Christopher Gish v. Randall Hepp

• Court: Court of Appeals for the Seventh Circuit
• Decided: April 3, 2020
• Docket: 19-1476
• Status: Published

Opinion

In the

United States Court of Appeals For the Seventh Circuit ____________________ No. 19-1476 CHRISTOPHER R. GISH, Petitioner-Appellant, v.

RANDALL HEPP, Warden, Respondent-Appellee. ____________________

Appeal from the United States District Court for the Western District of Wisconsin. No. 3:15-cv-730 — James D. Peterson, Chief Judge. ____________________

ARGUED NOVEMBER 7, 2019 — DECIDED APRIL 3, 2020 ____________________

Before HAMILTON, SCUDDER, and ST. EVE, Circuit Judges. SCUDDER, Circuit Judge. Christopher Gish pleaded guilty to first-degree reckless homicide in Wisconsin state court for killing his longtime girlfriend and the mother of his children. He appealed, claiming that his trial counsel provided ineffec- tive assistance by failing to investigate an involuntary intoxi- cation defense. Police found Gish disoriented and delirious on the night of the killing, and he claimed that rare side effects from taking prescription Xanax affected his ability to 2 No. 19-1476

appreciate the wrongfulness of his conduct. After the Wiscon- sin Court of Appeals rejected the claim and affirmed his con- viction, Gish turned to federal court and wound his way through a thicket of habeas proceedings. The district court held an evidentiary hearing but denied relief because Gish failed to show that his counsel’s deficient performance re- sulted in prejudice: even if counsel had investigated involun- tary intoxication, that defense was so unlikely to succeed that Gish still would have pleaded guilty. We affirm. I A Early in the morning on July 14, 2012, Wisconsin police found Christopher Gish soaking wet, unable to answer ques- tions, and wandering in an unsteady manner on railroad tracks near the Milwaukee airport. The officers took Gish to the hospital, where he told paramedics that he had blacked out. He then proceeded to make a series of nonsensical state- ments suggesting that he did not understand his wherea- bouts. At one point, for instance, Gish stated that “all I saw was red” and “you are in my bedroom, why are you in my room?” Upon ascertaining Gish’s home address, the police entered and found his longtime girlfriend and the mother of his children, Margaret Litwicki, stabbed to death in a bed- room. Once Gish’s condition stabilized, he agreed to an inter- view with the police. A videotape showed that Gish gained lucidity over the course of the questioning. Initially Gish de- nied any memory of the previous night, but later in the inter- view he confessed to stabbing Litwicki multiple times in his bedroom. He said he attacked Litwicki because he suspected No. 19-1476 3

that she was having an affair and believed she might take his kids from him. Wisconsin authorities charged Gish with first-degree in- tentional homicide, which carries a mandatory sentence of life imprisonment. See WIS. STAT. §§ 939.50(3)(a), 940.01(1)(a). Na- than Opland-Dobs served as Gish’s court-appointed counsel. Gish told Opland-Dobs that he had taken prescription Lamic- tal and Xanax before the homicide and thought those medica- tions may have induced his erratic behavior in a way that would afford some legal defense to the charge. Opland-Dobs researched the effects of Lamictal, but not Xanax—a choice he later said he could not explain. He ulti- mately determined that any Lamictal-based defense would be futile and so advised Gish. When prosecutors later offered to accept a plea to first-degree reckless homicide, which carries a maximum sentence of 60 years, see WIS. STAT. §§ 939.50(3)(b), 940.02(1), Opland-Dobs advised Gish to take it. Gish agreed, pleaded guilty, and received a sentence of 40 years’ imprisonment and 20 years’ extended supervision. B With the assistance of new counsel, Gish filed a direct ap- peal in Wisconsin state court. Counsel then filed what Wis- consin law calls a “no-merit report”—the functional equiva- lent of an Anders brief in federal criminal practice—represent- ing that any appeal would be meritless and requesting per- mission to withdraw as Gish’s appointed lawyer. See WIS. STAT. § 809.32 (setting out Wisconsin’s procedure for filing no- merit reports); accord Anders v. California, 386 U.S. 738, 744 (1967) (advising that “if counsel finds his case to be wholly 4 No. 19-1476

