Livingstone v. North Belle Vernon Borough

91 F.3d 515, 1996 WL 426800

• Court: Court of Appeals for the Third Circuit
• Decided: July 31, 1996
• Docket: 95-3252
• Status: Unknown

Opinion

OPINION OF THE COURT

LOUIS H. POLLAK, District Judge.

This is the second time that this matter has come before this court.

Appellants Frances and Joseph Livingstone commenced this civil rights suit, pursuant to 42 U.S.C. § 1983, in January, 1991 against defendants North Belle Vernon Borough, Fayette City Borough, Washington Township, Officer Darhl Snyder, Officer Raymond Moody, and Officer Frank E. Monack. The defendants moved for summary judgment, asserting that the Livingstones’ claims were barred by an agreement said to have been made in 1990 in which the Livingstones waived any civil claims in exchange for the termination of a criminal prosecution of Frances Livingstone. (Agreements like this one, in which a criminal defendant waives potential civil claims in exchange for the dismissal of the case against her, are called “release-dismissal agreements.”)

The district court granted summary judgment in favor of the defendants. In Livingstone v. North Belle Vernon Borough, 12 F.3d 1205 (3d Cir.1993) (in banc) (“Livingstone /”), this court reversed, finding that there was a genuine issue of material fact as to whether the Livingstones had concluded the release-dismissal agreement voluntarily. 12 F.3d at 1214. In that opinion, we also observed that the agreement raised a number of other possible legal questions, including whether its enforcement would be in the public interest and whether it was invalidated by the municipalities’ failure to formally ratify it.

On remand, the district court granted partial summary judgment in favor of the defendants as to the latter two questions, finding that the agreement’s enforcement would be in the public interest and that no formal ratification was necessary. The district court then conducted a jury trial devoted to the single question whether the Livingstones had concluded the release-dismissal agreement voluntarily. The jury found that the Livingstones did indeed voluntarily conclude the release-dismissal agreement. Accordingly, the district court ordered that judgment be entered in favor of the defendants and against the Livingstones as to all of the Livingstones’ claims. This appeal followed.

I. Factual Background and Procedural History

A. Livingstone I.

In Livingstone I, we compendiously summarized this case’s elaborate factual and procedural history. It seems efficient to reproduce that summary here verbatim. (Except as noted, all footnotes and all emendations are from Livingstone I.)

This civil suit filed by Frances and Joseph Livingstone against the police officers and municipalities arose from the conduct of the police officers on the night of January 12-13, 1989, at the Livingstone home in Washington Township, Pennsylvania. During a family argument between Carrie Livingstone, age twenty-two, who was unmarried and living at her parents’ home with her fourteen-month-old son, and Joseph, her father, Mr. Livingstone struck Carrie on the face, causing her lip to split and bleed. Carrie ran out of the house and to the community ambulance service across the street, where an employee called the police. When Officer Frank Mo-nack arrived, Carrie told him that her father had struck her and that her parents were holding her son without her consent.

Monack, who was at that time an officer in the Washington Township Police Department and is now Chief of Police, radioed for assistance pursuant to an inter-municipal police cooperation agreement. Raymond Moody, who was and is the Chief of Police for the Borough of Fayette City, and Darhl Snyder, an officer in the North Belle Vernon Police Department, responded. They proceeded to the Livingstone home where Mr. Livingstone permitted them to enter for the purpose, he later testified, of discussing possible criminal charges against him arising out of the incident. Following a brief discussion, Mo-nack and Snyder accompanied Mr. Livingstone outside, and Monack told him to go to the nearby police station to make a statement. 1 No charges were filed against Mr. Livingstone that evening or at any later time.

Monack and Snyder then reentered the Livingstone household, this time in search of Carrie’s son and admittedly without a warrant or court order. 2 Mrs. Livingstone had retreated to the back bedroom with her grandson, and had locked and barricaded the door. When she refused to open the door, Monack picked the lock and then tried to push the door open. From the partially opened door, Mrs. Livingstone hit him with a fishing rod and scratched him. Monack and Snyder broke the door down to enter the room, and then Monack told Mrs. Livingstone she was under arrest.

Mrs. Livingstone testified that both men struck her, causing her to lose consciousness and sustain bruises, lacerations, lost teeth, and head injuries. According to defendants, they used force only for the purpose of getting handcuffs on her after she struck the officer, and a stun gun to subdue her because she was screaming and kicking. Snyder held her down while Mo-nack used the gun. Mrs. Livingstone claims that Monack then said “you want a thrill, I’ll give you a thrill” and applied the stun gun between her legs. A medical examination conducted at the hospital that night notes a bum in the vulval area.

The officers removed Mrs. Livingstone, handcuffed, from the house. She states that they dragged her outside and dropped her several times, banging her head, and then left her lying in cold muddy water for hours. The officers claim that her thrashing caused them all to fall, and that she refused to get up.

On January 13, 1989, the morning after the altercation, Mrs. Livingstone was charged by Monack, on behalf of the Washington Township Police Department, with disorderly conduct, aggravated assault, terroristic threats, resisting arrest, and interference with custody. At a pre liminary hearing on April 18, 1989, Mrs. Livingstone was held over for a jury trial on all but the terroristic threats charge, and the aggravated assault charge was reduced to simple assault.

The trial in Fayette County Court of Common Pleas began on February 13, 1990, with attorney Thomas R. Ceraso representing Frances Livingstone and Jack R. Heneks, Jr., an Assistant District Attorney, representing the Commonwealth of Pennsylvania. Carrie' Livingstone testified for the prosecution, followed by Monack, Snyder, Moody, Police Chief Robert Matthews of Washington Township, and Evelyn Rehe of the community ambulance service. The Commonwealth rested, and Mrs. Livingstone demurred to all of the charges. The demurrer was granted on the charge of interference with custody on the ground that there were no facts showing danger to the child, but was denied as to the other charges.

Thereafter, Joseph Livingstone and his son, James, testified for the defense. Before Mrs. Livingstone was to take the stand (and presumably would have testified about her claims with regard to police use of a stun gun on her private parts), the trial judge, Judge Cicchetti of the Court of Common Pleas, met with Heneks and Cer-aso to discuss whether the matter could be resolved. 3 After settlement negotiations, a conference was held in camera with Judge Cicchetti. Present were Moody, Monack, Matthews (now deceased), the Living-stones, Ceraso, and Heneks.

Ceraso summarized the arrangement by stating that the defense would move for a judgment of acquittal after James Livingstone finished his testimony; that expenses for the physical damage to the Livingstone house and for Mrs. Livingstone’s reasonable medical care would be paid; and that once those bills were paid, the Livingstones would release any civil claims. Ceraso stated on the record:

there will be an agreement on the part of my client, Mrs. Livingston[e], and also her husband, Joe Livingston[e], who is present, that upon payment of reasonable medical bills that w[e]re associated with the incident that occurred, based on my forwarding those to Washington Township with confirmation, together with bills reflecting damage incurred at the household of Mr. and Mrs. Livingston[e], that Washington Township will cause the same to be paid. At the time of final payment of those bills, there will be a full and complete release signed with reference to any civil action on the part of Mr. and Mrs. Livingston[e]. It’s also my understanding that at that time there will also be a release signed by Washington Township, or any of its proper officials, or any member of the police force necessary to release Mr. and Mrs. Livingston[e] from any liability_

App. at 1109.

In response to the judge’s inquiry, the parties voiced an expression of assent. The court asked whether “you all think this is in the best interest for everyone” and Matthews, Monack, and Heneks said they did. App. at 1112. When they returned to the courtroom, Ceraso moved for a judgment of acquittal on the criminal charges against Mrs. Livingstone, which the court granted.

It is undisputed that the settlement agreement was never reduced to writing. The Livingstones never submitted for payment any medical bills or household repair bills, and no payments have been made. The Board of Supervisors of Washington Township took no action to officially ratify the agreement until almost two years after the conference in chambers, and only then after this suit was started. 4 There is no evidence that the other two municipal defendants, North Belle Vernon Borough and Fayette City Borough, have ever taken any action to ratify any putative settlement.

On January 14,1991, almost a year after the criminal trial, the Livingstones filed this action against the police officers and the three employer municipalities. The complaint consisted of seven claims: a federal claim filed pursuant to 42 U.S.C. § 1983 (1988), and state law claims alleging assault and battery, malicious prosecution, malicious abuse of process, invasion of privacy, intentional infliction of emotional distress, and conversion.

The defendants filed motions to dismiss or, in the alternative, for summary judgment. The district court referred the case to a magistrate judge, who ordered the parties to engage in discovery [footnote omitted] and to submit briefs and materials in support of the motions.

In their briefs in support of summary judgment, defendants argued that the suit was barred by the release-dismissal agreement reached during the in camera conference before Judge Cicchetti. The Liv-ingstones claimed that they never intended to waive their rights to sue, pointing out that the agreement was never reduced to writing and that Washington Township never made the contemplated payments. They also contended that the agreement was never properly entered into by the municipalities, as the Washington Township Board of Supervisors never formally approved it and the other boroughs’ governing bodies never considered it, and that the agreement was invalid and unenforceable under Pennsylvania law.

On April 8, 1992, the magistrate judge submitted a Report and Recommendation recommending that summary judgment be granted for the defendants on the basis of the release-dismissal agreement. Although the court acknowledged that Washington Township may not have formally approved the agreement, it noted that two supervisors, a quorum, had approved it, thereby satisfying Pennsylvania law. Without comment on the absence of the other municipalities and officers from the agreement, the magistrate judge concluded that plaintiffs had contracted with all of the defendants, and thus the plaintiffs’ civil suit was barred; that the agreement comported with due process because the plaintiffs understood that they were waiving their rights to assert future civil claims and had entered into the release voluntarily; and that there was sufficient consideration because the plaintiffs, in exchange for the surrender of their potential civil claims, had secured the dismissal of the criminal charges and a promise by the defendants not to sue them.

Objections were filed but the district court adopted the magistrate judge’s opinion as its own and granted summary judgment for all defendants. This timely appeal followed.

Livingstone I, 12 F.3d at 1206-09.

