Constitutional Validity of Pleas

Constitutional Validity of Pleas

Rule 11(a) lists the pleas a defendant may enter, but the Constitution adds a separate requirement about the quality of the choice. Whatever label a plea carries, a guilty plea supports a conviction only if it represents a voluntary and intelligent choice among the alternative courses of action open to the defendant.¹ North Carolina v. Alford applied that standard and traced it to Boykin v. Alabama and to earlier decisions, so it is an established test and not something devised for the case.² The test governs every guilty plea, including a plea entered by a defendant who professes innocence. The later sections on the Rule 11 colloquy and on plea agreements therefore measure their procedures against it.

A standard this exacting fits what a plea accomplishes. A defendant who pleads guilty consents to judgment without a trial and surrenders the rights a trial would have secured, among them the jury, the confrontation of accusers, the opportunity to present witnesses in defense, the privilege against self-incrimination, and proof beyond a reasonable doubt.³ The stakes are also systemic. Missouri v. Frye observed that guilty pleas produce about ninety-seven percent of federal convictions and ninety-four percent of state convictions, so the validity of the plea decides the validity of nearly every conviction the system produces.⁴

The plea also reshapes what the defendant may contest afterward, which is another reason its validity carries so much weight. Blackledge v. Perry recounts that Tollett v. Henderson described a guilty plea as a break in the chain of events that came before it, so that a defendant who pleads may not later raise independent claims about constitutional violations that preceded the plea and is limited to attacking the voluntary and intelligent character of the plea itself.⁵ Because no trial occurs, nothing has tested the earlier steps, and the plea's own validity becomes the one gate left between the defendant and a conviction. Blackledge marks the edge of that rule. The defendant there pleaded guilty to a felony charge the State brought after he had already been charged with a misdemeanor, and the Court held that the plea did not bar a federal habeas attack, because the claim went to the State's very power to bring him into court on the felony charge.⁶ The claim in Tollett, a challenge to how the grand jury was selected, concerned a defect a new indictment could have cured, which is why it fell within the break-in-the-chain rule and Perry's claim did not.

The standard has two operative parts, and each answers a different question. Voluntariness asks whether the defendant was pushed: because the plea is the defendant's own admission, it must be free of compulsion within the meaning of the Fifth Amendment.⁷ Intelligence asks whether the defendant understood, since a waiver of rights is effective only when made with sufficient awareness of the relevant circumstances and likely consequences. The split earns its place because the two failures leave different evidence. Pressure shows up in threats, promises, and the conditions of custody, while misunderstanding shows up in what the judge and counsel said, or failed to say, about the charge and the sentence. A plea must satisfy the test as a whole, and both halves are required.

The rest of this section takes the test in four steps. The first identifies which trial rights a plea surrenders and why the federal standard reaches state courts as well. The second turns to Boykin v. Alabama and the rule that a waiver cannot be presumed from a silent record. The third defines voluntariness and sorts the pressure that invalidates a plea from the pressure that does not, and the fourth addresses whether the choice was intelligent and how the quality of counsel's advice bears on that question. The inquiry starts with what a plea gives up, because a waiver can be judged only once the right waived is known.

Trial Rights Waived by Pleading

A waiver can be judged only after the right waived has been identified, so the voluntary-and-intelligent test begins with an inventory of what a guilty plea gives up. The inventory is long because of what a plea is. In Boykin v. Alabama the Court explained that a guilty plea is more than an admission that the accused committed certain acts: it is itself a conviction, and nothing remains but to enter judgment and fix punishment.¹⁰ Once the plea completes the conviction, the protections a trial would have supplied on the way to a verdict are gone, because no trial occurs for them to govern. Those protections come from two parts of the Bill of Rights. The Sixth Amendment lists the guarantees of an accused in a criminal prosecution: a speedy and public trial by an impartial jury, notice of the accusation, confrontation of the witnesses against him, compulsory process for witnesses in his favor, and the assistance of counsel.¹¹ The Fifth Amendment adds the privilege against compelled self-incrimination, and a plea implicates that privilege most directly.¹²

