Three cases are listed as hot this week. The cases all focus on the limits imposed on prosecutors either by the federal constitution, judicial order or the canons & rules of ethics.

In the first of the cases defining the limits on prosecutions isOhio v. Lamar. The Ohio Supreme Court in Lamar examines how little a prosecutor must divulge concerning potential witnesses. Without ever once citing United States v. Jencks, the Lamar Court held that what is commonly referred to as Jencks material need not be turned over to meet state and federal constitutional requirements. Summaries of statements of witnesses without the names attached suffice for Brady under the unique circumstances of Lamar.

InOsband v. Woodford the Ninth Circuit again evaluates whether a district court's protective order preventing the prosecution of any trial counsel file turned over in federal habeas discovery was permissible. The order below was "designed to ensure that the prosecution on retrial will not use the discovery permitted in the habeas proceeding to circumvent the more limited discovery available in criminal prosecutions." Noting that Circuit precedent clearly placed the decision in the hands of "the very broad discretion" of the district court in ordering discovery, the court affirms the order.

In the third hot listed case of the week the Colorao Supreme Court clarifies how far a prosecutor may go before risking disciplinary action inIn the Matter of Pautler. Attorney Pautler was brought before the disciplinary board after he had impersonated a public defender in order to secure the surrender of a person accused of capital murder. ThePautlerCourt notes that attorney Pautler is unrepentant and states that he would repeat his actions if the unique circumstances underlying this matter would arise again. ThePautlerCourt holds, however, that the unique situation involved in the instant circumstance and Pautler's long history of public service, among other mitigating factors, justifies a suspended sentence (3 months suspension of license) instead of disbarment.

Two noncapital Supreme Court cases are also noted this week. InUnited States v. Cotton etal.a unanimous Court held failure to object may waive putative "Apprendi" error. InAlabama v. Sheltona sharply split Court held that the right to counsel exists even where the sentence imposed is a suspended sentence.

The Focus article of the week is by Samuel R. Gross, The Risks of Death: Why Erroneous Convictions are Common in Capital Cases, 44 BFLR 469 (1996). Mr. Gross's article is oft cited in the current debate on how to improve the death penalty and reduce the chances of executing the innocent and worthy of a quick review as part of the actual nuts and bolts as to how the system is failing.

In other news of the week, the execution of Johnny Martinez in Texas has quietly begun to raise, once again, about the quality of Texas's rules and protocols in capital cases. The family of Martinez's victim, Clay Peterson, pleaded for Martinez's life to be spared. TheTexas Board of Pardon and Parole, by a 9-8 vote, decided, however, to turn a deaf ear to the victim's family members and to deny clemency. Two more executions (including that of Napoleon Beazley) are scheduled in Texas before the end of the month and five additional executions scheduled for June.

Execution Information

Since the last edition the following have been executed:

May

22 Johnny Martinez Texas (see above)

The following executions dates for this summer are considered serious:*

May

28 Napoleon Beazley Texas----juvenile

30 Stanley Baker Jr. Texas

June

5 Christopher Simmons Missouri---juvenile

11 Willie Mac Modden Texas

13 Daniel Reneau Texas

25 Robert Coulson Texas

26 Jeffrey Williams Texas

27 Gary Etheridge Texas

July

23 Randall Cannon Oklahoma

August

8 T.J. Jones Texas---juvenile

14 Javier Medina Texas----for. natl.

28 Toronto Patterson Texas---juvenile

HOT LIST

Ohio v. Lamar, 95 Ohio St. 3d 181; 2002 Ohio 2128; 2002 Ohio LEXIS 1125 (OH 05/15/2002) Relief denied on another "Lucasville riot" defendant. Claims rejected include: [1] failure to disclose exculpatory evidence; [2] judicial bias; [3] limitations on voir dire; [4] selective prosecution based on race as he was the only person charged for "killing only inmates;" [5] prejudicial joinder; [6] use of prejudicial hearsay; [7] sufficiency; [8] newly discovered evidence; [9] improper penalty phase arguments; [10] improper penalty phase instructions; and [11} prosecutorial misconduct in its remarks, statements and openings/closings.

In his first proposition of law, LaMar asserts that the state failed to disclose material evidence tending to exculpate him in the murders, thereby denying him his constitutional right to a fair trial. See Brady v. Maryland(1963), 373 U.S. 83, 87, 83 S. Ct. 1194, 10 L. Ed. 2d 215. Prior to trial, LaMar's counsel asked the state to produce all evidence of an exculpatory nature, as required by Crim.R. 16(B)(1)(f). The state's response did not provide a complete list of inmate statements, [*15] purportedly because of concerns about the safety of inmates who had given statements to law enforcement about the events of the SOCF riot. Instead, the state submitted to the trial court an eleven-page document containing summaries of inmate interviews.

At a pretrial hearing, the trial court read summaries of inmate statements from the bench and, with a few exceptions, did not identify who made them. The contents of the summaries varied widely, ranging from useless information (e.g., "He was in a cell adjoining K-2-36 but gave no useful information") to pure speculation (e.g., "He was in an adjoining cell to K-2-36 and heard a commotion and thought Weaver hung himself") to statements with exculpatory value (e.g., "he observed Eskridge kill Svette in L-corridor"; "he was in an adjoining cell * * * and identified Greg Curry telling Keith LaMar to kill Weaver because he was a snitch"). Many of the statements, however, corroborated LaMar's participation in the killings, while other statements named other assailants without eliminating LaMar as a participant.

After reading all of the summaries, the trial court granted a continuance to the defense and authorized funds for a second defense [*16] investigator to conduct additional interviews of inmate witnesses. The trial court did not, however, require the prosecution to provide the full statements (i.e., contents of the statement with the name of the inmate who gave it) to the defense. Instead, the court ordered the prosecution to provide the defense with the names of forty-three inmates who gave statements to law enforcement. Although the prosecution complied with the trial court's directive and provided a list of forty-three names, it did not match the statements with the names supplied.

Suppression by the prosecution of evidence that is favorable to the accused and "material either to guilt or to punishment" is a violation of due process. Brady, 373 U.S. at 87, 83 S. Ct. 1194, 10 L. Ed. 2d 215. Evidence suppressed by the prosecution is "material" within the meaning of Bradyonly if there exists a "reasonable probability" that the result of the trial would have been different had the evidence been disclosed to the defense. Kyles v. Whitley(1995), 514 U.S. 419, 433-434, 115 S. Ct. 1555, 131 L. Ed. 2d 490; see, also, United States v. Bagley(1985), 473 U.S. 667, 105 S. Ct. 3375, 87 L. Ed. 2d 481. [*17] As the United States Supreme Court has stressed, "the adjective ["reasonable"] is important. The question is not whether the defendant would more likely than not have received a different verdict with the evidence, but whether in its absence he received a fair trial, understood as a trial resulting in a verdict worthy of confidence." Kyles, 514 U.S. at 434, 115 S. Ct. 1555, 131 L. Ed. 2d 490; see, also, Strickler v. Greene(1999), 527 U.S. 263, 289-290, 119 S. Ct. 1936, 144 L. Ed. 2d 286.

Assuming arguendothat the prosecution "suppressed" the evidence LaMar complains of within the meaning of Brady, n2 we find no due process violation. On the record before us, we find no reasonable probability of a different trial outcome had the defense received the full statements. Many of the statements identified LaMar as a participant in the murders. And statements identifying other inmates as participants did not exculpate LaMar because each victim had been attacked by multiple assailants. Cf. State v. Waddy(1992), 63 Ohio St.3d 424, 433, 588 N.E.2d 819 (holding evidence not to be material within the meaning of Bradywhen the [*18] evidence did not eliminate the defendant as the perpetrator). Finally, with respect to murders in cellblock L-6, none of the statements assisted LaMar's alibi defense (i.e., that LaMar was in the recreation yard at the time of the killings). In short, nothing in the contents of the statements "could reasonably be taken to put the whole case in such a different light as to undermine confidence in the verdict." Kyles, 514 U.S. at 435, 115 S. Ct. 1555, 131 L. Ed. 2d 490.

As a separate Bradyclaim, LaMar argues that the trial court should have found a Bradyviolation and granted his motion for a new trial after defense counsel discovered that the prosecution had provided complete inmate statements--the same material the defense had requested at the pretrial hearing in LaMar's case--in the criminal trials of two other SOCF inmates (Derek Cannon and Rasheem Matthews). LaMar argues that the disclosure in these two noncapital cases undercuts the prosecution's proffered reason (i.e., inmate safety) for failing to disclose the information in his case. As we stated above, however, the statements were not material to LaMar's case within the meaning of Brady. We therefore reject this argument.

In his final Bradyclaim, LaMar claims that he should have received a new trial in light of evidence the defense discovered after trial. The evidence consisted of (1) Ohio State Highway Patrol summaries of two statements given by inmate Willie Kastner to investigators and (2) transcripts of interviews conducted by law enforcement officers with inmates David Hackett, Tyronne Golphin, Gerald Kelly, William Turner, and Daniel Davidson. These witnesses identified [*20] several persons other than LaMar as being involved in the murders of Vitale, Staiano, Depina, and Svette. LaMar further notes that the interviews with Hackett and Davidson suggest that Vitale and Staiano may not have been dead at the time their assailants left L-6. Because the state failed to disclose these potentially exculpatory statements, LaMar argues that there has been a Bradyviolation warranting a new trial.

We reject this Bradyclaim for reasons similar to the ones we stated above. While these statements identified other inmates, none of them exonerated LaMar. At best, these witnesses established that there were several persons who joined in beating the L-6 victims to death. The state's theory all along was that LaMar was one of many assailants who participated in murdering the victims in L-6. Therefore, the existence of this evidence does not undermine our confidence in the trial outcome.

Finding no Bradyviolation that would warrant reversal, we reject LaMar's first proposition of law.

Osband v. Woodford, 2002 U.S. App. LEXIS 9166, (9th Cir 05/15/2002) District court's protective order on trial counsel's files (barring admissibility at any retrial) affirmed.

The State argues that the district court's denial of the motion to reconsider conflicts with Ninth Circuit authority. We disagree. In declining to reconsider the magistrate's protective order, the district court properly applied the law of this circuit as set forth in our en banc decision in McDowell II.

The protective order [*15] in McDowell Iand IIwas virtually identical to the protective order in this case. The order "stated that the Attorney General could use any documents produced from the trial counsel's file only for purposes of the pending habeas litigation." McDowell II, 197 F.3d at 1254. "By the terms of the order, this restriction was to continue in effect after the conclusion of the habeas proceedings and would apply 'in the event of a retrial of all or any portion of [the petitioner's] criminal case.'" Id. After the district court granted the writ of habeas corpus and reaffirmed that the order was to remain in effect, the State filed a motion for reconsideration under Federal Rule of Civil Procedure 59(e), arguing that the order was improper. On appeal from the denial of that motion, a three-judge panel of this court agreed with the State that the protective order "misapprehended the law regarding the scope of the attorney-client privilege and was thus an unintentional abuse of discretion." McDowell I, 173 F.3d at 1191. In the panel's view, the district court's order constituted "an unwarranted anticipatory interference with the prerogatives of the [*16] state courts" because it "effectively enjoins California courts from adjudicating a state law issue concerning [the petitioner's] waiver of the attorney-client privilege" and attempts "to retain continuing supervisory jurisdiction over the conduct of the retrial in state court." Id. at 1191. Such an order, the panel wrote, "contravenes basic principles of comity and federalism." Id.

