specially concurring:
It is with considerable grief that I write to specially concur in the result denying punitive damages against the City of Houston. I fully concur in the majority opinion on all other issues, but must specially concur on the issue of punitive damages be*1231cause the majority suggests that City of Newport v. Fact Concerts, Inc., 453 U.S. 247, 101 S.Ct. 2748, 69 L.Ed.2d 616 (1981), might allow punitive damages against a municipality in a section 1983 suit if the facts were particularly egregious. Would that it were so, for then I could, with clear judicial conscience, urge taxing the City of Houston with punitive damages. If there were any narrow gap around Newport for an egregious case, this one would slip through; I am aghast at the thought that any violation of constitutional rights more appalling, more threatening than the one that occurred here might actually exist. Sadly, I view Newport as presenting an impenetrable barrier to punitive damages. Would that it were not so, for now I must, with troubled human conscience, concur in this unfortunate result.
The majority quotes a portion of a footnote from Newport as admitting an exception for egregious cases:
29. It is perhaps possible to imagine an extreme situation where the taxpayers are directly responsible for perpetrating an outrageous abuse of constitutional rights. Nothing of that kind is presented by this case. Moreover, such an occurrence is sufficiently unlikely that we need not anticipate it here.
Newport, 101 S.Ct. at 2760 n.29. The majority states that “[t]he plight of Randy Webster, however reprehensible, however tragic, does not rise to the level of outrageous conduct to which Justice Blackmun referred .... [W]e believe that it would take a far more serious violation than that we confront to ground punitive damages against Houston.” Majority Opinion, supra at 1229. Among the multitude of legal standards of conduct, courts have required findings of malicious, wilfull, or intentional conduct before exacting punitive damages. If Newport had in fact created a new standard, “an outrageous abuse of constitutional rights,” I would have no problem in holding as a matter of law that this case is such an outrageous abuse.
I believe the majority opinion has misinterpreted this footnote; rather than creating an exception, this footnote precludes any exception. It appears in a section of the opinion addressing a possible retribution justification for punitive damages. Justice Blackmun argues that retribution against the taxpayers would make no sense except in some rare case where the taxpayers were actually culpable,1 and that did not seem likely enough to merit consideration.2 The bottom line of the opinion admits of no exceptions: “[W]e hold that a municipality is immune from punitive damages under 42 U.S.C. § 1983.” 101 S.Ct. at 2762.3
The language of Newport is clear and absolute, and, accordingly, so is our duty in this case. If I merely disagreed with the Supreme Court’s rationale in Newport, I would concur without further comment, and in fact, I agree with that rationale and the result in that case. However, after due deliberation and with great respect for the august institution above me, I am of the belief that the holding of Newport sweeps more broadly than its rationale would justify. Because of the gravity of the constitutional violation in the present case, I would be derelict in my duty if I did not express my views before acquiescing in this result.
I am in agreement with the Supreme Court in Newport not only in regard to rationale and result, but also in regard to the Court’s overall approach to section 1983. What disturbs me is that the very method*1232ology and rationale that dictate the result in Newport dictate a contrary result now. To illustrate this I shall follow the methodology of Newport and at each point indicate how the differences between Newport and the instant case, under the same form of analysis, lead to a different conclusion now. First I shall summarize the factual differences between the cases. Next, I shall cover the common law background and legislative history of section 1983. Finally, I shall discuss the policies underlying imposition of punitive damages.
I. FACTS OF NEWPORT AND WEBSTER V. CITY OF HOUSTON
My interest in contrasting the facts of these cases is not solely to highlight the disparity in gravity of constitutional violations. Though that disparity tugs at the emotions, it is not a sufficient basis for protesting binding precedent. Rather, I want to establish the factual predicate for the arguments that follow. My goal is to show that the salient facts supporting the rationale in Newport are not present here; in fact, the opposite facts supporting an opposite result are present.
A. State of Nature in the State of Texas: Webster v. City of Houston
Randy Webster, a seventeen-year-old, stole a van in the City of Houston. Three police officers, guardians of the peace, gave chase. At the conclusion of a high-speed chase, Randy lost control of the van and attempted to surrender to the police. He had his hands up and was trying to get out of the stolen van, unarmed, offering no resistance, when the police, to whom he wished to surrender, forced him to the ground and shot him in the back of the head. The police, rightly nervous about the consequences of killing an unarmed and unresisting teenager who was trying to surrender peaceably, planted a weapon near the corpse. The purpose of planting this “throw-down” gun was to add plausibility to their subsequent planned perjury — they did not shoot an unarmed, unresisting boy, no, they merely defended themselves against an armed and dangerous fleeing felon.
