Harvard Law Review Harvard Law Review Harvard Law Review

Constitutional Law

Constitutional Waivers by States and Criminal Defendants

Where there is a right, there is (usually) a way to waive it.1×1. See, e.g., Jason Mazzone, The Waiver Paradox, 97 Nw. U. L. Rev. 801, 801 (2003) (pointing out that certain constitutional rights, like First Amendment rights and other noncriminal protections, are not freely waivable). Where there is a waiver, the question of its validity arises. And to answer that question, despite the range of constitutional rights available to be waived, the Court has returned time and time again to the same paradigmatic definition2×2. See William J. Stuntz, Waiving Rights in Criminal Procedure, 75 Va. L. Rev. 761, 796 (1989) (dubbing this standard “the starting point for modern constitutional waiver doctrines”). first articulated in Johnson v. Zerbst3×3. 304 U.S. 458 (1938). : a valid waiver is the “intentional relinquishment or abandonment of a known right or privilege,”4×4. Id. at 464. or to use the Court’s subsequent rephrasing, the waiver must be “voluntary [and] knowing.”5×5. Brady v. United States, 397 U.S. 742, 748 (1970); see, e.g., Maryland v. Shatzer, 559 U.S. 98, 104 (2010); Edwards v. Arizona, 451 U.S. 477, 483 (1981) (quoting Arizona v. Edwards, 594 P.2d 72, 78 (Ariz. 1979)); see also Edward L. Rubin, Toward a General Theory of Waiver, 28 UCLA L. Rev. 478, 491–93 (1981). Specifically, when it comes to criminal constitutional protections, the Court has almost uniformly applied this familiar requirement regardless of the substantive right, from the Sixth Amendment right to counsel, which Johnson addressed,6×6. See Johnson, 304 U.S. at 458. to the privilege against self-incrimination7×7. See Miranda v. Arizona, 384 U.S. 436, 475 (1966). and the right to a jury trial.8×8. See Brady, 397 U.S. at 748. There are some exceptions. See, e.g., Schneckloth v. Bustamonte, 412 U.S. 218, 248–49 (1973) (consent searches).

That canonical definition of waiver, as well as the broader notion that there is some uniformity to waiver standards across constitutional rights, has figured in the Court’s conversations about waivers of a very different kind of right: state sovereign immunity. The voluntariness and knowingness concerns underlying Johnson consistently emerge in the Court’s analyses of immunity waivers.9×9. See infra Part II, pp. 2556–71. And occasionally, the Court has gone so far as to draw explicit comparisons to other rights when shaping its sovereign immunity jurisprudence. The debates surrounding Parden v. Terminal Railway of the Alabama State Docks Department,10×10. 377 U.S. 184 (1964), overruled by Coll. Sav. Bank v. Fla. Prepaid Postsecondary Educ. Expense Bd., 527 U.S. 666 (1999). where the Court found that a state had constructively waived its immunity by participating in a federally regulated activity,11×11. See id. at 192. occurred partly on such terms. Justice White’s dissent argued that the holding rendered impossible the state’s “intentional relinquishment . . . of a known right or privilege.”12×12. See id. at 200 (White, J., dissenting) (quoting Johnson v. Zerbst, 304 U.S. 458, 464 (1938)). Other Justices later continued that line of reasoning,13×13. See, e.g., Emps. of the Dep’t of Pub. Health & Welfare v. Dep’t of Pub. Health & Welfare, 411 U.S. 279, 296 (1973) (Marshall, J., concurring in the result) (“[T]he concept of implied consent or waiver relied upon in Parden approaches . . . the outer limit of the sort of voluntary choice which we generally associate with the concept of constitutional waiver.” (citing, inter alia, Johnson, 304 U.S. at 464)); Edelman v. Jordan, 415 U.S. 651, 673 (1974) (“Constructive consent is not a doctrine commonly associated with the surrender of constitutional rights . . . .”); Welch v. Tex. Dep’t of Highways & Pub. Transp., 483 U.S. 468, 473 (1987) (plurality opinion) (quoting Edelman, 415 U.S. at 673); see also Stewart A. Baker, Federalism and the Eleventh Amendment, 48 U. Colo. L. Rev. 139, 163 (1977) (describing the Court’s “useful” analogy of state immunity to individual rights). and when the Court overruled what was left of Parden in College Savings Bank v. Florida Prepaid Postsecondary Education Expense Board,14×14. 527 U.S. 666. it invoked Johnson once again, noting that state sovereign immunity is no less constitutionally protected than the right to trial by jury in criminal cases.15×15. See id. at 681–82 (stating that “Parden-style waivers are simply unheard of in the context of other constitutionally protected privileges,” id. at 681, and that “[t]he classic description of an effective waiver of a constitutional right is the ‘intentional relinquishment or abandonment of a known right or privilege,’” id. at 682 (quoting Johnson, 304 U.S. at 464)). Lower courts have picked up the trend of drawing connections between sovereign immunity waivers and the classic individual waiver standard, even repeating Johnson’s language to articulate the rule for states.16×16. See, e.g., Gruver v. La. Bd. of Supervisors for the La. State Univ. Agric. & Mech. Coll., 959 F.3d 178, 182 (5th Cir. 2020) (“As is usually true for waivers, any [sovereign immunity] waiver must be knowing and voluntary.”); Pettigrew v. Okla. ex rel. Okla. Dep’t of Pub. Safety, 722 F.3d 1209, 1213 (10th Cir. 2013) (quoting Coll. Sav., 527 U.S. at 675); Lombardo v. Pa. Dep’t of Pub. Welfare, 540 F.3d 190, 198 (3d Cir. 2008) (applying the Johnson standard); Pace v. Bogalusa City Sch. Bd., 403 F.3d 272, 277 (5th Cir. 2005) (stating that consent to suit must be “knowing and voluntary,” and citing Johnson to define waiver); Barbour v. Wash. Metro. Area Transit Auth., 374 F.3d 1161, 1166–67 (D.C. Cir. 2004) (explaining that waiver of sovereign immunity must be “knowing,” id. at 1167); Arecibo Cmty. Health Care, Inc. v. Puerto Rico, 270 F.3d 17, 29 (1st Cir. 2001) (comparing a state’s choice to waive its immunity with “any case of a knowing and intelligent waiver of rights”); Garcia v. SUNY Health Scis. Ctr. of Brooklyn, 280 F.3d 98, 114 (2d Cir. 2001).

This Note takes a closer look at that oft-invoked comparison. It begins by taking a step back. Part I considers what we would theoretically expect from waiver law if we accepted the Court’s suggestion that constitutional waivers, no matter the underlying right, must at minimum meet Johnson’s requirements of voluntariness and knowingness. It argues that because of relative social and structural positions, courts can more readily expect states to act knowingly and voluntarily in waiver settings than they can criminal defendants. Legal presumptions should therefore establish a double standard that corresponds to those baseline expectations: it should be harder to show knowing, voluntary waivers by criminal defendants than by states.

Part II turns to the law in practice and shows that the Court has established precisely the opposite double standard: while it is easy for criminal defendants to waive their rights, it is nearly impossible for states to do so. The Note shows this phenomenon through the lens of three waiver mechanisms that the Court has analyzed in both the sovereign and criminal contexts: waivers implied by conduct, bargained-for waivers, and waivers through litigation. Across all three, the Court has returned — both implicitly and explicitly — to the same concepts of voluntariness and knowingness as touchstones for the validity of sovereign immunity and criminal protection waivers.17×17. Courts have used “intentional” and “voluntary” interchangeably. See Rubin, supra note 5, at 492 & n.63. This Note understands intentionality as a distinct concept that encompasses voluntariness but also includes an element of knowingness. See id. (adopting this interpretation). Because voluntariness and knowingness remain the core notions underlying intentionality, however, and because they are the core concepts underlying Johnson, this Note will focus on them. However, it has applied those concepts very differently in each context. “Voluntary and knowing” is defined liberally as applied to criminal defendants but defined narrowly for states. As a descriptive matter, this discrepancy calls into question both the Court’s suggestion that there is uniformity to its evaluation of constitutional waivers and lower courts’ continued invocation of the analogy between sovereign and criminal waiver rules.

