CITY OF ST. LOUIS v. PRAPROTNIK, 485 U.S. 112 (1988)

Justice O’Connor announced the judgment of the Court and delivered an opinion, in which the Chief Justice Rehnquist, Justice White, and Justice Scalia join.

[1]This case calls upon us to define the proper legal standard for determining when isolated decisions by municipal officials or employees may expose the municipality itself to liability under 42 U.S.C. § 1983.

I

[2]The principal facts are not in dispute. Respondent James H. Praprotnik is an architect who began working for petitioner city of St. Louis in 1968. For several years, respondent consistently received favorable evaluations of his job performance, uncommonly quick promotions, and significant increases in salary. By 1980, he was serving in a management-level city planning position at petitioner’s Community Development Agency (CDA).

[3]The Director of CDA, Donald Spaid, had instituted a requirement that the agency’s professional employees, including architects, obtain advance approval before taking on private clients. Respondent and other CDA employees objected to the requirement. In April 1980, respondent was suspended for 15 days by CDA’s Director of Urban Design, Charles Kindleberger, for having accepted outside employment without prior approval. Respondent appealed to the city’s Civil Service Commission, a body charged with reviewing employee grievances. Finding the penalty too harsh, the Commission reversed the suspension, awarded respondent back pay, and directed that he be reprimanded for having failed to secure a clear understanding of the rule.

[4]The Commission’s decision was not well received by respondent’s supervisors at CDA. Kindleberger later testified that he believed respondent had lied to the Commission, and that Spaid was angry with respondent.

[5]Respondent’s next two annual job performance evaluations were markedly less favorable than those in previous years. In discussing one of these evaluations with respondent, Kindleberger apparently mentioned his displeasure with respondent’s 1980 appeal to the Civil Service Commission. Respondent appealed both evaluations to the Department of Personnel. In each case, the Department ordered partial relief and was upheld by the city’s Director of Personnel or the Civil Service Commission.

[6]In April 1981, a new mayor came into office, and Donald Spaid was replaced as Director of CDA by Frank Hamsher. As a result of budget cuts, a number of layoffs and transfers significantly reduced the size of CDA and of the planning section in which respondent worked. Respondent, however was retained.

[7]In the spring of 1982, a second round of layoffs and transfers occurred at CDA. At that time, the city’s Heritage and Urban Design Division (Heritage) was seeking approval to hire someone who was qualified in architecture and urban planning. Hamsher arranged with the Director of Heritage, Henry Jackson, for certain functions to be transferred from CDA to Heritage. This arrangement, which made it possible for Heritage to employ a relatively high-level “city planning manager,” was approved by Jackson’s supervisor, Thomas Nash. Hamsher then transferred respondent to Heritage to fill this position.

[8]Respondent objected to the transfer, and appealed to the Civil Service Commission. The Commission declined to hear the appeal because respondent had not suffered a reduction in his pay or grade. Respondent then filed suit in federal district court, alleging that the transfer was unconstitutional. The city was named as a defendant, along with Kindleberger, Hamsher, Jackson (whom respondent deleted from the list before trial), and Deborah Patterson, who had succeeded Hamsher at CDA.

[9]At Heritage, respondent became embroiled in a series of disputes with Jackson and Jackson’s successor, Robert Killen. Respondent was dissatisfied with the work he was assigned, which consisted of unchallenging clerical functions far below the level of responsibilities that he had previously enjoyed. At least one adverse personnel decision was taken against respondent, and he obtained partial relief after appealing that decision.

[10]In December 1983, respondent was laid off from Heritage. The layoff was attributed to a lack of funds, and this apparently meant that respondent’s supervisors had concluded that they could create two lower-level positions with the funds that were being used to pay respondent’s salary. Respondent then amended the complaint in his lawsuit to include a challenge to the layoff. He also appealed to the Civil Service Commission, but proceedings in that forum were postponed because of the pending lawsuit and have never been completed. Tr. Oral Arg. 31-32.

[11]The case went to trial on two theories: (1) that respondent’s First Amendment rights had been violated through retaliatory actions taken in response to his appeal of his 1980 suspension; and (2) that respondent’s layoff from Heritage was carried out for pretextual reasons in violation of due process. The jury returned special verdicts exonerating each of the three individual defendants, but finding the city liable under both theories. Judgment was entered on the verdicts, and the city appealed.

[12]A panel of the Court of Appeals for the Eighth Circuit found that the due process claim had been submitted to the jury on an erroneous legal theory and vacated that portion of the judgment. With one judge dissenting, however, the panel affirmed the verdict holding the city liable for violating respondent’s First Amendment rights. 798 F.2d 1168 (1986). Only the second of these holdings is challenged here.

[13]The Court of Appeals found that the jury had implicitly determined that respondent’s layoff from Heritage was brought about by an unconstitutional city policy. Id., at 1173. Applying a test under which a “policymaker” is one whose employment decisions are “final” in the sense that they are not subjected to de novo review by higher-ranking officials, the Court of Appeals concluded that the city could be held liable for adverse personnel decisions taken by respondent’s supervisors. Id., at 1173-1175. In response to petitioner’s contention that the city’s personnel policies are actually set by the Civil Service Commission, the Court of Appeals concluded that the scope of review before that body was too “highly circumscribed” to allow it fairly to be said that the Commission, rather than the officials who initiated the actions leading to respondent’s injury, were the “final authority” responsible for setting city policy. Id., at 1175.

[14]Turning to the question of whether a rational jury could have concluded that respondent had been injured by an unconstitutional policy, the Court of Appeals found that respondent’s transfer from CDA to Heritage had been “orchestrated” by Hamsher, that the transfer had amounted to a “constructive discharge,” and that the injury had reached fruition when respondent was eventually laid off by Nash and Killen. Id. at 1175-1176, and n.8. The court held that the jury’s verdict exonerating Hamsher and the other individual defendants could be reconciled with a finding of liability against the city because “the named defendants were not the supervisors directly causing the layoff, when the actual damages arose.” Id. at 1173, n.3. Cf. Los Angeles v. Heller, 475 U.S. 796 (1986).