This crime, heinous as it is, is not what arouses my ire, nor is it the basis for municipal liability. What is outrageous, what makes the city liable, what is a perversion of public duty almost beyond belief is that the policemen acted in accordance with tacit Houston Police Department policy. The evidence showed that 75-80% of Houston policemen carried throw-down guns and that at least one incident like this had happened before. The heirarchy of the police department was aware of and indirectly condoned the practice. Most outrageous is that the practice was taught at the police academy, passed on as oral tradition from instructor to cadet.
A key point about this case is that Houston’s tacit policy did not originate in any memo, it cannot be traced to the acts of any specific persons. Unlike the shooting, the “promulgation” of the policy of using throw-down guns does not involve any discrete actions by any discrete policeman, but rather the collective acts of omission of every policeman, every city official, every person who knew of the practice.
B. Rock and Roll Will Never Die: City of Newport v. Fact Concerts, Inc.
Fact Concerts, Inc. (“Fact”) had entered into a contract with the City of Newport to hold a jazz festival at a local concert site. The mayor and council became alarmed when they heard that Blood, Sweat and Tears would perform.4 When Fact refused to remove Blood, Sweat and Tears from the bill, the city cancelled the contract on a pretext and announced the cancellation widely over the local media one day prior to the beginning of the concert. Fact obtained injunctive relief from a state court and the show went on. Unfortunately, due to the adverse publicity, ticket sales were off.
*1233These were the facts on which the Supreme Court held that under no circumstances could punitive damages be assessed against a municipality.
C. Compare and Contrast: Lost Sales v. Lost Lives
Though the general tenor of the two cases is manifestly different, it is useful to summarize the salient differences for the purpose of my subsequent legal analysis. In Newport the relevant activity was operating a concert site, a paradigmatic proprietary municipal activity; in this case the relevant activity is operating a police force, a paradigmatic governmental activity. In Newport the challenged activity was an ad hoc decision; here it is a widely followed municipal policy. In Newport there were specifically identifiable wrongdoers; here it is impossible to point to any person or persons responsible for the offending policy. In Newport the very nature of the wrong makes it public, whereas here the very nature of the wrong tends towards concealment. Finally, in Newport the consequence of the wrong was lost ticket sales; here it is needlessly, tragically, lost life.
With this factual background for Newport and our current case in mind, I now turn to an examination of the Supreme Court’s inquiry into punitive damages against a municipality under section 1983. The Court’s approach was two-phased. First, it investigated whether the legislative history, viewed in light of the then-existing common law background, allowed any finding of legislative intent to allow punitive damages against a municipality on those facts. It then examined the policies behind section 1983 to see if they would support punitive damages on those facts. I shall follow this two-phase approach, applying each phase first to Newport, then to our present case.
II. LEGISLATIVE HISTORY OF SECTION 1983
The Court’s underlying approach to the legislative history of section 1983 is in two parts. First, the Court presumes that Congress was well aware of the common law immunities that existed at the time, and that section 1983 embodies them unless there is some evidence of congressional intent otherwise. Second, because there is virtually no legislative history for section 1983 itself, the Court looks to the rejected Sherman Amendment to the Act to discern Congress’ attitudes towards section 1983.
A. General Background
Before applying the Court’s legislative intent approach to the two cases, it is helpful to look at two broad points of historical background. First consider the general history of common law municipal immunities. Initially a general policy of immunity prevailed — no citizen could sue the sovereign save by leave of the sovereign, and a municipal corporation was an instrument of the sovereign and thereby immune from liability. As municipal corporations became more complex, this unitary view of immunity broke down. The justification for extending immunity to municipalities did not apply when municipalities acted like corporations rather than arms of the sovereign. Thus municipal liability was bifurcated depending on whether the municipal activity was of a governmental nature or a proprietary, commercial nature. When the municipality was acting in a governmental manner, it was totally immune from all liability. When the municipality was acting in a proprietary manner, it was immune from punitive damages but liable for compensatory damages.
The second historical background point is the context in which Congress enacted section 1983.
