CRUEL TIMING: RETROACTIVE APPLICATION OF STATE CRIMINAL PROCEDURAL RULES TO DIRECT APPEALS

CRUEL TIMING: RETROACTIVE APPLICATION OF STATE CRIMINAL PROCEDURAL RULES TO DIRECT APPEALS

In 1987, the Supreme Court held that the Constitution requires federal and state courts to retroactively apply all new federal-constitutional rules of criminal procedure to direct appeals of convictions. Since then, the Court has not addressed whether the U.S. Constitution also requires state courts to retroactively apply new criminal procedure rules derived from state law on direct review. This issue is particularly significant because state jurisdictions prosecute the overwhelming majority of criminal cases in the United States. Most state courts have continued using a type of case-specific retroactivity analysis that usually leads them to reject retroactive application of new state rules to direct appeals. So the extent of a new criminal procedural rule’s retroactive effect, and potentially the result of a defendant’s appeal, is tied to the source of the new rule: federal or state law. This Note examines the Supreme Court’s retroactivity jurisprudence and assesses the different approaches state courts have taken in determining whether to retroactively apply new state rules to direct appeals. Ultimately, this Note argues that the Due Process Clause requires state courts to retroactively apply all new procedural rules, whether derived from federal or state law, to direct appeals of convictions, because failure to do so results in unequal treatment of similarly situated defendants.

Introduction

Consider two similar cases: Carla Carlson and Mark Marks are both convicted of murder on similar evidence. Both appeal. Marks’s appeal is heard first. The state’s highest court reverses Marks’s conviction and creates a new rule that the type of evidence used to convict Marks and Carlson is inadmissible because it was discovered through police conduct that violated state law. The remaining evidence is insufficient to warrant a conviction, so the prosecution dismisses Marks’s case and he is released from custody. Then Carlson’s appeal is heard, and she claims the appellate court should apply this new rule to require the same relief as Marks received. If the new rule is rooted in the U.S. Constitution, the court will apply it to Carlson’s appeal. 1 See Griffith v. Kentucky, 479 U.S. 314, 322 (1987) (holding U.S. Constitution requires federal and state courts to apply newly declared federal-constitutional rules of criminal procedure to criminal cases on direct review). Whether Griffith v. Kentucky also requires that courts retroactively apply all new federal rules, not just those derived from the U.S. Constitution, is disputed. The First, Sixth, Seventh, and Ninth Circuits have found that it does, while the Third Circuit has ruled that it does not. See infra section III.A (discussing dispute and its ramifications for retroactivity of state criminal procedural rules, and arguing Griffith applies to all new federal rules). But if the rule is rooted in state law, most courts will likely reject Carlson’s request and uphold her conviction, even though under current law Carlson’s conviction was based on inadmissible evidence and her conviction is no different from Marks’s, except in that it was not reversed. 2 See Cowans v. Bagley, 624 F. Supp. 2d 709, 735–40 (S.D. Ohio 2008) (stating most state courts generally do not retroactively apply new state procedural rules on direct review and giving examples); Taylor v. State, 10 S.W.3d 673, 680–81 (Tex. Crim. App. 2000) (same). The extent of a new criminal procedural rule’s retroactive effect is thus tied to the rule’s source: federal or state law. Carlson is denied the benefit of a new state-law rule because her conviction was both too early and too late. Had the state’s highest court heard Carlson’s case before Marks’s, Carlson would go home free and not Marks. If Carlson’s trial occurred after the new rule’s announcement in Marks’s case, the rule would apply to Carlson’s trial in the first place. 3 This hypothetical is loosely based on Modesitt v. State, 578 N.E.2d 649 (Ind. 1991), and Mason v. Duckworth, 74 F.3d 815 (7th Cir. 1996). In both cases, the defendants were convicted partly because of witnesses’ unsworn out-of-court statements that were inconsistent with the witnesses’ trial testimony. In Modesitt, the Indiana Supreme Court overruled a previous decision by holding that such statements were inadmissible hearsay, and reversed Modesitt’s conviction. 578 N.E.2d at 652–54. Months later, the Indiana Court of Appeals rejected Mason’s request for similar relief and upheld his eighty-year sentence even though the evidence that helped convict him was no longer admissible in Indiana under Modesitt. See Mason, 74 F.3d at 817–18 (discussing Indiana Court of Appeals’s decision). The Seventh Circuit judge who wrote the opinion affirming the denial of Mason’s habeas corpus petition noted,
One should never underestimate the importance of good timing. Had John Mason’s ex-wife Patricia been murdered after September 27, 1991 . . . his chances of being a free man today—instead of a guest of the State of Indiana at one of its penal institutions—would be appreciably better. But Mason’s timing was bad . . . .
Id. at 816.
Carlson suffers only because her trial happened in this exact timeframe. Does this treatment of Carlson violate the U.S. Constitution? This Note explores that question.

In Griffith v. Kentucky in 1987, the Supreme Court ruled that the Constitution requires that federal and state courts apply newly declared federal-constitutional rules of criminal procedure to criminal cases on direct review. 4 479 U.S. at 320–22 (applying Batson rule prohibiting race-based peremptory challenges to jurors and reversing conviction on direct appeal). For definitions of “direct review” and “collateral review,” see infra notes 19–21 and accompanying text. The Court based its decision on two constitutional principles: Article III’s “cases” and “controversies” requirement 5 The Constitution authorizes the Supreme Court to hear only actual “cases” and “controversies.” See U.S. Const. art. III, § 2 (establishing scope of federal judicial power). and the principle of treating similarly situated defendants equally. 6 Griffith, 479 U.S. at 323. It is not clear which constitutional provision the equal treatment principle comes from. See infra note 56 (discussing debate about source of equal treatment principle and showing principle is likely rooted in Due Process Clause or Equal Protection Clause). But the Court has not addressed whether and under what circumstances the Constitution requires that courts apply newly declared state-law rules of criminal procedure, derived from state constitutions, statutes, or common law, 7 For an explanation of the differences between these three types of rules, see infra section I.A. For a discussion of potential differences in how they should be treated, see infra section II.D. to criminal cases on direct review. 8 See Griffith, 479 U.S. at 322 (saying nothing about state-law rules); see also, e.g., Cowans v. Bagley, 624 F. Supp. 2d 709, 734–35 (S.D. Ohio 2008) (stating Griffith did not address retroactive application of state-law rules); Guzman v. Greene, 425 F. Supp. 2d 298, 316–17 (E.D.N.Y. 2006) (same). State and federal courts universally hold that Griffith’s bright-line rule of retroactivity does not bind them with regard to any kind of state law. The vast majority of state courts have instead adopted a case-specific retroactivity analysis that usually hinges on the perceived importance of the rule in question in the fact-finding process. 9 See infra section II.B (describing and examining constitutional and normative justifications for majority approach to retroactivity of state rules). Consequently, courts that use this case-specific analysis almost never find a newly declared state rule of criminal procedure to be sufficiently important to the fact-finding process as to merit retroactive application to criminal cases on direct review. 10 See infra notes 98–104 and accompanying text (noting how rarely courts using case-specific analysis retroactively apply new rules and describing types of rules courts never retroactively apply). In some of these cases, applying the newly declared state rule to a conviction on direct review would likely lead to a reversal, increasing the likelihood of either a not-guilty verdict (as in Carlson’s case) or a lower sentence on remand. 11 See, e.g., Murtishaw v. State, 255 F.3d 926, 955–56 (9th Cir. 2001) (declining to retroactively apply new requirement mandating trial courts instruct juries that honest, unreasonable belief of need for self-defense negates malice when evidence warrants it); Mason v. Duckworth, 74 F.3d 815, 818 (7th Cir. 1996) (upholding conviction after declining to retroactively apply new hearsay rule, that would have excluded incriminating evidence). A new criminal procedural rule’s impact on defendants thus hinges on whether the rule comes from federal 12 A majority of circuits to confront the issue have held that Griffith v. Kentucky requires courts to retroactively apply all new federal criminal procedural rules (i.e., constitutional, statutory, and common law) to direct appeals. Only the Third Circuit has limited Griffith’s application to federal-constitutional rules. See infra section III.A (describing circuit split). or state law.

American retroactivity jurisprudence is notoriously convoluted and has changed significantly over the last fifty years. 13 See Guzman, 425 F. Supp. 2d at 316 (“The failure of the Supreme Court to state that its holding applies only to new constitutional rules has created some confusion.”); Harry A. Olivar, Jr., Criminal Procedure: Retroactivity of Batson v. KentuckyGriffith v. Kentucky, 107 S. Ct. 708 (1987), 10 Harv. J.L. & Pub. Pol’y 773, 776 (“The doctrine of retroactivity is a thorny one, as can be seen from the Court’s past struggles in this area.”). Some courts have appeared unsure of how to navigate this jurisprudence, mistakenly applying the wrong cases or conflating precedents about substantive rules and precedents about procedural rules. See, e.g., Guzman, 425 F. Supp. 2d at 316 (failing to distinguish between cases involving substantive rules and those involving procedural rules). This Note aims to guide courts’ and practitioners’ navigation of Supreme Court retroactivity jurisprudence. State criminal procedure law represents one small branch of this jurisprudence that is rarely discussed in academic scholarship, which has primarily focused on retroactivity of federal-constitutional rules instead, 14 See, e.g., A. Christopher Bryant, Retroactive Application of “New Rules” and the Antiterrorism and Effective Death Penalty Act, 70 Geo. Wash. L. Rev. 1, 1–6 (2002) (focusing on retroactivity of federal-constitutional rules); Richard H. Fallon, Jr. & Daniel J. Meltzer, New Law, Non-Retroactivity, and Constitutional Remedies, 104 Harv. L. Rev. 1731, 1731–34 (1991) (same); Meir Katz, Plainly Not “Error”: Adjudicative Retroactivity on Direct Review, 25 Cardozo L. Rev. 1979, 1980 (2004) (same); Paul J. Mishkin, Foreword: The High Court, The Great Writ, and the Due Process of Time and Law, 79 Harv. L. Rev. 56, 58–72 (1965) (same); Kermit Roosevelt III, A Little Theory Is a Dangerous Thing: The Myth of Adjudicative Retroactivity, 31 Conn. L. Rev. 1075, 1077–79 (1999) [hereinafter Roosevelt, A Little Theory] (same); Walter V. Schaefer, The Control of “Sunbursts”: Techniques of Prospective Overruling, 42 N.Y.U. L. Rev. 631, 631–35 (1967) (same). but merits close examination because state jurisdictions prosecute the overwhelming majority of criminal cases in the United States. 15 Daniel C. Richman, Kate Stith & William J. Stuntz, Defining Federal Crimes 2 (2d ed. 2014) (noting local jurisdictions prosecute five to six million cases a year, while federal jurisdictions prosecute roughly 60,000 a year). This Note aims to fill a gap in this scholarship by asking where state criminal procedural law fits into the Supreme Court’s retroactivity jurisprudence and assessing how state courts decide whether to retroactively apply new state criminal procedural rules. This Note concludes that denying a convicted defendant the benefit on direct appeal of a new criminal procedure ruling only because the defendant’s trial occurred too early (before the state’s highest court declared the new rule) and her appeal occurred too late (after the state’s highest court issued the ruling in another defendant’s case) violates a constitutional principle articulated in Griffith by treating similarly situated defendants unequally.

Part I recounts the development of the Supreme Court’s retroactivity jurisprudence and discusses its nuances that are relevant to this Note. Part II discusses in what circumstances and how often courts deal with retroactivity of state criminal procedural rules on direct review, describes the different approaches state courts have taken to this issue, and examines arguments in favor of and against these approaches. Part III argues that the U.S. Constitution requires courts to retroactively apply all new state criminal procedural rulings to defendants’ direct appeals.

I. Retroactivity on Direct Review Generally

“The question of retroactivity is what to do when the law changes. More precisely, it is to whom the new law should be applied, and to whom the old.” 16 Roosevelt, A Little Theory, supra note 14, at 1075. Section I.A explains some background law that is important for understanding retroactivity doctrine. Section I.B sets forth the historical development of the U.S. Supreme Court’s retroactivity doctrine. Section I.C discusses Griffith v. Kentucky and its two rationales. Section I.D introduces the issue of whether the U.S. Constitution requires retroactive application of any state rules.

A. Background Law

Before beginning an in-depth discussion of retroactivity, it is important to note four distinctions that courts treat as significant in deciding whether to retroactively apply a new rule. 17 See infra note 57 and accompanying text (defining “new rule”). The first is the difference between direct review and collateral review. 18 See infra section I.C.2 (explaining significance of distinction between direct and collateral review in retroactivity law). “Direct review” refers to a court’s review of a criminal defendant’s initial (referred to as “direct”) appeal after she was convicted in trial court. 19 See Richard H. Fallon, Jr. et al., Hart and Wechsler’s The Federal Courts and the Federal System 1266 (7th ed. 2014) (defining direct review). When a defendant exhausts her right to direct appeal, the conviction is deemed “final.” 20 Id. at 1292. “Collateral review” refers to a court’s review of a criminal defendant’s habeas petition after her conviction is final. 21 See id. (distinguishing collateral attacks on final convictions from direct appeals).

The second is the difference between federal and state law. State courts have adopted varying doctrines for determining whether to retroactively apply new state rules, most of which are different from the doctrine federal courts use to decide whether to retroactively apply a new federal rule. 22 See supra notes 1–3 and accompanying text (discussing Carla Carlson hypothetical, in which retroactive effect of new evidence rule was tied to whether rule was derived from federal or state law); infra Appendix (displaying different retroactivity doctrines applied by federal and state courts).

The third is the difference between constitutional rules, statutory rules, and common law rules. 23 References to “rules of criminal procedure” in this Note refer to all three of these rules. Constitutional rules are rules a court derives from the U.S. or a state constitution. 24 See Obergefell v. Hodges, 135 S. Ct. 2584, 2607 (2015) (declaring new rule derived from U.S. Constitution); Batson v. Kentucky, 476 U.S. 79, 82–84 (1986) (holding U.S. Constitution’s Equal Protection Clause requires criminal defendant be allowed to establish prima facie case of racial discrimination based on prosecution’s use of peremptory challenges to strike members of defendant’s race from jury venire). Statutory rules are rules a court derives from a statute. 25 See, e.g., Texas Dep’t of Hous. & Cmty. Affairs v. Inclusive Cmtys. Project, Inc., 135 S. Ct. 2507, 2514 (2015) (interpreting Fair Housing Act to allow impact-based discrimination claims); Bailey v. United States, 516 U.S. 137, 144 (1995) (interpreting federal statute criminalizing “use” of firearm during and in relation to drug trafficking to require evidence of active employment of firearm by defendant, not mere possession). Common law rules are rules “whose content cannot be traced directly by traditional methods of interpretation to . . . statutory or constitutional commands.” 26 Fallon et al., supra note 19, at 635. For an example of a case that created a common law rule of criminal procedure, see Modesitt v. State, 578 N.E.2d 649, 652–54 (Ind. 1991) (overruling previous decision by holding out-of-court declarant-witness statements not made under oath and inconsistent with declarant-witness’s testimony are inadmissible hearsay).