The privilege comes first in Boykin's list, and the reason is nearly mechanical. A defendant who goes to trial may stay silent and leave the State to prove guilt with other evidence, whereas a defendant who pleads guilty supplies the proof personally and so becomes a witness against himself within the meaning of the Fifth Amendment.¹³ The privilege protects an accused's communications, as distinct from physical evidence such as fingerprints or a blood sample, and an admission of guilt spoken in open court is a communication in the plainest sense.¹⁴ The privilege carries real weight at trial because the State may not make silence expensive: in Griffin v. California the Court held that a prosecutor's comment on the accused's failure to testify violates the Self-Incrimination Clause, since comment makes the assertion of the privilege costly.¹⁵ A plea surrenders that whole protection in a single statement. The privilege also explains why voluntariness is measured against the Fifth Amendment, because the privilege is fulfilled only when the defendant may remain silent unless he chooses to speak in the "unfettered exercise of his own will."¹⁶

The other two rights Boykin names, trial by jury and confrontation, depend on a trial taking place, and a plea removes the trial.¹⁷ The Sixth Amendment promises an impartial jury, and a defendant who admits guilt in open court leaves no question for a jury to decide.¹⁸ The confrontation guarantee entitles the accused to be confronted with the witnesses against him, and Rule 11 describes the trial-stage version of that right as the right to confront and cross-examine adverse witnesses.¹⁹ After a plea no prosecution witness takes the stand, so the State's evidence is never tested by cross-examination.

Boykin's three rights are a floor and not the full list. Rule 11(b)(1)(E) names, as trial rights the defendant must be told about, the rights to confront and cross-examine adverse witnesses, to be protected from compelled self-incrimination, to testify and present evidence, and to compel the attendance of witnesses.²⁰ The last two items trace to the Sixth Amendment's compulsory process guarantee, which gives the accused the means to obtain witnesses in his favor, and Washington v. Texas treated that guarantee as binding on the states through the Fourteenth Amendment.²¹ A defendant who pleads gives up the defense case as well: no one is subpoenaed and no defense witness is heard. Justice Douglas, concurring in Santobello v. New York, collected the same roll call, describing a guilty plea as a waiver of the rights to a jury trial, to confront accusers, to present witnesses in defense, to remain silent, and to be convicted only on proof beyond a reasonable doubt.²² That last right rests on due process rather than on the Sixth Amendment's text, because, as In re Winship held, the Due Process Clause protects the accused against conviction except on proof beyond a reasonable doubt of every fact necessary to constitute the crime.²³ A plea eliminates the prosecution's burden entirely, since the defendant's admission replaces the evidence the burden would have required the State to produce.

Two Sixth Amendment guarantees are affected differently, because the plea does not simply extinguish them. The right to counsel carries over to the plea decision itself, since the need for a lawyer peaks at the moment of choosing whether to plead. United States v. Wade grounded the guarantee in the principle that the accused need not stand alone against the State at any stage where counsel's absence might derogate from the right to a fair trial, and its discussion of Hamilton v. Alabama shows why arraignment qualifies, since available defenses can be lost for good if they are not raised there.²⁴ A defendant may still decline a lawyer at the plea stage, and the warnings required for that choice are lighter than those required to represent oneself at trial. In Iowa v. Tovar the Court found the Sixth Amendment satisfied when the trial court informs the accused of the nature of the charge, of the right to be counseled on the plea, and of the range of allowable punishments.²⁵ The lighter warnings follow from what a lawyer does at this stage, which is to advise on whether to plead, so the defendant needs to know that help is available without hearing every detail of what it would involve. The same list shows that the notice guarantee carries over as well, since a plea can be intelligent only if the defendant knows what the charge is.

Every right on this list binds state courts as well as federal ones, and that fact controls how a state-court plea is tested. The Fourteenth Amendment forbids a state to deprive any person of liberty without due process of law.²⁶ A Bill of Rights guarantee that is fundamental and essential to a fair trial is made obligatory on the states through that clause.²⁷ Each of Boykin's three rights has its own incorporation authority: the privilege through Malloy v. Hogan, the jury right through Duncan v. Louisiana, and confrontation through Pointer v. Texas.²⁸ Malloy held that the Fifth Amendment privilege is protected by the Fourteenth Amendment against abridgment by the states.²⁹ Boykin drew the consequence for pleas: "The question of an effective waiver of a federal constitutional right in a proceeding is of course governed by federal standards."³⁰ The reason is structural. If each state set its own test for what counts as giving up a federal right, a lenient local rule would shrink a guarantee the Constitution extends nationwide, and a defendant's protection would depend on the courthouse in which the plea happened to be entered.

