Batson v. Kentucky is widely regarded as a failure. In the thirty-plus years since it was decided by the Supreme Court, the doctrine has been subjected to unrelenting criticism for its inability to stop the dis­criminatory use of peremptory challenges. The scholarly literature is nearly unanimous: Batson is broken. But this Article approaches Batson from a different perspective, focusing on Batson’s appellate virtues rather than its trial shortcomings. This change in focus reveals a number of ways in which the Batson doctrine provides opportunities on appeal that do not exist at trial. In short, this Article argues that appellate Batson punches far above its trial weight.

Batson’s appellate virtues have been overlooked by the literature, and this Article’s first task is to illustrate them. This Article’s second project is to reorient the discussion about Batson by placing the doctrine in the proper context. In comparison to other antidiscrimination claims—and to other postconviction claims, more broadly—Batson has a real luster. Though not often acknowledged as such, Batson is the one meaningful doctrine for fighting discrimination in the jury-selection process and in the criminal justice system more generally. Enormous pressure is put on Batson as a result, and maybe Batson is not up to the task. But with Batson’s appellate dimension, the doctrine is more up to the task than previously thought. This Article’s final goal, in light of Batson’s appellate virtues, is to suggest a reconceptualization of Batson as not merely a jury-selection doctrine but rather a multi­purpose vehicle capable of fighting discrimination wherever it occurs in the trial pro­cess—even if the discrimination takes place outside of jury selection. For appellate judges who want to correct the injustice of a trial stained by discrimination, a broad-based Batson doctrine may be their best, last, and only hope.


    1. Batson’s Practical Failings
      1. Batson’s Oddities
      2. Batson’s Antidiscrimination Comparators
      3. Swain v. Alabama
      4. Fair Cross-Section
      5. Ham and Ristaino
    1. Difference in Remedy Between Trial and Appellate–Habeas Litigation
    2. Difference in Evidence Between Trial and Post-Trial Proceedings
      1. Jury-Selection Notes
      2. The Prosecutor’s Behavior in Other Cases and Outside of Court
      3. Policies, Trainings, and Internal Memos
      4. Comparative Juror Analysis
      5. Batson Reconstruction Hearings
    3. Difference in Judicial Approach to Trial and Appeal
    1. Batson as a Multipurpose Vehicle to Combat Discrimination
    2. Appellate Batson’s Symbolism, Rhetoric, and Power
    3. The Gap Between the Trial and Appellate Doctrine Is a Feature, Not a Flaw, of Batson
      1. Evidence: Leveling Up and Leveling Down
      2. Remedy: Leveling Up and Leveling Down
      3. Broader Reflections on the Trial–Post-Trial Gaps



Batson v. Kentucky 1 476 U.S. 79 (1986). is well known, much condemned, but misunder­stood. Academic and judicial commentators emphasize Batson’s short­comings. 2 See infra section I.A. They say it fails to stop peremptory strikes that are moti­vated by race, gender, and other prohibited characteristics. 3 See infra notes 14–20 and accompanying text. Despite its promise, Batson permits anyone who is so inclined to make prejudicial peremptory strikes, so long as the striker takes a few perfunctory steps to conceal her intent, or so the argument goes. 4 See infra notes 26–32 and accompanying text. A chorus of voices has asserted that dis­criminatory jury selection is every bit as problematic today as it was at Batson’s inception back in 1986, 5 See, e.g., Miller-El v. Dretke, 545 U.S. 231, 270 (2005) (Breyer, J., concurring) (“On the other hand, the use of race- and gender-based stereotypes in the jury-selection process seems better organized and more systematized than ever before.”). and that the only option left for pre­venting the discriminatory use of peremptory strikes, now that Batson has struck out, is to ban peremptories altogether. 6 See infra notes 12–13 and accompanying text.

This Article comes at Batson from a different direction. It acknowl­edges Batson’s failings as a trial doctrine—its inability to prevent and remedy strikes in real time—but shifts the focus to Batson’s virtues in appellate and postconviction proceedings. Rather than dismissing Batson as an abject failure, this Article compares it to other equal protection and antidiscrimination claims that litigants use in their post-trial litigation. This shift in focus is part of the process of resuscitating Batson’s reputa­tion. Indeed, too little attention has been given to how Batson operates post-trial, when it is the lone meaningful doctrine for fighting discrimina­tion in the justice system—the only doctrine defendants can plead and actually win. 7 See infra section I.C.

As the lone meaningful antidiscrimination doctrine, Batson has been placed under the pressure of enormous expectations, and the doctrine admittedly may not be up to the task. But Batson may be more up for the challenge than previously thought, especially if the focus shifts from Batson’s trial failings to its post-trial potential. This Article argues for a re-evaluation of Batson in light of the fundamental divergence between trial and appellate Batson. It argues that there is great potential for the latter, not only as a doctrine that fights discriminatory peremptory strikes but as a multipurpose vehicle for attacking all forms of discrimination that man­ifest themselves at trial. Batson is a rare invitation for judges—especially appellate judges—to denounce structural discrimination, and unlike other doctrines, Batson’s automatic-reversal remedy allows judges to at­tach consequences to their words. 8             See infra section II.A.

This Article proceeds in three parts. Part I examines the substantial literature criticizing Batson and discusses why Batson is such a surprising standard-bearer of antidiscrimination law. Part II illustrates several signif­icant ways that post-trial Batson claims provide opportunities for liti­gants that were not available at trial. These include a remedy, post-trial, that elevates the value of a Batson win beyond what it would have been at trial. They also include the increased ability to bring in extra-record evi­dence that was not, and could not have been, available at trial. In addition, Part II identifies idiosyncrasies of Batson’s pleading structure that make it well suited to survive in the difficult habeas landscape.

Part III discusses the implications of the trial–post-trial divide within the Batson doctrine. Batson should bear more weight in the fight against discrimination by incorporating into the Batson claim any evidence of the prosecutor’s racism at trial—even evidence from proceedings outside of jury selection. That is the first implication of recognizing post-trial Batson’s unique virtues. Second, this Part notes how Batson presents fundamen­tally different issues to appellate judges than to trial judges. For appellate judges, Batson is the rare opportunity to declaim on struc­tural issues of racism, democracy, and civics—an opportunity that trial judges do not have or want. Finally, Part III argues that the gap between trial and post-trial Batson cannot be closed without undermining key te­nets of the doctrine. The divergence, as uncomfortable as it makes peo­ple, is built into the doctrine.

Batson appeals are extremely hard to win. There is no denying that. But for all of Batson’s failings, it is still the strongest antidiscrimination doctrine available to litigants, and it provides a unique opportunity for appellate judges to take aim at all manner of discrimination that may have taken place at trial, even outside of the voir dire process.

I. Batson’s Critiques and Batson’s Proper Context

a. Batson’s Practical Failings

In 1986, when the Supreme Court decided Batson v. Kentucky, the an­nounced goal was to end the race-based use of peremptory strikes. 9 476 U.S. 79, 85 (1986) (describing the foundation of “the Court’s unceasing efforts to eradicate racial discrimination in the procedures used to select” jurors). The aims of the doctrine could not have been higher. 10 See Katherine Goldwasser, Limiting a Criminal Defendant’s Use of Peremptory Challenges: On Symmetry and the Jury in a Criminal Trial, 102 Harv. L. Rev. 808, 835 (1989) (explaining the Court’s hope that Batson would “work to bring us closer to our ideal” that “race is irrelevant”). Batson was the Court’s official acknowledgement that discrimination in jury selection was an assault on defendant, juror, and justice alike. 11 Batson, 476 U.S. at 86–87. From the be­ginning, though, Batson was received with skepticism by those most intent on eliminating discrimination. In his Batson concurrence, Justice Thurgood Marshall predicted that the decision “will not end the racial discrimi­nation that peremptories inject into the jury-selection process. That goal can be accomplished only by eliminating peremptory chal­lenges entirely.” 12 Id. at 102–03 (Marshall, J., concurring). And, in the three decades since, Justice Marshall’s assessment has echoed through judicial opinions and academic articles to the point that it is considered a mainstream view of Batson. 13 See, e.g., Miller-El v. Dretke, 545 U.S. 231, 267 (2005) (Breyer, J., concurring) (“Today’s case reinforces Justice Marshall’s concerns.”); Crittenden v. Chappell, 804 F.3d 998, 1020 (9th Cir. 2015) (McKeown, J., dissenting) (“Justice Marshall was prescient in his concurrence in Batson . . . .”); Morgan v. Commonwealth, 189 S.W.3d 99, 115–16 (Ky. 2006) (Graves, J., concurring) (discussing Justice Marshall’s Batson concurrence and expressing “hope [that this decision] may be one step closer to the inevitable implosion of the current peremptory challenge system”), overruled by Shane v. Commonwealth, 243 S.W.3d 335 (Ky. 2006); People v. Brown, 769 N.E.2d 1266, 1272 (N.Y. 2002) (Kaye, C.J., concurring) (“My own years on this extraordinary Court, dealing with countless Batson challenges, have brought me far closer to the perception of Justice Thurgood Marshall . . . .”); Nancy S. Marder, Batson v. Kentucky: Reflections Inspired by a Podcast, 105 Ky. L.J. 621, 626 (2017).

Batson is a great disappointment, if not an outright failure, accord­ing to many thoughtful critiques. The doctrine is “toothless” 14 Alafair S. Burke, Prosecutors and Peremptories, 97 Iowa L. Rev. 1467, 1469 (2012) (internal quotation marks omitted) (quoting Leonard L. Cavise, The Batson Doctrine: The Supreme Court’s Utter Failure to Meet the Challenge of Discrimination in Jury Selection, 1999 Wis. L. Rev. 501, 501). and a “charade”; 15 Minetos v. CUNY, 925 F. Supp. 177, 185 (S.D.N.Y. 1996). “impotent in preventing discrimination”; 16 Charles J. Ogletree, Just Say No!: A Proposal to Eliminate Racially Discriminatory Uses of Peremptory Challenges, 31 Am. Crim. L. Rev. 1099, 1105 (1994); see also Camille A. Nelson, Batson, O.J., and Snyder: Lessons from an Intersecting Trilogy, 93 Iowa L. Rev. 1687, 1689 (2008) (“[T]he original goals articulated twenty years ago in Batson remain unfulfilled. Batson’s promise of protection against racially discriminatory jury selection has not been realized.”). blind to “the in­equities that flow from racial- and gender-based discrimination”; 17 Ulysses Gene Thibodeaux, The Changing Face of Jury Selection: Batson and Its Practical Implications, 56 La. B.J. 408, 409 (2009). “‘almost surely a failure’ and an ‘enforcement nightmare.’” 18 Joshua Revesz, Comment, Ideological Imbalance and the Peremptory Challenge, 125 Yale L.J. 2535, 2535 (2016) (first quoting Kenneth J. Melilli, Batson in Practice: What We Have Learned About Batson and Peremptory Challenges, 71 Notre Dame L. Rev. 447, 503 (1996); and then quoting William T. Pizzi, Batson v. Kentucky: Curing the Disease but Killing the Patient, 1987 Sup. Ct. Rev. 97, 134); see also Note, Judging the Prosecution: Why Abolishing Peremptory Challenges Limits the Dangers of Prosecutorial Discretion, 119 Harv. L. Rev. 2121, 2134 (2006) (“Although the Supreme Court has expanded Batson’s scope significantly since 1986, the doctrine has been largely ineffective.” (footnote omitted)). To Justice Stephen Breyer, “the use of race- and gender-based stereotypes in the jury-selec­tion process seems better organized and more systematized than ever before.” 19 Miller-El v. Dretke, 545 U.S. 231, 270 (2005) (Breyer, J., concurring); see also State v. Saintcalle, 309 P.3d 326, 334 (Wash. 2013) (“Twenty-six years later it is evident that Batson, like Swain before it, is failing us.”). The criticism of Batson is so persistent that it seems everyone who writes about the doctrine must emphasize its failings. 20 See, e.g., Theodore McMillian & Christopher J. Petrini, Batson v. Kentucky: A Promise Unfulfilled, 58 UMKC L. Rev. 361, 369 (1990) (discussing the “extremely difficult time” courts have had in applying Batson). Professor Russell Covey identified a number of articles and books criticizing Batson, including: David Cole, No Equal Justice 120 (1999); Albert W. Alschuler, The Supreme Court and the Jury: Voir Dire, Peremptory Challenges, and the Review of Jury Verdicts, 56 U. Chi. L. Rev. 153, 209 (1989); Lonnie T. Brown, Jr., Racial Discrimination in Jury Selection: Professional Misconduct, Not Legitimate Advocacy, 22 Rev. Litig. 209, 213 n.16 (2003); and Daniel M. Hinkle, Peremptory Challenges Based on Religious Affiliation: Are They Constitutional?, 9 Buff. Crim. L. Rev. 139, 199 (2005). Russell D. Covey, The Unbearable Lightness of Batson: Mixed Motives and Discrimination in Jury Selection, 66 Md. L. Rev. 279, 284 n.17 (2007).

Against this bleak backdrop of Batson scholarship, this Article pre­sents a more positive account—an account informed by observing liti­gators’ relative enthusiasm for other antidiscrimination doctrines. Be­fore presenting this positive account of Batson, however, it is important to lay out some of the most important critiques of Batson for the reader.

One important complaint about Batson is the ease with which Batson’s prohibitions are evaded. 21 See Burke, supra note 14, at 1470 (“[T]he burden of rebutting th[e] [prima facie] inference in stage two is extremely low.”); Christie Stancil Matthews, Missing Faith in Batson: Continued Discrimination Against African Americans Through Religion-Based Peremptory Challenges, 23 Temp. Pol. & C.R. L. Rev. 45, 53 (2013) (“[Batson’s failure] is mainly attributable to . . . the overly-wide latitude lower courts routinely give to proffered ‘race-neutral’ reasons . . . ; the failure by many courts to conduct a proper, searching in­quiry into pretext in the third step; and the inability of the test to adequately account for subconscious racial bias.”). Batson intended to provide litigants with a mechanism to prevent, identify, and disallow discriminatory per­emptory strikes. It established a three-step, burden-shifting framework borrowed from employment discrimination. 22 Johnson v. California, 545 U.S. 162, 168 (2005) (reciting the Batson framework); Robin Charlow, Tolerating Deception and Discrimination After Batson, 50 Stan. L. Rev. 9, 23–24 (1997) (noting that the Batson framework was derived from the Supreme Court’s employment discrimination doctrine). At step one of Batson, the person challenging the peremptory strike must make a prima facie show­ing that the strike was motivated by the juror’s race, gender, or other protected characteristic. 23 Two notes on nomenclature. First, Batson initially applied to race-based discrimi­nation only. See Batson v. Kentucky, 476 U.S. 79, 82, 96–98 (1986). The doctrine has since expanded to encompass other types of illegal discrimination. See infra notes 44–48 and accompanying text. For the sake of readability, this Article sometimes refers only to “race-based” discrimination, but that reference encompasses all forms of illicit dis­crimination. When further distinctions need to be made, this Article uses more specific language. Second, throughout, this Article speaks of the defendant as the one bringing the Batson challenge and the prosecutor as the one who made the peremptory strike. This is short­hand. Prosecution and defense alike can raise Batson challenges. Georgia v. McCollum, 505 U.S. 42, 59 (1992). So can civil litigants. Edmonson v. Leesville Concrete Co., 500 U.S. 614, 631 (1991). If there is an inference of discrimination, the trial judge moves to step two, where she asks the prosecutor to explain what motivated the strike. 24 Johnson, 545 U.S. at 168. Once that answer is provided, the inquiry moves to the third step, where either party may present additional evi­dence and where the judge must decide, in light of all the evidence, whether the strike was motivated by discriminatory intent. 25 Id.

As critics have noted, the trouble with this framework is at step two: The prosecutor can make up any justification she wants for the strike, and those justifications can be impossible to disprove. 26 Jeffrey Bellin & Junichi P. Semitsu, Widening Batson’s Net to Ensnare More than the Unapologetically Bigoted or Painfully Unimaginative Attorney, 96 Cornell L. Rev. 1075, 1075 (2011) (“[W]e conclude that Batson is easily avoided through the articulation of a purportedly race-neutral explanation for juror strikes.”); Nelson, supra note 16, at 1692 (describing “the ease with which a prosecutor may overcome a defendant’s showing of a prima facie case of purposeful discrimination in the jury-selection process”); id. at 1701 (“As seen in Snyder v. Louisiana, the neutral-explanation test in Batson is too deferential to prosecutors and allows for the use of pretextual, dubious, and inconsistent prosecutorial responses.”); Ogletree, supra note 16, at 1107 (“[I]n many jurisdictions . . . Batson has been more or less undermined by prosecutors who fabricate facially neutral reasons for striking minority jurors, and trial courts that have difficulty evaluating such reasons. . . . ‘[So] Batson lacks the . . . “teeth” required to ensure that black jurors are not excluded on the basis of race.’” (quoting Brian Wilson, Recent Case, Batson v. Kentucky: Can the ‘New’ Peremptory Challenge Survive the Resurrection of Strauder v. West Virginia?, 20 Akron L. Rev. 355, 364 (1986))); id. at 1123 (“To effectively prevent the racial use of the peremptory challenge, some change in law or procedure will have to ad­dress the problem at its source: the willingness of trial judges to accept pretexts for racially discriminatory peremptory strikes.”). This is especially true because there is no requirement that the prosecutor’s explanation be logical or plausible, so long as the prosecutor can convince the judge that it is sincerely held. 27 See Harris v. Haeberlin, 752 F.3d 1054, 1059 (6th Cir. 2014) (“The justification need not be persuasive; in fact, if true, it may even be ‘only a frivolous or utterly non­sensical justification.’” (quoting Johnson, 545 U.S. at 171)). But see Purkett v. Elem, 514 U.S. 765, 768 (1995) (per curiam) (“[I]mplausible or fantastic justifications may (and probably will) be found to be pretexts for purposeful discrimination.”). Bizarre, trivial justifications may count as “race-neutral.” Examples include using a strike because a juror “[w]ore a beret one day and a sequined cap the next,” 28           Bellin & Semitsu, supra note 26, at 1094 (internal quotation marks omitted) (quoting Smulls v. Roper, 535 F.3d 853, 856 (8th Cir. 2008) (en banc)). or “[m]entioned the word ‘gov­ernment’ twice in his answers,” 29 Id. at 1095 (internal quotation marks omitted) (quoting United States v. Ervin, 266 F. App’x 428, 436 (6th Cir. 2008)). or “[l]acked outside hobbies and inter­ests,” 30 Id. (citing Lewis v. Bennett, 435 F. Supp. 2d 184, 191 (W.D.N.Y. 2006)). or “[l]acked ‘hope in the legal system.’” 31 Id. at 1091 (citing People v. Hamilton, 200 P.3d 898, 934 (Cal. 2009)); see also Purkett, 514 U.S. at 769 (finding “[t]he prosecutor’s proffered explanation . . . that he struck juror number 22 because he had long, unkempt hair, a mustache, and a beard” to be legitimate at step two); People v. Randall, 671 N.E.2d 60, 65–66 (Ill. App. Ct. 1996) (“Surely, new prosecutors are given a manual, probably entitled, ‘Handy Race-Neutral Explanations’ or ‘20 Time-Tested Race-Neutral Explanations.’”); Bellin & Semitsu, supra note 26, at 1091–96 (collecting justifications from cases and studies); J. Thomas Sullivan, Lethal Discrimination, 26 Harv. J. Racial & Ethnic Just. 69, 95 (2010) (“If trial judges cyni­cally accept unreasonable, but facially race-neutral, explanations for peremptory striking of minority venirepersons, the application of deferential standards of appellate review effectively insulates racially-discriminatory practices in jury selection from meaningful ap­pellate review.”). Anyone with even a modicum of savvy can choose a justification that is not observable on the record—such as the claim that the juror was not making good eye con­tact—thereby making it impossible for trial judges, and later appellate judges, to disprove the justification. 32 See Mimi Samuel, Focus on Batson: Let the Cameras Roll, 74 Brook. L. Rev. 95, 97–98 (2008) (explaining prosecutors’ reliance on “intangibles such as eye contact, tone of voice, demeanor, posture, and laughing or coughing” put courts “at a tremendous disadvantage” in “discern[ing] whether the given reasons are in fact discriminatory because the courts have little or no evidence” for “assessing the validity of the reason”). The prosecutor has so much freedom that she practically cannot get caught unless she picks a demon­strably false or explicitly race-based justification. The ease of inventing pretexts to satisfy step two can make the entire Batson frame­work feel like a farce. A very significant and legitimate criticism, to be sure.