The fourth is the difference between substantive and procedural rules. The Supreme Court has said that a criminal rule is substantive “if it alters the range of conduct or the class of persons that the law punishes.” 27 Schriro v. Summerlin, 542 U.S. 348, 353 (2004). The clearest example of a substantive rule is a statute criminalizing a particular type of activity, like murder or robbery. Other examples include rules rooted in the Eighth Amendment’s prohibition of cruel and unusual punishment. See Montgomery v. Louisana, 136 S. Ct. 718, 732–36 (2016) (holding new rule, which declared Eighth Amendment bars mandatory life sentences without parole for juvenile offenders, was substantive). Conversely, “rules that regulate only the manner of determining the defendant’s culpability are procedural.” 28 Schriro, 542 U.S. at 353. Obvious examples of procedural rules include rules about admissibility of evidence and rules allocating decisionmaking authority among the judge and jury. Other examples include: a rule that the Sixth Amendment’s jury-trial guarantee prohibits sentencing judges from finding aggravating circumstances necessary for the imposition of the death penalty, see id.; a rule prohibiting courts from instructing a jury to disregard mitigating factors not found by a unanimous vote, see Beard v. Banks, 542 U.S. 406, 408 (2004); and a rule forbidding prosecutors from suggesting to a capital jury that it is not responsible for a death sentence, see Sawyer v. Smith, 497 U.S. 227, 232–33 (1990). Classifying a rule as either substantive or procedural is sometimes difficult. 29 See, e.g., Katharine A. Ferguson, Note, The Clash of Ring v. Arizona and Teague v. Lane : An Illustration of the Inapplicability of Modern Habeas Retroactivity Jurisprudence in the Capital Sentencing Context, 85 B.U. L. Rev. 1017, 1032 (2005) (discussing difficulty of classifying rule as substantive or procedural). Some new rulings that courts have declined to retroactively apply to direct appeals arguably can be classified as either substantive or procedural. See, e.g., State v. Hudson, 415 S.E.2d 732, 751 (N.C. 1992) (declining to retroactively apply new ruling eliminating perjury as aggravating factor in sentencing). But this distinction is important because courts are far more likely to retroactively apply substantive criminal rules than procedural ones. 30 See, e.g., Teague v. Lane, 489 U.S. 288, 305–11 (1989) (ruling only new constitutional rules that are substantive, not procedural (unless they are rare “watershed” procedural rules), apply retroactively on collateral review); see also, e.g., People v. Vasquez, 670 N.E.2d 1328, 1333 (N.Y. 1996) (“[C]ases on direct appeal are generally decided in accordance with the law as it exists at the time the appellate decision is made . . . .”). This is because a new substantive rule can exculpate a convicted defendant or forbid a punishment imposed upon her, whereas a new procedural rule only governs the manner in which criminal proceedings are conducted. 31 See, e.g., Schriro, 542 U.S. at 352 (“[Substantive] rules apply retroactively because they ‘necessarily carry a significant risk that a defendant stands convicted of an act that the law does not make criminal’ or faces a punishment that the law cannot impose upon him.”(quoting Bousley v. United States, 523 U.S. 614, 620–21 (1998))). Refusal to retroactively apply an exculpatory criminal substantive rule to a defendant means a continual deprivation of her liberty or property, despite the fact that the state’s highest court has decided that her actions are not a crime. In most cases, state courts apply new state criminal substantive rules to direct appeals 32 See, e.g., Vasquez, 670 N.E.2d at 1333 (“[C]ases on direct appeal are generally decided in accordance with the law as it exists at the time the appellate decision is made . . . .”); State v. Lagundoye, 674 N.W.2d 526, 531 (Wis. 2004) (“[A] new rule of substantive criminal law is presumptively applied retroactively to all cases, whether on direct appeal or on collateral review.”). For an exception, see People v. Douglas, 617 N.Y.S.2d 733, 739 (App. Div. 1994) (refusing to retroactively apply new rule requiring proof defendant had knowledge of weight of controlled substance as element of possessory offense due to number of cases that would be affected).  and do not apply new state criminal procedural rules to direct appeals. 33 See, e.g., People v. Mitchell, 606 N.E.2d 1381, 1384 (N.Y. 1992) (declining to retroactively apply newly declared right of defendant to be presented for certain voir dire questioning); State v. Cowans, 717 N.E.2d 298, 314 (Ohio 1999) (declining to retroactively apply rule requiring trial court to determine, when capital defendant evinces desire to waive mitigation evidence, that waiver is knowing and voluntary and defendant is competent to effect waiver).

Federal and state courts have built their retroactivity doctrines around these four distinctions. All four affect whether a court will retroactively apply a rule. For example, courts almost always retroactively apply substantive rules (like a rule that the First Amendment bars criminalization of flag-burning) on both direct and collateral appeals, 34 See infra Appendix (showing courts generally apply both federal and state rules to direct and collateral appeals). but they usually do not retroactively apply procedural rules (like a rule governing jury selection) on collateral appeals. 35 See, e.g., Teague, 489 U.S. at 305–11 (holding new federal procedural rules will not be retroactively applied on collateral review unless they are “watershed” rules). The Appendix displays the different types of new rules (whether they are substantive or procedural, federal or state, or derived from a constitution, statute, or common law) and the extent to which courts retroactively apply them to convictions on direct appeal. Part II and Part III focus on state criminal procedural rulings (the Appendix’s bottom row), but it is important to understand where these rulings fit within the larger picture of retroactivity jurisprudence.

Section I.B discusses the historical development of retroactivity doctrine.

B. The History of Retroactivity

At common law, all judicial decisions were applied to all direct and collateral appeals, keeping with Blackstone’s logic “that the duty of the court was not to ‘pronounce a new law, but to maintain and expound the old one.’” 36 Linkletter v. Walker, 381 U.S. 618, 622–23 (1965) (quoting 1 William Blackstone, Commentaries 69 (15th ed. 1809)). Judicial decisions were merely evidence of an independent and unchanging common law, so there was no “new” law to apply retroactively. 37 See Roosevelt, A Little Theory, supra note 14, at 1082–83 (describing Blackstonian view of law). Over time, however, legal philosophers began to promote the more realistic view that when judges overrule a previous decision or announce a decision that is unexpected in light of preexisting law, they are not merely discovering law, but in fact “making” new law. 38 See Linkletter, 381 U.S. at 623–24 (describing gradual adoption in nineteenth and early twentieth century of view “judges do . . . something more than discover law; they make it interstitially by filling in with judicial interpretation the vague, indefinite, or generic statutory or common-law terms that alone are but the empty crevices of the law”). In reaction to such views, the Supreme Court began to take a more nuanced approach to the retroactive application of new judicial rulings, considering factors such as whether the case before the court was on direct or collateral appeal, whether the party opposing retroactive application had relied on a previous rule in a consequential way, and public policy. 39 See id. at 624–28 (describing Supreme Court’s retroactivity jurisprudence through 1965); Chicot Cty. Drainage Dist. v. Baxter State Bank, 308 U.S. 371, 374 (1940) (“[Existence of prior rule] is an operative fact and may have consequences which cannot justly be ignored. The past cannot always be erased by a new judicial declaration.”). See generally Fallon et al., supra note 19, at 1292–94 (describing development of retroactivity doctrine). In the late 1960s, the Supreme Court held that the “Constitution neither prohibits nor requires retrospective effect” of any new decision of constitutional criminal procedure 40 A “constitutional rule of criminal procedure” is a criminal procedural rule derived from and rooted in the U.S. Constitution. on direct or collateral review, thus leaving the Court to decide the question as a matter of policy. 41 Linkletter, 381 U.S. at 629; see also Stovall v. Denno, 388 U.S. 293, 300–01 (1967) (“We also conclude that, for these purposes, no distinction is justified between convictions now final, as in the instant case, and convictions at various stages of trial and direct review.”). The Warren Court created several additional rules of retroactivity in the late 1960s. See Desist v. United States, 394 U.S. 244, 256–57 (1969) (Harlan, J., dissenting) (listing rules). These rules were created in a period when the Court greatly expanded protections available to criminal defendants. See James B. Beam Distilling Co. v. Georgia, 501 U.S. 529, 537–38 (1991) (describing creation of these rules and motivation behind them); see also Fallon et al., supra note 19, at 1293 (“The retroactivity question grew in significance during the 1960s as a result of the sharp increase in the number of habeas petitions filed, the expansive scope of habeas review recognized by the Warren Court, and that Court’s broad and novel criminal procedure decisions . . . .”).
The Court embraced selective prospectivity (applying a rule to the party before the court, but not to other litigants on direct or collateral review) as a method of allowing for the expansion of such protections without “seriously disrupt[ing] the administration of our criminal laws [by] . . . requir[ing] the retrial or release of numerous prisoners found guilty by trustworthy evidence in conformity with previously announced constitutional standards.” Johnson v. New Jersey, 384 U.S. 719, 731 (1966). Justice Harlan objected to what he described as “incompatible rules and inconsistent principles” that created “doctrinal confusion.” Desist, 394 U.S. at 258 (Harlan, J., dissenting) (“‘Retroactivity’ must be rethought.”); see also Mackey v. United States, 401 U.S. 646, 675–95 (1971) (Harlan, J., concurring in part and dissenting in part) (describing problems with Court’s convoluted retroactivity jurisprudence and promoting uniform rule of retroactivity on direct review and prospectivity of procedural rules on collateral review).
Toward that end, it created a three-prong test (the “Stovall test”) for claims of retroactivity of new constitutional rules of criminal procedure, considering “(a) the purpose to be served by the new standards, (b) the extent of the reliance by law enforcement authorities on the old standards, and (c) the effect on the administration of justice of a retroactive application of the new standards.” 42 Stovall, 388 U.S. at 297.

In 1982, the Supreme Court changed course again, ruling that retroactivity analysis for convictions on direct review must be different than for convictions on collateral review. 43 See United States v. Johnson, 457 U.S. 537, 562–63 (1982) (holding Supreme Court decision construing Fourth Amendment must be applied retroactively to all convictions on direct review, subject to one exception). The Court openly adopted the logic of Justice Harlan’s dissents in Desist and Mackey. Id. at 548 (“We now agree with Justice Harlan that retroactivity must be rethought.” (citation omitted) (internal quotation marks omitted) (quoting Desist, 394 U.S. at 258 (Harlan, J., dissenting))). This set the stage for Griffith v. Kentucky in 1987.

C. Griffith: The Retroactive Application of Constitutional Criminal Procedural Rules on Direct Review

In Griffith v. Kentucky, the Court held that the U.S. Constitution requires that courts retroactively apply all new federal-constitutional criminal procedural rules to direct appeals of convictions, explicitly overruling the Court’s previous contrary rulings and exceptions. 44 479 U.S. 314, 322–23, 328 (1987). The Court also eliminated the exception to retroactive application that had previously existed for cases in which the new rule constituted a “clear break” with the past. Id. at 326–27. The two Griffith defendants had been convicted at trial in 1984, and courts rejected their direct appeals in 1985. 45 Id. at 318–19; Brief for Respondent at 1, 3, Griffith v. Kentucky, 479 U.S. 314 (1987) (No. 85-5221), 1986 WL 727575. Both defendants petitioned for writs of certiorari to the U.S. Supreme Court. 46 Griffith, 479 U.S. at 318–20. In June 1986, while these petitions were pending, the Court held in Batson v. Kentucky that criminal defendants can establish a prima facie case of racial discrimination based on the prosecution’s use of peremptory challenges to strike members of the defendant’s race from the jury venire, and shift the burden to the prosecution to produce a neutral explanation for its challenges. 47 476 U.S. 79, 89–96 (1986). Two weeks later, the Court granted certiorari to the two Griffith defendants. 48 Griffith, 479 U.S. at 318–20. In January 1987, the Court applied the new Batson rule and reversed the two convictions, remanding the cases to allow lower courts to determine whether the record disclosed any improper use of the prosecution’s peremptory challenges of black jurors. 49 Id. at 328. The remedy a court uses when it retroactively applies a new rule on direct appeal is highly case specific. Courts sometimes find that a failure to use a subsequently announced criminal procedural rule constitutes harmless error and decline to remand the case. See, e.g., United States v. Lopez-Pena, 912 F.2d 1542, 1550 (1st Cir. 1989) (applying retroactively new ruling barring magistrates from presiding over jury selection in felony cases, but finding magistrate’s violation of rule was harmless error, and denying defendant’s request for remand). Courts typically do not overturn guilty pleas in the wake of a new criminal procedural rule that would have weakened the prosecution’s evidence at trial. See, e.g., Brady v. United States, 397 U.S. 742, 757 (1970) (holding withdrawal of guilty plea in federal court not allowed “simply because it later develops that the State would have had a weaker case than the defendant had thought or that the maximum penalty . . . has been held inapplicable in subsequent . . . decisions”).

To date, however, the Supreme Court has not addressed the central focus of this Note: whether Griffith applies to new criminal procedure rulings that are rooted not in the U.S. Constitution but rather in state law. 50 “State law” refers to a rule derived from a state constitution, statute, or judicial decision. Addressing this issue requires an explanation of Griffith’s reasoning.

1. The Two Rationales of Griffith. — Griffith’s holding was based on two constitutional principles. The first is that Article III requires that the Court only adjudicate specific “cases” and “controversies.” 51 Griffith, 479 U.S. at 322–23; see also U.S. Const. art. III, § 2. Therefore, the “nature of judicial review requires” that the Court apply current federal-constitutional law (including a newly declared rule) to cases that have not exhausted the full course of appellate review and instruct lower federal and state courts to do the same. 52 Griffith, 479 U.S. at 322–23 (stating nature of judicial review prevents Court from “‘fishing one case from the stream of appellate review, using it as a vehicle for pronouncing new constitutional standards, and then permitting a stream of similar cases subsequently to flow by unaffected by that new rule’” (quoting Mackey v. United States, 401 U.S. 667, 679 (1971) (Harlan, J., concurring in the judgment))).

