Indiana Law Review Quieting the Guilty and Acquitting the Innocent: A Close Look at a New Twist on the Right to Silence Gordon Van Kessel* Table of Contents Introduction 926 A. The Ascending Right to Silence 926 B. Conventional Foundations ofthe Right to Silence 929 C. A New Twist on the Right to Silence 930 D. Significance ofthe "Anti-Pooling" Theory 932 E. Analyzing Games, Markets, and "Anti-Pooling" Theories 935 I. Assumptions and Implications of the "Anti-Pooling" Theory 939 A. Assumptions ofthe "Anti-Pooling" Theory and Use of Market Models 939 B. The Adverse Inference Problem: The "Anti-Pooling" Model as Both Resting on and Undermining the Right to Silence 941 C. "Anti-Pooling" by Using the Stick Rather than the Carrot 943 D. The Consequences ofSticks and Carrots 948 II. Validity of Market Analogies, Game-Theoretic Assumptions, and the "Anti-Pooling" Theory 953 A. Validity ofMarket Analogies 953 1 . The One-Shot Trial and Juror Awareness 953 2. Tasting the Fruit and Testing the Cars 953 3. The Limits of Real World "Anti-Pooling" 954 B. Assessing Benefits to the Innocentfrom "Anti-Pooling" 954 1 . Identifying Innocent Suspects Who Would Be Harmed by Fewer Guilty Suspects Remaining Silent 955 2. Assessing the Reaction of the Guilty to Elimination of the Right to Silence 956 3. The Market Analogy and Silent Lemon-Sellers 956 4. The Market Analogy and Truthful or Unbelievable Lemon-Sellers: Confessions and Refutable Fabrications 958 G Distinguishing Between Pretrial and Trial Silence 960 D. Empirical Evidence Testing the "Anti-Pooling" Theory 962 E. The Limited Value ofComparative Studies 965 F. Assessing the Costs ofthe Right to Silence 969 1 . Helping the Guilty to Help Themselves 970 2. Helping the Innocent by Choosing to Speak 971 III. Implications of the "Anti-Pooling" Theory Regarding Reform of the Right to Silence 972 * Professor of Law, University of California, Hastings College of the Law. The author wishes to thank Hastings Professors Vikram Amar, Rory Little, and other colleagues who offered helpful comments and suggestions. 926 INDIANA LAW REVIEW [Vol. 35:925 A. The Many Faces ofthe Right to Silence 972 B. Enhancing the Right to Silence 974 1. The Consequences of an Expanded Right to Silence 977 2. Enhancing the Right to Counsel 977 3. Limiting Deception and Requiring Disclosure of Police Evidence 979 C. Restricting the Right to Silence 982 D. Recognizing the Harm Caused by Uncontradictable Lies and the Importance of Unrehearsed Statements 982 E. Shaping the Right to Silence Through "Anti-Pooling" Measures Which Increase Accurate Factfinding 985 Conclusion 986 Introduction A. The Ascending Right to Silence The right to silence is on the upswing on both sides of the Atlantic. Throughout Europe, there is near-universal recognition of a right to silence and a privilege against self-incrimination that applies to both the pretrial and trial stages of a criminal case. 1 Those aspects of the right to silence that require advice ofthe right and prohibit adverse inferences from silence also are generally accepted. Most civil law countries ofcontinental Europe have adopted rules that require suspects be informed ofthe right to remain silent prior to questioning as well as rules that prohibit courts from considering defendant's silence as evidence ofguilt, 2 although in practice such guarantees often are not as strong as 1 . The European Court ofHuman Rights has repeatedly stated that, although the European Convention on Human Rights contains no explicit guarantee ofa right to silence, "there [could] be no doubt that the right to remain silent under police questioning and the privilege against self- incrimination are generally recognized international standards which lie at the heart of the notion of a fair procedure under Article 6 [which guarantees the right to a fair and public hearing]." Murray v. United Kingdom, 22 Eur. Ct. H.R. 29, 1 45 (1996); see also Saunders v. United Kingdom, 23 Eur. Ct. H.R. 313 (1997); Funke v. France, 16 Eur. Ct. H.R. 297 (1993). The court's language in Funke, connecting the right to silence with police interrogation, and its later reliance on the privilege in Saunders, when dealing with official compulsion under oath, suggests the court perceives different roles for the silence right and the privilege. See European Convention for the Protection of Human Rights and Fundamental Freedoms, Nov. 4, 1950, 213 U.N.T.S. 221. There is, however, an explicit privilege against self-incrimination in the United Nations International Covenant on Civil and Political Rights (ICCPR). Article 14(3) states that "[i]n the determination of any criminal charge against him, everyone shall be entitled to the following minimum guarantees . . . (g) Not to be compelled to testify against himself or to confess guilt." International Covenant on Civil and Political Rights, Dec. 19, 1966, Article 14, 999 U.N.T.S. 171, 177. 2 . See CraigM . Bradley, The Emerging International Consensus as to Criminal Procedure 2002] THE RIGHT TO SILENCE 927 in America 3 due to differences in legal and social cultures and between adversary and inquiry procedures. 4 Only in England, Israel, and a few other countries are factfinders legally permitted to draw inferences of guilt from silence during police questioning and at trial. 5 In America, the right to silence is also on firm ground. Miranda rules, once thought to be in jeopardy, have been extended by the U.S. Supreme Court in some respects 6 and recently were reaffirmed and strengthened in Dickerson v. United States. 7 Furthermore, the Supreme Court recently renewed its Rule, 1 4 MICH. J. Int'L L. 171,21 9-20 ( 1 993); Stephen C. Thaman, Miranda in Comparative Law, 45 St. LOUIS U. L.J. 581 (2001); Gordon Van Kessel, European Perspectives on the Accused as a Source ofTestimonial Evidence, 100 W. Va. L. Rev. 799, 809, 821-23, 832 (1998). 3. See Miranda v. Arizona, 384 U.S. 436 (1966). 4. Unlike Miranda rules, for example, in most European countries a defendant' s assertion of the right to silence generally does not operate to shut down interrogation and the police may continue to ask questions. See Van Kessel, supra note 2, at 810, 821-23, 832. Furthermore, in criminal trials in continental Europe it is a rare event for the defendant not to speak and respond to questions. Id. at 833. In some countries, particularly in France, the right to silence has more theoretical than practical significance. A French lawyer recently told me that I am wasting my time on the right to silence which is regarded in France as a foreign, English-style concept. For a fascinating example of this point in the context ofa French murder trial, see Ren6e Lettow Lerner, The Intersection ofTwo Systems: An American on Trialfor an American Murder in the French Cow d 'Assises, 2001 U. III. L. Rev. 791, 812 (pointing out that the spotlight of the French trial is "squarely on the defendant" and describing how the presidingjudge closely examines the accused to serve the central purpose of the French trial—finding out what happened and why). 5. See Van Kessel, supra note 2, at 821-23, 832. The European Court of Human Rights has found that with certain protections such use of silence does not violate the right to a fair trial under the Convention. See also Condron v. United Kingdom, 3 1 Eur. Ct. H.R. 1 (2000) (holding that permitting the factfinder to consider defendant's silence is not of itself incompatible with the right to a fair hearing provided silence is not the sole or main basis for the conviction); Murray v. United Kingdom, 22 Eur. Ct. H.R. 29, ^| 47 (1996) (holding that adverse inferences may be drawn from silence "in situations which clearly call for an explanation" from the defendant if the assistance of a lawyer is provided when defendant must decide whether to speak). 6. The Supreme Court has held that when a defendant asserts his right to counsel, he may not be subjected to further police questioning until counsel has been made available to him unless defendant independently initiates further conversations with the police. Edwards v. Arizona, 45 1 U.S. 477, 484 (1981). The Edwards prohibition on future questioning was later extended to offenses wholly unrelated to the crime as to which the suspect has requested counsel. Arizona v. Roberson, 486 U.S. 675 (1988). Professor Yale Kamisar described Edwards as "in effect [establishing] a new 'prophylactic rule' that built on and reinforced Miranda's 'prophylactic rules,'" and regarded Roberson as reaffirming and reinvigorating Edwards. Yale Kamisar, Confessions, Search and Seizure and the Rehnquist Court, 34 TULSA L.J. 465, 474, 499 (1999). 7. See Dickerson v. United States, 530 U.S. 428, 432 (2000) (holding that because Miranda is a "constitutional decision" it may not be overruled by Congress and that "Miranda and its progeny in this Court govern the admissibility ofstatements made during custodial interrogation in both state and federal courts"). Affirming Miranda 's constitutional foundations substantially 928 INDIANA LAW REVIEW [Vol. 35:925 commitment to the rule against adverse inferences from silence at the guilt phase of the criminal trial that was established by the Warren Court in Griffin v. California* and even applied the rule to the sentencing stage. 9 Only Justice Thomas urged the Court to reexamine Griffin} With the Supreme Court reaffirming Miranda and extending Griffin, at least for the present, the basic right to silence, with its warning requirements and its rule against adverse inferences, is secure throughout the criminal process from custodial interrogation through sentencing. Yet, however safe may be its core principles, the right to silence constantly is being attacked and defended, and many of its individual aspects are highly controversial. Rationales supporting the right to silence therefore remain critically important when courts and legislatures decide whether to expand or contract the right's particular guarantees that go beyond the simple right to silence warnings and the rule against adverse inferences. strengthened its practical effect by increasing the prospect of civil penalties against those who disregard its mandates. Prior to Dickerson, interrogating officers often would continue questioning despite a suspect's invocation ofMiranda rights. See Charles D. Weisselberg, Saving Miranda, 84 Cornell L. Rev. 109, 112 (1998) (presenting evidence "that police officers in some jurisdictions are systematically trained to violate Miranda"). California courts had condemned the practice of questioning "outside Miranda" but had permitted it in practice. See People v. Peevy, 953 P.2d 1212, 122S (Cal. 1998) (finding admissible for impeachment a statement obtained in deliberate violation ofMiranda, while noting that "it is indeed police misconduct to interrogate a suspect in custody who has invoked the right to counsel"); People v. Bradford, 929 P.2d 544 (Cal. 1997) (strongly disapproving continued questioning following defendant's request for counsel, but affirming his conviction). However, with the prospect of civil rights suits for violation of Miranda's standards, California law enforcement agencies have altered their practices and police no longer engage in questioning a suspect once he states that he wishes to remain silent or to consult with counsel. See Cal. Attorneys for Criminal Justice v.Butts, 195F.3d 1039 (9th Cir. 1999) (establishing a clear rule that continued questioning after defendant's invocation of the right to counsel constitutes a violation of the Fifth Amendment and bars any claim of qualified immunity); Cooper v. Dupnik, 963 F.2d 1220 (9th Cir. 1992) (rejecting a claim ofqualified immunity in a § 1983 action brought on behalf of a suspect who was interrogated after he had requested counsel, but who was never tried, finding police conduct to be coercive and a violation ofthe defendant's constitutional rights). 8. 380 U.S. 609 (1965). 