[15]The dissenting judge relied on our decision in Pembaur v. Cincinnati, 475 U.S. 469 (1986). He found that the power to set employment policy for petitioner city of St. Louis lay with the mayor and aldermen, who were authorized to enact ordinances, and with the Civil Service Commission, whose function was to hear appeals from city employees who believed that their rights under the city’s Charter, or under applicable rules and ordinances, had not been properly respected. 798 F.2d at 1180. The dissent concluded that respondent had submitted no evidence proving that the mayor and aldermen, or the Commission, had established a policy of retaliating against employees for appealing from adverse personnel decisions. Id. at 1179-1181. The dissenting judge also concluded that, even if there were such a policy, the record evidence would not support a finding that respondent was in fact transferred or laid off in retaliation for the 1980 appeal from his suspension. Id. at 1181-1182.

[16]We granted certiorari, 479 U.S. 1029 (1987), and we now reverse.

* * * * *

[17]Two terms ago, in Pembaur, supra, we undertook to define more precisely when a decision on a single occasion may be enough to establish an unconstitutional municipal policy. Although the Court was unable to settle on a general formulation, Justice Brennan’s plurality opinion articulated several guiding principles. First, a majority of the Court agreed that municipalities may be held liable under § 1983 only for acts for which the municipality itself is actually responsible, “that is, acts which the municipality has officially sanctioned or ordered.” 475 U.S. at 480. Second, only those municipal officials who have “final policymaking authority” may by their actions subject the government to § 1983 liability. Id. at 483. Third, whether a particular official has “final policymaking authority” is a question of state law. Ibid. Fourth, the challenged action must have been taken pursuant to a policy adopted by the official or officials responsible under state law for making policy in that area of the city’s business. Id. at 482-483, and n.12.

[18]The Courts of Appeals have already diverged in their interpretations of these principles. Compare, for example, Williams v. Butler, 802 F.2d 296, 299-302 (CA8 1986) (en banc), cert. pending sub nom. City of Little Rock v. Williams, No. 86-1049, with Jett v. Dallas Independent School Dist., 798 F. 2d 748, 759-760 (CA5 1986) (dictum). Today, we set out again to clarify the issue that we last addressed in Pembaur.

B

[19]We begin by reiterating that the identification of policymaking officials is a question of state law. “Authority to make municipal policy may be granted directly by a legislative enactment or may be delegated by an official who possesses such authority, and of course, whether an official had final policymaking authority is a question of state law.” Pembaur v. Cincinnati, 475 U.S. at 483 (plurality opinion).^[1] Thus the identification of policymaking officials is not a question of federal law and it is not a question of fact in the usual sense. The States have extremely wide latitude in determining the form that local government takes, and local preferences have led to a profusion of distinct forms. Among the many kinds of municipal corporations, political subdivisions, and special districts of all sorts, one may expect to find a rich variety of ways in which the power of government is distributed among a host of different officials and official bodies. See generally C. RHYNE, THE LAW OF LOCAL GOVERNMENT OPERATIONS §§ 1.3-1.7 (1980). Without attempting to canvass the numberless factual scenarios that may come to light in litigation, we can be confident that state law (which may include valid local ordinances and regulations) will always direct a court to some official or body that has the responsibility for making law or setting policy in any given area of local government’s business.^[2]

[20]We are not, of course, predicting that state law will always speak with perfect clarity. We have no reason to suppose, however, that federal courts will face greater difficulties here than those that they routinely address in other contexts. We are also aware that there will be cases in which policymaking responsibility is shared among more than one official or body. In the case before us, for example, it appears that the mayor and aldermen are authorized to adopt such ordinances relating to personnel administration as are compatible with the City Charter. See St. Louis City Charter, art. XVIII, § 7(b), App. 62-63. The Civil Service Commission, for its part, is required to “prescribe … rules for the administration and enforcement of the provisions of this article, and of any ordinance adopted in pursuance thereof, and not inconsistent therewith.” § 7(a), App. 62. Assuming that applicable law does not make the decisions of the Commission reviewable by the mayor and aldermen, or vice versa, one would have to conclude that policy decisions made either by the mayor and aldermen or by the Commission would be attributable to the city itself. In any event, however, a federal court would not be justified in assuming that municipal policymaking authority lies somewhere other than where the applicable law purports to put it. And certainly there can be no justification for giving a jury the discretion to determine which officials are high enough in the government that their actions can be said to represent a decision of the government itself.

[21]As the plurality in Pembaur recognized, special difficulties can arise when it is contended that a municipal policymaker has delegated his policymaking authority to another official. 475 U.S. at 482-483, and n.12. If the mere exercise of discretion by an employee could give rise to a constitutional violation, the result would be indistinguishable from respondeat superior liability. If, however, a city’s lawful policymakers could insulate the government from liability simply by delegating their policymaking authority to others, § 1983 could not serve its intended purpose. It may not be possible to draw an elegant line that will resolve this conundrum, but certain principles should provide useful guidance.

[22]First, whatever analysis is used to identify municipal policymakers, egregious attempts by local government to insulate themselves from liability for unconstitutional policies are precluded by a separate doctrine. Relying on the language of § 1983, the Court has long recognized that a plaintiff may be able to prove the existence of a widespread practice that, although not authorized by written law or express municipal policy, is “so permanent and well settled as to constitute a custom or usage’ with the force of law.” Adickes v. S. H. Kress & Co., 398 U.S. 144, 167-168 (1970). That principle, which has not been affected by Monell or subsequent cases, ensures that most deliberate municipal evasions of the Constitution will be sharply limited.

[23]Second, as the Pembaur plurality recognized, the authority to make municipal policy is necessarily the authority to make final policy. 475 U.S. at 481-484. When an official’s discretionary decisions are constrained by policies not of that official’s making, those policies, rather than the subordinate’s departures from them, are the act of the municipality. Similarly, when a subordinate’s decision is subject to review by the municipality’s authorized policymakers, they have retained the authority to measure the official’s conduct for conformance with their policies. If the authorized policymakers approve a subordinate’s decision and the basis for it, their ratification would be chargeable to the municipality because their decision is final.

C

[24]Whatever refinements of these principles may be suggested in the future, we have little difficulty concluding that the Court of Appeals applied an incorrect legal standard in this case. In reaching this conclusion, we do not decide whether the First Amendment forbade the city from retaliating against respondent for having taken advantage of the grievance mechanism in 1980. Nor do we decide whether there was evidence in this record from which a rational jury could conclude either that such retaliation actually occurred or that respondent suffered any compensable injury from whatever retaliatory action may have been taken. Finally, we do not address petitioner’s contention that the jury verdict exonerating the individual defendants cannot be reconciled with the verdict against the city. Even assuming that all these issues were properly resolved in respondent’s favor, we would not be able to affirm the decision of the Court of Appeals.