The Note concludes by confronting the puzzle presented by Parts I and II. If the Court’s canonical definition of waiver requires that they be knowing and voluntary, and if states are more likely than criminal defendants to be able to waive their rights knowingly and voluntarily, why does current law presume the opposite? Part III offers a potential explanation. It suggests that the Court’s approach rests on implicit hierarchies in its understanding of state and criminal protections. Specifically, even as the Court purportedly seeks to be equally protective of various constitutional rights,18×18. See Coll. Sav., 527 U.S. at 682 (“State sovereign immunity, no less than the right to trial by jury in criminal cases, is constitutionally protected.”). it has placed states’ rights on a higher plane than individual criminal protections by prioritizing preservation of sovereign immunity and minimizing the societal costs of doing so.

I. The Double Standard in Theory

This Part begins by taking the Court at its word. It assumes we can compare waivers by both states and criminal defendants, and it accepts that valid constitutional waivers must be knowing and voluntary. Because that validity standard is abstract, courts must identify judicial presumptions or rules to apply it, and those presumptions should vary across different contexts depending on our baseline expectations of the rights holder’s ability to act voluntarily and knowingly. Before diving into the Court’s current approach, it is worth considering what those baselines are and what standards one would expect them to produce.

First, states are far more likely than individual defendants to know their rights and understand the scope of their waivers. States have attorneys, information, financial resources, institutional knowledge, and experience. Moreover, states participate in producing the legislation that compels or bargains for their waivers.19×19. See Vicki C. Jackson, Principle and Compromise in Constitutional Adjudication: The Eleventh Amendment and State Sovereign Immunity, 75 Notre Dame L. Rev. 953, 1007 (2000) (noting that “the political structure of the national political branches provides multiple opportunities for consideration of the interests of the states in their governmental capacities”). Indeed, it is a foundational tenet of our political process that Congress represents state interests.20×20. See id. at 956. There is no analogous representation for criminal defendants. States thus not only are better equipped to understand the consequences and nature of a waiver, but also are in the room where decisions about those waivers are made. Meanwhile, our radically underfunded and overburdened criminal defense system21×21. See generally, e.g., The Const. Project, Justice Denied 52–70 (2009). regularly saddles defendants with incompetent counsel.22×22. See, e.g., Erwin Chemerinsky, Remarks, Lessons from Gideon, 122 Yale L.J. 2676, 2679 (2013) (suggesting that in many situations self-representation may have been more effective than was representation by ineffective counsel). The reality of who enters that system — for example, a disproportionate number of individuals with lower education levels and/or mental illnesses23×23. Cf. Nat’l Rsch. Council, The Growth of Incarceration in the United States 65–67, 204–05 (Jeremy Travis, Bruce Western & Steve Redburn eds., 2014). — exacerbates the structural information disadvantages criminal defendants face. The presumption that such defendants have any legal knowledge on which to fall back is untenable.

Second, states are better positioned to voluntarily waive their rights. Voluntariness can be analyzed as freedom from coercion: the more vulnerable to coercion the rights holder, the weaker the presumption that they acted voluntarily.24×24. Cf., e.g., Jessica Wilen Berg, Understanding Waiver, 40 Hous. L. Rev. 281, 308 (2003); Michael E. Tigar, The Supreme Court, 1969 Term — Foreword: Waiver of Constitutional Rights: Disquiet in the Citadel, 84 Harv. L. Rev. 1, 25 (1970). A waiver might be coercive because of the nature of the choice it invokes (is it too good to pass up?) or the environment in which the choice is made.25×25. See Mitchell N. Berman, Coercion, Compulsion, and the Medicaid Expansion: A Study in the Doctrine of Unconstitutional Conditions, 91 Tex. L. Rev. 1283, 1291–92, 1321–22 (2013). The Court has recognized the former in both sovereign and criminal waiver scenarios26×26. See, e.g., Coll. Sav. Bank v. Fla. Prepaid Postsecondary Educ. Expense Bd., 527 U.S. 666, 686–87 (1999); Brady v. United States, 397 U.S. 742, 749–50 (1970). but has applied the latter primarily, if exclusively, to criminal defendants, who often waive their rights in inherently coercive contexts.27×27. For discussion of waivers in custodial interrogations and in the face of a prosecutor seeking to compel a guilty plea, see infra sections II.A.1, pp. 2557–60, and II.B.1, pp. 2565–66. Considering the enormous gap between the power and resources of criminal defendants and prosecutors,28×28. See The Const. Project, supra note 21, at 61–64 (describing resource gap). This Part seeks to draw only general conclusions about these baselines. But it is crucial to notice that the power differential between the state and criminal defendants — and the latter’s vulnerability to coercion — is made far worse by the racism that pervades the penal system. To cite just one example, in the plea bargaining process, white defendants systematically get offered more lenient charges and sentences than do Black defendants. See Carlos Berdejó, Criminalizing Race: Racial Disparities in Plea-Bargaining, 59 B.C. L. Rev. 1187, 1215–17 (2018). The system thus reduces the already minimal leverage that Black defendants have when deciding whether to waive their rights. it is also difficult to presume that a defendant could adequately resist coercion that does manifest. By contrast, we can more safely presume that state decisionmakers have the political clout to make voluntary decisions under pressure from the federal government.29×29. For example, when it comes to cooperative federalism programs (a subset of which require immunity waivers), the state “agent” providing consent is usually an agency or the governor, see Bridget A. Fahey, Consent Procedures and American Federalism, 128 Harv. L. Rev. 1561, 1573–75, 1612–13 (2015); both political entities act with the state’s imprimatur and so have far more political and institutional power than does a criminal defendant.

Using these baselines, we would expect courts to be more willing to presume that states, not criminal defendants, validly waived their rights, since the risk that the former acted unknowingly or involuntarily is lower than the corresponding risk for the latter. In terms of legal rules, this would translate into a double standard: a waiver test for states that is easy to satisfy, and a waiver test for criminal defendants that is hard to satisfy. But as Part II will show, while we have a double standard today, it is the precise opposite of the one we might expect.

II. The Double Standard in Practice

This Part moves to the descriptive. Several constitutional provisions protect criminal defendants.30×30. See, e.g., William J. Stuntz, The Uneasy Relationship Between Criminal Procedure and Criminal Justice, 107 Yale L.J. 1, 53, 72 (1997) (citing the Fourth, Fifth, and Sixth Amendments). Recognizing that defendants may waive the rights conferred by those provisions, the Court has had to develop standards to evaluate such waivers. It has almost universally turned to Johnson as a starting point. In addition, the Constitution affords states protections: the Court has embraced state immunity from suit by private individuals31×31. State immunity does not extend to suits by other states or the federal government. See South Dakota v. North Carolina, 192 U.S. 286, 317–18 (1904); United States v. Texas, 143 U.S. 621, 642–46 (1892). as a right located in the Constitution with the Eleventh Amendment as the textual hook.32×32. See, e.g., Franchise Tax Bd. v. Hyatt, 139 S. Ct. 1485, 1493–96 (2019) (canvassing constitutional roots of state immunity); Alden v. Maine, 527 U.S. 706, 730–35 (1999). Note that there is, however, a debate about the constitutional grounding of Eleventh Amendment immunity. See, e.g., Martha A. Field, The Eleventh Amendment and Other Sovereign Immunity Doctrines: Part One, 126 U. Pa. L. Rev. 515, 541–44 (1978) (questioning sovereign immunity’s constitutional foundation). Because sovereign immunity is waivable,33×33. The ability to waive immunity and consent to suit was acknowledged in the earliest sovereign immunity cases. See, e.g., Clark v. Barnard, 108 U.S. 436, 447–48 (1883); Hans v. Louisiana, 134 U.S. 1, 17 (1890). This Note treats waiver and consent interchangeably as situations in which the state, through its own actions or decisions, gives up its immunity. But see Jonathan R. Siegel, Waivers of State Sovereign Immunity and the Ideology of the Eleventh Amendment, 52 Duke L.J. 1167, 1187 (2003) (suggesting that waiver and consent are different and were governed by separate rules, at least prior to 1945). here too the Court has fashioned standards for waiver validity.

