From Casetext: Smarter Legal Research

Willis v. Cockrell

United States District Court, W.D. Texas, Pecos Division
Aug 9, 2004
No. P-01-CA-20, CAPITAL HABEAS (W.D. Tex. Aug. 9, 2004)

Opinion

No. P-01-CA-20, CAPITAL HABEAS.

August 9, 2004


AMENDED ORDER GRANTING PETITION FOR WRIT OF HABEAS CORPUS


INTRODUCTION

Ernest Willis brings this petition under 28 U.S.C. § 2254 for a writ of habeas corpus challenging his conviction and sentence of death in Texas state court for the murder of Elizabeth Grace Belue. The parties filed cross motions for summary judgment. After an extensive review of the state court determination, the parties' briefing and the applicable law, the Court finds that Willis's petition for writ of habeas corpus should be granted because both his conviction and sentence were obtained in violation of the United States Constitution. Specifically, the Court grants Willis's petition on the following grounds: 1) Willis's due process rights were violated by the State's administration of medically inappropriate antipsychotic drugs without Willis's consent; 2) the State suppressed evidence favorable and material to the sentencing determination; 3) Willis received ineffective assistance of counsel at the guilt-innocence phase; and 4)Willis received ineffective assistance of counsel at the sentencing phase. On all other grounds, Willis's petition is denied.

Petition, (Docket No. 13), filed Dec. 12, 2001 [hereinafter Pet.].

Respondent's answer and motion for summary judgment (Docket No. 19) and Petitioner's Reply (Docket No. 22). Petitioner's Reply indicates that Petitioner believes that the record before the Court is satisfactory and thus this petition is ripe for decision. See Pet.'s Reply, at 2.

FACTUAL AND PROCEDURAL BACKGROUND

In the early morning hours of June 11, 1986, a fire destroyed a home in Iraan, Texas. At the time, the house was occupied by four people: Elizabeth Belue, Gail Allison, Ernest Willis and Billy Willis. All were guests of the resident tenants of the house, Michael and Cheryl Robinson. The Robinsons were not home at the time. Two of the guests, Elizabeth Belue and Gail Allison, died in the fire due to smoke inhalation. Their remains were found in two of the bedrooms of the three bedroom house. The other two guests who survived the fire were Petitioner, Ernest Willis, and his cousin, Billy Willis. The Willis cousins did not know Belue or Allison prior to the day of the fire. Billy Willis escaped the fire when he jumped, naked, out of a bedroom window.

According to Ernest Willis, on the night of the fire he was sleeping on the sofa in the living room. Willis further claims that the smell of fire awakened him and that he ran through the house trying to awaken the occupants but could not enter the bedrooms due to the fire and smoke. Willis claims that when his attempts to reach the others failed, he ran out the front door and around the outside breaking windows in an attempt to secure an escape route for those still inside. The State of Texas disputes Willis's version of the events.

Willis was ultimately arrested and charged with the murder of Elizabeth Belue. The indictment charged Petitioner with intentionally and knowingly causing the death of Elizabeth Belue in the course of committing arson on a habitation. According to the State, Willis intentionally poured a flammable liquid accelerant on the floor of the house and set it afire. But even if one relies exclusively upon the testimony of witnesses presented by the prosecution at trial, numerous discrepancies remain regarding the events leading up to the fatal fire. The State did not present at trial a theory of Willis's alleged motive.

Both of the individuals who survived the fire, Billy and Ernest Willis, were initially suspects.

After a jury trial before the Honorable Brock Jones of the District Court of Pecos County, Texas, 112th Judicial District, Willis was convicted on August 4, 1987 of capital murder and sentenced to death for Belue's murder. Willis's sentencing phase was held on August 5, 1987. Willis's conviction was affirmed on direct appeal on June 7, 1989, and on October 9, 1990, the United States Supreme Court denied certiorari. Willis then filed for state post-conviction relief on October 8, 1991. On June 7, 2000, following five days of hearing, Judge Jones of the Texas trial court issued detailed findings of fact and conclusions of law and recommended granting relief to Willis. On December 13, 2000, the Texas Court of Criminal Appeals ("CCA") denied Willis all relief.

Willis v. State, 785 S.W.2d 378, 387 (Tex.Crim.App. 1989), reh'g denied, (Jan. 17, 1990), cert. denied, 498 U.S. 908 (1990).

Willis v. Texas, 498 U.S. 908 (1990).

Judge Jones was the judge for both Willis's trial and his state post-conviction hearing.

Willis then filed the instant petition alleging the following claims for relief: 1) Willis is innocent and thus the Eighth and Fourteenth Amendments require that his conviction and sentence be vacated; 2) the State's wrongful administration of antipsychotic medications to Willis violated his right to due process and other constitutional rights, including the right to counsel and the right to confront witnesses; 3) defense counsel rendered ineffective assistance at the guilt-innocence phase; 4) defense counsel rendered ineffective assistance at the sentencing phase; 5) the prosecution suppressed evidence material to the sentencing determination; and 6) the cumulative effect of error outlined in the above claims violated due process.

In support of his argument for habeas relief grounded in actual innocence, Willis relies upon evidence he introduced at the state post-conviction hearing supporting his account of the pertinent events. But, as will be detailed in Section IV addressing the innocence claim, Judge Jones rejected the innocence claim based upon insufficiency of the evidence Willis offered in support.

STANDARD OF REVIEW

The federal habeas statute, as amended by the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA), 28 U.S.C. § 2254, provides that:

An application for a writ of habeas corpus on behalf of a person in custody pursuant to the judgment of a State court shall not be granted with respect to any claim that was adjudicated on the merits in State court proceedings unless the adjudication of the claim —
(1) resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States; or
(2) resulted in a decision that was based on an unreasonable determination of the facts in light of the evidence presented in the State court proceeding.

A state court's decision is deemed contrary to clearly established federal law if the state court arrives at a conclusion opposite to that reached by the Supreme Court on a question of law or if the state court decides a case differently than the Supreme Court on a set of materially indistinguishable facts. Under the "unreasonable application" clause, a federal habeas court may grant the writ if the state court identifies the correct governing principle from the Supreme Court's decisions but unreasonably applies that principles to the facts of the prisoner's case.

Williams v. Taylor, 529 U.S. 362 (2000).

Id.

Pursuant to section 2254(e)(1), state court findings of fact are presumed to be correct, and the petitioner bears the burden of rebutting the presumption of correctness by clear and convincing evidence. When the state habeas judge also served as the trial judge, as Judge Jones did in this case, the state judge's factual findings are entitled to particular deference.

28 U.S.C. § 2254(e). See also Pondexter v. Dretke, 346 F.3d 142, 146 (5th Cir. 2003); Valdez v. Cockrell, 274 F.3d 941, 947 (5th Cir. 2001).

See Davis v. Blackburn, 789 F.2d 350, 352 (5th Cir. 1986); Vuong v. Scott, 62 F.3d 673, 684 (5th Cir. 1995), cert. denied, 516 U.S. 1005 (1995).

LEGAL ANALYSIS

Before addressing each of Willis's claims, the determination must first be made whether the Texas trial court's post-conviction factual findings are properly before this Court in light of the CCA's denial of relief.

I. The State Trial Court's Post-Conviction Factual Findings are Properly Before the Court

At the outset, the Court notes that the posture of this dispute, cross-motions for summary judgment, indicates the parties' agreement that the state trial court's post-conviction findings of fact are properly before this Court on habeas review. Neither party requested an evidentiary hearing. Moreover, the Court, on independent review, finds the state trial court's factual findings are properly considered here, even in light of the Texas CCA's denial of relief.