On rehearing en banc, we disagreed, holding that "the district court did not commit clear error when it limited access to the file pursuant to the terms of the protective order." McDowell II, 197 F.3d at 1256. In support of our holding, we emphasized that district courts have "very broad discretion in fashioning discovery orders under Fed. R. Civ. P. 26(c), and the protective order did not fall clearly outside the bounds of that authority." Id. Therefore, we held that the district court did not commit clear error in denying the motion to reconsider the protective order.

Applying McDowell II, we hold in this case that the district court did not commit clear error in denying the motion to reconsider the protective order. While a petitioner in [*17] a habeas corpus action who raises a Sixth Amendment claim of ineffective assistance of counsel waives the attorney-client privilege as to the matters challenged, see Wharton, 127 F.3d at 1203, McDowell IImakes clear that it is within the discretion of the district court to issue an order limiting that waiver to the habeas proceeding in which the ineffective assistance question is raised.

The State argues that the protective order may unduly limit its ability to use the attorney-client-privileged materials in the event of a retrial in state court, and that the state rather than the federal court should determine the scope of the waiver on retrial. It is true that the order is designed to ensure that the prosecution on retrial will not use the discovery permitted in the habeas proceeding to circumvent the more limited discovery available in criminal prosecutions. See McSurely v. McClellan, 138 U.S. App. D.C. 187, 426 F.2d 664, 671-722 (D.C. Cir. 1970) ("Civil discovery may not be used to subvert limitations on discovery in criminal cases, either by the government or by private parties."); Campbell v. Eastland, 307 F.2d 478, 487 (5th Cir. 1962) [*18] ("A litigant should not be allowed to make use of the liberal discovery procedures applicable to a civil suit as a dodge to avoid the restrictions on criminal discovery and thereby obtain documents he would not otherwise be entitled to for use in his criminal suit."). It is also true that if Osband is granted a new trial because of the ineffectiveness of his trial counsel, the order will limit the ability of the State to use against him material that would have remained undisclosed but for the violation of his constitutional rights by his earlier counsel.

In the posture of this case, however, the question is not whether the federal district court on habeas was correct in entering a protective order limiting the use of attorney-client material on retrial. The question before us is narrower. We are asked only to decide whether it was clear error for the district court to deny the motion for reconsideration of such a protective order. The en banc panel in McDowell IIprovided the answer when presented with a virtually identical order: "The question being a debatable one, the district court did not commit clear error when it limited access . . . pursuant to the terms [*19] of the protective order." 197 F.3d at 1256.

The State finds support for its argument in Anderson v. Calderon, 232 F.3d 1053 (9th Cir. 2000). Anderson, decided by a three-judge panel after our en banc decision in McDowell II, wrote that a district court correctly denied a capital habeas petitioner's request for a protective order comparable to the orders entered in McDowelland in this case. Id. at 1099-1100 . The Andersonpanel set forth, in language that tracks verbatim the language of the vacated panel opinion in McDowell I, its disapproval of the protective order sought by Anderson as "an unwarranted anticipatory interference of the state courts" that would "contravene basic principles of comity and federalism." Compare id. at 1099-1100 with McDowell I, 173 F.3d at 1191.

We refuse to accord to Andersonthe weight the State would give it. To the extent Andersonis inconsistent with the statement of Ninth Circuit law set forth in McDowell II, we, of course, are bound by the en banc decision. "An appellate panel simply cannot modify an En banc decision." Ewing v. Williams, 596 F.2d 391, 397 (9th Cir. 1979); [*20] see also Ponderosa Dairy v. Lyons, 259 F.3d 1148, 1155 (9th Cir. 2001) ("Only an en banc panel may overturn existing Ninth Circuit precedent.") (quoting Jeffries v. Wood, 114 F.3d 1484, 1492 (9th Cir. 1997)). McDowell IIplainly holds that protective orders like the one issued in this case do not fall outside the bounds of the very broad discretion of the district courts, and that it is not clear error to deny a motion to reconsider such an order.

In the Matter of Pautler, 2002 Colo. LEXIS 368 (Co 05/13/2002) Censure of prosecutor in a capital prosecution upheld when he imitated a public defender in order to negotiate the surrender of the accused.

We are unpersuaded by Pautler's assertion that his deception of Neal was "justified" under the circumstances, and we underscore the rationale set forth in People v. Reichman, 819 P.2d 1035 (Colo. 1991). There, a district attorney sought to bolster a police agent's undercover identity by faking the agent's arrest and then filing false charges against him. Id. at 1036. The DA failed to notify the court of the scheme. Id. We upheld a hearing board's imposition of public censure for the DA's participation in the ploy. Id. at 1039. n5

To support our holding in Reichman, we cited In re Friedman, 76 Ill. 2d 392, 392 N.E.2d 1333, 30 Ill. Dec. 288 (Ill. 1979). There, a prosecutor instructed two police officers to testify falsely in court in an attempt to collar attorneys involved in bribery. Friedman, 392 N.E.2d at 1334. [*16] A divided Illinois Supreme Court found such advice violated the ethics code despite the undeniably wholesome motive. Id. at 1336. Similarly, in In re Malone, 105 A.D.2d 455, 480 N.Y.S.2d 603 (N.Y. App. Div. 1984), a state attorney instructed a corrections officer, who was an informant in allegations against correctional officers abusing inmates, to lie to an investigative panel. Id. at 604-05. The instruction was purportedly to save the testifying officer from retribution by the other corrections officers. Id. Again, despite the laudable motive, the New York court upheld Malone's censure for breaking the code. Id. at 607-08.

Thus, in Reichman, we rejected the same defense to Rule 8.4(c) that Pautler asserts here. We ruled that even a noble motive does not warrant departure from the Rules of Professional Conduct. Moreover, we applied the prohibition against deception a fortiori to prosecutors:

District attorneys in Colorado owe a very high duty to the public because they are governmental officials holding constitutionally created offices. This court has spoken out strongly against misconduct by public officials who are lawyers. The respondent's [*17] responsibility to enforce the laws in his judicial district grants him no license to ignore those laws or the Code of Professional Responsibility.

Reichman, 819 P.2d at 1038-39 (citations omitted).

We stress, however, that the reasons behind Pautler's conduct are not inconsequential. In Reichman, we also stated, "While the respondent's motives and the erroneous belief of other public prosecutors that the respondent's conduct was ethical do not excuse these violations of the Code of Professional Responsibility, they are mitigating factors to be taken into account in assessing the appropriate discipline." Id. at 1039. Hence, Reichman unambiguously directs that prosecutors cannot involve themselves in deception, even with selfless motives, lest they run afoul of Rule 8.4(c).

B. Imminent Public Harm Exception

Pautler requests this court to craft an exception to the Rules for situations constituting a threat of "imminent public harm." In his defense, Pautler elicited the testimony of an elected district attorney from a metropolitan jurisdiction. The attorney testified that during one particularly difficult circumstance, a kidnapper [*18] had a gun to the head of a hostage. The DA allowed the kidnapper to hear over the telephone that the DA would not prosecute if the kidnapper released the hostage. The DA, along with everyone else involved, knew the DA's representation was false and that the DA fully intended to prosecute the kidnapper. Pautler analogizes his deceptive conduct to that of the DA in the hostage case and suggests that both cases give cause for an exception to Rule 8.4(c).

We first note that no complaint reached this court alleging that the DA in the kidnapper scenario violated Rule 8.4(c), and therefore, this court made no decision condoning that DA's behavior. But assuming arguendo that the DA acted in conformity with the Rules, one essential fact distinguishes the hostage scenario from Pautler's case: the DA there had no immediately feasible alternative. If the DA did not immediately state that he would not prosecute, the hostage might die. In contrast, here Neal was in the midst of negotiating his surrender to authorities. Neal did make references to his continued ability to kill, which Pautler described as threats, but nothing indicated that any specific person's safety was in imminent danger. More [*19] importantly, without second guessing crime scene tactics, we do not believe Pautler's choices were so limited. Pautler had several choices. He had telephone numbers and a telephone and could have called a PD. Indeed, he attempted to contact attorney Plattner, an indication that communicating with a defense attorney was not precluded by the circumstances. Pautler also had the option of exploring with Neal the possibility that no attorney would be called until after he surrendered. While we do not opine, in hindsight, as to which option was best, we are adamant that when presented with choices, at least one of which conforms to the Rules, an attorney must not select an option that involves deceit or misrepresentation. n6 The level of ethical standards to which our profession holds all attorneys, especially prosecutors, leaves no room for deceiving Neal in this manner. Pautler cannot compromise his integrity, and that of our profession, irrespective of the cause. C. Duress and Choice of Evils

Pautler further argues that the traditional defenses of duress n7 and "choice of evils" n8 provide examples of appropriate defenses to allegations of ethical misconduct. He also refers the court to the comment after Rule 4.1 where attorneys permissibly "misrepresent" their client's position as part of "generally accepted conventions in negotiations." Colo. RPC 4.1 cmt. Pautler does not assert that any of these exceptions apply in his case, but that their existence demonstrates that exceptions are, at times, available to the otherwise strictly interpreted ethics rules.

This court has never examined whether duress or choice of evils can serve as defenses to attorney misconduct. n9 We note that the facts here do not approach those necessary for either defense: Pautler was not acting at the direction of another person who threatened harm (duress), nor did he engage in criminal conduct to avoid imminent public injury (choice of evils).

A review board in Illinois examined a similar scenario and decided against such an exception. In re Chancey, No. 91 CH348, 1994 WL 929289, [*22] at *7 (Ill. Att'y Reg. Disp. Comm'n Apr. 21, 1994). In Chancey, a prosecutor with an impeccable reputation drafted a false appellate court order for the sole purpose of deceiving a dangerous felon who had abducted his own child and taken her abroad. Id. at **2 -4. Chancey signed a retired judge's name to the order. Id. at *3. He never intended to file the order and did not file the order, nor was the order ultimately used to deceive the felon. Id. Despite its non-use, and despite Chancey's undeniably worthy motive, the Illinois board reprimanded Chancey for his deceit. Id. at *7. Rather than consider an exception in light of valid concerns over the safety of an abducted child, the board insisted on holding attorneys, especially prosecutors, to the letter of the Rules. Further, the board observed, and we agree, that motive evidence was only relevant in the punishment phase, as either a mitigating or aggravating factor. Id.