A second, related criticism concerns Batson’s requirement that de­fendants prove intentional discrimination. 33 See Hernandez v. New York, 500 U.S. 352, 372 (1991) (O’Connor, J., concurring) (noting that parties must prove that “the prosecutor intentionally discriminated” on the basis of race in order to prove a Batson violation); United States v. Green, 599 F.3d 360, 377 (4th Cir. 2010) (“When a party challenges his opponent’s exercise of a peremptory challenge on equal protection grounds, the party bears the burden of proving intentional discrimination.”); see also Anna Roberts, Disparately Seeking Jurors: Disparate Impact and the (Mis)use of Batson, 45 U.C. Davis L. Rev. 1359, 1367 (2012) (observing that Batson requires defendants to establish a prima facie case of purposeful discrimination). As with all equal protection claims, Batson requires the court to determine the prosecutor’s subjective intent; it is not enough to show that the prosecutor’s actions have a dis­parate impact. 34 See Roberts, supra note 33, at 1367. The intent requirement has rankled commentators because it is so difficult to prove. And the problem of proof is particularly pronounced, given that the prosecutor is given an opportunity, at step two, to muddy the waters by providing false justifications. Further, the intent requirement has been criticized for creating a certain social awk­wardness for judges, insofar as it requires them to say the prosecutor was racist. 35 See, e.g., Coombs v. Diguglielmo, 616 F.3d 255, 264 (3d Cir. 2010) (“No judge wants to be in the position of suggesting that a fellow professional—whom the judge may have known for years—is exercising peremptory challenges based on forbidden racial considerations.”); State v. Saintcalle, 309 P.3d 326, 338 (Wash. 2013) (“A requirement of conscious discrimination is especially disconcerting because it seemingly requires judges to accuse attorneys of deceit and racism in order to sustain a Batson challenge.”); see also Adam M. Gershowitz, Prosecutorial Shaming: Naming Attorneys to Reduce Prosecutorial Misconduct, 42 U.C. Davis L. Rev. 1059, 1085 (2009) (“[R]epeated contact may lead to a close relationship and bond between the judge and the prosecutor. It therefore makes sense that the trial judges they appear in front of day after day would be reluctant to take prosecutors to task publicly.”). According to this critique, the awkwardness of branding the prosecutor racist raises the stakes for all involved and is another impedi­ment to finding a Batson violation. 36 See supra note 35. There is arguably a difference between a single race-based act in a lifetime—enough to trigger a Batson violation—and the character of being a racist. But the point stands that the intent requirement nudges the Batson inquiry from looking at an individual act toward looking at the prosecutor’s character. As will be discussed later on, there is a sharp distinction between Batson’s focus on intent and the Sixth Amendment fair cross-section doctrine’s lack of regard for intent when addressing racial underrep­resentation in the jury pool. Later on, this Article also describes how Batson’s focus on intent may benefit defendants, by allowing them to bring in a broader range of evidence about the prosecutor’s misdeeds. See infra section I.C.2.

Still another major critique of Batson is the deference appellate courts owe to trial court determinations. 37 See Nancy S. Marder, Batson Revisited, 97 Iowa L. Rev. 1585, 1592 (2012) [hereinafter Marder, Batson Revisited] (“Another reason that Batson and its progeny have been so ineffective is that . . . appellate judges are deferential to trial judges’ determi­nations.”); Nancy S. Marder, Justice Stevens, the Peremptory Challenge, and the Jury, 74 Fordham L. Rev. 1683, 1708 (2006) [hereinafter Marder, Justice Stevens, the Peremptory Challenge, and the Jury] (“Although appellate review could correct this problem, at least by ensuring consistency within a circuit, appellate review of Batson challenges tends to be quite deferential to the trial judge.”); Thibodeaux, supra note 17, at 410 (“A trial court ruling on discriminatory intent, however, must be sustained unless it is clearly erroneous. Given the propensity for affirmance under this standard or an abuse of discretion stan­dard, a trial court’s ruling is virtually immune to reversal.”); see also Daniel R. Pollitt & Brittany P. Warren, Thirty Years of Disappointment: North Carolina’s Remarkable Appellate Batson Record, 94 N.C. L. Rev. 1957, 1959 (2016) (“North Carolina’s highest court has never once in those thirty years found a substantive Batson violation.”). According to this critique, because Batson claims are fact-intensive, the trial court is entitled to so much deference that there is nothing left for the appellate court to do. 38 See Marder, Justice Stevens, the Peremptory Challenge, and the Jury, supra note 37, at 1708–09. A trial judge could thus doom a Batson claim by making an unfavorable determination that cannot effectively be reviewed. This Article pushes back on the conventional wisdom by pointing out how appellate Batson may offer more opportunities to litigants than the trial doctrine did. Al­though appellate judges are required to defer to many aspects of the trial court’s Batson rulings, they have found ways to breathe life into Batson claims, even under the extreme deference required by the Antiterrorism and Effective Death Penalty Act (AEDPA). 39 Antiterrorism and Effective Death Penalty Act of 1996, 28 U.S.C. § 2254(d) (2012); see also Davis v. Ayala, 135 S. Ct. 2187, 2199 (2015) (describing the highly defer­ential standard of review under AEDPA). One of the main projects of this Article is to show how unexpectedly emboldened appellate courts are when it comes to finding Batson violations, in no small part due to the civic, political, and moral stakes of the Batson doctrine.

Another vein of criticism concerns Batson’s limited scope. Critics complain that Batson does not protect enough categories of people. 40 See, e.g., Jonathan Grossman, Sixth Dist. Appellate Program,
Wheeler/Batson Developments 4 (2006),
wheeler_materials.web.pdf [] (collecting cases showing cat­egories not covered by Batson, including “[p]oor people,” “[y]oung adults,” “[p]eople who do not speak English,” “[p]eople who have been arrested, been victims, or believe in law and order,” and “[p]eople with long hair and beard”).
Supreme Court decisions have recognized that the Batson doctrine now covers race-, 41 See Batson v. Kentucky, 476 U.S. 79, 98 (discussing race-based peremptories). gender-, 42 See J.E.B. v. Alabama ex rel. T.B., 511 U.S. 127, 129–30 (1994) (extending Batson to cover gender-based peremptories). and ethnicity-based peremptories. 43 Hernandez v. New York, 500 U.S. 352, 355 (1991) (extending Batson to cover ethnic origin). State courts and lower federal courts have extended Batson to include other groups. 44 Grossman, supra note 40, at 3–4 (listing cases from state and federal courts that have expanded the Batson doctrine to cover additional groups such as “Hispanics,” “[h]omosexuals,” and religious groups). Well-reasoned criticisms call for extending Batson to strikes based on sex­ual orientation, 45 See Tania Tetlow, Solving Batson, 56 Wm. & Mary L. Rev. 1859, 1868 n.32 (2015) (“Circuits are split about the application of the Batson rule to sexual orientation, but if that is recognized as a suspect category, it should be included in this analysis as well.”); John J. Neal, Note, Striking Batson Gold at the End of the Rainbow?: Revisiting Batson v. Kentucky and Its Progeny in Light of Romer v. Evans and Lawrence v. Texas, 91 Iowa L. Rev. 1091, 1113 (2006) (advocating for sexual-orientation protection); see also Cal. Civ. Proc. Code § 231.5 (West 2016) (protecting against discrimination under a number of cate­gories, including race, color, religion, ancestry, disability, genetic information, and sexual orientation); SmithKline Beecham Corp. v. Abbott Labs., 740 F.3d 471, 474 (9th Cir. 2014) (holding that equal protection prohibits strikes based on sexual orientation). disability, 46 See Mary A. Lynch, The Application of Equal Protection to Prospective Jurors with Disabilities: Will Batson Cover Disability-Based Strikes?, 57 Alb. L. Rev. 289, 293 (1993). religion, 47 See Matthews, supra note 21, at 63–64 (“The Supreme Court should specifically address the constitutionality of religion-based strikes, and such strikes should be pro­hibited, given that religion, like race and gender, has traditionally been deemed a suspect classification.”); Courtney A. Waggoner, Note, Peremptory Challenges and Religion: The Unanswered Prayer for a Supreme Court Opinion, 36 Loy. U. Chi. L.J. 285, 328 (2004) (“The Supreme Court should . . . rule against the government’s ability to exercise per­emptory challenges based upon a potential juror’s religious affiliation.”). and other characteristics. 48 Lynch, supra note 46, at 318 (“Relying on Batson and its progeny—which has tended to extend rather than limit Batson’s scope—many lower courts . . . have decided to extend Batson to discrimination claims of gender-, ethnic-, and age-based peremptory strikes. Other courts have limited Batson to race-based strikes.”); cf. Revesz, supra note 18, at 2537 (suggesting “peremptory-challenge procedures” should be eliminated for their ideological effect on juries because they “produce juries that are considerably more con­servative than a random sampling of Americans”).

There is some tension between the criticism that Batson is ineffectual and the criticism that it covers too few characteristics. If the doctrine were toothless, it should not matter how much or how little it purports to protect. One way to resolve this tension, however, is to think of Batson’s symbolic importance, independent of its effectiveness. The doctrine lim­its what characteristics may lawfully be used to exclude citizens from par­ticipation in the basic civic institution of the jury. Batson’s protection of a particular group is an affirmation that the group members belong as full participants in society. 49 See Duncan v. Louisiana, 391 U.S. 145, 156 (1968) (noting the Framers’ “in­sistence upon community participation in the determination of guilt or innocence”); Strauder v. West Virginia, 100 U.S. 303, 308 (1879) (describing harm that accrues to African American citizens from being excluded from jury participation). As argued below, this democratic symbolism is no small part of Batson’s success as a doctrine, especially on appeal and in postconviction proceedings.

B. Batson’s Oddities

The above account catalogues Batson’s functional failings. Another level of criticism takes aim at the doctrine’s theoretical coherence. For better or worse, Batson is something of a doctrinal oddball. This section reflects on several of Batson’s most prominent oddities. These quirks do not, individually or collectively, preclude Batson from playing a promi­nent role in antidiscrimination law, but they do make Batson something of a surprising choice for this role, especially compared to doctrines that are more theoretically straightforward in their attacks on systemic discrimination.

First is the question of what right (or rights) Batson is supposed to protect. One might be surprised not to find a straightforward answer to this basic question. The Batson Court justified its decision in terms of pro­tecting the constitutional rights of defendants and jurors. 50 Batson v. Kentucky, 476 U.S. 79, 85–87 (1986) (referring to the rights of the defendant and the juror). Subsequent Supreme Court cases have emphasized the jurors’ rights over the de­fendants’ (and included all manner of litigants). 51 Pamela S. Karlan, Race, Rights, and Remedies in Criminal Adjudication, 96 Mich. L. Rev. 2001, 2015 (1998) [hereinafter Karlan, Race, Rights, and Remedies] (“Over the past decade, the Court has moved toward the view that the victim in Batson cases is the ex­cluded juror.”). As scholars have noted, this change in emphasis has implicated another doctrinal oddity: The jurors’ rights are vindicated not by the jurors themselves but by third parties, namely, the litigants. 52 Powers v. Ohio, 499 U.S. 400, 409 (1991) (“An individual juror does not have a right to sit on any particular petit jury, but he or she does possess the right not to be excluded from one on account of race.”). Such third-party standing is rather unusu­al in constitutional litigation, yet it is the foundation on which Batson rests. 53 Edmonson v. Leesville Concrete Co., 500 U.S. 614, 629 (1991) (discussing a third party’s ability to vindicate a juror’s rights). Among its unusual implications, the reliance on third-party standing means that jurors can effectively be barred from service, with no remedy, if both the defendant and the prosecutor choose not to challenge the exclusion. 54 In Swain litigation, see infra section I.C, it used to be a problem that both sides would agree to wholesale exclusion of jurors of a particular race. However, a judge could raise a Batson objection sua sponte. See, e.g., People v. Rivera, 852 N.E.2d 771, 784 (Ill.), modified on denial of reh’g, (Ill. 2006); People v. Bell, 702 N.W.2d 128, 134 (Mich.), amended on denial of reh’g by 704 N.W.2d 69 (Mich. 2005).

The remedy for a Batson violation also raises theoretical concerns. At trial, the remedy is to reseat the juror who was struck or, more frequent­ly, to draw an entirely new venire. 55 See, e.g., Batson, 476 U.S. at 100 n.24 (expressing “no view on whether it is more appropriate . . . for the trial court to discharge the venire and select a new jury from a panel . . . or to disallow the discriminatory challenges and resume selection with the improperly challenged jurors reinstated on the venire” (citation omitted)); United States v. Robinson, 421 F. Supp. 467, 474 (D. Conn. 1976), mandamus granted sub nom. United States v. Newman, 549 F.2d 240 (2d Cir. 1977). Post-trial, the remedy is an automatic reversal of the conviction. 56 See Tankleff v. Senkowski, 135 F.3d 235, 248 (2d Cir. 1998) (“Because the effects of racial discrimination during voir dire ‘may persist through the whole course of the trial proceedings,’ we hold that a Batson/Powers claim is a structural error that is not subject to harmless error review.” (quoting Powers, 499 U.S. at 412)); Rosa v. Peters, 36 F.3d 625, 634 n.17 (7th Cir. 1994); Ramseur v. Beyer, 983 F.2d 1215, 1225 n.6 (3d Cir. 1992) (en banc); Brian R. Means, Postconviction Remedies § 30:2, Westlaw (database updated June 2017) [hereinafter Means, Postconviction Remedies] (stating that the use of peremptory chal­lenges on the basis of race or gender gives rise to structural error). But cf. Karlan, Race, Rights, and Remedies, supra note 51, at 2019 n.87 (providing examples of when courts have found no structural error as long as some African Americans remain in the jury pool). A fascinating academic literature explores the potential mismatch between right and remedy. 57 Professor Karlan’s writing on the relationship between the Batson right and the Batson remedy is extremely useful in drawing out the swirl of interests at work. See Pamela S. Karlan, Batson v. Kentucky: The Constitutional Challenges of Peremptory Challenges, in Criminal Procedure Stories 381, 396–97 (Carol S. Steiker ed., 2006) [hereinafter Karlan, Batson] (discussing Batson amicus briefs that expressed the view that the constitutional Batson remedy would occur at the trial level and would involve relatively small transaction costs); see also Karlan, Race, Rights, and Remedies, supra note 51, at 2004 (exploring “the complications that arise in the definition of rights and in the operation of remedies when the Equal Protection Clause is used in criminal adjudication”). If the juror is the one whose rights are violated, how does drawing a new venire vindicate the juror’s rights, as the struck juror will not be in the new venire? 58 Of course, if the improperly struck juror has not left the courtroom, courts can also reseat the juror, which would at least remedy the wrong suffered by that juror. But that is less typical. The reason that juror is not often reseated is in large part due to the logistics of jury selection. The juror might have already been instructed to leave the courtroom by the time the defense makes the Batson challenge, as there is no requirement that the challenge be made immediately after the strike. See United States v. Rodriguez, 917 F.2d 1286, 1288 n.4 (11th Cir. 1990) (noting that a Batson challenge is timely when made before jurors take the oath and the trial begins), abrogated by Powers v. Ohio, 499 U.S. 400 (1991); United States v. Thompson, 827 F.2d. 1254, 1257 (9th Cir. 1987) (allow­ing a Batson challenge “just after” the jury was sworn in). And if the juror is removed and then reseated, that juror may develop suspicions about who struck her and why. If Batson protects the defendant’s rights, on what theoretical basis does it do so? Is it because a black juror is presumed to vote differently from a white one, so the race-based strike is presumed to have affected the ver­dict by changing the racial composition of the jury? And does that com­mit the courts to the uncomfortable position that jurors act differently based on their race? Does it mean that only defendants whose cases would have come out differently with a different jury are entitled to relief? Or is the defendant given a new venire or new trial because of a principled harm: She was tried by a jury tainted by discrimination? But why should this symbolic harm result in the powerful medicine of throw­ing out a conviction? Later on, this Article returns in more depth to the implications of the almost-unanimous view that Batson violations require automatic reversal, free of any harmless error analysis. 59 See infra section II.A; see also Karlan, Race, Rights, and Remedies, supra note 51, at 2016 (“The movement toward a juror-centered view of the Batson right was tempting in part because it enabled the Court to finesse the question whether the race or sex of a jury’s members affects trial outcomes.”).

Another oddity—not at all to say flaw—of Batson’s antidiscrimination bona fides is its availability to prosecutors 60 Georgia v. McCollum, 505 U.S. 42, 59 (1992) (holding that “the Constitution prohibits a criminal defendant from engaging in purposeful discrimination on the ground of race in the exercise of peremptory strikes”). and civil litigants, 61 Edmonson v. Leesville Concrete Co., 500 U.S. 614, 616 (1991) (holding that even in civil cases, “race-based exclusion violates the equal protection rights of the challenged jurors”). rather than just criminal defendants. There is nothing inconsistent with allow­ing prosecutors and civil litigants to counteract discrimination in the justice system, but there is some level of irony that a doctrine created to defend African American defendants against deep-seated, institutional racism now allows prosecutors to prevent the removal of white jurors or corporations to prevent the removal of male jurors. The unusual nature of Batson as a doctrine that protects jurors and all litigants—not just crim­inal defendants—adds to Batson’s effectiveness.

The final oddity to mention involves the theatricality of Batson’s pleading framework. This theatricality, this Article argues, is not inci­dental to the success of the doctrine. 62 See infra section III.C. Unlike any other forum, Batson puts government officials on the spot to account, in a public manner, for their discriminatory actions. This morality play takes the structural racism that pervades the justice system and gives it a human face: the prosecu­tor’s. This personification makes the racism easier to envision. It also makes the taint of racism seem more limited and manageable than when it is conceptualized in institutional terms. 63 One of the reasons that Batson is a more appealing claim than the fair cross-section or discriminatory-prosecution claims, see infra section I.C, is that it is individ­ualized to a single unconstitutional act—a single strike. An appellate court that finds a fair cross-section violation in the drawing of a county’s jury pool will potentially call into doubt hundreds of other convictions—or it would if there were not such steep hurdles to pleading fair cross-section claims that were not litigated at trial. See infra section I.C. Likewise, a court finding that a prosecutor has engaged in selective pros­ecution or selective administration of the death penalty would have to confront the systemic impli­cations of such a finding. Even if other cases cannot be reopened for procedural reasons, the fact that a prosecutor was making charging decisions based on race undermines the fairness of all the convictions to have come from that prosecutor’s office. Finding a Batson violation in a single case does not bring along with it such systemic implications, even if evidence of a prosecutor’s violations in one case may be used in other cases to show the prosecutor’s bias. This theater is an important, and unusual, aspect of constitutional litigation. And, as this Article ar­gues later on, Batson’s pageantry has the key benefit of allowing judges—especially appellate judges—to declaim on broad principles of justice. 64 See infra section III.C. Batson transforms the peremptories into a morality play about which cit­izens may be excluded from civic participation and on what grounds, and it affords appellate judges the rare prompt to talk about structural dis­crimination in the justice system.

C. Batson’s Antidiscrimination Comparators

The critiques of Batson’s failings and its quirks are in many respects quite reasonable. But they fail to consider Batson in comparison to other antidiscrimination claims. Although Batson is flawed in many respects, it is arguably the only meaningful vehicle for challenging racial discrimina­tion in jury selection and in the justice system more generally. 65 See infra notes 104–107 and accompanying text (noting the number of times that people have won Batson claims). This sec­tion discusses Batson’s less auspicious comparators in antidiscrimination claims, in an effort to show Batson’s comparative strengths.