The practice with which Congress was especially concerned was the systematic refusal of local law enforcement officials to enforce state criminal laws against members of the Ku Klux Klan and others who attacked blacks and Republicans. State law did not sanction or permit such discrimination, and in some cases the highest state officials were Republicans who opposed that discrimination, but the individual state employees who actually *1234had to carry out the administration of the criminal law pursued that practice nonetheless. As Representative Perry noted, “Sheriffs, having eyes to see, see not; judges, having ears to hear, hear not .... In the presence of these gangs all the apparatus and machinery of civil government, all the processes of justice, skulk away as if government and justice were crimes and feared detection.” The Forty-second Congress regarded such a failure to enforce state laws as the quintessential violation of the equal protection clause. In some instances those state employees, especially sheriffs and deputies, actually joined in the criminal activities. The states themselves were criticized, not for actively supporting the violence or formally sanctioning the lack of law enforcement, but for “failing” or “neglecting” to curb this practice of non-enforcement by local judges and sheriffs.
Schnapper, Civil Rights Litigation After Monell, 79 Colum.L.Rev. 213, 229 (1979) (footnotes omitted). Thus, the primary concern of section 1983 is lawless violence against defenseless citizens in violation of their constitutional rights, aided and abetted by the local authorities.
B. Legislative History as Applied to Newport
In applying the two-part legislative history analysis in Newport, the Court first sketched the brief history of municipal immunities just presented, 101 S.Ct. at 2756 & n.19. Running a concert hall is clearly a proprietary activity, so the Court noted that the appropriate common law background was one of liability for compensatory damages, but immunity from punitive damages. 101 S.Ct. at 2756.
Given this common law background, the Court then looked for evidence of congressional intent to alter the common law scheme of municipal immunities. Because almost no legislative history on section 1983 itself exists, the Court examined the history of the Sherman Amendment on the theory that congressional attitudes towards issues in the Sherman Amendment suggest the likely congressional intent in section 1983. The original intent of the Sherman Amendment was to make municipalities liable for losses caused by mob violence. 101 S.Ct. at 2758-59. During debate on this amendment, it was made clear that the liability extended only to compensatory damages, not to punitive damages. The amendment was rejected.
The Newport Court “doubt[ed] that a Congress having no intention of permitting punitive awards against municipalities in the explicit context of the Sherman Amendment would have meant to expose municipal bodies to such novel liability sub silentio under § 1 of the Act.” 101 S.Ct. at 2759. The Court further pointed out that Congress rejected the Sherman Amendment in part because of the strain it would place on the municipal fisc and the unfairness of forcing innocent taxpayers to pay for the “deeds of persons over whom they had neither knowledge or control.” Id. The Court saw “no reason to believe that Congress’ opposition to punishing innocent taxpayers and bankrupting local governments would have been less applicable with regard to the novel specter of punitive damages against municipalities.” Id.
C. Legislative History as Applied to This Case
In this section I will apply the same two-part legislative history analysis of Newport to the facts of this case: first I will examine the common law scheme of immunities and then look for evidence of congressional intent to alter that scheme.
The questionable activity in this case involves running a police department, a clearly governmental function. As mentioned above, the common law scheme granted municipalities total immunity from liability for their governmental activities. The second part of the inquiry is whether Congress showed any intent to alter the common law scheme of immunities, and, if so, to what degree. As Monell v. Department of Social Services, 436 U.S. 658, 98 S.Ct. 2018, 56 L.Ed.2d 611 (1978), so clearly points out, with section 1983 Congress intended to make municipalities liable for policies vio*1235lating civil rights. Thus we have, in the case of governmental activities, a clear congressional intent to alter the existing scheme of common law immunities. Given that Congress intended to strip away at least some of the municipal common law immunities, the second step of the inquiry now becomes whether they intended to abolish ail such immunities, in the context of municipal governmental activities.
As a preliminary point, because Newport establishes that Congress did not intend to abrogate immunity from punitive damages in a municipality’s proprietary capacity, it might initially seem strange to believe that Congress would intend to allow liability for punitive damages in the traditionally more protected governmental capacity. This is not a strange thought, however, when considered in light of the core purposes of section 1983 — prevention of governmental misconduct caused by aiding and abetting violence through failure to exercise the machineries of justice. Because section 1983 is closely tied to the prevention of official misconduct, it should be and is more sensitive to municipal dysfunction in the governmental area than it is to dysfunction in the proprietary area. For this reason, a policy under which policemen are permitted literally to get away with murder by perpetrating a fraud in a subsequent investigation is more offensive to the principles of section 1983 than is a municipal breach of a contract for a concert site that results in lost ticket sales. Accordingly, it is not at all hard to believe Congress might eliminate more immunities in the sensitive governmental sphere of activity.