9. See Mitchell v. United States, 526 U.S. 3 1 4 ( 1 999) (holding that in federal court a guilty plea does not waive the privilege at the sentencing phase and reaffirming and extending the rule of Griffin such that the sentencingjudge may not draw an adverse inference from defendant's silence in determining the facts about the crime which bear upon the severity ofthe sentence). While the dissenters in Mitchell disagreed with the majority's description ofthe no-adverse-inference rule as "an essential feature of our legal tradition," they acknowledged that it "may be true" that the rule has found "wide acceptance in our legal culture" which they found an "adequate reason not to overrule" it. Id. at 331-32 (Scalia, J., dissenting). 10. Id. at 341 (Thomas, J., dissenting). 2002] THE RIGHT TO SILENCE 929 B. Conventional Foundations ofthe Right to Silence Defenders of the right to silence generally rely on conventional rationales articulated by the U.S. Supreme Court that involve a complex set of values such as upholding fairness, personal dignity, free will, and avoiding torture, inhumane treatment, and the cruel trilemma of self-accusation, perjury or contempt. 11 Of late, the Court has emphasized the deterrence of government coercion 12 and the maintenance ofour adversary system ofjustice which prohibits the government from making a defendant the unwilling "instrument of his or her own condemnation." 13 The traditional view recognizes that the right to silence may help the guilty avoid conviction but concludes that it is the price which must be paid for the right's many benefits. 14 Suggestions that the right to silence also helps the innocent are more controversial. Scholar Jeremy Bentham advocated that only the guilty exercise the right while benefit from it 15 and others contend that on occasion even the innocent may be helped by the opportunity to seek refuge in silence. 16 The Supreme Court has remained somewhat ambivalent as 1 1 . See Miller v. Fenton, 474 U.S. 104, 116 (1985) (referring to the voluntariness inquiry as having a "hybrid quality" and a "complex of values"); Murphy v. Waterfront Com'n of N.Y. Harbor, 378 U.S. 52, 55 (1964) (referring to the privilege against self-incrimination as reflecting "our respect for the inviolability of the human personality" and "our fear that self-incriminating statements will be elicited by inhumane treatment and abuses," and "our sense of fair play which dictates 'a fair state-individual balance by requiring the government to leave the individual alone until good cause is shown for disturbing him'"). The concern for privacy interests was mentioned in Murphy but has since been downplayed by the Supreme Court. See United States v. Doe, 465 U.S. 605 (1984) (explaining that the Fifth Amendment does not create a zone of privacy that protects an individual from the compelled production by the government of personal records); Fisher v. United States, 425 U.S. 391 (1976) (rejecting the contention that the Fifth Amendment somehow independently protects privacy). 12. See Colorado v. Connelly, 479 U.S. 157, 167 (1986) (describing the purpose of the voluntariness rule as deterring future constitutional violations and preventing fundamental unfairness in the use ofevidence, rather than excluding "presumptively false evidence"). The Court also held that a waiver of Miranda rights cannot be involuntary absent official compulsion or coercion and stated that the "sole concern" of the Fifth Amendment privilege is government coercion. Id. at 170. 13. Mitchell, 526 U.S. at 325 (citing Rogers v. Richmond, 365 U.S. 534, 541 (1961)). 1 4. See Ullmann v. United States, 350 U.S. 422, 428 ( 1 956) (noting that while the "privilege may, on occasion, save a guilty man from his just deserts, . . . [i]t was aimed at a more far-reaching evil—a recurrence of the Inquisition and the Star Chamber, even if not in their stark brutality"). 1 5. See Jeremy Bentham, A Treatise on Judicial Evidence 24 1 (M. Dumont ed., Fred B. Rothman & Co. 1 98 1 )( 1 825). 16. Scholars recently have sought to justify the privilege on the ground that it protects the innocent by offering them a refuge from speaking in a way that might lead to unreliable verdicts. See Peter Arenella, Foreword: O.J. Lessons, 69 S. Cal. L. Rev. 1233, 1250 (1996) (arguing that the privilege protects three types offactually innocent defendants—those who fear taking the stand because they will be impeached by their prior convictions, those whose nervousness, appearance, 930 INDIANA LAW REVIEW [Vol. 35:925 to whether the privilege helps the innocent avoid conviction or otherwise leads to more reliable verdicts. 17 C. A New Twist on the Right to Silence Recently, other voices have offered a new twist on the right to silence which proposes an unconventional way in which the right benefits the innocent. Professors Daniel Seidmann and Alex Stein have proposed a behavioral or "game-theoretic model" 18 as a foundation for an innocent-benefit theory that they contend has been largely ignored or underestimated by academics but which offers a better justification for the right than conventional rationales. 19 According to Seidmann and Stein, the right to silence is justified primarily on the ground that it benefits the innocent, not because they may use it themselves, but because of its use by the guilty. Through encouraging the guilty to remain silent, the right assists factfinders in identifying those who are unjustly suspected or accused of criminal conduct. By remaining silent, the guilty separate themselves from the innocent, rather than lie and "pool" their false or lack ofmental agility might enable a prosecutor to make them appear guilty through artful cross- examination, and those whose truthful direct testimony would incriminate, despite their factual innocence); Stephen J. Schulhofer, Some Kind Wordsfor the Privilege Against Self-incrimination, 26 Val. U. L. Rev. 311, 329-31 (1991) (noting that in light of the "realities of trial practice and risks to the innocent that all lawyers understand," trial lawyers can think ofnumerous reasons why they would advise an innocent client not to take the stand). See also Craig M. Bradley, Griffin v. California: Still Viable After All These Years, 79 MiCH L. Rev. 1290, 1293-94 (1981) (making similar arguments). 17. Compare Murphy v. Waterfront Comm'n of N.Y. Harbor, 378 U.S. 52, 55 (1964) (stating that the privilege reflects "our distrust of self-deprecatory statements and our realization that the privilege, while sometimes 'a shelter to the guilty,' is often 'a protection to the innocent.'"), with Connelly, 479 U.S. at 170 (stating that the sole concern of both the Due Process Clause and the Fifth Amendment is to deter government coercion, rather than to assure that statements of suspects are either reliable or the product of the suspect's free will) and Baxter v. Palmigiano, 425 U.S. 308, 3 1 9 (1 976) (stating that the privilege "has little to do with a fair trial and derogates rather than improves the chances for accurate decisions"). In Withrow v. Williams, 507 U.S. 680, 682-83 (1993), the Court shifted back to trustworthiness as a basis for the privilege when it refused to extend the restrictions of Stone v. Powell, 428 U.S. 465 (1976), on federal habeas corpus review of state convictions regarding Fourth Amendment violations to Miranda violations, partly on the ground that Miranda is related to the correct ascertainment ofguilt and braces against the admission of unreliable statements. 18. Daniel J. Seidmann & Alex Stein, The Right to Silence Helps the Innocent: A Game- Theoretic Analysis ofthe Fifth Amendment Privilege, 1 14 HARV. L. Rev. 430, 438 (2000). 1 9. In their support for the right to silence, Seidmann and Stein seek to drive another nail in Bentham's coffin and to bury even deeper the suggestion ofeliminating the rule against adverse inferences from silence. Id. at 433. They view the "conventional wisdom" that the right to silence helps only the guilty as a "facially compelling" but ultimately a flawed argument often voiced by "law and order" conservatives. Id. at 435, 451-55. 2002] THE RIGHT TO SILENCE 93 1 stories with true accounts offered by innocent suspects during pretrial interrogation and innocent defendants at trial. 20 Inducing this "anti-pooling effect" enhances the credibility of innocent suspects 21 and increases the likelihood of their acquittal. 22 In this way, the "good" that guilty suspects consume by remaining silent does not remain private, but is shared by innocent suspects who are not subjected to the "negative externalities" flowing from perjured accounts by the guilty. 23 To accomplish this goal, silence must be seen by the guilty as an attractive alternative to fabrication. 24 Thus, calls to abandon the right to silence and to permit adverse inferences from its exercise should be rejected. The innocent as well as the guilty have an interest in maintaining the right as a refuge during pretrial questioning and as a viable alternative to perjury at trial. 25 Seidmann and Stein believe that their "anti-pooling" rationale offers a better explanation for the present ramifications ofthe right to si lence than conventional justifications which have been accepted and relied on by decisions of the Supreme Court. 26 They contend that the "anti-pooling" rationale forms the primary basis for retaining the right to silence principles that prohibit use of silence as evidence ofguilt through disallowing adverse inferences from exercise of the right both during custodial interrogation and at trial. 27 This rationale explains why the right is limited to testimonial evidence, to the single sovereign context, to criminal cases and to the custodial interrogation and trial contexts. 28 Indeed, their "anti-pooling" rationale suggests that the right to silence in America might be expanded and made even more attractive to guilty suspects. On the other hand, Seidmann and Stein appear to believe that when the "anti-pooling" rationale does not apply, there is no valid reason to recognize right to silence protections against adverse inferences from its exercise. 29 20. Id. at 433, 459-60. 21. Mat 433. 22. Id. at 451. 23. Mat 457-58. 24. Id. at 433, 438. 25. Id. at 453-54 n.79. 26. Id. at 474-75, 489. Indeed, Seidmann and Stein agree with many of Bentham's criticisms of the accepted justifications for the privilege such as "individualistic notion[s] of fairness" and avoidance ofthe cruel "trilemma" ofself-accusation, contempt, or perjury. Id. at 452- 53. 27. For example, Seidmann and Stein criticize the fairness and reliance foundations ofDoyle v. Ohio, 426 U.S. 610 (1976), but they accept its rule that silence following Miranda warnings should not be the subject of adverse comment or inferences. Seidmann & Stein, supra note 18, at 453-54 nn.79, 491. 28. Seidmann & Stein, supra note 18, at 474-75. 29. Seidmann and Stein believe that if the guilty "cannot fabricate evidence in a way that harms the innocent, then they should not be exempted from potential self-incrimination," and that "[o]nly the existence ofameaningful fabrication alternative should therefore activate the privilege." Id. at 480. Nor do they believe that the innocent are in need of the right to silence protections 932 INDIANA LAW REVIEW [Vol. 35:925 D. Significance ofthe "Anti-Pooling" Theory Seidmann and Stein's "anti-pooling" theory represents a unique and ambitious effort to justify the right to silence on a ground that even the most ardent conservatives accept as paramount—acquittal of the innocent through accurate factfinding. But the new theory has profound implications for the right to silence. If valid, it suggests that the right should not only be maintained, but expanded to encourage even more guilty defendants to claim it.30 A favorable attitude of courts and legislatures toward the right to silence may well lead to even broader protections in the context of police interrogation. For example, relying on the notion that Miranda established a prophylactic rule rather than a constitutional right, the Supreme Court has declined to apply the fruit of the poisonous tree doctrine to Miranda violations involving failure to give the required warnings 31 and has permitted the use ofstatements not permitted in the prosecutor's case because ofMiranda defects to impeach a testifying defendant.32 However, the Court's affirmation ofMiranda as constitutionally based has cast doubt on the continued validity ofthese rules. 