1. Swain v. Alabama. — Swain v. Alabama 66 380 U.S. 202 (1965). was the immediate pre­decessor to Batson. Swain prohibited race-based peremptory strikes but only in the most extreme circumstances. Under Swain, the defendant could prevail on equal protection grounds only by showing that the pros­ecutor had a virtually unbroken pattern of striking African American jurors “in case after case, whatever the circumstances, whatever the crime and whoever the defendant or the victim may be” such that “no Negroes ever serve on petit juries.” 67 Id. at 223. If the prosecutor struck black jurors most, but not all, of the time, or if the prosecutor struck black jurors only in cases in which the defendant was also black, that pattern would not be enough to satisfy Swain. Likewise, even if the prosecutor’s strikes were persistent enough to satisfy Swain, there was still the problem of proof. The defendant was required to obtain and analyze extensive data on strikes in other cases to establish the prosecutor’s unbroken pattern. 68 Bruce J. Winick, Prosecutorial Peremptory Challenge Practices in Capital Cases: An Empirical Study and a Constitutional Analysis, 81 Mich. L. Rev. 1, 11 (1982) (“Not only does such a showing require transcribing voir dires of a large number of cases, itself an expensive undertaking, but additional investigation would be required to ascertain the race of each venireperson . . . . [This] is beyond the ability and resources of virtually all defendants.”). The defendant could not use any evidence from his own case about how the prosecutor had used his strikes. 69 See Swain, 380 U.S. at 223. These extensive data were extreme­ly difficult to gather, even when they existed, and in many instances there was simply no record of whom a prosecutor had struck in other cases, much less that person’s race. 70 See People v. Wheeler, 583 P.2d 748, 767–68 (Cal. 1978) (“[Under Swain,] the defendant would be required to somehow obtain and analyze the records of an undetermined number of individual trials . . . . But he would have no practical way of discovering which of the excused jurors were black, or of proving their race even if he could learn of it . . . .”).

The Court in Batson specifically addressed these shortcomings of Swain and attempted to make such equal protection challenges easier to plead and win. 71 See Batson v. Kentucky, 476 U.S. 79, 92–93 (1986) (finding that the prevailing interpretation of Swain places a “crippling burden of proof” on defendants and rejecting this formulation as inconsistent with equal protection jurisprudence). As will be shown later on in this Article, Swain was not wholly discarded. To this day, litigants wishing to raise Swain claims can do so instead of, or in addition to, Batson claims. They just have to refer to the prosecutor’s behavior in other cases. Indeed, the more Batson claims rely on extra-record evidence to show discriminatory intent, the more Batson begins to resemble Swain. 72 See infra section II.B (discussing extra-record Batson evidence).

2. Fair Cross-Section. — Another comparator to Batson is the fair cross-section requirement of the Sixth Amendment, as articulated by the Supreme Court in Duren v. Missouri. 73 See 439 U.S. 357, 363–64 (1979); see also Berghuis v. Smith, 559 U.S. 314, 319 (2010) (describing the test set out in Duren). Like Swain, but unlike Batson, fair cross-section claims are systemic challenges. The Sixth Amendment guar­antees criminal defendants the right to be tried by an impartial jury of their peers, 74 U.S. Const. amend. VI. and the fair cross-section doctrine requires that the veni­re—the pool of jurors summoned to appear for service—is sufficiently representative of the community. If the jury-summons system results in significant and repeated underrepresentation of a cognizable group, the fair cross-section requirement may be violated. 75 See Taylor v. Louisiana, 419 U.S. 522, 530 (1975). Even if there is a persistent underrepresentation caused by the summons system, the state is given an opportunity to justify it on some grounds, such as a “significant state interest.” See Duren, 439 U.S. at 367–69.

In theory, this doctrine has great potential. While it may be impossi­ble to root out discrimination in peremptory strikes, the effect of this discrimination could be blunted by having jury pools that have represen­tatively large numbers of minority jurors—the goal of the fair cross-section doctrine. Or, put another way, if minority jurors are under­represented in the jury pool, it is easier for the prosecutor to use her peremptory strikes to make the jury entirely white than if there are more minority jurors in the venire. An additional virtue of the fair cross-section doctrine, at least in principle, is that there is no need to prove discriminatory intent; all that is required is a significant and systemic underrepresentation of a cognizable group. 76 Duren, 439 U.S. at 368 n.26 (noting that equal protection challenges require dis­criminatory purpose but that, “[i]n contrast, in Sixth Amendment fair-cross-section cases, systematic disproportion itself demonstrates an infringement of the defendant’s interest in a jury chosen from a fair community cross section”). That is true because the Sixth Amendment right, unlike its Fifth Amendment analogue, does not require a showing of pur­poseful discrimination.

Despite these theoretical advantages, however, subsequent Supreme Court decisions have severely degraded the fair cross-section doctrine described in Duren. For example, the Court has hampered fair cross-section challenges by its insistence that defendants identify, with particu­larity, what mechanism of the jury-summons process is responsible for the underrepresentation in the venire. 77 Berghuis, 559 U.S. at 332 (noting, albeit in the deferential posture of federal habeas review, that “[n]o ‘clearly established’ precedent of this Court supports Smith’s claim that he can make out a prima facie case merely by pointing to a host of factors that, individually or in combination, might contribute to a group’s underrepresentation”). This is a difficult task for de­fendants to undertake because while the underrepresentation can be relatively easily observed, it may be that no one knows why this un­derrepresentation is occurring (or it may be that a confluence of factors leads to the underrepresentation). Identifying the particular cause of the underrepresentation is made even more difficult by the policies of many clerks of court to abstain from tracking demographic information about their jury lists—a decision that appears to have been made with the intent of foreclosing fair cross-section challenges. 78 See Rosa Holdeman, Inst. for Court Mgmt., Hispanic Representation in Jury Panels of the Superior Court of California, County of Orange Is Unknown 35 (2009),
2009/Holdeman_HispanicRepJuryPanels.ashx [] (“The cur­rent policy of not collecting race and ethnicity data is primarily based on the fear of po­tential jury challenges.”).

And still another flaw of the fair cross-section doctrine, as if another flaw were required, is the application of the doctrine to groups that make up a small percentage of the population. Many courts measure the “abso­lute disparity,” which is the number of percentage points between a group’s representation in the jury-eligible population and that group’s representation in the venire. 79 See, e.g., United States v. Royal, 174 F.3d 1, 6–8 (1st Cir. 1999) (noting that six circuits have endorsed the absolute-disparity or closely related “absolute impact” method of calculating misrepresentation). (If African Americans make up seventeen percent of the jury-eligible population and just twelve percent of the venire, the “absolute disparity” would be five percentage points.) Although there is no hard-and-fast cutoff, many published decisions parrot the no­tion that the defendant must show an absolute disparity of seven-plus percentage points—sometimes the threshold is quoted as ten percentage points—in order to make out a fair cross-section violation. 80 United States v. Hernandez-Estrada, 749 F.3d 1154, 1161, 1164 (9th Cir. 2014) (en banc) (noting and ultimately rejecting circuit law creating a 7.7% threshold); United States v. Grisham, 63 F.3d 1074, 1078–79 (11th Cir. 1995) (noting a ten percent threshold for absolute disparity). This is a nearly insurmountable bar to fair cross-section protection for groups that make up less than ten percent of the population. It means that even if such group members are completely excluded from the venire, that ex­clusion will not cross the ten percent threshold needed to trigger a fair cross-section claim. 81 E.g., Hernandez-Estrada, 749 F.3d at 1161 (“Indeed, we have specifically high­lighted the fact that if a minority group makes up less than 7.7% of the population in the jurisdiction in question, that group could never be underrepresented in the jury pool, even if none of its members wound up on the qualified jury wheel.”).

In sum, these and other flaws with the fair cross-section doctrine have left commentators to bemoan the doctrine’s demise. 82 See Mary R. Rose & Jeffrey B. Abramson, Data, Race, and the Courts: Some Lessons on Empiricism from Jury Representation Cases, 2011 Mich. St. L. Rev. 911, 959 (noting the “enormous investigative burden on the defense” imposed by the requirement to identify the cause of underrepresentation); David M. Coriell, Note, An (Un)fair Cross Section: How the Application of Duren Undermines the Jury, 100 Cornell L. Rev. 463, 465 (2015) (“Placing the burden on the defendant to prove how a specific jury-selection pro­cedure is responsible for nonrepresentative jury venires is a high bar that often renders the fair cross section guarantee illusory.”). A basic search of fair cross-section decisions over the last decade—both state and federal—reveals just a single fair cross-section victory. 83 The Westlaw search queried <“fair cros!” /s violat!> in the “Holding” field. For the period running January 1, 2008 through November 4, 2017, one fair cross-section vic­tory resulted: Garcia-Dorantes v. Warren, 801 F.3d 584, 587 (6th Cir. 2015), cert. denied, 136 S. Ct. 1823 (2016). Though the search was not exhaustive, and there may be a number of victories that never made it into Westlaw’s database, the result is suggestive of the doc­trine’s anemia.

3. Ham and Ristaino. — Still another way to combat discrimination in jury selection is to allow robust questioning about jurors’ racial biases. If racial bias emerges from the questioning in voir dire—and it is an open question whether jurors would admit to as much in court—that could be the basis of for-cause and peremptory strikes. Vigorous ques­tioning seems like an innocuous way to confront racial discrimination, and it has the benefit of not requiring the court to decide who can and cannot serve on the jury. All the court has to do is allow the parties to ask questions to get more detailed information about the jurors. Despite the upside of such questioning, the Supreme Court has pulled back on the notion that the Constitution protects a defendant’s right to ask these questions about racial biases.

In Ham v. South Carolina, a black defendant alleged that his prosecu­tion for possession of marijuana was retaliation for his work as a civil rights activist. 84 409 U.S. 524, 524–25 (1973). The trial judge denied the defendant’s request to have the jurors questioned about racial prejudice. 85 Id. at 525–26. The Supreme Court reversed Ham’s conviction on due process grounds because of the judge’s “refusal to make any inquiry as to racial bias of the prospective jurors.” 86 Id. at 529. Three years later, however, the Supreme Court significantly limited this holding in Ristaino v. Ross. 87 424 U.S. 589 (1976). In Ristaino, a black defendant accused of attacking a white security guard was not allowed to question jurors about racial prejudices. 88 Id. at 590. The Supreme Court held it was not a constitutional vio­lation for the judge to prevent such questioning. 89 Id. at 597. The earlier deci­sion in Ham was limited to circumstances in which “[r]acial issues . . . were inex­tricably bound up with the conduct of the trial”—and that was not the case in Ristaino. 90 Id.

Later cases have tinkered around the edges with defining when questioning is constitutionally required. 91 See Turner v. Murray, 476 U.S. 28, 33 (1986) (declaring the capital nature of the trial to be a circumstance under which questioning about bias must be allowed). But the message is clear: The Constitution will not assist a defendant in questioning her way to a racially impartial jury, except in certain unusual circumstances. 92 See, e.g., Nancy Lewis Alvarez, Note, Racial Bias and the Right to an Impartial Jury: A Standard for Allowing Voir Dire Inquiry, 33 Hastings L.J. 959, 961 (1982) (noting that the Ristaino line of cases “unduly limit[s]” the constitutional right to question jurors about bias). A basic search of a decade’s worth of cases citing Ristaino turned up one victory. 93 For the period running January 1, 2008 through November 4, 2017, a Westlaw search of all cases citing Ristaino v. Ross resulted in just one defense victory: United States v. Bates, 590 F. App’x 882, 884, 886 (11th Cir. 2014).

*   *   *

The picture does not get any rosier if one looks beyond jury selection to other attempts to address race in the justice system. Racial disparities in charging and sentencing have an enormous effect on the criminal justice system. In McCleskey v. Kemp, the Supreme Court was con­fronted with strong statistical evidence of racial discrimination in capital sen­tencing decisions. 94 481 U.S. 279, 286–87, 295 (1987). But McCleskey effectively closed the door on discrimi­natory-prosecution claims, holding that a strong statistical correlation was not enough to prove discrimination; 95 Id. at 297. the defendant had to show that discrimination was the cause of the sentencing decision in his particular case—an almost impossible burden. The Court’s decision in United States v. Armstrong made discriminatory prosecution even harder to prove by raising the bar for obtaining the very discovery necessary to meet the bar set by McCleskey. 96 517 U.S. 456, 468 (1996) (“The justifications for a rigorous standard for the elements of a selective-prosecution claim thus require a correspondingly rigorous standard for discovery in aid of such a claim.”); see also Melissa L. Jampol, Goodbye to the Defense of Selective Prosecution, 87 J. Crim. L. & Criminology 932, 932 (1997) (arguing that “the Supreme Court’s decision in United States v. Armstrong imposes a barrier that is too high for almost any defendant alleging selective prosecution to obtain discovery, thus making the already difficult claim of race-based selective prosecution virtually impossible to prove”). A basic search of a decade’s worth of “discriminatory charging” and “selective prosecution” decisions turned up one trial victory and one intermediate appellate court victory—both of which were reversed by reviewing courts. 97 In the holding field, the search queried <“discriminatory charging” or “selective prosecution”> for the period January 1, 2008, through November 4, 2017. There was one trial court victory: State v. Pope, 713 S.E.2d 537, 540 (N.C. Ct. App. 2011). There was one appellate court victory: Lovill v. State, 287 S.W.3d 65, 81 (Tex. App. 2008), rev’d, 319 S.W.3d 687 (Tex. Crim. App. 2009). One additional case satis­fied the threshold for ordering discovery. 98 Commonwealth v. Bernardo B., 900 N.E.2d 834, 846 (Mass. 2009).

Arguably, even this dismal state of affairs is more encouraging than the Fourth Amendment’s position on racial discrimination. In Whren v. United States, the Supreme Court held that the subjective intent of an officer was not relevant to the Fourth Amendment inquiry into the con­stitutionality of a traffic stop. 99 517 U.S. 806, 813 (1996). While selective enforcement of the laws violates the Constitution, the Court held, the way to challenge that “is the Equal Protection Clause, not the Fourth Amendment.” 100 Id. According to Whren, “Subjective intentions play no role in ordinary, probable-cause Fourth Amendment analysis.” 101 Id.

Swain, Duren, Ham–Ristaino, McCleskey, and Whren—this is the anti­discrimination company Batson keeps. Of course, the Supreme Court will occasionally call out racial discrimination in a sui generis criminal case, as it did during the October 2016 term in Pena-Rodriguez v. Colorado 102 137 S. Ct. 855, 868 (2017) (reversing the conviction because a juror admitted that bias against the defendant and a witness guided his decision). and Buck v. Davis. 103 137 S. Ct. 759, 775 (2017) (reversing on grounds that the defense expert asserted defendant’s future dangerousness was increased by his race). But the Court appears to have little appetite for anti­discrimination claims that apply to more than one or two idiosyncratic cases.

Given the lack of alternatives to fight racial discrimination in the jus­tice system, Batson becomes all the more important, despite its flaws. It is the only doctrine in this beleaguered group under which defendants ac­tually have a chance to prevail. The same type of basic Westlaw query that turned up one fair cross-section victory, one jury-questioning victory, and zero discriminatory-prosecution victories in over a decade, turned up forty Batson wins over the same period. 104 For the period from January 1, 2008 through November 4, 2017, the following search was queried on Westlaw to count the number of defense victories: <Holding: (Batson /s violat!)>. This search resulted in forty Batson victories (including remands for further evidentiary proceedings), suggesting Batson is more successful than its peers. (And there are a number of additional Batson wins, even among the cases cited in this Article, that were not captured by this very elementary search. 105 See, e.g., Snyder v. Louisiana, 552 U.S. 472 (2008); Shirley v. Yates, 807 F.3d 1090, 1106 n.16 (9th Cir. 2015); Adkins v. Warden, Holman CF, 710 F.3d 1241, 1244 (11th Cir. 2013); United States v. McAllister, 693 F.3d 572, 581 (6th Cir. 2012); Harris v. Hardy, 680 F.3d 942, 945 (7th Cir. 2012); Rice v. White, 660 F.3d 242, 243 (6th Cir. 2011); United States v. Rutledge, 648 F.3d 555, 557 (7th Cir. 2011); Reynoso v. Hall, 395 F. App’x 344, 345 (9th Cir. 2010); Reed v. Quarterman, 555 F.3d 364, 365 (5th Cir. 2009); Dolphy v. Mantello, 552 F.3d 236, 237 (2d Cir. 2009); Paulino v. Harrison, 542 F.3d 692, 695 (9th Cir. 2008); Harris v. Haeberlin, 526 F.3d 903, 905 (6th Cir. 2008); Rizo v. Kernan, Nos. EDCV 03-787-JAK (AJW), EDCV 03-822-JAK (AJW), EDCV 03-824-JAK (AJW), EDCV 03-901-JAK (AJW), 2016 WL 8606275, at *10 (C.D. Cal. Oct. 13, 2016), adopted by No. EDCV-03-787 JAK (AJW), 2017 WL 1115150 (C.D. Cal. Mar. 24, 2017); Mitcham v. Davis, 103 F. Supp. 3d 1091, 1097 (N.D. Cal. 2015); City of Seattle v. Erickson, 398 P.3d 1124, 1131 (Wash. 2017); People v. Gutierrez, 395 P.3d 186, 189 (Cal.), reh’g denied by 2017 Cal. LEXIS 5957 (Cal. July 26, 2017). ) Although this search method is crude, it provides some data for the anecdotal observa­tion that winning Batson decisions are more commonly encountered than other types of antidiscrimination victories. 106 Two caveats: First, some may consider forty, or even one hundred, wins across a decade as essentially zero—just like the other antidiscrimination doctrines. But defense appellate victories are rare enough that it does not seem reasonable to equate forty with zero, at least in this author’s view. Second, the method of searching for published and unpublished opinions does not account for trial victories if there is no written decision. There are likely more Batson trial victories than there are fair cross-section, discriminatory-charging, or other types of antidiscrimination trial victories. This assumption is based on the fact that Batson trial claims, more than other types of claims, can be made spon­taneously, without extensive briefing and pleading. When a Batson trial victory occurs, there is unlikely to be an appeal by the prosecution; rather, jury selection just restarts. See infra section II.A. Whereas, if defendants were winning other types of more research-intensive claims at trial, those decisions would be appealed and result in written opinions. For this reason, this Article’s simplistic method of searching for Batson wins may actually underestimate Batson wins to a greater extent than it underestimates other doctrines’ wins. It is worth adding that Batson claims, not just Batson victories, are probably more numerous than other types of antidiscrimination claims because Batson can easily be raised and preserved in any case in which there was jury selection. Fair cross-section and discriminatory-prosecution claims, on the other hand, require significant extra-record research to raise and preserve—a point this Article re­turns to shortly. 107 Does it erode the meaningfulness of Batson if it has both more wins and more claims than its comparator doctrines? That might depend on how many more claims there really are—and there are no practical ways to estimate this. This Article argues that what makes Batson meaningful is, first, that litigants can actually deploy it, and second, that there are a non-negligible number of Batson wins. It is still a long shot, to be sure, but Batson wins are not unheard of, whereas other types of antidiscrimination wins effectively are.

All of these factors contribute to making Batson a more vibrant and significant doctrine than its antidiscrimination competitors. In criminal cases overrun by racial discrimination, Batson may be the only arrow in the quiver for a judge—especially an appellate judge—who wants to remedy the blight of discrimination. But commentary on Batson has been too quick to dismiss Batson’s virtues, too quick to lose focus of Batson’s proper context, because the commentary has focused too little on what Batson can do post-trial.