Exam facts that place a plea in a state court invite the mistake of testing the waiver against the state's own plea rules alone. Federal standards govern whether the waiver of these federal rights was effective, and state procedure may add protection without lowering that floor.³¹

None of these rights prevents a defendant from pleading guilty, because even the most basic of them can be waived. United States v. Mezzanatto observed that the most basic rights of criminal defendants are subject to waiver, and it cited Boykin for the proposition that a knowing and voluntary guilty plea waives the privilege, the jury right, and confrontation.³² United States v. Ruiz restated the package: a plea forgoes the fair trial and its accompanying guarantees, and the Constitution insists that the plea be voluntary and that the related waivers be knowing and made with sufficient awareness of the likely consequences.³³ The measure of waiver is the one Johnson v. Zerbst used for counsel, an intentional relinquishment or abandonment of a known right or privilege.³⁴ Justice Powell, concurring in Edwards v. Arizona, put the test in plain terms: whatever the right, the question is whether the actor fully understands it and voluntarily intends to relinquish it.³⁵ That is the voluntary-and-intelligent standard seen from the waiver side.

The waiver also reaches only as far as the defendant's choice does. In Blakely v. Washington the defendant pleaded guilty to kidnapping, and the facts he admitted supported a maximum of 53 months, yet the judge imposed 90 months on a finding of deliberate cruelty that the defendant had not admitted, a sentence the Court held violated the jury right.³⁶ The Court explained that after a guilty plea the State may still seek judicial sentence enhancements, but only if the defendant stipulates to the relevant facts or consents to judicial factfinding.³⁷ A plea therefore waives the jury as to what the plea admits, and any further waiver has to come from the defendant. What remains is how a court may find that a waiver of this kind actually occurred, and Boykin answers that a silent record cannot supply it, which the next section takes up.

The Silent-Record Rule

The last section listed what a guilty plea surrenders. The next question is how a court can know the defendant really surrendered those rights. Boykin v. Alabama answers it, and its facts show the problem at its starkest. A grand jury returned five indictments against the 27-year-old defendant for common-law robbery, an offense punishable by death in Alabama. The trial court found him indigent and appointed counsel, and three days later, at his arraignment, he pleaded guilty to all five.³⁹ So far as the record showed, the judge asked him no questions about the pleas, and he never addressed the court.⁴⁰ The pleas did not end the case, because Alabama required a jury to fix punishment even after a guilty plea. At that proceeding the State presented its eyewitnesses and defense counsel cross-examined only briefly. The defendant neither testified nor offered evidence about his background, and the jury imposed death on each indictment.⁴¹

A federal question about a plea normally reaches the Supreme Court only if it was raised below, and a defendant who said nothing at arraignment might have been expected to lose on that ground. Alabama's automatic-appeal statute for capital cases removed the obstacle. It required the reviewing court to comb the record for any error prejudicial to the defendant, even one counsel never raised, and it was the only vehicle for plain-error review in Alabama criminal appeals.⁴² Plain error is an obvious mistake apparent from the record itself, which a court may correct although nobody objected. With the question properly presented, the Court held that it was error, plain on the face of the record, for the trial judge to accept the guilty plea without an affirmative showing that it was intelligent and voluntary.⁴³

The missing questions mattered because of what a plea is. A guilty plea is itself a conviction, so nothing remains after it except entering judgment and fixing punishment, and the plea carries the weight of a verdict.⁴⁴ For the standard governing its validity, the Court borrowed from the waiver-of-counsel cases. Carnley v. Cochran had said that presuming waiver from a silent record is impermissible, and Boykin held that the same standard decides whether a guilty plea was voluntarily made.⁴⁵ This silent-record rule reached the three rights the Court named: the privilege against compulsory self-incrimination, the right to trial by jury, and the right to confront one's accusers. The Court refused to presume the waiver of any of them from silence.⁴⁶

The refusal follows from the definition of waiver. Under Johnson v. Zerbst a waiver is an intentional relinquishment of a known right, and courts indulge every reasonable presumption against finding one.⁴⁷ Silence cannot meet that definition, because a defendant who says nothing may have understood every right at stake or none of them, and the record cannot tell the two apart. Boykin named the danger directly: ignorance, incomprehension, coercion, terror, inducements, and subtle or blatant threats can all hide behind a bare plea and leave no trace.⁴⁸ A one-word plea looks the same whether it was informed or extracted, and for that reason the evidence of understanding has to come from somewhere other than the plea itself.

