Plaintiff, Gardner C. Coughlen, appeals the district court's order granting summary judgment to defendants in this Sec. 1983 action. Plaintiff alleges that defendants, police officers Jim Coots, Mark Jump, and Daniel Farrell unlawfully arrested him, used excessive force, and maliciously prosecuted him in violation of his federal constitutional rights and Kentucky law. Because the district court's analysis of the validity of the release-dismissal agreement upon which it based its grant of summary judgment did not comport with Supreme Court guidelines, we reverse and remand.

On the evening of January 15, 1991, plaintiff, an Illinois businessman, was staying at a hotel in Covington, Kentucky, in order to attend a meeting. After dinner and drinks with several associates, he returned to his hotel. According to plaintiff, as he walked through the parking garage, he pulled on the barrier arm of one of its gates and apparently broke it. The incident prompted hotel security officers to summon Covington police.

Officers Coots, Jump, and Farrell found plaintiff in the hotel lobby, took him into custody, and escorted him down an elevator to a patrol car. In the process, plaintiff suffered injuries, including a cut that required twelve stitches.

As might be expected, the manner in which these injuries were inflicted is disputed. Defendants maintain that plaintiff was intoxicated and resisted arrest so violently that they had to use force and chemical mace to subdue him. According to the officers, plaintiff's injuries were inflicted when he banged himself against the door and interior of the patrol car. For his part, plaintiff asserts that he did not resist arrest, but was beaten and maced by the officers on the way to the patrol car without any provocation from him.

Plaintiff was charged with assault of a police officer, resisting arrest, and public intoxication. Upon the advice of his legal counsel, plaintiff signed an agreement, under the terms of which he agreed not to sue the officers or the city of Covington for any cause of action arising out of his arrest, and to plead guilty to the charge of public intoxication. In return, the prosecutor agreed to dismiss the assault and resisting arrest charges. However, after plaintiff's criminal charges were resolved according to the terms of the agreement, he lodged brutality complaints against the officers with the Covington Chief of Police and the FBI. When these complaints were not handled to his satisfaction, he filed this suit.

Defendants moved for summary judgment, arguing that the release barred the suit. Plaintiff then sought to amend his complaint to add an additional party and another theory of recovery. The district judge asked plaintiff whether he would consent to reinstatement of the criminal charges against him in exchange for voiding the release, but plaintiff declined. On May 15, 1992, the district court granted the officers' motion for summary judgment and denied plaintiff's motion to amend.

The district court rejected plaintiff's argument that the release was invalid, citing the Supreme Court's decision in Town of Newton v. Rumery, 480 U.S. 386, 107 S. Ct. 1187, 94 L. Ed. 2d 405 (1987), for the proposition that "[s]uch releases have been held not to be against public policy." However, a careful examination of Rumery shows that the district court's reading of the case goes too far.

In Rumery, a majority of the Court held that such agreements are not per se invalid as contrary to public policy. Rumery, 480 U.S. at 392-94, 107 S.Ct. at 1191-93; id. at 399, 107 S.Ct. at 1195 (O'Connor, J., concurring in part and in the judgment). The five justices rejected the argument that it was inherently coercive to present a criminal defendant with a choice between facing criminal charges and waiving his right to sue under Sec. 1983. Id. at 393, 107 S.Ct. at 1192. The opinion noted that in other contexts, such as plea bargaining, "criminal defendants are required to make difficult choices that effectively waive constitutional rights" and it found "no reason to believe that release-dismissal agreements pose a more coercive choice than other situations we have accepted." Id. (citations omitted).

But, while the Rumery majority rejected the lower court's holding that release-dismissal agreements are per se invalid, it recognized that "in some cases these agreements may infringe important interests of the criminal defendant and of society as a whole." Id. at 392, 107 S.Ct. at 1192. Justice O'Connor described such cases in her concurrence:

Permitting such releases may tempt public officials to bring frivolous criminal charges in order to deter meritorious civil complaints. The risk and expense of a criminal trial can easily intimidate even an innocent person whose civil and constitutional rights have been violated. The coercive power of criminal process may be twisted to serve the end of suppressing complaints against official abuse, to the detriment not only of the victim of such abuse, but also of society as a whole.