[25]The city cannot be held liable under § 1983 unless respondent proved the existence of an unconstitutional municipal policy. Respondent does not contend that anyone in city government ever promulgated, or even articulated, such a policy. Nor did he attempt to prove that such retaliation was ever directed against anyone other than himself. Respondent contends that the record can be read to establish that his supervisors were angered by his 1980 appeal to the Civil Service Commission; that new supervisors in a new administration chose, for reasons passed on through some informal means, to retaliate against respondent two years later by transferring him to another agency; and that this transfer was part of a scheme that led, another year and a half later, to his lay off. Even if one assumes that all this was true, it says nothing about the actions of those whom the law established as the makers of municipal policy in matters of personnel administration. The mayor and aldermen enacted no ordinance designed to retaliate against respondent or against similarly situated employees. On the contrary, the city established an independent Civil Service Commission and empowered it to review and correct improper personnel actions. Respondent does not deny that his repeated appeals from adverse personnel decisions repeatedly brought him at least partial relief, and the Civil Service Commission never so much as hinted that retaliatory transfers or layoffs were permissible. Respondent points to no evidence indicating that the Commission delegated to anyone its final authority to interpret and enforce the following policy set out in article XVIII of the city’s Charter, § 2(a), App. 49:

“Merit and fitness. All appointments and promotions to positions in the service of the city and all measures for the control and regulation of employment in such positions, and separation therefrom, shall be on the sole basis of merit and fitness….”

[26]The Court of Appeals concluded that “appointing authorities,” like Hamsher and Killen, who had the authority to initiate transfers and layoffs, were municipal “policymakers.” The court based this conclusion on its findings (1) that the decisions of these employees were not individually reviewed for “substantive propriety” by higher supervisory officials; and (2) that the Civil Service Commission decided appeals from such decisions, if at all, in a circumscribed manner that gave substantial deference to the original decisionmaker. 798 F.2d at 1174-1175. We find these propositions insufficient to support the conclusion that Hamsher and Killen were authorized to establish employment policy for the city with respect to transfers and layoffs. To the contrary, the City Charter expressly states that the Civil Service Commission has the power and the duty:

“To consider and determine any matter involved in the administration and enforcement of this [Civil Service] article and the rules and ordinances adopted in accordance therewith that may be referred to it for decision by the director [or personnel], or on appeal by any appointing authority, employee, or taxpayer of the city, from any act of the director or of any appointing authority. The decision of the commission in all such matters shall be final, subject, however, to any right of action under any law of the state or of the United States.” St. Louis City Charter, art. XVIII, § 7(d), App. 63.

[27]This case therefore resembles the hypothetical example in Pembaur: “If [city] employment policy was set by the [Mayor and Aldermen and by the Civil Service Commission], only [those] bod[ies’] decisions would provide a basis for [city] liability. This would be true even if the [mayor and aldermen and the Commission] left the [appointing authorities] discretion to hire and fire employees and [they] exercised that discretion in an unconstitutional manner ” 475 U.S. at 483, n. 12. A majority of the Court of Appeals panel determined that the Civil Service Commission’s review of individual employment actions gave too much deference to the decisions of appointing authorities like Hamsher and Killen. Simply going along with discretionary decisions made by one’s subordinates, however, is not a delegation to them of the authority to make policy. It is equally consistent with a presumption that the subordinates are faithfully attempting to comply with the policies that are supposed to guide them. It would be a different matter if a particular decision by a subordinate was cast in the form of a policy statement and expressly approved by the supervising policymaker. It would also be a different matter if a series of decisions by a subordinate official manifested a “custom or usage” of which the supervisor must have been aware. See supra, at 127. In both those cases, the supervisor could realistically be deemed to have adopted a policy that happened to have been formulated or initiated by a lower-ranking official. But the mere failure to investigate the basis of a subordinate’s discretionary decisions does not amount to a delegation of policymaking authority, especially where (as here) the wrongfulness of the subordinate’s decision arises from a retaliatory motive or other unstated rationale. In such circumstances, the purposes of § 1983 would not be served by treating a subordinate employees’ decision as if it were a reflection of municipal policy.

[28]Justice Brennan’s opinion, concurring in the judgment, finds implications in our discussion that we do not think necessary or correct. See post at 142-147. We nowhere say or imply, for example, that “a municipal charter’s precatory admonition against discrimination or any other employment practice not based on merit and fitness effectively insulates the municipality from any liability based on acts inconsistent with that policy.” Post, at 145, n.7. Rather, we would respect the decisions, embodied in state and local law, that allocated policymaking authority among particular individuals and bodies. Refusals to carry out stated policies could obviously help to show that a municipality’s actual policies were different from the ones that had been announced. If such a showing were made, we would be confronted with a different case than the one we decide today.

[29]Nor do we believe that we have left a “gaping hole” in § 1983 that needs to be filled with the vague concept of “de facto final policymaking authority.” Post at 144. Except perhaps as a step towards overruling Monell and adopting the doctrine of respondeat superior, ad hoc searches for officials possessing such “de facto” authority would serve primarily to foster needless unpredictability in the application of § 1983.

* * * * *

[30]Accordingly, the decision of the Court of Appeals is reversed, and the case is remanded for further proceedings consistent with this opinion.

It is so ordered.

Justice Kennedy took no part in the consideration or decision of this case.

Justice Brennan, with whom Justice Marshall and Justice Blackmun join, concurring.

[31]Despite its somewhat confusing procedural background, this case at bottom presents a relatively straightforward question: whether respondent’s supervisor at the Community Development Agency, Frank Hamsher, possessed the authority to establish final employment policy for the city of St. Louis such that the city can be held liable under 42 U.S.C. § 1983 for Hamsher’s allegedly unlawful decision to transfer respondent to a dead-end job. Applying the test set out two Terms ago by the plurality in Pembaur v. Cincinnati, 475 U.S. 469 (1986), I conclude that Hamsher did not possess such authority and I therefore concur in the Court’s judgment reversing the decision below. I write separately, however, because I believe that the commendable desire of today’s plurality to “define more precisely when a decision on a single occasion may be enough” to subject a municipality to § 1983 liability, ante, at 123, has led it to embrace a theory of municipal liability that is both unduly narrow and unrealistic, and one that ultimately would permit municipalities to insulate themselves from liability for the acts of all but a small minority of actual city policymakers.