frivolous, after a conscientious examination of it, he should so advise the court and request permission to withdraw”). Gish responded to the no-merit report by insisting that he had a non-frivolous basis for appeal. He claimed that his trial counsel, Opland-Dobs, provided ineffective assistance by fail- ing to pursue the affirmative defense of involuntary intoxica- tion, a complete defense to homicide under Wisconsin law. Gish emphasized that he told Opland-Dobs all about the Xanax he had taken before the homicide and suggested that the medication may have affected his ability to discern right from wrong. See WIS. STAT. § 939.42(1). He supported this con- tention with police reports describing his delirium shortly af- ter the homicide, medical records showing he had been pre- scribed Xanax, and information about Xanax’s side effects that he had found online and in textbooks. Gish then went a step further: he insisted that, had he known an involuntary intoxication was viable, he would have rejected the govern- ment’s plea and instead gone to trial. Appellate counsel responded by emphasizing that Gish never once suggested to his trial counsel, Opland-Dobs, that either the Xanax or Lamictal so affected his mental state as to prevent him from understanding the wrongfulness of his con- duct. So, appellate counsel put it, “there wasn’t anything to investigate.” The Wisconsin Court of Appeals evaluated Gish’s ineffec- tive assistance claim under the familiar standards of Strickland v. Washington, 466 U.S. 668 (1984). Gish had to show that Op- land-Dobs’s performance “fell below an objective standard of reasonableness,” id. at 688, and resulted in prejudice, meaning that there was “a reasonable probability that, but for counsel’s No. 19-1476 5

unprofessional errors, the result of the proceeding would have been different,” id. at 694. The Wisconsin court denied relief, concluding that any contention of ineffective assistance was so lacking—having no “arguable merit”—that Gish could not even clear Strickland’s first hurdle of showing that Opland-Dobs’s performance was deficient. Indeed, the court wholesale adopted Gish’s appel- late counsel’s version of events, disregarding Gish’s allega- tions in their entirety and even refusing to consider the police reports and other documents Gish submitted in support of his ineffective assistance claim. In effect, then, the Wisconsin court affirmed Gish’s conviction for the same reason sug- gested by his appellate counsel—“there wasn’t anything to in- vestigate.” The Wisconsin Supreme Court denied review, and Gish then turned his attention to securing relief in federal court. II A Invoking 28 U.S.C. § 2254, Gish petitioned the district court for federal habeas relief, renewing his claim that Op- land-Dobs provided ineffective assistance of counsel by fail- ing to investigate a Xanax-based involuntary intoxication de- fense. To secure relief, Gish had to establish that the Wiscon- sin Court of Appeals’s decision “was contrary to, or involved an unreasonable application of, clearly established Federal law,” or “was based on an unreasonable determination of the facts in light of the evidence presented in the State court pro- ceeding.” 28 U.S.C. § 2254(d)(1)–(2). Although ultimately denying relief, the district court did so only after holding an evidentiary hearing, taking 6 No. 19-1476

testimony, and receiving other evidence on the merits of Gish’s contention that Opland-Dobs should have pursued an involuntary intoxication defense. The district court deter- mined the evidentiary hearing was warranted, and indeed necessary, because Gish, despite offering his prescription rec- ords, the police reports, and information about the side effects of Xanax, never had a reasonable opportunity to develop the factual basis for his claim on direct appeal in the state court. Even more, the district court found that Gish’s allegations, if true, supported his claim that Opland-Dobs performed defi- ciently. The state court’s back-of-the-hand rejection of Gish’s ineffective assistance claim, the district court concluded, re- flected an unreasonable application of Strickland, for Gish had brought forth enough evidence on direct appeal to reasonably question the adequacy of Opland-Dobs’s representation in the trial court. B Several witnesses testified at the evidentiary hearing. Gish testified on his own behalf and called pharmacology consult- ant James T. O’Donnell and his trial counsel Nathan Opland- Dobs. For its part, the state called Kayla Neuman, a chemist in the toxicology section of the Wisconsin State Laboratory of Hygiene, and Detective Brent Hart, who had interviewed Gish the morning he was apprehended. The district court heard conflicting evidence about whether Gish took Xanax on the day he killed Litwicki. On the one hand, Gish testified that he told Opland-Dobs he had taken both Xanax and Lamictal on the day of the homicide. But Gish plainly stated in the interview with Detective Hart the morning of the homicide that he had last taken Xanax “[a] couple days” before, which, given the half-life of Xanax, No. 19-1476 7