In determining how much immunity Congress intended to abrogate, I need to examine how Congress thought imposing liability would assist eradicating the evils of governmental misconduct. Again, I follow the technique of Newport and examine the Sherman Amendment. Senator Sherman, speaking on behalf of his amendment, stated:
Let the people of property in the southern States understand that if they will not make the hue and cry and take the necessary steps to put down lawless violence in those States their property will be holden responsible, and the effect will be most wholesome.
Cong. Globe, 42d Cong., 1st Sess. 761 (1871), quoted in Monell, 436 U.S. at 667, 98 S.Ct. at 2024. The Sherman Amendment, then, envisioned altering municipal behavior by placing liability on the citizens and relying on them to encourage alterations in behavior.
This Amendment, of course, was defeated. That defeat should not be construed as a rejection of this deterrent mechanism, however. The Amendment was rejected because of the novel source of liability, not because of the fact of liability or functioning of the deterrent mechanism. See Monell, 436 U.S. at 669-83, 98 S.Ct. at 2024-32. The Sherman Amendment was rejected because it imposed an affirmative duty on local governments to prevent violence by third parties. The portion of the Act containing what is now section 1983 raised no such problems, because it merely prevented local governments from violating the Constitution. “So long as federal courts were vindicating the Federal Constitution, they were providing the ‘positive’ government action required to protect federal constitutional rights and no question was raised of enlisting the States in ‘positive’ action.” Monell, 436 U.S. at 680-81, 98 S.Ct. at 2030-31. So by imposing liability on municipalities in section 1983, Congress intended to deter future deprivations of constitutional rights by placing an economic burden on the taxpayers, who would cause the municipality to change its evil ways.
III. PUBLIC POLICY IN SECTION 1983
A. Policy, Proprietary Functions and Newport
The second half of the Court’s two-phase section 1983 methodology is to investigate the policies behind section 1983 and see if those policies suggest any result. In Newport, the first phase of investigating legislative history showed that the common law scheme of immunities allowed compensatory damages but no punitive damages for *1236proprietary municipal activities, and the Court found no evidence of congressional intent to alter that scheme. The purpose of the second phase inquiry into policy, then, is to see if there are policy arguments strong enough to dictate a result counter to Congress’ intent.
The Court considered two rationales for punitive damages and found that neither was harmonious with the policies of section 1983. The first potential rationale was retribution. The Court felt it made no sense to exact punitive damages from a municipality for retributive purposes because the actual economic burden would fall on the taxpayers, who were blameless. 101 S.Ct. at 2759-60.5 This portion of Newport is valid under any facts and I have no quarrel at all with the Court’s treatment of retribution.
The Court then considered whether punitive damages against a municipality would deter “the malicious conduct of their policy-making officials.” 101 S.Ct. at 2760. First, the Court noted, punitive damages against the municipality probably would not deter wrongdoing individuals. Second, a deterrent effect is already present through existing sanctions, such as public reaction, ^reprimand from superiors, and so on. Third, damages levied directly against the offending officials serve as a more effective deterrent. Finally, the Court observed the potential for devastating financial liability for municipalities, resulting in “a serious risk to the financial integrity of these governmental entities.” 101 S.Ct. at 2760-61. Given all this, the Court concluded that the deterrence rationale did not support punitive damages.
B. The Core of Section 1983, Governmental Functions and Houston
The second phase inquiry in the present case has a different posture than it had in Newport because here the results of the first phase were different. The common law background was that municipalities were totally immune from damages for their governmental activities, and there was a definite congressional intent to alter that common law scheme. The legislative history of the Sherman Amendment shows that Congress’ motivation for altering the common law scheme was to deter governmental deprivation of citizens’ constitutional rights under color of state law. The second phase inquiry into policy now is to determine whether subjecting Houston to punitive damages in this case would promote Congress’ deterrent goals.
1. The Arguments Against Punitive Damages in Newport Do Not Apply Here. —The question of whether punitive damages would deter misconduct was considered in Newport, and there the Supreme Court found they would not help deter misconduct. The great factual differences between Newport and the case before me now, however, make Newport’s arguments against punitive damages inapplicable here. The primary factual difference making Newport’s arguments inapplicable is that Newport involved the wrongful acts of individual officials, whereas this case deals with a wrongful policy of concealing police murder, not traceable to any single individual, but rather the outgrowth of the collective inaction of the entire police department.