Having rejected silence, the Court told trial judges what to do instead. When a defendant faces death or imprisonment, the judge owes the greatest care in canvassing the matter with him, so that he understands what the plea means and what follows from it.⁴⁹ A judge who does this accomplishes two things at once. The defendant is actually informed, which serves the intelligence half of the validity test. The judge also leaves a record adequate for any later review, which forestalls collateral proceedings that try to reconstruct a courtroom exchange from murky memories.⁵⁰ The second function explains why the burden of making the record rests on the court. The judge is the one person at arraignment who can create it cheaply and reliably, and if the cost of silence fell on the defendant, anyone never told of the rights would have to prove, years afterward, what a conversation that never happened would have contained. Justice Harlan, joined by Justice Black, dissented because the Court was in effect fastening Rule 11's prophylactic requirements on the states as a matter of due process, for a petitioner who had never alleged that his plea was involuntary or uninformed.⁵¹ The dissent shows what the holding demands, since Boykin did not have to prove that anyone pressured or misled him and the constitutional defect was the absence of any showing at all.⁵²

Miranda v. Arizona supplies a parallel from a different setting, and it shows the same logic operating outside the plea context. There the Court placed a heavy burden on the government to demonstrate that a suspect knowingly and intelligently waived the privilege against self-incrimination and the right to counsel, and it held that a valid waiver will not be presumed from the suspect's silence or from the fact that a confession was eventually obtained.⁵³ The burden sits with the government because it is responsible for the isolated setting in which the waiver supposedly occurred and has the only means of producing reliable evidence of what was said there.⁵⁴ Miranda's own confession opened with a typed paragraph reciting that it was made with full knowledge of his legal rights, yet the officers admitted at trial that he had never been advised of his right to an attorney, and the Court concluded that his statement was obtained under circumstances that did not meet constitutional standards.⁵⁵ A recital of knowledge is only words on paper, and it does not show what the defendant was told or understood. Miranda is an analogy and not a plea case, since it concerns custodial interrogation and says nothing about the content of a plea canvass. It contributes the principle that a waiver of constitutional rights must be shown by the record, and neither silence nor the bare fact that a statement or plea was obtained can stand in for that showing.

The rule concerns the record, and Boykin did not announce a fixed script for the judge to recite. The exact content of the questions comes from Rule 11, which a later section takes up. The following year Brady v. United States described the requirement that a plea be voluntary and intelligent as long recognized, and it identified the new element Boykin added as the requirement that the record affirmatively disclose a plea entered understandingly and voluntarily.⁵⁶ Justice Douglas, who wrote Boykin, later summarized the consequence in his concurrence in Santobello v. New York: after McCarthy and Boykin, trial courts must interrogate defendants who plead guilty so that the waiver of fundamental rights affirmatively appears in the record.⁵⁷ North Carolina v. Alford suggested that a state might also show a knowing waiver through counsel's later testimony that the rights were explained before the plea. That question seldom arises today, because every jurisdiction requires the judge to advise the defendant of at least the three Boykin rights.⁵⁸

The exam trap lies in facts that place a lawyer beside a defendant who pleads guilty but say nothing about any inquiry by the judge. The instinct is to assume the plea was understood because pleas usually follow advice. The silent-record rule forbids that step, because silence is not waiver, and the analysis rests on what the record shows. Appointed counsel stood beside Boykin at his arraignment, and the pleas still fell for want of any showing.⁵⁹

Voluntariness and Pressure

The silent-record rule tells a court how to show that a plea was valid, and this section turns to the first element of validity itself, which is voluntariness. Brady v. United States explains why the Fifth Amendment is involved at all. The center of a guilty plea is the defendant's admission in open court that the acts charged were committed, so the defendant stands as a witness against himself, and the minimum requirement is that the plea be the voluntary expression of the defendant's own choice.⁶¹ For that reason the requirement is grounded in due process and likened to the rule that confessions must be voluntary.⁶² The difficulty is that almost no plea is entered free of pressure, so the doctrine has to say which pressures count as compulsion and which are only the ordinary weight of a bad situation.