33 Furthermore, states may expand their own versions ofthe right to silence. Minnesota, for example, requires that confessions be recorded and imposes individual criminal and civil liability on law enforcement officers for violation of the right to consult with counsel by failing to honor a request to speak with a lawyer by any person in their custody. 34 Finally, the number of erroneous convictions being brought to light by newly- found DNA evidence has resulted in calls to reform interrogation practices that which prohibit adverse inferences from failing to testify on the ground that many innocent defendants may remain silent for fear of prior conviction impeachment. Id. at 494. 30. Seidmann and Stein seem to favor strengthening all rules which induce an "anti-pooling" effect through making silence an "attractive alternative" to fabrications. Id. at 433. Currently, only a minority of suspects assert their Miranda rights during custodial interrogation. See infra note 92 and accompanying text. 3 1 . Oregon v. Elstad, 470 U.S. 298, 306-07 ( 1 985) (stating that since Miranda sweeps more broadly than the Fifth Amendment privilege against self-incrimination and may be triggered even in the absence of a violation of the privilege, its "preventive medicine" provides a remedy even to one who has suffered no constitutional harm). 32. Oregon v. Hass, 420 U.S. 714 (1975) (permitting impeachment of a defendant where warnings were given, but interrogation continued after defendant asked for counsel); Harris v. New York, 401 U.S. 222, 226 (197 1) (holding that statements inadmissible in the prosecution's case-in- chiefbecause obtained in violation of Miranda may, ifnot coerced or involuntary, be used to attack the credibility of the defendant if he takes the stand). 33. See infra notes 90, 217. 34. See Peter Erlinder, Getting Serious About Miranda in Minnesota: Criminal and Civil Sanctionsfor Failure to Respond to Requests for Counsel, 27 WM. MITCHELL L. REV. 941, 970 (2000) (noting that this responsibility can be "vindicated" by either a private consultation in the place ofconfinement or by telephone access to counsel in a reasonably confidential setting, but that when a person in custody requests access to counsel, the law requires consultation with counsel to be provided before questioning can continue). 2002] THE RIGHT TO SILENCE 933 are claimed to lead to false confessions. 35 Seidmann and Stein's theory that the right to silence helps the innocent is likely to bolster efforts to strengthen and expand the right in context of custodial interrogation. Conversely, in attacking the conventional foundations ofthe right to silence and urging the acceptance of a heretofore largely unrecognized rationale as its primary basis, Seidmann and Stein are placing the right to silence in a precarious position. If the newly proposed foundation is shown to be infirm, the authors have undermined the traditional and currently accepted rationales for the right without offering any solid alternative support. 36 The right to silence is constantly being challenged, particularly aspects ofthe 35. See Richard A. Leo & Richard J. Ofshe, Missing the Forestfor the Trees: A Response to Paul Cassell 's "Balanced Approach " to the False Confession Problem, 74 DENV. U. L. Rev. 1135, 1137-39 (1997) (arguing that "there is compelling and abundant evidence that false confessions occur regularly" and that those that are noticed are only the tip ofthe false confession iceberg); Richard J. Ofshe & Richard A. Leo, The Decision to Confess Falsely: Rational Choice and Irrational Action, 74 Denv. U. L. Rev. 979, 983 (1997) [hereinafter Ofshe & Leo, Decision to Confess] (contending that while the third degree has "virtually disappeared," police-induced false confessions still occur regularly and are a serious problem for the American criminal justice system). Ofshe and Leo blame deceptive interrogation techniques, such as leading the suspect to believe that the evidence against him is overwhelming and his fate is certain and that there are advantages in confessing. Id. at 985-86. But Professor Paul Cassell has vigorously disputed the notion that false confessions occur frequently and has criticized Leo and Ofshe for failing to consider the costs of lost convictions that might follow from restrictions on police questioning. See Paul G. Cassell, Balanced Approaches to the False Confession Problem: A Brief Comment on Ofshe, Leo and Alschuler, 74 DENV. U. L. Rev. 1123 (1997) [hereinafter Cassell, Balanced Approaches). The debate has continued focusing on a study of what Ofshe and Leo describe as sixty cases of "police-induced false confessions in the post-Miranda era." Richard A. Leo & Richard J. Ofshe, The Consequences of False Confessions: Deprivations of Liberty and Miscarriages ofJustice in the Age ofPsychological Interrogation, 88 J. CRIM. L. & CRIMINOLOGY 429, 433 ( 1 998) [hereinafter Leo & Ofshe, Consequences ofFalse Confessions]. See also Paul G. Cassell, Protecting the Innocentfrom False Confessions andLost Confessions—andfrom Miranda, 88 J. CRIM. L. & CRIMINOLOGY 497 ( 1 998) [hereinafter Cassell, Protecting the Innocent]. See also Paul G. Cassell, The Guilty and the "Innocent": An Examination ofAlleged Cases of Wrongful Convictionfrom False Confessions, 22 Harv. J.L. & PUB. Pol'y 523 (1999) [hereinafter Cassell, Wrongful Conviction]', Richard A. Leo & Richard J. Ofshe, Using the Innocent to Scapegoat Miranda. Another Reply to Paul Cassell, 88 J. Crim. L. & CRIMINOLOGY 557 (1998) [hereinafter Leo & Ofshe, Scapegoat). Occasionally, there are calls to do away with interrogation entirely. See Thaman, supra note 2, at 620-24 (calling for eliminating interrogation of suspects as a means of investigation). 36. Seidmann and Stein deal only with the evidentiary aspect of the right to silence which prohibits adverse inferences; they accept the validity of that aspect of the right to silence which exempts a person from contempt for refusal to incriminate oneself, noting that even the most ardent critics ofthe right to silence do not call for removal ofthe contempt exemption. Seidmann & Stein, supra note 18, at 440 n.36. 934 INDIANA LAW REVIEW [Vol. 35:925 right during custodial interrogation. 37 Even the rule against adverse inferences has been questioned by judges and scholars who have proposed forms ofpretrial judicial examination of the accused conducted by magistrates at which defendants would be afforded counsel, but warned that silence could lead to adverse inferences at trial. 38 Recently, Professor Alschuler looked to Scottish procedure and suggested ajudicially supervised, deposition-style examination at which the accused would remain unsworn but subject to adverse inferences for silence. 39 Professor Akhil Amar would even require the accused to testify under 37. For example, a request for counsel during custodial interrogation currently has a more powerful bite than a refusal to speak or answer questions. There is no per se rule against later questioning by the police following an indication ofa desire to remain silent by a suspect provided officials initially cease questioning. Michigan v. Mosley, 423 U.S. 96 (1975). However, once a suspect requests a lawyer, there can be no further questioning until counsel has been made available to him unless he first initiates further conversations with the police. Edwards v. Arizona, 45 1 U.S. 477 (1981). This stronger medicine applies even to questioning concerning offenses wholly unrelated to the crime as to which the suspect had requested counsel. Arizona v. Roberson, 486 U.S. 675 (1988). While protections against further questioning after a request for counsel have been incorporated into the Sixth Amendment, they are "offense specific" in the Sixth Amendment context such that a request for counsel does not prohibit continued questioning concerning uncharged crimes. McNeil v. Wisconsin, 501 U.S. 171, 175 (1991). Justice Kennedy, however, has urged the Court to bring Fifth Amendment rules into line with Sixth Amendment standards such that prohibitions on further questioning under Miranda are also offense-specific in the sense ofnot applying to crimes unrelated to those as to which the suspect had requested counsel. Id. at 1 83 (Kennedy, J., concurring). Justice Kennedy also urges the overruling ofArizona v. Roberson. See id. Furthermore, Justice Kennedy, joined by Justices Scalia and Thomas, views as "questionable" importation of the broad rule prohibiting further questioning after a request for counsel into the Sixth Amendment context where it is triggered by a request for counsel at arraignment or other judicial proceeding although defendant has agreed to be questioned without a lawyer. They find "difficult to understand" a rule that operates "to invalidate a confession given by the free choice of suspects who have received proper advice of their Miranda rights but waived them nonetheless." Texas v. Cobb, 532 U.S. 162, 174-75 (2001) (Kennedy, J., concurring). 38. Paul G. Kauper, Judicial Examination ofthe Accused—A Remedyfor the Third Degree, 30 MICH. L. Rev. 1224 (1932); Roscoe Pound, Legal Interrogation of Persons Accused or Suspected ofCrime, 24 J. CRIM. L. & CRIMINOLOGY 1014(1 934); John H. Wigmore, Nemo Tenetur Seipsum Proedre, 5 Harv. L. Rev. 71, 85-88 (1892). See also MARVIN E. FRANKEL, Partisan Justice 98-99 ( 1 980); Walter V. Schaefer, The Suspectand Society 7 1 , 77-8 1 ( 1 967); Lloyd Weinreb, DENIAL OF Justice 1 63-64 ( 1 977); Marvin Frankel, The Searchfor Truth: An Umpireal View, 123 U. Pa. L. Rev. 1031, 1053 (1975); Henry J. Friendly, The Fifth Amendment Tomorrow: The Casefor Constitutional Change, 37 U. ClN. L. REV. 67 1 , 685, 700-0 1,713-16(1 968); R. Kent Greenawalt, Silence as a Moral and Constitutional Right, 23 Wm. & MARY L. REV. 15, 51-65 ( 1 98 1 ); Yale Kamisar, Kauper 's "Judicial Examination ofthe Accused" Forty Years Later —Some Comments on a Remarkable Article, 73 MICH. L. REV. 15 (1974). 39. Albert W. Alschuler, A Peculiar Privilege in Historical Perspective: The Right to Remain Silent, 94 MICH. L. REV. 2625, 2670-71 (1996). See also WILLIAM T. PlZZl, TRIALS Without Truth: Why Our System of Criminal Trials Has Become an Expensive Failure 2002] THE RIGHT TO SILENCE 93 5 oath by threat of contempt but with limited testimonial immunity which would not extend to fruits of such compelled testimony. 40 Consequently, however secure may be the core protections of the right to silence, the rationales supporting the right remain critically important to the future development of its many aspects. The "anti-pooling" theory therefore deserves a close inspection. E. Analyzing Games, Markets, and "Anti-Pooling Theories First, this Article outlines the assumptions on which Seidmann and Stein base their "Anti-Pooling" theory and describes how they use game and market models to apply and test the assumptions and draw conclusions regarding the effect ofthe right to silence. A troubling aspect ofthe "anti-pooling" theory that the authors do not discuss will also be addressed—specifically that the theory's reliance on the rule against adverse inferences from silence, when carried to its logical conclusion, contains the seeds of its own destruction. Since the theory rests on the proposition that the no-adverse-inference rule leads to more guilty people remaining silent and factfinders believing more innocent suspects who speak, ultimately, the theory will lead to factfinders becoming more skeptical of those who refuse to speak and this will tend to undermine the very right to silence principles on which the theory rests. Second, assuming that "anti-pooling" is a desirable goal and might be furthered by inducing the guilty to refrain from lying by remaining silent, this Article points out that, in contrast to European countries, the United States already has considerable "anti-pooling" incentives apart from the right to silence. Particularly in the trial context, the costs of speaking in America are very high. By penalizing those who speak, we induce the guilty to remain silent by using the penalty for speaking as a stick and the safety of silence as a carrot. With strong "anti-pooling" measures in the form of potent impediments to speaking already in place in American criminal trials, we may not need a powerful right to silence in order to achieve the "anti-pooling" that Seidmann and Stein believe is so important to the credibility of innocent suspects. Next, this Article inquires into the validity of the market analogies and assumptions on which the "anti-pooling" theory rests. First, I will contend that the market analogy has little practical relevance in the real world ofthe American criminal trial which normally is a concentrated, one-shot process where factfinders do not accumulate market-savvy by continuous exposure to the marketplace of exonerating statements. Second, even if factfinders might gain some market experience, they have no way oftesting the products they chose or reject—the exonerating statements ofcriminal defendants. They have no way of knowing whether the shrinking pool of suspects and defendants claiming and What WeNeed to Do to Rebuild It 68 ( 1 999) (urging the adoption of a system involving "some formal pressure" on suspects to cooperate with the police). 