According to Craker v. Procunier, the Fifth Circuit requires that deference is owed to the state court's post-conviction factual findings when denial by the Texas Court of Criminal Appeals was not inconsistent with those factual findings. This must be the case because, for example, the appellate court might hold that the facts determined by the trial court did not warrant relief based on the appropriate legal standards, and such a holding would not be inconsistent with those factual findings. Despite the deference typically afforded the state court's post-conviction factual findings, in some circumstances the state trial court's findings do not survive the CCA's denial of relief. In Micheaux v. Collins, the Fifth Circuit held that the state trial court's findings did not survive the CCA's denial of relief where 1) the CCA denied relief without written order and 2) the factual findings were directly inconsistent with the CCA' peremptory denial of relief.

Craker v. Procunier, 756 F.2d 1212, 1213-14 (5th Cir. 1985). See also Westley v. Johnson, 83 F.3d 714, 721 n. 2 (5th Cir. 1996).

Westley, 83 F.3d at 721 n. 2.

Micheaux v. Collins, 944 F.2d 231, 232 (5th Cir. 1991) (en banc).

Id. See also Singleton v. Johnson, 178 F.3d 381, 384, 85 (5th Cir. 1999). In Walbey v. Dretke, 2004 WL 909736 (5th Cir. 2004) (per curiam) (unpublished), the Fifth Circuit applied Micheaux instead of Craker even though the CCA had issued a written order. However, the written order in Walbey was silent as to the state trial court's findings of fact and did not state whether the CCA accepted or rejected the factual findings of the trial court. In addition, the Walbey court stated that the facts found by the state trial court were directly inconsistent with the CCA's denial of habeas relief. In Walbey, the CCA's opinion contained no specific factual findings or reasoning to support its ultimate conclusion, and thus "the terse opinion of the . . . CCA here is the functional equivalent of a denial without written order." Id. at *3. The Fifth Circuit remanded the case to the district court for an evidentiary hearing. Although unpublished because it provided no change or explanation of a generally established rule of law, Walbey is mentioned here because it demonstrates a helpful application of the distinction between Micheaux and Craker.

The CCA's order in the instant case more closely resembles Craker on the " Craker/Micheaux continuum." This Court discusses the CCA's analysis of each of Petitioner's claims in the relevant section in this opinion. Generally though, for two of the claims before the Court — prosecutorial suppression of evidence and wrongful administration of antipsychotic drugs — the CCA identified a legal principle and found that the facts as found by the trial court did not meet the legal standard. For the other two claims — ineffective assistance of counsel at the guilt-innocence phase and at the sentencing phase — the CCA discussed facts from the record different than, but not inconsistent with, the facts relied upon by the trial court. Then, based on a determination of those different facts as legally significant, and on the basis of legal standards the CCA employed, the CCA denied relief. Because the CCA's opinion in this case included legal reasoning and discussion of the facts, it is not the functional equivalent of denial without written order. And for all four of the above claims, the CCA's opinion was based on the use of, in whole or in part, an erroneous legal standard irrespective of the relevant facts used in relation to that legal standard. Therefore, this Court must defer to the post-conviction factual findings of the state trial court.

Walbey, 2004 WL 909736 at *2.

II. The State Trial Court's Post-Conviction Findings of Fact

Here, the Court provides a summary of the state trial court's post-conviction factual findings. The relevant facts will be reiterated or developed for the analysis of each of Petitioner's claims in the appropriate section, as well.

A. The State Unnecessarily Medicated Willis While Incarcerated and During Trial

Willis was arrested and incarcerated at Pecos County Jail on October 22, 1986. Willis was not taking any antipsychotic medications at the time of his arrest and initial incarceration in the Pecos County Jail. The State began administering Haldol (the brand name for the generic drug Haloperidol) to Willis on February 23, 1987. As of March 23, 1987, the State began administering 40 milligrams ("mg.") of Haldol per day to Willis; and on May 30, 1987, the State began administering between 8 mg. and 32 mg. of Perphenazine per day to Willis.

The State continued to daily administer these doses of Haldol and Perphenazine to Willis throughout the course of his trial, including the jury selection, guilt-innocence and penalty phases. These proceedings began on July 8, 1987 and concluded on August 5, 1987. Willis was formally sentenced on August 5, 1987. The State continued to administer Haldol and Perphenzine to Willis until August 27, 1987. The following day, Willis was transported from Pecos County to the Texas Department of Corrections ("TDC") in Huntsville. Willis has not been administered antipsychotic medication at any time since August 27, 1987 — either during subsequent stays in the Pecos County Jail (pursuant to bench warrants) or while in the custody of TDC.

This factual finding implies a lack of medication beyond the date of the trial court's post-conviction factual findings. While the record suggests that the finding remains true long after the trial court's hearing and until today, this Court makes no such finding and instead defers to the trial court's finding and that relevant period of time.

There are multiple reasons the medications administered to Willis were inappropriate according to Judge Jones. First, the dosages for Haldol (40 mg. per day) and Perphenazine (8 mg. — 32 mg. per day) that the State gave to Willis during the course of the trial were high doses, even for acutely psychotic patients. The maximum dose of Haldol for a severely psychotic person is 15 mg. per day. Willis received more than twice that amount at 40 mg. per day. Second, Willis was administered two antipsychotic medications. Judge Jones found that the combination of two different antipsychotic drugs has more than an additive effect on a patient and that the administration of antipsychotic drugs to a non-psychotic individual increases the side-effects of the drugs.

Judge Jones also found that common side effects of antipsychotic medication include: flat or little facial expression, inexpressiveness, rigidity of the facial muscles, fixed gaze, drowsiness, confusion and diminished ability to communicate with others. Judge Jones stated that all of these side effects were exhibited by Willis during his trial, and Willis's demeanor at the evidentiary hearing on his habeas petition was markedly different from his demeanor at trial. Judge Jones found that Willis's expression, from the moment he stepped into the courtroom for voir dire throughout the entire trial, reflected an apparent indifference to the proceedings. Judge Jones found that Willis's demeanor at trial was a direct result of the antipsychotic medications he was receiving, and was "absolutely typical" of known side effects of antipsychotic medications. Finally, Judge Jones found that the prosecution seized upon Willis's demeanor in the guilt-innocence and punishment phases of the trial, asking the jury to draw inferences of guilt and future dangerousness from Willis's lack of apparent feeling or emotion.

Judge Jones also found that, while an individual's I.Q. is typically stable throughout one's life, Willis's intelligence test at the time of trial was significantly lower than at the time of the evidentiary hearing on the habeas petition. Ex parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 10.

Id.

Judge Jones also made findings regarding the medical justifications for the antipsychotic medications. Judge Jones found that the State's administration of the drugs to Willis was without any medical need. Antipsychotic medications like Haldol and Perphenazine are not justified unless a patient is suffering "psychotic symptoms" as a result of a "lifelong" psychotic disorder. "Psychosis is a very, very serious psychiatric condition . . . manifest by symptoms such as schizophrenia, derangement, hallucinations, delusions, paranoia, and formal thought disorder." Judge Jones found that nothing in any of Willis's records, or his social or medical history, indicates that he needed to take antipsychotic medications. Furthermore, the record does not show that the State established the requisite "overriding justification" and "medical appropriateness" findings before administering the mind-dulling or psychotropic drugs to Willis during his trial. Finally, the state court found that although Willis did not affirmatively object to the medication, his failure to object was not consent.

Id. at 11.

Id.

As will be discussed later, although not so determined by Judge Jones, the evidence suggests that Willis was actually medicated without his knowledge for symptoms he did not manifest.

B. Findings of Ineffective Assistance of Counsel at the Guilt-Innocence and Sentencing Phases

Judge Jones found ineffective assistance of counsel at multiple stages in Willis's representation.