The doctrine tolerates pressure because the plea system runs on it. A defendant who sees little chance of acquittal gains a smaller exposure and an earlier start on correction, while the State gains a prompter punishment and saves trial resources for cases in which guilt is truly disputed. This mutuality of advantage helps explain why well over three-fourths of convictions rest on pleas.⁶³ If every plea motivated by the wish to avoid a harsher penalty were involuntary, the government would have to forbid guilty pleas, fix one invariable penalty for each crime, move sentencing to an authority ignorant of how the conviction arose, and bar pleas to reduced charges and lesser included offenses.⁶⁴ The confession cases show why a rigid definition of compulsion cannot be imported unchanged. Bram v. United States set the test for confessions: a confession must be free and voluntary, extracted by no threats or violence and obtained by no direct or implied promise, however slight, and Malloy v. Hogan carried that test forward when it made the privilege binding on the states.⁶⁵ Applied literally to pleas, a no-promises rule would condemn every bargain, because bargaining consists of promising a benefit in return for the plea.⁶⁶ The plea cases therefore keep the concern with compulsion but allow inducements.

Brady's own facts show where the line falls. He was charged under the federal kidnapping statute, and because the indictment alleged that the victim was not released unharmed, he faced a possible death sentence if the jury recommended it.⁶⁷ He first pleaded not guilty with competent counsel at his side, changed his plea after learning that a codefendant who had already confessed would plead guilty and be available to testify against him, and entered the plea after the trial judge twice questioned him about its voluntariness.⁶⁸ He later argued that the fear of death made the plea involuntary, relying on United States v. Jackson, which had barred imposition of the statute's death penalty because of its tendency to encourage guilty pleas and jury waivers.⁶⁹ The Court read Jackson narrowly: it had not held the statute inherently coercive of pleas and had created no new test, so a plea remained valid if both voluntary and intelligent.⁷⁰ Applying that test, the Court held that a plea is not invalid merely because it was entered to avoid the possibility of a death penalty.⁷¹ The Court even assumed that Brady would not have pleaded but for the death-penalty provision and still found no compulsion, because nothing showed he was so gripped by fear of death or hope of leniency that he could not, with counsel's help, rationally weigh the advantages of trial against those of pleading.⁷²

That reasoning yields the voluntary-plea standard, which asks what induced the plea. A plea entered by a defendant who is fully aware of the direct consequences, including the actual value of any commitments made by the court, the prosecutor, or the defendant's own counsel, must stand unless it was induced by threats, misrepresentation, or improper promises.⁷³ The three exceptions rest on different reasons:

• Threats include actual or threatened physical harm and mental coercion that overbears the defendant's will, and the objection is that the State's pressure has displaced the defendant's own judgment.⁷⁴
• Misrepresentation includes promises that go unfulfilled or cannot be fulfilled, and the objection is that the defendant is weighing a trial against a deal that does not exist.⁷⁵
• Improper promises are those with no proper relationship to the prosecutor's business, such as a bribe, and the objection is that the prosecutor's legitimate currency is charges and sentencing recommendations rather than favors unrelated to the case.⁷⁶

The misrepresentation category also answers an incentive problem. The prosecutor controls the charges and the recommendations and wants the plea, which creates a temptation to oversell a deal, and a rule that lets a defendant undo a plea built on a false or undeliverable promise takes away the gain from overselling. Santobello v. New York adds that when a plea rests in any significant degree on a prosecutor's promise, so that the promise is part of the inducement, the promise must be fulfilled, and the essence of the promises that induced the plea must be made known.⁷⁷ Justice Douglas's concurrence in that case collected the older decisions treating pleas as voidable: threats of physical harm, threats to use false testimony, and, in Machibroda v. United States, a threat of additional prosecutions made after the defendant said he would tell his lawyer about private promises the prosecutor had made in exchange for the plea.⁷⁸