40. See Akhil Reed Amar, The Constitution and Criminal Procedure: First Principles 56-57 (1997). For a critical analysis ofProfessor Amar's proposals, see Stuart Taylor, Jr., Rethinking the Fifth Amendment (Again), LEGAL Times, July 17, 1995, at 27. 936 INDIANA LAW REVIEW [Vol. 35:925 innocence is due to more guilty suspects remaining silent (thus increasing the proportions of innocents in the pool ofthose making exonerating statements) or to other factors such as more guilty suspects confessing, more plea bargaining, or fewer innocent suspects arrested or prosecuted. Third, this Article points out that in today's real world of police interrogation the pool of exonerating statements by guilty and innocent suspects is rather large, and any incremental increase in the pool of exonerating statements that might be caused by elimination ofthe right to silence would not be likely to decrease the factfinder's perception of the credibility of claims of innocence. Assuming that "anti-pooling" can affect the factfinder's evaluation ofclaims of innocence, the Article takes a close look at the validity of some of the assumptions of the "anti-pooling" theory in terms of the degree the innocent might be helped through the exercise of the right to silence by those who otherwise would lie. This will entail looking at both the number of innocent suspects who are in a position to be helped by "anti-pooling" and the number of guilty people who, without the right to silence, would speak in a way that would harm innocents. First, only those innocent suspects in a position to benefit from "anti- pooling" are those who are faced with evidence ofmoderate strength and whose stories are unconvincing. Second, according to the "anti-pooling" theory, only the guilty who, absent the right to silence, would tell convincing stories would be in a position to spread the benefits of silence to the innocent. This category is quite limited. Only a very small proportion of guilty suspects and defendants who now claim the right to silence would tell convincing stories and confuse factfinders if the right were eliminated. In short, few guilty fabricators help the innocent by exercising their right to silence. Seidmann and Stein assume that the right to silence causes the guilty to switch from telling lies to remaining silent, but not from telling the truth to silence. They further assume that without a right to silence, suspects and defendants who would have remained silent would have no choice but to tell lies that would "pool" with the true claims of innocent suspects thereby increasing the likelihood of their conviction. However, it is likely that even with the prospect of adverse inferences from silence, many suspects would continue to remain silent, particularly career criminals and defendants in weak cases where, without a confession, the prosecution may not be able to satisfy its burden of proof. It is even less likely that the threat of adverse inferences from silence at trial would convince all defendants to take the stand because of the perils of testifying, including the prospect of aggressive cross-examination which may expose prior convictions and other highly damaging evidence. Furthermore, even ifthe threat ofadverse inferences would convince more to speak, many would either confess or fabricate ineffectively, particularly in the context ofpretrial interrogation where the guilty are less likely to be able to convincingly shape their denials as they would at trial. To support their claim that the right to silence causes the guilty to switch from lying to remaining silent, but not from confessing to remaining silent, Seidmann and Stein dismiss evidence of a reduction in confessions following implementation ofMiranda. Instead, they look to British studies finding that the 2002] THE RIGHT TO SILENCE 93 7 1 994 Criminal Justice and Public Order Act (CJPOA), 41 which permitted adverse inferences from silence during pretrial interrogation and at trial, caused more suspects and defendants to speak, but it did not increase the confession rate. However, it is dangerous to draw conclusions concerning alterations in particular aspects of the right to silence from foreign legal systems with very different procedural rules and professional cultures. The rule against adverse inferences cannot meaningfully be analyzed in isolation but must be considered in relation to other aspects of the silence right, as well as the procedural context in which they operate. The many differences between British and American rights and procedures suggest that limiting the right to silence in America may well have different consequences than it will across the ocean. In sum, in light ofthe small number of innocent suspects who are in a position to be helped by "anti-pooling" and the limited number ofguilty people who, without the right to silence, would speak in ways that would harm the innocent, any benefit to the innocent from the "anti-pooling" effect ofthe guilty choosing to speak rather than to remain silent most likely is marginal at best. Next, I will inquire into the costs of the right to silence stemming from the fewer guilty suspects that speak to the police or to juries. Seidmann and Stein acknowledge that the right to silence reduces the conviction rate and results in the acquittal of some guilty defendants. 42 However, they claim that drawing meaningful conclusions from a cost-benefit analysis is difficult when it is not known how many innocent suspects may be jailed without the silence right or how many guilty are now freed on account of it. They conclude that the "requisite cost-benefit calculation" is beyond the scope of their study. 43 Nevertheless, they suggest that the social benefits from fewer wrongful convictions strongly outweigh the social costs do more wrongful acquittals. 44 While the authors recognize that "anti-pooling" also might be brought about by increasing incentives to tell the truth and confess, 45 they assert that a "much cheaper" and more preferable way to "purge the lemons" is to pay potential producers of false statements to remain silent by giving them the right to do so "without sustaining punishment or adverse inference." 46 However, a meaningful 41. Criminal Justice and Public Order Act, 1994, c. 33, §§ 34-39 (Eng.) [hereinafter CJPOA]. 42. See Seidmann & Stein, supra note 1 8, at 499-500. Seidmann and Stein state that "the right to silence reduces convictions of both innocent and guilty defendants." Id. at 473. 43. Id. 44. Id. 473-74. Seidmann and Stein contend that the prevention of wrongful convictions is an "immensely greater value to society than prevention of wrongful acquittals," hence retention of the silence right would be "the socially optimal choice." Id. at 494. 45. Seidmann and Stein recognize that "the desired separation" also could be achieved by inducing more guilty suspects to confess rather than lie through such measures as more prosecutions for perjury and paying for true statements by plea bargaining, but they dismiss such prospects on the ground that they "generally incur greater social costs than do incentives for silence." Id. at 434, 460-61. 46. Id. at 461. 938 INDIANA LAW REVIEW [Vol. 35:925 analysis of the practical consequences of the right to silence requires some attempt to assess its costs both in the form of lost convictions of the guilty and diminished help for the innocent who benefit from guilty suspects either confessing or making false but refutable (and ultimately incriminating) statements. 47 Since the amount of assistance that silence by the guilty provides to the innocent through "anti-pooling" is marginal at best, the right to silence may overprotect in a way that helps many guilty, but very few innocents, avoid conviction. Finally, I will look at implications of the "anti-pooling" theory that suggest that the right to silence is such a good thing for the innocent that it should be enlarged to better protect them. I will argue that expanding the right to silence by adoption ofrules that induce more suspects to request counsel at interrogation, limit deception, or require disclosure of prosecution evidence prior to questioning, would deter some criminals from confessing and assist others in fabricating effectively. By doing so, it would undermine the goal of truth discovery, particularly in marginal cases where police may have strong suspicions but not enough evidence to persuade a jury beyond a reasonable doubt, that is, in the very cases in which confessions are most needed to convict the guilty. 48 Consequently, Seidmann and Stein's analysis suggests good reasons to be skeptical of proposals that would expand the right to silence in the pretrial context, particularly those that would formalize the interrogation process by means of lawyers armed with knowledge ofpolice evidence and sworn to use all legal means to prevent the prosecution from proving its case beyond a reasonable doubt. However, Seidmann and Stein's "anti-pooling" analysis is helpful in focusing attention on the harm caused by convincing lies, which can frustrate accurate factfinding in more important ways than their general "pooling" effect in the marketplace of exonerating statements. False statements claiming innocence, which are plausible and not subject to effective contradiction, may not only lead to the release of the guilty, but may also contribute to the arrest and conviction of the innocent by "specific pooling" (the creation of case-specific factual conflicts). This is more directly detrimental to accuracy than the diminished credence given to statements of the innocent from the mere fact that a few more guilty people lie. Furthermore the "anti-pooling" theory is helpful in emphasizing the importance of unrehearsed statements, particularly the defendant's story prior to an opportunity to contrive a response to prosecution evidence. Yet liberal admissibility ofsuch statements should be a two-way street in which defendants' early claims of innocence can be offered by the defense, as well as by the prosecution. In sum, Seidmann and Stein's "anti-pooling" analysis shows that lies come 47. False statements help the innocent in that the police will likely investigate these statements, learn oftheir falsity, and allow the prosecutor to use them as impeachment during trial. 48. Seidmann & Stein, supra note 18, at 461. Furthermore, according to Seidmann and Stein's "anti-pooling" theory, in weak or marginal cases the exercise ofthe right to silence does not benefit the innocent but merely helps the guilty avoid conviction. 