Nor does the commentary to Colo. RPC 4.1 persuade us that an exception to Colo. RPC 8.4(c) is appropriate. If anything, the fact that the commentary to Rule 4.1 made explicit an already acknowledged exception demonstrates that, where [*23] applicable, the Rules and commentary set forth their own exceptions. Neither Colo. RPC 8.4(c), nor its comment, contain any such exception. On a related point, the hearing board noted, "Both of the rules under which Pautler was charged are imperative, not permissive in application. Compliance with their mandatory provisions is required and is not subject to the exercise of discretion by the lawyer."

D. Role of Peace Officer

Finally, Pautler contends that this court has never addressed whether district attorneys, "while functioning as peace officers," may employ deception to apprehend suspects. He suggests that because peace officers may employ lethal force when pursuing a fleeing, dangerous felon, it would be absurd to sanction an officer who instead uses artifice, simply because that officer is also a licensed attorney. We disagree.

The Rules of Professional Conduct apply to anyone licensed to practice law in Colorado. See In re C de Baca, 11 P.3d 426, 429-30 (Colo. 2000) (ruling that lawyers must adhere to the Rules of Professional Conduct even when suspended from the practice of law). The Rules speak to the "role" of attorneys in society; however, we [*24] do not understand such language as permitting attorneys to move in and out of ethical obligations according to their daily activities. Pautler cites Higgs v. District Court, 713 P.2d 840 (Colo. 1985), for the proposition that this court has provided a test for distinguishing when prosecutors act as "advocates" and when they act as "investigators," for purposes of governmental immunity. Id. at 853. Such test exists, but we hold here that in either role, the Rules of Professional Conduct apply. The obligations concomitant with a license to practice law trump obligations concomitant with a lawyer's other duties, even apprehending criminals. Moreover, this case does not confront us with the propriety of an attorney using deceit instead of lethal force to halt a fleeing felon. We limit our holding to the facts before us. Until a sufficiently compelling scenario presents itself and convinces us our interpretation of Colo. RPC 8.4(c) is too rigid, we stand resolute against any suggestion that licensed attorneys in our state may deceive or lie or misrepresent, regardless of their reasons for doing so.

The complaint also charges Pautler with violating [*25] Rule 4.3:

In dealing on behalf of a client with a person who is not represented by counsel, a lawyer shall state that the lawyer is representing a client and shall not state or imply that the lawyer is disinterested. When the lawyer knows or reasonably should know that the unrepresented person misunderstands the lawyer's role in the matter, the lawyer shall make reasonable efforts to correct the misunderstanding. The lawyer shall not give advice to the unrepresented person other than to secure counsel.

Colo. RPC 4.3. This rule targets precisely the conduct in which Pautler engaged. At all times relevant, Pautler represented the People of the State of Colorado. n10 The parties stipulated that Neal was an unrepresented person. Pautler deceived Neal and then took no steps to correct the misunderstanding either at the time of arrest or in the days following. Pautler's failure in this respect was an opportunity lost. Where he could have tempered the negative consequences resulting from the deception, he instead allowed them to linger.

While it is unclear whether Pautler actually gave advice to Neal, he certainly did not inform Neal to retain counsel. In addition, Pautler went further than implying he was disinterested; he purported to represent Neal. Without doubt, Pautler's conduct violated the letter of Colo. RPC 4.3.

For reasons substantially similar to those above, we refuse to graft an exception to this rule that would justify or excuse Pautler's actions. Instead, we affirm the ruling of the hearing board finding a violation of Colo. RPC 4.3 and turn now to consider the sanction imposed.

The hearing board suspended Pautler for three months and then stayed that suspension during twelve months of probation. During the probationary period, Pautler was to fulfill various conditions including retaking the MPRE. We review this sanction under a reasonableness standard. C.R.C.P 251.27(b).

The board rendered its decision after reviewing the ABA Standards for Imposing Lawyer Sanctions (1991 & Supp. 1992) (ABA Standards). Those standards require examination of the duty violated; the lawyer's mental state; the potential or actual injury caused by the lawyer's misconduct; and the existence of aggravating [*27] or mitigating factors. ABA Standards 3.0.

The board found that Pautler violated duties to the legal system, the profession, and the public. It also ruled that his mental state was "not only knowing, it was intentional." Further, the board found actual injury to the administration of justice in that Pautler's conduct "contributed to a perceived lack of trust between Neal and his lawyers, adversely impacted subsequent judicial proceedings and resulted in additional hearings to explore factual and legal issues created by the deceptive conduct." The board ruled the harm was perhaps unquantifiable, but certainly present. The board also found substantial "potential injury" because, had Neal discovered Pautler's deception, the "negotiating gains made by Sheriff Moore might be lost, Neal could terminate communication and resume or escalate his murderous crime spree." The board also considered the implications of whether Pautler actually became Neal's lawyer.

Addressing mitigating factors, the board acknowledged Pautler's praiseworthy motive, but also found a "secondary" motive: to keep Neal "talking about his crimes without the benefit of requested legal representation and thereby gain [*28] an advantage in subsequent legal proceedings." Other mitigating factors included Pautler's full cooperation with the Office of Attorney Regulation, see ABA Standards 9.32(e), and his lack of prior discipline, see id. at 9.32(a). Among the aggravators, the board found Pautler's substantial experience with the law, see id. at 9.22(i), and, most importantly, his lack of remorse, see id. at 9.22(g). While the board ultimately ruled that the mitigating factors outweighed the aggravating factors, they declined to depart from the presumptive sanction of suspension.

We conclude that the hearing board's discipline was reasonable. Pautler violated a duty he owed the public, the legal system, and the profession. His role of prosecutor makes him an instrument of the legal system, a representative of the system of justice. The fact that he lied for what he thought was a good reason does not obscure the fact that he lied -in an important circumstance and about important facts. To the extent Pautler's misconduct perpetuates the public's misperception of our profession, he breached public and professional trust. See generally ABA Standards 5.0-7.0.

Second, the record [*29] supports the board's finding that Pautler acted intentionally. He intended to deceive Neal into believing not only that the attorney on the telephone was a PD, but that the attorney represented him. Because Pautler's conscious objective was to accomplish the result, his mental state was intentional. See ABA Standards definitions.

Third, we agree that the evidence before the hearing board supported the finding of actual, unquantifiable harm. We do not agree, however, that the evidence also supported a finding of potential harm. n11

As to the aggravating factors, we do not find adequate support in the record for the board's finding that Pautler harbored a secondary, ulterior motive. While it is undoubtedly true that Pautler sought to keep Neal on the telephone until he surrendered, no evidence suggested he did so in an effort to gain a tactical advantage in subsequent criminal proceedings. Pautler never attempted to elicit incriminating statements from Neal. Indeed, Neal had already confessed to the crimes in substantial detail, both over the telephone and in the taped confession he left at the Belleview apartment; there was little need for additional evidence. For purposes of aggravation and mitigation, we conclude that Pautler's only motive was Neal's surrender to law enforcement.

However, we do find an additional aggravating circumstance: Pautler's post-incident conduct. An attorney's post-incident conduct also bears upon aggravation and mitigation. See ABA Standards 9.22(j) (indifference in making restitution is an aggravating factor); id. at 9.32(d) (timely good-faith effort to make restitution or to rectify consequences of misconduct is a mitigating factor). After the immediacy of the events [*31] waned, Pautler should have taken steps to correct the blatant deception in which he took part. Instead, he dismissed such responsibility believing that the PD's office "would find that out in discovery." Although we do not agree that Pautler's subsequent failure to correct the deception was evidence of a secondary, ulterior motive, as the hearing board found, we do find that such conduct was an independent aggravating factor.

In mitigation, we credit Pautler's commendable reputation in the legal community, his lack of prior misconduct, and his full cooperation in all these proceedings. In addition, we believe Pautler's motivation to deceive Neal was in no way selfish or self-serving. He believed he was protecting the public.

In light of the various factors bearing on Pautler's discipline, we do not find the hearing board's sanction unreasonable. See C.R.C.P. 251.27(b). Other attorneys participating in deceit and misrepresentation have received suspensions. See, e.g., In the Matter of Gibson, 991 P.2d 277, 279 (Colo. 1999) (ordering thirty-day suspension when attorney deceived client to hide the fact that the attorney had neglected his client's tort claim); [*32] People v. Casey, 948 P.2d 1014, 1015 (Colo. 1997) (affirming forty-five-day suspension when an attorney "represented" a teenager in criminal charges knowing the teen was using an assumed name).

In sum, we agree with the hearing board that deceitful conduct done knowingly or intentionally typically warrants suspension, or even disbarment. See ABA Standards 7.2 ("Suspension is generally appropriate when a lawyer knowingly engages in conduct that is a violation of a duty owed to the profession . . . ."); id. at 5.11(b) ("Disbarment is generally appropriate when . . . a lawyer engages in any other intentional conduct involving dishonesty, fraud, deceit, or misrepresentation. . . ."). We further agree that the mitigating factors present in Pautler's case outweigh the aggravating factors, and affirm the imposition of a three-month suspension, which shall be stayed during twelve months of probation. This sanction reaffirms for all attorneys, as well as the public, that purposeful deception by lawyers is unethical and will not go unpunished. At the same time, it acknowledges Pautler's character and motive.

SUPREME COURT

United States v. Cotton etal., --- US --- (5/20/2002) Failure to object may waive putative "Apprendi" error.

A federal grand jury returned an indictment charging respondents with conspiracy to distribute and to possess with intent to distribute a detectable amount of cocaine and cocaine base. Respondents were convicted and received a sentence based on the District Courts finding of drug quantityat least 50 grams of cocaine basethat implicated the enhanced penalties of 21 U.S.C. 841(b). They did not object in the District Court to the fact that the sentences were based on a quantity not alleged in the indictment. While their appeal was pending, this Court decided, in Apprendi v. New Jersey, 530 U.S. 466, 490, that [o]ther than the fact of a prior conviction, any fact that increases the penalty for a crime beyond the prescribed statutory maximum must be submitted to a jury, and proved beyond a reasonable doubt. In federal prosecutions, such facts must also be charged in the indictment. Id., at 476. Respondents then argued in the Fourth Circuit that their sentences were invalid under Apprendi, because the drug quantity issue was neither alleged in the indictment nor submitted to the petit jury. That court vacated the sentences on the ground that it had no jurisdiction to impose a sentence for an offense not charged in the indictment.