North Carolina v. Alford confirms the standard from the bargaining side. Alford, charged with first-degree murder, pleaded guilty to second-degree murder, which capped his exposure at thirty years, and he did so to avoid a possible death sentence.⁷⁹ The court of appeals had held the plea involuntary because fear of the death penalty was its principal motivation, and the Supreme Court held that standard erroneous.⁸⁰ The question remained whether the plea represented a voluntary and intelligent choice among the alternatives open to the defendant, and the fact that Alford would not have pleaded except for the chance to limit his penalty did not show otherwise, especially because competent counsel had advised that the plea would be to his advantage.⁸¹ The two cases differ in the form the benefit took, a lower statutory ceiling for pleading in Brady and a reduced charge in Alford. They share the facts that explain the result: a strong prosecution case, competent counsel, and a defendant able to weigh trial against plea.⁸² Brady lists everyday situations that do not differ in kind, such as a plea made on advice that the judge will probably be more lenient than a jury, a plea to a lesser included offense, and a plea to some counts with the others dropped. In each, the defendant might never plead without the prospect of a lighter penalty, and none is compelled.⁸³ Alford also raises a separate question about a defendant who protests innocence while pleading, which belongs to the factual-basis requirement taken up later.⁸⁴

Brady's answer to the confession cases turns on setting. Bram involved a confession given in custody, alone and without counsel, where even a mild promise of leniency could bar the statement because a person in that position is too sensitive to inducement and the effect of the promise is too hard to assess.⁸⁵ Brady, by contrast, first pleaded not guilty, met no threats or promises in face-to-face dealings with the authorities, had competent counsel and full opportunity to compare trial with plea, and entered the plea in open court before a judge alert to the law governing pleas.⁸⁶ The difference follows from how a promise does harm. A promise endangers a choice when it prompts an impulsive response to an advantage that only seems real, and a lawyer's advice, time to reflect, and a courtroom proceeding supply the information and the interval that expose such an advantage. Counsel's presence can therefore dissipate pressure that a lone suspect in a police station has no way to counteract.

Courts have also declined to turn bargaining pressure into a freestanding ground for relief. In Bordenkircher v. Hayes the prosecutor made clear at the outset of negotiations that he would seek a recidivist indictment if the defendant did not plead, and the Court rejected the claim that a charging decision influenced by the hope of a plea violates due process.⁸⁷ In the give-and-take of plea bargaining no punishment or retaliation is present so long as the accused is free to accept or reject the offer, and defendants advised by competent counsel are presumptively capable of intelligent choice in response to prosecutorial persuasion.⁸⁸ That open offer differs from the threat alleged in Machibroda, which came after the defendant spoke of revealing the bargain to his lawyer, because Hayes knew the true terms of the offer when he decided and the choice remained his.⁸⁹ United States v. Mezzanatto makes the same point from the defendant's side: bargaining necessarily presses defendants to give up fundamental rights, the government may encourage a plea by offering substantial benefits, and the answer to possible prosecutorial abuse is case-by-case inquiry into fraud or coercion, not a ban on the practice.⁹⁰

Two further limits on what counts as pressure remain. The first concerns the source. Compulsion for this purpose means compulsion by the government, so an ultimatum from a defendant's relative, however harsh, does not make a plea involuntary.⁹¹ The Miranda-waiver cases use the same vocabulary, describing a voluntary waiver as a free and deliberate choice rather than the product of intimidation, coercion, or deception, and treating coercive police activity as a necessary predicate to involuntariness under due process; Colorado v. Connelly arose from a confession and not a plea, but it shows what voluntary means for any waiver.⁹² The second limit separates a coerced plea from a coerced confession that preceded a plea. In McMann v. Richardson the defendant confessed, pleaded guilty, and later claimed that the coerced confession had driven the plea, and the Court held that the plea stood because the decision to plead was itself free.⁹³ The judge tests voluntariness directly as well: Rule 11(b)(2) requires the court to address the defendant personally and determine that the plea did not result from force, threats, or promises other than those in a plea agreement, which gives a threatened defendant the chance to say so in open court.⁹⁴ The colloquy is taken up in detail in a later section.

Exam facts that dwell on a defendant's fear of a heavier sentence invite the conclusion that the plea was coerced. Students lose this issue when they treat a plea motivated by a lighter penalty as compelled, because the question is whether a threat, a misrepresentation, or an improper promise produced the plea, or whether a defendant with counsel could weigh real alternatives and chose the benefit.⁹⁵

Voluntariness is only the first half of the validity test. A plea can be free of improper pressure and still fail if the defendant did not understand what was being surrendered, and Edwards v. Arizona, though a Miranda case, stressed that the voluntariness of an admission and a knowing and intelligent waiver are discrete inquiries.⁹⁶ The next section takes up the second inquiry and the role counsel plays in satisfying it.