would suggest that its effects had worn off by the time of the killing. In much the same vein, a nurse who treated Gish at the hospital wrote in his patient visit records that Gish re- ported having sold his Xanax and Lamictal pills—suggesting that perhaps he had never taken them at all in the days before the homicide. And the district judge heard testimony that the police found no Xanax in a search of Gish’s home. The district court also heard expert testimony about the possible effects of Xanax. Both parties’ experts agreed that Xanax can trigger hallucinations, agitation, rage, and hostile behavior. The state’s expert, Neuman, added that mixing Xanax with Lamictal can amplify these effects. Gish’s expert, O’Donnell, testified that the police finding Gish in a tempo- rary delusional state was more consistent with Xanax intoxi- cation than with the effects of mental illness. O’Donnell added that Gish could not appreciate the criminality of his conduct, but the district court found that conclusion speculative, backed by no medical evidence, and therefore not credible. Finally, the district court heard from Gish and Opland- Dobs regarding their plea discussions. For the most part, their accounts aligned: Gish testified that he had asked Opland- Dobs to consider defenses based on Xanax and Lamictal. Op- land-Dobs did not dispute that aspect of Gish’s testimony, ad- mitted that he failed to investigate Xanax, and expressed re- gret for that failure. He conceded that, given the evidence he had available to him in representing Gish, investigating Xanax would have been “appropriate” and he “didn’t give it enough consideration.” Opland-Dobs offered no justification for this failure, saying, “[w]hy I didn’t follow up on the Xanax, I can’t explain,” because ignoring that path “doesn’t seem like what I should have done.” 8 No. 19-1476

On the question of prejudice, Gish testified that he only pleaded guilty on the assumption that he would have had a “zero percent chance” of being acquitted at trial. He explained that there was “no sense” in “putting the family through” a trial “that was just a wish-wash,” where he believed he had no chance of prevailing. But Gish was equally clear that his decision may have been different had Opland-Dobs pursued the involuntary intoxication defense and told him it had some chance of prevailing. Even if that defense were a weak one, giving him as low as a “one-percent chance” of acquittal, Gish insisted he would have “always take[n] the chance” and rolled the dice at trial. C Aided by the evidentiary hearing, the district court pro- ceeded to the merits of Gish’s ineffective assistance claim. The court made quick work of Strickland’s deficient performance prong by assuming that Opland-Dobs’s complete and admit- ted failure to evaluate a Xanax-based intoxication defense was unreasonable. Moving to Strickland’s prejudice prong, the court concluded that Gish fell short of showing he would have forgone the plea deal and gone to trial had Opland-Dobs pursued the defense. While Gish so testified, the district court was not willing to credit that testimony over other evidence pointing in the opposite direction. The district court placed particular emphasis on Gish’s statements to Detective Hart not only that he had last taken Xanax “[a] couple days” before the homicide, but also that he did not regret killing Litwicki in light of her alleged infidelity. The district judge likewise highlighted Gish’s statement to the nurse that he had sold his prescriptions—a fact corroborated by the police’s failure to find any trace of Xanax in Gish’s No. 19-1476 9

home. Considering this evidence in its totality, the district court determined that Gish had no reasonable prospect at trial of demonstrating the essential element of the intoxication de- fense—that he failed to appreciate right from wrong at the time of the homicide. The district court also found that the state’s plea offer was reasonably attractive, as it guaranteed Gish a maximum of 60 years rather than life imprisonment. Gish now appeals. III A We begin with the decision of the Wisconsin Court of Ap- peals, the last state court to consider (at least a portion of) Gish’s ineffective assistance claim on the merits in a reasoned opinion. See Wilson v. Sellers, 138 S. Ct. 1188, 1192 (2018). Gish needs to show, as the district court recognized, that the Wis- consin court’s decision “was contrary to, or involved an un- reasonable application of, clearly established Federal law,” or “was based on an unreasonable determination of the facts in light of the evidence presented in the State court proceeding.” 28 U.S.C. § 2254(d)(1)–(2). In answering that question, we must “train [our] attention on the particular reasons—both le- gal and factual—why state courts rejected [Gish’s] federal claims.” Wilson, 138 S. Ct. at 1191–92. Where, as here, the state court issued an explanatory opinion, we “review[] the specific reasons given by the state court and defer[] to those reasons if they are reasonable.” Id. at 1192. The Wisconsin Court of Appeals rejected Gish’s ineffective assistance claim on the ground that “there wasn’t anything [for his trial counsel, Nathan Opland-Dobs] to investigate.” With nothing to investigate, the reasoning ran, Opland-Dobs 10 No. 19-1476