Held:

1. A defective indictment does not deprive a court of jurisdiction. Ex parte Bain, 121 U.S. 1, the progenitor of the Fourth Circuits view that the indictment errors are jurisdictional, is a product of an era in which this Courts authority to review criminal convictions was greatly circumscribed. It could examine constitutional errors in a criminal trial only on a writ of habeas corpus, and only then if it deemed the error jurisdictional. The Courts desire to correct obvious constitutional violations led to a somewhat expansive notion of jurisdiction, Custis v. United States, 511 U.S. 485, 494, which is not what the term means today, i.e., the courts statutory or constitutional power to adjudicate the case, Steel Co. v. Citizens for Better Environment, 523 U.S. 83, 89. Because subject-matter jurisdiction involves a courts power to hear a case, it can never be forfeited or waived. Thus, defects require correction regardless of whether the error was raised in district court. But a grand jury right can be waived. Post-Bain cases confirm that indictment defects do not deprive a court of its power to adjudicate a case. See, e.g., Lamar v. United States, 240 U.S. 60. Thus, this Court some time ago departed from Bains view that indictment defects are jurisdictional. Stirone v. United States, 361 U.S. 212; Russell v. United States, 369 U.S. 749, distinguished. Insofar as it held that a defective indictment deprives a court of jurisdiction, Bain is overruled. Pp. 35.
2. The omission from a federal indictment of a fact that enhances the statutory maximum sentence does not justify a court of appeals vacating the enhanced sentence, even though the defendant did not object in the trial court. Under Federal Rule of Criminal Procedure 52(b)s plain-error test, where there is an (1) error, (2) that is plain, and (3) that affects substantial rights, an appellate court may correct an error not raised at trial, but only if (4) the error seriously affects the fairness, integrity, or public reputation of judicial proceedings. Johnson v. United States, 520 U.S. 461, 466467 (internal quotation marks omitted). The Government concedes that the indictments failure to allege a fact that increased the sentences was plain error. But, even assuming the error affected respondents substantial rights, it did not seriously affect the fairness, integrity, or public reputation of judicial proceedings. The evidence that the conspiracy involved at least 50 grams of cocaine base was overwhelming and essentially uncontroverted. It is true that the Fifth Amendment grand jury right serves a vital function in providing for a body of citizens that acts as a check on prosecutorial power, but that is no less true of the Sixth Amendment right to a petit jury, which must find guilt beyond a reasonable doubt. The petit jurys important role did not, however, prevent the Johnson Court from applying the longstanding rule that a constitutional right may be forfeited in criminal as well as civil cases by the failure to make timely assertion of the right. Yakus v. United States, 321 U.S. 414, 444. The real threat to the fairness, integrity, or public reputation of judicial proceedings would be if respondents, despite the overwhelming and uncontroverted evidence that they were involved in a vast drug conspiracy, were to receive a sentence prescribed for those committing less substantial drug offenses because of an error that was never objected to at trial. Pp. 69.

261 F. 3d 397, reversed and remanded.

Rehnquist, C.J., delivered the opinion for a unanimous Court.

Alabama v. Shelton, --- US --- (5/20/2002) Right to counsel exists even where the sentence imposed is a suspended sentence.

Defendant-respondent Shelton represented himself in an Alabama Circuit Court criminal trial. The court repeatedly warned Shelton about the problems self-representation entailed, but at no time offered him assistance of counsel at state expense. He was convicted of misdemeanor assault and sentenced to a 30-day jail term, which the trial court immediately suspended, placing Shelton on two years unsupervised probation. The Alabama Supreme Court reversed Sheltons suspended jail sentence, reasoning that this Courts decisions in Argersinger v. Hamlin, 407 U.S. 25, and Scott v. Illinois, 440 U.S. 367, require provision of counsel in any petty offense, misdemeanor, or felony prosecution, Argersinger, 407 U.S., at 37, that actually leads to imprisonment even for a brief period, id., at 33. The State Supreme Court concluded, inter alia, that because a defendant may not be imprisoned absent provision of counsel, Sheltons suspended sentence could never be activated and was therefore invalid.

Held:A suspended sentence that may end up in the actual deprivation of a persons liberty may not be imposed unless the defendant was accorded the guiding hand of counsel in the prosecution for the crime charged. Argersinger, 407 U.S., at 40. Pp.419.

(a)The controlling rule is that absent a knowing and intelligent waiver, no person may be imprisoned for any offense unless he was represented by counsel at his trial. Argersinger, 407 U.S., at 37. Pp.56.

(b)Applying this actual imprisonment rule, the Court rejects the argument of its invited amicus curiae that failure to appoint counsel to an indigent defendant does not bar the imposition of a suspended or probationary sentence upon conviction of a misdemeanor, even though the defendant might be incarcerated in the event probation is revoked. Pp.616.

1. The Sixth Amendment does not permit activation of a suspended sentence upon an indigent defendants violation of the terms of his probation where the State did not provide him counsel during the prosecution of the offense for which he is imprisoned. A suspended sentence is a prison term imposed for the offense of conviction. Once the prison term is triggered, the defendant is incarcerated not for the probation violation, but for the underlying offense. The uncounseled conviction at that point result[s] in imprisonment, Nichols v. United States, 511 U.S. 738, 746; it end[s] up in the actual deprivation of a persons liberty, Argersinger, 407 U.S., at 40. This is precisely what the Sixth Amendment, as interpreted in Argersinger and Scott, does not allow. P.6.
2. The Court rejects the first of two grounds on which amicus resists this reasoning, i.e., amicus attempt to align this case with Nichols and with Gagnon v. Scarpelli, 411 U.S. 778. Those decisions do not stand for the broad proposition that sequential proceedings must be analyzed separately for Sixth Amendment purposes, with the right to state-appointed counsel triggered only in proceedings that result in immediate actual imprisonment. The dispositive factor in Gagnon and Nichols was not whether incarceration occurred immediately or only after some delay. Rather, the critical point was that the defendant had a recognized right to counsel when adjudicated guilty of the felony for which he was imprisoned. See Nichols, 511 U.S., at 743, n.9. Here, revocation of probation would trigger a prison term imposed for a misdemeanor of which Shelton was found guilty without the aid of counsel, not for a felony conviction for which the right to counsel is unquestioned. See id., at 747; Gagnon, 411 U.S, at 789. Far from supporting amicus position, Gagnon and Nichols simply highlight that the Sixth Amendment inquiry trains on the stage of the proceedings corresponding to Sheltons Circuit Court trial, where his guilt was adjudicated, eligibility for imprisonment established, and prison sentence determined. Nichols is further distinguishable because the Court there applied a less exacting standard allowing a trial court, once guilt has been established, to increase the defendants sentence based simply on evidence of the underlying conduct that gave rise to his previous conviction, 511 U.S., at 748, even if he had never been charged with that conduct, Williams v. New York, 337 U.S. 241, and even if he had been acquitted of a misdemeanor with the aid of appointed counsel, United States v. Watts, 519 U.S. 148, 157. That relaxed standard has no application here, where the question is whether the defendant may be jailed absent a conviction credited as reliable because the defendant had access to counsel. Pp.69.
3. Also unpersuasive is amicus contention that practical considerations weigh against extension of the Sixth Amendment appointed-counsel right to a defendant in Sheltons situation. Based on figures suggesting that conditional sentences are commonly imposed but rarely activated, amicus argues that the appropriate rule would permit imposition of a suspended sentence on an uncounseled defendant and require appointment of counsel, if at all, only at the probation revocation stage, when incarceration is imminent. That regime would unduly reduce the Sixth Amendments domain. In Alabama, the probation revocation hearing is an informal proceeding, at which the defendant has no right to counsel, and the court no obligation to observe customary rules of evidence. More significant, the defendant may not challenge the validity or reliability of the underlying conviction. A hearing so timed and structured cannot compensate for the absence of trial counsel and thereby bring Sheltons sentence within constitutional bounds. Nor does this Court agree with amicus that its holding will substantially limit the States ability to impose probation. Most jurisdictions already provide a state-law right to appointed counsel more generous than that afforded by the Federal Constitution, while simultaneously preserving the option of probationary punishment. See id., at 748749, n.12. Even if amicus is correct that some States cannot afford the costs of the Courts rule, those jurisdictions have recourse to the option of pretrial probation, whereby the prosecutor and defendant agree to the defendants participation in a pretrial rehabilitation program, which includes conditions typical of post-trial probation, and the adjudication of guilt and imposition of sentence for the underlying offense occur only if the defendant breaches those conditions. This system reserves the appointed-counsel requirement for the few cases in which incarceration proves necessary, see Gagnon, 411 U.S., at 784, while respecting the constitutional imperative that no person be imprisoned unless he was represented by counsel, Argersinger, 407 U.S., at 37. Pp.916.

(c)The Court does not rule on Alabamas argument that, although the Sixth Amendment bars activation of a suspended sentence for an uncounseled conviction, the Constitution does not prohibit, as a method of effectuating probationary punishment, the imposition of a suspended sentence that can never be enforced. There is not so much as a hint in the Alabama Supreme Courts decision that Sheltons probation term is separable from the prison term to which it was tethered. Absent any prior presentation of the novel position the State now takes, this Court resists passing on it in the first instance. It is for the State Supreme Court to consider before this Court does whether the suspended sentence alone is invalid, leaving Sheltons probation term freestanding and independently effective. See Hortonville Joint School Dist. No. 1 v. Hortonville Ed. Assn., 426 U.S. 482, 488. Pp.1618.

Affirmed.

CAPITAL CASES (Favorable Disposition)

See above

CAPITAL CASES (Unfavorable Disposition)

Tennessee v. Stevens, 2002 Tenn. LEXIS 223 (Tenn 5/14/2002) On direct appeal the state supreme court holds: "(1) the trial court did not abuse its discretion in limiting the testimony of defendant's crime scene expert to his analysis of the evidence at the crime scene; (2) the trial court's exclusion of the testimony of Corey Milliken's foster father regarding Milliken's prior bad acts constituted harmless error; (3) the trial court applied hearsay and other evidentiary rulings in an unbiased and even-handed manner; and (4) the sentence of death is not disproportionate to the sentence imposed in similar cases. For all other issues not specifically discussed in this opinion, we agree with and affirm the judgment of the Court of Criminal Appeals."

Smithers v. Florida, 2002 Fla. LEXIS 886 (FL 05/16/2002) Relief denied on the following claims: "(1) the trial court erred by denying Smithers' motion to sever the two offenses; (2) the trial court erred by denying Smithers' motion to suppress his confession; (3) fundamental error occurred when defense counsel waived Smithers' presence for the pretrial motion in limine hearing; (4) the trial court erred in finding HAC for the Roach murder; (5) the trial court erred in finding CCP for the Cowan murder; and (6) the trial court erred by failing to declare a mistrial during the penalty phase when one of the State's witnesses introduced lack of remorse as a consideration."

Louisiana v. Carmouche, 2002 La. LEXIS 1520 (LA 05/15/2002) "The principal issues in this appeal involve: (1) whether the trial court should have granted the defense's challenge for cause regarding a particular prospective juror; (2) whether the trial court should have granted the defense's motion for a mistrial or to strike the entire venire due to a different prospective juror's mention of a "sanity test" taken by the defendant; (3) whether the trial court should have granted the defense's motion for a mistrial or to strike certain jurors who asked why the defendant was not wearing handcuffs; and (4) whether the trial court should have appointed a sanity commission during trial."