2002] THE RIGHT TO SILENCE 939 in various forms and that the need to distinguish between them is important. Convincing falsehoods can be highly beneficial for the guilty but highly harmful to the innocent, and rebuttable and ultimately incriminating falsehoods can be as important to accurate factfinding as confessions. Thus, leaving in place the rules against adverse inferences and Miranda's basic right to silence warning, we should shape the right to silence and associated guarantees applicable to police interrogation with a focus on permitting procedures that tend to induce guilty suspects to tell the truth, and avoiding procedures that give them the opportunities and tools that would further the creation of uncontradictable fabrications. Such reforms would offer fewer benefits to the guilty than would an expanded right to silence, while protecting the innocent in more significant ways than merely reducing the number of lies in the marketplace ofexonerating statements. I. Assumptions and Implications of the "Anti-Pooling" Theory A. Assumptions ofthe "Anti-Pooling" Theory and Use ofMarket Models Seidmann and Stein's "anti-pooling" model is based on a number of assumptions. First, innocent suspects almost invariably will speak and assert their innocence both during pretrial interrogation and at trial. 49 Second, the fate of innocent suspects often depends on the credibility oftheir true stories. Next, in their efforts to appear innocent, guilty suspects "pool" their false stories with true accounts offered by innocent suspects and harm the innocent by diminishing the credibility of their stories. 50 In this way, when guilty suspects perjure themselves, they "impose negative externalities" on innocent suspects. 51 However, virtually all suspects seek to be released and exonerated and generally 49. Id. at 433. While Seidmann and Stein recognize that the innocent may exercise the right to silence in exceptional cases, id. at 464, they accept Bentham's claim that innocent suspects rarely exercise the right to silence. Id. at 436, 455 n.82 (putting aside exceptional cases and noting that "[t]he existence ofsilent innocents does not enter into our model"). The innocent suspect's choice to speak is rational since "an innocent suspect is ... at least as well offtelling the truth as exercising the right to silence." Id. at 466. Contrast the Supreme Court's observations in Mitchell v. United States, 526 U.S. 3 14, 329 ( 1 999) that the rule against adverse inferences from silence "is ofproven utility." The Court in Ullmann v. United States, 350 U.S. 422, 426 (1956), noted that people "too readily assume that those who invoke [the privilege] are either guilty of crime or commit perjury in claiming the privilege." Later the Court quoted Wigmore's observation that "the layman's natural first suggestion would probably be that the resort to privilege in each instance is a clear confession of crime." Lakeside v. Oregon, 435 U.S. 333, 340 n.10 (1978). But times have changed: "It is far from clear that citizens, and jurors, remain today so skeptical of the principle or are often willing to ignore the prohibition against adverse inferences from silence." Mitchell, 526 U.S. at 330. 50. Seidmann & Stein, supra note 1 8, at 433, 457-58. 51. Id. at 442-43, 458. 940 INDIANA LAW REVIEW [Vol. 35:925 will act rationally in pursuit ofthis objective. 52 Additionally, although the guilty "typically" will choose to speak rather than to exercise the right to remain silent, and despite that "silence is usually the better choice,"53 in their rational pursuit of exoneration, the guilty will refrain from "pooling" by exercising the right to silence if it appears in their interest to remain silent rather than to lie. 54 When silence is not penalized by adverse inferences of guilt, it appears as an attractive alternative to fabrication, and guilty suspects will perceive (correctly in most cases) that they are better offremaining silent than speaking. 55 Further, when the guilty "rationally exercise" the right to silence to reduce the risk of their own conviction, they also reduce the risk that innocent suspects will be wrongfully convicted. 56 Thus, by refraining from "perjuriously pooling with innocents," the guilty "minimize[s] the risk" of wrongful conviction ofthe innocent. 57 Finally, in light ofthe considerable benefits flowing to the innocent from the exercise of the right to silence by the guilty, retaining the silence right is a cheap price to pay the guilty for withholding their fabrications and increasing the prospect that innocent defendants will be acquitted. 58 In essence, Seidmann and Stein view the right to silence and its embodiment in the privilege against self-incrimination as a means of helping the innocent in their struggle to make themselves believed. Their "anti-pooling" theory posits that the right to silence induces the guilty to choose silence over fabrication, which results in fewer false statements compared with true ones, which causes factfinders to be more accepting ofthe statements of innocent suspects and leads to a greater likelihood of their acquittal. Seidmann and Stein apply and test the foregoing assumptions in theoretical laboratories of markets and games. In these "game-theoretic" models, participants in the justice system exercise rational choices based on their own welfare and on their perception ofthe choices others in the system would make given the particular rules ofthe game or marketplace. 59 The process of criminal 52. See id. at 442, 448. However, Seidmann and Stein acknowledge that the guilty often do not act rationally since they generally speak when it would be in their best interest to remain silent. Under "stressful interrogation" and with "asymmetric information," "guilty suspects often choose the worst possible move, which brings about the worst possible outcome." Id. at 464. 53. Id. at 448. Seidmann and Stein nonetheless concede that silence does have its price. At the pretrial stage, silence is "tantamount to admitting guilt and challenging the police to obtain evidence that will convict," and at trial the damage from silence is even more serious. Id. at 446-47. 54. Id. at 448. However, the fact that most guilty suspects speak to police while total silence usually is their best choice seems inconsistent with the assumption ofthe game-theoretic model that "each player's strategy is that player's best move in light of the strategies actually chosen by the other players." Id. at 465-66. 55. Id. at 465 (noting that experienced suspects and those receiving legal advice are more likely to exercise their silence right). 56. Id. at 499. 57. Id. at 457-58. 58. Id. at 473, 494. 59. Id. at 433-34. Seidmann and Stein describe their game-theoretic method, otherwise 2002] THE RIGHT TO SILENCE 94 1 investigation and trial is seen as a market for self-exonerating stories in which suspects and defendants seek to sell their accounts to factfinders—police and prosecutors at the pretrial stage and the jury at trial. 60 Since the innocent must compete with the guilty in this enterprise, measures should be adopted that "drive false statements out of the market." 61 According to Seidmann and Stein's used car market analogy, 62 with lemon- sellers (guilty fabricators) and apple-sellers 63 providing the same stories about their cars, buyers (factfinders) are confused and may disbelieve apple-sellers, thereby convicting innocent defendants. But if lemon-sellers do not make claims ofgood quality, their false statements will not be pooled with the true claims of apple-sellers. Buyers then will give greater credence to valid claims of good quality thereby increasing the chance that they will buy them (acquit apple- sellers). 64 B. The Adverse Inference Problem: The "Anti-Pooling" Model as Both Resting on and Undermining the Right to Silence Seidmann and Stein concede that the right to silence in the form of a prohibition on adverse inferences from silence, along with aiding the innocent in avoiding unjust convictions, to some extent helps the guilty to escape conviction. But in order to help the innocent in the manner suggested by the theory, must the right to silence necessarily also hurt the guilty in a way that undermines the very basis of the right? Consider a society in which all innocent suspects asserted their claims of innocence and all guilty remained silent. If factfinders became aware of this phenomenon, they would believe all exonerating statements. According to Seidmann and Stein's model, as a society moves in this direction and as fewer known as the Bayesian Nash Equilibrium Tool, as focusing on a person's rational choice in a strategic situation in which that person's welfare or best choice depends on his or her perception about the choices that others in the game will make, and explores the effects of altering the rules of the situation (the game) on all participants. Id. at 441, 465. The theory assumes that the belief of all players, including the suspect and the factfinder, as to how others will respond are correct such that "each player's strategy is that player's best move in light ofthe strategies actually chosen by the other players." Id. at 465-66. 60. Id. at 460. 61. Id. 62. Id. at 459. 63 . For the purposes of this Article, apple-sellers are innocent suspects who tell the truth to factfinders. Seidmann and Stein did not use this term in their article. 64. According to Seidmann and Stein, with no anti-pooling protections, buyers ofused cars will pay no more than average value and quality car owners will take their cars off the market, eventually creating a single "market for lemons." Id. at 459. But innocent suspects cannot readily opt out ofthe market and go home. They are undergoing custodial interrogation or on trial. They must try to sell their cars, and it is unreasonable to assume that they will not convey their cars' attributes to prospective sellers merely because owners of lemons will lie about their cars. Id. 942 INDIANA LAW REVIEW [Vol. 35:925 guilty suspects try to lie their way out ofaccusations, factfinders will give greater credence to the accounts ofinnocent suspects. Increasing the proportion ofguilty persons who refuse to speak will increase the likelihood that the factfinder will accept the accounts ofthe innocent (who virtually always speak and are assumed to speak the truth) and will convict them less often. In fact, the authors visualize a perfect "right-to-silence regime" in which "neither pooling nor the ensuing wrongful convictions materialize" for the reason that "innocents still tell the truth, whereas guilty suspects separate themselves by rationally exercising the right." 65 In this ideal world, where measures have been adopted which "drive false statements out ofthe market," 66 the guilty will separate themselves from the innocent by remaining silent, and the jury will draw "a favorable inference from any exculpatory statement, and innocent suspects (who alone make such statements) are thus acquitted." 67 But does this model also depend on factfinders being more skeptical ofthose who assert their right to silence and refuse to speak, with police and prosecutors less inclined to release them during the pretrial stage and juries more inclined to convict them at trial? 68 If so, the model depends on factfinders being more inclined to employ adverse inferences from silence at the same time that it relies on a broad right to silence unencumbered by adverse inferences as a safe harbor for the guilty who otherwise would lie and confuse things. It might be argued that with guilty suspects remaining silent more frequently, factfinders would give more weight to innocent accounts while not changing their attitude toward those who claim the right to silence, thus avoiding any adverse inferences. But the used car analogy and notions ofhuman behavior point in the opposite direction. With fewer lemon-sellers falsely touting their cars and buyers placing greater credence in the true claims of apple-sellers, buyers naturally would be more skeptical of car sellers who remained silent and refused to provide any information regarding the history or condition of their cars. In a society in which only innocent suspects claimed innocence and all guilty suspects remained silent, factfinders would believe all exonerating statements, and convict all silent defendants. As a justice system moves in this direction and as fewer guilty suspects claim innocence, the model posits that factfinders will tend to believe more innocent accounts, or as Seidmann and Stein put it, will be more hesitant to "rationally discount the probative value ofuncorroborated exculpatory statements." 