The second constitutional principle supporting Griffith is that courts must treat similarly situated defendants equally. 53 Id. at 323. The Court said that “the problem with not applying new rules to cases pending on direct review is ‘the actual inequity that results when the Court chooses which of many similarly situated defendants should be the chance beneficiary’ of a new rule.” 54 Id. (quoting United States v. Johnson, 457 U.S. 537, 556 n.16 (1982)). The Court did not say whether Griffith stands fully and equally on each constitutional principle it discussed or whether it only speaks to the combination of both points. 55 See infra section III.B (analyzing whether Griffith’s equal treatment principle is sufficient alone to require Griffith rule). This could affect whether Griffith applies to state law because the “equal treatment” principle almost certainly applies to state courts, 56 The Court did not cite a specific constitutional provision in support of this principle. There has been some debate about which constitutional provision this principle comes from. An amicus brief filed on Griffith’s behalf suggests that this principle is derived from the Equal Protection Clause. See U.S. Const. amend. XIV (containing Equal Protection Clause); Brief for the Lawyers’ Committee for Civil Rights Under Law as Amici Curiae Supporting Petitioner at 24–26, Griffith v. Kentucky, 479 U.S. 314 (1987) (No. 85-5221), 1986 WL 727572 (arguing Equal Protection Clause requires Supreme Court to treat all similarly situated defendants equally and apply remedies to equal protection violations consistently). Some state courts have disagreed with this contention. See Commonwealth v. Waters, 511 N.E.2d 356, 357 (Mass. 1987) (stating “Griffith . . . is not based on equal protection . . . grounds”); Burr v. Kulas, 532 N.W.2d 388, 391–92 (N.D. 1995) (rejecting Equal Protection Clause challenge based in part on Griffith to prospective application of new rule of state civil law); State v. Stilling, 770 P.2d 137, 143 (Utah 1989) (rejecting Equal Protection claim based on Griffith).
Alternatively, some scholars believe Griffith’s equal treatment principle is most likely rooted in the Fifth Amendment’s Due Process Clause. See, e.g., Russell M. Coombs, A Third Parallel Primrose Path: The Supreme Court’s Repeated, Unexplained, and Still Growing Regulation of State Courts’ Criminal Appeals, 2005 Mich. St. L. Rev. 541, 608 (arguing Supreme Court’s most likely justification for imposing Griffith rule on state courts would be due process); Leonard N. Sosnov, No Mere Error of State Law: When State Appellate Courts Deny Criminal Defendants Due Process, 63 Tenn. L. Rev. 281, 317–18 (1996) (regarding principle of treating similarly situated defendants equally as rooted in Due Process Clause).
Both the Equal Protection Clause and the Due Process Clause apply to state governments. See, e.g., Malinski v. New York, 324 U.S. 401, 415 (1945) (Frankfurter, J., concurring) (“To suppose that ‘due process of law’ meant one thing in the Fifth Amendment and another in the Fourteenth is too frivolous to require elaborate rejection.”). So this uncertainty does not affect any of the analysis in section III.A unless Griffith’s principle of equal treatment of similarly situated defendants is rooted in a part of the Constitution that does not apply to states, such as Article III. See Coombs, supra, at 597–98 (discussing possibility that Griffith’s equal treatment principle is rooted in Article III). If this were true, it would undercut the argument that Griffith imposed retroactivity requirements on state courts. This seems unlikely though, because the Supreme Court clarified years later that Griffith does impose some retroactivity requirements on state courts. That suggests Griffith’s holding is not based solely on Article III, but also on a constitutional provision that applies to states. See Powell v. Nevada, 511 U.S. 79, 84 (1994) (holding Griffith requires state courts, in addition to federal courts, to retroactively apply new federal-constitutional rules); Coombs, supra, at 598 (admitting Powell v. Nevada likely means Griffith’s equal treatment principle applies to state courts). For simplicity’s sake, this Note assumes that Griffith’s principle of equal treatment is rooted in the Due Process Clause. But, again, section III.B’s constitutional analysis would be the same if the equal treatment principle were instead rooted in the Equal Protection Clause.
but Article III’s Cases and Controversies Clause does not.

2. The Aftermath of Griffith: The Distinction Between Direct and Collateral Review. — Two years after Griffith, the Court ruled in Teague v. Lane that new constitutional rules 57 A new rule is one that breaks new ground or imposes a new obligation on states or the federal government. “To put it differently, a case announces a new rule if the result was not dictated by precedent existing at the time the defendant’s conviction became final.” Teague v. Lane, 489 U.S. 288, 301 (1989). are not to be applied retroactively on collateral review, with two exceptions: substantive rules and the rare procedural rules that are “components of basic due process.” 58 Id. at 310–13. Since Teague, the Supreme Court has said that only watershed rules of criminal procedure that implicate the fundamental fairness and accuracy of the criminal proceeding qualify as components of basic due process, and it is unlikely that any such rules have yet to emerge. Whorton v. Bockting, 549 U.S. 406, 417 (2007). Accordingly, the Court has rejected every claim that a new procedural rule qualified as a component of basic due process. Id. at 418. The Court based this decision on the difference between collateral and direct review. Collateral review is not meant to serve a “perceived need to assure that an individual accused of crime is afforded a trial free of constitutional error.” 59 Teague, 489 U.S. at 308. As the Court has said more recently: “[H]abeas corpus is a guard against extreme malfunctions in the state criminal justice systems, not a substitute for ordinary error correction through appeal.” Harrington v. Richter, 131 S. Ct. 770, 786 (2011) (internal quotation marks omitted) (quoting Jackson v. Virginia, 443 U.S. 307, 332 n.5 (1979) (Stevens, J., concurring in the judgment)). Instead, the threat of habeas is meant to incentivize trial and appellate courts to conduct proceedings in accordance with “established constitutional standards.” 60 Teague, 489 U.S. at 306. To perform this function, “the habeas court need only apply the constitutional standards that prevailed at the time the original proceedings took place.” 61 Id.

In tandem, Griffith and Teague have come to signify that the Court’s many noteworthy constitutional criminal procedural rulings in the last twenty-five years have applied retroactively to direct appeals, but not on state post-conviction proceedings or federal habeas corpus review, both forms of collateral review. 62 See, e.g., Ring v. Arizona, 536 U.S. 584, 609 (2002) (“[A] sentencing judge, sitting without a jury, [may not] find an aggravating circumstance necessary for imposition of the death penalty.”). The Supreme Court decided in Schriro v. Summerlin that the new rule announced in Ring was procedural and thus did not apply retroactively on collateral review. 542 U.S. 348, 353, 358 (2002). For example, a defendant whose death penalty sentence, ordered by a judge who found an aggravating circumstance necessary for imposition of the death penalty, became final on June 23, 2002, would be denied a new sentencing hearing, but a defendant in an identical situation whose death sentence had been scheduled to become final on June 25, 2002, would be constitutionally guaranteed to have his sentence invalidated. 63 Ring was decided on June 24, 2002. 563 U.S. at 584. Although this may seem just as inequitable or arbitrary as the situation the Supreme Court addressed in Griffith, the distinction between direct and collateral review has become a cornerstone of the Supreme Court’s retroactivity jurisprudence. 64 See Mary C. Hutton, Retroactivity in the States: The Impact of Teague v. Lane on State Postconviction Remedies, 44 Ala. L. Rev. 421, 423 (1993) (“[W]hen considered in light of . . . Griffith v. Kentucky . . . Teague dramatizes that the posture of a case on either direct or collateral review is now the determinative factor for the Court.”).

Since Teague, the Supreme Court has expanded the list of new rulings that courts must retroactively apply to direct appeals. In 1993, the Court held in Harper v. Virginia Department of Taxation that new federal-constitutional rules must be applied on direct appeal in civil cases, as well as criminal ones. 65 509 U.S. 86, 95–96 (1993) (describing facts of Harper). In 2008, the Court found that state courts can retroactively apply federal-constitutional rules that the U.S. Supreme Court does not, but must retroactively apply any federal rule that the Supreme Court does—so the U.S. Constitution imposes a minimum retroactivity requirement on state courts when it comes to federal-constitutional rules, but not a maximum requirement. 66 See Danforth v. Minnesota, 552 U.S. 264, 267–68 (2008) (allowing Minnesota to retroactively apply federal-constitutional rule of criminal procedure on collateral review, even though Teague dictates federal courts would not do so). But the Court has not said whether the constitutional principles articulated in Griffith similarly require retroactive application of new criminal procedure rulings derived from a source other than the U.S. Constitution—like new interpretations of state constitutions or federal or state statutes, or new state common law rules—on direct review. 67 See, e.g., Cowans v. Bagley, 624 F. Supp. 2d 709, 735 (S.D. Ohio 2008) (noting ambiguity); Guzman v. Greene, 425 F. Supp. 2d 298, 316 (E.D.N.Y. 2006) (same).

D. The Next Frontier for Retroactivity: Does the U.S. Constitution Require Retroactive Application of State Law?

In the last fifty years, the Supreme Court has issued several rulings about and changed its approach to the retroactivity of federal-constitutional rules. 68 See supra sections I.A–.C (discussing evolution in Supreme Court’s retroactivity jurisprudence). But the Court has not decided a case about the retroactivity of state law since 1973, when it upheld the Florida Supreme Court’s refusal to apply a new criminal substantive rule, derived from Florida’s constitution, to a conviction on collateral review. 69 Wainwright v. Stone, 414 U.S. 21, 23–24 (1973). The Supreme Court has never required a state court to retroactively apply any newly declared state law, whether it be a state-constitutional, statutory, or common law ruling. 70 See, e.g., Guzman, 425 F. Supp. 2d at 316–17 (stating Supreme Court has never required state court to retroactively apply state rule); Taylor v. State, 10 S.W.3d 673, 680–81 (Tex. Crim. App. 2000) (stating Constitution imposes no retroactivity requirements on state law). Put another way, the Supreme Court has never said that retroactivity of a state rule is a federal question. 71 On direct appeal from state courts, the Supreme Court can only review decisions that turn on questions of federal law. Federal questions involve “the validity of construction of the Constitution, [laws], or authority of the Federal government.” Murdock v. City of Memphis, 87 U.S. (20 Wall.) 590, 618 (1875). Requiring state courts to retroactively apply new state rules could have a much bigger impact on the American criminal justice system than Griffith did, because all but a fraction of criminal cases in the United States are prosecuted by state jurisdictions. 72 See Richman, Stith & Stuntz, supra note 15 (giving numbers of federal and state prosecutions).

A case that reached the Supreme Court in 1999 suggested for the first time that the Due Process Clause might require that state courts apply at least some new state laws retroactively. In Fiore v. White, two defendants, Fiore and Scarpone, had each been convicted of violating a Pennsylvania statute. 73 See Fiore v. White (Fiore I), 528 U.S. 23, 24–25 (1999). The Pennsylvania Supreme Court later overturned Scarpone’s conviction, holding that the statute did not apply to Scarpone or Fiore. 74 Id. The Pennsylvania courts then denied Fiore’s request to reconsider his identical conviction, even though the Pennsylvania Supreme Court had just declared his actions were not criminal. 75 Id. The U.S. Supreme Court granted certiorari to “decide when, or whether, the Federal Due Process Clause requires a State to apply a new interpretation of a state criminal statute retroactively to cases on collateral review.” 76 Fiore v. White (Fiore II), 531 U.S. 225, 226 (2001).

The Court did not resolve this issue, instead remanding the case and then reversing Fiore’s conviction on other grounds. 77 The Court certified Fiore’s case to the Pennsylvania Supreme Court to determine whether the latter court’s decision in Scarpone’s case (that Fiore and Scarpone’s conduct was not a crime) constituted a change in Pennsylvania law, or merely a clarification of existing law “at the date Fiore’s conviction became final.” Fiore I, 528 U.S. at 29 (emphasis added). The Pennsylvania Supreme Court responded that its decision in Scarpone’s case was merely a clarification of the statute’s previous meaning. Fiore v. White (Fiore II), 757 A.2d 842, 848–49 (Pa. 2000). Once the Pennsylvania Supreme Court made this clear, there was no longer a “new rule” that Fiore sought to have the court retroactively apply to him, and therefore no retroactivity issue. According to Pennsylvania’s highest court, Fiore’s conduct was never illegal. The Supreme Court thus held that the Due Process Clause barred Pennsylvania from convicting Fiore for conduct that its criminal statute did not prohibit at the time of his conviction. The Court did not rule on whether state courts must retroactively apply a “new” rule. Fiore II, 531 U.S. at 228–29. Fiore’s facts, in particular the presence of two identical convictions and a new substantive rule, are extreme, so its broader applicability is unclear. But it suggests for the first time that the Fourteenth Amendment’s Due Process Clause may require state courts to retroactively apply newly declared state laws in certain circumstances. 78 See Guzman v. Greene, 425 F. Supp. 2d 298, 316–17 (E.D.N.Y. 2006) (stating Fiore suggests Due Process Clause imposes retroactivity requirements on state law). Since courts are far more likely to retroactively apply new rulings on direct, rather than collateral, review, 79 See supra section I.C.2 (discussing significance of distinction between direct and collateral review). the fact that the Court granted certiorari to consider whether states must retroactively apply a new criminal substantive rule on collateral review suggests states might be required to apply such a rule on direct review. 80 See Guzman, 425 F. Supp. 2d at 316–17 (“[Fiore’s focus on finality] suggest[ed] that a state should apply a new judicial construction of a state criminal statute to cases that are pending on direct review.”). So at least in some circumstances—but perhaps only extreme ones—retroactivity of a state rule may be a federal question. 81 Fiore is also interesting because of its focus on the distinction between a newly announced rule that explains what the law always was and one that “changes the law” could potentially have led to a far broader ruling about retroactivity. Had the Pennsylvania Supreme Court said that its newly announced rule “changed the law,” the U.S. Supreme Court would have likely had to answer two questions: (1) Can a court’s decision “change the law” when it interprets a statute or only discover what the law had always been, and if the answer is the latter, then (2) are state courts constitutionally permitted to decide that their state’s law is such that state court decisions can change it? Compare Harper v. Va. Dep’t of Taxation, 509 U.S. 86, 106–07 (1993) (Scalia, J., concurring) (arguing courts can only discover what the law has always been, and thus new judicial rulings should always apply retroactively), with Harper, 509 U.S. at 115–17 (O’Connor, J., dissenting) (“‘[W]hen the Court changes its mind, the law changes with it.’” (quoting James B. Beam Distilling Co. v. Georgia, 501 U.S. 529, 550 (1991) (O’Connor, J., dissenting))). These questions are beyond the scope of this Note, but they linger in the background of retroactivity law.

Part II addresses where state criminal procedural rulings fit within the Supreme Court’s retroactivity jurisprudence, and it discusses the different approaches state courts have taken to state-law retroactivity issues.