State v. Anderson, No. SC83680 (Mo. 05/14/2002) On direct appeal the state supreme court holds, most notably, that: (1) sufficient African-American representation of the jury pool existed; (2) Anderson was competent to stand trial; (trial court did not abuse its discretion in striking, for cause, two venire persons who indicated they could not consider the death penalty and then later gave contradictory responses; (4) Anderson plausibly cannot claim that he was prejudiced significantly by the prosecutor's use of billing records regarding the amount of money the defense expert made from her work for the public defender's officE; (5) Anderson failed to object, at trial, to certain instances during the prosecutor's closing arguments in the guilt and penalty phases and cannot now claims should have resulted in a mistrial; (6) the trial court did not err in failing to grant a mistrial on its own motion after two officers testified that Anderson did not make statements to them about the murders because they did not question him about the murders; (7) The evidence supports all three statutory aggravating factors the jury found regarding Anderson's killing of Debbie Rainwater. These factors were not duplicative, and even if they were, no prejudice resulted because Anderson does not challenge one of the factors, and only one valid statutory aggravating factor was needed for the jury to consider imposing the death penalty; (8) the trial court did not commit plain error in sustaining the state's objections to some of the defense's questions during voir dire, particularly where the defense was allowed to rephrase these questions; and (9) there is no merit in Anderson's claims that the state's decision to seek and impose the death penalty in his case was based on the fact that he was an African-American man who killed a Caucasian woman.

Ohio v Yarbrough, 95 Ohio St. 3d 227; 2002 Ohio 2126; 2002 Ohio LEXIS 1133 (OH 05/15/2002) On direct appeal a litany of a claims are denied including, mot notably: [1] admission of three statements that putatively violated hearsay rule; [2] certain out-of-court statements incriminating others should have beenadmittedas statements against interest; [3] sufficiency of the evidence; [4] failure to strike life hesitant jurors; [5] method of empaneling venire systematically under-represented African-Americans; [6] prosecutorial misconduct (notably certain comments) tainted both phases of trial; [7] that the two death specifications arose from "one indivisible act"; [8] failing to consider mitigating factors and improperly considered nonstatutory aggravating circumstances; [9] ineffectiveness relating to failure to ask certain questions in voir dire, failure to impeach with prior inconsistent statements, and failure ot objections, and [10] failure to request an instruction defining "for hire," "witness," and "criminal proceeding," appellant fails to show prejudice.

OTHER NOTABLE CASES

Aron v. United States, No. 99-14518 (11th Cir. 05/13/2002) Conclusion that the district court's dismissal of Aron's 28 U.S.C. § 2255 motion as untimely is due to be reversed and the case remanded for an evidentiary hearing on whether Aron exercised due diligence in discovering the facts relating to his claims of ineffective assistance of appellate counsel.

AMENDED OR DELAYED PUBLICATION CASES

No cases noted.

FOCUS

With the recent publication of the Illinois Commision on Capital Punishment it seems somehow appropriate to turn to one of the seminal works on errors in capital cases Samuel R. Gross, The Risks of Death: Why Erroneous Convictions are Common in Capital Cases, 44 BFLR 469 (1996). Mr. Gross's article is oft cited in the current debate on how to improve the death penalty and reduce the chances of executing the innocent and worthy of a quick review as part of the actual nuts and bolts as to how the system is failing.

III. THE PRODUCTION OF ERRORS

A. Investigation

1. Clearance Rates and Pressure on the Police . We often talk of a miscarriage of justice as an error at trial, but that's a mistake. The error occurs much earlier, in the investigation of the crime, when the police identify the wrong person as the criminal. If *476 they gather enough evidence against this innocent suspect, the error will ripen into a criminal charge; if that charge survives the formal and informal processes of pre-trial screening, it will go to trial and a jury may confirm the mistake by a wrongful conviction. For the police, the issue is: What factors make mistakes more likely? For the prosecutor, the judge, and the jury the issue is: What factors reduce the likelihood that they will correct police errors?

Most crimes are never solved. In 1993, a mere 21% of all serious crimes known to the police were "cleared"-which usually means that a suspect was arrested; of serious violent crimes, 44% were cleared. [FN35] But even these low figures only tell half the story. Most crimes are not "known to the police"-in 1993, only 35% of all crimes, and 42% of crimes of violence, were reported. [FN36] In other words, only about 18% of all crimes of violence are solved by the police, including about 13% of robberies, 18% of rapes, and 6% of burglaries. [FN37]

On the whole, the crimes that are reported to the police have better evidence than those that are not reported. Cases with extremely strong evidence-those in which the culprit is caught in the act, or seen and identified by several people-are almost always reported. If the victim has to take the initiative to notify the police, he'll be more likely to do so if he thinks there's a good chance that the criminal will be caught. If the police do hear about a robbery, or a rape, or a burglary, for which the identity of the criminal is not immediately obvious, their investigation will usually be perfunctory: put out a call to other officers to try to spot the criminal in flight; interview the witnesses at the scene; collect immediately available physical evidence; that's it. If a suspect doesn't emerge from this process it is unlikely that the case will ever be prosecuted. Police detectives do not have the time to conduct detailed investigations of every reported felony, and in the usual run-of-the- mill case there is little pressure on them to do so. The net result is that in general the felonies that are prosecuted are likely to be those in which the evidence of guilt is strongest.

Homicides are different. First, almost every homicide is reported *477 to the police when the body of the deceased person is found. There are exceptions- cases in which the body is not discovered, and others in which the cause of death is mistaken for accident or suicide or natural causes-but they could not account for more than a few percent of the total. [FN38] Second, most homicides known to the police are cleared-66% in 1993, [FN39] more in previous years. [FN40] Overall, the proportion of all homicides that are solved is about four times higher than the comparable proportion for other violent crimes. [FN41] A study of robbery investigations in Chicago in 1982- 83, by Franklin Zimring and James Zuehl, provides an excellent illustration: 13% of all robberies reported to the police were solved within two months (including a somewhat lower proportion of robberies with injuries to the victims), compared to 57% of robbery killings . [FN42] This difference cannot be explained by superior evidence-on the contrary, robbery homicides will usually have weaker evidence, since the victim is dead-but must be due to a systematic difference in the investigation by the police.

Many homicides, perhaps most, are easy to investigate. Most homicides are committed by relatives, friends or acquaintances of the victim, and many if not most occur during arguments or brawls. [FN43] In a typical ordinary homicide-a killing by a friend as a result of a drunken fight-the killer is known from the start. But the police get the hard murders as well as the easy ones, and there is much more pressure to solve these cases than non-homicidal crimes. The relatives of the victim care more, the prosecutor cares *478 more, the public is much more likely to be concerned, and the police themselves care more. Death produces strong reactions-in this context, a desire to punish and to protect. Other outrageous crimes can have the same effect- kidnappings, for example, or serial rapes-but they are rare. Homicide is common.

For the most part, the pressure to solve homicides produces the intended results. The police spend more time, they are more persistent, they have more resources at their disposal, and they catch more of the criminals. An investigation that would be closed without arrest if it were a mere robbery, may end in a conviction if the robber killed one of his victims. But that same pressure can also produce mistakes. If the murder cannot be readily solved, the police may be tempted to cut corners, to jump to conclusions, and-if they believe they have the killer-perhaps to manufacture evidence to clinch the case. The danger that the investigators will go too far is magnified to the extent that the killing is brutal and horrifying, and to the extent that it attracts public attention-factors which also increase the likelihood that the murder will be treated as a capital case.

The murder of 10-year-old Jeanine Nicarico is a good example. [FN44] In February 1983 she was abducted from her home in Naperville, Illinois, raped and killed-a crime of stunning brutality. The murder was the subject of a long, frustrating, unsuccessful investigation-a humiliating public failure. Thirteen months after the murder-and less than two weeks before the local prosecutor stood for reelection-three men were indicted: Rolando Cruz, Alejandro Hernandez and Stephen Buckley. Cruz and Hernandez were convicted and sentenced to death; their convictions were reversed by the Illinois Supreme Court. They were convicted again, but this time only Cruz was sentenced to death. Again the convictions were reversed. Finally, at Cruz's third trial-over twelve years after the murder-the case fell apart when a police officer admitted he had lied under oath, and the judge entered a judgment of acquittal. What seems to have happened is this: under intense pressure, the police convinced themselves that they knew who killed Jeanine Nicarico, and they manufactured evidence to convince prosecutors and to use in court. If the criminal had taken jewelry from the Nicarico home rather than a child-or even if he had knocked out a family member or set the home on fire-there would probably have been a minimal investigation, no arrests, no trial, and no Search Term Begin erroneous convictions Search Term End .

*479 In some highly charged murders, the police manufacture a case out of whole cloth. When Ronda Morrison was murdered on November 1, 1986, in Monroeville, Alabama, there were no suspects, and an eight-month investigation turned up no leads. [FN45] Then the police arrested a man by the name of Ralph Myers in connection with a different killing in a nearby county, and pressured him into saying that he drove Walter McMillian-a local resident-to the scene of the crime, and saw him shoot Ms. Morrison. Myers initially denied that he knew McMillian, or anything whatever about the killing, but eventually he gave in and said what he was told to say. McMillian was convicted and sentenced to death; he spent six years on death row before the frame-up was exposed. It's easy to see the hand of racism in this case. Apparently McMillian was chosen for the role of killer because he was a black man in rural Alabama who was known to have carried on an extra-marital affair with a white woman. But the nature of the crime was also an essential ingredient. Even the most racist police would hardly go to all that trouble for anything less than a heinous crime, and they'd be most likely to do it for capital murder. [FN46]

2. Eyewitness Identification . Most miscarriages of justice for crimes other than homicide are caused by eyewitness misidentifications. In Rattner's sample of wrongful convictions, 52% of the errors for which the cause could be determined were caused by misidentifications, [FN47] and other researchers concur that eyewitness error is by far the most common cause of convictions of innocent defendants. [FN48] On the other hand, eyewitness error was a factor in *480 only 16% of Bedau and Radelet's cases of errors in potentially capital prosecutions. [FN49] This disparity may not prove that there is a systematic difference in the causes of errors between capital and other non-capital cases-as Bedau and Radelet point out, the samples are too unsystematic for firm conclusions [FN50] -but it certainly does suggest it. If we assume that among the murder cases in Rattner's sample-which make up 42% of the total-eyewitness misidentifications caused 17% of the errors (a slightly higher proportion than Bedau and Radelet report), that implies that over 80% of the errors in non-murder cases were due to misidentifications. [FN51]