Intelligent Choice and Counsel

Voluntariness asks whether improper pressure produced the plea. The second half of the validity test asks whether the defendant understood what the plea meant. The law asks that question because a guilty plea is more than an admission of past conduct. It is also the defendant's consent to a judgment of conviction without any trial, which waives the right to trial before a jury or a judge. Waivers of constitutional rights must be knowing, intelligent acts done with sufficient awareness of the relevant circumstances and likely consequences, which is the knowing-and-intelligent standard.⁹⁸ A waiver is the intentional relinquishment of a known right, so whether a given waiver was intelligent depends on the facts of the particular case, including the accused's background, experience, and conduct.⁹⁹ A person cannot give up a right without knowing it exists, so the inquiry looks at what this defendant actually grasped rather than at what an ideal defendant would have known.

The cases state the test for the whole plea as whether it "represents a voluntary and intelligent choice among the alternative courses of action open to the defendant."¹⁰⁰ The intelligent half has practical components. The defendant must be competent, and for a plea that requires no more than competence to stand trial, meaning a rational as well as factual understanding of the proceedings and the ability to consult with a lawyer with a reasonable degree of rational understanding.¹⁰¹ Beyond competence the defendant needs information, but the amount is modest. In Iowa v. Tovar, a defendant who pleaded guilty without a lawyer received what the Constitution required when the court told him the nature of the charges, his right to be counseled about the plea, and the range of allowable punishments.¹⁰² The required awareness is therefore general: the charge, the rights given up, and the penalties the conviction can bring.

Most defendants cannot assemble even that information alone. They do not know the elements of the offense, cannot estimate their sentencing exposure, and cannot judge how likely a conviction is, so their understanding comes largely from a lawyer's advice. In Estelle v. Smith the Court recognized the point in a related setting: a layperson may not know the scope, nuances, and boundaries of a constitutional privilege, and a major decision should not be forced on a defendant without counsel's guidance.¹⁰³ If advice is the channel through which understanding reaches the defendant, defective advice contaminates the understanding. The concurring opinion in Padilla v. Kentucky states the mechanism: incompetent advice distorts the defendant's decisionmaking, and a defendant who pleads on counsel's express misrepresentation has not assumed the risk of an unknown consequence in the way one who pleads without definitive information has.¹⁰⁴

Hill v. Lockhart supplies the working standard. Hill pleaded guilty to murder and theft under an agreement in which the State recommended concurrent sentences of 35 and 10 years, and he later claimed that his court-appointed lawyer had told him he would be parole-eligible after one-third of his sentence when, because of a prior felony, he had to serve one-half.¹⁰⁵ The Court held that where a defendant is represented by counsel and pleads on counsel's advice, the voluntariness of the plea depends on whether the advice was within the range of competence demanded of attorneys in criminal cases, and that a defendant who pleads on such advice may attack the plea's voluntary and intelligent character only by showing the advice fell outside that range-of-competence standard.¹⁰⁶ The standard measures the competence of the advice and leaves its accuracy aside. A plea trades a trial for a bargain, and a defendant who relies on the good-faith judgment of a reasonably competent lawyer accepts the inherent risk that the judgment about the facts, or about what a court will do, may later prove mistaken.¹⁰⁷ If every forecast that failed could undo a plea, no bargain would ever be final.

Hill then held that the two-part test of Strickland v. Washington governs ineffective-assistance claims that arise from the plea process.¹⁰⁸ That test requires the defendant to show deficient performance, meaning representation that fell below an objective standard of reasonableness under prevailing professional norms, and then to show that the deficiency prejudiced the defense.¹⁰⁹ The deficiency prong asks whether the lawyer acted as a reasonably competent criminal attorney would. The prejudice prong needed adapting, because a plea involves no trial whose result could have changed. In the plea setting the defendant must show a reasonable probability that, but for counsel's errors, he would not have pleaded guilty and would have insisted on going to trial.¹¹⁰

The reason for the prejudice requirement is that an error by counsel, even a professionally unreasonable one, does not warrant undoing a judgment if the error had no effect on it.¹¹¹ In a plea case the conviction rests on the decision to plead, so the question becomes whether the error bore on that decision, and a defendant who would have pleaded anyway was not harmed by the error. That is why Hill lost: the Court affirmed because he had not alleged the kind of prejudice that would have entitled him to a hearing, namely that with adequate advice he would have pleaded not guilty and gone to trial.¹¹²

