ROVNER, Circuit Judge.
This appeal concerns a civil action brought under 42 U.S.C. § 1983 by Christopher Parish seeking damages for his wrongful conviction based on a violation of the Due Process Clause. The appeal, brought by Parish who prevailed in his § 1983 action, focuses on the adequacy of the damages award and the district court's restriction of evidence relating to it.
Parish's ordeal began on Halloween 1996, when he was arrested by Elkhart police officers as he prepared to take his three children trick-or-treating, and charged with attempted murder and armed robbery. He was 20 years old at the time. Parish maintained his innocence of the charge and informed the officers that he was at home and then at a family gathering in Chicago on the day that the crime was committed.
Before trial, Parish was offered a plea deal of one year in prison to avoid the risk of 50 years' imprisonment, but he refused to plead guilty. At the criminal trial, the evidence demonstrated that two African-American males, one tall and one shorter, had burst into an apartment at an Elkhart housing project and demanded drugs and money. When Michael Kershner, an occupant of the apartment, reached for his AK-47 semiautomatic rifle, a struggle ensued and eventually one of the intruders shot Kershner in his abdomen. Although severely injured, Kershner survived the attack. The assailants fled with a taser, an FK-19 rifle, and laundry money possessed by one of the occupants of the apartment. They left behind a hat that the taller of the two intruders had been wearing, which was a "custom" black leather baseball cap with rhinestones in the pattern of a "J" on it.
The Elkhart Police Department assigned Detective Steve Rezutko as the lead investigator of the shooting. One of the eyewitnesses to the incident was interviewed that night and stated that there were two men involved, that he had seen them around but did not know their names, and that one of them looked like a guy he knew, Chris Parish. Another eyewitness to the incident, Nona Canell, was provided with a photo book, and she identified the photo of Parish as one of the intruders, stating that when she saw his photo she was still in shock and in a reflex at seeing his photo she "dropped the book...." No physical evidence at the scene tied Parish to the crime. Based mainly on the eyewitness testimony, the jury convicted Parish of the crimes.
For the next eight years, Parish pursued available avenues of appeal. Eventually, after eight years of imprisonment, the Indiana appellate court in 2005 overturned his conviction and ordered a new trial based on the failure of his attorney to properly investigate and the introduction of an improper jury instruction. Parish was 30 years old at the time of his release from prison.
The government offered Parish a plea that would have resolved his case with no additional jail time, but Parish refused that offer. The government then decided not to proceed with a new criminal trial and dismissed the case.
Parish subsequently brought a civil action against a number of defendants alleging a deprivation of his constitutional right to due process, ultimately dismissing all defendants except the City of Elkhart and Steve Rezutko. The jury found in favor of Parish, and no challenge has been raised to the finding of liability. In fact, that determination is well-supported by the evidence introduced at trial.
Although the jury found in favor of Parish, it awarded him only $73,125 in compensatory damages and $5,000 in punitive damages for the eight years he was wrongly imprisoned. That verdict is astoundingly low for cases of wrongful conviction. In a motion for a new trial, Parish presented the court with evidence indicating that the average jury award was nearly $950,000 per year of wrongful imprisonment, with a median of nearly $790,000 per year. The award here of approximately $9,000 per year is an extreme outlier. In fact, the district court accepted that point, noting that the defendants had essentially conceded that the verdict in this case was "out of line" with many other cases in which a person was incarcerated due to a violation of the right to a fair trial.
Parish argued in his motion for a new trial that as to the issue of compensatory damages, the damage award was so low as to lack any rational basis, and that the trial court erred in improperly limiting the evidence that Parish could introduce at trial. The court denied that motion, and we review that denial for abuse of discretion. Whitehead v. Bond, 680 F.3d 919, 927-28 (7th Cir.2012).
The defendants and the court, in response to the motion for a new trial, asserted as a justification for the low damages that the jury probably determined that Parish was guilty of the underlying offense. A jury that believed the plaintiff was guilty of the crime would award lower damages because the imprisonment is attributable to the person's own actions as well as the civil defendants' misbehavior and even a fair prosecution and trial may well have resulted in the person's imprisonment. In light of the relatively uncontested nature of the damages and the undeniable hardship of imprisonment, the theory advocated by the defendants and the court is the only sensible explanation for an otherwise unfathomably-low damages amount.
It also is the reason why a reversal of the damages award is necessary in this case. One of Parish's arguments for reversal is that the district court refused to allow him to present significant evidence that he was not guilty of the offense. Parish sought to introduce substantial evidence casting doubt on the veracity of the eyewitness identifications and on the reliability of the investigation as a whole. That evidence included identification of other individuals as possible perpetrators, and recantations by the eyewitnesses of their identification of Parish. The trial court's refusal to allow such evidence was error. Moreover, the acknowledgment by the defendants and the court of the primacy of the innocence issue makes clear that the error was not harmless.