With this background history as predicate, the court in Livingstone I then proceeded to review the elements of proof for a showing of voluntariness, finding that the parties seeking to enforce the release-dismissal agreement bore the burden of showing that the Livingstones’ assent was “voluntary, deliberate and informed.” 12 F.3d at 1211. We concluded that the defendants had not met this burden with the certainty called for on summary judgment, given that Mrs. Livingstone was confused as to the terms of the release-dismissal arrangement, that the claimed release-dismissal agreement was never written down, and that the asserted agreement — assuming there was a meeting of the minds — was made, if at all, during a brief and ambiguous oral colloquy. See id. at 1211-14. Accordingly, we reversed the grant of summary judgment and directed that the case be remanded for further proceedings.

B. Post-Livingstone I Proceedings

Following remand, the defendants filed a motion for summary judgment challenging the Livingstones’ complaint on a variety of grounds other than the release-dismissal agreement. Their motions were referred to a magistrate judge, whose Report and Recommendation (“R & R”) the district court then adopted without substantive comment. In accordance with the recommendations of the magistrate judge, the district court ordered that (1) summary judgment be entered in favor of all defendants as to the Living-stones’ claims of malicious prosecution, abuse of process, and invasion of privacy; 5 and (2) summary judgment be entered in favor of Officer Moody as to the assault and battery claims. The district court denied summary judgment as to the Livingstones’ constitutional claims, their claims of assault and battery against Officers Monack and Snyder, their claim of intentional infliction of emotional distress, and their conversion claim. App. at 326, 352.

The Livingstones then filed a motion for partial summary judgment renewing two arguments that they had already unsuccessfully made to the district court in the first round of the litigation. These arguments were that the release-dismissal agreement was unenforceable because (a) the municipalities had not ratified it, as (assertedly) required by Pennsylvania law, and (b) the release-dismissal agreement had been concluded in a manner which violated “the Pennsylvania Rules of Criminal Procedure and public policy.” App. at 355. 6 The defendants responded with cross-motions for partial summary judgment that asserted that no ratification was necessary, because the Livingstones had failed to submit their medical bills to Washington Township for payment. The district court granted summary judgment in favor of the defendants on both the ratification question and the public-interest question.

The district court then conducted a jury trial limited to one question, whether the Livingstones voluntarily entered into the release-dismissal agreement. After several days of trial, including extended testimony by Thomas R. Ceraso (Mrs. Livingstone’s lawyer at her criminal trial), the jury found that the Livingstones did indeed enter into the agreement voluntarily. Accordingly, the district court entered judgment in favor of the defendants and against the Livingstones. The Livingstones moved for a new trial, for judgment notwithstanding the verdict, and to amend the court’s judgment to require Washington Township to pay household damages and medical bills to the Livingstones (apparently in order to enforce the terms of the release-dismissal agreement). These motions were denied, and this appeal followed.

On appeal, the Livingstones assert that the district court (1) erred in finding that the agreement was valid and enforceable even though the municipalities had not ratified it; (2) erred in ruling that the enforcement of the agreement was in the public interest as a matter of law; and (3) made a number of errors at the voluntariness proceeding. 7 We will discuss these questions in that order.

As to the second and third of these arguments, the district court, appellants, and appellees all assume that identical legal standards govern the enforcement of the release-dismissal agreement as to the Livingstones’ section 1983 claims and as to their state-law claims. However, as we suggested in Livingstone I, see 12 F.3d at 1209 n. 6, this is not necessarily the case. Federal common law governs the enforceability of the release-dismissal agreement as to the Livingstones’ section 1988 claims, while we must look to Pennsylvania law to assess the enforceability of the agreement as to their state-law claims. Thus, the legal standards applicable to the Livingstones’ state-law claims will be discussed separately.

II. Jurisdiction and Standard of Review

We have appellate jurisdiction pursuant to 28 U.S.C. § 1291. As to the municipal ratification and public-interest questions, on which the district court granted summary judgment, our review is plenary. See Erie Telecommunications Inc. v. City of Erie, 853 F.2d 1084, 1093 (3d Cir.1988). The appellants also assert that, at the voluntariness proceeding, the district court (1) gave an incorrect jury instruction, (2) erred in declining to give a requested jury instruction, and (3) made a number of incorrect evidentiary rulings. To the extent that appellants claim that a jury instruction failed to state the proper legal standard, our review is plenary. See Government of Virgin Islands v. Isaac, 50 F.3d 1175, 1180 (3d Cir.1995). To the extent that appellants contest the district court’s refusal to give particular jury instructions, our review is for abuse of discretion. See id. The evidentiary rulings that the appellants challenge are all discretionary rulings of the type that we review for abuse of discretion. This includes rulings as to the relevance of evidence and as to its prejudicial effect, see In re Japanese Electronic Products, 723 F.2d 238, 257, 260 (3d Cir.1983), rev’d on other grounds, sub nom. Matsushi-ta Electrical Industrial Co. v. Zenith Radio Corp., 475 U.S. 574, 106 S.Ct. 1348, 89 L.Ed.2d 538 (1986), and as to waiver of the attorney-client privilege, see United States v. Bilzeñan, 926 F.2d 1285, 1293 (2d Cir.1991).

III. Municipal Ratification

The terms of the alleged release-dismissal agreement, as recited by Ceraso, were that:

there will be an agreement on the part of my client, Mrs. Livingston[e], and also her husband, Joe Livingston[e], who is present, that upon payment of reasonable medical bills that w[e]re associated with the incident that occurred, based on my forwarding those to Washington Township with confirmation, together with bills reflecting damage incurred at the household of Mr. and Mrs. Livingston[e], that Washington Township will cause the same to be paid. At the time of final payment of those bills, there will be a full and complete release signed with reference to any civil action on the part of Mr. and Mrs. Livingston[e]. It’s also my understanding that at that time there will also be a release signed by Washington Township, or any of its proper officials, or any member of the police force necessary to release Mr. and Mrs. Livingston[e] from any liability....

Appellees North Belle Vernon Borough and Officer Darhl Snyder’s App. at 30. In short, the arrangement was apparently that, after the prosecution of Mrs. Livingstone was terminated, the Livingstones would submit bills for property damage and for medical costs to Washington Township. Once the Township paid these bills, the Livingstones, the municipalities, and the police officers would then sign full mutual releases of civil claims.

The Livingstones concede that they never submitted their bills to Washington Township, as apparently required by the terms of the agreement. The district court found that their failure to do so rendered it impossible for the municipal defendants to ratify the release-dismissal agreement, as “the public fisc cannot be allocated for an indefinite amount to a private party.” App. at 377-78. The district court did not, however, discuss an antecedent question: whether (and how) a contract was ever formed between the Liv-ingstones and the municipalities. Conceptually, it would hardly be possible for the Liv-ingstones to have rendered impossible the performance of a contract that was never formed. 8

Under Pennsylvania law, a township cannot enter into a binding contract except by a vote of the township’s supervisors. See Abington Heights School District v. Township of South Abington, 72 Pa.Cmwlth. 388, 456 A.2d 722, 724 (1983). North Belle Vernon Borough and Fayette City Borough never conducted such a vote, and Washington Township only did so after the present suit was filed. 9 On appeal, the Livingstones assert that the failure of the municipalities to ratify the release-dismissal agreement meant that no contract was ever formed between the Livingstones and the municipalities, and that the release-dismissal agreement is therefore unenforceable.

We will not address this question, however, because we find that it was not necessary for the municipalities to be parties to the release-dismissal agreement in order for it to be enforceable. It would suffice for the municipalities to have been third-party beneficiaries of an agreement concluded between the Livingstones and the Commonwealth of Pennsylvania; and, indeed, the colloquy before Judge Cicchetti suggests that this is what was intended (assuming, of course, that a valid agreement was formed at all). The principal parties negotiating the purported release-dismissal agreement were the Livingstones (through Mrs. Livingstone’s attorney, Ceraso) and the Commonwealth (through Heneks, an assistant district attorney). The agreement’s terms appear to have been that the Commonwealth would not oppose Mrs. Livingstone’s motion for a judgment of acquittal. In exchange, the Living-stones would submit their medical and household damages bills to Washington Township, and, when those bills were paid, would sign a full release of civil liability with all of the municipalities and police officers involved, reciprocal releases of civil liability being signed by those police officers and municipalities with potential claims against the Livingstones.

Although the municipalities and police officers were clearly intended to benefit from this agreement, the agreement’s success did not require them to be parties to it. The Livingstones were not harmed by the municipalities’ lack of party status. If Washington Township did not pay the Livingstones’ actual expenses, or if one of the municipalities or officers refused to sign (or to negotiate in good faith towards) a release, the Living-stones would have lost nothing. The criminal charges against Mrs. Livingstone could not have been reinstituted; moreover, the Livingstones would presumably have been free to file a civil action against any of the municipalities or police officers that failed to cooperate as anticipated. 10

An implicit term of this release-dismissal agreement is necessarily that the Livingstones could bring a civil suit against the municipalities or police officers only after the Livingstones had made a good-faith effort to negotiate towards reciprocal releases and those negotiations had failed. This term follows from the duty of good faith and fair dealing, Restatement (Second) of Contracts § 205 (1981), and that duty’s correlative obli gation not to act so as to defeat an agreement’s objective. The record indicates that the Livingstones did not make any effort to negotiate towards such reciprocal releases. Hence, assuming that the release-dismissal agreement is otherwise valid and enforceable — the question that we will address next — the Livingstones’ failure to seek mutual releases would seem to bar their suit. 11

The Livingstones also question whether North Belle Vernon Borough and Fayette City Borough — which we will refer to, for brevity, as “the two boroughs” — had the same status under the release-dismissal agreement as did Washington Township. In the voluntariness proceeding in the district court, counsel for the Livingstones had requested that a specific question on the verdict form address the status of the two boroughs under the agreement. The district court declined to include such a question on the form, finding that Ceraso’s statements in the colloquy before Judge Cicchetti included all three municipalities, and that all three therefore had the same status for purposes of the voluntariness question. In response to the objections of the Livingstones’ counsel to this ruling, the district court permitted him to argue to the jury that the ambiguous nature of the agreement between the Living-stones and the two boroughs rendered the release-dismissal agreement involuntary as a whole. App. at 804-06.

Although the Livingstones’ argument focuses on whether the release-dismissal agreement was voluntary as to the two boroughs, this issue cannot be completely disentangled from that of whether the agreement addressed the boroughs at all. The colloquy before Judge Cicchetti is far from a model of clarity on this question. 12 During the colloquy, Ceraso stated that he had

no objection if those police departments or those municipalities also wish to be included in the release, and we would then have reciprocal releases from them, and we would let that up to their individual counsel to make that decision, but we certainly would have no objection in doing that so it would be reciprocal on both sides.