1. Failure to Investigate Willis's Demeanor and Discover the Administration of Antipsychotic Drugs

Judge Jones found that defense counsel took no steps to determine the cause of Willis's appearance or demeanor during the course of trial. As a result, defense counsel never learned that the State was administering high doses of antipsychotic medication to Willis during his incarceration at Pecos County Jail both before and during trial. Defense counsel did not speak with any person with medical training concerning Willis's physical and emotional appearance. Defense counsel did not attempt to review Willis's Pecos County Jail medical records.

Judge Jones found that Willis's defense counsel not only had the right to access those records, but that it was "rudimentary" and "basic" for counsel to gather such records. In addition, defense counsel recognized a problem with Willis's demeanor and suspected that the problem could be related to medication that Willis was taking but, nevertheless, failed to investigate Willis's demeanor and failed to gather medical records. Had defense counsel gathered Willis's Pecos County Jail records, counsel would have known Willis was unnecessarily receiving large doses of Perphenazine and Haldol prior to and during his trial.

Ex Parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 17.

2. Failure to Object to the State's Use of Willis's Demeanor and the State's Descriptions of Willis as an Animal

Judge Jones found that the State referred to Willis's demeanor during trial as evidence of guilt and dangerousness and the State urged jurors to infer a lack of remorse based on Willis's demeanor. Defense counsel did not object to any of these references by the prosecution. The state trial court found that the prosecution characterized Willis as a "pit bull," an "animal," and a "rat," during voir dire, closing arguments and at the penalty phase.

Id. The Court of Criminal Appeals, in its decision affirming Willis's conviction on direct appeal, held that failure to object to an impermissible jury argument generally waives any error. See Willis, 785 S.W.2d at 385.

Ex Parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 18.

Based upon these findings, Judge Jones concluded as a matter of law that defense counsel's failure to object to the State's use of Willis's demeanor contributed to defense counsel's failure to meet the standard of reasonableness required for effective assistance of counsel. The Court considers the legal conclusions related to the factual findings in the relevant section below.

3. Failure to Cross-Examine Aggravating Evidence and to Present Mitigating Evidence

Id. at 19-22.

Judge Jones made the following findings of fact with respect to defense counsel's failure to cross-examine purported aggravating evidence and failure to present mitigating evidence on Willis's behalf. The penalty phase of Willis's trial lasted less than half a day. The transcript from the penalty phase consumes barely ten pages. The prosecution called two witnesses, both local law enforcement officers, who testified that Willis had a bad reputation in the unspecified communities in which he resided. On cross-examination, defense counsel asked these witnesses a total of two questions. Defense counsel knew in advance who the State's witnesses would be and what the subject matter of their testimony would be. Counsel did not investigate the veracity of the witnesses or otherwise develop evidence or arguments to respond to the government's penalty phase case.

Judge Jones also found that Willis's case was his counsel's first capital trial. The defense did not prepare for the penalty phase, did not meet with Willis in advance of the penalty phase, introduced no evidence, and presented no witnesses whatsoever on Willis's behalf. Despite being unprepared, defense counsel did not request a continuance or a recess to prepare for the penalty phase. In fact, defense counsel met with Willis less than three hours prior to July 1987, when jury selection commenced. Defense counsel spoke to four or five people who knew Willis but failed to follow-up on the limited information those individuals had pertaining to Willis.

Judge Jones found that defense counsel could have presented the following mitigating evidence but did not do so: testimony of at least five Pecos County Law Enforcement Officers that Willis was a respectful and well-behaved prisoner who was not the type to act violently or misbehave; testimony of other individuals that Willis was non-violent; testimony that Willis turned himself in when he learned of the outstanding indictment against him; testimony of heroic acts by Willis who, for example, saved the life of a drowning boy and assisted his infant niece who had been severely burned in a car fire; testimony of family and friends describing Willis as a caring family man and responsible individual. The state trial court found that the above mitigating evidence was readily accessible and available to defense counsel at little or no cost. Every character witness who testified at the post-conviction hearing stated that he or she would have been willing to testify on Willis's behalf at his trial.

Defense counsel contacted none of these witnesses. Some of the witnesses were present in the courtroom for portions of Willis's trial. Ex Parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 21.

C. Prosecution's Failure to Disclose Pretrial Psychological Report

At the post-conviction hearing in state court, Judge Jones heard evidence concerning a pretrial psychological report finding that Willis was not a future danger. The report was submitted to the prosecution and never turned over to the defense before or during trial. The findings of fact are summarized below. Based upon these findings of fact, Judge Jones held that the evidence suppressed by the prosecution was both favorable and material and that Willis was entitled to habeas relief for due process violations.

See Brady v. Maryland, 373 U.S. 83, 87 (1963) (prosecutorial suppression of evidence that is favorable to an accused "violates due process where the evidence is material either to guilt or punishment, irrespective of the good faith or bad faith of the prosecution").

On December 2 and 3, 1997, before the post-conviction evidentiary hearing at the state trial court, Willis was interviewed by Dr. Mark Cunningham, a clinical and forensic psychologist. During this interview, Willis stated that he recalled having been examined by a psychologist while awaiting trial in the Pecos County Jail. No reference to a report of a pretrial psychological or psychiatric examination existed in the trial transcript, the trial exhibits, the case files of Willis's trial counsel, or the court's files. Consequently, an investigation was conducted to determine whether Willis's recollection was accurate.

As a result of the investigation, it was determined Dr. Jarvis Wright, a forensic psychologist, examined Willis on July 12, 1987 and prepared a written report memorializing his findings. Dr. Wright forwarded a copy of his report (the "Wright report") to Willis's post-conviction counsel in December 1997.

Dr. Wright conducted the examination and prepared the written report on behalf of the prosecution. Before Willis's trial, the District Attorney's office contacted Dr. Wright and requested a psychological examination of Willis. On July 12, 1987, Dr. Wright examined Willis, who was then in the custody of the Pecos County Jail, to determine: 1) Willis's competency to stand trial; 2) Willis's sanity and the presence or absence of mental illness; and 3) the likelihood that Willis would present a future danger. Shortly after the examination, Dr. Wright orally reported his findings directly to J.W. Johnson in the District Attorney's office. Dr. Wright informed Johnson that, based on the evaluation of Willis, he "didn't think this was a good death penalty case," as he found no evidence to support a conclusion of future dangerousness for the purposes of the Texas capital sentencing statute. Furthermore, Dr. Wright determined that Willis was competent to stand trial and did not exhibit any form of mental illness or mental retardation. At the time of Willis's trial, Dr. Wright did not discuss the psychological examination of Willis with anyone other than Johnson.

At the time Dr. Wright conducted the examination of Willis, there was a pending motion for a psychiatric evaluation. After the evaluation and report by Dr. Wright, the State withdrew its motion for a psychiatric evaluation and stated that no expert testimony of Willis's mental state would be offered at trial. Pet. at 156.

Ex Parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 2.

On Monday, July 20, 1987, the first day of testimony in Willis's trial, Dr. Wright sent, by Federal Express, a final copy of the Wright report and the Wright invoice from his office in San Angelo, Texas, to the District Attorney's office in Fort Stockton, Texas. On Tuesday, July 21, 1987, at 2:41 p.m., the Federal Express package with the Wright report and the Wright invoice arrived at Johnson's office. Albert Valadez, the assistant prosecutor in Willis's trial, accepted and signed for this Federal Express package.

Willis's trial lasted two and one-half weeks.

Ex Parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 3. Federal Express records, as well as Dr. Wright's records, are the source of all facts relating to the delivery and receipt of the Wright report.

During the state habeas hearing, the State repeatedly denied that the prosecution had any knowledge of the Wright report, a claim belied by the facts presented during hearing.