The same framework reaches the bargaining stage. In Missouri v. Frye the Court treated plea negotiation as a critical stage at which counsel must perform effectively, and it held that a lawyer who never communicated formal plea offers to the client, so that the offers lapsed, performed deficiently.¹¹³ The prejudice showing differs from Hill because the complaint is a lost opportunity rather than a bad plea. The defendant must show a reasonable probability that he would have accepted the earlier offer, and also that neither the prosecution nor the court would have prevented the offer from being implemented, since a defendant has no right to be offered a plea or to have the judge accept one.¹¹⁴ A Hill claim asks whether the defendant would have gone to trial, and a Frye claim asks whether the defendant would have taken the better deal.

Counsel's advice is only one source of a defendant's understanding, and the next question is what the State and the court must supply. Hill answers that the Court had never held the Constitution to require the State to give a defendant information about parole eligibility for a plea to be voluntary, and that such a requirement would conflict with the federal rules then governing pleas.¹¹⁵ The two poles of Hill therefore differ in one fact. Hill's claim could succeed only through counsel's misadvice about parole, while a defendant who complained solely that neither the State nor the court told him his parole date, with no claim that his lawyer misinformed him, would have no constitutional objection on that ground. The reason a precise forecast cannot be required is practical: the exact sentence is not set until weeks after the plea, usually after a presentence report, so the judge can describe only the possible penalties and the defendant must decide without knowing the most important variable.¹¹⁶

Courts give the required awareness a boundary by separating direct from collateral consequences. A direct consequence is a sanction that follows directly, immediately, and largely automatically from the conviction, while a collateral consequence is one that is not certain to follow, does not affect the length or nature of the sentence, or lies outside the court's control, and a defendant must be aware of the former and not the latter.¹¹⁷ Brady's own formulation speaks of a defendant fully aware of the direct consequences of the plea.¹¹⁸

Immigration consequences have been pulled out of the collateral category as far as counsel's duty is concerned. Padilla held that effective counsel must advise a client about the deportation consequences of a conviction, and that the client must still establish prejudice from the misadvice.¹¹⁹ The likelihood of removal can easily decide whether a defendant pleads or goes to trial, and Rule 11(b)(1)(O), added in 2013, now requires the judge to warn a defendant who is not a citizen before the plea that a conviction could lead to removal, denial of citizenship, and denial of entry into the United States in the future.¹²⁰ The dissent in Padilla would have kept a claim of that kind under due process rather than the Sixth Amendment, which shows that an unintelligent-plea argument can be framed either way.¹²¹

Even for direct matters the defendant need not know everything. In United States v. Ruiz the Court held that the Constitution does not require prosecutors to disclose impeachment information before a plea agreement.¹²² The Court gave three reasons. First, impeachment information matters to the fairness of a trial, and the law ordinarily treats a waiver as knowing if the defendant fully understands the nature of the right and how it would likely apply in general, even without the specific detailed consequences.¹²³ Second, earlier cases accepted pleas despite a defendant's misapprehension of the quality of the State's case, of the likely penalties, or of the admissibility of a confession.¹²⁴ Third, the value of such information depends on what the defendant already knows of the case, and other safeguards, including Rule 11, reduce the risk that innocent people plead guilty.¹²⁵ The Court has likewise refused to hold that a defendant's ignorance that a prior coerced confession could not be used against him compromised the voluntariness of his later plea, as it held in McMann v. Richardson.¹²⁶

Exam facts often stress what the defendant did not know, and the tempting conclusion is that the plea must fail. Students lose this issue when they treat a gap in what the defendant knew as an unintelligent plea, because the question is whether the defendant understood the charge, the rights surrendered, and the direct consequences, and whether counsel's advice stayed within the range of competence.¹²⁷

Because a court cannot assume that counsel's advice was adequate, Rule 11(b)(1) requires the judge to address the defendant personally in open court and to determine that the defendant understands the rights waived, the nature of the charge, and the penalties it carries, among other matters.¹²⁸ That colloquy is a second check on understanding that operates whatever the lawyer said, and the next section takes it up in detail.