A look at the evidence allowed and that withheld from the jury on the question of responsibility for the crime reveals that the deck was effectively stacked against Parish. Significant testimony as to Parish's guilt of the crime, and particularly the testimony of eyewitnesses identifying him, was admitted whereas testimony as to his innocence, including statements by those same eyewitnesses expressing their doubts as to that identification, was excluded. The result was that the jury was deprived of significant probative evidence as to the issue of Parish's guilt or innocence. That appears to have been an unfortunate consequence of a decision made under the pressure of time, and not subsequently altered. The court recognized the potential for a reversible error when it was presented with extensive motions in limine on the eve of trial. In fact, the court encouraged the parties to continue the case to allow adequate time to delve into the issues presented in those motions. Parish was willing to do so, but the defendants objected to the option. The defendants, with their exposure to the case preparation, should have anticipated that such innocence evidence would be critical to a fair trial, but they nevertheless sought and obtained its exclusion. The court's concern that the decision might result in reversible error proved prescient.
As was mentioned, Parish was convicted largely based on the testimony of eyewitnesses identifying him as one of the perpetrators. The detective responsible for the investigation of the crime, Rezutko, had a troubling history as a lead investigator. At trial, Larry Towns, a former supervisor of Rezutko as a captain in the Elkhart Police Department, testified that he decided Rezutko should no longer be assigned as a lead investigator in homicide cases prior to October 1996 (when the Parish investigation occurred.) Towns detailed his concerns with how Rezutko conducted investigations, including the repeated use of suggestive photo lineups, steering of witnesses in identifications, and the unconventional solicitation of multiple statements from witnesses with additional helpful details included in the consecutive statements. Despite those concerns, the Elkhart Police Chief refused to remove Rezutko from the detective bureau entirely and allowed him to continue as lead investigator for non-homicide investigations.
As a result of Rezutko's investigation, two persons were convicted for the crimes of that night — Parish and Keith Cooper. Cooper was prosecuted and convicted as the second participant — and shooter. His conviction was subsequently overturned by the courts as well. Cooper was then offered a plea deal, that would allow him to remain free based on time served if he pled guilty, and he accepted that offer. Accordingly, Cooper stands convicted as one of the two perpetrators of the crime. In this § 1983 action, Parish sought to introduce testimony from Cooper that Cooper had never met Parish and had never committed any crime with him, but the court excluded that testimony.
Parish also had evidence pointing to another possible perpetrator. One of the assailants was wearing a hat with a "J" in rhinestones on it that was knocked off in the fracas. A DNA analysis of the hat matched Johlanis Ervin, a person who was tall and who was currently in prison charged along with his brother Michael, who was of shorter height, with murder. When two of the eyewitnesses were shown a picture of the Ervin brothers, they indicated that the brothers looked like the intruders and that they no longer had confidence in their identification of Parish.
The court's decision regarding the admissibility of that eyewitness testimony perfectly encapsulates the problems with this trial, and alone requires reversal. The court allowed the defendants to read the deposition testimony of eyewitness Canell identifying Parish as the perpetrator of the crime, including the dramatic testimony that she dropped the photo book when she saw Parish's picture. The court redacted, however, the deposition testimony in which Canell stated that Johlanis Ervin's brother Michael looks like Christopher Parish's twin, and that she is now unsure as to the identification. Similarly, the jury heard Kershner's testimony that he easily picked out Parish from the lineup and could have sworn it was him, but the court refused to allow the testimony that Kershner has since been shown pictures of the Ervin brothers and believes he may have misidentified Parish, and would no longer testify that Parish was the intruder. In addition, the jury did not hear their deposition testimony that — at Rezutko's urging — they had lied at the criminal trial about being certain of their identification of Parish as the perpetrator. That decision to redact only the exculpatory portions of the eyewitness testimony skewed the testimony to such an extent that it no longer resembled its true nature. It fundamentally misrepresented what the eyewitnesses actually believe, and it deprived the jury of critical information.
In choosing to so limit the evidence, the district court stated that the prejudice outweighed the probative value. A district court may exclude relevant evidence if the court determines that "its probative value is substantially outweighed by a danger of one or more of the following: unfair prejudice, confusing the issues, misleading the jury, undue delay, wasting time, or needlessly presenting cumulative evidence." Federal Rules of Evidence 403; Whitehead v. Bond, 680 F.3d 919, 930 (7th Cir.2012). The most relevant evidence is, by its nature, prejudicial, but it is only unfair prejudice that requires exclusion. Id.; United States v. Boros, 668 F.3d 901, 909 (7th Cir.2012). Evidence is unfairly prejudicial if it would cause the jury to decide the case on an improper or irrational basis, such as by appealing to the jury's emotions. United States v. Miller, 688 F.3d 322, 327 (7th Cir.2012). Both probative value and prejudice must be determined in context, and as the probative value increases, so does our tolerance of the risk of prejudice. Boros, 668 F.3d at 909; Whitehead, 680 F.3d at 930. We review a district court's Rule 403 determination for abuse of discretion. If we determine that the court abused its discretion in admitting or excluding evidence under Rule 403, we will reverse only if the error was not harmless. Boros, 668 F.3d at 910; Whitehead, 680 F.3d at 930.