II. Batson’s Appellate–Trial Divide

Batson appellate claims are extremely difficult to win. There should be no confusion on this point. It is hard enough to prevail on a Batson claim at trial, but the “great deference” that appellate courts must give to trial court Batson determinations makes it even more difficult. 108 See supra notes 37–39 and accompanying text. At the same time, appellate Batson offers a number of opportunities that are not available to litigants at trial. This Article argues that these opportunities cause appellate Batson to punch above its trial weight, and that this di­vergence between trial and post-trial Batson has not previously been acknowledged.

This may sound like a paradox. How can Batson appellate and post­conviction litigation provide more opportunities than Batson trial litiga­tion when the appellate courts owe deference to the trial judgments? The answer has several parts. First, the value of a Batson victory on appeal and habeas is far greater than it is at trial, so even if Batson claims are less likely to succeed post-trial, their victories are nonetheless more valuable. Second, the types of evidence and arguments that can be advanced in post-trial Batson claims differ from, and exceed, those that are available at trial. Indeed, the topic of Part II is one of the main contributions of this Article: an illustration of the ways in which Batson postconviction litiga­tion is more expansive than at trial. This divergence between trial and post-trial Batson has been largely overlooked, obscuring Batson’s po­tential as a doctrine that can go beyond jury selection. This Part fleshes out the divergence between trial and post-trial Batson and identifies ways in which the appellate side of the doctrine is growing in importance, even as the trial side of the doctrine is receding.

A. Difference in Remedy Between Trial and Appellate–Habeas Litigation

To understand why Batson might be more significant to appellate lit­igants than trial litigants, start with the remedy. In baseball terms, a trial Batson win is a single or maybe a double. Post-trial, a Batson win is a home run—or more. Consider, for example, that if a Batson violation is recog­nized at trial, the remedy is to return the struck juror to the box or, more often, to dismiss the current jurors and draw a new venire. 109 Batson v. Kentucky, 476 U.S. 79, 99 n.24 (1986). For the de­fendant, the benefits are uncertain and relatively modest. Reseating the struck juror may improve the ultimate outcome of the case. Or it may not. It is hard to gauge at the outset how any individual juror will vote. Redrawing an entirely new venire could result in a significantly better choice of jurors. Or it could result in a venire that is substantially the same as—or worse than—the original one. This is the sense in which the value of a trial Batson win is uncertain. Granted, if the prosecutor has al­ready struck six African American jurors by the time the court declares the Batson violation, there is good reason to think that the new venire will be more favorable to the defense (assuming the court does not permit the prosecutor to strike African American jurors like that again). But in situations in which there are one or two suspicious strikes, it is far less clear how and whether a Batson win will affect the outcome of the case. The only thing certain is that a Batson win at trial will set back the start of the trial, perhaps by a few hours, perhaps by a few days or weeks. Hardly a significant victory.

On appeal and in postconviction proceedings, however, Batson’s remedy is much more significant: automatic reversal of the conviction. The automatic nature of the reversal is significant in setting apart post-trial Batson from trial Batson and in distinguishing Batson from other ap­pellate and habeas doctrines. The key point is that Batson is a “structural error.” 110 Tankleff v. Senkowski, 135 F.3d 235, 248 (2d Cir. 1998); see also supra note 56 and accompanying text. Structural errors are those that affect not the trial itself but the framework and mechanics for holding the trial. 111 Tankleff, 135 F.3d at 240 (“Harmless error analysis is inappropriate in this con­text, however, because exclusion of jurors on the basis of race is a structural error that can never be harmless.”); id. at 248. When a court finds a structural error, it does not go through the steps of analyzing whether any harm accrued to the defendant because of the error. 112 See Arizona v. Fulminante, 499 U.S. 279, 310 (1991) (noting structural errors that include “unlawful exclusion of members of the defendant’s race from a grand jury”); Karlan, Race, Rights, and Remedies, supra note 51, at 2018 (“Despite the recent spread of harmless error doctrine throughout constitutional criminal procedure, the federal courts that have considered the question have generally treated Batson violations as structural and thus subject to per se reversal.”). But see id. at 2019 n.87 (noting that some outlier courts have applied harmless error review to Batson violations). Rather, the court must throw out the conviction, even if there is such overwhelming evidence against the defendant that she would have been convicted had the error not occurred. “Structural error” claims, like Batson and a few others, stand out for not being subject to “harmless error review.” 113 See Fulminante, 499 U.S. at 310. In Batson, the idea of quantifying any prejudice is particularly difficult because it is not clear that any one juror’s presence or absence would change the outcome, unless one presupposes that jurors are more likely to vote for par­ticular outcomes based on race. So a requirement to show prejudice would be concep­tually impossible, on top of all the other difficulties inherent in peering into the black box of the jury. The exemption is critically important for appellate and habeas litigants because harmless error analysis is a major hurdle. 114 See Cornell W. Clayton & J. Mitchell Pickerill, The Politics of Criminal Justice: How the New Right Regime Shaped the Rehnquist Court’s Criminal Justice Jurisprudence, 94 Geo. L.J. 1385, 1408 n.152 (2006) (discussing the harmless error hurdle); Brent E. Newton, A Primer on Post-Conviction Habeas Corpus Review, Champion, June 2005, at 16, 18 (on file with the Columbia Law Review and available in full on Westlaw) (discussing “hurdles,” including harmless error review). There are different standards regarding who bears the burden of proving the error was harmless and what that burden consists of. See, e.g., Steven H. Goldberg, Harmless Error: Constitutional Sneak Thief, 71 J. Crim. L. & Criminology 421, 428 n.77 (1980) (noting that the Supreme Court has promulgated dif­ferent standards as to the burden of proof in harmless error review). In other habeas doctrines, for example, a defendant will not be able to win relief even if she has the most well-documented, clear-cut constitutional violation, if the evidence against her is significant enough that she would have been con­victed had the error not occurred. 115 See Newton, supra note 114, at 18 (explaining the standard for harmless error in federal habeas review). Such is the challenge of harmless error review: It moots out wide swaths of constitutional rights for de­fen­dants who have overwhelming evidence against them. In capital cases, in which there are separate guilt and penalty trials, another threat from harmless error analysis is the potential for the court to find that a con­stitutional violation affected only the penalty phase of the trial, thus leaving the guilty verdict untouched. 116 See Linda E. Carter, Harmless Error in the Penalty Phase of a Capital Case: A Doctrine Misunderstood and Misapplied, 28 Ga. L. Rev. 125, 125–26 (1993). Courts have more latitude in de­fining what is harmless than in defining what is a constitutional violation, so they can sidestep deciding a constitutional question under the guise of saying that the harmlessness analysis was more clear-cut. 117 See, e.g., Lipscomb v. Carothers, No. 88-3788, 1989 WL 117192, at *1 n.2 (9th Cir. Sept. 29, 1989); cf. Metrolimo, Inc. v. Lamm, 666 So. 2d 552, 554 n.2 (Fla. Dist. Ct. App. 1995) (avoiding the underlying issue of comparative negligence because the error was harmless); State v. Martin, 761 A.2d 500, 501 (N.H. 2000) (avoiding the underlying evidentiary issue because of harmless error); Daniel Epps, Harmless Errors and Substantial Rights, 131 Harv. L. Rev. (forthcoming 2018) (manuscript at 2–3) (on file with the Columbia Law Review) (arguing that harmless error should be conceived as part of the constitutional right, rather than as a remedy).

Batson’s immunity from harmless error review gives it a special luster post-trial that it lacks at trial, and this helps to explain part of the diver­gence between trial and post-trial practice. For inmates serving long prison terms or facing execution, the opportunity to vacate the conviction and to start again with jury selection is one of the biggest legal victories they could hope to achieve. Unlike at trial, going back to the beginning is not a modest remedy with uncertain benefits. For the inmate, the restart is exceedingly valuable, and not just as a means of delay. At trial, there may have been numerous errors—constitutional, tactical, and oth­erwise—that led to the defendant’s conviction or sentence. Many of those errors may not present winnable claims on appeal, but Batson’s automatic reversal wipes clean the entire slate and allows the defendant a coveted second chance to defend her liberty. 118 Karlan, Batson, supra note 57, at 397 n.77 (noting that an amicus brief in Batson “predicted that because ‘the constitutional remedy will generally occur at the trial court level’ it would ‘involve little more than a new beginning to jury selection’—a relatively small ‘transactional cos[t]’ for ‘insuring a representative jury’” (quoting Brief of Michael McCray et al. as Amici Curiae, Batson v. Kentucky, 476 U.S. 79 (1986) (No. 84-6263), 1985 WL 667869)). By this measure, Batson is a far more significant victory on appeal and in habeas than at trial. 119 Scholars have noted that the automatic-reversal remedy may make judges less likely to acknowledge a Batson violation because they are afraid of the remedy. See Karlan, Race, Rights, and Remedies, supra note 51, at 2015 (“What Batson shows is that when courts cannot calibrate the remedy, they fudge on the right instead.”); Steven M. Shepard, Note, The Case Against Automatic Reversal of Structural Errors, 117 Yale L.J. 1180, 1187–88 (2008). They suggest this would be most likely in cases in which race played no obvious role in the case—such as ones in which prosecutors did not make racially charged state­ments to the jury or when the underlying facts of the crime do not implicate racial ten­sions. In such cases, a Batson win would seem like a windfall. See Karlan, Race, Rights, and Remedies, supra, at 2022 (“By contrast, in cases without an obvious racial salience, trial judges may view Batson objections as attempts to set up grounds for per se reversal and thus may accept somewhat dubious prosecutorial explanations . . . .”).

It could be argued that any constitutional violation is more signifi­cant on appeal than at trial. For example, a trial court would remedy a violation of the Confrontation Clause by excluding the unconfronted evidence or by issuing a curative instruction to the jury to ignore the evi­dence, but these trial remedies pale in comparison to the appellate remedy of throwing out the conviction. Does that mean the trial–post-trial imbalance Batson demonstrates is no different from the Confrontation Clause? No. Batson is different because its status as struc­tural error severs any connection between what happens on appeal and what would have happened at trial. The appellate Batson litigant actually gets more than she would have at trial. In the Confrontation Clause example, the defen­dant who shows a violation at trial is able to exclude some significant piece of evidence, and that presumably changes the likely outcome of the defendant’s case. The same theory holds on appeal: If the Confrontation Clause violation proves significant to the outcome of the case, the defendant gets a new trial and that piece of evidence will be excluded, presumably giving her a better chance at victory. In the Batson example, it is extremely unclear how the defendant would have been better off at trial if the particular struck juror had not been struck, a point noted in more depth above. 120 See supra notes 113–114 and accompanying text. But the appellate Batson litigant clearly benefits from the full reversal, even without showing that the struck jurors would have affected the outcome of her case. This distinc­tion makes the trial–post-trial gap in Batson more significant than in other constitutional doctrines. 121 One important caveat, of course, is that the “victorious” appellate defendant can still be retried. So one might argue that this appellate win, in all its drama, still only restores the defendant to the position she would have occupied had she won at trial. An obvious rejoinder, however, is that retrying someone years after the initial conviction is no easy feat and may benefit the defendant.

B. Difference in Evidence Between Trial and Post-Trial Proceedings

A further manifestation of the divide between trial and appellate Batson can be seen in the way in which habeas litigants use types of evi­dence that are not available to them at trial. Batson is even more expan­sive in this respect than other record-expanding habeas doctrines, like Brady v. Maryland 122 373 U.S. 83, 87 (1963) (holding that “the suppression by the prosecution of evi­dence favorable to an accused upon request violates due process where the evidence is material either to guilt or to punishment, irrespective of the good faith or bad faith of the prosecution”). and Strickland v. Washington. 123 466 U.S. 668, 698 (1984) (articulating “the general standards for judging ineffec­tiveness claims”). Because of Batson’s focus on the prosecutor’s intent, a surprisingly wide range of evidence becomes relevant—even evidence that was not in existence at the time of trial.

This section discusses several types of evidence and analytical meth­ods that are not available at trial but are available later on. Because much of the discussion concerns extra-record evidence, a brief definition of re­cord and extra-record evidence is useful. Batson is often considered a record-based doctrine. 124 See, e.g., Transcript of Oral Argument at 44, Johnson v. California, 541 U.S. 428 (2004) (No. 03-6539), 2004 WL 5607668 (“[T]he res gestae takes place in front of the court. It takes place in front of the parties therefore. Everything that that party needs is avail­able to the party.”); Matthew Seligman, Note, Harrington’s Wake: Unanswered Questions on AEDPA’s Application to Summary Dispositions, 64 Stan. L. Rev. 469, 485–86 (2012) (“A paradigmatic record-based claim is a Batson claim . . . . All the evidence re­quired to make out a Batson claim will be contained in the trial record—the pro­secution’s use of peremptory challenges, the race of the prospective jurors, any race-neutral expla­nation for the peremptory challenges  . . . , and so on.”). The evidence needed for a Batson claim typically comes from the record of the defendant’s case: the jurors’ question­naires and the transcripts of voir dire and the peremptory strikes. There is no need to conduct any research from outside of the case—research that would yield extra-record information. 125 Transcript of Oral Argument at 44, Johnson, 541 U.S. 428 (No. 03-6539), 2004 WL 5607668. Indeed, this was one of Batson’s chief innovations over Swain; it relieved defendants of the need to research the prosecutor’s behavior in other cases. 126 See supra notes 68–71 and accompanying text.

Even though Batson does not require the defendant to look beyond the record of the case, it also does not prevent her from doing so if there is extra-record evidence that reflects the prosecutor’s bias. This extra-record evidence can consist of the prosecutor’s Batson violations in other cases, or the prosecutor’s racist comments in the media, or internal office policies that promote racial discrimination—anything that speaks to the prosecutor’s intent. Another meaning of extra-record evidence is evi­dence that may relate to the defendant’s own case, but was not part of the record of the case at trial. A prime example can be seen in the prose­cutor’s notes from jury selection. These notes may be unavailable at trial insofar as they are considered privileged work product, yet they may be­come available to the defendant after trial in postconviction proceedings.

As this section shows, extra-record evidence has become increasingly important to Batson. As prosecutors grow more adept at producing sani­tized justifications for their strikes at step two of Batson, extra-record evi­dence of the prosecutor’s discriminatory intent is increasingly important in winning Batson claims. What is particularly significant, for purposes of this Article, is the way in which this extra-record Batson evidence has made a greater impact on post-trial litigation. Even though trial litigants are capable of incorporating most—but not all—types of extra-record evi­dence into their cases, the extra-record evidence and analytical meth­ods described here remain largely a tool of postconviction and appellate Batson. This divergence provides another example of the split between Batson trial and post-trial practice.

1. Jury-Selection Notes. — There is no better evidence of a prosecu­tor’s intent than her notes from jury selection. A number of significant Batson opinions—including the Supreme Court’s recent decision in Foster v. Chatman 127 136 S. Ct. 1737, 1755 (2016). —have used these jury-selection notes to demonstrate the prosecutor’s bias. But there is a significant difference in the availability of these notes at trial compared to in postconviction proceedings.

In Foster v. Chatman, the prosecutor’s jury-selection notes showed that race was a dominant factor in jury selection, despite the prosecu­tion’s repeated assertions to the contrary. 128 Id. at 1743–44, 1754–55. At trial, Foster asked for the notes, but the prosecutor refused to turn them over, and the court did not force the issue. 129 Transcript of Oral Argument at 19, Foster, 136 S. Ct. 1737 (No. 14-8349), 2015 WL 6694912 (“[T]he defense lawyers at trial did move for the prosecution’s notes. And the prosecution opposed that. They’re very strict in not—not giving up their notes.”). When the notes came out years later in postconvic­tion proceedings, they showed that the prosecutor had targeted jurors because of race. 130 Foster, 136 S. Ct. at 1743–44, 1754–55. For example, the notes contained the letter “B” next to the names of all the African American jurors—names that the prosecutor had highlighted in green pen. 131 Id. at 1744. The notes also showed that all five of the venire’s qualified black prospective jurors made it onto the prosecu­tor’s list of the six least acceptable prospective jurors. 132 Id. Three decades after the trial, the jury-selection notes revealed the prosecutor’s obsession with race and led the Supreme Court to throw out the defendant’s conviction. 133 Id. at 1755.

In a number of other cases from the Supreme Court, the lower fed­eral courts, and the states, jury-selection notes have also played a role in showing grotesquely racist practices in jury selection. 134 See, e.g., Miller-El v. Dretke, 545 U.S. 231, 266 (2005) (finding that the prosecutors noted the race of prospective jurors in an attempt to avoid impaneling minorities); see also Shirley v. Yates, 807 F.3d 1090, 1106 n.16 (9th Cir. 2015) (“Another category of circumstantial evidence not available in this case—contemporaneous notes—may provide far stronger evidence that an asserted reason in fact motivated the strike of a long-since-forgotten veniremember.”); In re Freeman, 133 P.3d 1013, 1018 (Cal. 2006) (noting the use of the prosecutor’s notes to corroborate testimony that he exercised three peremptory challenges because he believed the prospective jurors were Jewish). In a notable deci­sion by District Judge Lucy Koh, the notes revealed such attentiveness to race that the prosecutor struck white jurors because one was “inter­est[ed] in African American culture and had written a book on African American folklore” and another’s answers to voir dire questions sug­gested she might be married to a black man. 135 Mitcham v. Davis, 103 F. Supp. 3d 1091, 1097 (N.D. Cal. 2015) (granting habeas relief on a claim of ineffective assistance of counsel for failure to raise what likely would have been a winning objection under Wheeler, a state law analogue of Batson). By the same token, prosecutors have used jury-selection notes to defend themselves against Batson allegations when the notes show nondiscriminatory reasons for their strikes. 136 See, e.g., Adkins v. Warden, 710 F.3d 1241, 1245 (11th Cir. 2013) (explaining that jury-selection notes help the prosecutor explain his strikes in a nondiscriminatory way); United States v. Tindle, 860 F.2d 125, 128 (4th Cir. 1988) (noting that the prose­cutor submitted his notes to the judge in camera to rebut the defendant’s prima facie case of discrimination); McCrory v. Henderson, 871 F. Supp. 597, 601 (W.D.N.Y. 1995) (allow­ing the prosecutor to use notes on remand to try to provide a neutral justification at Batson step two for his peremptory strikes), rev’d on other grounds, 82 F.3d 123 (2d Cir. 1996); Transcript of Oral Argument passim, State v. Osorio, 973 A.2d 365 (N.J. 2009) (No. A-59-08), 2009 WL 2703124. But cf. Transcript of Oral Argument at 14, United States v. Stephens, 514 F.3d 703 (7th Cir. 2008) (No. 06-2892), 2007 WL 5514343 (“JUDGE: Yeah, but wouldn’t it be surprising if a prosecutor’s notes would be considered determinative of a challenge? . . . [I]f you were someone who really was biased or prejudiced which I would never imagine that in your office. You might make, you know, notes that cover your tracks . . . .”).