II. Retroactivity of State Criminal Procedural Rulings: Case-Specific Inquiry or Bright-Line Rule?

New state criminal procedural rulings represent one largely unexplored branch of the Supreme Court’s retroactivity jurisprudence described in Part I. Since Griffith v. Kentucky, at least dozens 82 See Cowans v. Bagley, 624 F. Supp. 2d 709, 735–40 (S.D. Ohio 2008) (providing thirteen cases where defendants requested retroactive application of new state rules); Guzman, 425 F. Supp. 2d at 316–17 (same); Taylor v. State, 10 S.W.3d 673, 679 (Tex. Crim. App. 2000) (providing nine cases where defendants requested retroactive application of new state rules). of convicted defendants like Carla Carlson 83 See supra notes 1–3 and accompanying text (discussing Carlson hypothetical). have requested that state appellate courts apply such rulings that benefit them—usually either new

interpretations of a state statute 84 See, e.g., People v. Carrera, 777 P.2d 121, 127 (Cal. 1989) (declining to retroactively apply newly declared statutory right to privacy of pretrial detainee to conviction on direct appeal); Cooper v. State, 889 P.2d 293, 307–10 (Okla. Crim. App. 1995) (declining, on direct appeal, to retroactively apply new statutory rule barring jury instruction on flight when defendant gave no explanation for flight prior to conviction). or a new common law rule, 85 See Murtishaw v. State, 255 F.3d 926, 955–56 (9th Cir. 2001) (declining to retroactively apply new requirement that trial court must instruct jury that honest, unreasonable belief of need for self-defense negates malice when evidence warrants it); Mason v. Duckworth, 74 F.3d 815, 818 (7th Cir. 1996) (discussing Indiana appellate court’s refusal to retroactively apply recent Indiana Supreme Court decision adopting new evidence rule); People v. Mitchell, 606 N.E.2d 1381, 1384 (N.Y. 1992) (declining to retroactively apply newly declared right of defendant to be presented for certain voir dire questioning). issued by a state’s highest court—to their direct appeals. 86 There appears to be no written opinion since Griffith v. Kentucky in which a state court considered whether to retroactively apply a new interpretation of a state constitutional provision to a criminal defendant’s direct appeal. Before Griffith, some state courts held that new state constitutional criminal procedural and substantive rules would be prospective due to the extensive reliance police, prosecutors, and courts placed on previous rulings. See, e.g., People v. Bustamante, 634 P.2d 927, 935 (Cal. 1981) (en banc) (finding decision, which announced defendant has right under California constitution to assistance of counsel at pre-indictment lineup, would not apply retroactively), superseded by constitutional amendment, Cal. Const. art. I, § 28(d); Franklin v. State, 257 So. 2d 21, 24 (Fla. 1971) (holding decision that state substantive criminal statute violated state constitution would only apply prospectively). State courts refuse these requests in the vast majority of instances. 87 Cowans, 624 F. Supp. 2d at 735–40; Taylor, 10 S.W.3d at 680–81. This Part discusses state courts’ decisions about whether to retroactively apply newly declared state criminal procedural rules on direct appeal and their practical repercussions. It also evaluates the rationales and normative arguments supporting the Stovall approach and a bright-line rule of retroactive application.

A. State Court Split

The vast majority of state courts treat the decision of whether to retroactively apply a new state criminal procedural ruling to a direct appeal as a case-specific inquiry. 88 Taylor, 10 S.W.3d at 680–81 (treating retroactive application as case-specific inquiry and showing most states do same); see also People v. Erickson, 513 N.E.2d 367, 374–75 (Ill. 1987) (treating retroactive application as case-specific inquiry); People v. Sexton, 580 N.W.2d 404, 409–16 (Mich. 1998) (same); State v. Hudson, 415 S.E.2d 732, 751 (N.C. 1992) (same); State v. Gordon, 913 P.2d 350, 353–55 (Utah 1996) (same). Most state courts appear to have adopted Teague’s framework on collateral review. See, e.g., Taylor, 10 S.W.3d at 683 (citing Teague as justification for barring retroactive application of new procedural rules on direct appeal). Most of these courts use a version of the Stovall v. Denno test 89 See 388 U.S. 293, 297 (1967) (articulating test), overruled by Griffith v. Kentucky, 479 U.S. 314 (1987); infra note 92 and accompanying text (explaining test). to decide whether a new ruling should be retroactively applied to a direct appeal, 90 Most state courts adopted the Stovall test before Griffith and maintained it after Griffith. Taylor, 10 S.W.3d at 680–81 (maintaining Stovall approach even after Griffith, and showing most states have done same); see also Erickson, 513 N.E.2d at 374–75 ([T]he proper test of retroactivity is the three-pronged test set forth in Stovall v. Denno . . . . ”); Sexton, 580 N.W. 2d at 409–16 (“While we acknowledge the reasoning and rationale of Griffith, we decline to apply it to the cases before us today. Instead, we find the analysis in Stovall v. Denno, to be more persuasive.”); Hudson, 415 S.E.2d at 751 (refusing to extend Griffith to state law); Gordon, 913 P.2d at 353–55 (same). At least one state retroactively applies all newly declared statutory rules but uses the Stovall approach when a new common law rule constitutes a clear break from past precedent. See Donaldson v. Superior Court, 672 P.2d 110, 117–18 (Cal. 1984) (explaining California approach to retroactivity). Potential distinctions between statutory and common law rulings for retroactivity analysis are discussed further in section II.D. even though the Supreme Court has abandoned this test for federal retroactivity. 91 See Griffith, 479 U.S. at 326–28, overruling Stovall, 388 U.S. 293 (rejecting case-specific analysis as inappropriate for direct review). The Stovall test considers: “(1) the purpose to be served by the new ruling, (2) the extent of the reliance by law enforcement authorities on the old stan­dards, and (3) the effect on the administration of justice of a retroactive application of the new standards.” 92 Stovall, 388 U.S. at 297. This test considers only the characteristics of the rule sought to be retroactively applied and not its effect in a particular case. So this test does not allow a court to retroactively apply the same rule in one case but not in another depending on each case’s particular facts. Most states that use the Stovall test place the greatest emphasis on its first prong, usually only retroactively applying rules they deem integral to the truth-finding function of a trial. 93 See, e.g., People v. Kaanehe, 559 P.2d 1028, 1034 (Cal. 1977) (“Decisions [are] made fully retroactive only where the right vindicated is . . . essential to the integrity of the fact-finding process. On the other hand, retroactivity is not customarily required when the [right] is . . . merely collateral to a fair determination of guilt or innocence.”); Erickson, 513 N.E.2d at 375 (describing first factor of Stovall test as concerned with whether new rule enhances truth-seeking process, and noting retroactive application is unlikely when new rule does not bear directly upon jury’s truth-seeking role); Taylor, 10 S.W.3d at 683 (“In line with . . . other jurisdictions, . . . a Stovall inquiry will generally center around whether the new rule significantly impacts the truth-finding function of the trial: the new rule should generally be retroactive if it does, non-retroactive if it does not.”); Farbotnik v. State, 850 P.2d 594, 603 (Wyo. 1993) (holding new rule, which required a complete record, not “sufficiently significant with respect to fact finding at trial to warrant retrospective application of that rule”). The final two Stovall factors are generally only relevant in extreme cases or where the first factor does not dictate a clear result. 94 See, e.g., In re Johnson, 475 P.2d 841, 844 (Cal. 1970) (“It is also clear that the factors of reliance and burden on the administration of justice are of significant relevance only when the question of retroactivity is a close one after the purpose of the new rule is considered.”); People v. Douglas, 617 N.Y.S.2d 733, 739 (App. Div. 1994) (refusing to retroactively apply new requirement for proof of knowledge of weight in drug possession cases because of huge number of cases potentially affected); Taylor, 10 S.W.3d at 683 (“The result may differ, however, if in a given situation, the State is found to have unusually strong reliance interests, or the burdens upon the administration of justice are shown to be unusually high.”).

In contrast, only Arizona, Florida, Iowa, and Wisconsin have adopted Griffith v. Kentucky’s bright-line rule of retroactivity for new state criminal procedure rulings on direct appeal. 95 See State v. Gardfrey, 775 P.2d 1095, 1097 n.1 (Ariz. 1989) (adopting bright-line retroactivity rule for state criminal procedural rulings on direct review ); Smith v. State, 598 So. 2d 1063, 1066 (Fla. 1992) (same); State v. Royer, 436 N.W.2d 637, 640 n.2 (Iowa 1989) (same); State v. Koch, 499 N.W.2d 152, 158 (Wis. 1993) (same). When these states’ highest courts declare a new criminal procedural rule, courts within these states must subsequently apply the rule to all direct appeals, regardless of the rule’s purpose, the extent of reliance on the previous rule, or the effect of retroactive application on the administration of justice.

The Stovall three-factor test and the Griffith bright-line rule produce very different results in practice. State courts employing the Stovall three-factor test rarely find a new state criminal procedural ruling to be sufficiently integral to the truth-finding process to merit retroactive application. 96 See, e.g., Erickson, 513 N.E.2d at 374–75 (declining to retroactively apply new statutory right to waive capital sentencing jury prior to commencement of trial); People v. Sexton, 580 N.W.2d 404, 409–16 (Mich. 1998) (declining to retroactively apply new rule that police must inform suspect when retained counsel is available for consultation); State v. Hudson, 415 S.E.2d 732, 751 (N.C. 1992) (declining to retroactively apply new rule eliminating perjury as aggravating factor in capital punishment case); State v. Gordon, 913 P.2d 350, 353–55 (Utah 1996) (declining to retroactively apply new common law rule prohibiting part-time prosecutor from representing criminal defendant as defense counsel). But see infra note 97 (discussing three cases that used Stovall test and retroactively applied new rules). These courts generally only retroactively apply new rules related to the admissibility of evidence at trial 97 Three cases serve as good examples of this. In People v. Favor, the New York Court of Appeals retroactively applied a new decision that established a criminal defendant’s right to be present during evidentiary hearings meant to decide whether to permit the prosecution to raise the defendant’s prior convictions and bad acts on cross-examination if the defendant elected to testify. 624 N.E.2d 631, 637–38 (N.Y. 1993).
In People v. Guerra, the California Supreme Court retroactively applied its own ban on hypnosis-induced testimony to a direct appeal and reversed two defendants’ rape convictions. 690 P.2d 635, 651–52 (Cal. 1984). Hypnosis-induced testimony from the victim of the alleged rape had been critical to the convictions. Id. at 640–42.
In Taylor, the Texas Court of Criminal Appeals (Texas’s highest criminal court) determined that its previous decision abolishing a juvenile exception to Texas’s common law “accomplice witness” evidentiary rule was sufficiently important to merit application. 10 S.W.3d at 681–85. The Texas “accomplice witness” rule entitles a criminal defendant to have the jury instructed that it may not convict the defendant on the basis of an alleged accomplice’s testimony unless there is other evidence tending to connect the defendant to the offense. Id. at 677 n.1. Taylor was accused of manipulating his eleven- and ten-year-old children into setting fire to their mother and stepfather’s home, with the intent to kill them and their son. Id. at 676. Taylor’s two children testified against him at trial, and Taylor’s request for an “accomplice witness” instruction was denied. Id. He was then convicted in part due to the testimony of his children. Id. A year after Taylor’s conviction, while his case was still pending direct appeal, the Texas Court of Criminal Appeals (the state’s highest court for criminal cases) abolished the juvenile exception in another case, ruling that it contradicted the purpose of the accomplice witness rule and was not supported by the courts’ previous decisions or Texas statutes. Id. at 676–77. Taylor then asked the intermediate appellate court to apply this new ruling to his direct appeal. Id. at 677. His case reached the Court of Criminal Appeals, which determined that the abolition of the juvenile exception to the accomplice witness rule significantly impacts the truth-finding process at trial and reversed Taylor’s conviction. Id. at 681–85.
—but they sometimes decline to apply even seemingly important evidence rules on direct appeal. 98 See, e.g., Mason v. Duckworth, 74 F.3d 815, 818 (7th Cir. 1996) (discussing Indiana appellate court’s refusal to retroactively apply new common law rule that prior inconsistent statements of witness must have been stated under oath in judicial proceeding to be admissible). In addition, Stovall courts almost never retroactively apply procedural rights newly granted to defendants that they deem collateral to a fair determination of the defendant’s guilt or innocence, like a right to be present for certain voir dire questioning 99 See People v. Mitchell, 606 N.E.2d 1381, 1384 (N.Y. 1992) (declining to retroactively apply new decision that sidebar conferences with prospective jurors on issue of bias violated defendant’s statutory right to be present for voir dire). and a rule requiring judges to conduct a voir dire to determine the voluntariness of statements a defendant made to private individuals when the issue is raised. 100 See People v. Carrera, 777 P.2d 121, 140 (Cal. 1989) (refusing to retroactively apply statutory right to privacy of pretrial detainee); Commonwealth v. Waters, 506 N.E.2d 859, 862–63 (Mass. 1987) (deciding not to retroactively apply new judge-made rule requiring judges to conduct voir dire to assess voluntariness of statements defendant made to private individuals when issue is raised). For other cases where state courts declined to retroactively apply newly declared procedural protections to defendants’ direct appeals, see Sexton, 580 N.W.2d at 409–16 (declining to retroactively apply judge-made rule requiring police to inform suspect when retained counsel is available for consultation); Gordon, 913 P.2d at 353–55 (refraining from retroactively applying new common law rule prohibiting part-time prosecutor from representing criminal defendant as defense counsel). Stovall courts also generally do not retroactively apply new rulings requiring particular jury instructions 101 See, e.g., Lackey v. Scott, 28 F.3d 486, 491 (5th Cir. 1994) (affirming Texas courts’ refusal to retroactively apply new rule requiring trial court to define “reasonable doubt” for jury); Murtishaw v. State, 773 P.2d 172, 178–79 (Cal. 1989) (declining to retroactively apply new requirement that trial court must instruct jury that honest unreasonable belief of need for self-defense negates malice in murder trial when evidence warrants it); Cooper v. State, 889 P.2d 293, 307–10 (Okla. Crim. App. 1995) (deciding not to bar jury instruction on flight, absent defendant’s explanation for flight, in direct appeal despite new statutory rule requiring courts to do so). or regulating sentencing procedures. 102 See, e.g., State v. Hudson, 415 S.E.2d 732, 751 (N.C. 1992) (refusing to retroactively apply new ruling eliminating perjury as aggravating factor in sentencing); State v. Cowans, 717 N.E.2d 298, 314 (Ohio 2000) (declining to retroactively apply rule requiring trial court to determine, when a capital defendant evinces desire to waive mitigation evidence, that waiver is knowing and voluntary and defendant is competent to effect waiver). Conversely, state courts employing the Griffith bright-line rule always retroactively apply state criminal procedural rules. 103 See supra note 95 and accompanying text (listing state courts that have adopted bright-line retroactivity rule for state criminal procedural rulings on direct review). Therefore, a state court’s choice of whether to employ the Stovall three-factor test or the Griffith bright-line rule often dictates whether the court will retroactively apply a given rule, and thus whether it will reverse a conviction.