Had Dr. Wright been called as a witness during the penalty phase of Willis's trial, he would have testified, based on his examination of Willis, that he "knew of no information" that would justify a conclusion that Willis would be dangerous in the future. Furthermore, the Wright report stated that if "sworn testimony indicates that [Willis's] behavior until the time of the current alleged offense was no worse than his previous behaviors, we could probably say with safety that the current alleged behavior was an isolated event which he probably will not repeat." Judge Jones found an abundance of available evidence, through the testimony of acquaintances of Willis and law enforcement officers, established that Willis had no history of violent behavior and that any prior episodes of misconduct were nonviolent.

Ex Parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 4, citing Dr. Wright's testimony at the state post-conviction hearing.

Id. at 2, quoting Def. E.H. Ex. 25, at 5-6.

Judge Jones therefore found that the prosecution failed to disclose the Wright report to the defense prior to or during Willis's trial. Although Willis's trial attorneys agreed to allow the prosecutors to conduct a pre-trial psychological examination of Willis to ensure his competency to stand trial, the prosecution did not reveal that an assessment of future dangerousness had also been done.

D. Facts Related to Willis's Innocence Claim

The state trial court did not resolve substantial factual disputes related to Willis's claim that he is actually innocent. Willis's version of the incident leading to arrest and the events surrounding the incident differ from the State's theory of the case. Because the factual dispute was not resolved by Judge Jones's findings of fact, the parties' factual allegations and corresponding arguments are presented in the next section addressing the innocence claim.

III. Innocence Claim

Due to other relief given on different grounds, it is not necessary for this Court to resolve the parties' dispute regarding Willis's claim of innocence. But, to provide a background for the other substantive claims, the Court discusses in detail the facts Willis alleges. The factual allegations recited here are from Willis's petition and were not included in Judge Jones's factual findings. Although Willis's allegations of innocence and factual allegations supporting the claim were presented to the state trial court, the state trial court only made one factual finding concerning the innocence claim. The state trial court found that David Long, who had confessed to the crime for which Willis was convicted and sentenced to death, refused to testify at the state evidentiary hearing. The state trial court determined that Long's prior confession, which was tape recorded by law enforcement officers, was not sufficiently corroborated to be admissible. Therefore, other than Long's confession, these facts related to Willis's innocence claim have been neither specifically rejected nor accepted by the state court, though the state court did say that the testimony was insufficient to support a finding that Willis is innocent.

Id. at 25.

Id. at 33.

A. The State's Theory of the Fire

At trial, the State's experts testified that the burn patterns and degree of burning indicated that a flammable liquid was poured on the floor of the house, throughout the living and dining areas, in front of the bedroom door jambs, around the front and back door entrances, and beneath and on top of the sofa in the living area. The State's experts also testified that the fire originated in the living area of the house and quickly, if not simultaneously, ignited the dining room and kitchen. Thereafter the fire spread to the bedrooms. The State's arson investigators testified that if Willis had been sleeping on the sofa he would have been burned.

The State asserted that Willis's version of events was incredible for two main reasons. First, while broken glass was found outside the house, none was found inside, and thus the State said that the evidence did not support Willis's claim that he ran around the outside of the house trying to break windows so that the people inside could escape. Second, Willis had no burn marks, no singed clothing, no singed hair, did not smell like smoke, and his clothing did not have cinder marks. Two days after the fire, Willis had a very bad burn mark on his shoulder which Willis claimed occurred in the fire but several witnesses, including Sheriff Wilson, stated Willis had no such injury the day of the fire.

A stain was found on the shirt, and the stain was identified as betadine, an antiseptic.

Other evidence included the fact that the day after the fire, Deputy Jackson, one of the investigators on the case, discovered that the front portion of the garden hose had been cut off. Jackson learned from the tenants that this was a new hose that had previously been intact. Later, Jackson found a smaller portion of the garden hose, a trace analysis of which indicated the presence of gasoline. No known accelerant was positively identified on Willis's pants.

B. Confession of David Martin Long

While the state court found the corroboration of Long's confession insufficient, the corroborating witnesses were: David Paulk, Amelia Fuentes, Billy Willis, George Wheat, Michael and Cheryl Robinson and Marshall Smyth. See Pet. at 44-48.

Long was an inmate confined at the same facility with Willis. He was convicted of capital murder on an unrelated charge and has since been executed. While incarcerated, Long repeatedly told George Wheat, the supervisor of Psychiatric Services at Ellis One Unit, that he had set the Iraan fire. Initially, Long only told Wheat that there was an inmate on death row who Long knew was innocent because that inmate had been convicted of a crime Long had committed. Over time, Long identified Willis as the innocent inmate. Though Wheat was initially skeptical of Long's confession, Wheat became satisfied that the confession was truthful. Wheat decided the information had to be disclosed, and Long signed a consent form for disclosure. Wheat then informed the Warden, Pecos County law enforcement authorities, Willis and Willis's counsel at the time, of Long's confession. On September 11, 1990, Deputy Jackson, one of the primary investigators of the Iraan fire, conducted a nearly three-hour long videotaped interview of Long.

Long and Willis first met during recreation time when Long asked Willis where he was from; Willis answered Pecos. Long said he knew Billy Willis from Pecos and Petitioner Willis said Billy Willis had testified at his trial. At this point, Long realized Petitioner Willis was convicted of the crime Long committed. Petitioner Willis was then transferred to a work program and so Willis and Long no longer communicated at recreation time. Long requested a legal visit with Willis but decided not to say anything until he saw how Willis's direct appeal resolved. Long contemplated not saying anything until the hour of his own execution. Long requested a second legal visit at which time Long asked about Willis's direct appeal. Willis said his conviction and sentence were affirmed. At this point, Long told Willis that he committed the Iraan fire.

Prior to the interview, Jackson read Long his Miranda rights.

The substance of Long's confession is as follows: Long set the fire because he wanted to hurt or kill Billy Willis, Petitioner's cousin. Billy and Long were longtime associates who participated in various criminal activities together, usually drug related. On June 10, 1986, Long drove to Iraan from Round Rock, Texas, where Long lived, to purchase some drugs from Billy. In his pick-up truck, Long carried a half-gallon bottle of Wild Turkey alcohol mixed with Everclear grain alcohol and some methamphetamine. Long arrived in Iraan sometime between 2:00 a.m. and 4:00 a.m. He parked his truck about a block away from the Robinson house where Billy was staying. He sat in the truck for about twenty minutes drinking the Wild Turkey and Everclear mixture and injecting himself with methamphetamine. He then went into the house with the Wild Turkey and Everclear mixture.

Long testified that as he was in the house he became overcome with anger, and poured the Wild Turkey and Everclear mixture on the carpet around the dining room table and around the living room. Long did not pour any of the mixture on the couch where Willis was sleeping, because he did not want to wake him. Long then used his Bic lighter to ignite some clothing draped over a piece of furniture in the living room. After setting the fire, Long left the house, returned to his truck, and drove a couple of blocks down the street. He then left Iraan. Long stated he used the same method to start the fire in Iraan as he did to start a fire in Bay City, Texas, that also killed someone. Finally, during his confession, Long described the Robinson house in great detail.

Long stated that "the feeling started coming over me, the bitterness that I have toward Billy, which I had not ever went down into detail about, things that happened in the past. And when this happens to me, I kind of like get locked in my mind and things go black and white, and I started feeling an extreme bitterness toward him, because at one time I was going to shoot him . . . because of some things that happened in the past. . . ." Pet. at 22, citing Def. E.H. Ex. 4 at 14, ll. 17-24 (Long).

Mrs. Amelia Fuentes, who lived across the street from the Robinson house, saw a vehicle traveling slowly past her house on Fifth Street before any of the police or fire vehicles arrived. She had never seen the vehicle before. During the investigation of the fire, Mrs. Fuentes told Deputy Jackson about the vehicle. He told her to forget about it. Pet. at 23, citing, Tr. at 146, ll. 19-23 (Fuentes), May 23, 1996.