The difference between trial and habeas access to the notes is sig­nificant. At trial, the prosecutor’s notes are generally protected from disclosure under the work-product privilege, based on the theory that the notes reflect the prosecutor’s thinking about tactics and strategy for liti­gating the case. 137 Some courts hold the notes are privileged. See Ted A. Donner & Richard K. Gabriel, Jury Selection Strategy and Science § 3:7, Westlaw (database updated Dec. 2016) (“A number of courts have likewise found that counsel’s notes, taken during the voir dire, should not be subject to production during a Batson hearing because they constitute attorney work product.”); see also People v. Trujillo, 15 P.3d 1104, 1107–08 (Colo. App. 2000). For examples of other cases in which the prosecutor’s notes were held to be privi­leged, see Foster v. State, 374 S.E.2d 188, 192 (Ga. 1988); People v. Mack, 538 N.E.2d 1107, 1115–16 (Ill. 1989); Thorson v. State, 721 So. 2d 590, 595–96 (Miss. 1998); State v. Antwine, 743 S.W.2d 51, 67 (Mo. 1987); Guilder v. State, 794 S.W.2d 765, 767–68 (Tex. App. 1990). Defendants generally do not ask for prosecutors’ notes mid-trial. Nor would courts be likely to grant such requests. 138 This is not to say that a judge would be powerless under Batson to order such disclosure. See infra note 142 and accompanying text. In post­con­viction proceedings, however, there is a different approach to the notes. Prosecutors may continue to assert that the notes are privileged work product, or they may claim that the notes do not properly fall within the scope of any postconviction discovery right. In many cases, though, the notes wind up being disclosed to the defendant on habeas review, because there is simply less justification for protecting the prosecutor’s strategy and tactics once the trial has already run its course. The disclo­sure of the notes to the defense sometimes comes about voluntarily. 139 In one remarkable case, a prosecutor disclosed the notes during the direct appeal upon discovering what the trial prosecutor had written in them. United States v. Baskerville, No. 03-836 (JAP), 2011 WL 159782, at *1 (D.N.J. Jan. 18, 2011) (noting that the government voluntarily moved for remand on direct appeal when the new prosecutor discovered the original prosecutor’s notes); Appellant William Baskerville’s First Step Supplemental Brief on Appeal & Supplemental Appendix at *23, United States v. Baskerville, 448 Fed. App’x 243 (3d Cir. 2011) (Nos. 11-1175, 07-2927), 2011 WL 858744 (“The prosecutors’ voir dire notes show that white jurors who had friends or family who served time in prison were graded ‘Excellent’ and ‘Very Good’ while black jurors with ‘bad apple’ relatives were graded ‘Strike’ or ‘Good.’”); see also Foster v. Chatman, 136 S. Ct. 1737, 1743–44 (2016) (explaining that the notes were turned over pursuant to a state pub­lic records act request); Majid v. Portuondo, 428 F.3d 112, 115 (2d Cir. 2005) (describing in camera review of the prosecutor’s notes on remand); id. at 119 (noting that the pro­secutor voluntarily provided voir dire notes to the defense on remand); Mitcham, 103 F. Supp. 3d at 1097 (turning over notes in postconviction proceedings). There is also the practical reality that the attorney representing the government on appeal will often not be the same as the trial attorney, so her personal interests in the privilege may be lessened. Often, the district attorney’s office is no longer representing the pro­secution in ha­beas proceedings, and the new agency handling the case—likely the state attorney general’s office—may have less investment in the notes’ secrecy.

Another avenue to accessing the notes is postconviction-discovery lit­igation. In these discovery proceedings, a state or federal judge must determine whether there has been a sufficient showing to entitle the de­fendant to access the notes. A range of views exists on what is necessary to trigger disclosure, and some courts have created special carve-outs for when the notes must be disclosed, as in instances when the prosecutor used her notes at trial to refresh her recollection of the strikes. 140 See, e.g., Goode v. Shoukfeh, 943 S.W.2d 441, 449 (Tex. 1997) (“We hold that an Edmonson movant has the right to examine the voir dire notes of the opponent’s attorney when the attorney relies upon these notes while giving sworn or unsworn testimony in the Edmonson hearing.”); see also Guilder, 794 S.W.2d at 772–73 (“[N]owhere in his brief does appellant tell us that the prosecutor-witness used his notes to refresh his memory for the purposes of testifying. Indeed, appellant nowhere raises the admissibility of the prose­cutor’s notes under [this rule].”); 4 Gregory B. Butler, Duncan Ross MacKay, Ann H. Rubin & Jason R. Gagnon, Successful Partnering Between Inside and Outside Counsel § 64:34 n.3, Westlaw (database updated May 2017) (discussing how notes, such as a “jury consultant’s report,” may have to be disclosed if the prosecutor relied on them “to refresh his or her recollection as to the basis of a peremptory challenge”). A popular way to balance the Batson and work-product interests in the notes is to have the court review them in camera to determine whether they contain any support for a Batson claim. 141 In a number of cases, notes are disclosed in camera. See Harris v. Haeberlin, 752 F.3d 1054, 1057 (6th Cir. 2014) (noting that the district court, on habeas remand, re­viewed the prosecutors’ notes); Tindle, 860 F.2d at 128; United States v. Garrison, 849 F.2d 103, 105 (4th Cir. 1988); Gibson v. Wetzel, No. 11-4550, 2016 WL 1273626, at *4 (E.D. Pa. Mar. 31, 2016). Rounding out this brief survey of dis­closure procedures is the example of the federal judge who ordered the notes disclosed in postconviction proceedings because, in his view, the work-product privilege cannot be used to conceal racial discrimination. 142 Courts have also held that there is no privilege over intentional discrimination. Johnson v. Finn, Nos. CIV S-03-2063 RBB JFM P, CIV S-04-2208 RBB JFM P, 2007 WL 3232253, at *3–4 (E.D. Cal. Oct. 31, 2007) (holding there is no work-product privilege when it comes to a violation of the Equal Protection Clause and therefore ordering discovery).

Amidst a jumble of conflicting rules about when a defendant may access the notes, the key point is that the notes are much more accessible post-trial than during trial. This is significant because the prosecutor’s notes are the best evidence of her intent, yet the notes are stuck in a legal limbo. No one denies their importance to resolving Batson claims, but they remain beyond the defendant’s grasp at trial. It is only in post-trial pro­ceedings that they may become available to the defendant. This gap in access to the notes provides one striking example of the divergence be­tween trial and post-trial Batson.

2. The Prosecutor’s Behavior in Other Cases and Outside of Court. — The prosecutor’s behavior in other cases or outside of court is another exam­ple of extra-record evidence used in post-trial proceedings but largely absent from Batson trial litigation. If a prosecutor has violated Batson in one case, that violation can be proof of her bias in other Batson cases. 143 Madison v. Comm’r, 761 F.3d 1240, 1252 (11th Cir. 2014) (noting “the Mobile County District Attorney’s Office’s well-documented history of racially discriminatory jury selection, including at Mr. Madison’s first trial”); Hightower v. Terry, 459 F.3d 1067, 1078 (11th Cir. 2006) (“Here, Briley himself authored a memorandum that spelled out the scheme to limit the number of blacks in the jury pool.”); Rizo v. Kernan, Nos. EDCV 03-787-JAK (AJW), EDCV 03-822-JAK (AJW), EDCV 03-824-JAK (AJW), EDCV 03-901-JAK (AJW), 2016 WL 8606275, at *10 (C.D. Cal. Oct. 13, 2016), adopted by No. EDCV-03-787 JAK (AJW), 2017 WL 1115150 (C.D. Cal. Mar. 24, 2017) (“[T]he prosecutor was found to have purposefully discriminated against at least one African-American prospective juror in petitioners’ first trial.”); cf. People v. Howard, 824 P.2d 1315, 1326 n.4 (Cal. 1992) (“How­ever, even if a trial court might properly consider a prosecutor’s past in deter­mining whether a prima facie case exists, a court obviously cannot consider a prosecutor’s future.”). The same is true of the prosecutor’s discriminatory statements or actions. Although these instances of discrimination occur outside the four cor­ners of the defendant’s case, they show the prosecutor’s bias. But, for reasons discussed below, this type of evidence is embraced more in post­conviction proceedings than at trial, thus furthering the gap between trial and appellate Batson. Indeed, there is an irony in the use of such extra-record evidence. 144 See supra notes 67–72 and accompanying text. It feels like Batson has taken a step toward Swain v. Alabama, the peremptory challenge doctrine that required defendants to investigate and describe all the prosecutor’s peremptory strikes in oth­er cases. 145 See Miller-El v. Dretke, 545 U.S. 231, 240 (2005) (“[A]lthough some false reasons are shown up within the four corners of a given case, sometimes a court may not be sure unless it looks beyond the case at hand. Hence Batson’s explanation that a defendant may rely on ‘all relevant circumstances’ to raise an inference of purposeful discrimination.” (quoting Batson v. Kentucky, 476 U.S. 79, 96–97 (1986))); supra section I.C.1 (describing Swain as one of Batson’s antidiscrimination comparators).

Evidence of prosecutors’ misbehavior in other cases can be quite egregious. In one example, a California prosecutor was found to have violated Batson in the defendant’s first trial and subsequent retrial. 146 Currie v. McDowell, 825 F.3d 603, 607–08, 610–11 (9th Cir. 2016) (“In Miller–El v. Cockrell, the . . . fact that prosecutors belonged to a[n] . . . office with a history of racial bias . . . bolster[ed] [the] finding of a prima facie case. In this instance, it is not only the same office, but the same prosecutor, who brings a history of Batson violations with him.” (citation omitted)). When the defendant’s second retrial reached the Ninth Circuit on federal habeas review, Judge Marsha Berzon commented at oral argument that the prosecutor’s history of bias was so pervasive that she would find a prima facie case of racial discrimination anytime the prosecutor struck a minority juror. 147 Oral Argument at 19:00, Currie, 825 F.3d 603 (No. 13-16187), http://www.ca9. (on file with the Columbia Law Review); cf. Brief of Appellant at 108, Flowers v. State, 158 So. 3d 1009 (Miss. 2014) (No. 2010–DP–01348–SCT), 2013 WL 9982812 (“Here, it was not only the same office that had engaged in discrimination in jury selection, but the same prosecutor; it was not only the same prosecutor, but the very same case.”).
Numerous other courts have considered a prosecutor’s behavior in other cases as evidence of discriminatory intent in the defen­dant’s particular case. 148 See, e.g., Miller-El v. Cockrell, 537 U.S. 322, 347 (2003) (inferring discrimination, in part, on the basis that the prosecutor’s office had a history of discrimination); Edwards v. Roper, 688 F.3d 449, 458 (8th Cir. 2012) (discussing evidence of seven other Batson reversals in St. Louis County prosecutions and citing a newspaper article about a St. Louis County prosecutor “who claimed that other attorneys advised him to strike blacks from juries”); Coombs v. Diguglielmo, 616 F.3d 255, 265 (3d Cir. 2010) (finding evidence of bias in the prosecutor’s statement to defense counsel, outside of court, that the de­fendant’s first jury had hung “only because of a sympathetic Black juror”); Michelle May, Cent. Cal. Appellate Program, How to Make a Batson Case on Appeal 29, http:// (on file with the Columbia Law Review) (noting that “evidence of the prosecutor’s peremptory challenges in other cases—or the DA’s office’s policies or actions in other cases” may be present in the record and, if it is, can be used as “a factor supporting a third-stage Batson showing”).

And it is not just Batson violations from other cases that can be used to show a prosecutor’s discriminatory intent in jury selection. Anything the prosecutor says or does, in court or out of court, can potentially show her bias and, thus, serve as evidence that a strike of hers was motivated by race. An unlucky prosecutor in Kentucky, for example, was caught on a hot mic during a court recess talking about jury selection in racial terms: “We’ve got [name deleted], 49, she’s the old lady, the black lady. The other one is already off.” 149 Harris v. Haeberlin, 752 F.3d 1054, 1056 (6th Cir. 2014) (emphasis added by Harris) (internal quotation marks omitted) (quoting Harris v. Haeberlin, 526 F.3d 903, 907 (6th Cir. 2008)). The recording was discovered soon after the de­fendant’s conviction and made part of the record to be considered on the direct appeal. 150 Harris, 526 F.3d at 913; see also Harris, 752 F.3d at 1056.

Racist comments made in closing argument provide other examples of evidence relevant to the Batson inquiry because they, again, reflect the prosecutor’s bias. 151 See Petition for Writ of Certiorari at 22, Foster v. Chatman, 136 S. Ct. 1737 (2016) (No. 14-8349), 2015 WL 2457657. The same is true of racist comments in the press, civil judgments for discriminatory employment practices, and member­ship in the Ku Klux Klan. 152 See Petition for a Writ of Certiorari at 21–22, Dressner v. Louisiana, 562 U.S. 1271 (2011) (No. 10-752), 2010 WL 4959741 (“[T]he prosecutors in this case were the very same ones that amici identified in Snyder who wore neckties depicting a hangman’s noose and a grim reaper in death penalty proceedings.”); see also Brief of Nine Jefferson Parish Ministers as Amici Curiae Supporting Petitioner at 13–14, Snyder v. Louisiana, 552 U.S. 472 (2008) (No. 06-10119), 2007 WL 2605448 (citing Jeffrey Gettleman, Prosecutors’ Morbid Neckties Stir Criticism, N.Y. Times, Jan. 5, 2003, at A14); Sheri Lynn Johnson et al., Racial Epithets in the Criminal Process, 2011 Mich. St. L. Rev. 755, 782 (“[A] prosecu­tor’s use of a racial epithet could be probative of his racial motivation in striking a minor­ity race juror.”). All of this is relevant to the Batson claim inso­far as it shows the prosecutor’s bias. 153 At the same time, extra-record evidence about the prosecutor’s past can some­times help her defend against claims of bias. A judge’s experience with the prosecutor—whether in previous cases or socially—can serve as evidence that the prosecutor did not harbor bias and was not guilty of a Batson violation. See Alaska Rent-a-Car, Inc. v. Avis Budget Grp., 738 F.3d 960, 966 (9th Cir. 2013) (noting that the judge and opposing counsel “both expressed their confidence that Avis’s lawyer was not racially motivated”); Adkins v. Warden, 710 F.3d 1241, 1246 (11th Cir. 2013) (showing how the judge invoked his “own personal experience with the prosecutor in other cases” to find the prosecutor’s nondiscriminatory explanations to be credible).

A detailed illustration of this extra-record evidence can be seen in the Supreme Court’s 2008 decision Snyder v. Louisiana. 154 552 U.S. at 472–73. In that case, nine local ministers filed an amicus brief cataloguing the many Batson violations that the prosecutor’s office committed. 155 Brief of Nine Jefferson Parish Ministers as Amici Curiae Supporting Petitioner, supra note 152, at 7–10. And Batson violations were only the start. The amicus brief provided statistics on the racial use of peremptories in a dozen other capital prosecutions in the office 156 Id. and presented the Justices with examples of racially offensive statements that the trial prosecutor and a colleague made to a journalist in a joint inter­view—statements that referred to “black on black” murder and “white man’s justice” and joked about seating “Nazis on capital juries.” 157 Id. at 13 (internal quotation marks omitted) (quoting Ivan Solotaroff, The Last Face You’ll Ever See: The Private Life of the American Death Penalty 52 (2001)). The brief further recounted that the trial prosecutor’s office displayed “a tiny model electric chair holding cut-out faces of five African-American con­demned men,” 158 Id. and that the prosecutors in the office had “appeared in capital courtroom proceedings with neckties depicting a grim reaper and a hangman’s noose.” 159 Id. With one example after another, the brief showed how extra-record evidence can be used to demonstrate the pros­ecutor’s discriminatory intent.

What must again be emphasized, for purposes of this Article, is that the use of this type of extra-record evidence is a greater part of Batson habeas litigation than Batson trial litigation. This is in part a practical matter. Trial attorneys are more likely to plead Batson claims on the fly, typically without thoroughly searching the prosecutor’s history of dis­crimination in other cases. At trial, given the time constraints and the relatively modest value of winning a trial Batson claim, litigators may feel their time is better spent on other areas of the case. Appellate litigants, on the other hand, are well equipped to conduct the type of record col­lection and investigation that is required to develop such extra-record evidence. And the enormous benefit of a Batson appellate victory pro­vides an incentive to vigorously pursue such claims.

There is another practical consideration related to the divergence between trial and post-trial use of this type of extra-record evidence: Some of the evidence may not come into existence until after the de­fendant’s trial. Maybe the prosecutor joined the Ku Klux Klan shortly after the defendant’s conviction. Maybe the prosecutor earned a number of Batson reversals in the years between the defendant’s conviction and the defendant’s postconviction proceedings. Obviously, trial counsel cannot use evidence that is not in existence at the time, but the postcon­viction proceedings happen so long after the trial that there is a lot of time for this new evidence to come into existence. As will be described in more depth later on, 160 See infra section III.C.1. it is Batson’s focus on the prosecutor’s intent that makes such late-arising, extra-record evidence relevant to the claim. 161 Cf. Stephen B. Bright & Katherine Chamblee, Litigating Race Discrimination Under Batson v. Kentucky, Crim. Just., Spring 2017, at 10, 11–12 (recommending the devel­opment of a “database to track the strikes of black jurors across cases” because “[w]hen prosecutors have discriminated in one case, chances are they have discriminated in many others . . . [and] it is much harder to explain a well-defined pattern of disproportionate strikes across cases”). A declaration from the prosecutor about her own prior practices is also relevant. See In re Freeman, 133 P.3d 1013, 1015 (Cal. 2006) (“One time . . . Judge Golde called me into chambers and asked rhetorically ‘Quatman, what are you doing?’ . . . He said I could not have a Jew on the jury . . . .” (internal quotation marks omitted) (quoting declaration of former prosecutor John R. Quatman)).

Litigation style may also explain why extra-record evidence takes on greater importance post-trial than at trial. At trial, a defense attorney must stand in the presence of the prosecutor—perhaps even a prosecu­tor she will have to work with on a regular basis—and accuse the prosecutor of racial discrimination. If the defense attorney begins to bring in evidence of the prosecutor’s racism from far and wide, the accu­sations become more and more characterological: It is not just that the prosecutor allowed race to influence this particular strike, but rather, the prosecutor has a racist character. Such broad accusations about the char­acter of the prosecutor are relevant to the Batson claim, of course, but many commentators have noted that defense attorneys may still find it awkward to make these accusations. 162 Cf. supra note 35 and accompanying text.

Appellate litigation, on the other hand, has a different flavor. The appellate proceedings are carried out mostly on paper, which gives them a more impersonal feel. And, as mentioned earlier, the trial prosecutor is likely not even the attorney representing the state on appeal. 163 See supra note 139 and accompanying text. Because of this dynamic, it may be easier for the appellate attorney to present a detailed history of the prosecutor’s racial prejudice.

Regardless of the precise reason, the difference between the trial and post-trial use of this extra-record evidence is significant in creating a gap between the two sides of the Batson doctrine.

3. Policies, Trainings, and Internal Memos. — Internal manuals, train­ing materials, and office policies provide still another category of extra-record material that can be used to show the prosecutor’s intent. 164 See, e.g., Cochran v. Herring, 43 F.3d 1404, 1412 (11th Cir. 1995) (noting prose­cutors’ informal policy of striking jurors based on race). This type of evidence has been part of jury-selection litigation going all the way back to Swain. Just as with the jury-selection notes, these materials are considered powerful, probative evidence, but there is little clarity about when a litigant is entitled to them. The recurring theme, however, is that appellate litigants have easier access to these materials than do trial litigants.

In Miller-El v. Dretke, the fact that the prosecutors’ files contained the Sparling Manual was evidence that “race was on their minds when they considered every potential juror.” 165 545 U.S. 231, 266 (2005); see also Miller-El v. Cockrell, 537 U.S. 322, 334–35 (2003) (explaining the District Attorney’s Office instructed prosecutors to use peremptory strikes against minorities (“Do not take Jews, Negroes, Dagos, Mexicans or a member of any minority”) and distributed a manual entitled Jury Selection in a Criminal Case with an article “outlining the reasoning for excluding minorities from jury service”); Reed v. Quarterman, 555 F.3d 364, 382 (5th Cir. 2009) (“The Court in Miller-El II relied, in part, on the Sparling Manual to glean the history of racial discrimination in the Dallas County District Attorney’s Office, and Reed presented this same document at his Batson hear­ing.”). The same was true in Tucker v. Thomas, in which postconviction counsel discovered a manual in the prosecutor’s file showing how to justify peremptory strikes in race-neutral ways. See First Amended Petition for Writ of Habeas Corpus at 40, Tucker v. Thomas, No. 1:07–CV–868, 2017 WL 4011249 (M.D.N.C. Sept. 11, 2017). The Sparling Manual proved important in Miller-El and other Texas cases because the manual explicit­ly in­structed prosecutors on how to misuse race in jury selection. 166 See Miller-El, 545 U.S. at 266 (explaining that the “prosecutors took their cues from a 20-year-old manual of tips on jury selection” that put an “emphasis on race”). Miller-El also included testimony from a former Dallas prosecutor who was told by a superior that if he allowed an African American on his jury he would be fired. 167 Id. at 264. The claim in Miller-El was pleaded at trial as a Swain claim because the trial took place before Batson was decided. Id. at 263.