No doubt the main reason for this difference is the absence of a live victim in most homicides. [FN52] Victims provide crucial identification evidence in most robberies and rapes, and so they make most of the mistakes, when mistakes are made. In the absence of a victim the police may have no eyewitness evidence, and therefore no room for eyewitness error. This is hardly an advantage for accuracy. Many, perhaps most eyewitness identifications of criminals by strangers are accurate. [FN53] Frequently they are corroborated or lead to other evidence that greatly reduces the likelihood of error- fingerprints, stolen property, reliable confessions, etc. [FN54] In addition, for about half of all violent crimes eyewitness identifications are extremely reliable because the crimes were committed by relatives, friends, or others who are known to the victims. [FN55] Murderers are even more likely to be known to their victims [FN56] but that may not always help because in the words of the immortal cliche, *481 "dead men don't talk." The absence of eyewitness evidence in many homicides drives the police to rely on evidence from other sources: accomplices; jail-house snitches and other underworld figures; and confessions from the defendants themselves. Not surprisingly, perjury by a prosecution witness is the leading cause of error in erroneous capital convictions, and false confessions are the third most common cause. [FN57]

So far I've compared homicides to other crimes of violence-crimes to which there are usually eyewitnesses. Other crimes, however, typically have no eyewitnesses-burglary, for example. Search Term Begin Erroneous convictions Search Term End may well occur in burglary cases, but if so they are rarely discovered and reported. Only one of a sample of 136 proven misidentifications was a burglary, [FN58] and Rattner does not even list the crime as a category for his sample of Search Term Begin erroneous convictions Search Term End . Eyewitness evidence may play a role in solving an occasional burglary-a homeowner may catch a glimpse of the fleeing criminal, and later identify his picture-and once in a while (in the absence of determinative physical evidence) this may lead to errors. The same thing may happen in those homicides in which the killer is not immediately known; indeed it is more likely, since killing is more often a crime of passion rather than of calculation. These are chance events, unforced errors. If the burglar is not easily identified, that's usually the end: hence a clearance rate for burglary of only 13% of crimes known to the police. [FN59] In homicide investigations the police work to solve the hard cases as well, and they often succeed. But in the process they force more errors-of the same sort, and other types-as they press on against increasing odds.

3. Perjury by the Real Killers, and by Others . The killer who blames his crime on others is a familiar character in fiction. The most famous, no doubt, is Macbeth, who planted bloody daggers on Duncan's grooms, and then killed them himself in a supposed fit of vengeful rage. The model for modern type of deceitful killer, however, is Mark Twain's Injun Joe, from The Adventures of Tom Sawyer, who murdered young Dr. Robinson and then pinned the killing on hapless drunk Muff Potter. [FN60] Similar things happen in real life. Some criminals implicate innocent defendants in order to divert suspicion from themselves. In other cases, false witnesses *482 who may have had no role in the crime lie for money or for other favors from the authorities. Both of these motives are more powerful in homicides than in other criminal cases, and especially in capital homicides.

First, the threat of being caught is much greater for a homicide than for almost any other crime. It's no news that the police work much harder to find killers than burglars or robbers, and that their interest increases in proportion to the brutality and notoriety of the crime.

Second , if the culprit is suspected and caught, he has more to fear in a capital case: he might get executed. The threat of death can be a powerful motivator, when it's concrete. The death penalty as an abstract prospect does not seem to deter many homicides. [FN61] Before the crime, the killer-if he thinks about it at all-no doubt expects to escape scot free; he's not likely to weigh the benefits of murder against the costs of the possible punishment. After the crime, however, there is more time to think, and the fear of conviction and execution may be vivid-especially if the police seem to be closing in.

Third , a perjurious killer may have to admit to crimes himself. He and the innocent defendant may in fact have been accomplices in some crime other than the murder, or he might have been caught in undeniably compromising circumstances, or he might have to admit to some level of guilt in order to make his accusation credible. If so, the real killer has more to gain in a capital case than under other circumstances. If he has to go to prison, the gain from cooperation is time vs. death, as opposed to less time vs. more time. But that may not be necessary: if he helps break a capital case, he may walk. For example, in December 1976 David Harris was arrested in Vidor, Texas, in connection with the murder of Dallas patrolman Robert Wood, and promptly blamed the killing on Randall Dale Adams, a passing acquaintance with whom he had spent most of the day leading up to the murder. [FN62] Harris admitted that he stole the car from which the patrolman was shot, was present at the murder, and kept the stolen car afterwards. [FN63] He was certainly liable for auto theft, and possibly as an accessory to murder. Instead, after Adams was convicted, all charges against Harris *483 were dropped. [FN64]

Fourth , if the witness is lying to get favors unrelated to the crime at issue, he'll do much better if it's a big case-which usually means a murder, or better yet, a capital murder. The typical witness in this category is the jail- house snitch. For example, in 1932 Gus Colin Langley was convicted of first degree murder in North Carolina based in part on testimony from his cellmate, who said that Langley had confessed to him. [FN65] Langley came within half an hour of electrocution, but was exonerated four years later and received a full pardon. His cellmate didn't have to wait that long; after his perjurious testimony, unrelated charges against him were dropped. [FN66]

Fifth , it's easier to lie about a capital case than most other crimes of violence: there's usually no live victim to contradict the false witness.

The overall result seems to be that witness perjury is a far more common cause of error in murders and other capital cases than in lesser crimes. Bedau and Radelet identified it as a factor in 35% of their erroneous capital convictions, [FN67] while Rattner lists perjury as the cause of only 11% of his errors. [FN68] But recall that 45% of Rattner's cases are murders. If perjury was as common among the murder convictions in Rattner's sample as among Bedau and Radelet's cases, then erroneous murder convictions could easily account for all the cases in which the error was caused by perjury.

The case of Paris Carriger is a good illustration of the role of perjury in capital prosecutions. [FN69] On March 14, 1978, Carriger was arrested for the brutal robbery murder of Robert Shaw, the owner of a jewelry store, on the previous day. The evidence against Carriger was provided by Robert Dunbar, a friend on whose property Carriger was living in a trailer. Dunbar-who had a great deal of experience as a police informant-called the police and said he *484 could identify Shaw's killer in return for immunity from prosecution for various felonies: another robbery he committed two days earlier, possession of a gun he had bought (which was illegal because he was a convicted felon), and attempting to dispose of the proceeds of the Shaw robbery-murder. The police agreed to these terms. Dunbar then told them that Carriger had come to him, confessed to the killing, and asked for help in disposing of bloody clothes and stolen jewelry; Dunbar corroborated the story by producing some of the loot, and leading the police to some of the clothes. Carriger was convicted and sentenced to death almost entirely on Dunbar's testimony. He steadfastly maintained his innocence, and claimed that Dunbar himself-a man with a long history of violence and deception-must have committed the murder. After the trial, Dunbar, who was soon jailed for other crimes, bragged that he had framed Carriger. In 1987 he confessed his own guilt to various people, including his parents and a clergyman. That same year he repeated his confession in court, and admitted that he had lied at Carriger's trial and that he had committed the murder himself; three weeks later he retracted that confession, but admitted that he was doing so for fear that he'd be prosecuted for the murder and executed himself. In 1991, shortly before he died in prison, Dunbar confessed again, to his cellmate. Dunbar's ex-wife, who had corroborated his original story and had given him an alibi, testified in 1987 that Dunbar had forced her to lie.

Despite these developments, Carriger remains on death row. He has come close to execution on several occasions while his attorneys have fought for a new trial. [FN70] Under the circumstances, this seems a modest goal, since, at a minimum, the new evidence that has turned up since the trial raises grave doubts about Carriger's guilt. But if Robert Shaw hadn't been killed, none of this would ever have happened in the first place. Dunbar would probably never have approached the police, they would hardly have given an ex-felon immunity from prosecution for three serious felonies in order to convict someone else of a single robbery, and the victim would have been available to contradict a false story.

4. Confessions . A typical robbery investigation is resolved by an eyewitness identification; a typical homicide investigation is resolved by a confession. Many confessions are straight-forward affairs-volunteered by suspects who are overcome by guilt, or believe they have nothing to lose. These are the easy cases, where *485 nothing has been done that might produce a false confession, and where more often than not there is strong corroborating evidence of guilt. Some confessions, however, are not so readily given, but are instead the end products of long, drawn out interrogations.

American police officers use all sorts of coercive and manipulative methods to obtain confessions. They confuse and disorient the suspect; they lie about physical evidence, about witnesses, about statements by other suspects; they pretend that they already have their case sealed and are only giving the suspect a chance to explain his side of the story; they pretend to understand, to sympathize, to excuse; they play on the suspect's fears, his biases, his loyalty to family and friends, his religion; they exhaust the suspect and wear him down; in some cases, they use violence, even torture. [FN71] These are powerful techniques. They work to get confessions from guilty defendants-and sometimes from innocent defendants as well.

From the point of view of the police, the main problem with interrogation is not that it occasionally produces errors, but that it's extremely time consuming. It's likely to take hours, perhaps days to break down a suspect who resists and insists on his innocence. Frequently several police officers cooperate in the effort, questioning the suspect simultaneously or in relays. As a result, extended interrogation is largely reserved for big cases in which confessions are necessary for successful prosecution. Typically, that means homicides, and especially the most heinous homicides, for reasons I've mentioned: these are the cases that the police are most anxious to solve, and yet, because the victim is dead, they frequently lack eyewitnesses. [FN72]

As with perjury, false confessions are a much more common cause of errors for homicides than for other crimes. They were a cause of 14% of Bedau and Radelet's errors in homicide and capital *486 cases, [FN73] but only 8% of the errors reported by Rattner. [FN74] Since 45% of Rattner's cases are homicides, this suggests that false confessions are three to four times more common as a cause of miscarriages of justice for homicide cases than for other crimes. [FN75]

The case of Melvin Reynolds is a good example, [FN76] but by no means unique. [FN77] On May 26, 1978, 4-year-old Eric Christgen disappeared in downtown St. Joseph, Missouri. His body later turned up along the Missouri river; he had been sexually abused and died of suffocation. The police questioned over a hundred possible suspects, including "every known pervert in town," to no avail. One of them was Melvin Reynolds, a 25-year-old man of limited intelligence who had been sexually abused himself as a child and who had some homosexual episodes as an adolescent. Reynolds, although extremely agitated by the investigation, cooperated through several interrogations over a period of months, including two polygraph examinations and one interrogation under hypnosis. In December 1978 he was questioned under sodium amytal ("truth serum") and made an ambiguous remark that intensified police suspicion. Two months later, in February 1979, the police brought the still cooperative Reynolds in for another round of interrogation-fourteen hours of questions, promises and threats. Finally, Reynolds gave in and said, "I'll say so if you want me to." In the weeks that followed, Reynolds embellished this concession with details that were fed to him, deliberately or otherwise. That was enough to convince the prosecutor to charge Reynolds, and to convince a jury to convict him of second degree murder. He was sentenced to life imprisonment. Four years later, Reynolds was released when another man-Charles Hatcher- confessed to three murders, including that of Eric Christgen. As the St. Louis Post-Dispatch put it in an editorial: "The Reynolds case says something about confessions and police methods used to extract them, even when these methods fall short of outright physical brutality. Under the stress of constant harassment, individuals can reach breaking *487 points."