Appellees North Belle Vernon Borough and Officer Darhl Snyder’s App. at 31. This statement can be construed either (1) as indicating that the Livingstones had undertaken to negotiate towards a civil release with the two boroughs (making the boroughs, with Washington Township, third-party beneficiaries of the release-dismissal agreement), or (2) as merely making an offer to those two municipalities.

The question of which of these readings of Ceraso’s remarks is correct was not argued before the district court. On remand, the district court should permit the parties to brief this question. In resolving this issue, the district court may consult all of the sources to which courts usually refer in determining the meaning of ambiguous contractual language, including, for instance, the course of the negotiations between the parties. 13

IV. The Public Interest

In Town of Newton v. Rumery, 480 U.S. 886, 107 S.Ct. 1187, 94 L.Ed.2d 405 (1987), a four-Justice plurality found that, as a matter of federal common law, a release-dismissal agreement will operate to bar a section 1988 claim unless “the interest in [the agreement’s] enforcement is outweighed in the circumstances by a public policy harmed by the enforcement of the agreement.” Id. at 392, 107 S.Ct. at 1191. Justice O’Connor, whose fifth vote was dispositive, noted in a concurring opinion that it is the burden of the defendants to demonstrate 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.” Rumery, 480 U.S. at 401, 107 S.Ct. at 1196.

In Cain v. Darby Borough, 7 F.3d 377 (3d Cir.1993) (in banc), cert. denied, 510 U.S. 1195, 114 S.Ct. 1303, 127 L.Ed.2d 655 (1994), this court addressed the circumstances in which enforcement of a release-dismissal agreement will be in the public interest. Cain made clear that the above-quoted passage from Justice O’Connor’s Rumery concurrence should not be read to suggest that the “prosecutorial overreaching” and “public interest” questions are to be analyzed separately; rather, “the concept of prosecutorial misconduct is embedded in a larger inquiry into whether enforcing the release would advance the public interest.” Id. at 380; see also Lynch v. City of Alhambra, 880 F.2d 1122, 1126 n. 6 (9th Cir.1989) (arguing that there is only one inquiry); but compare Woods v. Rhodes, 994 F.2d 494, 500-01 (8th Cir.1993) (apparently treating the analyses as distinct).

Cain found that a party seeking to demonstrate that the enforcement of a release-dismissal agreement is in the public interest must make two distinct showings, which we will call here Cain’s “objective” and “subjective” elements. Cain’s objective element requires both that “the facts known to the prosecutor when the agreement was reached” must have sufficed to support the prosecutor’s proffered public interest reason for concluding the agreement, and that this public-interest reason be a legitimate one. 7 F.3d at 381. Relevant public interests include the interest, cited by the Court in Rumery, in avoiding the costs and disruptions associated with defending “marginal” or “frivolous” civil rights actions, Rumery, 480 U.S. at 395, 107 S.Ct. at 1193, and the countervailing interest, also cited by the Court, in detecting and deterring official misconduct. See Rumery, 480 U.S. at 394, 107 S.Ct. at 1192-93; id. at 400, 107 S.Ct. at 1195-96 (O’Connor, J., concurring).

Cain’s subjective element is.its requirement that:

the public interest reason proffered by the prosecutor must be the prosecutor’s actual reason for seeking the release.... Any alternative to the ‘actual reason’ requirement creates the real danger that actions taken pursuant to an improper motive, such as to protect public officials from a meritorious civil rights lawsuit, may be legally excused because a court later finds that some ‘benefit’ might have been incidentally achieved.

7 F.3d at 381 (emphasis in original). The party seeking to enforce the release-dismissal agreement bears the burden of proof on both of these elements.

In the present case, the district court denied a motion by the Livingstones that sought to establish as a matter of law that the enforcement of the release-dismissal agreement was contrary to the public interest, and instead entered summary judgment against the Livingstones on this question, finding that the enforcement of the agreement was in the public interest as a matter of law. On appeal, the Livingstones challenge both rulings. They argue, first, that the district court erred in denying their motion for summary judgment, and that it should have found the release-dismissal agreement unenforceable as a matter of law. In the alternative, they contend that the district court erred in granting the appellees’ motion for summary judgment, because there was a genuine issue of material fact as to prosecu-torial motive.

The Livingstones’ two arguments are founded on distinct elements of the Cain analysis. Their argument that the district court erred in declining to find the release-dismissal agreement unenforceable as a matter of law is directed at Cain’s objective element; it challenges the district court’s conclusion that the facts known to the prosecutor at the time the release-dismissal agreement was concluded sufficed to establish that it was in the public interest to conclude such an agreement. Their argument that there is a genuine issue of material fact as to prose-cutorial motivation is directed at Cain’s subjective component, which requires that the prosecutor’s stated reason for concluding a release-dismissal agreement “must be the prosecutor’s actual reason for seeking the release.” Cain, 7 F.3d at 381. We will consider these arguments in that order.

A. Cain’s Objective Element

1. The District Court’s Analysis

The district court concluded that the facts known to the prosecutor at the time the agreement was concluded justified finding that the enforcement of the release-dismissal agreement would be in the public interest. The court explained its conclusion as follows:

Here, Judge Cicchetti, who presided over the criminal trial against Mrs. Livingstone and who supervised the execution of the release-dismissal agreement, stated in the colloquy that he was supportive of the agreement because he saw no benefit to a criminal trial and that it was in everyone’s best interest to resolve the matter. Mr. Heneks, the assistant district attorney who was assigned to the case stated that he believed that the Commonwealth would be well-served by the resolution as well. Later, in a sworn statement, Mr. Heneks indicated that continuation of the criminal trial would have created further conflict between the Livingstones and their daughter who had reconciled their differences since the night of the incident. In addition, he stated that the agreement saved the Commonwealth from spending further resources to prosecute. The reasons stated by Mr. Heneks are factors that were known to him at the time the agreement was executed. In the absence of evidence that the motivation was improper, we may accept his explanation. In addition, each reason constitutes an independent, legitimate reason which is directly related to his prosecutorial responsibilities. See Rumery, 480 U.S. at 398, 107 S.Ct. at 1194.

App. at 380-381. The court’s analysis posits three public-interest rationales for upholding the release-dismissal agreement: the agreement’s supervision by Judge Cicchetti, the desire of the Commonwealth to avoid further conflict between the Livingstones and their daughter, and the Commonwealth’s wish to avoid expending more of the Commonwealth’s resources to prosecute Mrs. Livingstone. We will consider these three rationales seriatim.

As to the first of the three rationales, it is of course true that (1) the Rumery plurality noted that judicial supervision of release-dismissal agreements can “help ensure that the agreements did not result from prosecu-torial misconduct,” 480 U.S. at 398 n. 10, 107 S.Ct. at 1195 n. 10, and (2) Justice O’Connor observed that such supervision can “bear on whether a release was voluntary and not the product of overreaching,” id. at 401-02, 107 S.Ct. at 1196. Judicial supervision can indeed be important in ensuring that an agreement was concluded voluntarily, and, to a-lesser extent, that the prosecutor’s stated reasons for seeking an agreement are genuine. Judicial supervision is less relevant to Cain’s objective inquiry, however, which focuses on the information known to the prose cutor. At best, judicial supervision may help to reinforce a subsequent court’s independent determination that a prosecutor had a sound public-interest reason for concluding a release-dismissal agreement. As will become clear, it seems unlikely that Judge Ciechetti’s supervision of the dismissal of the charges against Mrs. Livingstone played that role here.

Nor does Heneks’ asserted desire to avoid further stress to the Livingstone family serve a particularly strong public interest. It is, of course, commendable for prosecutors to give some thought to the welfare of the accused’s family. In practice, however, the public would be rightly surprised were a prosecutor to place these considerations above, for instance, the public interest in punishing crime, or the public interest, expressed in section 1983, in exposing official abuse. 14 We do not think that the public’s interest in avoiding strain to a defendant’s family can, standing alone, be a legitimate reason for concluding a release-dismissal agreement.

Nor, finally, does Heneks’ wish to avoid the cost of further prosecution carry much weight. A desire to avoid the cost of prosecution (and of a related civil suit) may be an acceptable public-interest rationale for some release-dismissal agreements. As Justice O’Connor observed in Rumery:

[P]rosecutors may legitimately believe that, though the police properly defused a volatile situation by arresting a minor mis-demeanant, the public interest in further litigation is outweighed by the cost of litigation. Sparing the local community the expense of litigation associated with some minor crimes for which there is little or no public interest in prosecution may be a legitimate objective of a release-dismissal agreement.

480 U.S. at 399-400, 107 S.Ct. at 1195. By definition, in any case in which a release-dismissal agreement has been concluded, the community will have avoided the cost of prosecution; thus, a prosecutor could assert that “the public interest in further litigation is outweighed by the cost of litigation” in any case. In order to ensure that such assertions do not act as a blanket exception to the public-interest element of Rumery, the courts must subject those assertions to close scrutiny. 15

2. Marginal or Frivolous Nature of the Livingstones’ Civil Rights Claims

The record does not indicate that Heneks considered whether the Living-stones’ civil rights claims were marginal or frivolous before concluding the agreement. In Cain, we stated that a prosecutor must conduct an “individualized analysis” of a defendant’s civil rights claims before concluding a release-dismissal agreement, 7 F.3d at 383, and that in order for a release-dismissal agreement to be enforceable “there must be a case-specific showing that the released civil rights claims appeared to be marginal or frivolous at the time the agreement was made and that the prosecutor was in fact motivated by this reason.” Id. 16

The question whether the facts known to Heneks could have supported the conclusion that the Livingstones’ civil rights claims were marginal or frivolous was not addressed in the district court. 17 On this record, resolution of that question in the defendants’ favor was a necessary predicate for a grant of summary judgment holding that the release-dismissal agreement was in the public interest. We will, therefore, vacate the district court’s grant of summary judgment and remand the case so that the parties can address the question whether the Livingstones’ civil rights claims were regarded — and, if so, whether they were properly regarded — by the prosecuting attorney as marginal or frivolous.

We think that, on remand, the parties will, at a minimum, wish to take account of the material in this record which suggests that, at the time Heneks agreed to the dismissal of the charges against Mrs. Livingstone, considerable information pointing in the direction of significant police misconduct had come to Heneks’ attention. 18 Of course, what is of record here cannot be deemed dispositive of the question whether Heneks could properly have concluded that the Livingstones’ civil rights claims were marginal or frivolous, for the reason that the proceedings in the district court have not been focused on that issue. Further evidence may be adduced on remand that casts the relevant events in a very different light. But we think it may be helpful to relate the principally salient matter appearing in our current record in order to provide a point of departure for the proceedings on remand.