69 But if so, factfinders naturally would be more skeptical of those who failed to speak and assert their innocence and more readily assume guilt from silence. In short, the authors' model suggests that the more often guilty suspects exercise their right to silence, the more it hurts them by weakening the rule 65. Id. at 503. 66. Id. at 460. 67. Id at 469. 68. Id. Seidmann and Stein suggest that remaining silent during pretrial questioning will cause the authorities to concentrate their efforts on guilty suspects. Id. at 447. 69. Id at 503. , 2002] THE RIGHT TO SILENCE 943 against adverse inferences from silence. 70 The model, which is founded on a robust right to silence in the form ofa prohibition on adverse inferences from its exercise, naturally creates its own counter pressures which bring about a greater likelihood of adverse inferences from silence, thus undercutting its own foundations. The more efficiently the model operates and the more guilty people choose silence over fabrication, the more precarious the right to silence becomes. Furthermore, once the guilty become aware ofthe increased credibility given to claims of innocence, they would tend to prefer lying over silence. This "free- rider" tendency would then undercut the central assumption ofthe "anti-pooling" theory—that if fewer guilty speak, the remaining speakers will be seen as more credible. 71 With the "anti-pooling" model containing the seeds of its own destruction and eventually collapsing of its own weight, things would tend to even out in the end. 72 C. "Anti-Pooling " by Using the Stick Rather than the Carrot Assuming that "anti-pooling" is a desirable goal and that it might be furthered by inducing the guilty to refrain from lying by remaining silent, similar results might be achieved by making the defendant an offer he cannot refuse. Instead of inducing the guilty to remain silent by utilizing the right to silence carrot, which involves maintaining an attractive safe harbor in silence, one might 70. Of course, a trend toward more guilty suspects choosing silence over fabrications that has the practical effect of weakening the prohibitions on using silence to infer guilt may result in more accurate factfinding. Id. The authors agree that in the real world, silence in the face of criminal accusations is highly probative evidence of guilt since innocent suspects virtually always proclaim their innocence. Certainly, silence has much more than some "tendency to make the existence of any fact that is of consequence . . . more probable . . . than . . . without the evidence." Fed. R. Evid. 401. Provided factfinders are aware of the fact of silence and its significance, the practical effect of rules prohibiting adverse inferences may be weakened such that the cost of the right to silence in terms of more acquittals of guilty suspects might be significantly reduced. However, if"anti-pooling" increases the significance ofsilence in the eyes ofthejury, making them more skeptical of those who choose it, one might contend that, without a change in present rules of evidence, the shift would result in the danger of convicting more innocent defendants who choose to remain silent to avoid impeachment with prior convictions or the chance of appearing unconvincing under vigorous cross-examination in the formal trial context 71 . I owe this "free-rider" insight to my colleague Professor Rory Little. 72. Much would depend on such factors as the extent to which the guilty exercise the right, whether factfinders are aware that claims of innocence are becoming more credible from the fact that fewer guilty people make such claims, and whether factfinders nevertheless are following judicial instructions against inferring guilt from refusals to speak or to testify. In light of these considerations, it seems that the weakening ofthe no-adverse-inference rule is more likely to occur with respect to trial silence as opposed to pretrial silence, since generally the ultimate factfinder—the jury—is unaware of whether the defendant during custodial interrogation proclaimed his innocence or remained silent, whereas his refusal to testify at trial is an evident fact of which the jury is always aware, although legally forbidden to consider. 944 INDIANA LAW REVIEW [Vol. 35:925 seek to increase the possible harm from speaking as a stick to persuade the guilty to keep their lies to themselves. Instead ofmaking silence safer, speaking could be made more dangerous. In fact, we are now taking both approaches with an emphasis on punishing those who choose to speak. While we view the opportunity to speak and to testify in court as a fundamental right ofthe accused, our present system penalizes speaking both pretrial and at trial—and the perils of talking are increasing. In the pretrial context, speaking to the police or to prosecutors clearly is very dangerous. While there are some small benefits in confessing at an early stage of an investigation, 73 these benefits are greatly overshadowed in most cases by the damage such statements cause during plea bargaining and at trial. Even exonerating statements usually work against most defendants who eventually must choose to either accept a plea bargain or go to trial since evidence rules permit the prosecutor to use them freely, but generally prohibit their use by the defendant to bolster his claim of innocence. Ordinarily, such statements are inadmissible hearsay when offered by an accused.74 In contrast to lawyers in 73. In federal cases, a defendant may receive a three-level reduction in sentence for acceptance of responsibility in a timely manner. See U.S. Sentencing Guidelines Manual § 3E1 . 1 (b) (200 1 ) [hereinafter U.S.S.G.]; United States v. Corona-Garcia, 21 F.3d 973, 980-8 1 (9th Cir. 2000). 74. Only in unusual cases are defendants able to introduce their exonerating stories at trial. For example, prior statements of an accused generally are not admissible as consistent statements under Fed. R. Evid. 801(d)(1)(B). See United States v. Nelson, 735 F.2d 1070 (8th Cir. 1984) (defendant's prior exculpatory statement was not admissible as a consistent statement under rule 801(d)(1)(B)). Such statement may be admissible to rehabilitate a defendant only if defendant testifies and is impeached by an allegation of an improper motive and the statement was made before the improper influence or motive was alleged to have arisen. See Tome v. United States, 5 1 3 U.S. 150 (1995). Nor are post-arrest statements by a defendant asserting innocence admissible under the state of mind exception. See United States v. Carter, 910 F.2d 1524 (7th Cir. 1990) (defendant's prior exculpatory statement was not admissible under the state of mind exception of 803(3) since it referred to a past, rather than to a then-existing, state of mind); United States v. Rodriguez-Pando, 841 F.2d 1014 (10th Cir. 1988) (tape recording of defendant's statement to police the day following his arrest in which defendant claimed that he had been coerced to act as he did was inadmissible on the ground it was a statement of memory of past events and beliefs). A limited avenue of admissibility is available if a defendant raises a mental defense, calls an expert witness, and seeks to elicit the statement as one of the bases for the expert's opinion. Fed. R. Evid. 703 provides that if the facts or data upon which an expert bases an opinion or inference are of a type reasonably relied upon by experts in a particular field in forming opinions or inferences upon the subject, the facts or data need not be admissible in evidence in order for the opinion or inference to be admitted. However, the rule was amended, effective December 2000, to limit the admission of facts which form the basis for expert opinion when offered by the party calling the expert. Under the new rule, facts that are otherwise inadmissible "shall not be disclosed to the jury by the proponent of the opinion or inference unless the court determines that their probative value in assisting the jury to evaluate the expert's opinion substantially outweighs their prejudicial effect." Fed. R. EVID. 703. 2002] THE RIGHT TO SILENCE 945 European countries who often will advise suspects to be cooperative and truthful, American defense lawyers virtually always advise suspects not to talk to police. 75 For example, when O.J. Simpson appeared with his lawyer at a bail hearing and attempted to explain why he had fled in the Bronco, his lawyer advised him to remain silent. However, he kept talking. Finally his lawyer warned Simpson, "I will not allow you to speak and I will resign as your lawyer ifyou continue to do so." 76 Virtually all criminal defense attorneys would view such advice as sound, indeed vital, under the circumstances. Ofcourse, another reason why pretrial silence is attractive in America is that it provides a safe harbor for the guilty. The right to silence caution has some effect, but the real bite comes from associated rules and practices, such as the strict cut-off rules which require terminating interrogation whenever a suspect declines to speak or requests a lawyer. While Miranda and its progeny do not require the presence of stationhouse lawyers 77 or even give defendants the right on request to see a lawyer unless interrogated, 78 the Miranda-Edwards rules require terminating interrogation whenever a suspect declines to speak or requests a lawyer. 79 Such rules contrast sharply with the continental and English 75. Watts v. Indiana, 338 U.S. 49, 59 (1949) (Jackson, J., concurring in part and dissenting in part). Justice Jackson's well known observation over forty years ago still states the accepted wisdom of criminal defense lawyers in this country: "[A]ny lawyer worth his salt will tell the suspect in no uncertain terms to make no statement to police under any circumstances." Id. 76. Before represented by counsel, Simpson had gone to police headquarters voluntarily, submitted to questioning for thirty-two minutes, gave a blood sample and returned home. Arenella, supra note 1 6, at 1 237. Arenella characterized the police questioning as "polite" and asks why the detectives did not engage in more prolonged and tougher interrogation since he had waived his right to silence and to counsel. While regarding this as "one ofthe mysteries ofthe case," Arenella asks whether detectives would have adopted the same polite and deferential style ofquestioning if they were dealing "with a more typical suspect." Id. at 1237 n.8. 77. See Davis v. United States, 512 U.S. 452, 460 (1994) (citing Miranda v. Arizona, 384 U.S. 436, 474 (1966) (noting that Miranda rejected the notion "that each police station must have a 'station house lawyer' present at all times to advise prisoners")). 78. See Duckworth v. Eagan, 492 U.S. 1 95 ( 1 989) (holding that Miranda does not require that lawyers be producible on call, but only that a suspect be informed that he has the right to counsel before and during questioning and that counsel would be appointed for him if he could not afford one). Thus, "[i]fpolice cannot provide appointed counsel, Miranda requires only that police not question a suspect " Id. at 204. See also Moran v. Burbine, 475 U.S. 412 (1986) (finding that Miranda had been waived despite fact that defendant had not been informed that counsel purporting to represent him had called police and requested that no questioning take place and that police had assured counsel that defendant would not be questioned). 79. Miranda, 384 U.S. at 436. See also Minnick v. Mississippi, 498 U.S. 146 (1990) (holding thai Edwards' protection does not cease once suspect has consulted with his attorney, such that once defendant has requested an attorney, interrogation must cease and police may not re- initiate questioning without an attorney present even though defendant has consulted with his attorney); Arizona v. Roberson, 486 U.S. 675 ( 1 988) (holding that once a suspect cuts offcustodial interrogation by invoking his right to counsel, he may not, as long as he remains in custody, be 946 INDIANA LAW REVIEW [Vol. 35:925 approaches which allow continued questioning in the face of a suspect's refusal to speak. 80 In the trial context, the dangers from testifying are considerable and they are on the rise. America's super-adversary trial procedure shields a defendant from inquiry by focusing the trial on the lawyers rather than on the accused who often appears set apart from the trial process. 