B. Plea Bargaining and Dismissal

Most prosecutions are resolved without trial. Eighty to 90% of convictions result from guilty pleas, [FN78] usually after plea bargains, and at least 80% of defendants who are not convicted obtain pre-trial dismissals rather than acquittals. [FN79] In other words, most of the work of sorting criminal cases after arrest is done pre-trial, by the exercise of prosecutorial discretion to dismiss, to reduce charges, or to recommend or agree to a particular sentence. The direct impact of this pre-trial process is much more important to the accuracy of criminal dispositions than anything that happens later on. And yet, in the most important cases-murders, and especially capital murders-pre-trial screening may be distorted in ways that produce two distinct types of error.

1. Guilty Pleas by Innocent Defendants . Threat is an essential part of all plea bargaining: take the deal or you'll do worse after conviction. There is, undeniably, a coercive aspect to this bargain-the defendant must risk a severe penalty in order to exercise his right to trial-and plea bargaining has been strongly criticized on that ground. [FN80] Nonetheless the Supreme Court has held that coercion of that sort does not violate the due process clause. [FN81] One of the few qualifications on this general rule is a restriction on the form of death penalty statutes. While it is permissible for a prosecutor to ask for death and then let the defendant plead guilty in return for a lesser penalty-even if the defendant continues to deny that he is guilty [FN82] -a statute that permits the defendant to avoid the risk of execution by pleading guilty is unconstitutionally coercive. [FN83]

One attack on plea bargaining is that the threat is too effective, that it drives some innocent defendants to plead guilty along *488 with the mass of guilty ones. [FN84] That may happen with some regularity for innocent defendants who are offered light deals: time-served, diversion, 6-months unsupervised probation, and so forth. But among the more serious criminal convictions with severe penalties of imprisonment or death-those convictions that show up in cases of proven miscarriages of justice-the picture is different. I know of exactly one reported miscarriage of justice based on a guilty plea for a non-homicidal crime-and that was a peculiar case, a defendant who pled guilty to a crime he did not commit along with one which he did commit. [FN85] The available collections of known errors are hardly representative samples of the universe of Search Term Begin erroneous convictions Search Term End , and errors based on guilty pleas are probably less likely to be discovered than those based on trials. [FN86] Even so, this is a stark contrast to the overwhelming proportion of all convictions that are based on guilty pleas.

Judging from the available evidence, innocent defendants almost never plead guilty when doing so entails a substantial term of imprisonment. Except in capital prosecutions. Radelet, Bedau and Putnam [FN87] list 16 cases of innocent homicide defendants who pled guilty; in most, fear of execution is given explicitly as the reason for the plea. This is, no doubt, another illustration of how death is different. It seems that innocent defendants will almost always risk additional years of their lives in order to seek vindication rather than accept disgrace coupled with a long term of imprisonment, but some will not go so far as to risk death.

The case of John Sosnovske is a good example. [FN88] In 1990, he was falsely implicated in the rape murder of Taunja Bennett by his girl friend, Laverne Pavlinac, who apparently was afraid of him and anxious to be rid of him. In the process, Ms. Pavlinac became entangled in her own lies, and claimed to have participated in the killing. Both were charged with murder. Ms. Pavlinac recanted her confession but was convicted and sentenced to life in prison. Following her conviction, Mr. Sosnovske-who was facing the death penalty-pled no contest and was also sentenced to life imprisonment. Both were freed in 1995 after another man, Keith Hunter Jesperson, confessed and also pled guilty to the same murder. The Portland Oregonian summed up the case:

*489 But there was, indeed, one lesson in this weird case that does bear remembering. The threat of the death penalty led an innocent man to forgo his day in court and accept a lesser sentence in exchange for not fighting the charge. Prosecutors, defense attorneys and judges need to recognize that risk and ensure the death penalty threat isn't made in a shaky case to exact a guilty plea. [FN89]

2. Failures to Dismiss Charges Against Innocent Defendants . The major filter that may prevent a charge based on questionable evidence from turning into a conviction is prosecutorial discretion to dismiss. Overall, dismissals of felony charges outnumber acquittals about 4 to 1. [FN90] Many cases are dismissed because of weak evidence despite the fact that the prosecutor is convinced that the defendant is guilty; other cases are dismissed because the prosecutor is convinced of the defendant's innocence, or has at least come to doubt his guilt. For homicides, and particularly in capital cases, both sorts of dismissals are less likely.

A prosecutor might dismiss a weak case (or offer a deal that's tantamount to dismissal) simply to improve her win/loss ratio at trial, but the major incentive is to conserve resources. Trials are time consuming and expensive; they are a scarce resource. Since most cases cannot be tried, it is obviously sensible to try to restrict trials to cases where the outcomes will be useful- i.e., convictions. If possible, likely losses at trial are avoided through plea bargaining; if not, they may be dismissed even if the prosecutor is convinced of the defendant's guilt. Regardless of their belief in the defendants' guilt, prosecutors focus on the easiest cases-the ones with the best evidence-since those are the cases where their limited resources will have the greatest impact. But homicides are different. Homicides (and other notorious crimes) are the cases for which resources are conserved. A dead loser will still be dismissed, but what if it's merely likely that the defendant will acquitted? If it's a robbery, the prosecutor may dump the case and try another; if it's a murder, she's more likely to forge ahead.

Prosecutors lose a much higher proportion of murder trials than other felony trials, about 30% vs. about 15%. [FN91] As Robert Scott and William Stuntz point out, [FN92] the most likely explanation is that in murder cases they are willing to go to trial with comparatively weak evidence. The main effect of this extra effort is that *490 guilty defendants are convicted who otherwise would never even be tried. But in some cases the evidence is weak because the defendant is not guilty, and some of those innocent defendants are not only tried but convicted. In other words (as with police investigations) as prosecutors work to obtain convictions in hard homicide cases they draw in cases where it's difficult to separate the innocent from the guilty.

Prosecutors also dismiss charges in some cases because they believe the defendant may be innocent, regardless of the evidence that is available to obtain a conviction. The rules of professional responsibility allow a prosecutor to consider her own view of the defendant's guilt in deciding whether to charge, but do not require her to do so. [FN93] Prosecutors have widely varying views on how to apply this vague standard, from those who say that they will never prosecute unless they are convinced beyond a reasonable doubt of the defendant's guilt, to those who believe that regardless of their own uncertainty, their task is to make a case and let the jury decide. [FN94] Whatever the prosecutor's position in the abstract, an actual decision to dismiss a serious charge that would probably have resulted in a conviction is always difficult-especially if the crime has attracted a lot of attention, or if a victim, or several, were killed.

On October 8, 1962, John Stinson, a banker in Evansville, Indiana, was robbed at gunpoint. [FN95] Six months later, by sheer chance, Charles Del Monico-a reputed gangster-was identified as the robber when a witness saw his picture in a newspaper after he had been indicted for extortion. Eventually six witnesses identified Del Monico, apparently all in error: he had credible alibi witnesses who said he was in Miami at the time of the robbery, he passed several polygraph and "truth serum" tests, and there was no evidence corroborating the eyewitness identifications. On the eve of trial, the prosecutor, U.S. Attorney Richard Stein dismissed the charges. In answer to an angry challenge from the victim and his friends, Stein agreed that Del Monico would probably have been convicted at trial. Nonetheless, Stein felt compelled to dismiss because he was persuaded that Del Monico was innocent by the lack of corroboration of the identifications, and by the mass of inadmissible polygraph and truth serum evidence: "We were afraid of a miscarriage of justice." [FN96] This was a gutsy move. But would Mr. Stein have *491 done the same if John Stinson had been killed? Or would he have decided to say: "Let the jury decide"?

The problem is not just public pressure. The evidence of innocence in the Del Monico case did not arrive on the prosecutor's door step on its own; it was presented by his attorneys. That's the rule in cases where prosecutors dismiss because they believe the defendant is innocent. Since everybody agrees that innocent defendants should not be charged, for cases in which the defendant is innocent there is no conflict between the sides. The trouble is identifying the cases in which that applies. The defense brings evidence to the prosecution, often at the cost of some tactical advantage at trial, in order to persuade the prosecutor that this is one of those cases. If there happens to be persuasive, independent evidence of innocence, no problem; but even when there isn't, the defendant himself does know if he's been wrongly accused. But there's the rub: as Scott and Stuntz point out, in the absence of persuasive corroboration, there is no way an innocent defendant can signal his innocence-not by saying so, not by rejecting plea bargains and insisting on trial-that a guilty defendant cannot duplicate. [FN97] The most valuable source of information is also the least credible.

Judge Frank Easterbrook, in a comment on Scott and Stuntz's article, argues that the prosecutor's difficulty is mitigated by the role of the defense attorney: "Just as investment bankers may put their reputations behind hard-to- verify claims of corporate operations, so lawyers may put their reputations behind proffered information . . . Prosecutors take seriously information coming from reputable counsel. Guilty defendants cannot copy the signal of innocence sent by careful honest lawyers." [FN98] There's something to this- defense attorneys do fill this role-but it's an imperfect system at best. It only works if the innocent defendant has a careful, honest and reputable lawyer-by no means a universal condition-and if the prosecutor will listen. In many capital cases, the last condition is undermined by the nature of the pre- trial process.

If the prosecutor insists on the maximum available penalty, then of course there can be no deal. This is a possible scenario in any prosecution, but it's rare: there's usually something the state is willing to give up to get the defendant to plead guilty. In many death cases, however, the prosecutor knows from the start that she wants the death penalty. As a result, there is no plea bargaining: there's nothing to bargain over. In the absence of plea bargaining there will be fewer open channels of communication between the *492 defense and the prosecution, so it may be harder for the defense attorney to get a serious hearing. Worse, in that context, the true value of a claim of innocence becomes harder to interpret. When plea bargaining is an option, a defense lawyer is not likely to commit her credibility to the argument "He didn't do it" unless the lawyer believes that it's true, since (quite apart from possible effects on her reputation) taking that position will undermine her ability to bargain convincingly for a lenient deal. When no deal is possible, arguing that her client is innocent is the only pre-trial move left. As far as this client is concerned, there's nothing to lose by making it. And, since a client's life is at stake, the defense attorney may be driven to make the claim whether or not she believes it. More important, the prosecutor knows that the defense attorney may feel obliged to argue that the defendant is innocent, whether or not she thinks it's true. When inflexible lines are drawn at the start-which is particularly likely in a capital prosecution of a heinous, gruesome and highly publicized murder-the defense attorney is less likely to be able to convince the prosecutor of anything, and especially not that her client has been wrongly accused. [FN99]

C. Trial

An innocent defendant who goes to trial faces a high risk of conviction. The best generalization about juries in criminal cases is that they usually convict. To be sure, the great majority of defendants should be convicted. The question is: can juries accurately sort the innocent from the guilty? Or, to put it in context, how often do juries spot innocent defendants that the prosecutors have missed? Unfortunately, juries approach this task with two severe handicaps: they have less information than the prosecutors or the *493 police, and they have essentially no experience. Given these limitations, it is unrealistic to expect juries to systematically correct errors in the earlier decisions to investigate, to arrest and to prosecute. [FN100]

This is bad news for homicide defendants. Whether it's because prosecutors take weaker cases to trial [FN101] or because they insist on the maximum penalty, [FN102] homicide defendants are more likely to face a jury than other criminal defendants. In 1992, for example, 12% of robbery convictions across the country were obtained at trials, of which 8% were jury trials, while 41% of murder convictions were after trial, including 33% that went to jury trial. [FN103] In other words, since pre-trial sorting does less to winnow homicide cases than other prosecutions, homicide defendants are more likely to face the chancy ordeal of trial.