* * * * *

[32]In my view, Pembaur controls this case. As an “appointing authority,” Hamsher was empowered under the City Charter to initiate lateral transfers such as the one challenged here, subject to the approval of both the Director of Personnel and the appointing authority of the transferee agency. The Charter, however, nowhere confers upon agency heads any authority to establish city policy, final or otherwise, with respect to such transfers. Thus, for example, Hamsher was not authorized to promulgate binding guidelines or criteria governing how or when lateral transfers were to be accomplished. Nor does the record reveal that he in fact sought to exercise any such authority in these matters. There is no indication, for example, that Hamsher ever purported to institute or announce a practice of general applicability concerning transfers. Instead, the evidence discloses but one transfer decision—the one involving respondent—which Hamsher ostensibly undertook pursuant to a city-wide program of fiscal restraint and budgetary reductions. At most, then, the record demonstrates that Hamsher had the authority to determine how best to effectuate a policy announced by his superiors, rather than the power to establish that policy. Like the hypothetical Sheriff in Pembaur’s footnote 12, Hamsher had discretionary authority to transfer CDA employees laterally; that he may have used this authority to punish respondent for the exercise of his First Amendment rights does not, without more, render the city liable for respondent’s resulting constitutional injury. The court below did not suggest that either Killen or Nash, who together orchestrated respondent’s ultimate layoff, shared Hamsher’s constitutionally impermissible animus. Because the court identified only one unlawfully motivated municipal employee involved in respondent’s transfer and layoff, and because that employee did not possess final policymaking authority with respect to the contested decision, the city may not be held accountable for any constitutional wrong respondent may have suffered.

III

[33]These determinations, it seems to me, are sufficient to dispose of this case, and I therefore think it unnecessary to decide, as the plurality does, who the actual policymakers in St. Louis are. I question more than the mere necessity of these determinations, however, for I believe that in the course of passing on issues not before us, the plurality announces legal principles that are inconsistent with our earlier cases and unduly restrict the reach of § 1983 in cases involving municipalities.

[34]The plurality begins its assessment of St. Louis’ power structure by asserting that the identification of policymaking officials is a question of state law, by which it means that the question is neither one of federal law nor of fact, at least “not in the usual sense.” See ante, at 124. Instead, the plurality explains, courts are to identify municipal policymakers by referring exclusively to applicable state statutory law. Ante, at 124. Not surprisingly, the plurality cites no authority for this startling proposition, nor could it, for we have never suggested that municipal liability should be determined in so formulaic and unrealistic a fashion. In any case in which the policymaking authority of a municipal tortfeasor is in doubt, state law will naturally be the appropriate starting point, but ultimately the factfinder must determine where such policymaking authority actually resides, and not simply “where the applicable law purports to put it.” Ante, at 126. As the plurality itself acknowledges, local governing bodies may take myriad forms. We in no way slight the dignity of municipalities by recognizing that in not a few of them real and apparent authority may diverge, and that in still others state statutory law will simply fail to disclose where such authority ultimately rests. Indeed, in upholding the Court of Appeals’ determination in Pembaur that the County Prosecutor was a policymaking official with respect to county law enforcement practices, a majority of this Court relied on testimony which revealed that the County Sheriff’s office routinely forwarded certain matters to the Prosecutor and followed his instructions in those areas. See 475 U.S. at 485; ibid. (White, J., concurring); id. at 491 (O’Connor, J., concurring). While the majority splintered into three separate camps on the ultimate theory of municipal liability, and the case generated five opinions in all, not a single member of the Court suggested that reliance on such extra-statutory evidence of the county’s actual allocation of policymaking authority was in any way improper. Thus, although I agree with the plurality that juries should not be given open-ended “discretion to determine which officials are high enough in the government that their actions can be said to represent a decision of the government itself,” ante, at 126 (emphasis added), juries can and must find the predicate facts necessary to a determination of whether a given official possesses final policymaking authority. While the jury instructions in this case were regrettably vague, the plurality’s solution tosses the baby out with the bath water. The identification of municipal policymakers is an essentially factual determination “in the usual sense,” and is therefore rightly entrusted to a properly instructed jury.

[35]Nor does the “custom or usage” doctrine adequately compensate for the inherent inflexibility of a rule that leaves the identification of policymakers exclusively to state statutory law. That doctrine, under which municipalities and States can be held liable for unconstitutional practices so well settled and permanent that they have the force of law, see Adickes v. Kress & Co., 398 U.S. at 167, has little if any bearing on the question whether a city has delegated de facto final policymaking authority to a given official. A city practice of delegating final policymaking authority to a subordinate or mid-level official would not be unconstitutional in and of itself, and an isolated unconstitutional act by an official entrusted with such authority would obviously not amount to a municipal “custom or usage.” Under Pembaur, of course, such an isolated act should give rise to municipal liability. Yet a case such as this would fall through the gaping hole the plurality’s construction leaves in § 1983, because the state statutory law would not identify the municipal actor as a policymaking official, and a single constitutional deprivation, by definition, is not a well settled and permanent municipal practice carrying the force of law.^[3]

[36]For these same reasons, I cannot subscribe to the plurality’s narrow and overly rigid view of when a municipal official’s policymaking authority is “final.” Attempting to place a gloss on Pembaur’s finality requirement, the plurality suggests that whenever the decisions of an official are subject to some form or review—however limited—that official’s decisions are nonfinal. Under the plurality’s theory, therefore, even where an official wields policymaking authority with respect to a challenged decision, the city would not be liable for that official’s policy decision unless reviewing officials affirmatively approved both the “decision and the basis for it.” Ante, at 127. Reviewing officials, however, may as a matter of practice never invoke their plenary oversight authority, or their review powers may be highly circumscribed. See n.4 supra. Under such circumstances, the subordinate’s decision is in effect the final municipal pronouncement on the subject. Certainly a § 1983 plaintiff is entitled to place such considerations before the jury, for the law is concerned not with the niceties of legislative draftsmanship but with the realities of municipal decisionmaking, and any assessment of a municipality’s actual power structure is necessarily a factual and practical one.^[4]

[37]Accordingly, I cannot endorse the plurality’s determination, based on nothing more than its own review of the City Charter, that the mayor, the aldermen, and the CSC are the only policymakers for the city of St. Louis. While these officials may well have policymaking authority, that hardly ends the matter; the question before us is whether the officials responsible for respondent’s allegedly unlawful transfer were final policymakers. As I have previously indicated, I do not believe that CDA Director Frank Hamsher possessed any policymaking authority with respect to lateral transfers and thus I do not believe that his allegedly improper decision to transfer respondent could, without more, give rise to municipal liability. Although the plurality reaches the same result, it does so by reasoning that because others could have reviewed the decisions of Hamsher and Killen, the latter officials simply could not have been final policymakers.