81 But a defendant who dares take the witness stand will face cross-examination by an aggressive prosecutor as well as the possibility ofperforming poorly before the body that will determine his fate. Most important, our rules of evidence operate to strongly discourage the defendant from taking the stand by saying to him, Ifyou testify, the jury will become aware ofyour felonious history, you may be prosecuted for perjury or your sentenced enhanced ifyou lie, and you will be cross-examined by an aggressive prosecutor;82 but if you questioned by the original interrogators or others about an offense wholly unrelated to the crime as to which he has already requested counsel, unless counsel has been provided him or the suspect himself initiates further communications with officials); Smith v. Illinois, 469 U.S. 91, 95 (1984) (quoting Fare v. Michael C, 442 U.S. 707, 719 (1979)) (characterizing Edwards as a "ridged prophylactic rule" which embodies two distinct inquiries: whether defendant actually invoked his right to counsel and if he did, the court may admit responses to further questioning only if defendant both initiated further discussions with the police and knowingly and intelligently waived the right he had invoked); Edwards v. Arizona, 451 U.S. 477 (1981) (announcing a stricter rule when defendant requests a lawyer: That once a suspect has asserted his right to counsel under Miranda, there can be no further interrogation until counsel has been made available to him unless defendant himself initiates further conversations with the police); Michigan v. Mosley, 423 U.S. 96, 103 (1975) (describing a person's "right to cut off questioning" as a "critical safeguard" and stating that Miranda requires a "fully effective means ... to notify the person of his right of silence and to assure that the exercise of the right will be scrupulously honored"). Although voluntary statements obtained in violation of these rules can be used to impeach, courts are beginning to permit civil actions against police for "going beyond Miranda" and continuing to ask questions after a suspect has asked for a lawyer. See Cooper v. Dupnick, 963 F.2d 1 220 (9th Cir. 1 992); see also Cal. Attorneys for Criminal Justice v. Butts, 1 95 F.3d 1 039 (9th Cir 1999) (establishing clear rule that continued interrogation after defendant's invocation of the right to counsel constituted a clear violation ofthe Fifth Amendment giving rise to civil liability in which qualified immunity is unavailable). 80. See Van Kessel, supra note 2, at 8 1 9-2 1 . 8 1 . Compared to European trials, for example, courtroom arrangement and choreography greatly limit exposure of the accused. American defense lawyers generally sit beside the accused, often between him and the jury, whereas in continental trials, lawyers usually sit in back of the accused and are restricted in prompting his responses. In England, the accused (who is placed in a dock at the center-rear ofthe courtroom) is even more separated from his barrister who sits in the front benches some distance away. For a description of the numerous incentives to speak at a French criminal trial which starkly contrasts with American practices, see Lerner, supra note 4. 82. The threat of felony conviction impeachment can be a powerful deterrent to taking the witness stand. A study ofAmerican jury trials found that a defendant was almost three times more 2002] THE RIGHT TO SILENCE 947 remain silent, neither your past nor your present silence will be mentioned by the judge or prosecutor, and if you wish, the jury will be cautioned against drawing adverse inferences. 83 Furthermore, once the accused takes the witness stand, he is open to impeachment by many types of evidence previously found to be illegally obtained and inadmissible, such as fruits of illegal searches or seizures and statements obtained in violation of Miranda.™ The threat of admissibility of evidence and prior bad acts also can keep a defendant from the witness stand. In the recent highly-publicized "road rage" case in which a driver was charged with reaching into the car of another driver and throwing her fluffy, white dog into oncoming traffic, the judge ruled that if the defendant testified that the dog had bitten him first, the prosecutor could call a witness to testify that he had seen the defendant beat a disabled dog to death. The defendant did not take the stand and likely to refuse to testify if he had a criminal record than if not. See Harry Kalven, Jr. & Hans Zeisel, The American Jury 146 (1966). 83. The legal prohibition on adverse inferences precludes any reference to defendant's failure to testify. See Carter v. Kentucky, 450 U.S. 288 ( 1 98 1 ) (holding that on defendant's request, the jury was to be instructed that silence must be disregarded); Griffin v. California, 380 U.S. 609, 614 (1965) (prohibiting both judicial instructions and prosecutorial comment which suggested that defendant's silence at trial could be used as evidence of guilt); United States v. Buege, 578 F.2d 187, 188 (7th Cir. 1978) (disallowing prosecutorial argument that certain evidence was "uncontradicted" when contradiction would have required defendant to take the stand and would draw attention to his failure to do so). There appears to be a trend toward greater acceptance ofthe right to silence by both the courts and the American public, which suggests thatjuries may be taking it more seriously. Until recently, the common perception has been that the right to silence at trial is rather anemic and generally of little consequence. Jurors, and even judges, have ordinarily expected the defendant to give evidence and have held it against the defendant if he does not take the stand. However, due to recent extensive media coverage of high profile trials, the public is being exposed to situations in which the defendant does not make pre-trial statements or testify at trial. When neither the judge nor the lawyers ask why the accused fails to talk, the public slowly becomes accustomed to a system in which the accused is a silent and passive observer ofthe courtroom action. In recent years, both the legal profession and the public have become more accustomed to criminal trials in which the defendant remains silent while his lawyers attack the prosecution's case, and they have become more comfortable with the notion that the accused is not expected to personally provide his version ofthe events. This changing perception was recognized recently by the Supreme Court. See supra note 49. 84. See Michigan v. Harvey, 494 U.S. 344 (1990); United States v. Havens, 446 U.S. 620 (1980); Oregon v. Hass, 420 U.S. 714 (1975); Harris v. New York, 401 U.S. 222 (1971). Furthermore, defendant's silence, whether before or after arrest, can be used to impeach his testimony as long as it was not in response to Miranda warnings. Fletcher v. Weir, 455 U.S. 603 (1982); Jenkins v. Anderson, 447 U.S. 231 (1980). However, the prosecutor cannot use illegally obtained evidence to impeach the credibility ofdefense witnesses other than the defendant. James v. Illinois, 493 U.S. 307 (1990). 948 INDIANA LAW REVIEW [Vol. 35:925 the witness never testified. 85 The impediments to testifying are increasing. In some jurisdictions, the danger ofimpeachment by prior convictions has become more serious,86 and the Supreme Court recently limited the ability of defendants to lessen the impact of such impeachment. 87 The Court upheld a rule penalizing a testifying defendant who attempts to "remove the sting" ofprior conviction impeachment by bringing out the fact of the conviction on direct examination. 88 By doing so, defendant waived the right to appeal the judge's adverse in limine ruling allowing such impeachment. 89 Finally, the Court recently permitted the prosecutor to comment to the jury regarding defendant's presence at trial that allowed him to tailor his testimony to fit the evidence which had been presented. 90 D. The Consequences ofSticks and Carrots What are the results of the present system which penalizes speaking and provides a safe harbor for silence both pretrial and at trial? To support their claim that the right to silence "cannot be responsible for many erroneous acquittals," Seidmann and Stein assert that "suspects do not exercise the right to silence very often either at interrogation or at trial." 91 But this does not appear to reflect the current situation in America where a substantial number ofsuspects assert their Miranda rights by refusing to answer police inquiries and an even 85. Evelyn Nieves, Driver Who Tossed Dog is Convicted ofCruelty, N.Y. TIMES, June 20, 2001, at A 12. The case was tried in San Jose, California, and the defendant was convicted. Id. 86. In California, for example, prior to 1982 the courts had restricted the prosecutor's impeachment of a defendant with previous convictions. However, an initiative entitled the "Victim's Bill of Rights" added Section 28(0 to Article I of the California Constitution and provided that "[a]ny prior felony conviction of any person in any criminal proceeding, whether adult or juvenile, shall subsequently be used without limitation for purposes of impeachment . . . in any criminal proceeding." People v. Collins, 722 P.2d 173, 175 n.l (Cal. 1986) (quoting Cal. Const, art. 1, § 2816); People v. Castro, 696 P.2d 1 1 1 (Cal. 1985). Another provision ofthe 1982 initiative permitting admission of all relevant evidence (with some exceptions) now permits impeachment with prior conduct which did not result in a conviction. See People v. Wheeler, 841 P.2d 938, 943 (Cal. 1 992) (holding that the new rule gives trial courts broad discretion to admit or exclude all acts of dishonesty or moral turpitude relevant to impeachment). 87. Ohler v. United States, 529 U.S. 753, 760 (2000). 88. Id. at 758. 89. Id. (holding that if, following ajudge's in limine ruling permitting impeachment use of defendant's prior convictions, defendant preemptively testifies to those convictions on direct examination, defendant thereby waives the right to challenge the judge's ruling on appeal). 90. See Portuondo v. Agard, 529 U.S. 61, 65-76 (2000) (holding that the prosecutor's comments in her summation calling the jury's attention to the fact that defendant had the opportunity to hear all other witnesses testify and to tailor his testimony accordingly did not violate defendant's Sixth Amendment right to be present at trial, his Sixth Amendment right to confront witnesses, or his Fifth and Sixth Amendment right to testify in his own behalf). 91 . Seidmann & Stein, supra note 1 8, at 448. 2002] THE RIGHT TO SILENCE 949 larger proportion of defendants refuse to testify at trial. While most suspects waive their Miranda rights and make statements, a substantial number do not. The proportion ofAmerican suspects who assert their right to silence and refuse to answer questions varies considerably but averages around twenty percent. 92 The frequency ofdamaging statements also varies, but most studies have found that confession rates declined following Miranda and that in the post-Miranda era, confessions are found in less than one-half of the cases. 93 A study of thirty-seven capital jury trials in California from 1988 to 92. In my 1986 review of American studies, I found that the proportion of suspects who remained silent varied from five percent to sixty percent. See Gordon Van Kessel, The Suspect as a Source of Testimonial Evidence: A Comparison ofthe English and American Approaches, 38 Hastings L.J. 1, 116-19(1986). Professor Cassell's more extensive review of American studies found that the percentages of those invoking Miranda vary widely (from seventy-seven percent to four percent) averaging about twenty percent. Paul G. Cassell, Miranda 's Social Costs: An Empirical Reassessment, 90 Nw. U. L. Rev. 387, 495 (1996). Other recent studies have come to similar conclusions. In a study of police interrogation practices at three police departments in Northern California during 1992 and 1993, Professor Richard Leo found that twenty-one percent of the suspects (thirty-eight out of 1 82) invoked their Miranda rights. Richard A. Leo, Inside the Interrogation Room, 86 J. Crim. L. &CRIMINOLOGY 266, 268, 275-76 ( 1 996). Cassell's 1 994 Salt Lake County study found that 16.3% of suspects given Miranda invoked their rights initially. Cassell, supra, at 496. 