Id. at 400, 107 S.Ct. at 1196 (O'Connor, J., concurring in part and in the judgment) (citation omitted). See also id. at 394, 107 S.Ct. at 1193 (plurality opinion) ("We can agree that in some cases there may be a substantial basis for [the] concern [that trumped-up criminal charges will be used against criminal defendants making civil rights claims against police].").

The Court concluded that the validity of such agreements should be determined by courts using a "case-by-case approach [which] appropriately balances the important interests on both sides of the question of the enforceability of these agreements." Id. at 399, 107 S. Ct. at 1195 (O'Connor, J., concurring in part and in the judgment); see id. at 392, 107 S.Ct. at 1191 (release-dismissal is unenforceable if "the interest in its enforcement is outweighed in the circumstances by a public policy harmed by enforcement of the agreement").

While the majority did not expressly enumerate those "important interests" which must be balanced when evaluating a particular release-dismissal agreement, it did conclude that the Rumery agreement was valid because it was voluntary, there was no evidence of prosecutorial misconduct, and enforcement of the agreement would not adversely affect relevant public interests. Id. at 398, 107 S.Ct. at 1195. Those circuit courts which have had occasion to apply Rumery have taken these three concerns to be the "important interests" that should be considered by a court when determining whether a specific agreement should be enforced.1

Justice O'Connor said that she was writing separately

to emphasize that it is the burden of those relying upon such covenants to establish that the agreement is neither involuntary nor the product of an abuse of the criminal process.

....

... The defendants in a Sec. 1983 suit may establish that a particular release executed in exchange for the dismissal of criminal charges was voluntarily made, not the product of prosecutorial overreaching, and in the public interest.

Id. at 399, 401, 107 S. Ct. at 1195, 1196 (O'Connor, J., concurring in part and in the judgment). The four Rumery dissenters joined Justice O'Connor's burden of proof analysis. Id. at 417, 107 S.Ct. at 1204 (Stevens, J., dissenting). Accordingly, a majority of the Court supported the proposition that the burden of proving the enforceability of such a release is upon the party asserting it as a defense to a Sec. 1983 claim.

In sum then, the Rumery opinion instructs us that before a court properly may conclude that a particular release-dismissal agreement is enforceable, it must specifically determine that (1) the agreement was voluntary; (2) there was no evidence of prosecutorial misconduct; and (3) enforcement of the agreement will not adversely affect relevant public interests. The burden of proving each of these points falls upon the party in the Sec. 1983 action who seeks to invoke the agreement as a defense.

Here, the district court did not conduct the analysis called for by Rumery. Instead, the court concluded that "such releases have been held not to be against public policy in ... Rumery," and, in effect, treated the release as presumptively valid.2

We must therefore remand this cause to the district court in order that it may make the specific determinations required by Rumery, as enumerated above. Should the court conclude on remand that the release portion of the agreement is invalid, then it follows that the provision precluding the government from further prosecution would be likewise negated.

Since the Supreme Court's opinion in Rumery offers little elaboration concerning the elements required to support such a release, we take this opportunity to comment on some aspects of this case. For example, we do not agree with defendants' contention that "the issue of whether or not there was use of excessive force is not part of the set of facts needed to uphold the release/covenant not to sue." While a district court conducting a Rumery analysis should not prejudge the civil plaintiff's Sec. 1983 claim, the existence of substantial evidence of police misconduct in a particular case is by no means irrelevant to a proper Rumery inquiry, since it could be probative of the motives of the prosecutor for seeking such an agreement, as well as the degree to which enforcing the agreement would serve the public interest.

Rumery indicates that under ordinary circumstances it is not improper for prosecutors to obtain releases as part of a plea bargain or dismissal of criminal charges. As both the plurality and Justice O'Connor recognized, many of the civil rights suits filed by criminal defendants are meritless. Rumery, 480 U.S. at 395, 107 S.Ct. at 1193 (plurality opinion); id. at 399, 107 S.Ct. at 1195 (O'Connor, J., concurring in part and in the judgment). Thus, enforcing releases in those circumstances furthers an important public interest by protecting law enforcement officials from the considerable burdens of defending against unwarranted civil claims. Id. at 395-96, 107 S.Ct. at 1193-94 (plurality opinion); id. at 399, 107 S.Ct. at 1195 (O'Connor, J., concurring in part and in the judgment).