[38]This analysis, however, turns a blind eye to reality, for it ignores not only the lower court’s determination, nowhere disputed, that CSC review was highly circumscribed and deferential, but that in this very case the Commission refused to judge a propriety of Hamsher’s transfer decision because a lateral transfer was not an “adverse” employment action falling within its jurisdiction. Nor does the plurality account for the fact that Hamsher’s predecessor, Donald Spaid, promulgated what the city readily acknowledges was a binding policy regarding secondary employment;^[5] although the CSC ultimately modified the sanctions respondent suffered as a result of his apparent failure to comply with that policy, the record is devoid of any suggestion that the Commission reviewed the substance or validity of the policy itself. Under the plurality’s analysis, therefore, even the hollowest promise of review is sufficient to divest all city officials save the mayor and governing legislative body of final policymaking authority. While clarity and ease of application may commend such a rule, we have remained steadfast in our conviction that Congress intended to hold municipalities accountable for those constitutional injuries inflicted not only by their lawmakers, but “by those whose edicts or acts may fairly be said to represent official policy.” Monell, 436 U. S., at 694. Because the plurality’s mechanical “finality” test is fundamentally at odds with the pragmatic and factual inquiry contemplated by Monell, I cannot join what I perceive to be its unwarranted abandonment of the traditional factfinding process in § 1983 actions involving municipalities.

Thus in this case, a policy prohibiting lateral transfers for unconstitutional or discriminatory reasons would not shield the city from liability if an official possessing final policymaking authority over such transfers acted in violation of the prohibition, because the CSC would lack jurisdiction to review the decision and thus could not enforce the city policy. Where as here, however, the official merely possesses discretionary authority over transfers, the city policy is irrelevant, because the official’s actions cannot subject the city to liability in any event.

[39]Finally, I think it necessary to emphasize that despite certain language in the plurality opinion suggesting otherwise, the Court today need not and therefore does not decide that a city can only be held liable under § 1983 where the plaintiff “prov[es] the existence of an unconstitutional municipal policy.” See ante, at 128. Just last Term, we left open for the second time the question whether a city can be subjected to liability for a policy that, while not unconstitutional in and of itself, may give rise to constitutional deprivations. See Springfield v. Kibbe, 480 U.S. 257 (1987); see also Oklahoma City v. Tuttle, 471 U.S. 808 (1985). That question is certainly not presented by this case, and nothing we say today forecloses its future consideration.

* * * * *

Justice Stevens, dissenting.

[40]If this case involved nothing more than a personal vendetta between a municipal employee and his superiors, it would be quite wrong to impose liability on the City of St. Louis. In fact, however, the jury found that top officials in the City administration, relying on pretextual grounds, had taken a series of retaliatory actions against respondent because he had testified truthfully on two occasions, one relating to personnel policy and the other involving a public controversy of importance to the Mayor and the members of his cabinet. No matter how narrowly the Court may define the standards for imposing liability upon municipalities in § 1983 litigation, the judgment entered by the District Court in this case should be affirmed.

* * * * *

[41]Both Pembaur and the plurality and concurring opinions today acknowledge that a high official who has ultimate control over a certain area of city government can bind the City through his unconstitutional actions even though those actions are not in the form of formal rules or regulations. See Pembaur v. Cincinnati, supra, at 479-481; ante, at 123 (plurality), at 139-140 (concurrence). Although the Court has explained its holdings by reference to the nonstatutory term “policy,” it plainly has not embraced the standard understanding of that word as covering a rule of general applicability. Instead it has used that term to include isolated acts not intended to be binding over a class of situations. But when one remembers that the real question in cases such as this is not “what constitutes City policy?” but rather “when should a City be liable for the acts of its agents?,” the inclusion of single acts by high officials make sense, for those acts bind a municipality in a way that the misdeeds of low officials do not.

[42]Every act of a high official constitutes a kind of “statement” about how similar decisions will be carried out; the assumption is that the same decision would have been made, and would again be made, across a class of cases. Lower officials do not control others in the same way. Since their actions do not dictate the responses of various subordinates, those actions lack the potential of controlling governmental decisionmaking; they are not perceived as the actions of the city itself. If a County police officer had broken down Dr. Pembaur’s door on the officer’s own initiative, this would have been seen as the action of an overanxious officer and would not have sent a message to other officers that similar actions would be countenanced. One reason for this is that the County Prosecutor himself could step forward and say “that was wrong”; when the County Prosecutor authorized the action himself, only a self-correction would accomplish the same task, and until such time his action would have County-wide ramifications. Here, the Mayor, those working for him, and the agency heads are high-ranking officials; accordingly, we must assume that their actions have City-wide ramifications, both through their similar response to a like class of situations, and through the response of subordinates who follow their lead.^[6]

[43]Just as the actions of high-ranking and low-ranking municipal employees differ in nature, so do constitutional torts differ. An illegal search (Pembaur) or seizure (Tuttle) is quite different from a firing without due process (Owen); the retaliatory personnel action involved in today’s case is in still another category. One thing that the torts in Pembaur, Tuttle, and Owen had in common is that they occurred “in the open”; in each of those cases, the ultimate judgment of unconstitutionality was based on whether undisputed events (the breaking-in in Pembaur, the shooting in Tuttle, the firing in Owen) comported with accepted constitutional norms. But the typical retaliatory personnel action claim pits one story against another; although everyone admits that the transfer and discharge of respondent occurred, there is sharp, and ultimately central, dispute over the reasons—the motivation—behind the actions. The very nature of the tort is to avoid a formal process. Owens’ relevance should thus be clear. For if the Court is willing to recognize the existence of municipal policy in a non-rule case as long as high enough officials engaged in a formal enough process, it should not deny the existence of such a policy merely because those same officials act “underground,” as it were. It would be a truly remarkable doctrine for this Court to recognize municipal liability in an employee discharge case when high officials are foolish enough to act through a “formal process,” but not when similarly high officials attempt to avoid liability by acting on the pretext of budgetary concerns, which is what the jury found based on the evidence presented at trial.