93. In my 1986 review of English and American confession rates, I noted that American studies conducted after Miranda became common knowledge detected an increase in refusals to answer questions and "some decline in confession rates." Van Kessel, supra note 92, at 1 28. Professor Paul Cassell's later, more extensive reviews ofAmerican studies and his 1994 Salt Lake County study also found declining confession rates following Miranda. Paul G. Cassell, All Benefits, No Costs: The GrandIllusion o/Miranda'j Defenders, 90 Nw.U.L. REV. 1084, 1091-92 ( 1 996) (concluding that studies before Miranda found that defendants made damaging statements in well over fifty percent ofthe cases, while after Miranda the rates dropped considerably such that the rates now vary from twenty percent to fifty percent); Cassell, supra note 92, at 483 (finding a sixteen percent nationwide drop in confession rates following Miranda when confessions were necessary for conviction in twenty-four percent ofthe cases, resulting in a 3.8% loss ofconvictions in serious cases (. 16 x .24)); Paul G. Cassell & Bret S. Hayman, Police Interrogation in the 1990s: An Empirical Study ofthe Effects o/Miranda, 43 UCLA L. REV. 839, 842 (1996) (summarizing their Salt Lake County study and finding a confession rate of thirty-three percent suggesting that "Miranda has reduced the confession rate"). But see Stephen J. Schulhofer, Miranda 's Practical Effect: Substantial Benefits and Vanishingly Small Social Costs, 90 Nw. U. L. Rev. 500, 545 (1 996) (disputing Cassell's calculations and concluding that with the "necessary adjustments" the confession rate change drops from sixteen percent to 4. 1% and the harm to law enforcement due to lost cases declines to "at most 0.78%"). Cassell responded to Schulhofer, contending that studies before Miranda found that defendants made damaging statements in well over fifty percent of the cases, while after Miranda the rates dropped considerably such that the rates now vary from twenty percent to fifty percent. Cassell, supra, at 1091-92. But see George C. Thomas, III, Law 's Social Consequences, 51 RUTGERS L. REV. 845, 852 n.36 (1999) (reading Cassell's studies as showing a more than fifty percent chance that interrogation will lead to incriminating statements). 950 INDIANA LAW REVIEW [Vol. 35:925 1 992 found that defendants made pretrial confessions in only twelve ofthe cases (thirty-two percent). 94 This post-Miranda decline is not surprising since the right to silence has become a familiar feature on the legal landscape. Moreover, those with felony conviction records, who are more likely to refuse to make statements, 95 are becoming more aware of the consequences of waiving their Miranda rights. As to exercising the right to silence at trial, with increasing frequency defendants are not taking the stand at trial as they once did. In colonial America, virtually all defendants testified at trial, and this trend continued throughout the first half of this century. Studies of trials in the 1920s and the 1950s show that very few defendants refused to testify at trial and that few were helped by such refusals. 96 However, following Griffin and the Supreme Court's decisions ofthe 1960s and 1970s which strengthened the right to silence, fewer and fewer defendants are testifying at trial. Professor Schulhofer's study of Philadelphia In a multiple regression analysis of FBI data, Cassell and Fowles found that national crime clearance rates fell precipitously in the two years immediately following Miranda and have remained at lower levels ever since and concluded that "Miranda has seriously harmed society by hampering the ability of the police to solve crimes." Paul G. Cassell & Richard Fowles, Handcuffing the Cops? A Thirty-Year Perspective on Miranda's Harmful Effects on Law Enforcement, 50 Stan.L.Rev. 1055, 11 32 (1998) [hereinafter Cassell & Fowles, Handcuffing the Cops?]. John J. Donohue used his own regression model with Cassell and Fowles' data and found a statistically significant post- 1966 effect only for total violent crime and for the individual crime of larceny and could neither substantiate nor reject the claims of Cassell and Fowles. John J. Donohue III, Did Miranda Diminish Police Effectiveness?, 50 STAN. L. REV. 1 147, 11 72 (1998). In response, Cassell and Fowles then contended that Donohue's figures largely supported their conclusions that crime rates fell substantially after Miranda and that Miranda was in large part a cause of this decline. Paul G. Cassell & Richard Fowles, Falling Clearance Rates after Miranda: Coincidence or Consequence?, 50 STAN. L. Rev. 1181 (1998) [hereinafter Cassell & Fowles, Falling Clearance Rates]. 94. Scott E. Sundby, The Capital Jury andAbsolution: The Intersection of Trial Strategy, Remorse, and the Death Penalty, 83 CORNELL L. REV. 1 557, 1 584- 1 585, tbl. 8 ( 1 998). Only five of the twelve voluntarily turned themselves in or were arrested on unrelated charges and brought up the killing on their own. Id. The aim ofthe study was to assess the effect of remorse on ajury's decision to impose a sentence of death or life without parole. Figures were drawn from the California segment ofthe Capital Jury Project (CJP) which involved a study ofthirty-seven capital cases which were tried to juries during the years 1988 to 1992. Id. 95. See Cassell, supra note 92, at 465. 96. Even studies oftrials in the 1 920s and the 1 950s reveal that very few defendants refused to testify at trial and that few were helped by such refusals. See ARTHUR TRAIN, The PRISONER AT THE Bar 209-1 2 ( 1 923) (referring to an empirical study revealing that only twenty-three out of300 defendants choose to remain silent at trial (twenty-one of these were convicted anyway)). In their study of the American jury, Kalven and Zeisel describe an empirical Chicago jury study of trials conducted during the middle and late 1 950s showing that ninety-one percent ofdefendants without prior records and seventy-four percent ofthose with prior records, chose to testify at trial. Kalven & Zeisel, supra note 82, at 146. 2002] THE RIGHT TO SILENCE 95 1 felony trials in the 1 980s illustrates the decline in defendant testimony. 97 Nearly one-half of felony defendants did not testify at trial and twenty-three percent of this group was acquitted, 98 while fifty-seven percent ofmisdemeanor defendants chose not to testify at trial and thirty-four percent were acquitted. 99 A study of thirty-seven capitaljury trials in California from 1988 to 1992 revealed that "with a few notable exceptions, most defendants did not testify." 100 Only twenty-seven percent testified at the guilt phase and only twenty-two percent testified at the penalty trial (four defendants testified at the penalty phase only, while four testified at both). Thus, only thirty-eight percent ofthe defendants took the stand either at the guilt or the penalty trial. While studies on the number ofdefendants who testify are few, these results are consistent with my own observations and inquiries with trial lawyers and judges: while much depends on the particular charge and defense, the nature of the evidence, and the defendant's criminal record, the extent ofrefusals to testify varies from one-third to well over one-half in somejurisdictions. The failure ofAmerican defendants to testify has become so common that even the public rarely notices when the defendant does not take the witness stand. For example, of those who have seen the movie, Reversal of Fortune, how many were aware, much less thought it unusual, that Claus von Bulow failed to tell his story to the jury in either trial? 101 European practice provides a stark contrast. In continental Europe, nearly all defendants choose to testify. 102 Likewise, in England, it is the rare case in which the accused does not take the stand and give evidence. 103 The CJPOA104 which permits adverse inferences from silence in Great Britain has had a "marked impact on both pre-trial and trial practices" with a notable reduction in the exercise of silence among suspects in police custody and more defendants 97. Schulhofer, supra note 1 6, at 329-30. 98. Id. at 329-30 (citing Stephen J. Schulhofer, Is Plea Bargaining Inevitable?, 97 Harv. L. Rev. 1037, 1080 (1984)). Ofthe 162 felony defendants tried by ajudge without ajury, seventy- nine did not testify and forty-four percent of those who remained silent were acquitted on the principal charge (some convicted on lesser counts). Id. at 330 n.72. 99. Id. at 329-30. Altogether, thirty-nine percent of those who remained silent avoided conviction on the principle charge. Id. at 330 n.73 (citing Stephen J. Schulhofer, No Job Too Small: Justice Without Bargaining in the Lower Criminal Courts, 1 985 AM. B. FOUND. RES. J . 5 1 9, 571). 1 00. Sundby, supra note 94, at 1 56 1 . 101. The movie portrayed the efforts of Alan Dershowitz and his Harvard law students in obtaining the reversal of the conviction of Doctor von Bulow for murdering his wife. Even the Rhode Island Supreme Court, in reversing his convictions for attempted murder, failed to mention the fact that he never took the stand at the trial. See State v. von Bulow, 475 A.2d 995 (R.I. 1 984). 102. See supra note 1 and accompanying text. 103. Graham Hughes, English Criminal Justice: Is It Better than Ours?, 26 ARIZ. L. REV. 507, 590-91 (1984) (remarking that in England the case in which the defendant fails to testify is "exceptional" whereas "defendant's silence is becoming the common practice in trials in the United States"). 104. CJPOA, 1994, c. 33, §§ 34-39 (Eng.). 952 INDIANA LAW REVIEW [Vol. 35:925 testifying at trial. 105 With respect to pretrial questioning, before the CJPOA, ten percent of defendants refused all questions, thirteen percent refused some questions, and seventy-seven percent answered all questions. After the passage ofthe CJPOA, only six percent of defendants refused all questions and only ten percent refused some questions with eighty-four percent answering all questions. 106 There are few English studies regarding the effect of the CJPOA on the decision to testify at trial, but a study from Northern Ireland on the effect ofan Order permitting adverse inferences from a defendant's silence supports the view that the effect has been considerable. 107 For those charged with scheduled offenses (terrorist cases), the proportion ofdefendants refusing to testify declined from sixty-four percent in 1987 to forty-six percent in 1991 . ,08 The percentage dropped further to twenty-five percent in 1995. 109 Those charged with non- scheduled offenses (non-terrorist cases), the proportion of defendants refusing to testify fell from twenty-three percent in 1987 to fifteen percent in 1991 . ,I0 By 1995, the number had dropped to three percent. m The authors of the study concluded that with respect to non-scheduled defendants, the Order has rendered the number refusing to testify "almost negligible" such that it is now "only in the exceptional case that such defendants will absent themselves from the witness box." 112 There are several reasons for the American trend toward more reliance on silence, but much has to do with the high costs encountered in speaking to the police and testifying at trial. Consequently, with strong "anti-pooling" measures already in place in the form of potent impediments to speaking, it is appropriate to ask whether we need a powerful right to silence in order to achieve "anti- pooling" that Seidmann and Stein believe is so important to the credibility of innocent suspects and defendants. In fact, with both the safe harbor in silence and the penalties associated with speaking, we may be guilty of "anti-pooling" overkill which can be highly harmful to innocents when silence replaces more than convincing fabrications. 113 105. Tom Bucke et al., The Right ofSilence: The Impact ofthe Criminal Justice and Police Order Act 1994, Home Office Research Study 199 at 69 (2000). 106. /