The following sections explore those standards in the context of three waiver mechanisms that the Court has evaluated for both states and criminal defendants. Section II.A addresses waivers that are implied by the rights holder’s conduct. For criminal defendants, these typically arise in the context of an interrogation;34×34. There is an ongoing debate about whether Miranda warnings are constitutionally required. See, e.g., Laurent Sacharoff, Miranda’s Hidden Right, 63 Ala. L. Rev. 535, 537 & n.6 (2012). But even if the warnings are not constitutionally required, this Note’s comparison would remain salient because it is the underlying privilege against self-incrimination, not the warnings, that suspects can waive, and that privilege is constitutional. That was Miranda’s first holding. See Miranda v. Arizona, 384 U.S. 436, 467 (1966); see also Charles D. Weisselberg, Saving Miranda, 84 Cornell L. Rev. 109, 112 (1998) (suggesting that the warning holding has overshadowed this initial one). for sovereigns, one can look to Parden-style constructive waivers. Section II.B addresses bargained-for waivers that are exchanged for a benefit. For criminal defendants, this exchange manifests in plea bargaining, through which individuals can waive their rights to a trial and against self-incrimination in exchange for leniency; for states, the exchange is effectuated by Spending Clause legislation, through which states can waive their immunity for federal funds. Unlike waivers implied by conduct, bargained-for waivers include an explicit act to consummate the exchange — either pleading on the record35×35. See Boykin v. Alabama, 395 U.S. 238, 242 (1969). or accepting funds.36×36. See Atascadero State Hosp. v. Scanlon, 473 U.S. 234, 246–47 (1985). Section II.C addresses waivers that occur in the course of litigation. For criminal defendants, the question arises when a state defendant seeks to raise a procedurally foreclosed claim on direct or collateral review; for states, it can arise when the state did not raise the immunity defense at trial but seeks to raise it on appeal. Each example reveals that while concerns about voluntariness and knowingness consistently animate the Court, in practice the Court has defined those concepts inconsistently to strengthen the presumption against state waivers and weaken it for criminal defendants.

A. Waiver Implied by Conduct

1. Individual Fifth Amendment Waivers. — Miranda v. Arizona37×37. 384 U.S. 436. held that waivers of the Fifth Amendment privilege against self-incrimination during custodial interrogations must meet the Johnson standard.38×38. Id. at 475 (“This Court has always set high standards of proof for the waiver of constitutional rights, and we re-assert these standards as applied to in-custody interrogation.” (citing Johnson v. Zerbst, 304 U.S. 458 (1938))); see also Berghuis v. Thompkins, 560 U.S. 370, 385 (2010) (“[A] waiver of Miranda rights need only meet the standard of Johnson v. Zerbst.”). The theory behind the Court’s other holding establishing the well-known Miranda warnings was that such interrogations are “inherently coercive,” so the right against self-incrimination might be waived even when the suspect does not actually want or intend to waive it. Miranda, 384 U.S. at 533 (White, J., dissenting); see id. at 467 (majority opinion); Erin E. Brophy & Wendy W. Huang, Custodial Interrogations, 88 Geo. L.J. 1021, 1021 (2000). Miranda originally read Johnson to provide robust protections. A valid waiver had to be the product of “free . . . choice” (voluntariness inquiry) and made with complete awareness of the nature of the right abandoned and the consequences of abandoning it (knowingness inquiry).39×39. Miranda, 384 U.S. at 465; see id. at 464–65, 469; see also Moran v. Burbine, 475 U.S. 412, 421 (1986). But since Miranda, the Court has drastically limited its scope by weakening both requirements.40×40. See generally, e.g., Leslie A. Lunney, The Erosion of Miranda: Stare Decisis Consequences, 48 Cath. U. L. Rev. 727 (1999) (arguing that the Court all but overruled Miranda through subsequent decisions); Charles J. Ogletree, Commentary, Are Confessions Really Good for the Soul? A Proposal to Mirandize Miranda, 100 Harv. L. Rev. 1826, 1839–42 (1987) (showing the Court narrowed Miranda in various ways); Charles D. Weisselberg, Mourning Miranda, 96 Calif. L. Rev. 1519, 1578–83 (2008) (describing the loosening of the Court’s waiver requirements). It began by rejecting a clear statement rule in North Carolina v. Butler.41×41. 441 U.S. 369 (1979). The Butler Court held that neither “[a]n express written [n]or oral statement” was necessary to validly waive the rights to silence and interrogation counsel.42×42. Id. at 373. The defendant had been handed his Miranda rights in writing, but there was a dispute about whether he could read, and he refused to sign a written waiver. Id. at 378 (Brennan, J., dissenting). Butler told agents that he understood his rights, but said: “I will talk to you but I am not signing any form.” Id. at 371 (majority opinion). As Justice Brennan pointed out, these actions and words had an “uncertain meaning.” Id. at 378 (Brennan, J., dissenting). Butler’s decision to “talk” without signing a form could mean he did not intend his statements to be admissible. In other words, Miranda waivers could be implied by a suspect’s conduct.

Two years later, in Edwards v. Arizona,43×43. 451 U.S. 477 (1981). the Court added what appeared to be a protective prophylactic rule: when an accused invokes his right to counsel, the interrogation must stop until counsel is made available.44×44. See id. at 485; Maryland v. Shatzer, 559 U.S. 98, 104 (2010) (first quoting Edwards, 451 U.S. at 484; and then quoting McNeil v. Wisconsin, 501 U.S. 171, 176 (1991)). But subsequent cases diluted that protection. Edwards has been interpreted to require the accused to “unambiguously” assert the right to counsel before reaping its advantages; otherwise, the right is presumed waived and the interrogation may continue.45×45. Davis v. United States, 512 U.S. 452, 459 (1994); see id. (“[H]e must articulate his desire to have counsel present sufficiently clearly that a reasonable police officer in the circumstances would understand the statement to be a request for an attorney.”). The Court has acknowledged that this rule disadvantages suspects who will not assert their desire for a lawyer out of “fear, intimidation, [or] lack of linguistic skills” but suggested that the Miranda warnings protect suspects from coercion.46×46. Id. at 460. That argument appears, however, to conflate knowing that you have a right to a lawyer with knowing that you have to unambiguously assert your desire for one to preserve that right; it also seems to conflate awareness of the right with ability to make a free and deliberate (voluntary) choice about whether to exercise it.47×47. Imagine the following scenario: Police read a suspect her Miranda warnings and inform her she has a right to a lawyer. The suspect wants a lawyer and mumbles that maybe she should talk to one. But under Davis v. United States, 512 U.S. 452, this is too ambiguous an assertion of her rights to stop the interrogation, id. at 462, so the officers continue to question her. She responds, perhaps thinking a lawyer will eventually be provided and assuming that the responses will not be admissible, or perhaps because she is intimidated. Her intentions notwithstanding, she has waived her right to counsel because she failed to ask for a lawyer explicitly, before answering any questions. Now, suspects presumptively can be deemed to have waived their right to interrogation counsel voluntarily, even if they tried to assert it but did so too ambiguously, or even if they were just too intimidated to assert it.48×48. This line of critique is doctrinal, but there is significant evidence that the problem is far worse in practice. Miranda warnings are regularly phrased too confusingly to provide notice about the existence of the right, much less information about how to assert or waive it. See, e.g., Michael D. Cicchini, The New Miranda Warning, 65 SMU L. Rev. 911, 915–28 (2012).

The requirement that suspects waive their rights with full awareness has lost its teeth too. The Court in Moran v. Burbine49×49. 475 U.S. 412 (1986). held that even though police failed to inform the accused that his attorney had called to speak to him pre-interrogation, his subsequent waiver of Miranda rights was knowing and therefore valid,50×50. See id. at 419–20; id. at 425 (refusing to adopt a rule that would require police to “inform a suspect of an attorney’s efforts to reach him”). as he was entitled only to information “essential . . . to understand[ing] the nature of his rights and the consequences of abandoning them.”51×51. Id. at 424. But that reasoning begs the question. Information that a lawyer had tried to halt his interrogation might have affected the accused’s assessment of his situation and, thereby, his view of the consequences of waiving his rights.52×52. See id. at 468 (Stevens, J., dissenting) (“If a lawyer is seen as an aid to the understanding and protection of constitutional rights . . . then today’s decision makes no sense at all.”). Still, the Court held that the waiver was knowing.53×53. See id. at 422 (majority opinion).