[44]Thus, holding St. Louis liable in this case is supported by both Pembaur and Owen. We hold a municipality liable for the decisions of its high officials in large part because those decisions, by definition, would be applied across a class of cases. Just as we assume in Pembaur that the County Prosecutor (or his subordinates) would issue the same break-down-the-door order in similar cases, and just as we assume in Owen that the City Council (or those following its lead) would fire an employee without notice of reasons or opportunity to be heard in similar cases, so too must we assume that whistleblowers like respondent would be dealt with in similar retaliatory fashion if they offend the Mayor, his staff, and relevant agency heads, or if they offend those lower-ranking officials who follow the example of their superiors. Furthermore, just as we hold a municipality liable for discharging an employee without due process when its city council acts formally—for a due process violation is precisely the type of constitutional tort that a city council might commit when it acts formally—so too must we hold a municipality liable for discharging an employee in retaliation against his public speech when similarly high officials act informally—for a first amendment retaliation tort is precisely the type of constitutional tort that high officials might commit when they act in concert and informally.

[45]Whatever difficulties the Court may have with binding municipalities on the basis of the unconstitutional conduct of individuals, it should have no such difficulties binding a city when many of its high officials—including officials directly under the mayor, agency heads, and possibly the mayor himself—cooperate to retaliate against a whistleblower for the exercise of his First Amendment rights.^[7]

I would affirm the judgment of the Court of Appeals.

City of St. Louis v. Praprotnik – Audio and Transcript of Oral Argument

Footnotes

1. Unlike Justice Brennan, we would not replace this standard with a new approach in which state law becomes merely “an appropriate starting point” for an “assessment of a municipality’s actual power structure.” Post, at 143, 145. Municipalities cannot be expected to predict how courts or juries will assess their “actual power structures,” and this uncertainty could easily lead to results that would be hard in practice to distinguish from the results of a regime governed by the doctrine of respondeat superior. It is one thing to charge a municipality with responsibility for the decisions of officials invested by law, or by a “custom or usage” having the force of law, with policymaking authority. It would be something else, and something inevitably more capricious, to hold a municipality responsible for every decision that is perceived as “final” through the lens of a particular factfinder’s evaluation of the city’s “actual power structure.” ↵

    

2. Justice Stevens, who believes that Monell incorrectly rejected the doctrine of respondeat superior, suggests a new theory that reflects his perceptions of the congressional purposes underlying § 1983. See post at 148, n.1. This theory would apparently ignore state law and distinguish between “high” officials and “low” officials on the basis of an independent evaluation of the extent to which a particular official’s actions have “the potential of controlling governmental decisionmaking,” or are “perceived as the actions of the city itself.” Post at 171. Whether this evaluation would be conducted by judges or juries, we think, the legal test is too imprecise to hold much promise of consistent adjudication or principled analysis. We can see no reason, except perhaps a desire to come as close as possible to respondeat superior without expressly adopting that doctrine, that could justify introducing such unpredicatability into a body of law that is already so difficult. As Justice Stevens acknowledges, see post at 148, n.1, this Court has repeatedly rejected his interpretation of Congress’ intent. We have held that Congress intended to hold municipalities responsible under § 1983 only for the execution of official policies and customs, and not for injuries inflicted solely by employees or agents. See, e.g., Monell v. New York City Dept. of Social Services, 436 U.S. at 658, 694 (1978); Pembaur v. Cincinnati, 475 U.S. 469, 478-480 (1986). Like the Pembaur plurality, we think it is self-evident that official policies can only be adopted by those legally charged with doing so. See supra, at 124, and n.1. We are aware of nothing in § 1983 or its legislative history, and Justice Stevens points to nothing, that would support the notion that unauthorized acts of subordinate employees are official policies because they may have the “potential” to become official policies or may be “perceived as” official policies. Accordingly, we conclude that Justice Stevens’ proposal is without a basis in the law. ↵

    

3. Indeed, the plurality appears to acknowledge as much when it explains that the “custom or usage” doctrine will forestall “egregious attempts by local government to insulate themselves from liability for unconstitutional policies,” and that “most deliberate municipal evasions of the Constitution will be sharply limited.” Ante, at 127 (emphases added). Congress, however, did not enact § 1983 simply to provide redress for “most” constitutional deprivations, nor did it limit the statute’s reach only to those deprivations that are truly “egregious.” ↵

    

4. The plurality also asserts that “when an official’s discretionary decisions are constrained by policies not of that official’s making, those policies, rather than the subordinate’s departures from them, are the act of the municipality.” Ante, at 127. While I have no quarrel with such a proposition in the abstract, I cannot accept the plurality’s apparent view that a municipal charter’s precatory admonition against discrimination or any other employment practice not based on merit and fitness effectively insulates the municipality from any liability based on acts inconsistent with that policy. Again, the relevant inquiry is whether the policy in question is actually and effectively enforced through the city’s review mechanisms. ↵

    

5. Although the plurality is careful in its discussion of the facts to label Director Spaid’s directive a “requirement” rather than a “policy,” the city itself draws no such fine semantic distinctions. Rather, it states plainly that Spaid “promulgated a secondary employment’ policy that sought to control outside employment by CDA architects,” and that “[respondent] resented the policy.” Brief for Petitioner 2-3 (emphasis added). ↵

    