The modus operandi of both fires was similar. Long set fire to the Bay City victim's trailer using liquor as an accelerant, as he claimed he did in the Iraan fire. The reason Long gave for killing the victim of the Bay City fire and for attempting to kill Billy Willis was the same, that he held a grudge against both and snapped in their presence. The Bay City fire confession was used by the State in Long's capital murder trial for a triple axe murder for which he was convicted and sentenced to death. Furthermore, in Long's direct appeal, the Court of Criminal Appeals upheld the admission of the Bay City confession and stated it was corroborated by other witness' testimony. Long v. State, 823 S.W.2d 259, 268 n. 12 (Tex.Crim.App. 1991).

C. Willis's Evidence Contradicting the State's Theory of the Fire

This opinion provides only a short summary of the evidence presented during the state post-conviction hearing that negates the State's theory of the fire. A full description of the evidence presented and a description of Mr. Smyth's qualifications and methodology can be found at Pet. at 25-36.

At the state post-conviction hearing, Marshall Smyth, a fire investigator, testified for Willis. Smyth's testimony corroborates Long's accounts, shows that the State's theory of the case was mistaken and supports Willis's version of the events. The State had a "pour pattern" theory of the fire, meaning that in every area of the house where there was burn damage, an accelerant had been poured. Under this theory, Willis could not have run out of the house because the floor would have been in flames. According to the pour pattern theory, Willis would have had to spread accelerant in or near bedrooms and exits for the fire to burn as it did.

Deputy Jackson, one of the State's arson experts, admitted during his interview of Long that he is "not much of a [sic] arson investigator." Pet. at 26, citing, Def. E.H. Ex. 4 at 74, ll. 5-6.

Smyth testified that the pour pattern theory was physically impossible, and that the burn damage to the house could not have been caused by an accelerant such as gasoline. Instead, Smyth testified that the burn damage throughout the house was the result of "flashover" conditions throughout the house during various points in the fire. Smyth also testified that, consistent with Long's account of the fire, the maximum "area of origin" of the fire was the living room and dining room. Smyth's tests showed that it would have taken approximately ten to eleven minutes for the fire to spread from the chair where it was ignited, according to Long, to the surrounding carpeting soaked with the alcohol mixture. That period of time is consistent with Willis's description that, once awakened, he ran through the house and then exited through the front door without serious injury.

"Flashover is a transition point, actually, in the development of a fire inside a compartment or room. And it's the point at which the burning materials in this fire became so strong that they form a gas cloud under the ceiling of the room. And this gas cloud thickens up. And at some point temperatures on the floor are raised to their — the ignition point of the materials due to the radiation of the heat from the gas cloud. And, at that point, all the combustible materials in the room essentially simultaneously burst into flame. So it's that transition point in the buildup of a fire from something less than full room involvement to the point where all the materials in the room are involved in the flame." Pet. at 28, citing Tr. at 32, ll. 9-33 (Smyth), Jan. 12, 1998.

Pet. at 28, n. 12, citing Tr. at 112, ll. 7-14 (Smyth), Jan. 12, 1998.

Other evidence disputes the State's theory of the case. The clothes Willis wore on the night of the fire were submitted to the State's lab; no accelerant was found on the clothes. In addition, accelerant was not found on the carpet samples from the Robinson house that were submitted to the lab, and the State never produced any evidence regarding the type of accelerant used to start the fire, according to the State's theory. Finally, consistent with Willis's statement regarding his actions upon discovering the fire — that he awoke to the house already on fire and ran around trying to rouse others — Willis left in the house his boots, socks and pain medication.

In addition, Willis argues that his post-fire demeanor, which the State used as evidence of guilt during Willis's trial, can be explained in a manner that also supports his innocence. At the time of the fire, Willis was receiving care for chronic back pain and had undergone four back surgeries as a result of injuries suffered in past years as an oil field worker. Willis's prior back surgeries resulted in a chronic condition called "arachnoiditis." As a result, at the time of the fire, Willis was taking prescription opiate drugs like Talwin and Percodan to make his back pain tolerable. Two days before the fire, on June 9, Willis went to the emergency room because of excruciating back pain. He was given an injection of Butorphanal and some Phernergan. Later that day he took at least nine 50 mg. tablets of Talwin.

For exact quantities of the drugs taken, see Pet. at 39-40.

The next day, the day before the fire, at around 4:00 a.m., Willis again went to the emergency room, where he was given a dose of Demerol and some Emet-Con, a drug for treatment of nausea. He returned to the Robinson house at 7:00 a.m., took three tablets of Talwin and a muscle relaxant. Two hours later, he was still in pain, and he took another three tablets of Talwin. An hour later he went to see Dr. Edwin Franks, who gave him a steroid injection and a prescription for additional back pain medications. Throughout the rest of the day, Willis took at least six more Talwin tablets and one Percodan tablet. In addition, between 6:00 p.m. and midnight, Willis drank approximately six cans of beer, and took more Talwin and Percodan before going to sleep.

Pet. at 40, citing Lipman Depo. at 17, ll. 18-19, Jun. 8, 1998. The levels of pain medication that Willis took on June 9-10, 1986 were not unusual for chronic back pain patients. Id. at 20, ll. 1-15.

Willis contends the drugs he took in the two days before the fire would have affected his outward appearance in the time period immediately after the fire. Specifically, he claims the drugs would make him appear unemotional and unexcited. Also, he claims the alcohol consumed would have contributed to his low affect and to the suppression of his coughing after the fire.

Id. at 25, ll. 9-27.

The State asserted at trial that Willis's account of the fire was not believable because Willis was not injured. Two days after the fire, Willis did have a very bad burn on his shoulder, but the State claimed the burn was not present the day of the fire and thus was not caused by the fire. During the state post-conviction hearing, Willis put forth evidence that blistering does not necessarily occur immediately as a result of thermal burning and thus the appearance of the burn on Willis two days after the fire was not unusual.

Pet. at 40, citing Lipman Depo. at 77, ll. 1-7, Jun. 8, 1998.

One of the state investigators, Deputy Jackson, testified at trial that Willis's account that he ran around the outside of the house breaking windows in an effort to help the people still inside could not be truthful because glass was found only on the outside of the house. Willis claims the windows to the Robinson house were a particular type that prevented the glass from falling into the house. Willis claims that the windows consisted of two panels, a lower portion and an upper. When opened, he claims, the lower portion slides above the upper portion, creating two layers of glass. Willis claims the windows were open the night of the fire and that when he broke the upper part of the window, the lower part, as a second layer, prevented the glass from falling inside the house.

Pet. at 42, citing Trial Tr., vol. 19 at 140, ll. 8-13 (Deputy Jackson).

Finally, Willis argues that there was no motive to support the State's theory of the fire and that Willis had no motive to set the fire. Willis argues that at no point in the investigation of the fire, the trial or the post-conviction proceedings did the State produce evidence of any motive. And Willis, who was forty-two years old at the time of the crime, had never before been charged with a violent crime.

D. Analysis of Willis's Innocence Claim

The state trial court rejected a finding of innocence in this case. The state trial court found that Willis "failed to produce sufficient evidence to corroborate the statement of Mr. Long," and thus found Mr. Long's confession inadmissible. At post-conviction proceedings, the state trial court therefore held that "the testimony in the record does not support a finding that Willis is innocent."

Ex Parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 6.

The state trial court does not cite any authority requiring that or explaining why the confession must be corroborated to be admissible. However, under Texas law, an extrajudicial confession of wrongdoing, standing alone, is not sufficient to support a conviction; other evidence must exist, demonstrating that a crime has in fact been committed. See Rocha v. State, 16 S.W.3d 1, 4 (Tex.Crim.App. 2000).

Ex Parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 33.