The most important item of evidence in this respect is the report of Dr. Noche, the emergency-room doctor who examined Mrs. Livingstone on the night of her encounter with the police. That report indicates that Mrs. Livingstone had first or second degree burns in her vaginal area. 19 The substance of Dr. Noche’s report was almost certainly known to Assistant District Attorney He-neks. 20 The record does not contain any plausible explanation of how this bum came to appear on Mrs. Livingstone’s genitalia— other than Mrs. Livingstone’s own explanation, which was that it was the result of the police’s deliberate misuse of a stun gun. 21 Nor is there any indication in the record that Heneks was aware of other evidence that contradicted the emergency-room report. 22

In short, it is difficult to escape the conclusion that Dr. Noche’s report significantly corroborates Mrs. Livingstone’s claim that the police deliberately applied a stun gun to her genitalia, an act that, if it did occur, would amount to an outrageous instance of police abuse. 23

It is possible that facts not in the present record would undermine some element of the foregoing analysis. 24 On remand, the parties should address (1) whether Heneks made a determination that the Liv-ingstones’ civil rights claims were marginal or frivolous, and, if so, on what basis he did so; (2) whether Heneks knew or should have known 25 of the foregoing evidence of police misconduct; and (3) if Heneks did know, or should have known, of that evidence, whether other facts available to Heneks in some way undercut it. We emphasize that it would not suffice for the defendant municipalities and police officers to demonstrate on remand that Heneks was aware of other evidence that merely contradicted the foregoing evidence of police misconduct, as this would only establish that there was substantial evidence on both sides of the misconduct question. 26 Instead, defendants would have to demonstrate that Heneks was aware of other evidence that rendered the foregoing evidence of police misconduct fundamentally untrustworthy.

It is conceivable that the district court may conclude that Heneks was not aware, and had no reason to be aware, of some of the foregoing evidence of official misconduct, and that, not being apprised of this evidence, Heneks reasonably determined that the Liv-ingstones’ civil rights claims were marginal or frivolous. That would not, however, be the end of the district court’s inquiry. The district court would then have to address the further question whether enforcement of a release-dismissal agreement in the face of substantial evidence of police misconduct would be compatible with Bumery and Cain, notwithstanding that the evidence of misconduct was not known, or reasonably knowable, by the prosecutor at the time the prosecutor entered into what might appear, in retrospect, to be an improvident agreement. 27

B. Cain’s Subjective Element: Prosecutorial Motivation

Cain’s subjective element requires that the public-interest reasons cited by a party seeking to enforce a release be those that actually motivated the prosecutor to conclude the release. That is, under Cain, a court may not enforce a release if it finds that the release was concluded for some reason different from that presented as justifying enforcement, even if the court believes that “some ‘benefit’ ” would be “incidentally achieved” by enforcement. Cain, 7 F.3d at 381.

The Livingstones challenge the district court’s determination that there was not a genuine issue of material fact as to whether the prosecutor’s stated reasons for concluding the release-dismissal agreement were his actual reasons for doing so. The district court based this determination on the following: (1) the charges against Mrs. Livingstone were filed the day after her encounter with the police; (2) the charges “correspond to the relevant conduct of Mrs. Livingstone according to the statement taken from Carrie Livingstone, and the affidavit which supports the complaint”; and (3) “discussions of settlement were initiated after nearly three days of testimony in the criminal trial.” App. at 379-80.

We do not quarrel with these three propositions. But they do not, in our judgment, constitute a sufficient predicate for the determination that there is no genuine issue of material fact with respect to whether the prosecutor’s stated reasons were his real reasons. As we have noted, on the record before this court it appears not unlikely that the prosecutor was aware of substantial evidence of police misconduct in the present case. This lends credence to the inference that the prosecutor’s decision to bring charges against Mrs. Livingstone, the manner in which he conducted the trial, and his decision to propose the conclusion of a release-dismissal agreement to the Living-stones may have been motivated by a desire to protect the relevant police officers and municipalities from liability. Such a motivation would render the agreement unenforceable. See Cain, 7 F.3d at 381.

None of the three propositions relied on by the district court eliminates the possibility that the prosecutor acted with an improper motive. As to the fact that the charges against Mrs. Livingstone were filed promptly, it is true that, had the charges against Mrs. Livingstone been brought well after the incident, or after the police learned that she intended to sue, this might have indicated prosecutorial misconduct. See Lynch v. City of Alhambra, 880 F.2d 1122, 1128-29 (9th Cir.1989). But the fact that the charges against her were brought promptly does not, conversely, demonstrate that no misconduct occurred. As to the fact that the charges against Mrs. Livingstone were supported by independent evidence, charges need not be fabricated in order for a release-dismissal agreement to be the product of an improper prosecutorial motive. The relevant question is instead whether the decision to pursue a prosecution, or the subsequent decision to conclude a release-dismissal agreement, was motivated by a desire to protect public officials from liability. Finally, the fact that the discussions of settlement were initiated “after nearly three days of testimony in the criminal trial” is subject to many interpretations. One interpretation which is at odds ■with summary judgment is that a purpose of the trial was to erode the Livingstones’ resistance to signing a release.

We therefore find that there is a genuine issue of material fact as to whether the prosecutor’s stated reasons for concluding the release-dismissal agreement at issue in the present case were his actual reasons. Thus, should the district court find that information known to the prosecutor could have sufficed to establish that there was a legitimate public-interest reason for concluding a release-dismissal agreement, it should then conduct a jury trial to determine whether the prosecutor’s stated reasons for concluding an agreement were his actual reasons for doing so. 28

V. Voluntariness

A. Standard of Proof of Voluntariness.

The district court instructed the jury that the defendant’s burden of proof in establishing the voluntariness of the release-dismissal agreement was one of “preponderance of the evidence.” The Livingstones challenge that instruction, arguing that the standard should have been one of “clear and convincing evidence.” We agree.

In Rumery, the Supreme Court had no occasion to consider the appropriate standard of proof; the language used by the Court in finding the release-dismissal agreement at issue in that case enforceable was consistent with either a preponderance standard or a standard of clear and convincing evidence. 29 Although we did not explicitly address the question of the appropriate standard of proof in Livingstone I, we did observe that oral release-dismissal agreements should be subjected to particularly exacting judicial scrutiny:

Ordinarily, the existence and terms of [a release-dismissal] agreement can be resolved by reference to a written document. While we do not hold that as a matter of law an oral agreement to waive the right to sue in exchange for the dismissal of criminal charges can never be valid, the absence of a written release-dismissal agreement requires even more scrupulous review by the courts than otherwise. No published opinion of any of the courts of appeals after Rumery has even considered, much less sustained, an oral release-dismissal agreement. Indeed, the Rumery Court never mentioned the possibility of an oral release-dismissal agreement. Justice Stevens, at least, assumed that such agreements were written. See Rumery, 480 U.S. at 417 n. 22, 107 S.Ct. at 1205 n. 22 (“A court may enforce such an agreement only after a careful inquiry into the circumstances under which the plaintiff signed the agreement and into the legitimacy of the prosecutor’s objective in entering into [it].” (emphasis added)) (Stevens, J., dissenting).

12 F.3d at 1212. We then noted a number of advantages of written agreements. These included the fact that they “allow the parties more opportunity for deliberate reflection,” id., and that a written document facilitates negotiation as to the agreement’s terms, see id. at 1213. We also observed that a written release-dismissal agreement may provide a subsequent court with evidence as to “the parties’ respective bargaining power.” Id. For example, if the attorney for the party forgoing civil claims prepared the agreement, this may support the conclusion that the agreement was voluntary; if the prosecutor did so, and if he presented it in a manner that discouraged negotiation, this may support the conclusion that it was not. See id.

In Addington v. Texas, 441 U.S. 418, 99 S.Ct. 1804, 60 L.Ed.2d 323 (1979), the Supreme Court set forth its methodology in assigning standards of proof:

The function of a standard of proof, as that concept is embodied in the Due Process Clause and in the realm of factfinding, is to “instruct the factfinder concerning the degree of confidence our society thinks he should have in the correctness of factual conclusions for a particular type of adjudication.” In re Winship, 397 U.S. 358, 370, 90 S.Ct. 1068, 1070, 25 L.Ed.2d 368 (1970) (Harlan, J., concurring). The standard serves to allocate the risk of error between the litigants and to indicate the relative importance attached to the ultimate decision. Id. at 423, 99 S.Ct. at 1808. The Court then placed the three standards of proof within this broad framework. The least demanding standard, that of a preponderance of the evidence, is appropriate to a “typical civil case involving a monetary dispute between private parties.” Id. Society’s concern with the outcome of such a case is “minimal”; thus, it is appropriate to adopt a standard that allocates the risk of error between the litigants “in roughly equal fashion.” Id. The standard of proof beyond a reasonable doubt, by contrast, is reserved for criminal cases, in which society wishes to “exclude as nearly as possible the likelihood of an erroneous judgment.” Id.

Intermediate between these two standards is the one applicable in cases in which “the interests at stake ... are deemed to be more substantial than mere loss of money.” Id. at 424, 99 S.Ct. at 1808. The standard has been known by a variety of names, but “usually employs some combination of the words ‘clear,’ ‘cogent,’ ‘unequivocal,’ and ‘convincing.’ ” Id. at 424, 99 S.Ct. at 1808. Examples of proceedings in which the Court has found a heightened standard of proof to be appropriate are proceedings to terminate parental rights, see Santosky v. Kramer, 455 U.S. 745, 102 S.Ct. 1388, 71 L.Ed.2d 599 (1982); involuntary commitment proceedings, see Addington, 441 U.S. at 432, 99 S.Ct. at 1812-13; and deportation proceedings, see Woodby v. INS, 385 U.S. 276, 285-86, 87 S.Ct. 483, 487-88, 17 L.Ed.2d 362 (1966).