Two decades after Miranda, Colorado v. Connelly54×54. 479 U.S. 157 (1986). confirmed how easy it had become to find a valid waiver. The Court held that Connelly, who was suffering from schizophrenia when he waived his Fifth Amendment privilege, had done so voluntarily, even if he could not exercise “free will.”55×55. Id. at 170; see id. at 174 (Brennan, J., dissenting) (noting Connelly had been hospitalized for severe delusions five times prior to his confession); see also id. at 173 (Stevens, J., concurring in the judgment in part and dissenting in part) (critiquing the Court’s belief that “a waiver can be voluntary even if it is not the product of an exercise of the defendant’s ‘free will’” as “incomprehensible” (citation omitted)). According to the Court, the voluntariness of Fifth Amendment waivers “depended on the absence of police overreaching”56×56. Id. at 170 (majority opinion). through measures like “physical or psychological pressure,”57×57. Id. (quoting Moran, 475 U.S. at 421). and “not on ‘free choice.’”58×58. Id. After Connelly, lack of state manipulation is thus not just a necessary but a sufficient condition for voluntary waiver, and mental incapacitation does not render a waiver unknowing.

The Court appears to have no plans to revisit a more protective waiver rule. Berghuis v. Thompkins59×59. 560 U.S. 370 (2010). concerned a suspect who answered several of the detectives’ questions and incriminated himself, but only after nearly three hours of silence in the interrogation room and conflicting evidence about whether he understood his rights.60×60. Id. at 393–94 (Sotomayor, J., dissenting). The Court took this to be “a ‘course of conduct indicating waiver’ of the right to remain silent.”61×61. Id. at 386 (majority opinion) (quoting North Carolina v. Butler, 441 U.S. 369, 373 (1979)). Under Berghuis, courts can thus presume that a suspect waived their right to silence unless they “unambiguously” invoked the right, and remaining silent is ambiguous.62×62. Id. at 381. Coupled with Edwards as currently interpreted, Berghuis allows courts to find that waivers of Miranda rights were voluntary and knowing unless the suspect explicitly stated they did not wish to waive the right.

2. State Constructive Waivers. — Until relatively recently, states could impliedly waive their sovereign immunity through their conduct. In Parden, the Court held that Alabama, by deciding to operate a railroad after Congress enacted the Federal Employers’ Liability Act63×63. 45 U.S.C. §§ 51–60. (FELA), had consented to suit under the FELA provision permitting suits against any “common carrier by railroad” engaging in interstate commerce.64×64. Id. § 61; see Parden v. Terminal Ry. of the Ala. State Docks Dep’t, 377 U.S. 184, 185–87 (1964), overruled by Coll. Sav. Bank v. Fla. Prepaid Postsecondary Educ. Expense Bd., 527 U.S. 666 (1999). While the Court has referred to Parden as a case about congressional abrogation of immunity, see Pennsylvania v. Union Gas Co., 491 U.S. 1, 14 (1989) (plurality opinion), that view has been rejected, see, e.g., Alden v. Maine, 527 U.S. 706, 737 (1999) (“Parden was based on concepts of waiver and consent.”); Blatchford v. Native Village of Noatak, 501 U.S. 775, 787–88 (1991) (noting that “Parden itself neither mentioned nor was premised upon abrogation,” id. at 787, and was in fact about waiver and consent). This Note will thus address Parden purely as a waiver doctrine. The Court’s theory was that states who chose to operate a railroad governed by the FELA had “ventur[ed] into the congressional realm” and, in so doing, accepted the conditions that Congress attached to participation in that realm.65×65. Parden, 377 U.S. at 196. In other words, the State voluntarily engaged in conduct that implied an intent to waive its immunity. The dissent disputed that such waivers could possibly meet Johnson’s requirements for constitutional waivers.66×66. See id. at 200 (White, J., dissenting).

Parden’s theory was short lived.67×67. For a critique of Parden’s demise in the context of the Court’s broader sovereign immunity jurisprudence, see Jackson, supra note 19, at 961–63. In Employees of the Department of Public Health & Welfare v. Department of Public Health & Welfare,68×68. 411 U.S. 279 (1973). the Court declined to extend it to allow suits against state employers under the Fair Labor Standards Act69×69. 29 U.S.C. §§ 201–219. (FLSA).70×70. See Emps., 411 U.S. at 286–87. While Employees occasionally verged on reading Parden as an abrogation case, e.g., id. at 287 (refusing to extend Parden unless Congress clearly intends to “lift[] the sovereignty of the States”); id. at 299 (Brennan, J., dissenting) (arguing that Parden stands for the proposition that “States surrendered their sovereignty . . . when they granted Congress the power to regulate commerce”), the majority recognized that Parden “turned on . . . waiver,” id. at 282 (majority opinion); cf. supra note 64. But see Baker, supra note 13, at 169 (arguing that Employees placed Parden “firmly” on the ground that Congress could abrogate sovereign immunity by its commerce power). The FLSA had a much broader reach than the FELA, which affected only a distinct commercial activity.71×71. Compare Emps., 411 U.S. at 285 (“We deal here with problems that may well implicate elevator operators, janitors, charwomen, security guards, secretaries, and the like in every office building in a State’s governmental hierarchy.”), with Parden, 377 U.S. at 187 (stating that the FELA applied to “common carrier[s] by railroad in interstate commerce”). And Congress’s intention to condition so much state conduct on a waiver of immunity simply by authorizing suit in any “court of competent jurisdiction”72×72. 29 U.S.C. § 216(b). was too ambiguous, since federal courts were presumably not “competent” to hear suits against states.73×73. Emps., 411 U.S. at 284; see id. at 285–87. In his concurrence, Justice Marshall added another layer of concern: Alabama had chosen to engage in the regulated conduct after the FELA was enacted, whereas in Employees, the State was already operating hospitals and schools when the FLSA was amended to apply to those entities. See id. at 296 (Marshall, J., concurring in the result). So, in this situation, “the State had no true choice at all and . . . did not voluntarily consent.” Id. Employees thus added a hurdle to valid waivers that maps well onto now-familiar concerns: clear congressional intent ensures that a state’s decision to engage in “waiving conduct” was done with full awareness of the consequences and was thus knowing.

The Court heightened that hurdle in Edelman v. Jordan74×74. 415 U.S. 651 (1974). by holding that Illinois’s participation in a Social Security Act program, which required states to agree to administer funds in accordance with federal regulations, did not waive its immunity.75×75. See id. at 673. The lower court had relied on Parden to hold otherwise. See id. at 671–72. The Court explained that Congress had not been clear enough about its intent to induce waiver,76×76. Id. at 672 (distinguishing Parden because the FELA “authorized suit by designated plaintiffs against a general class of defendants which literally included States or state instrumentalities”). then took a step beyond Employees’ rule. After noting that constructive consent was not “commonly associated” with other constitutional rights,77×77. Id. at 673. it said it would “find waiver only where stated ‘by the most express language or by such overwhelming implications from the text as [would] leave no room for any other reasonable construction.’”78×78. Id. (quoting Murray v. Wilson Distilling Co., 213 U.S. 151, 171 (1909)). Notably, the Court in Murray v. Wilson Distilling Co., 213 U.S. 151, was interpreting a statute to determine whether a suit was against the state, and not whether the statute was explicit enough in requiring that states consent to waiver of their sovereign immunity. See id. at 170–71. Once again, the Court was concerned that states might unknowingly waive their immunity by engaging in otherwise innocuous conduct.

Edelman’s dissenters highlighted how much the Court had shifted its conception of waiver validity in the decade since Parden. Justice Douglas contended that Edelman and Parden were indistinguishable: in both cases, the state chose to join a federal-state program at a time when precedent allowed plaintiffs to recover from the state.79×79. Edelman, 415 U.S. at 686–87 (Douglas, J., dissenting) (“Today’s holding . . . necessitates an express overruling of several of our recent decisions. But it was against the background of those decisions that Illinois continued its participation in the federal program . . . .” Id. at 687.). Justice Marshall more explicitly invoked Johnson’s core concepts. He argued that Illinois “voluntarily obligated itself to comply” with the program’s terms “with full knowledge” that aid recipients could bring suit.80×80. Id. at 693 (Marshall, J., dissenting) (emphases added). Both dissenters appeared motivated by the same concern underlying the majority’s clear statement rule: Did the state intend to waive its immunity by its conduct? The majority simply harbored a stronger presumption that the question would be answered in the negative.