Section II.B examines why the vast majority of state courts use the Stovall three-factor test to determine whether to retroactively apply new state criminal procedural rules to convictions on direct appeal and considers normative arguments supporting this approach.

B. The Majority Approach: The Stovall Three-Factor Test

The choice of most states to use the Stovall three-factor test 104 See supra note 92 and accompanying text (articulating three-factor test). reflects two beliefs: that the U.S. Constitution does not require state courts to retroactively apply new state criminal procedural rules in such cases; and that the Stovall three-factor test is superior to Griffith’s bright-line rule of retroactivity as a policy matter. 105 See, e.g., Taylor v. State, 10 S.W.3d 673, 682–83 (Tex. Crim. App. 2000) (explaining policy justification for Stovall three-factor test). This section examines the rationales behind both beliefs.

1. Griffith Has No Effect on State Law. — Defendants like Carlson 106 See supra notes 1–3 and accompanying text (discussing Carlson hypothetical). often argue that Griffith requires state courts to apply its bright-line retroactivity rule to criminal procedural rulings. 107 See, e.g., State v. Hudson, 415 S.E.2d 732, 751 (N.C. 1992) (describing defendant’s argument that Griffith requires state courts to retroactively apply new state criminal procedural rulings); Farbotnik v. State, 850 P.2d 594, 601–02 (Wyo. 1993) (same). State courts employing the Stovall three-factor test always reject this assertion, relying on the fact that the Supreme Court has never required a state court to retroactively apply any state ruling. 108 See, e.g., People v. Mitchell, 606 N.E.2d 1381, 1384 (N.Y. 1992) (stating Supreme Court does not require state court to retroactively apply any state law); State v. Cowans, 717 N.E.2d 298, 314 (Ohio 1999) (same); Taylor, 10 S.W.3d at 679 (“Conversely, Griffith and Teague do not bind the states on the retroactivity of new rules under state law.”). The Supreme Court last ruled about state-law retroactivity in 1973, when it held that the U.S. Constitution does not require state courts to retroactively apply any state-law ruling. See Wainwright v. Stone, 414 U.S. 21, 23–24 (1973) (holding U.S. Constitution did not require Florida Supreme Court to retroactively apply state criminal substantive rule to convictions on collateral review). While Griffith and Harper are ambiguous in regard to whether they apply to all federal law or only to federal-constitutional law, see infra section III.A, they do not purport to affect state law. See Harper v. Va. Dep’t. of Taxation, 509 U.S. 86, 100 (1993) (referencing state-law retroactivity, but not purporting to affect it); Griffith v. Kentucky, 479 U.S. 314, 314–28 (1987) (omitting discussion of state-law retroactivity). These state courts sometimes cite a line of dicta 109 This line is dicta because the case was about the retroactivity of a federal-constitutional rule. State-law retroactivity was not at issue. See Am. Trucking Ass’ns v. Smith, 496 U.S. 167, 171 (1990) (plurality opinion) (stating issue in case was whether to retroactively apply Supreme Court decision about meaning of U.S. Constitution). For a discussion of the difference between holding and dicta for precedential purposes, see Bradley Scott Shannon, The Retroactive and Prospective Application of Judicial Decisions, 26 Harv. J.L. & Pub. Pol’y 811, 845–51 (2003). from a 1990 plurality opinion written by Justice O’Connor to show that, even in the wake of Griffith, state courts still have the authority to determine the retroactivity of their own decisions. 110 See Am. Trucking Ass’ns, 496 U.S. at 177 (plurality opinion) (“When questions of state law are at issue, state courts generally have the authority to determine the retroactivity of their own decisions.” (citing Great N. Ry. Co. v. Sunburst Oil & Ref. Co., 287 U.S. 358, 364 (1932))). Accordingly, these courts choose to use the Stovall three-factor test, not Griffith’s bright-line retroactivity rule, for the policy reasons discussed in the next section.

2. Normative Arguments Supporting the Stovall Three-Factor Test. — The Stovall three-factor test 111 For an articulation of the test, see supra note 92 and accompanying text. reflects a belief that retroactive application on direct appeal, and the reversal and retrial it might require, is only appropriate when a new rule raises doubts about the accuracy of past convictions. 112 See, e.g., Taylor, 10 S.W.3d at 681 (“We reject Griffith’s approach as too rigid. Fairly assessing whether a new rule should be given retroactive effect requires examining the various interests in favor of and against retroactivity, and Stovall’s three-factor balancing test accurately encompasses the competing interests involved.”). It allows courts to only retroactively apply new rules that significantly impact the truth-finding process of a trial. 113 See supra notes 97–98 and accompanying text (detailing how rare circumstances are in which courts using Stovall test retroactively apply new rules). For examples of rules that courts using the Stovall three-factor test have not retroactively applied, see supra notes 99–102; see also, e.g., Donaldson v. Superior Court of L.A. Cty., 672 P.2d 110, 118–19 (Cal. 1983) (noting Stovall test usually denies retroactivity to search and seizure holdings but affords retroactivity to decisions that implicate question of guilt and innocence); Taylor, 10 S.W.3d at 683 (“[I]f the new rule does not significantly impact the jury’s truth-finding function, the Stovall factors will generally balance in favor of prospective application: ordinarily, concerns regarding reliance and administration of justice will be enough to tip the scales of the balancing test against retroactivity.”); Shannon, supra note 109, at 865–71 (discussing efficiency as benefit of purely prospective application of new rules). State courts’ emphasis on whether new procedural rulings impact the truth-finding process of a trial suggests a belief that retroactive application, and the reversal and re-trial it might require, are only appropriate when a new ruling affects the probability of a convicted defendant’s guilt. 114 This may explain the results in Taylor v. State and People v. Favor, two of the rare cases where a state court applied the Stovall three-factor test and determined a new state criminal procedural ruling merited retroactivity. See supra note 97 and accompanying text (providing details of these two cases). In Taylor v. State, the accomplice witness rule was the product of a legislative judgment that testimony of an individual who could be charged with commission of the same crime with which the defendant is charged is inherently suspect, and the jury should be prevented from convicting defendants where they find no evidence that corroborates the accomplice witness’s testimony. See 10 S.W.3d at 684–85 (explaining why accomplice witness rule should be retroactively applied). The Texas Court of Criminal Appeals’s elimination of the previously existing exception to this rule when the witness in question was a minor served this end and was therefore intended to substantially improve fact-finding at trial by reducing the likelihood that an innocent defendant would be wrongly convicted due to testimony from a witness who was not credible. Id.
The new ruling at issue in People v. Favor—establishing a criminal defendant’s right to be present during evidentiary hearings meant to decide whether to permit the prosecution to raise the defendant’s prior convictions and bad acts on cross-examination if the defendant elects to testify—was likely intended to ensure that New York courts did not admit unfairly prejudicial evidence against criminal defendants that could potentially result in false convictions, given the effect that evidence of defendant’s prior convictions and bad acts can have on a jury. See 624 N.E.2d 631, 637–38 (N.Y. 1993) (explaining why court had to retroactively apply new ruling and reverse defendant’s direct appeal).

In addition, nonretroactivity gives judges more freedom to proactively declare new procedural protections for defendants without worrying about the accompanying efficiency costs of retroactive application. 115 See Jenkins v. Delaware, 395 U.S. 213, 217–18 (1969) (describing this freedom as benefit of nonretroactivity). The efficiency costs of retroactively applying new rules on direct appeal are much lower than on collateral review, however, since defendants have a limited window of time between their conviction at trial and the moment their conviction becomes final. This drastically limits the class of defendants who can receive the benefit of a newly declared criminal procedural rule. The Warren Court described nonretroactivity as a technique that allowed it to implement criminal procedure reforms without having to apply these reforms to every single similarly situated defendant on direct appeal. 116 Id. at 218. The Miranda rule is one such reform the Court chose not to retroactively apply to avoid efficiency costs. See Johnson v. New Jersey, 384 U.S. 719, 721 (1966) (declaring Miranda rule did not apply retroactively to direct appeals). Justice Harlan alternatively described the Warren Court’s embrace of prospectivity as “the product of the Court’s disquietude with the impacts of its fast-moving pace of constitutional innovation.” Mackey v. United States, 401 U.S. 667, 672 (1971) (Harlan, J., concurring in the judgment). Courts justify providing the benefit of a new rule to the defendant who procures it, 117 These defendants are like Marks in the introductory hypothetical. See supra notes 1–3 and accompanying text (discussing hypothetical). but not to other similarly situated defendants, 118 These defendants are like Carlson in the introductory hypothetical. See supra notes 1–3 and accompanying text (discussing hypothetical). by saying that providing such a windfall to the former is necessary to provide an incentive to challenge bad precedent. 119 See, e.g., People v. Carrera, 777 P.2d 121, 143 (Cal. 1989) (describing unequal treatment of similarly situated defendants as “unavoidable consequence of dispensing justice on a societal basis”); Taylor, 10 S.W.3d at 682–83 (“In a sense, the party procuring the new rule gains a windfall, required to encourage challenges to bad precedent, that we are not required to extend to similarly-situated individuals.”); Proctor v. State, 967 S.W.2d 840, 845 n.5 (Tex. Crim. App. 1998) (arguing possibility of windfall provides necessary incentive). Griffith rejected this unequal treatment of similarly situated defendants in the context of new federal-constitutional rulings of criminal procedure, however. 120 See supra notes 53–54 and accompanying text (describing Griffith’s equal treatment principle). Section II.C discusses the state courts that, inspired by Griffith, retroactively apply all of their own state criminal procedural rules to direct appeals.

C. The Minority Approach: Griffith’s Bright-Line Rule

Only four states retroactively apply all new state-law criminal procedural rulings, without case-specific analysis, to direct appeals: Arizona, Florida, Iowa, and Wisconsin. 121 See supra note 95 (listing state courts that have adopted bright-line retroactivity rule for state criminal procedural rulings on direct review). The highest courts in each of these four states adopted the Griffith bright-line rule 122 For ease of reading, the “Griffith bright-line rule” refers to a rule requiring retroactive application of all criminal procedural rules. But it is not clear whether Griffith itself requires retroactive application only of federal-constitutional rules or of all federal rules. See infra section III.A (describing circuit split on Griffith’s application to federal-nonconstitutional rules, but arguing Griffith does apply to such rules). So one could argue that a state court’s decision to retroactively apply both state-constitutional and state-nonconstitutional rules represents an adoption of a rule broader than Griffith. in the six years following Griffith without direction from their state legislatures. 123 See State v. Gardfrey, 775 P.2d 1095, 1097 n.1 (Ariz. 1989) (adopting Griffith rule without mentioning direction from legislature); Smith v. State, 598 So. 2d 1063, 1066 (Fla. 1992) (same); State v. Royer, 436 N.W.2d 637, 640 n.2 (Iowa 1989) (same); State v. Koch, 499 N.W.2d 152, 158 (Wis. 1993) (same). None of these courts appear to distinguish between procedural rulings derived from a state constitution, a state statute, or from state common law. 124 See Gardfrey, 775 P.2d at 1097 n.1 (neglecting to reference difference between constitutional, statutory, or common law rules for retroactivity purposes); Smith, 598 So. 2d at 1066 (same); Royer, 436 N.W.2d at 640 n.2 (same); Koch, 499 N.W.2d at 158 (same). Arizona and Florida have retroactively applied new statutory rules. Gardfrey, 775 P.2d at 1097; Smith, 598 So.2d at 1066. Iowa has retroactively applied a new common law rule. Royer, 436 N.W.2d at 640 n.2. Wisconsin has retroactively applied a new constitutional rule, but it does not appear to have encountered a case that required it to explicitly decide whether to retroactively apply state statutory or common law rules since it adopted the Griffith rule. Koch, 499 N.W.2d at 158. This section examines the rationales behind these four states’ choices and normative arguments supporting the Griffith bright-line rule and retroactive application generally, as well as arguments against them.

None of these four state courts held that the U.S. Constitution required them to apply the Griffith rule for state rules of criminal procedure. Iowa and Arizona’s supreme courts explicitly stated that the U.S. Constitution does not require them to retroactively apply state rules. 125 Gardfrey, 775 P.2d at 1097 n.1; Royer, 436 N.W.2d at 640 n.2. The Iowa Supreme Court favorably cited Griffith’s Article III rationale about the “nature of judicial review” requiring a new rule to be applied to all similar cases pending on review. Id. Since Article III does not apply to state courts, the Iowa Supreme Court presumably either felt that Iowa’s constitution required this bright-line rule or simply thought it was good policy. The Arizona Supreme Court said that the “principles of fairness” articulated in Griffith mandated retroactive application. Gardfrey, 775 P.2d at 1097 n.1. The Florida Supreme Court appeared to hold that Florida’s constitution required the court to adopt the Griffith bright-line rule because the due process and equal protection provisions contained in Florida’s constitution embodied the same “principles of fairness and equal treatment underlying Griffith.” 126 Smith, 598 So. 2d at 1066. Meanwhile, the Wisconsin court stated that “the Griffith rule . . . is applicable in Wisconsin” but did not explain why. 127 Koch, 499 N.W.2d at 158.

One normative argument for a bright-line rule of retroactive application is Griffith’s equal treatment principle: Courts should not treat similarly situated litigants differently due only to timing. 128 See Sosnov, supra note 56, at 317–20 (“If one goal of the Due Process Clause is to impose fundamental fairness restraints on state action, including the state appellate process, the state arguably should be required to apply new judicially created state criminal rules to all defendants whose cases are pending on direct review.”). An obvious counterargument is that the Griffith bright-line rule does not solve the unequal treatment problem, but merely shifts its line of demarcation to the date at which convictions become final. 129 See Griffith v. Kentucky, 479 U.S. 314, 331–32 (1987) (White, J., dissenting) (arguing distinguishing between direct and collateral review is misguided and same problem of equal treatment will occur with respect to direct and collateral appeals); supra section I.C.2 (showing Griffith rule results in courts applying new procedural rules to direct, but not collateral appeals). In states that retroactively apply state criminal procedural rules on direct, but not collateral appeal, a defendant whose conviction becomes final one day before the new rule was announced will be denied the benefit of the new rule, only because the court processed his direct appeal faster than those of other defendants. So the Griffith rule does not necessarily solve one of the problems it is designed for, but merely shifts its effects onto a different stage of criminal litigation. Some defendants will still fail to get the benefit of new criminal procedural rules due to unlucky timing. But state courts’ adoption of the Griffith rule would at least produce consistency between federal and state law. In states using the Griffith rule, the line of demarcation at which defendants will no longer be able to get the benefit of a new criminal procedural rule is the same regardless of whether the new rule is derived from federal or state law.