could not have rendered ineffective assistance. It made no dif- ference, the Wisconsin court added, that Gish sought on ap- peal to support his claim with police reports and other evi- dence showing that his prescription Xanax may have ex- plained his delusional state at the time of the homicide. None of that evidence was before the trial court and that is all that mattered on the Wisconsin court’s reasoning. The district court was right to call the Wisconsin court’s decision an unreasonable application of Strickland’s deficient performance prong. Return to the state court’s insistence that Gish’s claim lacked merit because (and only because) he never put his evidence before the trial court. That reasoning fails to meet the claim Gish raised on direct appeal—ineffective assis- tance of his trial counsel, Nathan Opland-Dobs. As the Wis- consin court would have it, Gish—while being advised by Opland-Dobs—somehow and some way (and apparently on his own) had to put before the trial court evidence to support a claim that Opland-Dobs had violated the Sixth Amendment by not pursuing an involuntary intoxication defense. Yet the trial record lacked evidence of Gish’s ineffective assistance claim precisely because, by the very terms of the claim, Op- land-Dobs’s deficient performance occurred during the trial court proceedings. Gish, in short, necessarily needed to sup- port his claim with evidence outside the trial record, for there was no other way he could have demonstrated his ineffective assistance claim or rebutted his appellate counsel’s view (as reflected in the no-merit report) that the claim was frivolous. This is not the first time we have found fault with the exact reasoning the Wisconsin Court of Appeals employed in reject- ing Gish’s ineffective assistance claim. In Davis v. Lambert, we explained that “it would defy logic to deny [a state habeas No. 19-1476 11

petitioner] an evidentiary hearing on whether his counsel’s failure to investigate the witnesses violated Strickland on the ground that he did not fully present those witnesses’ testi- mony to the state courts.” 388 F.3d 1052, 1061 (7th Cir. 2004). Similarly, in Mosley v. Atchison, we concluded that a state court unreasonably applied Strickland’s performance prong by disregarding a defendant’s showing on appeal that his trial counsel failed to pursue two potential alibi witnesses and in- stead assuming that counsel’s choice reflected a strategic de- termination. 689 F.3d 838, 848 (7th Cir. 2012). We chart the same course here and have little difficulty concluding that the Wisconsin Court of Appeals’s denial of Gish’s ineffective assistance claim rooted itself in an “unrea- sonable application” of Strickland’s deficient performance prong as well as an “unreasonable determination of the facts in light of the evidence [Gish] presented in the State court pro- ceeding.” 28 U.S.C. § 2254(d)(1)–(2). Gish brought forth spe- cific evidence that, if accepted as true, would have demon- strated that Opland-Dobs rendered deficient performance in failing to pursue a potential involuntary intoxication defense. See Jones v. Wallace, 525 F.3d 500, 503 (7th Cir. 2008) (noting that where a petitioner in state custody is “not at fault for fail- ing to develop the factual record” of his ineffective assistance claim, we “look only to whether, if proven, his proposed facts would entitle him to relief”). The Wisconsin Court of Ap- peals’s contrary conclusion reflected an unreasonable appli- cation of Strickland. In these circumstances, the same error sat- isfies § 2254(d)(2), for the Wisconsin court’s categorical disre- gard of Gish’s evidence resulted in a rejection of his ineffec- tive assistance claim on an unreasonable view of the facts. At the very least, all of this was enough, as the district court rec- ognized, to warrant an evidentiary hearing—to afford Gish an 12 No. 19-1476