The Court reaffirmed the majority’s approach in a trio of cases that did not directly address constructive waiver. Pennhurst State School & Hospital v. Halderman81×81. 465 U.S. 89 (1984). reiterated that a state’s consent to suit must “be unequivocally expressed.”82×82. Id. at 99 (citing Edelman, 415 U.S. at 673); see also Welch v. Tex. Dep’t of Highways & Pub. Transp., 483 U.S. 468, 473 (1987) (plurality opinion) (emphasizing express statement requirement). Soon the Court added that even if a state generally waived its immunity via state statute or constitution, the waiver had to be specific to federal court.83×83. See Atascadero State Hosp. v. Scanlon, 473 U.S. 234, 241 (1985); cf. Smith v. Reeves, 178 U.S. 436, 441 (1900) (holding that a state statute expressed consent to suit only in state, not federal, court, even though the statute consented to suit in general terms). The second question in Atascadero State Hospital v. Scanlon, 473 U.S. 234, which involved spending clause waivers, id. at 246, is briefly considered in section II.B.2, pp. 2566–68. And Port Authority Trans-Hudson Corp. v. Feeney84×84. 495 U.S. 299 (1990). established that a state’s express statutory waiver must be read with a narrowing presumption.85×85. Id. at 306 (“[T]he Court has . . . construed such ambiguous and general consent to suit provisions, standing alone, as insufficient to waive Eleventh Amendment immunity.”). In each case, the Court phrased its animating concern differently: Did the state act willfully86×86. See Atascadero, 473 U.S. at 241. or “intend[]”87×87. Id. at 238 n.1; see also Port Auth., 495 U.S. at 308. to consent to suit? Was it haled into court against its own “will”?88×88. Pennhurst State Sch. & Hosp. v. Halderman, 465 U.S. 89, 100 (1984). But each was just a different articulation of the Court’s classic touchstones for valid waiver. In demanding increasingly clear statements from states, the Court built safeguards to increase confidence that the state wanted to waive its immunity (voluntariness), and that the state did so with full awareness that the statute would be read as a waiver (knowingness).

In College Savings, the Court invoked those same concerns to shut the door on constructive waiver.89×89. Coll. Sav. Bank v. Fla. Prepaid Postsecondary Educ. Expense Bd., 527 U.S. 666, 680 (1999) (“We have never applied the holding of Parden to another statute, and in fact have narrowed the case in every subsequent opinion in which it has been under consideration.”); see also Siegel, supra note 33, at 1205 (questioning if “any[]” of Parden “remained at all” by the time of College Savings). According to Justice Scalia, constructive waiver violated the principle that waivers must be “altogether voluntary.”90×90. Coll. Sav., 527 U.S. at 681 (quoting Beers v. Arkansas, 61 U.S. (20 How.) 527, 529 (1858)). It forced states into a choice between refraining from lawful activity and waiving their rights, which “automatically” rendered the decisions involuntary.91×91. Id. at 687. The Court turned to a comparison with “other constitutionally protected privileges”92×92. Id. at 681 (emphasis omitted). and concluded that, under Johnson’s “classic description” of waiver, constructive waiver was necessarily invalid.93×93. Id. at 682. The Court thus suggested that, in overturning the constructive waiver theory, it was restoring sovereign immunity to its rightful place as a right protected from unknowing, involuntary waivers, just like a criminal defendant’s right to a trial. But as the following section will argue, that was a misleading, if not ironic, framing of the problem. The Court had already applied Johnson to Miranda in ways that allowed the kind of waiver that it found impermissible in College Savings.

3. Comparing the Standards. — Throughout Parden and the cases that followed, the Court both implicitly and explicitly raised the same concerns that drove its Fifth Amendment waiver cases: namely, the voluntariness and knowingness of the waiver. But the similarities stop there. Consider first how the Court analyzed voluntariness in each context. College Savings defined constructive waivers as situations where the rights holder is merely “put on notice” that lawful conduct will waive the right; then, it held that such waivers could never meet the voluntariness requirement.94×94. Id. at 681; see id. at 681–82. But is that not precisely what the Court allowed time and time again in Miranda’s progeny? Take the following analogy95×95. Justice Scalia chose an alternate analogy in College Savings: What if Congress amended a securities law, informing the public that anyone who violated it waived their right to a jury trial if prosecuted? See id. at 681–82. The Court would indeed probably refuse to endorse something so explicit, but it is useful to consider the subtler example provided above. : an individual, having been informed that she has a right to silence, remains silent — that is, exercises her right — then, after hours of interrogation, answers a few questions. Voluntary waiver? Apparently so, says Berghuis, because the suspect could have just stayed silent.96×96. See Berghuis v. Thompkins, 560 U.S. 370, 386 (2010). But in College Savings, the Court was unsatisfied with the analogous argument that a state could have just stayed away from the waiving conduct; instead, it held that a choice between avoiding lawful conduct and waiving immunity rendered the decision presumptively involuntary.97×97. See Coll. Sav., 527 U.S. at 686–87. That articulation of voluntariness is a far cry from the concept’s application to Miranda waivers, where the Court had long abandoned equivalence between voluntariness and “free choice.”98×98. Colorado v. Connelly, 479 U.S. 157, 170 (1986).

Consider also the knowingness inquiry. When constructive waiver was permissible, the Court was uncomfortable with laws that did not make it glaringly obvious that the consequence of certain conduct would be a waiver of immunity.99×99. See Emps. of the Dep’t of Pub. Health & Welfare v. Dep’t of Pub. Health & Welfare, 411 U.S. 279, 284–85 (1973). But when it came to the Fifth Amendment, the Court was comfortable finding that a suspect’s failure to explicitly invoke a right was a valid waiver of it, even if the rights holder did not necessarily know what was required of her in order to preserve the right.100×100. See supra notes 45–48 and accompanying text. The Court was also comfortable finding that criminal suspects knowingly waived their rights after police actively deceived their attorneys,101×101. See Moran v. Burbine, 475 U.S. 412, 424 (1986). or after suspects acted while mentally incapacitated.102×102. See Connelly, 479 U.S. at 170–71.

Finally, comparing the Court’s background presumptions about ambiguity may most clearly show how much easier it is to waive criminal protections than sovereign immunity. Unless a criminal defendant unambiguously asserts her right to silence or counsel, her conduct is presumed to constitute a voluntary, knowing waiver.103×103. See Berghuis v. Thompkins, 560 U.S. 370, 386–87 (2010). Post College Savings, the Court is satisfied that a state’s waiver is voluntary and knowing only if the state unambiguously states it is waiving its immunity.104×104. See Coll. Sav. Bank v. Fla. Prepaid Postsecondary Educ. Expense Bd., 527 U.S. 666, 680 (1999). For a thorough explication of this inconsistency, see Aaron Tang, Reverse Political Process Theory, 70 Vand. L. Rev. 1427, 1454–60 (2017) (arguing that the Court devised a rule for Miranda waivers that is “exactly the opposite” of its clear statement approach to statutory waivers of sovereign immunity, id. at 1459).

B. Waiver by Bargain

1. Plea Bargains. — Criminal defendants can waive constitutional rights like their right to trial and privilege against self-incrimination in exchange for a reduced sentence or charge by pleading guilty. Here, too, Johnson guides the validity test: guilty pleas “must be voluntary[,] . . . knowing, intelligent acts done with sufficient awareness of the relevant circumstances and likely consequences.”105×105. Brady v. United States, 397 U.S. 742, 748 (1970) (citing, inter alia, Johnson v. Zerbst, 304 U.S. 458, 464 (1938)); see also Godinez v. Moran, 509 U.S. 389, 396, 400–02 (1993); McCarthy v. United States, 394 U.S. 459, 466 (1969). Courts must treat the accused with the “utmost solicitude” to ensure they have a “full understanding” of the plea and its consequences.106×106. Boykin v. Alabama, 395 U.S. 238, 243 (1969); see id. at 244. But when it comes to policing the terms of the plea bargain, that solicitude has waned.