I do not mean to say that the institution of trial by jury does not help reduce the incidence of Search Term Begin erroneous convictions Search Term End . It no doubt does fill that function, but by brute force: by making it more difficult for the prosecution to obtain any convictions, and by discouraging trials of the guilty and the innocent alike unless the evidence of guilt is very strong. The main benefit of this process is that feedback from court may improve pre-trial investigations and increase selectivity in charging-the stages of the process we have already discussed. If all works well, the result is that few innocent defendants are brought to trial; most defendants who are convicted are guilty; most who are acquitted are also guilty; and yet, if an innocent defendant is tried, he will probably be convicted.

Given this structure, trial plays a comparatively minor role in the production of errors in capital cases. To the extent that jury behavior at trial does matter, the question is: Do juries behave differently in homicide trials in general, and in capital homicides in particular, than in other criminal trials? There are several reasons to think that juries treat homicides and capital cases differently than other criminal cases, and most of them point in the direction of a higher likelihood of conviction.*494 1. Factors that Increase the Likelihood of Conviction .

a. Publicity . Most crimes, even most homicides, receive very little attention from the media. A few crimes, however, are heavily publicized. Many, perhaps most of these notorious crimes are homicides, and especially the unusual and heinous homicides which are likely to be charged as capital crimes. In those cases, most jurors will have heard all sorts of things about the case before they got to court, many of them inadmissible, misleading and inflammatory. They may have seen or heard or read police officers or other government officials declare the defendant guilty. They may have witnessed or felt a general sense of communal outrage. All this will make them more likely to convict. Courts may attempt to mitigate the impact of pre-trial publicity by various means-most effectively by changing the location of the trial-or they refuse to do so. [FN104] As a result, the records of Search Term Begin erroneous convictions Search Term End include scores of cases in which publicity and public outrage clearly contributed to the error-from the convictions of Leo Frank in 1913 [FN105] and the Scottsboro Boys in 1931 [FN106] , to the conviction of Rolando Cruz and Alejandro Hernandez in 1985. [FN107]

b. Death Qualification . In capital cases, juries decide the sentence as well as determine guilt or innocence. As a result, the jury selection process includes a unique procedure, "death qualification," that is designed to ensure that the jury is qualified for the sentence phase. Most jurors who are strongly opposed to the death penalty, and some who are strongly in favor, are excluded at the outset. [FN108] Many studies have shown that these exclusions make the jury more likely to convict. [FN109] In addition, the process of questioning *495 jurors about their willingness to impose the death penalty before the defendant has been convicted tends to create the impression that guilt is a foregone conclusion, and the only real issue is punishment. [FN110]

c. Fear of Death . In a capital case, avoiding execution can become the overriding imperative for the defense. In extreme cases, fear of death drives innocent defendants to plead guilty in return for a lesser sentence, even life imprisonment. If the defendant does not plead guilty, either because no plea bargain is offered or because he was unwilling to take it, the same pressure will be felt at trial. It may drive the defense to make tactical choices that compromise its position on guilt in order to improve the odds on penalty; in some cases, the defense may virtually concede guilt and focus entirely on punishment. It will certainly distract the defense from the issue of guilt and force it to spread its resources more thinly. This distraction might increase the chances of conviction even for those capital defendants who are represented by skillful lawyers with adequate resources; it will be far more damaging for the many capital defendants whose defense is shamefully inadequate. [FN111]

d. Heinousness . In theory, jurors are supposed to separate their decision on the defendant's guilt from their reaction to the heinousness of his conduct: If the evidence is insufficient, they should be just as willing to acquit a serial murderer as a shoplifter. Nobody believes this. Even in civil trials, where the jury is asked to decide cases by a preponderance of the evidence, there are indications that juries (and judges) are more likely to find defendants liable, on identical evidence, as the harm to the plaintiff increases. [FN112] In criminal trials the problem is worse, since the burden of persuasion is proof beyond a reasonable doubt. In a close criminal case the jury is supposed to release a defendant who is in their opinion, probably guilty. This is a distasteful task under any circumstances, *496 but it becomes increasingly unpalatable-and unlikely-as we move up the scale from non- violent crime, to violent crime, to homicide, to aggravated grisly murder.

2. Factors Which Decrease the Likelihood of Conviction.

a. Quality of Defense . Capital defendants, and to some extent homicide defendants in general, may be better represented than other criminal defendants. The attorneys who are appointed to represent them may be more experienced and skillful, and their defenders may have more resources at their disposal. Other things being equal, higher quality representation will decrease the likelihood of conviction, and may operate as a check on errors and misconduct that drive some innocent capital defendants to trial and to conviction.

b. Severity of the Penalty . Prosecutors, defense attorneys, and judges widely believe that some jurors are more reluctant to convict a defendant who might be executed than one who faces a less extreme punishment. In Adams v. Texas [FN113] the United States Supreme Court acknowledges this possibility and held that a juror could not automatically be excluded from service because of this reaction:

Nor in our view would the Constitution permit the exclusion of jurors from the penalty phase of a Texas murder trial if they . . . frankly concede that the prospects of death penalty may affect what their honest judgment of the facts will be or what they deem to be a reasonable doubt. Such assessments and judgments are inherent in the jury system. . . . [FN114]

To the extent that jurors do feel this way, they may be less likely to convict in capital trials than in other homicides.

3. Net Effects . When there are forces that push in one direction and forces that push in the other, it is sometimes possible to say that they cancel out. Not here. A sailboat that keeps getting hit by gusts from the east is no less likely to capsize if there are also gusts from the west.

The effects I have described are extremely variable. Publicity, death- qualification, the heinousness of a homicide-each of these may make a critical difference in a particular case, or it may not. On the other side, the protective features of capital trials are uneven *497 at best. Many capital defendants do not have quality representation, by any standard. [FN115] And the anxiety that jurors may feel when a defendant's life is at stake will be relieved if a jury decides (as they often do in deliberations on guilt ) that he will not be sentenced to death. If they do get that far, the competing impulse-to not free a man who has killed-may take over, in force.

I once saw a cartoon of two men in black robes, obviously judges, talking in a hall. One says: "Some days I'm feeling good and everyone gets probation, and some days I get up on the wrong side of bed and I throw the book at everybody. It all balances out." In statistical terms, the problem is increased variance: Since nobody gets the average punishment, the more the judge's sentences are spread out arbitrarily, the more of them are errors-and errors on one side don't balance out errors on the other. In capital trials, one particular type of error-an uncommon one-is overwhelmingly important: conviction of an innocent defendant. Given the nature of the problem, I hazard to add one more metaphor: If you're building a seawall, adding height to one part won't counteract cutting away at another.

OTHER NEWS

TheDeath Penalty Information Centerreports:

Gallup Poll Probes National Opinion on the Death Penalty

The most recent national death penalty poll conducted by the Gallup Organization found that, while the majority of Americans support capital punishment, they oppose executing those who are mentally retarded, mentally ill, or who are juveniles at the time of their crime. Only about half of Americans believe the death penalty is applied fairly. The poll found that:

Only 53% of those polled believe the death penalty is applied fairly, while 40% say it is applied unfairly. Among non-white respondents, 54% believe the death penalty is applied unfairly.

When given the sentencing alternative of life without the possibility of parole, 52% of Americans support the death penalty and 43% favor life imprisonment.

82% of respondents oppose the death penalty for the mentally retarded.

73% oppose the death penalty for those who are mentally ill.

69% of Americans oppose capital punishment for juvenile offenders.

In general, 72% of Americans favor the death penalty for defendants convicted of murder and 25% opposed it.

(Gallup News Service, May 20, 2002). To see the complete results of this poll, visit the Gallup Web site (this link requires a subscription). See also, Public Opinion.

NEW VOICES: U.S. Senators Call for Moratorium on Executions,

Death Penalty Review

In a recent Baltimore Sun opinion piece, U.S. Senators Russ Feingold (D-WI) and Jon Corzine (D-NJ) called for a national halt to executions while an independent review of the nation's death penalty is conducted. The Senators wrote, "The message is clear: In response to the glaring flaws in the administration of capital punishment, the nation should conduct a thorough, nationwide review of the death penalty. No executions should go forward while an independent, blue-ribbon commission examines the federal and state systems of capital punishment - systems so riddled with errors that for every eight people executed in the modern death penalty era, one person on death row has been found innocent." The Senators are co-sponsors of the National Death Penalty Moratorium Act. (Baltimore Sun, May 16, 2002). See also, Innocence and Recent Legislative Activity.

Georgia Judge Notes Expensive Bottom Line in Capital Cases

A recent death penalty case in Georgia has led Fulton County Superior Court Judge Stephanie Manis to question the value of expensive capital trials. "The death penalty has great popular appeal, but I don't think the taxpayers have looked at the bottom line," she said. "The death penalty is damn expensive." Expenses for the capital trial of Jamil Abdullah Al-Amin, who received a sentence of life without the possibility of parole, included the following:

The District Attorney dedicated four prosecutors and a full-time investigator to the case during the trial and in the months leading up to it. Independent of the time spent preparing for the case, the prosecutor's salaries exceeded $74,000 for the two months of trial and jury selection.

The District Attorney's Office spent approximately $34,000 for equipment, graphic design for court exhibits, and expert testimony. An additional $43,000 was spent on overtime for investigators.

$164,000 in fees and expenses were spent for one of the defense attorneys in the Al-Amin case. Two additional lawyers have yet to submit bills and these additional charges are expected to put total legal fees well above $200,000.

It cost more than $87,000 to select and sequester the jury for this case. This amount included $63,600 for hotel rooms, dinners and drinks for jurors, $6,000 for juror lunches and beverages, $2,500 for juror transportation, $765 for entertainment expenses, and $14,300 to copy the questionnaires used to pick the jury. This amount does not include the overtime paid to sheriff's deputies who guarded the jury and the court.

(Atlanta Journal-Constitution, May 12, 2002). See also, Costs.