opportunity to develop the merits of his claim, an opportunity he never received in state court. Like the district court, then, we proceed to the merits of Gish’s ineffective assistance claim. B In considering Gish’s claim, we need say very little on Strickland’s first prong. Opland-Dobs testified in the district court and admitted in no uncertain terms that he never as- sessed a Xanax-based involuntary intoxication defense. We can assume this admitted failure is enough for Gish to show deficient performance. See Pole v. Randolph, 570 F.3d 922, 943 (7th Cir. 2009) (opting to “assume that counsel’s performance was deficient and move on to the second part of the analysis” because the petitioner could not show prejudice). This brings us to the primary issue on appeal: whether Op- land-Dobs’s failure to pursue an involuntary intoxication de- fense prejudiced Gish. Our review proceeds de novo (and not under the deferential standard of § 2254(d)) because this di- mension of Gish’s claim is one the Wisconsin Court of Ap- peals never reached and considered. That court stopped at Strickland’s first prong. In these circumstances, the Supreme Court has instructed, we treat the two prongs of Strickland as divisible and review the prejudice prong by taking our own fresh look at the evidentiary record developed in the district court. See Rompilla v. Beard, 545 U.S. 374, 390 (2005) (reviewing Strickland prejudice de novo because the state court did not reach that issue); see also Thomas v. Clements, 789 F.3d 760, 766–67 (7th Cir. 2015) (collecting cases adhering to this same approach). The controlling substantive standard comes from Hill v. Lockhart, 474 U.S. 52 (1985). The Court decided Hill one year No. 19-1476 13

after Strickland and did so to articulate what a defendant must show to establish that his trial counsel rendered ineffective as- sistance in advising him to plead guilty. First, and in full alignment with Strickland, the defendant must show that his counsel’s performance fell below an objective standard of rea- sonableness. See id. at 58. Second, when it comes to prejudice, “the defendant must show that there is a reasonable probabil- ity that, but for counsel’s errors, he would not have pleaded guilty and would have insisted on going to trial.” Id. at 59. The Court went further and addressed how the inquiry changes where, as here, counsel allegedly failed to advise his client of an affirmative defense. See id. at 59–60. In those circum- stances, the Court explained, “the resolution of the ‘prejudice’ inquiry will depend largely on whether the affirmative de- fense likely would have succeeded at trial.” Id. at 59. The standards announced in Hill map directly onto Gish’s claim and put him under an obligation to make a twofold showing. First, Gish had to show that Opland-Dobs per- formed deficiently in failing to investigate the Xanax-based defense. Second, Gish had to demonstrate that there existed a reasonable probability that, had his counsel investigated the defense, he would have rejected the plea offer and proceeded to trial with a likelihood of succeeding on the defense. See id. at 59. Gish urges a slightly different standard—one informed not only by Hill but even more by the Supreme Court’s deci- sion in Lee v. United States, 137 S. Ct. 1958 (2017). Like Gish, Jae Lee pleaded guilty after his trial counsel advised him that going to trial would be risky, and following a conviction, re- sult in more jail time. See id. at 1963. But Lee had a considera- tion other than prison top of mind. He told his attorney he 14 No. 19-1476

was a noncitizen and “repeatedly asked him whether he would face deportation as a result of the criminal proceed- ings.” Id. Lee’s attorney reassured him that a guilty plea would not result in deportation. Lee relied on and followed the advice even though it was wrong. By pleading guilty to an aggravated felony, Lee faced mandatory deportation un- der the Immigration and Nationality Act—the precise out- come he wanted to avoid. See id. (citing 8 U.S.C. §§ 1101(a)(43)(B), 1227(a)(2)(A)(iii)). Lee later pursued federal habeas relief, arguing that his attorney had rendered ineffec- tive assistance of counsel that resulted in severe prejudice. See id. The Supreme Court agreed. Usually a defendant “without any viable defense will be highly likely to lose at trial,” and when “facing such long odds will rarely be able to show prej- udice from accepting a guilty plea that offers him a better res- olution than would be likely after trial.” Id. at 1966. For Lee, however, “avoiding deportation was the determinative fac- tor”—the variable of “paramount importance”—in deciding whether to plead guilty or go to trial, while the time he spent in prison was relatively inconsequential to his litigation strat- egy. Id. at 1967–69. Lee’s counsel eliminated any doubt on the point, testifying that Lee would have gone to trial had he been properly informed that deportation would follow as auto- matic consequence of pleading guilty. See id. at 1967–68. All of this led the Court to conclude that Lee “would have rejected any plea leading to deportation—even if it shaved off prison time—in favor of throwing a ‘Hail Mary’ at trial.” Id. at 1967. Lee’s laser focus on averting deportation, the Court underscored, showed that his counsel’s errors prejudiced him. Id. at 1967–68. No. 19-1476 15