To begin with, a waiver by plea bargain can be deemed voluntary despite enormous disparities between the choices offered to the defendant. Consider Brady v. United States.107×107. 397 U.S. 742. The Court upheld as voluntary Brady’s guilty plea to avoid a possible death penalty.108×108. See id. at 749. That defendant Brady “preferred” and so chose a guilty plea to a jury trial that could have resulted in capital punishment, id., explained the Court, did not render his plea involuntary, id. at 750. According to the Court, Johnson merely required that the defendant be aware of the plea’s direct consequences,109×109. The Court later read this to include a requirement that the defendant understand the individual elements of the offense with which he has been charged, such that he understands what exactly he is pleading guilty to. See Henderson v. Morgan, 426 U.S. 637, 644–46 (1976). and that the plea not be induced by threats, misrepresentations, or “improper” promises like bribes.110×110. Brady, 397 U.S. at 755 (quoting Shelton v. United States, 246 F.2d 571, 572 n.2 (5th Cir. 1957) (en banc), rev’d on confession of error on other grounds, 356 U.S. 26 (1958)); see also id. at 748. In Brady’s case, since the state did not induce the plea by “threaten[ing] physical harm or by mental coercion overbearing [his] will,” Brady was able to weigh his options and validly waive his rights.111×111. Id. at 750 (“[N]or is there evidence that Brady was so gripped by fear of the death penalty or hope of leniency that he did not or could not, with the help of counsel, rationally weigh the advantages of going to trial against the advantages of pleading guilty.”). No matter that the alternative was risking the death penalty; as with Miranda waivers,112×112. See Colorado v. Connelly, 479 U.S. 157, 170 (1986) (“The voluntariness of a waiver of this privilege has always depended on the absence of police overreaching . . . .”). the only question was whether the state had impermissibly interfered with the rights holder’s decisionmaking.

Bordenkircher v. Hayes113×113. 434 U.S. 357 (1978). diluted the voluntariness test even more by allowing prosecutors to effectively punish defendants for refusing a waiver by plea bargain.114×114. See Doug Lieb, Note, Vindicating Vindictiveness: Prosecutorial Discretion and Plea Bargaining, Past and Future, 123 Yale L.J. 1014, 1017, 1033–39 (2014) (explaining that “[t]he law specifically permits . . . penalizing defendants for going to trial in an effort to induce a guilty plea,” id. at 1017). See generally Christine Schaack McGoey, Note, Prosecutorial Vindictiveness and the Plea Bargaining Exception: Bordenkircher v. Hayes Four Years Later, 13 Loy. U. Chi. L.J. 1011 (1982). The Court held that a prosecutor could threaten to reindict the accused on a charge that would result in mandatory life imprisonment if the accused did not plead guilty to the original offense, which carried a far lower penalty.115×115. See Bordenkircher, 434 U.S. at 358–59, 365. In light of the practical realities of plea bargaining, and since the defendant had been informed of the plea’s terms, his waiver was valid even though it was rendered under threat of a new, more severe charge and punishment.116×116. See id. at 360–62.

On the same day it decided Brady, the Court weakened the knowingness requirement too. Parker v. North Carolina117×117. 397 U.S. 790 (1970). upheld a defendant’s guilty plea despite his counsel’s error in assessing the admissibility of the defendant’s earlier confession; the defendant’s partial ignorance as to the costs of refusing the waiver did not render the waiver unknowing.118×118. See id. at 797–98. United States v. Ruiz119×119. 536 U.S. 622 (2002). went further by holding that the Constitution does not require “preguilty plea disclosure of [exculpatory] impeachment information” by the prosecution, because such information implicates a trial’s fairness but not the plea’s voluntariness.120×120. Id. at 629. But what about the knowingness of a plea? The Court admitted that impeachment information might make the defendant more aware of the “likely consequences” of a plea but was apparently satisfied that this did not render the plea invalid.121×121. Id. Indeed, the Court had already found many forms of defendant “misapprehension” acceptable in guilty pleas.122×122. Id. at 630; see id. at 630–31 (noting that a court can accept guilty pleas “despite various forms of misapprehension under which a defendant might labor,” id. at 630, including misunderstanding the strength of the state’s case or the likely penalties, not recognizing a defense because of counsel’s failure to point it out, or missing a constitutional infirmity in grand jury proceedings (citing, inter alia, United States v. Broce, 488 U.S. 563, 573 (1989); Tollett v. Henderson, 411 U.S. 258, 267 (1973); Brady v. United States, 397 U.S. 742, 757 (1970))).

2. Spending Clause Waivers. — Pursuant to its Spending Clause power, Congress can attach conditions to its funding, including sovereign immunity waivers.123×123. See Coll. Sav. Bank v. Fla. Prepaid Postsecondary Educ. Expense Bd., 527 U.S. 666, 686 (1999) (citing South Dakota v. Dole, 483 U.S. 203 (1987)); see also Rebecca E. Zietlow, Federalism’s Paradox: The Spending Power and Waiver of Sovereign Immunity, 37 Wake Forest L. Rev. 141, 186 (2002). States may thus exchange such waivers for a benefit from the federal government. “The legitimacy of Congress’ power to legislate under the spending power” rests in part “on whether the State voluntarily and knowingly accept[ed] the terms” of its exchange.124×124. Pennhurst State Sch. & Hosp. v. Halderman, 451 U.S. 1, 17 (1981) (emphasis added). The state must be able to “ascertain what is expected of it”125×125. Id. in accepting the condition and “make an informed choice.”126×126. Id. at 25.

The Supreme Court has seldom addressed this test in the context of sovereign immunity.127×127. The Court had the opportunity to confront this in Edelman and Employees, but it analyzed both cases in the context of constructive waiver, avoiding the Spending Clause question. See Michael T. Gibson, Congressional Authority to Induce Waivers of State Sovereign Immunity: The Conditional Spending Power (and Beyond), 29 Hastings Const. L.Q. 439, 459–60 (2002); see also Emps. of the Dep’t of Pub. Health & Welfare v. Dep’t of Pub. Health & Welfare, 411 U.S. 279, 284–85 (1973). When it has chosen to do so, however, it has strictly construed the requirements that the state acted knowingly and voluntarily, just as it did with constructive waivers. In Atascadero State Hospital v. Scanlon,128×128. 473 U.S. 234 (1985). for example, the Court explained that “mere receipt of federal funds cannot establish” a waiver of immunity if the bargain terms are not clear — rather, Congress’s condition must be unambiguously expressed.129×129. Id. at 246; see id. at 246–47. The Court affirmed that standard in South Dakota v. Dole, 483 U.S. 203 (1987), which did not involve sovereign immunity. See id. at 207. This requirement sounds in knowingness. More recently, in Sossamon v. Texas,130×130. 563 U.S. 277 (2011). the Court held that a statute authorizing “appropriate relief against a government” did not require a waiver of sovereign immunity clearly enough to induce valid consent from a state accepting funds under that statute, because “appropriate relief” usually does not include damages against a state.131×131. Id. at 285 (quoting 42 U.S.C. § 2000cc-2(a)); see id. at 285–92. Though the Court phrased its concern in terms of “intent,”132×132. Id. at 288. its underlying goal can be understood to be the securing of knowing waivers: Could the Court be confident that the state knew exactly what the terms of the bargain were and, therefore, relinquished its immunity because it wanted to?