The Court has stated that, in civil actions between private litigants, a standard of proof greater than one of a preponderance of the evidence will only apply in cases in which “‘particularly important individual interests or rights are at stake.’ ” Grogan v. Garner, 498 U.S. 279, 286, 111 S.Ct. 654, 659, 112 L.Ed.2d 755 (1991) (quoting Herman & MacLean v. Huddleston, 459 U.S. 375, 389-90, 103 S.Ct. 683, 691, 74 L.Ed.2d 548 (1983)). Thus, a preponderance standard suffices even in cases in which “severe civil sanctions” may ultimately be imposed, if those sanctions do not implicate particularly important interests or rights. See Huddleston, 459 U.S. at 389, 103 S.Ct. at 691. 30

We find that the enforcement of the oral release-dismissal agreement at issue in this case would indeed implicate “important individual interests or rights.” Although the Livingstones’ section 1983 claims are in form claims for money damages, underlying them is the Livingstones’ interest in redressing a possible violation of their constitutional rights. 31 Moreover, section 1983 actions, when successful, do more than compensate injured plaintiffs: they serve the important public purpose of exposing and deterring official misconduct, and thereby protecting the rights of the public at large. In Rumery, all nine Justices recognized the importance of ensuring that release-dismissal agreements do not encroach upon this purpose. See Rumery, 480 U.S. at 395, 107 S.Ct. at 1193; id. at 400, 107 S.Ct. at 1195-96 (O’Connor, J., concurring); id. at 419, 107 S.Ct. at 1205-06 (Stevens, J., dissenting). 32

A clear-and-convincing standard appropriately allocates more of the risk of error associated with oral release-dismissal agreements to those who seek to enforce them. As we noted in Livingstone I, oral release-dismissal agreements raise particularly significant questions of voluntariness, as the lack of a written document may inhibit negotiation as to an agreement’s terms and render it difficult for prospective parties to reflect on those terms. We also observed in Livingstone I that an oral agreement ordinarily contains less evidence as to the course of the parties’ negotiations than does a writ ten agreement. As a result, there is a greater risk of error in a jury's evaluation of whether an oral release-dismissal agreement was concluded voluntarily.

We think that those seeking to enforce a release-dismissal agreement should bear this greater risk. Indeed, a "clear and convincing" standard will encourage prosecutors-who are likely to have comparatively frequent contact with release-dismissal agreements, and who have an interest in ensuring that those agreements are later found to be enforceable-to ensure that release-dismissal agreements are, whenever possible, written down. The standard will therefore have the salutary effect of reducing the overall risk of misunderstandings in the conclusion of release-dismissal agreements, and increasing the accuracy of juries' decisions as to whether a release-dismissal agreement was concluded voluntarily. 33

Since, when this case was first remanded, the parties challenging the enforceability of the Livingstones' oral release-dismissal agreement were only required to establish the voluntariness of the agreement under a preponderance-of-the-evidence standard, the jury's finding of voluntariness will be vacated. If, on this remand, it again becomes necessary to address the issue of voluntariness, the more demanding clear-and-convincing standard will be utilized. 34

B. Instruction on Existence of a Legitimate Criminal Justice Objective.

The Livingstones sought to have the district court instruct the jury that one of the factors for it to consider in determining whether they voluntarily entered into the release-dismissal agreement was "whether there is a l[e}gitimate criminal justice objective to support [the agreement's] validity." Livingstones' Proposed Jury Instruction 10, App. at 394. The district court declined to so instruct; the Livingstones contend that this was error.

Evidently the rationale for the proposed instruction was that the Livingstones sought to argue to the jury that elements of the public-interest analysis should enter into the jury's evaluation of whether the agreement was volimtary. We see no reason why the public-interest issue is pertinent to the jury's consideration of the vokmtariness issue, and we therefore think the district court was correct in concluding that such an instruction would have been inappropriate.

C. Admission of Ce'raso's Testimony.

The district court found that, by challenging the release-dismissal agreement, the Livingstones had waived any claim of attorney-client privilege as to the testimony of Ceraso, Mrs. Livingstone's lawyer at her criminal proceeding. 35 Accordingly, the dis- triet court permitted Ceraso to be deposed, and then allowed him to be called as a witness at the voluntariness proceeding. App. at 101. On appeal, the Livingstones argue that this decision was erroneous. We disagree.

“The attorney-client privilege is waived for any relevant communication if the client asserts as a material issue in a proceeding that: (a) the client acted upon the advice of a lawyer or that the advice was otherwise relevant to the legal significance of the client’s conduct.” Restatement of the Law Governing Lawyers § 130(1) (Final Draft No. 1, 1996); see also Rhone-Poulenc Rorer Inc. v. Home Indem. Co., 32 F.3d 851, 863 (3d Cir.1994) (“[A] party can waive the attorney client privilege by asserting claims that put his or her attorney’s advice in issue in the litigation.”). The Livingstones’ complaint states that Washington Township “may seek to assert as a possible defense a purported agreement not to sue and/or release,” but that the Township “will not be able to sustain its burden that the same was entered into in a knowing and voluntary fashion.” App. at 21. The complaint goes on to state that the agreement was not “knowing” because “[plaintiffs, at the time, were unaware that the same could be interpreted as foregoing a damage claim. They specifically were unaware of the precise extent of any claimed waiver.” App. at 22. The Living-stones made similar claims before the district court and on appeal.

Mrs. Livingstone was represented by counsel at her criminal trial; her attorney played a central role in the negotiation of the release-dismissal agreement. Under Rum-ery, the advice of counsel is an explicit, and important, element of the voluntariness analysis. See Rumery, 480 U.S. at 394, 107 S.Ct. at 1192-93; id. at 401, 107 S.Ct. at 1196 (O’Connor, J., concurring) (citing, as one of the factors bearing on the enforceability of a release-dismissal agreement, “importantly, whether the defendant was counseled”). 36 Mrs. Livingstone’s assertion that she did not appreciate the release-dismissal agreement’s legal implications is tantamount to a claim that her attorney did not give her accurate legal advice. It would be unfair to allow her to make this claim without permitting the opposing parties to investigate her attorney’s version of the relevant events. See United States v. Bilzerian, 926 F.2d 1285, 1292 (2d Cir.), cert. denied, 502 U.S. 813, 112 S.Ct. 63, 116 L.Ed.2d 39 (1991) (holding that a party who asserts a claim that “in fairness requires examination of protected communications” thereby waives the attorney-client privilege as to those communications). In the terms of the draft Restatement, Mrs. Livingstone has effectively asserted that the advice provided to her by her attorney is “relevant to the legal significance of [her] conduct.” Accordingly, we find no error in the district court’s ruling that the attorney-client privilege had been waived. 37

D. Exclusion of Expert Testimony.

The Livingstones challenge the district court’s decision, at the voluntariness proceeding, to exclude the testimony of then-expert, John Peters, who had prepared a report addressing the underlying liability of the police officers and of Washington Township. Mrs. Livingstone’s attorney, Ceraso, had testified that he had advised Mrs. Livingstone to conclude a release-dismissal agreement because any damages that she would recover in a subsequent civil suit would have been largely, or completely, offset by the damages that the police officers would recover, assuming that they filed counterclaims. App. at 767. The plaintiffs sought to introduce Peters’ testimony in order to demonstrate that Ceraso’s advice had been inaccurate. The district court found that Peters’ testimony was inadmissible under Rule 702, because it would not assist the trier of fact to understand the evidence, and under Rule 403, because of prejudice, confusion of the issues, misleading the jury, and waste of time. App. at 777.

We will address only the district court’s application of Rule 403, which we find was entirely appropriate. Peters’ report was quite likely to be prejudicial; it asserted, in considerable detail, that Washington Township and the police-officer defendants had violated Mrs. Livingstone’s civil rights. 38 A jury presented with a substantial amount of information on the merits of an 'underlying civil rights action might well look to those merits in making its decision on the distinct — and distinctly different — issue of vol-untariness, thus creating a significant risk of prejudice.

Rule 403 requires that a court balance the prejudicial effect of proposed evidence against its probative value. If evidence that a party to a release-dismissal agreement had received improper legal advice is of sufficient probative value, this analysis may well weigh in favor of admissibility. 39 The probative value of Peters’ testimony was not, however, high, as it did not engage Ceraso’s testimony directly. Ceraso’s advice to Mrs. Livingstone had addressed the net award of damages that she could expect from her potential civil suit against the police and their potential civil suit against her. Peters’ report only barely touched on the merits of a possible civil suit by the police against Mrs. Livingstone, and did not discuss the likely award of damages in either suit. 40 Thus, his testimony would not have greatly helped the jury to understand the correctness of Ceraso’s advice. 41

E. Exclusion of Trial Transcript.

At the trial of the voluntariness issue, counsel for the Livingstones sought to introduce into evidence an exchange between He-neks and Judge Cicehetti that occurred the day before the release-dismissal agreement purportedly was concluded. The district court found that this exchange was not relevant, and excluded it. The Livingstones appeal this ruling, asserting that Judge Cie-chetti’s comments in the exchange that they sought to introduce resembled his later remarks at the release-dismissal colloquy, and that the Livingstones might have been misled into believing that he was simply repeating his earlier comments. We agree with the district court’s finding that this exchange is not relevant. Judge Cicehetti’s comments in the portion of the exchange presented to the district court, App. at 790, bore little resem blance to his later comments at the release-dismissal proceeding, Appellees’ App. at 82.

VI. Application of Pennsylvania Law to the Livingstones’ State-Law Claims

As we noted in our discussion of the procedural history of this ease, the district court dismissed a number of the Livingstones’ state-law claims on grounds, such as the statute of limitations, unrelated to the release-dismissal agreement. The dismissal of those claims is not before us on appeal. The remaining state-law claims included claims of assault and battery against defendants Mo-nack and Snyder; a claim of intentional infliction of emotional distress against defendants Monack, Snyder, and Moody; and a claim of conversion against all defendants.

Neither the parties nor the district court have discussed what standard applies to determine the enforceability of the release-dismissal agreement as to the state-law claims. Instead, they have apparently assumed that the standard applicable to these claims is no different from that applicable to section 1983 claims. This is not necessarily true; the question whether the Livingstones have waived their claims under state law is itself one of state law, see Livingstone I, 12 F.3d at 1210 n. 6, and state law cannot be assumed to parallel federal law on this question.

In Livingstone I, we observed that the courts of Pennsylvania “frequently follow the principles set forth in the Restatement [of Contracts],” id., and suggested that the Pennsylvania Supreme Court might be likely to do as the United States Supreme Court did in Rumery, and look to the public-interest analysis in the Restatement of Contracts to determine when it is appropriate to enforce a release-dismissal agreement. See id. However, we did not then have occasion to decide precisely what standard Pennsylvania would apply to the enforcement of a release-dismissal agreement. That question is now before us. Indeed, that question subsumes two distinct questions: (1) What standard would Pennsylvania courts be likely to apply to determine whether the enforcement of a release-dismissal agreement is in the public interest? (2) What standard would Pennsylvania courts be likely to apply to determine the voluntariness of a release-dismissal agreement?