In Herrera v. Collins, the Supreme Court held that "[c]laims of actual innocence based on newly discovered evidence have never been held to state a ground for federal habeas relief absent an independent constitutional violation occurring in the underlying state criminal proceeding." In Herrera, the Court did assume "for the sake of argument . . . that a truly persuasive demonstration of `actual innocence' made after trial would render the execution of a defendant unconstitutional." Since Herrera, the lower courts dispute whether federal habeas relief is available based on a showing of innocence without a constitutional error at trial. While the Ninth and Seventh Circuits held that habeas relief is available based upon a post-conviction showing of innocence alone, the Fifth Circuit rejected this rule and holds that newly discovered evidence related to innocence is not sufficient grounds alone for habeas relief. Willis acknowledged that even if this Court found innocence, relief would nevertheless be unavailable to him under the law of this Circuit.

Id. at 417.

See Jackson v. Calderon, 211 F.3d 1148, 1164 (9th Cir. 2000), cert denied, 531 U.S. 1072 (2001); Carriger v. Stewart, 132 F.3d 463, 476 (9th Cir. 1997) (en banc), cert denied, 523 U.S. 1133 (1998); Milone v. Camp, 22 F.3d 693, 699 (7th Cir. 1994), cert denied, 513 U.S. 1076 (1995).

Lucas v. Johnson, 132 F.3d 1069, 1074 (5th Cir. 1998) (holding that "the existence merely of newly discovered evidence relevant to the guilt of a state prisoner is not a ground for relief on federal habeas corpus."). See also Robinson v. Johnson, 151 F.3d 256, 267 (5th Cir. 1998), cert denied, 526 U.S. 1100 (1999). The Fourth Circuit has likewise refused to recognize an actual innocence claim alone. See Royal v. Taylor, 188 F.3d 239, 243 (4th Cir. 1999).

The State did not address any of the factual allegations of innocence proffered by Willis. Instead, the State claims that because actual innocence is not a cognizable claim in habeas, Willis's innocence claim is barred by the nonretroactivity rule of Teague v. Lane. Teague prevents application of novel rules of law to petitioners whose convictions are final. There are two exceptions to the Teague rule. The first exception occurs when a new rule of law places "certain kinds of primary, private individual conduct beyond the power of the criminal law-making authority to proscribe." The second exception occurs when the new rule of law "requires the observance of those procedures that are implicit in the concept of ordered liberty."

489 U.S. 288 (1989).

See Williams, 529 U.S. at 380.

Teague, 489 U.S. at 307 (internal citations omitted).

Id.

If the Supreme Court were to find that an innocence claim were cognizable in habeas, this Court has no doubt that, for a petitioner who could make a showing of actual innocence, the first Teague exception would apply, and thus Teague would not bar relief. But under this Circuit's current jurisprudence, innocence alone is not a sufficient basis for federal habeas relief. While both parties' presentations to the Court in cross-motions for summary judgment raise strong reason to be concerned that Willis may be actually innocent, under Herrera and Lucas, innocence is not a cognizable claim in habeas; thus, it would be inappropriate for this Court to determine the issue. In any event, the determination is unnecessary because the Court must grant Willis's writ on other grounds.

The Teague exceptions are not part of section 2254(d)'s deference provisions. The Supreme Court has not yet resolved the tension between Teague and section 2254 in that regard.

Herrera v. Collins, 506 U.S. 390, 400 (1993); Dowthitt v. Johnson, 230 F.3d 733, 741-42 (5th Cir. 2000), cert. denied, 532 U.S. 915 (2001).

IV. Administration of Medically Inappropriate Antipsychotic Medications

During the evidentiary hearing on Willis's state habeas petition, evidence and testimony were presented concerning his claim that the State's wrongful administration of antipsychotic drugs denied Willis of due process and other constitutional rights. At the conclusion of the hearing, Judge Jones entered detailed findings of fact regarding the administration of the medication by the State, the effect on Willis and the lack of any justification for the medication. These findings were summarized above. Judge Jones then entered conclusions of law recommending relief be granted on the claim.

Judge Jones held that the administration of antipsychotic medication to Willis during his trial denied him the ability to assist in his own defense in violation of his right to counsel, and prejudicially affected his demeanor at trial in violation of substantive due process rights. In addition, the trial court held that the State can only administer medication to a defendant involuntarily if the standard articulated by the Supreme Court in Riggins is met: 1) administration of the drugs was "medically appropriate and, considering less intrusive alternatives, essential for the sake of [the defendant's] own safety or the safety of others; 2) administration of the drugs was medically appropriate and that the prosecution could not "obtain an adjudication of [the defendant's] guilt or innocence by using less intrusive means;" or 3) that the administration of medication was "necessary to accomplish an essential state policy."

See Riggins v. Nevada, 504 U.S. 127, 133, 142 (1992).

See id. at 131.

Id. at 135-36, 138. The state trial court based its analysis of this claim largely on Riggins. The CCA denied the claim based on a prior CCA opinion interpreting Riggins. In addition, both parties have extensively briefed Riggins. Though not raised by the state trial court, the CCA or either party, the Court notes that Riggins was decided in 1992, two years after Willis's conviction became final on direct appeal, on October 9, 1990. However, the Supreme Court's decision in Washington v. Harper, 494 U.S. 210, 222 (1990), was decided on February 27, 1990, before Willis's conviction became final. As explained in the text of this opinion, Harper explicitly states that State administered antipsychotic drugs must be medically appropriate. Furthermore, subsequent Supreme Court cases — namely Riggins and United States v. Sell, 539 U.S. 166 (2003) — state that the rule of law emanated from Harper. In Penry v. Lynaugh, 492 U.S. 302 (1989) ( Penry I), the Supreme Court held that dicta in Jurek v. Texas, 428 U.S. 262 (1976), established law for Teague purposes. Thus, the statements in Harper — that due process requires that state administered antipsychotic drugs be medically appropriate — are sufficient for Teague purposes in Willis's case, even if they are dicta.

Judge Jones found that the administration of the drugs to Willis was not medically appropriate, not essential for the safety of Willis or others, and not necessary to accomplish an essential state policy. Furthermore, Judge Jones held a showing of prejudice was not required because under Riggins, there is a "strong possibility" that trial defense was impaired.

Ex Parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 13, quoting Riggins, 504 U.S. at 138.

Judge Jones also found that the administration of antipsychotic medications to Willis violated Willis's right to confront witnesses because a defendant's physical presence and demeanor in the courtroom are essential to the exercise of his confrontation rights. The medication given to Willis left him unable to confer with counsel and unable to exhibit any emotive response to the testimony of adverse witnesses. Furthermore, Willis was prevented from reacting or responding to the proceedings and was not able to demonstrate sensitivity or compassion.

Id., citing Riggins, 504 U.S. at 142. See also Coy v. Iowa, 487 U.S. 1012, 1020 (1988).

Ex Parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 13.

Judge Jones found the administration of the medication also violated a number of other constitutional rights. First, the medicine prevented Willis from assisting in his own defense and denied him his Sixth Amendment right to the effective assistance of counsel. Willis was unable to communicate with counsel or make tactical decisions during trial. Thus the administration of the medication was an actual or constructive denial of the right to counsel by the State and not subject to a prejudice showing. Second, the medication of Willis and his resulting demeanor effectively forced Willis to testify against himself in violation of the Fifth Amendment. This is especially so because the prosecution used Willis's demeanor as evidence of his guilt. Finally, Judge Jones found the administration of the medication violated Willis's right to an individualized capital sentencing determination.

Id. at 14-16.