One might argue that consistency in retroactivity doctrine between federal and state law is desirable in itself. 130 Cf. infra Appendix (showing disparity in extent to which courts retroactively apply new federal and state rulings). It seems arbitrary to constitutionally guarantee defendants like Carlson 131 See supra notes 1–3 and accompanying text (providing introductory hypothetical). the benefit of a new rule if that rule comes from federal law, 132 See Griffith, 479 U.S. at 322 (holding criminal defendants get benefit of new federal-constitutional procedural rules on direct review). but not if the new rule comes from state law. 133 See People v. Sexton, 580 N.W.2d 404, 417–18 (Mich. 1998) (Brickley, J., dissenting) (arguing Michigan should adopt Griffith rule); Sosnov, supra note 56, at 317–20 (“If one goal of the Due Process Clause is to impose fundamental fairness restraints on state action, including the state appellate process, the state arguably should be required to apply new judicially created state criminal rules to all defendants whose cases are pending on direct review.”). There does not seem to be anything unique about a federal rule that dictates that it should apply to direct appeals, but a state rule should not. 134 Cf. David Lehn, Adjudicative Retroactivity as a Preclusion Problem: Dow Chemical Co. v. Stephenson, 59 N.Y.U. Ann. Surv. Am. L. 563, 565 (2004) (arguing such distinctions are arbitrary and should be irrelevant in retroactivity analysis). But there is arguably something unique about the character of a federal-constitutional rule. See Danforth v. Minnesota, 552 U.S. 264, 271 & n.5 (2008) (“[T]he source of a ‘new rule’ is the Constitution itself, not any judicial power to create new rules of law. Accordingly, the underlying right necessarily pre-exists our articulation of the new rule.”). A state judicial ruling is just as much “law” as a federal ruling. The Griffith bright-line rule is also far easier to consistently apply than the Stovall three-factor test because it does not require case-specific analysis of the extent to which a rule improves the truth-finding process. 135 See Sexton, 580 N.W.2d at 419–23 (Brickley, J., dissenting) (claiming Michigan courts inconsistently apply Stovall three-factor test and arguing Griffith rule is easier to apply). The Griffith rule does require case-specific analysis of the appropriate remedy once the court retroactively applies a new rule to a direct appeal, however. 136 See infra section III.C (discussing practical impact of Griffith rule, including remedy once rule is retroactively applied).

Scholars have advanced many arguments in favor of retroactive application generally. 137 Retroactivity on collateral review is a cause célèbre among academics. Most retroactivity scholarship argues that courts should retroactively apply new decisions in more situations than they do now. See, e.g., James S. Liebman & William F. Ryan, Some Effectual Power: The Quantity and Quality of Decisionmaking Required of Article III Courts, 98 Colum. L. Rev. 696, 873–74 (1998) (contending Article III does not permit habeas courts to defer to state court decisions that were erroneous when rendered); Kermit Roosevelt III, A Retroactivity Retrospective, with Thoughts for the Future: What the Supreme Court Learned from Paul Mishkin, and What It Might, 95 Calif. L. Rev. 1677, 1697 (2007) (“[C]ourts should decide cases according to their best current understanding of the law, and that the question of whether a rule has ‘retroactive effect’ should simply be excised from our doctrine.”); Shannon, supra note 109, at 836–71 (promoting rule of retroactive application in all cases). One common argument is that retroactive application is more compatible with the adjudicative function of courts than prospectivity. 138 E.g., Shannon, supra note 109, at 838–45; see also Mishkin, supra note 14, at 65 (observing prospectivity “smacks of the legislative process”). Declaring a new precedent and not adhering to it on a subsequent direct appeal arguably violates traditional conceptions of the adjudicative function. 139 See, e.g., Griffith v. Kentucky, 479 U.S. 314, 322 (1987) (saying “nature of judicial review” requires retroactivity on direct review); Shannon, supra note 109, at 838–45 (arguing prospectivity is legislative, not adjudicative, technique); Robert Von Moschzisker, Stare Decisis in Courts of Last Resort, 37 Harv. L. Rev. 409, 428 (1924) (arguing prospectivity “encroach[es] on the prerogatives of the legislative department of government”). According to this line of thought, an appellate court’s job should be to decide the legal issues before it, not grade the performance of the trial court given the state of the law at the time of trial. 140 Article III applies this principle to federal courts. See supra notes 51–52 (discussing first Griffith rationale). But it does not apply to state courts. Commentators typically do not argue that the U.S. Constitution requires state courts to adopt this principle, but rather that they should adopt it as a matter of good policy. See, e.g., Shannon, supra note 109, at 838–45 (discussing courts’ adjudicative role). Some also argue that a refusal to retroactively apply a new decision on direct review violates principles of stare decisis. 141 See Harper v. Va. Dep’t of Taxation, 509 U.S. 86, 106–10 (1993) (Scalia, J., concurring) (describing nonretroactivity as method of destroying stare decisis); Shannon, supra note 109, at 851–62 (arguing prospectivity and stare decisis are conflicting principles).

Some commentators have also argued that nonretroactivity 142 “Nonretroactivity” refers to the general concept of not retroactively applying new rules to past cases. It includes both selective and full prospectivity. is undesirable because it enables judicial activism. 143 See, e.g., Harper, 509 U.S. at 106–10 (Scalia, J. concurring) (rejecting nonretroactivity as “jurisprudential tool[] of judicial activism”); Mackey v. United States, 401 U.S. 667, 680 (1971) (Harlan, J., concurring in the judgment) (describing nonretroactivity as tool courts use “to cut . . . loose from the force of precedent, allowing [them] to restructure artificially those expectations legitimately created by extant law and thereby mitigate the practical force of stare decisis”). For further explanation of how nonretroactivity allows courts to minimize the efficiency costs of overruling precedent, see supra notes 115120 and accompanying text. These commentators favor retroactive application because they believe it deters judges from overruling precedent by forcing them to bear the efficiency costs of retroactive application. 144 See, e.g., Harper, 509 U.S. at 106–10 (Scalia, J. concurring) (favoring retroactivity because it cabins judicial activism). Many other scholars view this deterrence as a bad thing, arguing prospectivity is a device necessary to “augment[] the power of the courts to contribute to the growth of the law in keeping with the demands of society.” 145 Jonathan Mallamud, Prospective Limitation and the Rights of the Accused, 56 Iowa L. Rev. 321, 359 (1970); see also, e.g., Beryl Harold Levy, Realist Jurisprudence and Prospective Overruling, 109 U. Pa. L. Rev. 1, 6, 28 (1960) (describing nonretroactivity as “deliberate and conscious technique of judicial lawmaking” that “facilitat[es] more effective and defensible judicial lawmaking”).

Part II has thus far discussed state criminal procedural rulings generally. But some courts treat state-constitutional, statutory, and common law rules differently in retroactivity analysis. 146 See People v. Lopez, 27 Cal. Rptr. 2d 25, 26–27 (Ct. App. 1993) (noting statutory interpretations are not new rules of law, and thus court must apply them retroactively). See generally infra Appendix (displaying differences in how courts apply rulings derived from different sources). Section II.D discusses state courts’ treatment of criminal procedure rulings derived from different sources and potential differences in how and when state constitutional, statutory, and judge-made rules should be retroactively applied.

D. The Source of a State Criminal Procedural Rule: Does It Matter?

Many state courts largely appear to treat new state-constitutional, statutory, and judge-made rules equally in retroactivity analysis. 147 See Taylor v. State, 10 S.W.3d 673, 680–81 (Tex. Crim. App. 2000) (showing many states apply Stovall three-factor test to new statutory interpretations and judge-made rules). There is not much case law discussing the retroactivity of state-constitutional rules, but courts generally do not acknowledge the source of a state criminal procedural rule as being relevant in retroactivity analysis. See, e.g., Cowans v. Bagley, 624 F. Supp. 2d 709, 735–40 (S.D. Ohio 2008) (disregarding source of state rule as irrelevant in retroactivity analysis); People v. Sexton, 580 N.W.2d 404, 409–16 (Mich. 1998) (same); Taylor, 10 S.W.3d at 680–81 (same). Only a few courts have acknowledged differences between these rules. Some state courts retroactively apply decisions interpreting statutes that courts had not previously interpreted, reasoning that a first-time interpretation of a statute is not a “new rule” that triggers the Stovall three-factor test. 148 See, e.g., Taylor, 10 S.W.3d at 681–82 (“A first time interpretation, even if unanticipated by the parties in the case, cannot be considered a new rule because, presumably, the Legislature intended that interpretation when it enacted the statute.”). For example, California goes even further and applies all new statutory interpretations to direct appeals, even if they overrule previous decisions, 149 Lopez, 27 Cal. Rptr. 2d at 26–27 (“[W]henever a decision undertakes to vindicate the original meaning of an enactment, putting into effect the policy intended from its inception, retroactive application is essential to accomplish that aim.” (quoting People v. Garcia, 684 P.2d 826, 831 (Cal. 1984))). but takes a more nuanced approach to constitutional 150 See People v. Bustamante, 634 P.2d 927, 935–36 (Cal. 1981) (en banc) (declaring new California constitutional rule that defendant has right to counsel at pre-indictment lineup, but announcing it would only apply prospectively), superseded by constitutional amendment, Cal. Const. art. I, § 28. and judge-made rules. 151 California courts retroactively apply judge-made rules when the California Supreme Court resolves a conflict between lower court decisions, addresses an issue not previously presented to the courts, or when it vacates a Court of Appeal decision by a grant of review, because “[i]n each of these cases there was no clear rule on which anyone could have justifiably relied.” Lopez, 27 Cal. Rptr. 2d at 27–28 (citing People v. Guerra, 690 P.2d 635, 640 (Cal. 1984) (en banc)). Michigan has suggested that it might treat judge-made rules uniquely as well, giving consideration to whether they were foreshadowed by previous decisions. 152 See Sexton, 580 N.W.2d at 411–15 (adopting Stovall three-factor test and declining to retroactively apply new judge-made rule in part due to lack of sufficient foreshadowing). But see id. at 419 (Brickley, J., dissenting) (arguing Michigan Supreme Court never finds a new judge-made rule to have been sufficiently foreshadowed).

There is some logic to treating constitutional, statutory, and judge-made rules differently in retroactivity analysis. 153 See, e.g., Roosevelt, A Little Theory, supra note 14, at 1076 (arguing new statutory interpretations should certainly apply retroactively in all situations, but retroactivity of constitutional and judge-made rules is more complex); cf. Danforth v. Minnesota, 552 U.S. 264, 271 & n.5 (2008) (“[T]he source of a ‘new rule’ is the Constitution itself, not any judicial power to create new rules of law. Accordingly, the underlying right necessarily pre-exists our articulation of the new rule.”). State courts that distinguish between them often run into practical difficulties in identifying a new rule’s source of law, however. 154 See, e.g., People v. Carrera, 777 P.2d 121, 142–43 (Cal. 1989) (finding statute inspired but did not require new rule, so rule did not retroactively apply to direct appeals); Sexton, 580 N.W.2d at 419–23 (Brickley, J., dissenting) (identifying difficulties Michigan Supreme Court has had in determining whether given rule is judge-made or derived from constitution). Some courts have particularly struggled in determining whether rules intended to promote important constitutional or statutory interests are mandated or merely prophylactic. 155 See, e.g., Sexton, 580 N.W.2d at 419–23 (Brickley, J., dissenting) (identifying difficulties Michigan Supreme Court has had determining whether given rule is mandated by, or merely intended to promote, interests of state constitution or statutes); State v. Gordon, 913 P.2d 350, 353–55 (Utah 1996) (demonstrating obscurity about whether rule prohibiting part-time prosecutor from representing criminal defendant is rooted in constitutional right to counsel, or merely prophylactic). A full discussion of differences between rules derived from different sources of law and how these differences impact retroactivity analysis is outside the scope of this Note. 156 For such a discussion, see generally Roosevelt, A Little Theory, supra note 14, at 1075–85 (explaining differences between statutory, constitutional, and judge-made law for retroactivity purposes).

Part II has discussed how state courts have approached decisions about whether to retroactively apply new state criminal procedural rulings to direct appeals. Part III argues that the U.S. Constitution requires these courts to retroactively apply all such rules.

III. Due Process Requires State Courts to Retroactively Apply All New State Criminal Procedural Rulings to Direct Appeals of Convictions

This Part examines Griffith’s scope: What kinds of criminal procedural rules does the U.S. Constitution require courts to apply on direct review? The three most likely answers to this question are that Griffith only applies to federal-constitutional rules; that it applies to all federal rules, but no state rules; and that it applies to all federal and state rules. This Part argues that the third and broadest view of Griffith’s scope is correct. But a threshold question that must be resolved before exploring the third view is whether Griffith requires, at a minimum, that courts retroactively apply all federal rules, not merely federal-constitutional rules. This is because both of Griffith’s rationales—Article III and the equal treatment principle—apply to federal law, yet only the latter applies to state law. 157 See infra section III.B (discussing Griffith’s two rationales and their application to state law). So for Griffith to apply to state rules, it must also apply to all federal rules. Section III.A analyzes this threshold issue.