In excluding the evidence of innocence here, the court provided little explanation as to what prejudicial impact such testimony could present. The court at times focused on what evidence was known at the time of the crime, limiting the ability of Parish to produce evidence that was not known at that time. That limitation, however, is relevant only regarding the issue of liability for wrongful conviction. In assessing the adequacy of the damages amount, the trial court itself recognized that Parish's guilt or innocence was a significant issue for the jury in determining the appropriate damage award, and that issue is not dependent on what information was available at the time of the original trial.
At oral argument, the defendants stated that allowing the evidence would be prejudicial because it would entail two or three mini-trials as to the guilt of Cooper, the Ervins and Parish. The mini-trials of which the defendants complain, however, are merely the necessary engagements on the critical issues of the trial. As the court acknowledged, Parish's guilt or innocence was an issue in determining the appropriate damages. In preventing Parish from introducing evidence pointing to the guilt of others for the crime, and in prohibiting evidence establishing the unreliability of the eyewitness identification as to the perpetrators of the crime, the court did not avoid mini-trials on Parish's innocence — it simply guaranteed that the mini-trial on the issue would include only evidence supporting one side. Nor did it apparently consider a more nuanced limitation on the evidence that could be introduced. For instance, even if the court believed that references to Cooper and Ervin were prejudicial — a position for which we find no support — the court could have allowed the jury to hear that the eyewitnesses no longer believe that Parish is the perpetrator and would no longer testify to that effect, and excised references to the Ervins.
The defendants inexplicably argue that the evidence that Parish sought to introduce did not establish Parish's innocence, but rather addressed only the guilt or innocence of other persons — specifically Cooper and the Ervins. There is no merit to this facile argument. If Parish can establish that another individual is responsible for the crime, that certainly is evidence that Parish is not the one responsible — rather strong evidence at that. The evidence regarding Cooper indicated that the investigation and prosecution of the case was problematic and the eyewitness identification unreliable. DNA linking the "J" hat to Johlanis Ervin cast doubt on the Cooper conviction, as Johlanis was taller like Cooper. Although that might call into question the eyewitness identification regarding Cooper, the defendants argue that invalidating the eyewitness testimony as to Cooper has no relevance as to the identification of Parish as the other assailant. The eyewitnesses that identified Cooper, however, are the same persons who identified Parish, and the testimony involved their recollection and observation of the same event. It is disingenuous to assert that evidence indicating that their identification was wrong as to Cooper does not call into question the reliability of their identifications for the crime as a whole. Therefore, that evidence was not "unfairly" prejudicial. The evidence as to the reliability of the eyewitness identification as to the events of the night was relevant to Parish's guilt as well as Cooper's, and any danger that the jury would be swayed by emotion or bias upon hearing of the misidentification of Cooper is minimal at best.
As Cooper is the only person who stands convicted of the offense, his testimony that Parish did not commit any crimes with him is of obvious relevance. The court, however, refused to allow Parish to introduce Cooper's statement that he did not know Parish and that Parish had never committed any crime with him. In fact, it is difficult to see how that testimony would be unfairly prejudicial at all. The reluctance to allow such evidence appears to reflect the court's concern with the reliability of the evidence that Cooper was involved in the crime. Such doubts may well exist in light of the potential connection to the Ervins, although the Ervins have never been charged. But if the court doubted Cooper's guilt because of the potential involvement of the Ervins, then there is no justification for also preventing Parish from introducing the statement by Johlanis that he had never met Parish. Like Cooper, Johlanis Ervin would have testified that he did not know Parish and never committed any crimes with him. Therefore, both the person convicted of the offense and the person implicated by the DNA evidence would have testified that they did not know Parish and that he did not commit any crime with them, and neither was allowed. That evidence was relevant to Parish's innocence, and there is no basis for holding it unfairly prejudicial.
Finally, the defendants also contend that the low jury damages award was justified because Parish was never exonerated of the charge, and therefore the jury could properly take that into account in determining that he was responsible for the crime. Given the procedural progression of the case, this argument is insincere. Parish refused to enter into a plea bargain at every opportunity in the criminal case, choosing to assert his innocence at trial. Even after serving eight years in prison, when the government offered him a plea deal that would have allowed him to serve no additional time if he pled guilty to the charge, Parish declined the offer. The government then decided to dismiss the case, depriving Parish of the ability to obtain that "exoneration" that the defendants now claim he lacks. There can be no adverse inference of guilt from such a history. Moreover, even if the jury could consider that, it would not render the exclusion of the other evidence harmless. The testimony allowed as to the issue of Parish's guilt or innocence was overwhelmingly skewed; there is no question that the evidentiary errors were not harmless.
Because the district court's rulings improperly limited the introduction of evidence relating to Parish's innocence, and that evidence was critical to the damages issue, the award of damages cannot stand. The excluded evidence did not impact the jury's consideration of the liability issue and that issue is not before us on appeal, and therefore a new trial is required only as to the damages issue. See Cobige v. City of Chicago, IL, 651 F.3d 780, 785 (7th Cir.2011). Accordingly, the jury's determination of liability is affirmed, the award of damages is vacated, and the case remanded for a new trial as to the issue of damages only. Circuit Rule 36 shall apply on remand. Costs on appeal are to be taxed against appellees.