A. Public Interest

We have discovered no reported Pennsylvania cases addressing the question of when, if ever, it is in the public interest to enforce a release-dismissal agreement. Our analysis of the caselaw and policies of the Commonwealth of Pennsylvania has persuaded us, however, that the Supreme Court of Pennsylvania would apply a public-interest standard resembling that applied under federal law.

The courts of Pennsylvania have long declined to enforce contracts that are contrary to public policy. See, e.g., Kuhn v. Buhl, 251 Pa. 348, 96 A. 977 (1916) (finding unenforceable as against public policy an agreement between bidders for public lands under which one of them would, in exchange for a fee, withdraw its bid). After Kuhn, the Supreme Court of Pennsylvania accepted the standard set forth in section 320(1) of Tentative Draft No. 12 of the Restatement (Second) of Contracts (March 1, 1977) as its standard for the nonenforcement of contracts as against public policy. See Central Dauphin School District v. American Casualty Co., 493 Pa. 254, 426 A.2d 94, 96 (1981). Section 320(1) of the Tentative Draft was to emerge (with one minor stylistic alteration not relevant here) as Section 178(1) of the Restatement (Second) as finally adopted. It provides that “[a] promise or other term of an agreement is unenforceable on grounds of public policy if legislation provides that it is unenforceable or the interest in its enforcement is clearly outweighed in the circumstances by a public policy against the enforcement of such terms.” Restatement (Second) of Contracts § 178(1) (1981). We may reasonably conclude — as the Pennsylvania Superior Court has already concluded, see Donegal Mutual Insurance Co. v. Long, 387 Pa.Super. 574, 564 A.2d 937, 942 (1989) — that the Pennsylvania Supreme Court, having accepted tentative section 320(1) in Central Dauphin, would now accept permanent section 178(1).

In Rumery, the Supreme Court drew upon section 178(1) to fashion its federal common-law rule that a release-dismissal agreement will be unenforceable “if the interest in its enforcement is outweighed in the circumstances by a public policy harmed by enforcement of the agreement.” Rumery, 480 U.S. at 392 & n. 2, 107 S.Ct. at 1191-92 & n. 2. We think that the Pennsylvania Supreme Court would not only draw on section 178(1) in considering the enforceability of a release-dismissal agreement but, in construing that section’s open language, would look to Rum-ery and its progeny in the courts of appeals as persuasive authority. The Pennsylvania Supreme Court would, of course, also consider the policies of the Commonwealth of Pennsylvania, as expressed in the Commonwealth’s statutes and common law. The Pennsylvania Supreme Court has treated Pennsylvania’s common law as an important instrument for curbing official misconduct, paralleling at the level of state law the United States Supreme Court’s view of the policies underlying section 1983. See, e.g. Supervisors of Lewis Township v. Employers Mutual Casualty Co., 514 Pa. 242, 523 A.2d 719, 722 (1987) (finding that permitting insurance coverage of willful or fraudulent conduct on the part of a public official is contrary to Pennsylvania law and public policy, as personal financial liability is intended to deter official misconduct). We therefore conclude that the Pennsylvania Supreme Court would subject agreements that purport to waive tort liability to at least as careful scrutiny as the United States Supreme Court has applied to agreements purporting to waive liability under section 1983.

The Livingstones argue that the law of the Commonwealth of Pennsylvania strictly limits the private resolution of criminal charges. In support of this claim, they cite Pennsylvania Rule of Criminal Procedure 314, a rule permitting a form of court-supervised settlement in certain types of criminal cases. That rule provides:

When a defendant is charged with an offense which is not alleged to have been committed by force or violence or threat thereof, the court may order the case to be dismissed upon motion and a showing that:

(a) the public interest will not be adversely affected;

(b) the attorney for the Commonwealth consents to the dismissal;

(c) satisfaction has been made to the aggrieved person or there is an agreement that satisfaction will be made to the aggrieved person;

(d) there is an agreement as to who shall pay the costs.

Pa. R.Crim. P. 314. The Livingstones assert that the fact that this rule does not permit settlements in the case of offenses “alleged to have been committed by force or violence or threat thereof’ implies that such settlements are disfavored, or perhaps prohibited, under Pennsylvania law.

We are not persuaded that Rule 314 demonstrates that Pennsylvania would not permit release-dismissal agreements in other situations. A prosecutor who has sound public-interest reasons for declining to go forward with a prosecution, or for terminating a prosecution after it has begun, must have the authority to do so. As a corollary of this authority, a prosecutor presumably also has the authority to condition a dismissal on some undertaking by the defendant. 42

We find, however, that Rule 314 demonstrates that the courts of Pennsylvania would be likely to subject release-dismissal agreements to close scrutiny. Further, the fact that Rule 314(a) requires that a judge determine that “the public interest will not be adversely affected” by a dismissal demonstrates the Commonwealth’s commitment to reviewing release-dismissal agreements for their impact on the public interest. The Pennsylvania courts have also read Rule 314’s limitations on the circumstances in which criminal prosecutions may be dismissed to indicate that “the law does not favor out-of-court compromise over prosecution.” Commonwealth v. Pettinato, 360 Pa.Super. 242, 520 A.2d 437, 439 (1987) (concluding that an offer from a criminal defendant to pay a complainant a fee in exchange for her agreement not to testify was admissible into evidence in the defendant’s criminal trial; because Rule 314 strictly limits consensual dismissals in criminal cases, the civil rule of evidence barring the admission of offers of settlement into evidence did not apply).

In summary, then, we find that Pennsylvania would be likely to permit release-dismissal agreements to be enforced in some cases, but would monitor them closely to ensure that their enforcement is in the public interest. The federal rule, which places the burden of proving that a release-dismissal agreement is in the public interest on those seeking to enforce the agreement, has the same goals. Pennsylvania would therefore be likely to apply a very similar rule.

B. Voluntariness

The Court observed in Rumery that private citizens are permitted to waive their constitutional rights in many circumstances. For instance, criminal defendants may waive constitutional rights through plea bargaining, and the resulting agreements are ordinarily enforced if they are voluntary. Voluntary release-dismissal agreements, Rumery reasoned, should therefore also be permitted. See 480 U.S. at 393-94, 107 S.Ct. at 1192-93.

Pennsylvania, too, permits plea bargaining, see, e.g., Commonwealth v. Spence, 534 Pa. 233, 627 A.2d 1176, 1184 (1993), and will uphold a guilty plea if it is knowing and voluntary, see Commonwealth v. Alston, 473 Pa. 40, 373 A.2d 741, 743 (1977). We believe that the Pennsylvania Supreme Court would be likely to follow a line of logic similar to that of the Court in Rumery, and permit release-dismissal agreements upon a showing of voluntariness.

However, we anticipate that the Pennsylvania Supreme Court would be very attentive to how the voluntariness of a release-dismissal agreement is established. Such judicial attentiveness would be called for both because of the danger that such agreements will be concluded in improper circumstances, and because Pennsylvania has a policy of declining to enforce contracts concluded under duress or threat of prosecution. See, e.g., Germantown Mfg. Co. v. Rawlinson, 341 Pa.Super. 42, 491 A.2d 138,143 (1985) (applying a rule that threats of criminal prosecution constitute duress rendering a contract voidable, and stating: “It is an affront to our judicial sensibilities that one person’s ability to seek another’s prosecution can be bartered and sold the same as commodities in the market place. It is even more repugnant when the ford stench of oppression pervades the transaction.”). For reasons we have already discussed, the voluntariness of oral release-dismissal agreements is especially likely to be problematic, and — precisely because such agreements are not evidenced by a writing — determinations of the voluntariness of such agreements are particularly likely to be unreliable. See supra at 534-536. Accordingly, we predict that the Pennsylvania Supreme Court, when faced with the question, will subject the voluntariness of oral release-dismissal agreements to a heightened standard of proof, and we therefore conclude that the voluntariness of the release-dismissal agreement now before us must be demonstrated by clear and convincing evidence.

VII. Conclusion

For the reasons set forth above, we will vacate the judgment of the district court and remand for further proceedings in accordance with this opinion.

1 . The parties dispute whether Monack also told Mr. Livingstone he was under arrest.

2 . There is some dispute as to Moody’s whereabouts.

3 . Plaintiffs assert that the Commonwealth initiated settlement negotiations, but Heneks in his affidavit states that it was the trial judge who instigated the settlement discussions. We do not find this a material dispute. The effect of the agreement “is not dependent upon which party first suggests the release_” Boyd v. Adams, 513 F.2d 83, 88 (7th Cir.1975).

4 . There was testimony that on the day of the settlement, two of the three members of Washington Township’s Board of Supervisors were in the courthouse, were informed of the agreement, assented to it, and later telephoned the third supervisor, who also agreed. Washington Township has attached to its brief a document show ing an official ratification by its Board of Supervisors in February 1992, after this suit was filed.

5 . The R & R found that summary judgment should be granted as to the malicious prosecution claim because, as the underlying prosecution had ended with a compromise, it had not terminated favorably to Mrs. Livingstone. The R & R does not indicate whether this finding assumes the existence and validity of the release-dismissal agreement that is now before us, or — if it does not so assume — what the basis is for the finding that a compromise occurred.

6 . It is customary to refer to “the public interest,” rather than to "public policy," in discussing the enforceability of release-dismissal agreements, and this opinion will use the former terminology.

7 .The Livingstones also renew their argument that the district court erred in declining to require Washington Township to pay Mrs. Livingstone's household damages and medical bills. There is not now a controversy as to whether Washington Township will or must pay the Liv-ingstones' bills, as the Livingstones have not yet submitted those bills for payment. (We note that Washington Township has stated in its brief on appeal that it is willing to pay them if they are submitted.) We therefore agree with the district court that it is not at this time appropriate to order that Washington Township pay the Living-stones’ medical bills.

8 . Nor is it entirely clear how the performance of a contract between the Livingstones and either North Belle Vernon Borough or Fayette City Borough could have been rendered impossible by the Livingstones’ failure to submit their bills to Washington Township, the only entity that seems to have been prepared to assume the obligation of paying those bills.