A. Griffith Applies to All Federal Rules

Griffith indisputably requires courts to retroactively apply all federal-constitutional rulings of criminal procedure. 158 See 479 U.S. 314, 322–23 (1987) (stating U.S. Constitution requires courts to retroactively apply new federal criminal procedural rules rooted in U.S. Constitution on direct appeal). There is a circuit split on how broadly Griffith applies, however. The First, Fifth, Sixth, and Ninth Circuits have held that Griffith additionally applies to federal-nonconstitutional rulings. 159 See United States v. Mauldin, 109 F.3d 1159, 1161 (6th Cir. 1997) (applying new Supreme Court holding about meaning of “use” and “carry” in federal statute retroactively to defendant’s direct appeal); United States v. Rivas, 85 F.3d 193, 195 & n.1 (5th Cir. 1996) (same); United States v. Jones, 24 F.3d 1177, 1179 (9th Cir. 1994) (applying retroactively new Daubert test for admissibility of expert testimony to direct appeal and citing Griffith); United States v. Lopez-Pena, 912 F.2d 1542, 1545 (1st Cir. 1989) (applying new statutory interpretation retroactively to bar magistrate from presiding over jury selection to defendant’s direct appeal). The First Circuit has found “nothing in Griffith, either in terms or purport, distinguishing between constitutional and statutory interpretations.” Id. at 1545. The Fifth, Sixth, and Ninth Circuits stated without explanation that Griffith applied to federal law generally. See Mauldin, 109 F.3d at 1161 (stating, without reasoning, Griffith required court to retroactively apply new decision to defendant’s direct appeal); Rivas, 85 F.3d at 195 & n.1 (same); Jones, 24 F.3d at 1179 (same). The Fifth and Sixth Circuit opinions were about substantive criminal rules, but they still cited Griffith. See Mauldin, 109 F.3d at 1161 (applying retroactively new rule about essential elements of crime); Rivas, 85 F.3d at 195 (same). It is unclear whether either court would rule differently if faced with a procedural rule instead. But the Third Circuit has disagreed, ruling that “Griffith should be confined to constitutional rules of criminal procedure and thus does not require retroactive application of new procedural decisions not constitutionally grounded.” 160 Diggs v. Owens, 833 F.2d 439, 442 (3d Cir. 1987) (refusing to retroactively apply new ruling derived from Interstate Agreement on Detainers Act); see also United States v. Wilson, 429 F.3d 455, 458–60 (3d Cir. 2005) (refusing to retroactively apply new ruling derived from Rule 11 of the Federal Rules of Criminal Procedure, citing Diggs). The Third Circuit has historically been reluctant to retroactively apply new rules—it also found no due process violation in Fiore v. White and was reversed by the Supreme Court. 149 F.3d 221, 222 (3d Cir. 1998).

Whether Griffith applies to federal-nonconstitutional rulings turns on whether constitutional criminal procedural rules are special and should be retroactively applied in situations where an otherwise identical statutory or common law rule should not be. Since the 1960s, when modern retroactivity doctrine began to develop, most of the retroactivity issues the U.S. Supreme Court has addressed have involved new constitutional rules. 161 See Montgomery v. Lousiana, 136 S. Ct. 718, 732–36 (2016) (addressing new federal-constitutional rule); Harper v. Va. Dep’t of Taxation, 509 U.S. 86, 89 (1993) (same); Teague v. Lane, 489 U.S. 288, 305 (1989) (same); Griffith, 479 U.S. at 318 (same); Stovall v. Denno, 388 U.S. 293, 300 (1967) (same). This is likely because, since the start of the 1960s, criminal defendants’ constitutional rights have expanded much more than defendants’ statutory or common law rights, in ways that affect more cases. 162 See, e.g., Erwin Chemerinsky, The Roberts Court and Criminal Procedure at Age Five, 43 Tex. Tech. L. Rev. 13, 25 (2010) (describing Roberts Court’s expansion of defendants’ Confrontation Clause rights); Eric J. Miller, The Warren Court’s Regulatory Revolution in Criminal Procedure, 43 Conn. L. Rev. 1, 3–6 (2010) (describing Warren Court’s constitutional-rights-expanding jurisprudence in criminal procedure context). So the Supreme Court has developed federal retroactivity doctrine through the lens of constitutional rules and referred specifically to constitutional rules in Griffith, 163 Griffith, 479 U.S. at 318–22. Teague, 164 Teague, 489 U.S. at 305–10. and other cases. 165 E.g., Montgomery, 136 S. Ct. at 732–36.

But the Supreme Court has never suggested that nonconstitutional rulings are less worthy of retroactive application. For example, the source of a new rule was not a factor in the Stovall test the Supreme Court used before Griffith. 166 See supra note 42 and accompanying text (listing three Stovall factors). And Harper v. Virginia Department of Taxation, the Supreme Court’s most recent case about direct-appeal retroactivity, contained expansive language stating courts must apply all new rules of “federal [procedural] law” on direct review. 167 See 509 U.S. 86, 90 (1993) (“[W]e hold that this Court’s application of a rule of federal law to the parties before the Court requires every court to give retroactive effect to that decision.”); id. at 97 (“When this Court applies a rule of federal law to the parties before it, that rule is the controlling interpretation of federal law and must be given full retroactive effect in all cases still open on direct review . . . .”). Perhaps most significantly, Griffith’s two rationales apply with equal force to all new federal criminal procedure rulings. 168 See supra section I.C.1 (discussing Griffith’s two rationales). The “nature of [federal] judicial review” includes review for any kind of error, not merely constitutional error, and failure to retroactively apply a new nonconstitutional rule leads to the same unequal treatment of defendants discussed in Griffith. 169 See supra section I.C.1 (discussing Griffith’s two rationales); see also Lehn, supra note 134, at 565 (arguing such distinctions are arbitrary and should be irrelevant in retroactivity analysis). For these reasons, and because the vast majority of circuit courts agree, this Note assumes that Griffith extends to all new federal rulings of criminal procedure. The more novel question, which has not been addressed at length by courts or scholars, is whether Griffith extends to new state criminal procedure rules.

B. Griffith’s Equal Treatment Principle Applies to State Courts

Griffith is based on two constitutional principles. The first is Article III’s Cases and Controversies requirement: The nature of judicial review requires an appellate court to decide a case based on its best understanding of the governing legal principles. 170 See supra notes 51–52 and accompanying text (explaining Griffith’s Article III principle). The second is that courts must treat similarly situated defendants equally. 171 See supra notes 53–56 and accompanying text (explaining Griffith’s principle of equal treatment and uncertainty over which constitutional provision principle is derived from). Announcing and applying a new constitutional rule of criminal procedure to one defendant’s direct appeal, then refusing to apply the new rule in another defendant’s subsequent direct appeal, is inequitable because it arbitrarily rewards one defendant and punishes another.

Article III’s Cases and Controversies Clause does not impose requirements on state courts or state law. 172 See U.S. Const. art. III (creating U.S. Supreme Court and not mentioning state courts or state law). The Supreme Court is bound by the Cases and Controversies Clause when exercising its appellate jurisdiction over direct appeals of state decisions. But if state-law nonretroactivity is not a federal question—as the Supreme Court held in Wainwright v. Stone, 414 U.S. 21, 23–24 (1973)—the Supreme Court does not have authority to review it. Murdock v. City of Memphis, 87 U.S. (20 Wall.) 590, 592 (1875). So whether the Supreme Court can review state-law nonretroactivity today turns on whether state-law nonretroactivity implicates the constitutional principle of equal treatment discussed in Griffith and is thus a federal question. But the Fourteenth Amendment’s Due Process Clause does. 173 See U.S. Const. amend. XIV (“[N]or shall any State deprive any person of life, liberty, or property, without due process of law . . . .”); see also supra note 56 (discussing uncertainty over which constitutional provision Griffith’s equal treatment principle comes from, arguing it likely comes from either Due Process or Equal Protection Clause, both of which apply to states, and assuming for simplicity’s sake it comes from Due Process Clause). The Supreme Court has always interpreted the two Due Process Clauses identically, so if the Fifth Amendment requires federal courts to treat similarly situated defendants equally, the Fourteenth Amendment requires the same of state courts. 174 As Justice Frankfurter once said: “To suppose that ‘due process of law’ meant one thing in the Fifth Amendment and another in the Fourteenth is too frivolous to require elaborate rejection.” Malinski v. New York, 324 U.S. 401, 415 (1945) (Frankfurter, J., concurring). But see Ryan Williams, The One and Only Substantive Due Process Clause, 120 Yale L.J. 408, 415 (2010) (arguing only Fourteenth Amendment’s Due Process Clause encompasses substantive due process). In addition, Griffith does not indicate that its equal treatment principle only applies to federal-constitutional rules, or even just federal rules. 175 See infra notes 176–177 and accompanying text (showing Griffith’s focus on equal treatment of similarly situated defendants applies to state rules too). The Court said that “the problem with not applying new rules to cases pending on direct review is ‘the actual inequity that results when the Court chooses which of many similarly situated defendants should be the chance beneficiary’ of a new rule.” 176 Griffith v. Kentucky, 479 U.S. 314, 323 (1987) (quoting United States v. Johnson, 457 U.S. 537, 556 n.16 (1982)); see also Hankerson v. North Carolina, 432 U.S. 233, 247 (1977) (Powell, J., concurring in the judgment) (“[When] one chance beneficiary . . . enjoys retroactive application, while others similarly situated have their claims adjudicated under the old doctrine . . . [it] hardly comports with the ideal of administration of justice with an even hand.” (citations omitted) (internal quotation marks omitted) (quoting Desist v. United States, 394 U.S. 244, 255 (1969) (Douglas, J., dissenting))); Michigan v. Payne, 412 U.S. 47, 60 (1973) (Marshall, J., dissenting) (“Different treatment of two cases is justified under our Constitution only when the cases differ in some respect relevant to the different treatment.”); Desist, 394 U.S. at 258–59 (Harlan, J., dissenting) (“[W]hen a similarly situated defendant [to one who received the benefit of a new rule] comes before us, we must grant the same relief or give a principled reason for acting differently.”). This same inequity occurs when a state’s highest court declines to retroactively apply a state law to a defendant’s direct appeal. 177 See, e.g., supra notes 1–3 and accompanying text (discussing hypothetical in which one defendant gets benefit of new state criminal procedure ruling and another does not, due only to timing). Griffith’s focus on the actual inequity caused by selective prospectivity, 178 Selective prospectivity refers to the judicial practice of declaring and applying a new rule to one criminal defendant’s direct appeal, then declining to apply it to a subsequent defendant’s appeal. See supra notes 1–3 and accompanying text (providing example of selective prospectivity). as opposed to something intrinsic about federal law, suggests that the technique of selective prospectivity itself violates due process because it arbitrarily produces disparate results for similarly situated defendants. 179 See Griffith, 479 U.S. at 323 (stating problem with selective prospectivity is that it causes actual inequity). For further discussion of why selective prospectivity is incompatible with the judicial role, see, e.g., Harper v. Va. Dep’t of Taxation, 509 U.S. 86, 106–10 (1993) (Scalia, J., concurring). So Griffith’s equal treatment principle is implicated when any new criminal procedural ruling, not just one rooted in federal law, is applied to one direct appeal but not to a subsequent one. If this principle applies to all new federal and state rules, the remaining question is whether this principle, on its own, is sufficient to require the Griffith bright-line rule of retroactivity. Put another way: Does Griffith rest fully on its equal treatment principle? Or was the Article III principle also necessary to its holding? If the former is true, then Griffith’s equal treatment principle also requires that state courts retroactively apply all new state criminal procedural rules. 180 See supra notes 55–56 and accompanying text (discussing this ambiguity within Griffith).

There will not be a clear answer to this question until the Supreme Court revisits Griffith. 181 Griffith is not the only case in which the Supreme Court has cited two constitutional provisions in support of a new rule and not said whether each one individually is sufficient to merit the rule. See, e.g., Obergefell v. Hodges, 135 S. Ct. 2584, 2604–05 (2015) (holding Due Process Clause and Equal Protection Clause give same-sex couples fundamental right to marry). But Griffith’s language suggests that each of its two rationales were independently sufficient to justify its holding. It refers to both propositions as “basic norms of constitutional adjudication” and states that selective prospectivity of the constitutional rule in question violated both principles. 182 See Griffith, 479 U.S. at 322 (rejecting selectively prospective application of new federal-constitutional rule of criminal procedure). And in support of its equal treatment principle, Griffith cites Justice Marshall’s statement that “[d]ifferent treatment of two cases is justified under our Constitution only when the cases differ in some respect relevant to the different treatment,” 183 Id. at 327 (citing Michigan v. Payne, 412 U.S. 47, 60 (1973) (Marshall, J., dissenting)). Justice Marshall then argued that “a difference in the speed with which a judicial system disposes of an appeal” is not relevant to different treatment of two cases, and thus “considerations of fairness rooted in the Constitution lead me to conclude that cases in the pipeline when a new constitutional rule is announced must be given the benefit of that rule.” Payne, 412 U.S. at 60 (Marshall, J., dissenting). And Justice Marshall suggested in a footnote that this equal treatment principle is rooted in due process. See id. at 60 n.2 (mentioning Due Process Clause as barrier to unequal treatment of defendants). as well as a similar statement by Justice Powell, 184 Griffith, 479 U.S. at 327 (“‘[It] hardly comports with the ideal of administration of justice with an even hand,’ when ‘one chance beneficiary—the lucky individual whose case was chosen as the occasion for announcing the new principle—enjoys retroactive application, while others similarly situated have their claims adjudicated under the old doctrine.’” (internal quotation marks omitted) (quoting Hankerson v. North Carolina, 432 U.S. 233, 247 (1977) (Powell, J., concurring in the judgment))). both of which seem to say that treating similarly situated defendants unequally is unconstitutional in itself. One might argue that the Supreme Court would not have discussed the Article III principle at length if it was not necessary to Griffith’s holding. But presumably a procedural action is unconstitutional if it violates just one “basic norm of constitutional adjudication.” It seems wrong to say that the combination of two constitutional violations is necessary to merit a rule remedying the violations. It is possible that only Griffith’s Article III principle, and not its equal treatment principle, was significant enough to justify the Griffith bright-line rule independently. But there is no support for this in the text besides the fact that Griffith discusses its Article III principle first. 185 Id. at 322–23 (discussing Article III principle before equal treatment principle).

Case law outside the retroactivity context also supports the proposition that state court decisions that violate Griffith’s equal treatment principle are unconstitutional. The Supreme Court has ruled that inconsistently applying state criminal procedural rules, as well as applying procedural rules that lack precedential support, can violate the Due Process Clause. 186 See Staub v. City of Baxley, 355 U.S. 313, 318–19 (1958) (holding application of state procedural rule, which lacked support in state law and had been inconsistently applied, violated due process); cf. Brinkerhoff-Faris Tr. & Sav. Co. v. Hill, 281 U.S. 673, 680 (1930) (“The federal guaranty of due process extends to state action through its judicial [branch].”). Creating a new procedural rule and applying it to one direct appeal, but not a subsequent one, constitutes arbitrarily inconsistent application of state law. Just as a state’s interest in efficient administration of justice does not justify the application of procedural rules lacking precedential support, it also cannot justify applying an overruled state law to a direct appeal. 187 See, e.g., Staub, 355 U.S. at 318–19 (holding application of state procedural rule that lacked support in state law and had been inconsistently applied violated due process).

