of a rule by the American Law Institute in the Model Penal Code, a rule which accurately states Fourth Amendment limitations on the use of deadly force against fleeing felons: The use of deadly force is not justifiable ... unless (i) the arrest is for a felony, and (ii) the person effecting the arrest is authorized to act as a peace officer or is assisting a person whom he believes to be authorized to act as a peace officer; and (iii) the actor believes that the force employed creates no substantial risk of injury to innocent persons; and (iv) the actor believes that (1) the crime for which the arrest is made involved conduct including the use or threatened use of deadly force; or (2) there is a substantial risk that the person to be arrested will cause death or serious bodily harm if his apprehension is delayed. Model Penal Code § 3.07(2)(b) (Proposed Official Draft, 1962). Our holding here under the Fourth Amendment is not inconsistent with our holdings in Wiley v. Memphis Police Department, 548 F.2d 1247 (6th Cir.1977), and Beech v. Melancon, 465 F.2d 425 (6th Cir. 1972), or the three judge District Court opinion in Cunningham v. Ellington, 323 F.Supp. 1072 (W.D.Tenn.1971), in all of which youths were killed by the Memphis police while fleeing from the commission of a burglary. In each of those cases the Tennessee statute was drawn into question under the “cruel and unusual punishment” clause of the Eighth Amendment and under the Fourteenth Amendment as a matter of substantive due process. In none of these cases was a Fourth Amendment question raised, discussed, mentioned or decided. Fourth Amendment considerations were not argued. Moreover, in each of the cases the narrow question before the court was whether the police officer who shot the fleeing boy was entitled to a good faith privilege against liability based upon his reliance upon the Tennessee statute. In each case the court held, just as we held in our previous decision in this case, Garner v. City of Memphis, supra, that the officer is insulated from personal liability by a good faith privilege which entitles him to rely *248 upon the Tennessee statute. This is the ratio decidendi of each of those cases. In those cases it was unnecessary to reach the constitutionality of the statute in order to decide the question of the officers’ immunity, and in any event, no Fourth Amendment question was raised in any of the cases. III. In his opinion of July 8,1981, Judge Wellford held that although Owen v. City of Independence, 445 U.S. 622, 100 S.Ct. 1398, 63 L.Ed.2d 673 (1980), precludes the city of Memphis from claiming immunity based on the good faith of its police officers, that opinion left open the question whether the City could claim immunity for its good faith reliance on a facially valid state law in enacting City police regulations. Because he found no violation of Garner’s constitutional rights, the Judge did not have to answer this question. In light of our finding of a constitutional violation, we must reach this question; in doing so, we hold that there is no good faith immunity for municipalities under § 1983. The reasoning underlying the Supreme Court’s decision in Owen, supra, precludes a municipality’s claim of good faith immunity under § 1983 altogether. Justice Brennan, speaking for the Court in Owen, gave two major reasons why good faith immunity of city officials should not be extended to municipalities themselves. First he pointed out that at common law, which is the source of immunities under § 1983, there was no good faith immunity for governmental entities. Id. at 640, 100 S.Ct. at 1410. Sovereign immunity at common law was unrelated to the question of good faith and was waived when the government consented to suit as it does under § 1983. Immunity for discretionary functions, the only other governmental immunity at common law, involved concerns of separation of powers, unrelated to good faith. Because a municipality has no “discretion” to violate constitutional rights of its citizens, this traditional form of immunity does not come into play. There is no common law analogue which would suggest that municipalities have immunity for good faith reliance on state law under § 1983. Second, Justice Brennan discussed the public policy considerations which justify individual good faith immunity and found that they did not weigh heavily in favor of governmental immunity. The two considerations are (1) the injustice of forcing an individual whose position requires him to exercise discretion to bear the cost of his good faith reliance on a law or regulation; and (2) the danger that the threat of liability would deter individuals from executing the duties of their offices or even from seeking public office. Id. at 654, 100 S.Ct. at 1417. When a municipality is held liable, whether for the actions of its officials, or based on its own reliance on state law, no single individual or official must bear the cost. The cost is spread among the general public, which is ultimately responsible for the conduct of its officials. There is little danger that individuals will hesitate to carry out their duties or accept public office, when any liability for their reliance on state law will be paid from the public fisc. In a well-reasoned opinion, the Tenth Circuit sitting en banc, held that good faith reliance by a school district on the prior law of the circuit provided no independent protection from liability for wrongful dismissal of a teacher. Bertot v. School District No. 1, Albany County, 613 F.2d 245, 251 (10th Cir.1979). It held that the remedying of deprivations of fundamental constitutional rights must be of primary concern to courts and other governmental bodies. A rule imposing liability despite good faith reliance insures that if governmental officials err, they will do so on the side of protecting constitutional rights. It also serves the desirable goal of spreading the cost of unconstitutional governmental conduct among the taxpayers who are ultimately responsible for it. Id. at 252. Neither the District Judge nor the City of Memphis has offered any reason why the courts should expand the doctrine of good faith immunity under § 1983. The considerations which prompted the Supreme Court in Owen to deny good faith immunity *249 to municipalities for the acts of their officials apply with equal force to this case. Accordingly, the judgment of the District Court is reversed and the case remanded for further proceedings consistent with this opinion.