College Savings elaborated on another limit to Spending Clause waivers during its discussion of constructive waiver.133×133. See Gibson, supra note 127, at 465 (“Justice Scalia . . . went out of his way to justify the use of the Conditional Spending power as a counterweight to state sovereign immunity.”). Where Congress offers states a gift or gratuity in exchange for a waiver of sovereign immunity (and threatens to withhold it if the state retains its immunity), a state’s decision to accept the gift can entail a valid waiver.134×134. Coll. Sav. Bank v. Fla. Prepaid Postsecondary Educ. Expense Bd., 527 U.S. 666, 686 (1999) (citing Dole, 483 U.S. 203). But Congress may not just offer anything at all. There is a boundary between permissible and impermissible gifts, and it is delineated by voluntariness: in some cases, the withheld gift might be so large as to be a sanction that renders the waiver coercive.135×135. See id. at 687. The Court explained that where Congress threatens to exclude states from otherwise lawful activity for refusing to waive their immunity, it is automatically sanctioning the state and the resulting waiver is automatically involuntary. Id. That is constructive waiver and the point at which implied waivers by conduct and bargained-for waivers meet. The Court did not elaborate on or enforce that limit until after College Savings in a case that had nothing to do with sovereign immunity.136×136. See Nat’l Fed’n of Indep. Bus. v. Sebelius, 567 U.S. 519, 625 (2012) (Ginsburg, J., concurring in part, concurring in the judgment in part, and dissenting in part) (noting that this was the “first time ever” that the Court had found Congress’s exercise of the spending power too coercive). In NFIB v. Sebelius,137×137. 567 U.S. 519. Chief Justice Roberts suggested that a financial inducement becomes coercive at some point between “relatively mild encouragement” and “a gun to the head.”138×138. Id. at 581 (plurality opinion). Threatening twenty percent of a state’s budget by cutting Medicare funding was sufficient to invalidate the Spending Clause argument in NFIB. See id. Applying this rule to the immunity context, a state’s waiver of its sovereign immunity as a condition of accepting federal funds may not be truly voluntary if the benefit of the bargain is so significant, or the cost of declining so large, as to prevent meaningful choice.139×139. Cf. Berman, supra note 25, at 1346–47 (concluding that the ACA’s Medicaid provision was unconstitutionally coercive because it left states with no legitimate choice). But see Samuel R. Bagenstos, The Anti-leveraging Principle and the Spending Clause After NFIB, 101 Geo. L.J. 861, 865 (2013) (arguing that NFIB stands for a narrower principle).

3. Comparing the Standards. — For both types of bargained-for waivers, the conditions of the bargain must be clearly stated. But beyond that, the tests differ significantly. The difference is most apparent in the Court’s line-drawing on voluntariness. When it comes to criminal defendants bargaining with the state about their rights, the Court’s voluntariness/coercion inquiry focuses on the propriety of the state’s conduct and turns a blind eye to even the largest disparity between the choices offered to the rights holder.140×140. See Brady v. United States, 397 U.S. 742, 750 (1970). When it comes to sovereigns, the voluntariness of the bargained-for waiver turns almost entirely on that disparity,141×141. See Coll. Sav. Bank v. Fla. Prepaid Postsecondary Educ. Expense Bd., 527 U.S. 666, 687 (1999). and the state’s freedom from improper congressional interference with the decisionmaking itself is presumed.

Moreover, in the sovereign context, there is a per se rule against bargains that “sanction” states for failing to waive a right.142×142. Id. But the Court effectively permits the state to inflict a penalty of a heightened charge on criminal defendants who refuse to plead guilty.143×143. See sources cited supra note 114 and accompanying text. Why? Because “the simple reality [is] that the prosecutor’s interest at the bargaining table is to persuade the defendant to forgo his right to plead not guilty.”144×144. Bordenkircher v. Hayes, 434 U.S. 357, 364 (1978). Having condemned sanctions in College Savings as nullifying the voluntariness of state waivers, the Court nevertheless accepts sanctions when assessing criminal waivers. The result is that the government may lawfully punish individuals,145×145. See id. at 372–73 (Powell, J., dissenting). but not states, for exercising (or refusing to waive) their rights.

Finally, while both sets of rights holders must be made aware of the bargain’s terms, the Court has weakened a crucial aspect of the knowingness requirement for criminal waivers by accepting pleas by defendants from whom information about the consequences of going to trial was withheld.146×146. See supra p. 2566. The equivalent for states might be if the federal government withheld information about the consequences of a state’s refusing the waiver and accompanying funds. While there is no case testing such a situation, it is hard to imagine the Court accepting it.

C. Waiver by Litigation

1. Litigation Waiver of Criminal Protections. — If a criminal defendant is procedurally barred from raising a constitutional claim due to conduct in the course of litigation, what is the standard by which courts evaluate the defendant’s relinquishment of that claim? Here, the Court has not just made the classic waiver standard easier to satisfy; it has declined to apply it altogether, choosing instead to evaluate such waivers as “forfeitures” that need not be voluntary or knowing at all. The question typically arises for criminal defendants when they failed to raise an issue at trial. The traditional rule in federal court is that an “appellate court does not consider an issue not passed upon below,” though it has ultimate discretion.147×147. Singleton v. Wulff, 428 U.S. 106, 120 (1976); see id. at 121. The Federal Rules of Criminal Procedure allow for consideration of some “plain error[s]” in trials,148×148. Fed. R. Crim. P. 52(b). but the Court has clarified that the rules defining plain error are untethered from the conditions of the litigation waiver (was it voluntary or knowing?); they instead depend on the defaulted claim and the error.149×149. See United States v. Olano, 507 U.S. 725, 732–35 (1993).

When a defendant’s failure to satisfy a state procedural rule blocks them from raising a federal constitutional claim, they may challenge that default collaterally on habeas.150×150. Because the question of whether the constitutional right can be vindicated is a federal law question, a state procedural default does not necessarily end the matter. See Ralph S. Spritzer, Criminal Waiver, Procedural Default, and the Burger Court, 126 U. Pa. L. Rev. 473, 476 (1978). In Fay v. Noia,151×151. 372 U.S. 391 (1963). the Court applied Johnson to evaluate whether a federal court could hear such claims152×152. Id. at 439 (“The classic definition of waiver enunciated in Johnson v. Zerbst — ‘an intentional relinquishment or abandonment of a known right or privilege’ — furnishes the controlling standard.” (citation omitted) (quoting Johnson v. Zerbst, 304 U.S. 458, 464 (1938))). : unless a petitioner “knowingly” forwent vindication of her federal claims in state court in a way that “deliberate[ly]” bypassed state procedures, the claim could go forward.153×153. Id. But Wainwright v. Sykes154×154. 433 U.S. 72 (1977). soon overruled Fay.155×155. Id. at 87 (internal quotation marks omitted); see id. at 87–88. Today, federal courts hearing habeas petitions must honor state procedural defaults unless the petitioner can show that doing so would lead to “prejudice,”156×156. Id. at 117 (Brennan, J., dissenting). and that there was “cause” for her default.157×157. See id. at 116. So, when it comes to litigation conduct, criminal defendants need no longer voluntarily, knowingly waive a constitutional claim at trial or in state court to be barred from airing it; instead, the right is presumptively lost if it is procedurally blocked.158×158. See Spritzer, supra note 150, at 513–14. This is not just a weakening but rather a rejection of the classic waiver standard.

2. Litigation Waiver of Sovereign Immunity. — If a state fails to invoke the sovereign immunity defense at trial but tries to raise it on appeal, what is the standard by which a federal court would evaluate that procedural waiver? The answer, according to Edelman, is simple and forgiving: the defense remains available to the state no matter what, whether it voluntarily and knowingly withheld the defense or not. A state’s failure to invoke sovereign immunity at trial does not foreclose raising it on appeal, because immunity is like a jurisdictional defense.159×159. Edelman v. Jordan, 415 U.S. 651, 677–78 (1974); see also Calderon v. Ashmus, 523 U.S. 740, 745 n.2 (1998).

Admittedly, a state can still waive its immunity through litigation in other ways: it can file suit in federal court160×160. See Gardner v. New Jersey, 329 U.S. 565, 574 (1947). or remove an action to federal court.161×161. Lapides v. Bd. of Regents of Univ. Sys. of Ga., 535 U.S. 613, 624 (2002). But the waiver bar set by those cases is extraordinarily high because, in both situations, the state chooses to take advantage of the federal forum and simultaneously tries to deny that forum’s jurisdiction by raising the immunity defense.162×162. See id. at 616. Viewed from that angle, these waivers require more than voluntariness and knowingness; they involve state exploitation of legal loopholes. From the Court’s perspective, recognizing such waivers is necessary to close those loopholes.163×163. See Siegel, supra note 33, at 1217 (noting the Court’s emphasis on the “‘inconsistency and unfairness’ that would inhere in permitting a state to first invoke, and then claim immunity from, federal jurisdiction” (quoting Lapides, 535 U.S. at 622)).