The Philadelphia District Attorney’s Office provides yet another ex­ample of how internal trainings can be used to show a prosecutor’s racist intent. There, a prosecutor was caught in a taped training session ex­plaining to other prosecutors how to remove African Americans—especially African American women—from the jury. 168 See Wilson v. Beard, 426 F.3d 653, 655 (3d Cir. 2005) (discussing videotaped training of Assistant District Attorney Jack McMahon discussing race-based techniques for jury selection); Watkins v. Klopotoski, No. 08-5802, 2009 WL 6593918, at *4 n.4 (E.D. Pa. Dec. 11, 2009) (“[T]he Third Circuit determined that the now infamous McMahon tape, which depicted Philadelphia ADA Jack McMahon advocating the use of peremptory strikes against African American jurors in 1987, provided insight into McMahon’s own motivation for striking black jurors during his prosecution of a petitioner in 1984.”), adopted by No. 08-5802, 2010 WL 2431610 (E.D. Pa. June 14, 2010). The training came replete with instructions on how to avoid being caught by Batson. 169 Wilson, 426 F.3d at 658. Significantly, the timing of the video’s release suggests that it was disclosed for political reasons rather than out of any particular fealty to equal protection. Id. at 656. Simi­larly, North Carolina prosecutors allegedly took part in a “Top Gun” training academy “where they provided a cheat sheet of pat, ‘race-neutral’ expla­nations” for Batson challenges, according to a publication by one ACLU lawyer. 170 Cassandra Stubbs, ACLU Capital Punishment Project, Strengthening Batson Challenges with the RJA-MSU Study for Durham-Area Practitioners 7 (2016), http:// [
The list of attendees and the contents of the training have since been distributed within the criminal defense bar to assist in showing attendees’ bias. 171 See id.

In the context of the gap between trial and post-trial Batson, it is im­portant to point out that these internal materials are frequently consid­ered privileged work product and can be difficult to obtain at trial. To the extent that they are obtained and used in Batson cases, it is usually by postconviction litigants. When these policies, trainings, and manuals come to light, it is often through leaks rather than discovery or public records requests. 172 See supra note 169 and accompanying text; see also supra note 137 and accompanying text (providing several examples of the trial prosecutor’s notes being disclosed voluntarily). And the fact that the office-wide materials apply to so many prosecutors in the office means that once these manuals and train­ings are leaked to one defendant, they are likely to spread to many others as well. In this way, a trickle of information in one case can turn into a torrent of postconviction claims in other cases, as defendants who were prosecuted by the same offending prosecutor or prosecutor’s office learn of them.

4. Comparative Juror Analysis. — The next example of divergent evi­dence between trial and post-trial proceedings is conceptually quite different. Here, this Article considers an analytical tool—comparative juror analysis—that is a mainstay of Batson appellate litigation but that is not practically deployable at trial. Comparative juror analysis draws compar­isons between those jurors who were struck and those who were seat­ed. 173 See Miller-El v. Dretke, 545 U.S. 231, 241 (2005) (“If a prosecutor’s proffered reason for striking a black panelist applies just as well to an otherwise-similar nonblack who is permitted to serve, that is evidence tending to prove purposeful discrimination to be considered at Batson’s third step.”); United States v. Barnette, 644 F.3d 192, 201–02 (4th Cir. 2011) (considering a comparative-juror-analysis challenge to jury selection); Crittenden v. Ayers, 624 F.3d 943, 956 (9th Cir. 2010) (“Comparative juror analysis is an established tool at step three of the Batson analysis for determining whether facially race-neutral reasons are a pretext for discrimination.”). The purpose of the comparisons is to determine whether the prosecu­tor’s stated reasons for striking a particular juror hold up across the entire venire. 174 There are variations on this that this Article does not discuss in any depth here. Sometimes, the comparison group may not be seated jurors but rather jurors whom the prosecutor did not strike and who did not make it onto the panel (because, perhaps, the defendant struck them or the jury was accepted by both sides before those jurors came into the box). See, e.g., People v. Lomax, 234 P.3d 377, 411 n.14 (Cal. 2010) (“In general, a comparative juror analysis ‘compares panelists who were struck with those who were allowed to serve or were passed by the prosecution before being ultimately struck by the defense.’” (quoting People v. Lenix, 187 P.3d 946, 967 (Cal. 2008))). If they do not, that is a sign that the justification for the particular strike may be false. For example, if a prosecutor struck a black juror on the grounds that the juror lacked a college degree, one would expect the prosecutor to strike white jurors who lacked college degrees. If she did not strike white jurors without college degrees, that suggests the demand for a college degree was a pretext and the true mo­tivation for the strike was something else, perhaps race.

Because jurors typically differ on more than one dimension, compar­ative juror analysis can become quite complicated and time con­suming. A careful reading of the jurors’ questionnaires and voir dire answers is required to create the appropriate comparisons. The process is akin to multivariable regression analysis; litigants attempt to show that the two jurors being compared are largely similar except for one salient characteristic: race. 175 See Barnette, 644 F.3d at 201–02 (rejecting a comparative-juror-analysis challenge because the defendant did not adequately account for possible factors other than race). If the juror’s race begins to look like the best predictor of the prosecutor’s strikes, that is powerful evidence of racial discrimination.

The point worth emphasizing is comparative juror analysis’s relative prominence post-trial as compared to at trial. 176 Case law in many jurisdictions requires appellate courts to perform comparative juror analysis in evaluating Batson claims. Grossman, supra note 40, at 12 (“Federal courts have concluded that in order to determine if the prosecutor’s reasons were pretextual, a comparative analysis is appropriate.”); see also People v. Gutierrez, 395 P.3d 186, 204 (Cal. 2017) (Liu, J., concurring) (“[C]omparative juror analysis is an important tool in ferreting out improper discrimination, and the mandate to consider all relevant circumstances means a court must undertake comparative juror analysis even if it is raised for the first time on appeal.” (citations omitted)). Unlike the extra-record evidence mentioned above, the inputs for comparative juror analysis are all accessible to the parties at the time of trial. The juror questionnaires, the voir dire transcript, and the transcript of the prosecutor’s strikes pro­vide the data that are used to compare the jurors. 177 Sometimes, however, the race of the jurors is not actually recorded in the record on appeal. See, e.g., Williams v. Beard, 637 F.3d 195, 214 (3d Cir. 2011). But trial attorneys are not really in a position to use this information for the type of in-depth comparisons that are routine on appeal. 178 This is true even if trial courts are nominally required to perform the analysis. See Shirley v. Yates, 807 F.3d 1090, 1102 n.9 (9th Cir. 2015); Boyd v. Newland, 467 F.3d 1139, 1149 (9th Cir. 2006); Lenix, 187 P.3d at 961. It is not clear how such analysis can really be done at trial. See Reed v. Quarterman, 555 F.3d 364, 369–70 (5th Cir. 2009) (discussing comparative juror analysis on appeal and whether it should be procedurally barred because it was not raised at trial). This is not simply a matter of trial attorneys’ time limitations. 179 See supra section II.B.2. A bigger problem is that Batson chal­lenges are generally brought in the midst of jury selection, and at that time no one knows which jurors will end up on the panel and which will ultimately be struck. Without knowing the identities of the seated jurors—or the identities of all the struck jurors—the parties can­not effectively compare those who were seated with those were struck in the way that comparative juror analysis typically envisions. 180 Comparisons are not impossible, and some attorneys will point out comparisons among strikes—in the midst of voir dire—even though the final twelve jurors still have not been selected. E.g., People v. Crittenden, 885 P.2d 887, 903 (Cal. 1994) (noting a trial attorney’s fifty-juror comparison); Oral Argument at 16:10, United States v. Moore, 770 F.3d 809 (9th Cir. 2014) (No. 13-10464),
pk_id=0000013405 (on file with the Columbia Law Review) (recounting the juror-to-juror comparisons made at trial).
This is a fur­ther reason that comparative juror analysis, a critical tool of post-trial Batson, is not as significant at trial.

Granted, Batson challenges can be raised at the very end of jury se­lection or even on a motion for a new trial, and in such instances the identities of the seated jurors would be known. 181 See, e.g., Foster v. Chatman, 136 S. Ct. 1737, 1743 (2016) (mentioning a Batson challenge that was raised again in a new-trial motion); McCurdy v. Montgomery Cty., 240 F.3d 512, 522 (6th Cir. 2001) (same). The timing of peremptory challenges is also flexible, with some jurisdictions allowing “backstrikes,” for example. Bruce Hamilton, Note, Bias, Batson, and “Backstrikes”: Snyder v. Louisiana Through a Glass, Starkly, 70 La. L. Rev. 963, 964 n.8 (2010) (“A backstrike is a type of peremptory challenge used to strike jurors after they have been accepted onto the jury panel but before the panel has been sworn.”). But such delayed Batson challenges are unusual, 182 See Majid v. Portuondo, 428 F.3d 112, 127 (2d Cir. 2005) (“Batson hearings are typically conducted in association with, and at the same time as, jury selection.”). so the point remains: Comparative juror analy­sis, though a powerful tool for showing discriminatory intent, is largely an appellate-only 183 In this context, “appellate” also includes “postconviction” claims. method of analysis. This discrepancy widens the fis­sure between Batson claims at trial and Batson claims on appeal.

5. Batson Reconstruction Hearings. — Batson “reconstruction” hear­ings provide another example of how the trial and post-trial sides of Batson diverge. Reconstruction hearings have not been discussed in the academic literature, but they are of growing importance to Batson. 184 See, e.g., Madison v. Comm’r, 761 F.3d 1240, 1244 (11th Cir. 2014); Crittenden v. Ayers, 624 F.3d 943, 957–58 (9th Cir. 2010); Paulino v. Harrison, 542 F.3d 692, 695–96 (9th Cir. 2008); United States v. Ochoa-Vasquez, 428 F.3d 1015, 1045 n.40 (11th Cir. 2005); Green v. Travis, 414 F.3d 288, 300 (2d Cir. 2005); Pruitt v. McAdory, 337 F.3d 921, 929 (7th Cir. 2003). Here’s how they often come about: At trial, the defendant makes a Batson objection, and the trial judge finds no prima facie case of discrimina­tion. 185 See Pruitt, 337 F.3d at 927–28. As a result, the prosecutor is never asked to put his justifications for the strike on the record. 186 See id. at 928–29. Years later, an appellate court disagrees with the trial judge’s assessment and finds that the defendant did state a prima facie case of discrimination. 187 See id. at 928. The problem emerges that no one knows what the prosecutor’s justification was for the strike because she was never asked to offer one. Without knowing that justification, the Batson inquiry cannot move forward. For years, appellate courts would solve this problem by reviewing the record and guessing at what could have been the prosecutor’s reasons for the strike. The courts would then evaluate these imagined reasons at step three to see whether the prima facie case of discrimination was rebutted. 188 See Paulino, 542 F.3d at 700. Another option for the federal court is to grant a conditional writ and allow the state court to hold a hearing itself. Brian R. Means, Federal Habeas Manual § 13:13, Westlaw (database updated May 2017).

This practice of conjuring up reasons ran into conflict with the Supreme Court’s demand that the Batson inquiry seek out “actual an­swers” and “real reasons” for the strike—not just speculation. As the Court held in Johnson v. California, “[I]t does not matter that the prose­cutor might have had good reasons . . . [;] [w]hat matters is the real rea­son they were stricken.” 189 545 U.S. 162, 172 (2005) (alterations in original) (internal quotation marks omitted) (quoting Paulino v. Castro, 371 F.3d 1083, 1090 (9th Cir. 2004)). The reconstruction hearings were created as a way to produce these actual answers by reconstructing step two in a live evidentiary hearing. 190 Grossman, supra note 40, at 16 (critiquing the process of remanding cases for prosecutors to state the reason for a peremptory challenge because it is “nearly impossible for the prosecutor to genuinely remember the real reason” and the prosecutor may “try[] to invent more legitimate reasons after the fact”). To get the actual reasons for the strike, the prosecutor is called to the stand and asked about the strike or strikes. 191 Shirley v. Yates, 807 F.3d 1090, 1097–100 (9th Cir. 2015); Paulino, 542 F.3d at 701.

Simple as this sounds, the hearings have become very complicated, and they require a suspension of disbelief. Even though the hearings are supposed to reconstruct the trial proceedings, they actually diverge from trial practice in many significant ways. One divergence is that, unlike at trial, the prosecutor testifies as a sworn witness at the reconstruction hearing, and his testimony is subject to cross-examination. 192 This is more common than one would think. In 2005, the U.S. Supreme Court held that the California Supreme Court’s case law was demanding too much from litigants at step one. As a result, numerous California trial courts were improperly rejecting Batson claims at step one. See Johnson, 545 U.S. at 171. Another divergence involves the question of the prosecutor’s ability to remember her reasons for the strike. 193 See Green v. Travis, 414 F.3d 288, 293 (2d Cir. 2005) (“[T]he trial prosecutor . . . had little to no independent recollection of the characteristics or comments of any of the venirepersons at petitioner’s trial.”); Simmons v. Beyer, 44 F.3d 1160, 1167 (3d Cir. 1995) (“By the time of the reconstruction hearings, eleven years after Simmons’ trial, defense counsel did not recall how many African Americans were in the venire, how many were struck by the prosecution, or how many were seated as jurors. . . . The assistant prosecutor similarly did not remember . . . .”). But see Harris v. Haeberlin, 752 F.3d 1054, 1059 (6th Cir. 2014) (“Because circumstantial evidence may support a district court’s finding of intent, it is possible to reconstruct a meaningful Batson hearing even in the absence of a prosecutor’s independent recollection of his motives for making the challenged strike.”). At trial, there is no question that the prose­cutor is capable of remembering her true reasons. Even if she cannot recall the answer off the top of her head, she can consult her notes. On the contrary, in the reconstruction hearings, there are grave doubts about whether the prosecutor has any true memory of why she struck the juror. 194 See Johnson, 545 U.S. at 171 n.6 (stating a prosecutor’s refusal to justify making a strike after a court’s request for justification would support the inference of discrimination for the defendant in making a prima facie case); Yee v. Duncan, 463 F.3d 893, 899–900 (9th Cir. 2006) (finding criticism of the trial court’s consideration of circumstantial evi­dence misplaced because “demand[ing] trial courts to ignore evidence of the prosecutor’s ‘real’ intent when it is available” is contrary to the purpose of Batson to determine the reason for the prosecutor’s actions). Judge Stephen Reinhardt has provocatively suggested that prosecutors who destroy their notes should bear the burden if they cannot explain what they have done. See Shirley, 807 F.3d at 1106 n.16 (“Prosecutors who do not retain notes from voir dire run the risk that, as here, they will not be able to produce circumstantial evidence of their actual reasons for exercising a strike.”). Years or even decades may have passed since the strike. If the prosecutor has no notes of the strikes, she may have no way of refreshing her recollection. Even if the prosecutor has a memory of why she struck the juror, her memory may have been influenced by review of the voir dire transcript prior to the reconstruction hearing or by discussions the prosecutor had in preparation for the hearing. In other words, there are grave reasons to doubt that the answer provided by the prosecutor is the real reason, yet the reconstruction hearing treats the answer as if the prosecutor had said it in real time during the trial. 195 As the Johnson Court explained:
“The Batson framework is designed to produce actual answers to suspicions and inferences that discrimination may have infected the jury selection process. The inherent uncertainty present in inquiries of dis­criminatory purpose counsels against engaging in needless and imper­fect speculation when a direct answer can be obtained by asking a simple question.”
545 U.S. at 172; see also Harris v. Hardy, 680 F.3d 942, 955 (7th Cir. 2012) (“The ASA’s testimony at the Batson hearing suggests that he was not reciting his recollection of his reasons for the strike but rather was looking at the record and trying to come up with race-neutral reasons to justify the strike.”); Crittenden v. Ayers, 624 F.3d 943, 958 (9th Cir. 2010) (explaining reconstruction and providing examples of cir­cumstantial evidence the state may rely on); Dolphy v. Mantello, 552 F.3d 236, 240 (2d Cir. 2009) (“The district court may, in its discretion, hold a hearing to reconstruct the pro­secutor’s state of mind at the time of jury selection, and thereby determine whether the proffered race-neutral explanation for the striking of the African-American juror was pretextual . . . .”).

In many ways, the reconstruction hearing is a manifestation of Batson’s appellate ambitiousness. Batson’s strict adherence to its three-part framework, and its insistence on finding the actual reasons for the strikes, fuel the notion that the reasons for the strikes can be recon­structed so long after the trial simply by putting the prosecutor on the stand. In effect, the prosecutor is allowed to do what the appellate courts were prohibited from doing: hypothesize a reason for the strikes years after the fact. The reconstruction hearings provide yet another way in which post-trial Batson has expanded beyond the bounds of the trial doctrine.

C. Difference in Judicial Approach to Trial and Appeal

A further division between trial and post-trial Batson can be found in the growing importance of Batson procedural violations. While appellate judges must defer to the factual and credibility findings of trial judges, they have nonetheless found ways to identify procedural flaws in the way the Batson inquiry was executed. These procedural violations allow the appellate courts to engage with the Batson review in terms of legal ques­tions, rather than factual questions, so deference to the trial court is not required.

As commentators have noted, Batson claims involve many factual de­terminations on which appellate courts must defer to the trial courts. 196 See supra note 37 and accompanying text. Was the juror really slouched in her seat? Did the prosecutor appear genuine in explaining herself at step two? Did the questioning of the black jurors feel more aggressive than the questioning of white jurors? The conventional wisdom is that appellate judges must show great defer­ence on Batson claims because of all these credibility determinations. 197 See, e.g., United States v. Chinchilla, 874 F.2d 695, 697–98 (9th Cir. 1989) (“Since the district court’s determination of whether a peremptory challenge constituted pur­poseful discrimination turns on an evaluation of credibility of the prosecutor’s expla­nation, we should give those findings great deference.”). But appellate judges who see an injustice at trial and want to correct it have found ways to reframe the factual questions as legal and procedural ones. 198 See, e.g., Marder, Batson Revisited, supra note 37, at 1593 (“Of the eight cases that were not affirmed, five cases involved unusual procedures by the trial judge, one case was remanded for the prosecutor to give reasons, and two cases were before the Seventh Circuit on collateral review and the law had changed since the earlier rulings.”). Significantly, on these legal, procedural questions, appellate courts need not defer to the trial courts’ decisions. Essentially, appellate courts have found a way to avoid the deference they might otherwise owe to their trial colleagues by identifying errors in the way the trial court implemented Batson’s procedures. While they might not be able to sec­ond-guess trial courts’ credibility decisions, appellate courts can insist that the procedures for making these decisions be carried out correctly. After all, if these procedures were improperly carried out, the courts cannot be confident that the “real reasons” for the strikes were ever discovered.