However, these agreements should be scrutinized closely in cases where substantial evidence supports an allegation of police misconduct, in view of the potential for abuse of release-dismissal agreements by law enforcement officials. Id. at 394-95, 107 S.Ct. at 1192-93 (plurality opinion); id. at 400, 107 S. Ct. at 1196 (O'Connor, J., concurring in part and in the judgment). Examples of such prosecutorial misconduct would include situations where, following their use of excessive force, police officers file unfounded criminal charges as bargaining chips to cover up their own conduct or to induce the victim to give up his cause of action; or where a prosecutor, upon discovering that the victim has a meritorious civil claim, files frivolous criminal charges in order to protect the police officers. Accordingly, should a court conclude that a prosecutor secured a release-dismissal bargain in the face of substantial evidence of police misconduct, the court could take this as evidence of prosecutorial misconduct, since it would demonstrate that "the coercive power of criminal process" was being "twisted to serve the end of suppressing complaints against official abuse, to the detriment not only of the victim of such abuse, but also of society as a whole." Id. at 400, 107 S.Ct. at 1196 (O'Connor, J., concurring in part and in the judgment).

The least well-defined element of a Rumery analysis is the consideration of whether enforcement of the agreement will "adversely affect the relevant public interests." Id. at 398, 107 S.Ct. at 1195. The majority justices in Rumery suggested that this standard can be satisfied if the prosecutor demonstrates that obtaining the release was motivated by an independent, legitimate criminal justice objective. Id. at 398, 401-02, 107 S. Ct. at 1195, 1196-97. This does not appear to create a particularly difficult hurdle for the prosecutor to clear. Indeed, the reason proffered by the prosecutor for the release-dismissal agreement in this case--"to aid in the disposition of its heavy case load"--would probably suffice under ordinary circumstances. See id. at 401, 107 S.Ct. at 1196 (O'Connor, J., concurring in part and in the judgment) (citing "allocation of criminal justice resources" as a "legitimate criminal justice" concern).

Release-dismissal agreements also can be legitimate criminal justice tools in situations where police misconduct is alleged, but the prosecutor is genuinely unable to ascertain the truth surrounding the allegation. As the Ninth Circuit noted in Lynch v. City of Alhambra, 880 F.2d 1122 (9th Cir.1989), release-dismissal agreements can allow prosecutors to "achieve a rough substantial justice where the 'true' facts of the case are not known" by "allow[ing] everyone to declare the case a draw and go home." Id. at 1127 n. 8. Examples of other legitimate criminal justice objectives that come to mind are situations where the cost of prosecution would outweigh the benefit accruing to the public from a conviction; where the strength of evidence of criminal conduct is doubtful even though charges were filed in good faith; where witnesses or evidence are no longer available; where evidence is subsequently discovered that points to the criminal case defendant's innocence; or where criminal charges are not the product of prosecutorial misconduct and both sides benefit substantially from a balanced settlement in the sense that both avoid exposure to potential liabilities and expenses.

Rumery also requires a court reviewing a release-dismissal agreement to weigh or balance the interests on both sides of the issue of enforceability. Rumery, 480 U.S. at 392, 107 S. Ct. at 1191; id. at 399, 107 S.Ct. at 1195 (O'Connor, J., concurring in part and in the judgment). Therefore, in a case where there existed substantial evidence of police misconduct, a court reviewing the release-dismissal agreement could conclude that the public's interest in vindicating constitutional rights and deterring police misconduct3 by permitting a specific civil complaint to go forward, despite a release, outweighs a general prosecutorial interest in helping to manage a heavy case load.

For the foregoing reasons, the judgment of the district court is reversed and this cause is remanded for further proceedings consistent with this opinion. The district court should also reconsider plaintiff's motions to amend his complaint should it conclude that the release is not a bar to plaintiff's action.