6. That high officials may bind a municipality in ways that low officials may not should not surprise, for the pyramidal structure of authority pervades the law. For instance, the law of agency distinguishes between a general agent and a special agent; the former is “authorized to conduct a series of transactions involving a continuity of service,” while the latter is “authorized to conduct a single transaction or a series of transactions not involving continuity of service.” RESTATEMENT (SECOND) OF AGENCY §§ 3(1), (2) (1958). The distinction matters because only a general agent “subjects his principal to liability for acts done on his account which usually accompany or are incidental to transactions which the agent is authorized to conduct if, although they are forbidden by the principal, the other party reasonably believes that the agent is authorized to do them and has no notice that he is not so authorized.” Id. § 161. A special agent, to the contrary, “has no power to bind his principal by contracts or conveyances which he is not authorized or apparently authorized to make,” with some exceptions. Id. § 161A. A general agent thus binds his principal even through unauthorized acts precisely because those dealing with him perceive him as possessing broad authority to act on behalf of his principal. A special agent, possessing and known to possess only limited authority, cannot bind his principal for unauthorized acts because those dealing with him are on notice that his authority extends only so far. Likewise, a high municipal official can bind his principal (the city) for unauthorized actions because others—both lower officials and members of the public with whom he deals—perceive him as acting with broad authority and rely upon his actions in organizing their own behavior. The distinction between general agents and special agents has a firm “basis in the law.” See ante, at 125, n.2 (plurality). ↵

    

7. The plurality incorrectly claims that I have suggested “a new theory” for determining when a municipality should be bound by the acts of its agents. Ante, at 125, n.2. As both the plurality and the concurrence recognize, a municipality, like any institution, can only act through the agency of human beings. By holding that isolated actions of high officials may give rise to municipal liability, see, e.g., Owenv. City of Independence; Pembaur v. Cincinnati, the Court has indicated that the mere status of City officials matters in determining whether the City may be held liable for the officials’ actions. The argument of both the plurality and the concurrence that this principle should be applied only in the particular area of government that the erring official controls is unpersuasive, given the multifarious ways in which governmental agents may inflict constitutional harm. This case is a perfect example of why the “area-by-area” approach will not do; personnel actions may be taken in response to an employee’s protected speech by a number of high officials, none of whom possesses specific authority over “personnel” policy. Nevertheless, simply by virtue of their high rank, their actions may influence the actions of other municipal officials. It is that kind of influence that provides the common thread binding Monell and the later § 1983 municipal liability cases. In short, what the Court has characterized as “a new theory” is actually a way of understanding our precedents that will permit a judge to explain to a jury that “policy” means nothing if not “influence,” and that while the isolated gunshot of an errant police officer would not influence his colleagues, see Oklahoma City v. Tuttle, adverse personnel actions taken by a City’s highest officials in response to an employee’s Civil Service Commission appeals and his public testimony would set an example for other, lower officials to follow. ↵

Notes on City of St. Louis v. Praprotnik

1. Do a majority of the Justices agree that a single decision by an official with “final authority” over the decision may render the municipality liable under Section 1983?
2. Do a majority of the Justices agree how to determine which officials have “final authority”?
3. May a municipality avoid liability by providing that no action by any local official is final until approved by the mayor, and then ensuring that the mayor does not actually review any decisions? Compare Flanagan v. Munger, 890 F.2d 1557, 1568-69 (10th Cir. 1989) (Chief of Police is final policymaker where meaningful administrative review is illusory) with Ware v. Unified School Dist. No. 492, 902 F.2d 815, 818-19 (10th Cir. 1990) (School Superintendent is not final policymaker where school board retained and exercised authority to review decisions of Superintendent). How does Justice O’Connor’s approach in Praprotnik differ from Justice Brennan’s opinion as to this scenario?
4. One year after Praprotnik, the Supreme Court issued its decision in Jett v. Dallas Independent School District, 491 U.S. 701 (1989). Jett arose out of a suit by a white athletic director under 42 U.S.C. §§ 1981 and 1983 alleging discrimination in his reassignment to a position of lesser prestige. After holding that a municipality may not be held vicariously liable under § 1981, the Court remanded the case for a determination of whether the school district could be liable on the basis that the school Superintendent had final policymaking authority in the area of employee reassignments. The Court explained how the lower court was to resolve this issue:
   

   As with other questions of state law relevant to the application of federal law, the identification of those officials whose decisions represent the official policy of the local government unit is itself a legal question to be resolved by the trial judge before the case is submitted to the jury … [after] reviewing the relevant legal materials, including state and local positive law, as well as “‘custom and usage’ having the force of law,” Praprotnik, supra, at 124, n.1, 108 S. Ct. at 924, n.1, the trial judge must identify those officials who speak with final policymaking authority for the local government action alleged to have caused the particular constitutional or statutory violation at issue.

   Jett, 491 U.S. at 736. Did the Jett Court endorse the opinion of Justice O’Connor or the test proposed by Justice Brennan in Praprotnik? In Worsham v. City of Pasadena, 881 F.2d 1336, 1343-44 (5th Cir. 1989), Judge Goldberg’s dissenting opinion interpreted Jett as follows:

Jett clarifies that … [b]y proving a “custom or usage,” a plaintiff may demonstrate as a matter of fact that an official is invested with final policymaking authority….

* * * * *

Jett also characterizes the “final policymaker” determination as a threshold question of law for the trial judge. Because facts concerning custom or usage may play a role in such a determination, the trial judge, after Jett, is empowered to resolve factual disputes in its threshold inquiry. Jett thus envisions a role for the trial judge in this context similar, for example, to the role a judge plays in determining admissibility of certain evidence, or whether a matter is of public concern in the First Amendment context. These issues may involve disputed factual issues that the judge is empowered to resolve.

5. The Supreme Court again granted certiorari to determine when a local government official is a policymaker in Swint v. Chambers County Commission, 514 U.S. 35 (1995). Swint arose out of a Section 1983 action complaining of police raids on a nightclub. Under Mitchell v. Forsyth, 472 U.S. 511 (1985), the court of appeals had jurisdiction over the individual police officers’ interlocutory appeal from the district court’s denial of their motion for summary judgment seeking qualified immunity. The appellate court also reviewed and reversed the denial of summary judgment to the County Commission, which had been based on the district court’s finding that the sheriff was a policymaker for the county for the purpose of authorizing police raids.The Supreme Court, however, never reached the issue of whether the sheriff was a policymaker for the county, for the Court held that the district court’s denial of summary judgment to the County Commission was not an appealable order. First, unlike the rejection of the individual officers’ claimed entitlement to qualified immunity, the district court’s repudiation of the County Commission’s contention that the sheriff was not a policymaker was not an immediately appealable collateral order under 28 U.S.C. § 1291:
   

   The Commission’s assertion that Sheriff Morgan is not its policymaker does not rank, under our decisions, as an immunity from suit. Instead, the plea ranks as a “mere defense to liability.” Mitchell, 472 U.S. at 526. An erroneous ruling on liability may be reviewed effectively on appeal from final judgment. Therefore, the order denying the County Commission’s summary judgment motion was not an appealable collateral order.