Gish labors to situate himself like Lee. He does so mindful of Hill, but of the view that Lee modifies the prejudice ques- tion. In Gish’s view, Lee teaches that he could show prejudice by now contending in federal habeas that he would have gone to trial on a Xanax-based defense even if that defense had only one percent chance of success. We disagree and see Lee as reinforcing, not transforming, Hill. In Lee the Court took care to observe that defendants without a viable defense would “rarely” be able to show prej- udice from a guilty plea that reduces their sentencing expo- sure. See id. at 1966. Put most simply, the certainty of less jail time creates an incentive to avoid the longer shot of an acquit- tal at trial. See id. Lee was a rare exception: from Jae Lee’s per- spective, the consequences of pleading guilty and going to trial were “similarly dire”—he would be deported either way—so he was willing to bet on “even the smallest chance of success at trial.” Id. at 1966–67. Properly informed, Lee would have found nothing attractive about a plea offer that reduced his prison time (a relatively minor concern for him) but guar- anteed his deportation—the outcome he most wanted to avoid. Gish’s case is much different. The district court found that, unlike Jae Lee, Christopher Gish decided to plead guilty “based primarily on the prospects of success at trial.” Gish all but said so himself, testifying in the district court that he pleaded guilty because Opland-Dobs informed him that he had no chance of winning at trial. The district court further found that, in contrast with Lee’s persistent concern about de- portation, nothing in Gish’s communications with Opland- Dobs indicated that some factor other than the prospect of success would have motivated Gish to go to trial. 16 No. 19-1476

Unlike Lee, then, Gish wanted to consider an involuntary intoxication defense because he thought it might provide a basis to defeat the homicide charge. What is more, Gish, un- like Lee, said not a word—neither to his trial counsel nor to the district court—suggesting that he was willing to forgo a meaningful reduction in his sentencing exposure (from man- datory life imprisonment to a maximum of 60 years) to avoid collateral consequences. Put another way, the record shows that Gish thought about whether to plead guilty or to go to trial in just the way the Supreme Court in Lee described as paradigmatic for most defendants—by comparing the proba- bility of success at trial with the value of a reduced sentence from pleading guilty. On the record before us, then, we decline Gish’s invitation to deviate from the prejudice inquiry the Supreme Court ar- ticulated in Hill. The proper question therefore is whether there was a reasonable probability that Gish would have gone to trial on his affirmative defense, with the answer “de- pend[ing] largely on whether the affirmative defense likely would have succeeded at trial.” Hill, 474 U.S. at 59. C In the end, we agree with the district court that Gish’s Xanax-based involuntary intoxication defense had no reason- able prospect of success at trial. Even assuming he could mar- shal the evidence required to get a jury instruction on the de- fense, we see no likelihood the defense would have persuaded a jury that Xanax rendered him unable to appreciate the dif- ference between right and wrong at the time he stabbed Lit- wicki to death. Our confidence in this conclusion emerges from the detailed facts the jury would have learned: No. 19-1476 17

• Gish told a hospital nurse that he sold his pills and no longer had any. • Gish told Detective Hart that he last took Xanax “[a] couple days” before the homicide. • The police who searched Gish’s home found no trace of Xanax. • Even if Gish had taken Xanax the day of the homicide, it was unlikely that he was the rare patient who would have experienced effects so extreme as to prevent him from appreciating the wrongfulness of his conduct. The district court found that the little evidence Gish offered on that front (from his expert witness, James O’Donnell) lacked credibility. • In his interview with Detective Hart, Gish con- fessed to how he went about killing and abus- ing Litwicki—statements revealing an aware- ness of his own conduct. • Gish also offered a clear motive for the crime— that he suspected Litwicki was cheating on him and would take his kids away. The combined weight of these facts would have left Gish with no likely prospect of prevailing on an involuntary intox- ication defense and defeating the state’s robust case against him. By extension, then, and especially given Gish’s focus on offering a defense with a chance of succeeding, we have diffi- culty believing that Gish would have proceeded to trial and run the substantial risk of being convicted and receiving a mandatory sentence of life in prison. See Padilla v. Kentucky, 559 U.S. 356, 372 (2010) (emphasizing that a petitioner 18 No. 19-1476

challenging a guilty plea must show “that a decision to reject the plea bargain would have been rational under the circum- stances”); see also Woolley v. Rednour, 702 F.3d 411, 429 (7th Cir. 2012) (rejecting prejudice where the defendant had made the bare and unpersuasive allegation that wrongfully ex- cluded witness testimony could have led to acquittal). Because Gish cannot show prejudice from his trial coun- sel’s errors, we agree with the district court that he is not en- titled to habeas relief on his ineffective assistance claim. We therefore AFFIRM.