Thus, while the Court in Newport argued that punitive damages against the municipality would not deter individual misconduct, here we have municipal or group misconduct. Second, whereas in Newport the Court relied on sanctions from superiors as a sufficient deterrent, in this case the bad policy actually exists due to the collective inaction of the superiors. It seems strange to expect those same superiors to collectively sanction themselves for their continuing lassitude, and so Newport’s alternative sanction rationale does not apply. Third, in Newport the Court suggested that damages assessed directly against the offending official were a better deterrent. But while damages assessed directly against an individual might deter individual misconduct, at whom do we point the finger of guilt in *1237the ease of tragic collective apathy? All the individual police officers who knew of the policy but did nothing? The instructors at the police academy who allowed this horrible technique to be inculcated at the very time adherence to law and the Constitution and fundamental human morality should have been taught? The police chief? The city council? The citizens snug in their beds, oblivious to the fact that young men were being shot down by their “protectors”? Finally, the Court in Newport argued that punitive damages for lost ticket sales might deprive the citizens of other services because of financial burdens from increased liability. But in this core section 1983 case, that is exactly what Congress wanted. It is necessary that the threatened damages cause some deprivation for the populace so that they will be nudged out of their blissful ignorance, “and the effect will be most wholesome.”6
2. Punitive Damages in This Case Would Promote the Policies of Section 1983. —The factual differences between this case and Newport not only make Newport’s arguments against punitive damages inapplicable, they also affirmatively suggest that in this case punitive damages would help deter the conduct section 1983 was designed to prevent. Before turning to the specific factual differences that support punitive damages, I must discuss generally the theories behind deterrence and damages. Though it is a little unsettling to talk about dry economics in a setting as emotional as this one, that is what Congress’ deterrence rationale demands.
Under this economic view of damages, the key datum is that every action has social costs and social benefits. The correctness of a decision is judged by whether the social costs are less than the social benefits.7 Problems arise, however, when the decisionmaker will receive the benefits but other parties bear the cost. In that case the decisionmaker will decide to follow the course bringing benefits, regardless whether that course is socially correct. The function of damages is to force the decisionmaker to consider the costs of his or her actions as well as the benefits and promote socially correct decisionmaking.8 The point of damages, then, is to “internalize” costs to the decisionmaker.9
The choice between compensatory and punitive damages should be made based on whether one or the other will properly internalize the social costs and produce socially correct decisions. If all the parties bearing all the costs and enjoying all the benefits are before the court, then compensatory damages will properly account for all costs and benefits. For example, in the case of a city worried about violence at a music concert, the cost of cancelling the concert is the economic detriment to the promoter — lost ticket sales. The benefit of cancelling is potential avoidance of violence. If only compensatory damages are awarded, the city can make the correct decision. Compensatory damages force the city to internalize the cost of its decision, so it bears the loss rather than the promoter, and can make the right decision. Thus, in Newport the Court correctly denied punitive damages. All parties bearing all costs were before the court and only compensatory damages were proper.
*1238If, however, the situation is such that not all of the costs will come before the court, then compensatory damages will be inadequate to internalize the costs and punitive damages are needed. This is much more likely to happen when the relevant decision is one promulgating a policy, as in the present case, rather than one regarding a discrete event, as in Newport. The full costs of a policy might not come before the courts for several reasons.
The full costs of a policy might not be before a court because not all instances of the policy’s application are known. For example, if a policy results in ten citizens suffering a cost of $10, but only one of those instances is discovered, compensatory damages would not internalize all the costs. If a court can tell that nine instances of application are likely to remain undiscovered, when one does come before it, the court should assess $90 punitive damages to internalize the costs of the undiscovered applications. But this illustrates precisely one of the differences between Newport and this case. The very nature of the wrong in Newport was public — it was broadcast widely over the local media. The very nature of the wrong here is one of concealment — only one case of use of throw-down guns has ever been before the court, and when 75-80% of the Houston police were carrying throw-down guns, it is hard to avoid the inference that some cases are not coming before the courts.10
Another instance in which punitive damages are needed to internalize social costs is if some social costs are known but are diffused widely through society. This point highlights the key difference between Newport and this case, the reason it would be an injustice to award punitive damages in Newport and ' a greater injustice not to award them here. In Newport the total social damages from the incident are summed up in the lost profits of the promoters and some minor dissatisfaction of those music lovers who would have bought tickets but for the city council’s actions. This case, however, produces social costs of the gravest nature. Without trivializing the most grevious injury done to the Webster family, it makes a travesty of the most fundamental values of our society to ignore the damage done to them by Houston’s policy. The most primal reason for banding together in social groups is protection from violence. When members of the very institution created to protect us from violence instead inflict it, the social fabric is torn. This is but a tear, however, because we can all understand the possibility of uncontrollable renegades. But when the crime is concealed in accordance with the tacit policy of that institution, the social fabric is ripped into tattered rags.