   514 U.S. at 43. The Court likewise ruled that the court of appeals did not have pendent appellate jurisdiction to review denial of the county’s summary judgment motion:

   The parties do not contend that the District Court’s decision to deny the Chambers County Commission’s summary judgment motion was inextricably entwined with the court’s decision to deny the individual defendants’ qualified immunity motions, or that review of the former decision was necessary to ensure meaningful review of the latter…. Nor could the parties so argue. The individual defendants’ qualified immunity turns on whether they violated clearly established federal law; the County Commission’s liability turns on the allocation of law enforcement power in Alabama.

   514 U.S. at 51.

6. The Supreme Court took its next stab at resolving the policymaker debate in McMillian v. Monroe County, Alabama, 520 U.S. 781 (1997), a Section 1983 action arising out of the Sheriff of Monroe County’s concealment of exculpatory evidence. As a result of the sheriff’s actions, the plaintiff served six years on death row before his murder conviction was reversed. The parties agreed that the sheriff had final policymaking authority in matters of law enforcement; they differed, however, as to whether the sheriff was a policymaker for the State of Alabama or for the county.In a 5-4 decision, the Court held that for purposes of the particularized activity of law enforcement, the sheriff represented the State rather than the county. Deeming the issue to rest upon “the definition of the official’s function under relevant state law,” 520 U.S. at 785, the majority observed that the Alabama Constitution had been amended to add sheriffs to the list of officials designated as members of the executive department and transferred the power to impeach sheriffs from county courts to the State Supreme Court. The Court also referenced provisions of the Alabama Code that conferred upon state judges the authority to supervise sheriffs and gave sheriffs the authority to enforce state criminal law in their counties. However, the Court also entertained the plaintiff’s evidence that the county’s insurance policy arguably covered some of the claims, but concluded that the insurance policy displayed uncertainty as to whether the courts would consider the sheriff a county policymaker. 520 U.S. at 792 n.7.Justice Ginsburg railed against the majority’s reliance upon provisions of state law labeling sheriffs as officials of the executive department:
   

   [D]esignations Alabama attaches to sheriffs in its laws and decisions are not dispositive of a court’s assessment of Sheriff Tate’s status for § 1983 purposes. Cf. Regents of Univ. of Cal. v. Doe, 519 U.S. n.5 (1997) (slip op., at 4-5, n.5); Howlett v. Rose, 496 U.S. 356, 376 (1990) (defenses to § 1983 actions are questions of federal law); Martinez v. California, 444 U.S. 277, 284 and n.8 (1980) (state law granting immunity to parole officials does not control question whether such officers have immunity under § 1983). If a State’s designation sufficed to answer the federal question at issue, “States would then be free to nullify for their own people the legislative decisions that Congress has made on behalf of all the People.” Howlett, 496 U.S. at 383.

   520 U.S. at 798-99. Justice Ginsburg noted that Alabama sheriffs are elected by residents of the county, have power to execute law enforcement policies only within the geographic boundaries of their respective counties, are paid by the county, and have their offices provided, furnished and equipped by the county, all of which is directly contrary to the categorization of sheriffs as members of the executive department.

7. Is every action by a local governmental official with final authority municipal “policy” within the meaning of Section 1983? In Auriemma v. Rice, 957 F.2d 397 (7th Cir. 1992), the court of appeals held that even if the Superintendent of Police had final authority over promotion decisions, the alleged discriminatory actions of the Superintendent in promoting and demoting senior officials did not constitute city policy:
   

   If it were enough to point to the agent whose act was the final one in a particular case, we would have vicarious liability … That a particular agent is the apex of a bureaucracy makes the decision “final” but does not forge a link between “finality” and “policy.” Unless today’s decision ought to govern tomorrow’s case under a law or custom with the force of law, it cannot be said to carry out the municipality’s policy.

   * * * * *

   “[R]esponsibility for making law or setting policy”—the objective under Praprotnik of our search through local law—is authority to adopt rules for the conduct of government. Authority to make a final decision need not imply authority to establish rules. In Chicago, it does not. The Superintendent of Police in Chicago had no power to countermand the statutes regulating the operation of the department. The chief has “[c]omplete authority to administer the department in a manner consistent with the ordinances of the city, the laws of the state, and the rules and regulations of the police board.” …. If, in the course of selecting senior staff, Rice discriminated on account of race or policy, he violated rather than implemented the policy of Chicago.

   Id. at 400-01. How is the court to determine which decisionmakers with final authority are policymakers? May a municipality shield itself from Section 1983 liability for constitutional violations by enacting an ordinance that prohibits all officials with final authority from taking any actions that contravene the Constitution? Compare Greenboro Prof. Fire Fighters Ass’n v. Greensboro, 64 F.3d 962, 965 (4th Cir. 1995) (“When a final decision by an employee implements municipal policy, then municipal liability may follow. But if a final decision does not implement municipal policy, or is contrary to it, then it is not imputable to the municipality.”) with Gonzalez v. Ysleta Independent School Dist., 996 F.2d 745, 754 (5th Cir. 1993) (“[T]he existence of a well-established, officially adopted policy will not insulate the municipality from liability where the policy-maker herself departs from these formal rules.”).

8. Does a determination that a local government actor had final authority preclude the municipality from resorting to Parratt v. Taylor, 451 U.S. 527 (1981) to argue that the existence of an adequate post-deprivation remedy affords due process? See Wilson v. Civil Town of Clayton, Indiana, 839 F.2d 375, 380 (7th Cir. 1988) (“In a procedural due process case such as this, resolution of the Monell issue will also resolve the Parratt issue. Because a municipality may only be liable for ‘acts which the municipality has officially sanctioned or ordered,’ … its liability can never be premised on the result of a random and unauthorized act.”).