Some state courts have summarily rejected arguments that state-law retroactivity is a federal question by citing a line of dicta from a 1990 plurality opinion written by Justice O’Connor. 188 See supra note 110 and accompanying text (discussing Justice O’Connor’s line of dicta). For examples of state court decisions citing Justice O’Connor’s line, see People v. Mitchell, 606 N.E.2d 1381, 1384 (N.Y. 1992); Bush v. State, 428 S.W.3d 1, 13 (Tenn. 2014). As her only support for her proposition, Justice O’Connor cited a 1932 case. 189 Am. Trucking Ass’ns v. Smith, 496 U.S. 167, 177 (1990) (plurality opinion) (citing Great N. Ry. Co. v. Sunburst Oil & Ref. Co., 287 U.S. 358, 364 (1932)). But Griffith fundamentally altered the doctrine of retroactivity-on-direct-review, giving it constitutional underpinnings. 190 See supra section I.C.1 (explaining Griffith’s two constitutional rationales). A 1932 decision, rendered before the Supreme Court gave much attention to retroactivity, 191 Cf. Fallon et al., supra note 19, at 1293 (discussing how retroactivity issues grew in significance in second half of twentieth century, forcing Supreme Court to give them more attention). is not helpful in discerning what the Constitution has to say about retroactivity post-Griffith. Similarly, dicta in a plurality opinion written and joined by Justices who dissented in Griffith does not merit dismissal of questions about Griffith’s ramifications for state-law retroactivity. 192 Chief Justice Rehnquist and Justice White each dissented in Griffith v. Kentucky. 479 U.S. 314, 329 (1987) (Rehnquist, C.J., dissenting); id. at 329–34 (White, J., dissenting). Chief Justice Rehnquist and Justice O’Connor joined Justice White’s opinion, which supported continued use of the Stovall three-factor test and “recognize[d] no distinction for retroactivity purposes between cases on direct and collateral review.” Id. at 333 (White, J., dissenting). Chief Justice Rehnquist, Justice White, and Justice Kennedy (who was not on the Court at the time of Griffith) joined Justice O’Connor’s American Trucking Associations v. Smith plurality opinion. 496 U.S. at 171 (plurality opinion). Understandings of what states cannot do because of the Due Process Clause evolve over time. 193 See, e.g., Lawrence Rosenthal, Does Due Process Have an Original Meaning? On Originalism, Due Process, Procedural Innovations . . . and Parking Tickets, 60 Okla. L. Rev. 1, 2, 28–47 (2007) (arguing Framers envisioned due process as evolving concept and “mandate for the courts to evolve a common law governing the manner in which persons may be deprived of life, liberty, or property”). Griffith and its progeny are the proper touchstone for determining what the Constitution requires with respect to retroactivity on direct review.

Perhaps the strongest argument against extending Griffith to state criminal procedural rules is that there is no constitutional right to a criminal appeal. 194 See Jones v. Barnes, 463 U.S. 745, 751 (1983) (“There is, of course, no constitutional right to an appeal . . . .”); Ross v. Moffitt, 417 U.S. 600, 611 (1974) (“[I]t is clear that the State need not provide any appeal at all.”). It seems illogical that the U.S. Constitution would not require states to provide defendants the right to appeal convictions but would require states to follow certain regulations if they do choose to provide a right to appeal, such as having to retroactively apply new criminal procedural rulings to direct appeals. 195 Coombs, supra note 56, at 585–89. But cutting against this argument is the fact that the Supreme Court has imposed several constitutional requirements on state appellate courts in the past, even despite this possible logical inconsistency. 196 See M.L.B. v. S.L.J., 519 U.S. 102, 107 (1996) (holding states may not deny mother, due to her poverty, appellate review of sufficiency of evidence on which trial court based parental termination decree); Powell v. Nevada, 511 U.S. 79, 84 (1994) (holding Griffith requires state appellate courts to retroactively apply new federal-constitutional rules to defendants’ direct appeals); Brecht v. Abrahamson, 507 U.S. 619, 636 (1993) (clarifying U.S. Constitution requires state appellate courts to use harmless error appellate review standard created in Chapman v. California, 386 U.S. 18 (1967)); Douglas v. California, 372 U.S. 353, 356–58 (1963) (holding Fourteenth Amendment’s Due Process Clause guarantees defendant-appellants assistance of counsel in first appeal of right state chooses to provide). The Supreme Court has never clarified where its asserted power to regulate state appellate proceedings comes from. 197 See Coombs, supra note 56, at 608–10 (explaining and complaining about this uncertainty). But if such a power exists, the U.S. Constitution could require states to retroactively apply their new criminal procedural rules to direct appeals they choose to provide, even though it does not require them to provide rights to appeal in the first place.

Another argument against extending Griffith to state criminal procedural rules stems from federalism concerns. It may be normatively preferable for states to generally make their own decisions about state-law matters, like the retroactive effect of their own rules. 198 Cf., e.g., Gonzalez v. Raich, 545 U.S. 1, 42 (2005) (O’Connor, J., dissenting) (“One of federalism’s chief virtues, of course, is that it promotes innovation by allowing for the possibility that ‘a single courageous State may, if its citizens choose, serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country.’” (quoting New State Ice Co. v. Liebmann, 285 U.S. 262, 311 (1932) (Brandeis, J., dissenting))). Some commentators also question the Supreme Court’s frequent imposition of new due process requirements on state courts and its unclear articulations of the scope of its power to do so. 199 See, e.g., Coombs, supra note 56, at 609–16 (criticizing Supreme Court’s frequent imposition of new due process requirements on state courts). But the point of the Fourteenth Amendment’s Due Process Clause is to apply due process requirements to states. The Supreme Court has appellate jurisdiction over state court decisions for the precise purpose of ensuring that state courts follow federal law. 200 See, e.g., Testa v. Katt, 330 U.S. 386, 390–91 (1947) (holding state courts are required to enforce federal law because federal and state governments constitute single system, in which federal laws are supreme where valid). Invalidation of state courts’ procedural doctrines that violate due process requirements is a longstanding tradition. 201 See, e.g., Brinkerhoff-Farris Tr. & Sav. Co. v. Hill, 281 U.S. 673, 681–82 (1930) (“But, while it is for the state courts to determine the adjective as well as the substantive law of the state, they must, in so doing, accord the parties due process of law.”). In addition, federal intervention in state court proceedings is most justifiable when it protects citizens’ federal rights in litigation against state governments (like criminal prosecutions). 202 See Michigan v. Long, 463 U.S. 1032, 1067–68 (1983) (Stevens, J., dissenting) (arguing federal courts should be most inclined to overturn state judgments when citizens have been deprived of federal rights); Union Pac. R.R. Co. v. Pub. Serv. Comm’n, 248 U.S. 67, 70 (1918) (suggesting it is most important for Court to inquire into validity of state judgments when states deprive private parties of constitutional rights). The U.S. Constitution binds both federal and state courts with its guarantee of due process; 203 E.g., Testa, 330 U.S. at 391 (“[T]he Constitution and the laws passed pursuant to it are the supreme laws of the land, binding alike upon states, courts, and the people, any Thing in the Constitution or Laws of any State to the Contrary notwithstanding.” (internal quotation marks omitted) (quoting U.S. Const. art VI, § 2)). if this guarantee requires federal courts to consistently apply new rules to direct appeals, state courts should be required to do so as well.

The Supreme Court has still never held that the U.S. Constitution requires state courts to retroactively apply any state law. 204 See supra section I.D (discussing Supreme Court cases about retroactivity of state law). Until it does, most state courts will likely employ the Stovall three-factor test, not Griffith’s bright-line retroactivity rule, to determine whether to retroactively apply new state criminal procedural rulings. 205 See supra section II.B (showing most state courts use Stovall three-factor test). The Griffith bright-line retroactivity rule favors defendants and, if adopted by states, would likely cause more reversals of convictions than the Stovall three-factor test. 206 See supra notes 96–103 and accompanying text (showing use of Stovall three-factor test rarely results in retroactive application and subsequent reversal, and choice of Stovall test or Griffith rule often dictates whether ruling will be retroactively applied). But, contrary to what some state courts argue, 207 See People v. Sexton, 580 N.W.2d 404, 415 (Mich. 1998) (stating retroactive application of requirement that police inform suspect when retained counsel is available for consultation would undermine validity of large number of convictions and burden criminal justice system with numerous retrials). the Griffith rule would not result in mass reversals, as section III.C shows.

C. Practical Impact of Applying the Griffith Bright-Line Rule

The Griffith bright-line rule would not create as great a burden on the administration of justice as some state courts believe. 208 See, e.g., id. (stating retroactive application of requirement that police inform suspect when retained counsel is available for consultation “would be extremely disruptive to the administration of justice”); Taylor v. State, 10 S.W.3d 673, 681 (Tex. Crim. App. 2000) (dismissing Griffith rule as too rigid and placing too great a burden on administration of justice). Retroactive application of a new criminal procedural ruling to a conviction on direct appeal does not necessarily mean the court will reverse the conviction. To reverse, the appellate court must find that the trial court’s failure to apply the newly declared rule was a reversible error. 209 See, e.g., United States v. Lopez-Pena, 912 F.2d 1542, 1550 (1st Cir. 1989) (applying retroactively new ruling that magistrates cannot preside over jury selection in felony cases, but finding magistrate’s violation of rule was harmless error, and denying defendant’s request for remand); Taylor, 10 S.W.3d at 683 (“Even if the rule is retroactively imposed, a defendant may yet be denied relief on the basis that the error was not properly presented or was harmless.”). If the appellate court finds that the trial court’s failure to apply the new rule was not reversible in the defendant’s individual case, the conviction will be affirmed. This protects against the potential efficiency costs of mass retrials. 210 These efficiency costs appear to be a major reason why most state courts use the Stovall three-factor test instead of the Griffith bright-line rule. See, e.g., Sexton, 580 N.W.2d at 415 (discussing efficiency costs as disadvantage of Griffith rule).

Additionally, some state courts have held that even if a trial court’s failure to apply a new ruling was harmful error, the court may grant the defendant a different remedy than would typically be granted when the new ruling is violated. 211 See Taylor, 10 S.W.3d at 682 (“[T]he harshness of a retroactive rule [for conducting criminal prosecutions] can sometimes be mitigated by requiring a different remedy than would ordinarily be accorded the rule in its prospective application.”). For an example of a proposed alternative remedy for a violation of a new rule, see Michigan v. Payne, 412 U.S. 47, 64–65 (1973) (Marshall, J., dissenting) (proposing alternative remedy for retroactive constitutional violation, which would require states to present evidence that new, harsher sentence, given to defendant who had successfully appealed, was based on defendant’s postsentence conduct). Thus, the primary practical difference between the Griffith and Stovall approaches to retroactivity is that the Griffith bright-line rule leads to an assessment of the rule’s effect in the defendant’s particular case, whereas the Stovall approach only examines whether the rule itself merits retroactive application due to its purpose. 212 See supra note 92 and accompanying text (demonstrating Stovall three-factor test considers rules generally, not case-by-case, such that courts cannot retroactively apply same rule in one case but not another). Therefore, the Griffith bright-line rule is more flexible at the remedial stage 213 The remedial stage is all that matters to defendants. A defendant does not gain anything from retroactive application unless it results in a reversal of her conviction or a reduced sentence. than Stovall because it takes into account the circumstances of individual cases. 214 See Lopez-Pena, 912 F.2d at 1452 (affirming conviction, despite retroactively applying new rule, due to harmless error in instant case). As such, while the Griffith bright-line rule leads to more reversals than Stovall, 215 See supra section II.A.2 (showing different results of Griffith and Stovall tests). it does not create an overwhelmingly great burden on the administration of justice. It provides reversals only for defendants whose cases would be significantly affected by new procedural rules, even if the rule might not be considered particularly important to the integrity of the fact-finding process in a vacuum. So when taking into account the remedial stage of criminal appeals, the Griffith rule does not function as a bright-line rule, but rather as a case-specific inquiry that accounts for defendants’ individual circumstances better than the Stovall three-factor test does.

Conclusion

This Note seeks to address the unequal way in which state courts treat defendants like Marks and Carlson. 216 See supra notes 1–3 and accompanying text (presenting hypothetical about Marks and Carlson). Should courts deny Carlson the benefit of a new rule simply because she was both too early (because her trial occurred before the newly declared rule) and too late (because her direct appeal occurred after the state’s highest court declared the new rule)? 217 See supra notes 1–3 and accompanying text (presenting hypothetical about Marks and Carlson). The majority of state courts use the Stovall three-factor test and only grant defendants the benefit of a new rule when the rule significantly impacts the truth-finding process at trial—reflecting a policy judgment that retroactive application and the efficiency costs it imposes on the administration of justice is only appropriate when the new rule raises significant doubts about the accuracy of past verdicts. 218 See supra section II.B (describing majority of state courts’ approach). Four state courts use the Griffith bright-line rule for state criminal procedural rulings. 219 See supra section II.C (describing minority of courts who have adopted Griffith bright-line rule). This Note assesses both approaches and concludes that Griffith’s equal treatment principle requires state courts to retroactively apply new state criminal procedural rulings to direct appeals. The extent of a new criminal procedural rule’s retroactive effect should not hinge on whether the rule is derived from federal-constitutional, federal-nonconstitutional, or state-law principles. Selective prospectivity in applying criminal procedural rules violates the U.S. Constitution because it treats similarly situated defendants unequally.

Appendix

Type of New Ruling Constitutional Statutory Common Law
Federal Substantive Teague v. Lane: courts must retroactively apply these. 220 489 U.S. 288, 305 (1989). Courts must retroactively apply these. 221 See, e.g., Schriro v. Summerlin, 542 U.S. 348, 351 (2004) (“New substantive rules generally apply retroactively. This includes decisions that narrow the scope of a criminal statute by interpreting its terms . . . .” (citation omitted)). No case law since Griffith. 222 New rulings of this sort do not occur because federal courts cannot create substantive criminal common law other than barring actions that interfere with the court’s exercise of its authority (like contempt). United States v. Hudson, 11 U.S. (7 Cranch) 32, 32–34 (1812).
Federal Procedural Griffith: courts must retroactively apply these. 223 Griffith v. Kentucky, 479 U.S. 314, 322 (1987). Circuit split on whether courts must retroactively apply these. 224 See supra notes 159–160 and accompanying text (describing circuit split). No case law since Griffith.
State Substantive No case law since Griffith. State courts usually retroactively apply these. 225 See note 32 and accompanying text (listing examples and one exception). State courts usually retroactively apply these. 226 See supra note 225 (noting state courts generally apply new substantive rules on direct appeal).
State Procedural No case law since Griffith. 227 See supra note 86. Most state courts use version of Stovall test. 228 See supra section II.A (asserting most state courts use version of Stovall test in assessing retroactivity of new state criminal procedural decisions).  Five states always retroactively apply these. 229 See supra note 95 (listing state courts that have adopted bright-line retroactivity rule for state criminal procedural rulings on direct review). Most state courts use version of Stovall test. 230 See supra section II.A (showing most state courts use version of Stovall test in assessing retroactivity of new state criminal procedural decisions).  Four states always retroactively apply these. 231 See supra note 95 (listing state courts that have adopted bright-line retroactivity rule for state criminal procedural rulings on direct review).