9 . As we noted in Livingstone I, Washington Township’s board of supervisors also apparently gave some type of informal assent to the agreement at the time that it was concluded. See 12 F.3d at 1208 n. 4. There is some question as to whether this assent would satisfy the requirements of Pennsylvania law, however.

10 . There is some indication that the Livingstones may have reasonably believed that Washington Township was a party to the release-dismissal agreement. Immediately after Ceraso's remarks, quoted above, in the in-camera colloquy, Mr. Heneks said: "Your Honor, that is my understanding of the agreement that has been reached between the Washington Township Police Department and the Washington Township authorities that are present.” Appellees’ App. at 32. This statement might well have led the Living-stones to believe that Washington Township was a party to the agreement.

We will not decide here when, and whether, Washington Township became a party to the agreement. But, even if a mistake did occur as to the Township’s status as a party, it would not be enough to render the agreement voidable. In the words of the Restatement (Second) of Contracts, "Where a mistake of both parties at the time a contract was made as to a basic assumption on which the contract was made has a material effect on the agreed exchange of performances, the contract is voidable by the adversely affected party....” Restatement (Second.) of Contracts § 152(1) (1981). The agreed exchange of performances would not have been affected by the non-participation of Washington Township. As noted in the text, had Washington Township declined to pay the Livingstones' bills, or declined to sign a waiver, the Livingstones would have remained free to bring a civil suit, and hence would not have been harmed.

11 .The duty of good faith and fair dealing exists not only under federal common law, but also under Pennsylvania law. See, e.g. Somers v. Somers, 418 Pa.Super. 131, 613 A.2d 1211, 1214 (1992) ("In the absence of an express provision, the law will imply an agreement by the parties to a contract to do and perform those things that according to reason and justice they should do in order to carry out the purpose for which the contract was made and to refrain from doing anything that would destroy or injure the other party’s right to receive the fruits of the contract”) (quoting Frickert v. Deiter Brothers Fuel Co., 464 Pa. 596, 347 A.2d 701, 705 (1975) (Pomeroy, J., concurring)). Thus, the foregoing analysis applies both to the Livingstones' claims under federal law and to their state-law claims.

If the parties to an agreement intend for a third party to be a beneficiaty, the third party is entitled to enforce its rights under that agreement. See Restatement (Second) of Contracts § 304 (1981); see also Fizz v. Kurtz, Dowd & Nuss, Inc., 360 Pa.Super. 151, 519 A.2d 1037, 1039 (1987) (holding that a beneficiaty may enforce its right to performance if "recognition of a beneficiary’s right to performance is appropriate to effectuate the intention of the parties") (internal quotation omitted). Because the municipalities and police officers appear to have been intended beneficiaries of the alleged agreement, they have standing to insist that the Livingstones have a legal duty to perform their obligations under that agreement.

12 . This is an illustration of why release-dismissal agreements should be concluded in writing. See infra, pp. 534-536.

13 . We note that Ceraso seems to have had at least apparent authority to negotiate a release-dismissal agreement on Mrs. Livingstone’s be half. As a matter of contract interpretation, then, she would be bound by the terms of the agreement as he negotiated them, even if his understanding of the agreement’s terms differed from hers. (In his testimony before the district court, Ceraso also repeatedly referred to himself as representing Mr. Livingstone. See, e.g., App. at 713, 714, 715, 716, 719. The record does not make clear, however, whether (and, if so, when) an attorney-client relationship actually arose between Ceraso and Mr. Livingstone.)

A release-dismissal agreement is not, of course, an ordinary contract. Thus, if Mrs. Livingstone’s understanding of the two boroughs’ status under the agreement differed from Ceraso’s, this fact could be taken into account in the voluntariness analysis. In this regard, we observe that, should it become necessary on remand to conduct a new voluntariness proceeding, the district court should permit the Livingstones to argue that the release-dismissal agreement was voluntary as to certain of the municipalities but not as to others.

14 . Moreover, as we observed in Livingstone I, "whatever opportunity existed for further rupture in the Livingstone family had already been accomplished by the testimony of all family members save Mrs. Livingstone.” 12 F.3d at 1215.

15 . In the present case, the claim that the public interest in further prosecution was outweighed by the prosecution’s cost seems far from persuasive. As we observed in Livingstone I, "the agreement was struck, if at all, towards the conclusion of Mrs. Livingstone’s defense,” so that "the cost of the prosecution had already been incurred.” 12 F.3d at 1215. Any cost savings associated with terminating a prosecution that was nearing its end are likely to have been relatively small. On the other hand, the public interest in prosecuting Mrs. Livingstone may well have been substantial. Mrs. Livingstone was charged with assaulting a police officer, conduct which it is ordinarily thought to be in the public interest to punish.

16 . Cain implied that prosecutors might be excused from its requirement that they determine whether released civil rights claims appear marginal or frivolous in cases in which other public-interest reasons support enforcement of a release-dismissal agreement. See id. (referring to a showing that "the release-dismissal agreement advanced any other public interest”). This exception must necessarily be reserved for cases, like Rumery, in which unusually strong public interests support the release-dismissal agreement. See infra note 27. If prosecutors could routinely avoid assessing the merits of civil rights claims merely by citing such considerations as the avoided costs of trial — considerations that, as we have already noted, could be cited in virtually any case — Cain would have little force.

Indeed, in Cain, the prosecutor had required that a criminal defendant sign a civil release before entering Accelerated Rehabilitative Disposition (ARD). ARD is a program under which persons subject to relatively minor charges can undergo a probationary period in lieu of trial, and is intended precisely for cases in which the costs to the public of further prosecution outweigh the benefits of such prosecution. See Cain, 7 F.3d at 382-83. Thus, by consenting to ARD, the prosecutor in Cain had implicitly decided that it was not in the public interest to incur the costs of trial. Nevertheless, the Cain court found that the prosecutor was required to assess the merits of civil rights claims before concluding a release.

17 .Under Cain, we will only enforce a release-dismissal agreement on the basis of public-interest reasons for enforcing a release-dismissal agreement that were actually considered by the prosecutor who concluded that agreement. It is not now clear whether Heneks considered the marginal or frivolous nature of the Livingstones’ claims. Heneks’ affidavit discussing the circumstances under which the release-dismissal agreement was concluded does not state that he considered whether the Livingstones' possible civil rights claims were marginal or frivolous. App. at 627. However, the affidavit also does not foreclose the possibility that Heneks considered this question; the affidavit was signed in November 1991, some two years before Cain was decided, so that Heneks (and the appellants) can perhaps be excused for failing to address this issue in the affidavit. The Livingstones have seemingly never raised the applicability of this element of Cain in the district court. In these circumstances, we do not think that the appellees can fairly be treated as having waived the opportunity to address this matter in the district cotut. (Of course, if, on remand, it emerges that Heneks did not consider the marginal or frivolous nature of the Livingstones’ civil rights claims, that will end the district court’s inquiry, and the release-dismissal agreement will, under Cain, be unenforceable.)

18 . As a general matter, civil rights claims based on substantial evidence of official misconduct will not be either marginal or frivolous. But this may not be true in every case. For instance, if the official involved would clearly have absolute immunity for the alleged misconduct, then a subsequent civil rights suit might indeed be marginal, whether or not there is substantial evidence that the misconduct occurred.

19 . The report states:

Pelvic examination consisted of examination of the vulvar area and vaginal canal. The area is rather red and there is whitish bum like lesions. A magnifying lense [sic] was used and this appeared to be a first or ... second degree bum of the right side of the introitus. The left side is erythematous but does not have the same lesion as the right side. The canal itself is not red.

App. at 549. Webster's Third New International Dictionary (1966) defines "erythematous" as meaning "relating to or marked by erythema.” The latter word is in turn defined as “abnormal redness of the skin due to capillary congestion (as in inflammation).”

20 . At a deposition, Officer Monack stated, when asked by counsel for the Livingstones whether Monack had known the nature of Mrs. Livingstone’s claims of injury: "I would imagine. I think for some reason Mr. Ceraso asked Mr. Heneks if we could just go on the medical reports versus having to call the doctor in.” App. at 495. Officer Monack's testimony appears to describe an evidentiary stipulation of the sort that frequently occurs at trial. This stipulation presumably related to the emergency-room report, which was, according to Ceraso, App. at 227, 769-72, the only medical record available to him. It seems improbable that Heneks would not have familiarized himself with the contents of the report before stipulating to its authenticity.

Heneks was also one of the principal negotiators of the release-dismissal agreement. In those negotiations, Washington Township had indicated that it would pay Mrs. Livingstone’s medical bills. It would be curious indeed for Heneks to have conducted this negotiation without knowing the substance of the medical report itself. Heneks certainly had ample opportunity to learn what the report said; he was in regular communication with both Ceraso and Monack, both of whom knew the report’s contents. App. at 227, 495.

21 . Counsel for appellees asserted at oral argument that it is the position of the police-officer defendants that a stun gun had been used, but on Mrs. Livingstone's thigh. It is, perhaps, conceivable that one of the police officers inadvertently applied the stun gun to Mrs. Livingstone's genitalia, and that neither of the police officers involved in the encounter with Mrs. Livingstone noticed this error. Such a scenario, however, hardly seems likely; a trained police officer would presumably exercise great caution in the use of a stun gun, and would be likely to notice if he (or another officer) used such a gun improperly.

22 . Heneks did state, in an affidavit, that “I believed we had a sufficient case for conviction." App. at 625. But this, of course, does not imply that no police misconduct occurred; for instance, Mrs. Livingstone and Officer Monack may both have been at fault — i.e., each may have assaulted the other.

It is also true that Mrs. Livingstone’s own lawyer, Ceraso, testified that he had believed during the criminal trial that Mrs. Livingstone was unlikely to win a large damages award in a civil rights suit. App. at 227, 767. Upon close examination, however, Ceraso’s doubts seem to be more in the nature of pragmatic concerns about the likely extent of Mrs. Livingstone’s recovery than a belief that she did not have a viable cause of action. For instance, Ceraso noted that there were no follow-up medical records documenting her damages claims. App. at 227.

Finally, the Federal Bureau of Investigation did apparently inquire into whether criminal charges should be brought against Officer Monack based on the events of the night of January 12, 1989, and decided that they should not. App. at 457, 473. But, of course, criminal and civil standards of proof are different, and the section 1983 public interest in preventing police misconduct is by no means vindicated whenever an official body declines to bring criminal charges.