This point about procedural errors is not a small one. Procedural vi­olations are particularly significant in Batson, as compared to in other doctrines, because the Supreme Court has specifically decreed the three-step framework courts must employ to produce the actual reasons for the strikes. 199 See supra note 195. The question of whether the prosecutor had discriminatory intent cannot be separated from the process used to ascertain those in­tentions. Examples of these procedural violations include cases in which the trial judge applied too demanding a threshold for the prima facie case at step one, 200 See Johnson v. Finn, 665 F.3d 1063, 1068 (9th Cir. 2011) (concluding that the trial court improperly conflated the “strong likelihood” and “reasonable inference” standards). combined step two and step three into one, 201 See Purkett v. Elem, 514 U.S. 765, 768 (1995); United States v. McAllister, 693 F.3d 572, 581 (6th Cir. 2012); United States v. Rutledge, 648 F.3d 555, 559–60 (7th Cir. 2011); McCurdy v. Montgomery Cty., 240 F.3d 512, 521–22 (6th Cir. 2001); Addison v. State, 962 N.E.2d 1202, 1210 (Ind. 2012). offered her own suppositions about the reasons for the strikes before asking the prosecutor, 202 E.g., Currie v. McDowell, 825 F.3d 603, 611 (9th Cir. 2016) (finding it “troubling that [the prosecutor’s] explanations for the strike were largely adopted from reasons the trial judge had already suggested[] during his discussion of Batson step one”); see also Shirley v. Yates, 807 F.3d 1090, 1107 (9th Cir. 2015) (faulting the district judge for concluding, without seeking an actual answer, that it “could have been reasonable” for the prosecutor to strike a juror for lack of education). prevented relevant evidence from being presented, 203 E.g., Coombs v. Diguglielmo, 616 F.3d 255, 263 (3d Cir. 2010) (finding the trial court “effectively omitted the third step of the Batson inquiry by unreasonably limiting the defendant’s opportunity to prove that the prosecutor’s proffered reasons for striking Black jurors were pretextual”); Hardcastle v. Horn, 368 F.3d 246, 251 (3d Cir. 2004) (noting the trial court denied the prosecutor’s request for permission to state her reasons for chal­lenged strikes on the record). and misallocated the burden of proof. 204 E.g., United States v. Kimbrel, 532 F.3d 461, 467 (6th Cir. 2008). Many Batson victories—including remands for further fact-finding—involve these procedural violations. Not surprisingly, these claims about Batson procedural violations are largely confined to appeal. This is not to say that trial litigants are prohib­ited from complaining about some procedural violation the court has committed, but such complaints at trial would require the trial judge to reverse herself, so they are more feasible to raise on appeal.

The development of “procedural” Batson violations is also related to the intense postconviction pressures applied by AEDPA, 205 Pub. L. No. 104-132, 110 Stat. 1214 (1996) (codified as amended in scattered titles of the U.S.C.). a significant impediment to federal habeas relief. Passed by Congress in 1996, AEDPA prohibits federal courts from granting habeas relief to reverse a state con­viction unless the state court’s ruling on the legality of that conviction is “contrary to . . . clearly established Federal law, as determined by the Supreme Court of the United States.” 206 28 U.S.C. § 2254(d)(1) (2012). The term “clearly established Federal law” has come to mean the holdings of Supreme Court cases—not dicta, not logical extensions, not anything else. 207 Williams v. Taylor, 529 U.S. 362, 412 (2000) (O’Connor, J.); see also Howes v. Fields, 565 U.S. 499, 505 (2012). As a result of AEDPA, federal courts cannot recognize a constitutional violation in any state prisoner’s case unless the state courts’ decisions denying relief were so egregiously wrong as to have transgressed a direct holding of the Supreme Court.

Although AEDPA’s “clearly established” limitation is a major hurdle to federal habeas relief, this limitation does not apply equally to all types of habeas claims. The doctrines it hits hardest are those that rely on standards like “reasonableness” or on interest balancing. 208 In Yarborough v. Alvarado, the Court explained that:
“[T]he range of reasonable judgment can depend in part on the nature of the relevant rule. If a legal rule is specific, the range may be narrow. Applications of the rule may be plainly correct or incorrect. Other rules are more general, and their meaning must emerge in application over the course of time. Applying a general standard to a specific case can de­mand a substantial element of judgment. As a result, evaluating whether a rule application was unreasonable requires considering the rule’s spec­ificity. The more general the rule, the more leeway courts have in reach­ing outcomes in case-by-case determinations.”
541 U.S. 652, 664 (2004).
If the claim asks the state court to appraise the gestalt of some issue, it is difficult to show that the state court got it objectively and egregiously wrong. 209 For instance, the constitutional prohibition on shackling a defendant in front of the jury requires the courts to make a judgment call about whether “an essential state in­terest” was served by the shackling. Holbrook v. Flynn, 475 U.S. 560, 568–69 (1986). For an example in which the Court applied the AEDPA standard to a “general” test, see Yarborough, 541 U.S. at 664–65. On the other hand, with doctrines that are more rule-like—and especially with those that have multiple, clearly defined steps—it is easier to show that the state court misapplied Supreme Court law. Batson falls into this latter category with its mandatory, three-step burden-shifting framework. If the state court failed to follow any of Batson’s three steps, the federal court can point to the transgression of “clearly established Federal law” as grounds for intervening. 210 28 U.S.C. § 2254(d)(1). Moreover, Batson has the added benefit that the three-part framework has been a core part of the doctrine from the beginning, so there is no doubt that the law was clearly established at the time of the state court decision—still another requirement of AEDPA. 211 Williams v. Runnels, 432 F.3d 1102, 1105 n.5 (9th Cir. 2006) (“[T]he Supreme Court clearly indicates in Johnson that it is clarifying Batson, not making new law.” (citing Johnson v. California, 545 U.S. 162, 168–70 (2005))).

Consider two caveats about these procedural Batson violations. First, there are many Batson claims in which the appellate courts do end up deferring to the decision of the trial judge without searching out some procedural violation. 212 See, e.g., Batson v. Kentucky, 476 U.S. 79, 98 n.21 (1986) (explaining that the trial judge’s factual findings regarding discrimination should ordinarily be given “great deference”); United States v. Barnette, 644 F.3d 192, 209, 213–14 (4th Cir. 2011) (affirming the district court decision, even though the lower court conducted an in cam­era review of prosecutors’ notes, which is disfavored in Batson proceedings, and committed harmless error in refusing to provide defendant with the unredacted jury questionnaires); supra note 37 and accompanying text. In discussing procedural violations, this Article describes how appellate judges who want to intervene have found ways to do so in spite of the deference they might owe to the trial courts. This is not to suggest that all or most appellate judges take this route. Second, this argument does not imply that Batson claims are somehow immune from the limita­tions created by AEDPA. Many Batson claims are felled by the “clearly established” bar, especially when the state court decided an issue for which there is not a Supreme Court case directly on point. 213 See, e.g., McDaniels v. Kirkland, 813 F.3d 770, 775–76 (9th Cir. 2015) (en banc) (holding that, at time of state court decision, there was no clearly established federal law requiring sua sponte comparative juror analysis). This Article simply argues that, holding AEDPA’s limitations constant across all habe­as doctrines, there are some advantages Batson enjoys be­cause of its three-step, mechanical test. These advantages boost the use­fulness of Batson claims relative to other habeas doctrines and raise the stature of post-trial Batson relative to trial Batson.

III. Implications

A number of implications flow from Batson’s special strength in ap­pellate and habeas proceedings. These implications should affect the way litigants plead Batson claims, judges decide those claims, and academic commentators talk about Batson.

A. Batson as a Multipurpose Vehicle to Combat Discrimination

The point of discussing Batson relative to other antidiscrimination doctrines is not just to burnish the doctrine’s reputation. Rather, the goal is to suggest how Batson can be deployed as a multipurpose antidis­crimination doctrine capable of protecting against an array of racist acts, even those outside jury selection. For example, in Foster v. Chatman, the prosecutor employed overtly racist messaging in his closing argument. As the cert petition noted, “The prosecutor . . . argued that the jury should impose a death sentence to ‘deter other people out there in the pro­jects.’” 214 Petition for Writ of Certiorari, supra note 151, at i. Such a racist comment could be challenged on its own as a form of prosecutorial misconduct, but the showing needed for such a claim is difficult, and the misconduct claim is subject to harmless error review. 215 See Means, Postconviction Remedies, supra note 56, § 46:18 (“[E]ven if a pro­secutor’s comments are inappropriate, they alone do not justify the reversal of a criminal conviction obtained in an otherwise fair proceeding. Rather, the remarks must be ex­amined within the context of the trial to determine whether the prosecutor’s behavior amounted to prejudicial error.” (footnote omitted)). Batson, however, provides a more straightforward alternative that is not subject to harmless error review. In other words, the racist statement can be reconceptualized as proof of the prosecutor’s discrimi­natory intent in jury selection.

Snyder v. Louisiana revealed a similar dynamic. The prosecutor re­peatedly compared the black defendant to O.J. Simpson, despite de­mands by the defendant and the trial court that he not make such a comparison. 216 Brief of Petitioner at 2–3, Snyder v. Louisiana, 552 U.S. 472 (2008) (No. 06-10119), 2007 WL 2605447. When the case reached the Supreme Court, several Justices seized on the O.J. Simpson comparison as evidence of racism’s impact on the trial. 217 Transcript of Oral Argument at *37, Snyder, 552 U.S. 472 (No. 06-10119), 2007 WL 4252698. Although the Court’s decision in Snyder did not mention O.J. Simpson, the racist prosecutorial statements seemed to be on the Justices’ minds as they considered the Batson claim. 218 See Snyder, 552 U.S. at 474–86. And with good reason: The prosecutor’s statements to the jury reveal his focus on race at the time of trial and are thus relevant to understanding the moti­vation behind his peremptory strikes.

The potential synergy between Batson and other antidiscrimination doctrines is not limited to the prosecutorial-misconduct examples above. Claims that the prosecutor’s charging and sentencing decisions were ra­cially motivated could also be reconceptualized as evidence of Batson vio­lations, assuming of course that there were already some red flags to raise questions about the prosecutor’s use of peremptory challenges. As noted earlier, the difficulties are immense when it comes to pleading discrimi­natory charging or discriminatory sentencing. But any of the evidence that would support such a discriminatory-charging or -sentencing claim could also show the prosecutor’s racist intent at jury selection. For exam­ple, in a case recently before the Supreme Court, the prosecutor revealed in postconviction discovery that one of the reasons he sought the death penalty against the defendant was that the defendant was not a citizen. 219 Petition for a Writ of Certiorari at 17, Ayestas v. Davis, No. 16-6795, 2016 WL 8652345 (Nov. 7, 2016). This national-origin discrimination could support a Batson claim, and putting this evidence of discrimination inside the Batson challenge could give the claim greater legs because of Batson’s appellate virtues.

This is not to say that a fake Batson label should be slapped onto other antidiscrimination claims. Rather, the goal is to illustrate the true capaciousness of Batson. The doctrine can accommodate any type of evi­dence that speaks to the prosecutor’s discriminatory intent, even evi­dence that comes to light after the jury has been selected or after the trial has concluded. By acknowledging the breadth of Batson, litigants may find that evidence of discrimination—which could be pleaded as a stand-alone claim of prosecutorial misconduct of some other sort—would also, or better, fit within the Batson framework.

This more expansive view of Batson is also relevant to the oft-suggested “solution” to Batson’s shortcomings: the elimination of all per­emptory strikes. Dating back to Justice Marshall’s concurrence in Batson, commentators have suggested that eliminating peremptory strikes altogether is the only way to prevent them from being used in a discrim­inatory manner. 220 See supra notes 12–13 and accompanying text. Practitioners, even those concerned about ra­cial discrimination, often push back by pointing out that peremptory challenges may be the only way defendants can free themselves of jurors who harbor racist (or antidefendant) biases. 221 See supra section I.C. To this conversation, this Article adds the consideration that eliminating peremptory strikes altogether would also eliminate any value Batson may have as a multipurpose vehicle for fighting discrimination.

To the extent that Batson sweeps in—or could sweep in—claims of discrimination from anywhere in the trial, it would be a loss to defen­dants and the justice system for Batson to disappear along with the elim­ination of peremptory strikes. This is not to say that Batson’s utility as a multipurpose antidiscrimination vehicle justifies the continued use of peremptory strikes. There are many factors to consider in that debate, including the likelihood that eliminating peremptory strikes would result in more aggressive racial discrimination in other parts of the jury-selec­tion system, such as in the jury-summons process or the use of for-cause challenges—doctrines even less equipped than Batson to resist it. But the ongoing debate should consider the harm that would accrue from losing Batson’s ability to fight racial discrimination wherever it rears up in the trial.

B. Appellate Batson’s Symbolism, Rhetoric, and Power

In oral argument and in written decisions, appellate judges have taken on a striking tone of moral outrage toward Batson violations. 222 See infra notes 226–231 and accompanying text. Batson has come to be seen not simply as a doctrine that protects black defendants from discrimination but rather as a doctrine that guarantees the bedrock fairness of the judicial system for all litigants. Batson’s de­velopment into a guarantor of civic and democratic virtue has super­charged the rhetoric and symbolism of the doctrine, but this transfor­mation has occurred in ways that are far more accessible to appellate judges than their trial colleagues, thus furthering the divide between Batson’s meaning during trial and its meaning after. Understanding the nature of this schism is essential to appreciating Batson’s potential as an appellate doctrine.

As the Batson case law has developed over the years, it has trans­formed from a doctrine that protected black defendants from the elimi­nation of black jurors into a doctrine that now protects all parties—defendant, prosecutor, and civil litigants—against the removal of jurors for any of an expanding list of characteristics. 223 See Karlan, Batson, supra note 57, at 406–07. Batson is nearly as much about democracy and political community as it is about race. The fact that Batson speaks in terms of democracy and the justice system’s integrity makes it more politically powerful in taking on issues of racial discrimi­nation. Batson violations have become assaults on the judiciary itself because they undermine the fairness of the jury verdicts on which everything else relies. 224 See, e.g., People v. Gutierrez, 395 P.3d 186, 190 (Cal. 2017) (“Taints of discrim­inatory bias in jury selection—actual or perceived—erode confidence in the adjudicative process, undermining the public’s trust in courts. . . . The error is structural, damaging the integrity of the tribunal itself.” (citations omitted)).

The outrage that Batson violations provoke goes beyond those char­acteristics formally protected by the doctrine: race, gender, and national origin. That is because there is something unseemly about manipulating the jury—a body that is supposed to be representative of the popu­lation—into a body that over- and underrepresents whole groups of people. 225 Powers v. Ohio, 499 U.S. 400, 402 (1991) (“Jury service is an exercise of respon­sible citizenship by all members of the community, including those who otherwise might not have the opportunity to contribute to our civic life.”); United States v. Moore, 651 F.3d 30, 104 (D.C. Cir. 2011) (noting how Batson rights “extend ‘to those citizens who desire to participate “in the administration of the law, as jurors,” as well as to . . . eradicating discrimination from our civic institutions [that] suffers whenever an individual is ex­cluded . . . on account of his race’ or other suspect characteristic” (quoting Johnson v. California, 545 U.S. 162, 172 (2005))), aff’d in part sub nom. Smith v. United States, 568 U.S. 106 (2013). How many tweaks can be made to this civic institution before it stops being representative at all? Such concerns go beyond Batson and equal protection, yet they are part of what judges must wrestle with in deciding Batson claims. Judges face these questions because prosecutors’ justifications for strikes often violate basic notions of how the justice sys­tem ought to work, even if they do not violate Batson itself. At one oral argument, the appellate panel appeared incensed by the trial prosecu­tor’s proffered reason that he struck a juror because she disagreed that those brought to trial are probably guilty: 226 Oral Argument at 16:00, Williams v. Pliler, 616 F. App’x 864 (9th Cir. 2015) (No. 14-16393), (on file with the Columbia Law Review).

Judge Michael Daly Hawkins: Do you think it’s appropriate for a prosecutor, who’s prosecuting a criminal case, to, in effect, take the position that he prefers jurors who do not believe in the presumption of innocence?

Deputy Attorney General: That particular question was not actually asked, your honor.

Judge Hawkins: Well what was asked—he stated his reason—one of his reasons was her response to the question, ‘If the prosecution brings someone to trial that person is probably guilty,’ and she checked, ‘Disagree strongly.’ I would hope every American citi­zen would check the box that way. 227 Id.; see also Williams, 616 F. App’x at 870 (finding a Batson violation when a juror was struck because of her “agreement with the presumption of innocence that is basic to our criminal justice system, not any specific bias against the prosecution”); United States v. Mitchell, 502 F.3d 931, 1005 (9th Cir. 2007) (Reinhardt, J., dissenting) (“A strike based solely on a juror’s participation . . . in a prior acquittal . . . threatens the institution of the jury . . . [but] nonetheless survives the de minimis burden placed on the prosecution at the second [step] of the Batson analysis.”); Moran v. Clarke, 443 F.3d 646, 661 (8th Cir. 2006) (Beam, J., dissenting) (“Justice Thurgood Marshall in his eloquent Batson concurrence stated ‘Our criminal justice system requires not only freedom from any bias against the accused, but also from any prejudice against his prosecution. Between him and the state the scales are to be evenly held.’” (quoting Batson v. Kentucky, 476 U.S. 79, 107 (1986) (Marshall, J., dissenting))); Oral Argument at 11:06, Allen v. Benedetti, 629 Fed. App’x 814 (9th Cir. 2015) (No. 14-16671),
id=0000014900 (on file with the Columbia Law Review) (including the deputy attorney gen­eral’s observation of the trial judge’s comment that “I’ve never heard somebody striking somebody as a teacher”).

Ninth Circuit Judge Stephen Reinhardt has taken issue with striking jurors because they did not go to college, especially because of the disparate impact along racial lines: “Does it bother you that a prosecutor would say people who didn’t go to college shouldn’t be on juries?” he asked a prosecutor in one argument. 228 Oral Argument at 28:11, Shirley v. Yates, 807 F.3d 1090 (9th Cir. 2015) (No. 13-16273), (on file with the Columbia Law Review). But see Grossman, supra note 40, at 9 (discussing “[v]iews on the legal system” as grounds for satisfactory step two answers). A thought-provoking, if obscure, example of Batson’s penumbra comes from a county court in New York, where a reviewing court found a “Batson-like” violation because all li­censed hun­ters had been struck from the jury. 229 People v. Robar, 29 Misc. 3d 693, 698–99 (N.Y. Cty. Ct. 2010) (“This court is not yet ready to find . . . that licensed hunters are a . . . protected class distinct under Batson/
Luciano, but this court . . . [finds] they are a class governed by the Civil Rights Clause, which guarantees the right . . . for a defendant to be tried by a jury of his peers.”).
Striking someone for believing in the presumption of innocence or for lacking a college edu­cation is not a violation of the Equal Protection Clause, but it is nonetheless offensive. 230 In the context of the oral arguments described above, the judges’ concerns centered around the pretextual nature of such excuses. See supra notes 227–229 and accompanying text. At a time of enormous divisions and partisan­ship in American society, it is troubling to think about allowing the bed­rock civic institution of the jury to be further splintered. One wonders, for example, whether the jury could function as a civic institution if a prosecutor set out to cleanse it of all registered Democrats or everyone receiving public assistance. The civic and philosophical issues raised by Batson are inherently tied up in the way the doctrine is litigated, and they have created an exoskeleton of symbolism on top of Batson’s equal pro­tection bones.

Significantly, for purposes of this Article, Batson’s symbolism pro­vides appellate judges with an unrivaled opportunity to declaim on the structural issues facing the court system. And these judges have the op­portunity to make such pronouncements in the context of a structural error claim—an unusual context that means they are actually empowered to grant relief to the defendant. Their pronouncements on race are given all that much more weight because they are accompanied by the act of throwing out a conviction. And pronounce they have. Appellate deci­sions about Batson warn of nothing less than the demise of “multiracial de­mocracy” and the “proud” idea of the “melting pot” if Batson’s rules are not enforced. 231 Edmonson v. Leesville Concrete Co., 500 U.S. 614, 630–31 (1991) (“If our society is to continue to progress as a multiracial democracy, it must recognize that the automatic invocation of race stereotypes retards that progress and causes continued hurt and injury.”); People v. Garcia, 92 Cal. Rptr. 2d 339, 346 (Ct. App. 2000) (“Our jury venires daily include Cubans named O’Rourke, Indonesians named Opdyke, and Anglos named Gomes. . . . The country is a melting pot—and proud of it—and . . . the great folly of ste­reotyping is that nowhere on earth have race and ethnicity become harder to determine than they are here.”); see also State v. Saintcalle, 309 P.3d 326, 329 (Wash. 2013) (en banc) (“This appeal raises important questions about race discrimination in our criminal justice system.”), abrogated by City of Seattle v. Erickson, No. 93408-8, 2017 WL 2876250 (Wash. July 6, 2017). Batson is uniquely suited to these proclamations about civics and fairness because it is seen as a protector of the whole system, not just the fairness of the particular trial. 232 See J.E.B. v. Alabama, 511 U.S. 127, 141–42 (1994) (“All [potential jurors] . . . have the right not to be excluded summarily because of discriminatory and stereotypical presumptions . . . . Striking individual jurors on the assumption that they hold particular views simply because of their gender is ‘practically a brand upon them, affixed by the law, an assertion of their inferiority.’” (footnote omitted) (quoting Strauder v. West Virginia, 100 U.S. 303, 308 (1880))); SmithKline Beecham Corp. v. Abbott Labs., 740 F.3d 471, 485 (9th Cir. 2014) (“Strikes exercised on the basis of sexual orientation continue this deplor­able tradition of treating gays and lesbians as undeserving of participation in our nation’s most cherished rites and rituals. They [indicate] . . . our judicial system treats gays and lesbians differently[,] . . . depriv[ing] individuals of the opportunity to participate in per­fecting democracy . . . .”).