The Court of Criminal Appeals overruled the trial court's recommended relief on this claim in one paragraph. Citing a Texas case, the CCA held that because there was no motion to terminate medication or an objection to the medication in the record, Willis "has not demonstrated his treatment was involuntary." The CCA based its ruling on a legal determination that a showing of involuntariness requires an objection in the record. Because the CCA's overruling of the trial court was not inconsistent with the trial court's findings, but instead, a determination that the facts did not warrant relief under the legal standard, this Court must defer to the state trial court's findings of fact. Thus, the issues before this Court are: 1) whether the CCA's holding that an objection is a necessary condition for a finding of involuntariness is contrary to, or an unreasonable application of, clearly established federal law and 2) whether the CCA's implicit determination that the State can administer antipsychotic medication to pre-trial inmates with no established medical need is contrary to, or an unreasonable application of, clearly established federal law.

Ex Parte Willis, No. 27, 787-01, Order at 2 (Tex.Crim.App. 2000), citing Ex Parte Thomas, 906 S.W.2d 22 (Tex.Crim.App. 1995).

Though the CCA addressed Willis's medication claim on the merits, the CCA did not address the lack of medical justification for the antipsychotic drugs the State administered to Willis. However, the state trial court made a finding of fact as to the lack of medical justification, and Willis raised the lack of medical justification on appeal to the CCA as part of his claim for relief. Thus, the CCA's rejection of Willis's medication claim is an implicit finding that the lack of medical justification is not a ground for relief.

As explained below, the CCA erred on both grounds. First, the State cannot administer antipsychotic drugs unless medically appropriate according to Supreme Court holdings, and thus the CCA's denial of relief when Willis was medicated with antipsychotic drugs without medical justification is contrary to clearly established federal law. Second, the CCA's determination that an objection is a necessary condition of involuntariness is contrary to clearly established federal law regarding waiver of constitutional rights. In this case, no evidence exists that either Willis or defense counsel knew of the existence or nature of the medication.

A. Administration of Medically Inappropriate Drugs

After the state habeas hearing, the trial court found the State administered the antipsychotic drugs to Willis without any medical need. This determination is supported by the record. Three experts testified during the state habeas hearing that they were unable to find any evidence of psychosis or other mental disorder in Willis's medical or behavioral history. Furthermore, none of the records from the Pecos County Jail indicate that Willis was suffering from a psychotic disorder or exhibiting symptoms of psychosis. Numerous medical intake forms for Willis's admission to the Pecos County Jail state that Willis had never been treated for mental illness. One of the forms was filled out after Willis received antipsychotic medication in the Pecos County Jail. Disciplinary records from the Pecos County Jail state that Willis is negative for a history of mental illness. The report of the psychological exam administered to Willis at the time of trial stated there was no evidence that Willis was psychotic. Additionally, Willis's eleven-year records from TDC do not contain any evidence of a psychotic disorder.

Ex Parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 11.

See e.g., Crowder Dep.; Lipman Dep.; Cunningham Dep.

See e.g., Tr. at 267, ll. 1-14 (Cunningham).

Def. E.H. Ex. 30.

Def. E.H. Ex. 29 at 93, 117, 123, 129, 142; Crowder Dep., ll. 16-21.

Tr. at 177, ll. 2-14 (Wright).

Lipman Dep. at 44, ll. 16; Crowder Dep. at 42, ll. 18-23 ("We never see any psychosis appear in his extensive TDC records."); Crowder Dep. at 51, ll. 16-21.

A significant liberty interest exists in avoiding unwanted administration of antipsychotic drugs under the Due Process Clause of the Fourteenth Amendment. But, due process will allow "a mentally ill inmate to be treated involuntarily with antipsychotic drugs where there is a determination that . . . the treatment is in the inmate's medical interest."

Harper, 494 U.S. at 222; Parham v. J.R., 442 U.S. 584, 600-601 (1979). See also Youngberg v. Romeo, 457 U.S. 307, 316 (1982) (core liberty protected by due process, freedom from bodily restraint, survives criminal conviction, incarceration and involuntary commitment).

Riggins, 504 U.S. at 135 (internal quotations and citations omitted) (stating the Supreme Court's holding in Harper, 494 U.S. at 227).

In upholding a state procedure for involuntary medication of antipsychotic drugs in Washington v. Harper, the Supreme Court was careful to recognize that the state procedure required that the administration of medication be medically appropriate. Because the state procedure at issue in Harper recognized the petitioner's medical interests, it met the requirements of the Due Process Clause. In a lengthy footnote, the Court detailed that it would not adopt the State's procedure if the procedure did not require a finding of medical appropriateness before antipsychotic medication can be involuntarily administered.

Id. at 223.

Id. at 223, n. 8. See also id. at 227 (holding that the Due Process Clause permits the State to treat a prison inmate "who has a serious mental illness with antipsychotic drugs against his will, if the inmate is dangerous to himself or others and the treatment is in the inmate's medical interest.") (emphasis added). "[W]e hold that the regulation before us is permissible under the Constitution. It is an accommodation between an inmate's liberty interest . . . and the State's interests in providing appropriate medical treatment. . . ." Id. at 236 (emphasis added). The dissent in Harper explains the majority's decision as follows: "[A]lthough the Court does not find, as Harper urges, an absolute liberty interest of a competent person to refuse psychotropic drugs, it does recognize that the substantive protections of the Due Process Clause limit the forced administration of psychotropic drugs to all but those inmates whose medical interests would be advanced by such treatment." Id. at 243 (Stevens, J., dissenting).

The rule of Harper was reiterated in Riggins where a state involuntary medication procedure was found inadequate. "Under Harper, forcing antipsychotic drugs on a convicted prisoner is impermissible absent a finding of overriding justification and a determination of medical appropriateness." The Riggins Court noted that a pretrial detainee — the petitioner in Riggins — enjoyed as much constitutional protection as the convicted prisoner at issue in Harper.

504 U.S. 127 (1992).

Id. at 135 (emphasis added) (internal quotations and citations omitted).

Id., citing Bell v. Wolfish, 441 U.S. 520, 545 (1979).

Applying the rule of Harper to the procedures employed by the State, the Riggins Court held that the State "certainly would have satisfied due process if the prosecution had demonstrated, and the District Court had found, that treatment with antipsychotic medication was medically appropriate and, considering less intrusive alternatives, essential for the sake of Riggins' own safety or the safety of others." Alternatively, the State "might have been able to justify medically appropriate, involuntary treatment with the drug by establishing that it could not obtain an adjudication of Riggins' guilt or innocence by using less intrusive means."

Id. (emphasis added).

Id. (emphasis added).

The rule of Harper was reaffirmed again in United States v. Sell. There, the Supreme Court addressed whether a State may forcibly administer antipsychotic drugs to a criminal defendant in order to render him competent to stand trial. The Court held a four-part test must be met to involuntarily medicate a criminal defendant: 1) important governmental interests are at stake; 2) involuntary medication will significantly further those interests; 3) involuntary medication is necessary to further those interests; and 4) "administration of the drugs is medically appropriate, i.e., in the patient's best medical interest in light of his medical condition."

539 U.S. 166 (2003).

Id. at 177.

Id. at 180.

Id. at 181. "[T]he court must conclude that involuntary medication will significantly further those concomitant state interests. It must find that administration of the drugs is substantially likely to render the defendant competent to stand trial. At the same time, it must find that administration of the drugs is substantially unlikely to have side effects that will interfere significantly with the defendant's ability to assist counsel in conducting a trial defense, thereby rendering the trial unfair." Id., citing Riggins, 504 U.S. 142-45 (emphasis in original).

Id. ("[t]he court must find that any alternative, less intrusive treatments are unlikely to achieve substantially the same results. And the court must consider less intrusive means for administering the drugs.").

Id.

The State argues that Harper and Riggins are not on point because both cases addressed situations in which a formal objection was made to the medication. The argument is inapposite. The Court in both Harper and Riggins assumed the medication was medically appropriate. Thus, assuming the medication is medically appropriate, the issue in both cases became what procedures must the State go through in order to medicate an inmate against his will. In this case, on this record, no such assumption can be made. Indeed, the record does not support that assumption, and the state court found that in fact the medication was not medically appropriate. Harper followed by Riggins and Sell make clear that medical appropriateness is always a condition precedent to the involuntary administration of antipsychotic drugs to inmates.