23 . A further item of evidence, also available to Heneks, should have suggested to him that not only the police-officer defendants, but also Washington Township, may have been responsible for violating Mrs. Livingstone's civil rights. This evidence was the testimony of Police Chief Matthews at Mrs. Livingstone's criminal trial, which suggested that the Township had a policy of returning children to their mothers, irrespective of “the welfare of the child or anything else.” App. at 425. As Magistrate Judge Mitchell observed, in a (subsequently adopted) Report and Recommendation finding it inappropriate to enter summary judgment in favor of the municipalities as to the Livingstones’ Fourth Amendment claims, this testimony could be read to suggest that it was Washington Township’s policy to act without securing a warrant — as apparently occurred in this case. App. at 332. (This is not, of course, the only interpretation of Matthews’ testimony.)

24 . For instance, it is at least conceivable that accurate information available to Heneks (but somehow omitted from the present record) indicated that the medical report of Mrs. Livingstone's bums was fabricated.

25 . The pertinence of what a prosecutor should have known stems from Cain's instruction that prosecutors examine the public interests at stake "on a case-by-case basis” before concluding release-dismissal agreements. 7 F.3d at 382. Charging prosecutors with constructive knowledge of any evidence of which they should have been aware helps to ensure that a would-be plaintiff is not harmed by a prosecutor's failure to carry out this duty.

26 . This distinguishes the present case from situations in which the principal facts are not at all clear, and mutual misconduct is one conceivable explanation. In such a case, as the Ninth Circuit observed in Lynch v. City of Alhambra, 880 F.2d 1122 (9th Cir.1989), there is at least an argument for the use of release-dismissal agreements:

One rationale for the use of these agreements is that they achieve a rough substantial justice where the "true” facts of the case are not known. For example, if the prosecutor is confronted with conflicting stories of police misconduct, and is genuinely unsure as to whom to believe, the execution of a release-dismissal agreement allows everyone to declare the case a draw and go home, thereby avoiding the risk and expense of going to tried.

Id. at 1127 n. 8. In a case in which there is substantial evidence that official misconduct has occurred, it is less appropriate to declare a case a draw. This is particularly true when one of the parties purportedly agreeing to a draw is the subject of a pending criminal prosecution.

27 . Cf. Coughlen v. Coots, 5 F.3d 970, 975 (6th Cir.1993) (finding that enforcement of a release may not be in the public interest in a case in which there is substantial evidence of police misconduct). For purposes of this analysis, it may be useful to distinguish between those interests of the public that are served simply by the termination of a prosecution (such as avoiding the costs of that prosecution), and those that can only be served by enforcing a release-dismissal agreement. A possible example of the latter type of interest is provided by Rumery itself. In Rum-ery the prosecutor's reason for concluding a release-dismissal agreement was to abort two trials — Rumeiy’s criminal trial and the civil suit expected to be brought by Rumeiy — which would have required the testimony of a potential witness whose testimony was needed by the prosecutor in another trial and for whom testifying was likely to be traumatic. Only enforcement of the release-dismissal agreement could have served this interest. See Seth F. Kreimer, Releases, Redress and Police Misconduct: Reflections on Agreements to Waive Civil Rights Actions in Exchange for Dismissal of Criminal Charges, 136 U. Pa. L.Rev. 851, 932-35 (1988) (arguing that the courts should pay close attention to this class of public interests in analyzing the enforceability of release-dismissal agreements).

28 . In Livingstone I, we observed that, although the ultimate question of whether enforcement of a release-dismissal agreement is in the public interest is a question of law for the court, "there may be factual issues intertwined with the legal issues, such as whether the public interest reason proffered by the prosecutor is the actual reason that motivated the prosecutor to enter into the release-dismissal agreement.” 12 F.3d at 1215. Prosecutorial motivation is, accordingly, a jury question.

The question whether there is substantial evidence of police misconduct, by contrast, is not. The process of weighing the evidence of police misconduct against the prosecutor’s asserted reasons for concluding a release-dismissal agreement is part of the broad task of balancing the public interests that favor and that disfavor enforcement. That task is one for the court. See Berry v. Peterson, 887 F.2d 635, 637 (5th Cir.1989).

29 . Both Justice Powell’s plurality opinion and Justice O’Connor’s concurring opinion found that there was strong evidence that the agreement at issue in Rumery had been concluded voluntarily. Indeed, Justice Powell said that it was “clear” that Rumery entered into the agreement voluntarily, 480 U.S at 398, 107 S.Ct. at 1194, while Justice O’Connor described the evidence that Rumery entered into the agreement voluntarily as "convincing." Id. at 403, 107 S.Ct. at 1197.

30 . An apt example is in a § 10(b) securities fraud claim, in which the Court has found that a preponderance standard is appropriate, because in such a case there is no reason to accord special deference to the interests of either plaintiffs or defendants. Id. at 390, 107 S.Ct. at 1190-91.

31 . The first amendment's petition clause protects a citizen's right of access to governmental mechanisms for the redress of grievances, including the right of access to the courts for that purpose. See Bieregu v. Reno, 59 F.3d 1445, 1453 (3d Cir.1995); San Filippo v. Bongiovanni, 30 F.3d 424, 439 n. 18, 443 (3d Cir.1994), cert. denied, -U.S. -, 115 S.Ct. 735, 130 L.Ed.2d 638 (1995).

32 .It is appropriate to include the public’s interest in detecting and deterring official abuse among the “particularly important individual interests or rights” that are at stake in a proceeding to determine the enforceability of a release-dismissal agreement. Cf. New York Times Co. v. Sullivan, 376 U.S. 254, 285-86, 84 S.Ct. 710, 729, 11 L.Ed.2d 686 (1964) (finding that actual malice must be demonstrated with "convincing clarity” in a libel action, and citing the public’s interest in protecting those who would criticize official conduct from the expense and risk of litigation).

33 . We note that there is some analogy between the rule that we apply today and the Statute of Frauds, which requires that certain important contracts be in writing in order to be enforceable. The Statute of Frauds is intended, inter alia, to encourage caution in consummating important transactions, and to ensure that there is reliable evidence that a transaction occurred and of its terms. See Restatement (Second) of Contracts § 110 statutory note at 286 (1981). The surrender of civil rights claims in a release-dismissal agreement is an important transaction. Moreover, that transaction ordinarily occurs while criminal charges are pending against one of the parties. Encouraging the parties to write their agreement down renders it less likely that one of them will act in haste. Cf. Restatement (Second) of Contracts § 112 cmt. a (1981) (noting that the inclusion of suretyship contracts in the Statute of Frauds "serves the cautionary function of guarding the promisor against ill-considered action").

Section 139 of the Restatement (Second) of Contracts, "Enforcement by Virtue of Action in Reliance," provides a list of circumstances relevant to whether an oral agreement should be enforced despite the Statute of Frauds. This list includes "the extent to which.., the making and terms [of the agreementj are ... established by clear and convincing evidence." See Restatement (Second) of Contracts § 139 (1981) (emphasis added).

34 . Appellants do not raise the question whether all oral release-dismissal agreements are unenforceable as a matter of law. Livingstone I declined to reach this issue, see Livingstone I, 12 F.3d at 1212, and so do we; our discussion of the standard of proof of voluntariness should not be read to indicate that we have decided this question. We also note that we need not, and do not, address the appropriate standard of proof for enforcement of a written release-dismissal agreement.

35 . Ceraso described himself in his testimony at the voluntariness proceeding as effectively representing both Mr. and Mrs. Livingstone in the 35. Ceraso described himself in his testimony at the voluntariness proceeding as effectively repre- senting both Mr. and Mrs. Livingstone in the in the

36 . This fact distinguishes the present case from Rhone-Poulenc, in which we found that a client who made reference to the advice of counsel in seeking to establish his state of mind did not thereby waive the attorney-client privilege as' to that advice. In Rhone-Poulenc, only the client’s state of mind, and not the substance of counsel’s advice, was an explicit element of the relevant legal analysis. See 32 F.3d at 864. In Rumery, by contrast, the Supreme Court made the advice of counsel an explicit element of the voluntariness analysis.

37 . We emphasize that this waiver is a limited one. See Greater Newburyport Clamshell Alliance v. Public Service Company of New Hampshire, 838 F.2d 13, 22 (1st Cir.1988) (finding, after conducting a balancing analysis, that the plaintiffs had waived the attorney-client privilege in some respects, but only as to information for which the defendants had "shown a true need and without which they would be unfairly prejudiced in their defense").

38 . His report concluded, for instance, that Officer Monack "knew or should have known that he was not properly trained to use and/or carry the Nova 5000 [stun gun],” App. at 635; that "Washington Township ... knew or should have known that its police were not properly trained to carry the Nova 5000,” and that this knowledge “amounted to deliberate indifference,” App. at 634; that "Officer Monack illegally re-entered the Livingstone residence, after he left the house,” App. at 636; that "Officers Monack, Snyder and Moody used unreasonable and excessive force on Frances Livingstone,” App. at 640; and that "the charges filed against Mrs. Livingstone were ‘cover charges' to justify both her arrest and the officers' excessive and unreasonable use of force." App. at 647.

39 . Otherwise, it would be difficult or impossible for a plaintiff to establish that incorrect legal advice had rendered a release-dismissal agreement involuntary.

40 . Peters’ report did state that the criminal charges against Mrs. Livingstone had been "manufactured.” App. at 647. The reasoning that led him to this conclusion was far from clear, however; indeed, his report stated that this opinion "will be further developed upon reading the Pennsylvania Crimes Code regarding these offenses." Id.

41 . We also note that there is some authority suggesting that the qualifications of experts on police practices should be subjected to close scrutiny. See, e.g., Berry v. City of Detroit, 25 F.3d 1342, 1349 (6th Cir.1994), cert. denied, U.S.-, 115 S.Ct. 902, 130 L.Ed.2d 786 (1995). The district court did not reach the question of Mr. Peters' qualifications to testify as an expert.

42 . Consider what would occur if a case like Rumery arose in the Commonwealth of Pennsylvania. Rumery had been charged with witness-tampering. Because he had allegedly threatened the witness with death, this was an offense "committed by force or violence or threat thereof”; thus, Rumery would not have been entitled to a dismissal under Rule 314. Nevertheless, given the strength of the public interests cited by the Rumery Court as supporting enforcement of the release-dismissal agreement in that case, we predict that the Pennsylvania Supreme Court would conclude that Rule 314 does not deprive prosecutors of the power to conclude release-dismissal agreements in at least some situations involving violent crimes.