It is only by threat of punitive damages that we can be sure policymakers will be cognizant of this grave social cost. It is only by threat of punitive damages that we can arouse policymakers from their dozing to eradicate policies such as this. If ever there were a need for punitive damages to encourage sound decisionmaking, this is it. It is time that the courts have a showdown with the “throw-down.”
CONCLUSION
In Newport, a case within the penumbra of section 1983, the Supreme Court rightly perceived that punitive damages were called for by neither the history nor policy of the statute. Sadly, I feel, the Court erred in issuing a holding exceeding the facts before it. Today I must live with the consequences of that holding and concur in the majority’s denial of punitive damages, in obedience to the mandate from my superiors.
The majority opinion reads Newport as holding that in an outrageous case punitive damages may be assessed against a city. If Newport could be read that way, I would have no problem in finding as a matter of law that the tortious conduct of the city of Houston was outrageous in the extreme and *1239would therefore affirm the case on the basis that the city is liable for punitive damages in an outrageous case. My difficulty and problem with the majority opinion is that I do not believe that Newport holds anything other than that there can never be an assessment of punitive damages against a city.
The very reason men and women come together in society is so that each person, through sacrifice of some individual freedom to the state, might better be able to pursue personal goals, protected by the state from violence by others. It is an outrage when agents of the state, rather than protecting the citizenry from violence, violate the right to life of a citizen. It is an unspeakable perversion, however, when the faceless minions of the state, the very ones charged with preventing and investigating murders, have a covert policy of concealing such a crime committed by one of their own, of perpetrating a fraud on any court that might investigate.
The role of the courts in society is a delicate one, that of a physician to the body politic. For a minor headache like Newport the aspirin of compensatory damages is a sufficient remedy. But the existence of a police policy of concealing police murder is a cancer in the vitals of society, not unlike the disease of the Klan in 1870’s. Congress has prescribed deterrence and punitive damages are our sharpest scalpel. How can we be true to our oaths when denied the use of the tools of our trade? Who will cure us now?
. It might be possible to read footnote 29 as allowing an exception where the electorate of a political unit votes affirmatively in favor of outrageous conduct. Even in that situation there would be some difficulty in assessing damages where a minority of the electorate voted against the outrageous conduct.
. “Damages awarded for punitive purposes, therefore are not sensibly assessed against the governmental entity itself.” 101 S.Ct. at 2760 (emphasis in original). On the alternative deterrence rationale for punitive damages, the Court concluded that “the deterrence rationale of § 1983 does not justify making punitive damages available against municipalities.” Id
. Cf. Thomas v. City of New Orleans, 687 F.2d 80, 84 n.2 (5th Cir. 1982) (dicta) (reading Newport as absolute bar to punitive damages against municipalities).
. The mayor and council considered Blood, Sweat and Tears to be a rock group, and such music was thought conducive to poor audience behavior.
. It was at this point the Court wrote footnote 29, which the majority read as allowing an exception. See supra at 1229. But see supra, p. 1231.
. The Court’s fear of damages causing loss of services in Newport seems less pressing here. For one thing the category of cases that would merit punitive damages seems rather small. See infra III.B.2. For another, if the deterrent mechanism actually works, there will be few instances for assessing damages.
. This form of analysis is agnostic as to the distribution of costs and benefits. Thus the Court’s rhetorical argument in Newport that punitive damages would unjustifiably enrich an already fully compensated plaintiff is inapplicable here. See, e.g., Coase, The Problem of Social Cost, 3 J.L. & Econ. 1 (1960); Calabresi & Melamed, Property Rules, Liability Rules and Inalienability: One View of the Cathedral, 85 Harv.L.Rev. 1089 (1972).
. The assumption of an economically rational decisionmaker, though often unjustified in the case of individuals, seems quite appropriate in the context of municipal policymaking.
. See, e.g., G. Calabresi, The Cost of Accidents, 68-129 (1970).
. In fact, the record revealed that at least one other case had not been before the courts. See Majority Opinion, supra at 1223.