3. Waivers by Litigation Compared. — The contrast is stark in this pairing. On direct appeal, the standard applied to states who default on their constitutional claim is higher than is the classic waiver requirement: even a voluntary, knowing waiver at trial is not enough to bar an immunity defense on appeal. But the Court has abandoned any pretense of requiring voluntariness or knowingness for litigation waivers of criminal defendants’ constitutional claims. On direct appeal, a court’s decision to hear procedurally barred claims is the exception, not the rule. The standard for collateral review of state procedural defaults is equally strict: a rights holder can find that her constitutional claims were waived in state court without any knowledge or intent whatsoever.

III. The Double Standard Reversed

Together, Parts I and II present a puzzle. When assessing state and criminal waivers, the Court has invoked the same underlying requirements of voluntariness and knowingness. Often, it has done so under the guise of seeking equality of treatment across constitutional rights, be they state or individual. In practice, the Court has applied those concepts in ways that make it far easier for criminal defendants than states to waive their rights. But if what matters is protection from unknowing, involuntary waivers, we would expect the presumption against waivers of criminal protections to be stronger than that against waivers of sovereign immunity.164×164. See supra Part I, pp. 2554–56. The current rules appear to ignore the relative social and structural positions of states and criminal defendants. This Part considers why that might be.

To begin, if current standards reflect the opposite of the rights holders’ presumed capacities to knowingly, voluntarily waive their rights, it must be that the Court’s controlling concern is not waivers’ validity as classically defined. The first implication of the Court’s chosen double standard is thus that Johnson is canonical in name, but not in practice.

What other values might underlie its approach? One possibility is that the Court does not think it is equally important to protect all constitutional rights from invalid waivers. Specifically, perhaps it sees states’ rights as being of a higher order because they form the core of our constitutional structure. Scholars have suggested that the Eleventh Amendment is “an expression of the principles of federalism” and an attempt to protect states from federal interference.165×165. Margreth Barrett, Comment, The Denial of Eleventh Amendment Immunity to Political Subdivisions of the States: An Unjustified Strain on Federalism, 1979 Duke L.J. 1042, 1061; see also, e.g., Alden v. Maine, 527 U.S. 706, 713–15, 722 (1999). To allow states to waive their immunity too easily, then, would not just place their rights as sovereigns at risk; it would disturb the federal balance. This structural concern may also explain why the Court has looked to federal sovereign immunity as a strong analog for state sovereign immunity, even as it invokes comparisons to other rights.166×166. See Coll. Sav. Bank v. Fla. Prepaid Postsecondary Educ. Expense Bd., 527 U.S. 666, 682 (1999) (noting that federal sovereign immunity is “the closest analogy to the present case”); see also Gregory C. Sisk, A Primer on the Doctrine of Federal Sovereign Immunity, 58 Okla. L. Rev. 439, 443 (2005) (calling state sovereign immunity the “jurisprudential cousin” of federal sovereign immunity). Perhaps the Court is motivated to place states on a similar playing field to that of the federal government by shielding states from federal judicial intrusion.167×167. This explanation would also fit within the Court’s jurisprudence in the late twentieth century, which scholars have dubbed the “federalist revival.” See, e.g., Vicki C. Jackson, Federalism and the Uses and Limits of Law: Printz and Principle?, 111 Harv. L. Rev. 2180, 2213 (1998).

But the federalism concern explains only the Court’s reluctance to allow state waivers. For it to explain the contrasting leniency toward criminal waivers would require accepting that criminal constitutional protections do not serve a similarly integral structural function. Yet the ultimate goal of federalism is arguably to protect popular sovereignty and individual rights.168×168. See generally Akhil Reed Amar, Of Sovereignty and Federalism, 96 Yale L.J. 1425 (1987). And criminal constitutional rights exist to protect individuals when they are at their most vulnerable, coming head-to-head with the State.169×169. E.g., Miranda v. Arizona, 384 U.S. 436, 442–43 (1966). From this perspective, allowing the State (via its prosecutorial agents) to easily induce waivers of those rights directly undermines an important structural feature of our government too: the balance between the state and individual rights.

In order to accept this structural motivation, then, one would have to make a value judgment that federalism is simply more integral to our constitutional structure than is the protection of defendants from state abuses of power. The theory would be that it is more necessary to protect states from the federal government than it is to protect individuals from states. States’ rights are just different.170×170. Someone taking this position might add that this difference is reflected in sovereign immunity’s strong, preconstitutional roots. See generally, e.g., Bradford R. Clark, The Eleventh Amendment and the Nature of the Union, 123 Harv. L. Rev. 1817 (2010). The comparisons drawn in this Note suggest that the Court has taken some version of this position. Nonetheless, it has not explicitly defended this broader thesis, but has implied instead that the rights to a criminal trial and sovereign immunity are and should be treated equally.171×171. See Coll. Sav., 527 U.S. at 682. Given that it has not abided by that position in practice, its approach seems ripe for a defense.

The Court’s decisions suggest a supplemental explanation for its chosen double standard: a desire to avoid the societal costs of limiting criminal waivers. The Court frequently emphasizes that waivers are crucial to the criminal process. It has invoked concerns that inhibiting Fifth Amendment waivers would undermine society’s “substantial interest in securing admissions of guilt,”172×172. Moran v. Burbine, 475 U.S. 412, 427 (1986); see also Berghuis v. Thompkins, 560 U.S. 370, 382 (2010) (finding that suppressing confessions in cases where assertion of the right was ambiguous “would place a significant burden on society’s interest in prosecuting criminal activity”). and has voiced similar doubts in its plea-bargaining decisions.173×173. See Missouri v. Frye, 566 U.S. 134, 144 (2012) (noting that plea bargaining “is the criminal justice system” and not just an adjunct to it (quoting Robert E. Scott & William J. Stuntz, Plea Bargaining as Contract, 101 Yale L.J. 1909, 1912 (1992))); Bordenkircher v. Hayes, 434 U.S. 357, 363 (1978); see also Lieb, supra note 114, at 1039 (pointing out the Court’s struggle to regulate plea bargaining as it recognizes that practice’s centrality to the operation of the criminal system). In short, the Court fears that if it makes it too hard for defendants to waive their rights, the criminal process will grind to a halt and preclude findings of guilt. Perhaps this explains why it will more readily allow criminal waivers than state immunity waivers.

But that explanation is also incomplete, because it ignores the fact that limiting state immunity waivers incurs similar societal costs: states become harder, if not impossible, to sue, which precludes redress for individuals harmed by the government, effectively permitting states to violate the law with impunity.174×174. Cf. Erwin Chemerinsky, Against Sovereign Immunity, 53 Stan. L. Rev. 1201, 1202, 1214 (2001) (“Sovereign immunity . . . prevents accountability, even when the government egregiously violates the Constitution and federal laws.” Id. at 1214.). If the criminal waiver decisions were any indication, one would expect the Court to respond to this cost by making sovereign waivers easier too. That is not what has happened. The Court barely mentions the accountability cost in the sovereign immunity cases and certainly has not used it to find more sovereign waivers.175×175. 1 See id. at 1224 (noting that the Supreme Court has been unpersuaded by this accountability-based criticism of sovereign immunity and has only been “expanding . . . the reach” of the doctrine). The waiver double standard must, then, incorporate a second implicit hierarchy. Just as it views state and criminal rights differently, the Court views state violations of the law differently from individual violations. Its waiver jurisprudence seems to rest, in part, on a judgment that it is costlier to undermine (and thus more important to protect) the process by which we hold individuals accountable for their wrongs than it is to block the process by which we adjudicate state wrongs.

IV. Conclusion

The Supreme Court’s approaches to waivers of state sovereign immunity and criminal constitutional protections are conceptually linked. But in practice, the Court has defined waiver validity differently, if not contradictorily, in each context. Johnson’s classic requirement has been diluted as applied to criminal constitutional protections; by contrast, the presumptions against waivers of sovereign immunity are extremely powerful. If the Court were concerned with whether a waiver was knowing and voluntary, however, its presumptions would be flipped. There would be a double standard, but it would be one that made it harder for criminal defendants than states to waive their rights. The Court’s approach may best be understood, then, as resting not on broadly applicable principles of valid constitutional waiver, but rather on implicit assumptions about the importance and relative costs of protecting certain state versus criminal rights. But to the extent that is the case, the Court has yet to explicitly defend its current rules in that light. If the double standard is here to stay, such a defense may be overdue.