This system-wide perspective is an aspect of Batson that appellate, more than trial, judges have access to. Appellate decisions are reasoned and crafted in a manner that is fundamentally different from the deci­sions of trial courts, which helps account for appellate Batson’s unusual power. In general, appellate decisions are longer, more philosophical, and more far-reaching than trial court decisions. The existence of con­currences and dissents allows judges to stray further from the specific facts of the case and into the broader principles that govern—or ought to govern—the case. Moreover, appellate decisions are binding on future cases and, as a result, are capable of bringing forth systemic changes that go far beyond the individual case.

Trial judges, on the other hand, are generally not in a position to engage with Batson on these high-flying terms. Trial courts do not have the time or the platform to pronounce on Batson’s implications for civics and race; their Batson decisions are typically oral rulings from the bench, and even when they are written, they are unlikely to be published in any official reporter. 233 Karlan, Race, Rights, and Remedies, supra note 51, at 2022 (noting that “virtually no Batson decisions at the trial court level are reported”). But see McKinney v. Artuz, 326 F.3d 87, 92 (2d Cir. 2003) (quoting the trial judge’s suggestion that “it’s high time the Court of Appeals follows the wisdom of Thurgood Marshal[l] and decides all peremptory challenges are intrinsically prejudicial and should be eliminated as an archaic tool for more racist times” (alteration in original) (internal quotation marks omitted)). Nor are they precedential. These factors mean trial judges have less incentive to make the big pronouncements that their appellate colleagues do.

Granted, of course, there are many appellate Batson decisions that are quite mundane, devoid of anything approaching political or moral philosophy. 234 See, e.g., United States v. Platt, 608 F. App’x 22, 25 (2d Cir. 2015); United States v. Carter, 483 F. App’x 70, 74 (6th Cir. 2012). And, at the same time, there are some Batson trial decisions that declaim on big issues. 235 See State v. Evans, 998 P.2d 373, 377 (Wash. Ct. App. 2000) (quoting the trial judge’s language about “protect[ing] the right of jurors to participate in the civic process and . . . ensur[ing] that our justice system is free from any taint of racial bias”). But the point stands that there is an underlying structural difference in the way Batson claims present them­selves to appellate judges compared to trial judges.

Appellate courts may also treat Batson claims differently from their trial colleagues because of their distance in space and time from the Batson violations. At trial, as noted earlier, the Batson inquiry is intensely personal. 236 See supra note 35 and accompanying text. A defense attorney who wants to challenge a strike must ar­gue that it was motivated by discriminatory intent, and the trial judge must decide the issue on the spot. On appeal, however, the social awk­wardness of the Batson inquiry is lessened: The litigation is done mostly on paper rather than in person; years or even decades may have passed so the trial prosecutor may no longer be part of the case; and even if the prosecutor is part of the case, the appellate judges are less likely to be familiar with any of the lawyers. 237 There are a variety of reasons that the trial attorney may not be the one handling the appeal. It could be because the trial attorney has retired, or because there is a division of labor in her office that assigns appeals to a special unit, or because the appeal is han­dled by an entirely different agency, such as the state attorney general. See supra note 139 and accompanying text. The appellate deliberative process is slow and secluded, with groups of judges deciding cases behind closed doors rather than a single judge making decisions in the heat of the moment. That is the difference in space.

The difference in time is significant, too, especially in old capital cases that slowly move through the appellate and habeas pipelines. Judges today may be more sensitive to, and aware of, racial discrimination than judges were ten or twenty years ago. (At least, one hopes.) As a re­sult, an appellate court in 2018 considering whether Batson was violated may be applying contemporary understandings of race to a strike that took place in the late 1980s or early 1990s. 238 See, e.g., Foster v. Chatman, 136 S. Ct. 1737, 1747 (2016) (reviewing a Batson claim twenty-nine years from the time of trial); Madison v. Comm’r, 761 F.3d 1240, 1244 (11th Cir. 2014) (reviewing a Batson claim nineteen years from the time of voir dire). Significantly, Batson asks judges to determine whether the prosecutor’s strike was motivated by race, full stop. It is not a question of whether the decision would have been considered racist at the time. That means appellate judges apply their own contemporary understandings of racism to cases from a differ­ent era. 239 People v. Gutierrez, 395 P.3d 186, 207 (Cal. 2017) (Liu, J., concurring) (“Foster and Miller-El involved trials that took place over 30 years ago. I would surmise and hope, though I do not know for sure, that such brazenly unlawful practices are rare today.” (citations omitted)); see also Batson v. Kentucky, 476 U.S. 79, 87 (1985) (noting how con­fidence in the justice system suffers when there is discriminatory jury selection). To the extent that society’s understanding of racism has grown more sophisticated over time, an appellate judge looking back at the trial from today’s vantage point might be more inclined to see racial intent than the trial judge was at the time.

To be sure, there are countervailing forces that make it is easier for trial judges, rather than appellate judges, to find Batson violations, not least of which is the trial judge’s firsthand view of what transpired in court. Batson, after all, affords “great deference” to the trial judge’s cred­ibility determinations. 240 Batson, 476 U.S. at 98 n.21. This does not diminish the argument about Batson’s unique post-trial virtues. The point is not the relative likelihood of winning a Batson claim at trial versus on appeal. Rather, the point is that post-trial Batson provides unappreciated opportunities that trial Batson does not.

C. The Gap Between the Trial and Appellate Doctrine Is a Feature, Not a Flaw, of Batson.

If there is a divergence between trial and post-trial Batson, the temp­tation is to find a way to reconcile the two sides of the doctrine. The trouble is that it is not possible to reconcile the two halves of the doctrine without undermining key tenets of Batson. The following section de­scribes why the gap between trial and post-trial Batson will persist and why it is preferable to imposing an artificial symmetry on the two halves of Batson.

1. Evidence: Leveling Up and Leveling Down. — First, consider the gap between the evidence that can be used at trial and the evidence that can be used later on. In other areas of the law, the type of evidence and ar­guments that can be used post-trial are dictated by what was available at trial. Batson postconviction litigation could be made like Brady v. Maryland or Strickland v. Washington, which limit new evidence to material that could have been obtained at trial, if only the prosecutor had not failed to disclose it or defense counsel had not failed to uncover it. 241 See supra notes 122–123 and accompanying text; see also Kyles v. Whitley, 514 U.S. 419, 419 (1995) (“[C]onstitutional error results . . . if there is a ‘reasonable prob­ability’ that, had the evidence been disclosed to the defense, the result of the proceeding would have been different.”). There is a certain logic to this limitation, because it avoids having a claim at trial that cannot be fully proven until the postconviction proceed­ings. 242 This leveling down would be in line with concerns that Justice Clarence Thomas has raised in several Batson decisions: “We have no business overturning a conviction, years after the fact and after extensive intervening litigation, based on arguments not presented to the courts below.” Snyder v. Louisiana, 552 U.S. 472, 489 (2008) (Thomas, J., dissenting); see also Miller-El v. Dretke, 545 U.S. 231, 284 (2005) (Thomas, J., dissenting) (“Without the questionnaires never submitted to the trial court, Miller-El comes nowhere near establishing that race motivated any disparate questioning or treatment, which is pre­cisely why the majority must strain to include the questionnaires within the state-court re­cord.”). Justice Thomas was speaking about the deference that federal courts must show to state courts under AEDPA, but the same principle would apply even if the appellate and trial courts were both state or both federal. However, the trouble with a Brady– or Strickland-style rule for Batson evidence is that it would bar late-arising evidence that may speak directly to the prosecutor’s intent but was not in existence at the time of trial or may have been legally privileged. Examples include racist state­ments made by the prosecutor in or outside court, Batson violations charged to the prosecutor in cases that were decided after the defend­ant’s trial, and jury-selection notes that were deemed privileged from dis­closure at trial but became available later on. 243 See supra section II.B.1–.3. Leveling down might also prevent the use of comparative juror analysis on appeal, if some important element of it could not have been carried out at trial—even though com­parative juror analysis is effectively required of Batson liti­gants on appeal.

This leveling down of post-trial Batson would force courts to blind themselves to evidence of the prosecutor’s discriminatory intent, and this blindness would directly conflict with Batson’s goal of determining the actual reason behind the strike. As the Supreme Court explained in Foster v. Chatman: “We have ‘made it clear that in considering a Batson objection, or in reviewing a ruling claimed to be Batson error, all of the circumstances that bear upon the issue of racial animosity must be con­sulted.’” 244 Foster v. Chatman, 136 S. Ct. 1737, 1748 (2016) (quoting Snyder, 552 U.S. at 478); see also Johnson v. California, 545 U.S. 162, 172 (2005) (emphasizing that “[w]hat mat­ters is the real reason they were stricken” (internal quotation marks omitted) (quoting Paulino v. Castro, 371 F.3d 1083, 1090 (9th Cir. 2004))). It bears repeating that the question Batson poses is not whether the trial court made the right decision in light of the evidence availa­ble to it at the time of trial. Rather, the question Batson poses is whether the prosecutor’s strike was motivated by race—that is the importance of the Supreme Court’s insistence on the search for “actual answers.” 245 Johnson, 545 U.S. at 171; see also People v. Chism, 324 P.3d 183, 244 (Cal. 2014) (Liu, J., concurring in part and dissenting in part) (“[A]ppellate review of a Batson ruling is not merely an exercise in evaluating the trial court’s performance based on arguments put forth by the parties. Instead[,] . . . Snyder and Miller-El require appellate courts to con­sider all relevant circumstances . . . in determining whether the strike of a particular juror was improperly motivated.”). A lim­itation on late-arising evidence would flout that mandate.

This is not to say that Batson ushers all relevant evidence into state and federal habeas proceedings, much less direct appeals. Each of these procedural postures—direct appeal, state habeas, and federal habeas—has its own procedural rules governing what new evidence can be brought into the case, and these apply to all claims, including Batson. 246 The most prominent of these in federal habeas review of state convictions is Cullen v. Pinholster, which limits a federal court’s review of the conviction to the evidence that was before the state courts as they evaluated the conviction. 563 U.S. 170, 181–82 (2011). State habeas proceedings are more permissive in allowing for an expansion of the record, but they nonetheless have their own limitations on what new evidence can come into the case. See Stephanie Roberts Hartung, Missing the Forest for the Trees: Federal Habeas Corpus and the Piecemeal Problem in Actual Innocence Cases, 10 Stan. J. C.R. & C.L. 55, 63 (2014). In a direct appeal, there is a strong presumption against introducing any evidence not in the record. Id. at 59–60. The point is that holding these evidentiary limitations constant, Batson has a broader sweep than other record-expanding doctrines because it is not focused on whether the evidence could have been available at trial. What makes evidence relevant to Batson is its ability to speak to the intent of the prosecutor. Perhaps ironically, it is this focus on the prosecutor’s in­tent—a much-bemoaned aspect of the Batson inquiry—that sweeps so much evidence into post-trial Batson claims.

If it would not work to level down Batson’s post-trial use of evidence, what about leveling up the trial doctrine’s use of it? One could decree that trial litigants must have access to anything that will (or could) even­tually be part of the habeas case. Perhaps that decree would succeed in giving trial attorneys access to materials such as jury-selection notes or internal memoranda, which exist at the time of trial but are often con­sidered legally privileged. But aside from this category of materials, level­ing up trial Batson (obviously) could not produce evidence at trial that has not yet come into existence—for example, the prosecutor’s racist statements in subsequent cases or examples of Batson violations from cases that come after the defendant’s trial. So, leveling up could not on its own close the gap.

2. Remedy: Leveling Up and Leveling Down. — Nor is there much to be done to reconcile the value of the trial and post-trial remedies. That is true because going back to the beginning of jury selection is simply more significant on appeal and habeas than at trial. The only way to make a trial Batson win as valuable as a post-trial win is to alter the remedy, and that would involve an overhaul of settled Batson law. The trial remedy could be leveled up, it is true. For example, Professor Charles Ogletree has proposed that Batson violations could result in dismissal of the charg­es with prejudice against refiling them. 247 Ogletree, supra note 16, at 1117. This would certainly raise the significance of Batson’s trial remedy, likely beyond even the appellate and post-trial remedies. It would mean that once a Batson violation was de­clared, the defendant could never be prosecuted for that crime. But such an innovation would require a sea change in the doctrine and would probably not be politically possible.

Leveling down the appellate remedy would also close the gap. Such leveling down could be accomplished by declaring that Batson violations are not structural error and, thus, must have prejudiced the defendant in order to justify a reversal. Forcing a defendant to show prejudice would greatly reduce the significance of an appellate Batson win, but it would mire the doctrine in the impossible question of demonstrating that the presence or absence of any particular juror affected the outcome. In­deed, one of the reasons Batson must be a structural error is that there is no feasible way to detect or calculate the prejudice that accrues from the taint to the jury. 248 See Review Proceedings, 37 Geo. L.J. Ann. Rev. Crim. Proc. 805, 849 n.2554 (2008). As noted earlier, a few state court opinions have required showings of prejudice. See Karlan, Race, Rights, and Remedies, supra note 51, at 2019 n.87. Batson prejudice also arises in the context of claims that assert counsel was constitutionally ineffective for failing to raise a Batson objection at trial. Strickland v. Washington requires a showing of deficient performance of counsel and a resulting prejudice to the defendant. 466 U.S. 668, 687 (1984). Some courts have required a showing not only that the Batson objection would have succeeded but also that the defendant had a reasonable probability of a different outcome at the retrial. Other courts are willing to find the Strickland prejudice require­ment satisfied by the grant of a new trial. See Amy Knight Burns, Note, Insurmountable Obstacles: Structural Errors, Procedural Default, and Ineffective Assistance, 64 Stan. L. Rev. 727, 747–48 (2012). Because automatic reversal is a foundational as­pect of Batson, it seems exceedingly unlikely it could be eliminated in some effort to reconcile trial and post-trial Batson.

3. Broader Reflections on the Trial–Post-Trial Gaps. — The discomfort about allowing appellate Batson to outpace trial Batson is understandable. It upends the typical logic of the appellate system to have a claim that cannot be fully decided by the trial judge and, instead, ripens on appeal or habeas. Judges and prosecutors may naturally worry that even if the Batson objections are resolved at trial, the objections could arise with new force on appeal, endangering hard-earned convictions and injecting an air of unpredictability into every conviction. But the U.S. Supreme Court requires comparative juror analysis, even for the first time on appeal. As has been emphasized throughout, Batson is an absolutist doctrine fo­cused on the question of the prosecutor’s intent, not the question of how his intent would have been perceived at trial. If comparative juror analy­sis or some other late-arising evidentiary material speaks to that question, it is not for Batson to ignore that evidence just because it was not present­ed at trial. 249 As Justice Liu wrote in Chism, 324 P.3d at 244 (Liu, J., concurring in part and dissenting in part):
[A]ppellate review of a Batson ruling is not merely an exercise in evaluating the trial court’s performance based on arguments put forth by the parties. . . . [A]n appellate court is not precluded from con­sidering, and indeed must consider, grounds that the defendant did not bring to the trial court’s attention. . . . [T]here is no reason why jurors seated after a trial court’s ruling may be considered only if the defen­dant makes a renewed objection.”
A related concern is that these new Batson claims impose costs on the appellate system, even when they are not successful, because they still have to be adjudicated. But that is a complaint about postconviction litigation more generally, not about Batson.

Nor is this late-arising potential bad from a policy perspective. A prosecutor who is worried that today’s conviction could be undermined by tomorrow’s Batson appeal is not powerless. The surest way to avoid a Batson challenge in the first place is to accept the jury as-is, without exer­cising peremptory strikes. Short of that, a prosecutor who uses her strikes parsimoniously, and who voluntarily articulates the basis for the strikes, could do a lot to foreclose future Batson claims. It would be a good outcome if prosecutors were more circumspect about using their peremp­tories. It would be a welcome side effect of the trial–post-trial divergence if pro­secutorial behavior at trial were at least somewhat deterred by the un­certainty about future Batson litigation.

Finally, it is worth noting that there are other doctrines in which tri­al and post-trial practice dramatically diverge. A claim of judicial bias, for example, faces long odds at trial. The more biased the judge, the more likely the judge would reject the claim. But on appeal it could win. The denial of a public trial, another structural error, would be more likely to succeed on appeal than at trial. That is true not only because the judge who improperly closed the hearing would seem less likely to grant the public-trial objection but also because the public-trial right serves a pur­pose that is larger than the particular trial—preserving the public confi­dence in the integrity of the judiciary—and might thus have more purchase for appellate judges who must consider a wider range of cases. Juror-misconduct claims also fit within the trial–post-trial rubric, as the evidence needed to win on a juror-misconduct claim usually becomes available only when the record is expanded in postconviction litigation. 250 Misconduct allegations can conceivably be brought at trial, if the misconduct is detected.

What makes Batson stand out from the examples above is the duality between trial and post-trial proceedings. Batson can be raised at trial us­ing only the evidence on the record and then it can be raised in a very different way post-trial with the extra-record evidence. This postconvic­tion pivot distinguishes Batson from the claims discussed above, which can be alleged in only the most skeletal form at trial and must wait for record-expansion to be alleged in earnest. This pivot is a testament to the fact that Batson did not fully eclipse its extra-record-only predecessor, Swain; it just provided an easier route to get to the same place. Litigants can use a record-based Batson claim to preserve the issue for appellate and postconviction litigation, and then supplement that objection with later-acquired extra-record evidence that speaks to the prosecutor’s actu­al intent. And the courts must consider this additional evidence because the Supreme Court’s equal protection jurisprudence has explicitly set the focus on the prosecutor’s actual intent.


Batson has received significant and sustained criticism for what it fails to do. And while there is much that the critics get right about Batson, the criticism has failed to distinguish between the trial and post-trial sides of the doctrine. Unlike other criminal procedure doctrines, Batson plays out very differently on appeal than at trial. And, as this Article has shown, appellate Batson provides a number of opportunities for litigants and courts to address the stain of racial discrimination in ways that were not possible at trial. This Article is not an apology for Batson or an attempt to say that the doctrine is functioning fine. Rather, it aims to show Batson in a different light from the standard trial-focused critique. And it attempts to call attention to Batson’s strengths relative to other antidiscrimination doctrines.

Batson holds great potential to address the harm caused by racial discrimination in the justice system. Because it protects all litigants—criminal defendants, prosecutors, and civil litigants—and because it ul­timately protects jurors and the judicial system itself, it has survived intact over the years, even as courts have cast other antidiscrimination doctrines aside. One of the abiding ironies this Article discusses is that evidence of systemic racism can be used to prove a Batson violation, but the violation only necessarily implies a single act of wrongdoing. The doctrine is thus more palatable to the judiciary than fair cross-section, discriminatory-charging, and other claims that imply widespread discrimination. Perhaps in this unusual brew of universalism, third-party standing, and sys­temic-but-one-off jurisprudence, there is a model for other antidiscrimi­nation doctrines.