The State's motion for summary judgment was written before the Supreme Court's decision in Sell, 539 U.S. 166.

Harper, 494 U.S. at 223, n. 8 (". . . we will not assume that physicians will prescribe these drugs for reasons unrelated to the medical needs of the patients; indeed, the ethics of the medical profession are to the contrary."); Riggins, 504 U.S. at 133 (". . . we presume that administration of [antipsychotic drugs] was medically appropriate.").

Because Supreme Court precedents are unequivocal that antipsychotic medication administered by the State must be medically appropriate, the CCA's rejection of Willis's due process claim, when the record is clear that Willis was medicated with no medical need, is contrary to clearly established federal law.

B. Whether a Showing of Involuntariness Requires an Objection

The Court now addresses whether the CCA's holding that an objection is a necessary condition for a finding of involuntariness is contrary to, or an unreasonable application of, clearly established federal law. The state trial court made a factual finding that Willis did not consent to the medication. This finding was not rejected by the Court of Criminal Appeals. Instead, the CCA stated that because there was no objection on the record, Willis could not make a legal showing of involuntariness.

Ex Parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 9.

The State argues that the medication administered in both Harper and Riggins was determined involuntary because the inmate had objected to it on the record. But in neither of those cases did the Court require a recorded objection as a necessary element to a showing of involuntariness. The State also cites Richardson v. Johnson, and Adanandus v. Johnson, but neither of those cases included a finding of non-consent. Furtermore, in Adanandus, there was no finding that the petitioner had actually been medicated. Thus, neither case is instructive.

256 F.3d 257, 259 (5th Cir. 2001).

947 F. Supp. 1021, 1084 (W.D. Tex. 1996).

Id.

In all the cases uncovered by the Court in which antipsychotic medication was found to be voluntary, there was evidence in the record that the recipient knew of the medication and often requested it. There is no such evidence in the record for Willis's case. Also, the antipsychotic medication was given without medical need, strongly indicating that it was not just given involuntarily but also given without Willis's knowledge. The Court finds it unlikely that a reasonable and competent person would voluntarily take high doses of unnecessary antipsychotic drugs without evident medical need.

See e.g., Ex Parte Thomas, 906 S.W.2d 22; Fearance v. Scott, 56 F.3d 633 (5th Cir. 1995), cert. denied, 515 U.S. 1153 (1995); Adanandus v. Johnson, 947 F. Supp. 1021 (W.D. Tex. 1996).

In stating that a showing of involuntariness can only be made through an objection, the CCA cited only one case, its own decision in Ex Parte Thomas. There, the defendant initially requested the medication and later claimed to object to it. Defense counsel in that case was aware of the medication. Thus, the facts surrounding the voluntariness of the medication in Ex Parte Thomas are quite different than the facts surrounding involuntariness in the instant case.

Though not specifically found by the state trial court in post-conviction findings, there is evidence in the record that Willis was not aware he was taking antipsychotic medication. Willis was receiving several medications each day for back pain. The State notes that when Willis was given the medication, he placed his initials on the medication log sheet. The record though does not demonstrate that Willis knew the initials indicated anything other than receipt of his back pain medication, and because he expected to receive the back medication, Willis would not have had reason to question the medication. Because the State medicated Willis with antipsychotic drugs in the absence of any medical need, Willis would have had no reason to suspect the drugs were antipsychotics. The initials do not suggest Willis understood what medication he was receiving.

See Pet. at 78-81.

Ex Parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 11-12.

Also, Willis's Pecos County Jail medical records did not meet statutory requirements. See Pet. at 81, n. 37; 37 TEX. ADM IN. CODE § 273.4. See also Lipman Dep. at 38, ll. 16-39, l. 3 ("I can find no pharmacologically appropriate basis for [the] prescription" of the anti-psychotic medication to Willis in the Pecos County Jail records or other supporting documents.).

While the Supreme Court has not discussed the standard for involuntariness specifically in the context of involuntary medication, the Court has developed a standard for involuntariness used generally in a number of other contexts. In the context of right to counsel, the Supreme Court held that "[p]resuming waiver from a silent record is impermissible. The record must show, or there must be an allegation and evidence which show, that an accused was offered counsel but intelligently and understandably rejected the offer. Anything less is not waiver." The Supreme Court applied this standard for waiver to the guilty plea context. Also, the Supreme Court rejected state laws that denied the application of the right to speedy trial unless the defendant demanded trial, and instead the Court applied the same standard articulate above to the analysis of a waiver of the right to a speedy trial.

Carnely v. Cochran, 369 U.S. 506, 516 (1962).

Boykin v. Alabama, 395 U.S. 238, 242 (1969) (the Court noted that several constitutional rights are involved in a waiver that accompanies a guilty plea).

Barker v. Wingo, 407 U.S. 514, 524 (1972).

Id. at 526 (applying the standard used in Carnely, 369 U.S. at 516 and Boykin, 395 U.S. at 242).

Thus, the ordinary rule is that a court cannot infer a waiver of a constitutional right from the failure to object. In light of the constitutional rights implicated when a defendant is medicated with antipsychotic drugs, there is no reason to deviate from this established standard for waiver, nor is any such explanation given by the CCA. Because the CCA impermissibly deemed the medication voluntary from a silent record, a determination that Willis's medication was voluntary is an unreasonable application of clearly established Supreme Court precedents on waivers of constitutional rights.

Id. at 525 (". . . presuming waiver of a fundamental right from inaction, is inconsistent with this Court's pronouncements on waiver of constitutional rights.").

See Riggins, 504 U.S. at 142 (Kennedy, J. concurring) (noting that side effects of antipsychotic drugs can compromise the right of a criminal defendant to receive a fair trial. "The drugs can prejudice the accused in two principal ways: (1) by altering his demeanor in a manner that will prejudice his reactions and presentation in the courtroom, and (2) by rendering him unable or unwilling to assist counsel." Id. Justice Kennedy also stated that medication with antipsychotic drugs can effect a defendant's constitutional rights, his right to testify on his own behalf and his right to counsel. Id. at 142, 144.).

The trial court determined that the failure to object did not constitute consent.

The State then argues that even if the medication were involuntarily administered, Willis has not shown he was prejudiced because he has not demonstrated he was harmed in any manner. However, in Riggins, the Supreme Court held that once it has been established that a defendant was involuntarily medicated during a criminal trial without the proper due process considerations, because of the "substantial probability of trial prejudice," prejudice is presumed. Additionally, the Supreme Court's decisions in both Riggins and Harper recognized the severe effects of antipsychotic medications and the potentially debilitating effects of such medication on an accused's constitutional trial rights. The Supreme Court noted that it is possible for side effects to impact outward appearance, the content of testimony, the ability to follow the proceedings, the substance of communication with counsel, and comprehension at trial. Nevertheless, it is clear from the state trial court's findings of fact that Willis was actually prejudiced, both because of the effect of the medication on Willis's demeanor and because the prosecution used Willis's demeanor as evidence of guilt and future dangerousness. As to the effect on Willis's demeanor, the state court found Willis exhibited flat or little facial expression, inexpressiveness, rigidity of the facial muscles, a fixed gaze, drowsiness, confusion and diminished ability to communicate. Willis's demeanor was "markedly different" at the post-conviction hearing, when the antipsychotic drugs were no longer being given. As to the the prosecution's use of Willis's demeanor as evidence of guilt and future dangerousness, the trial court found the State asked the jury to infer guilt and propensity for future dangerousness from Willis's lack of feeling or emotion. Therefore, the Court finds that Willis was actually prejudiced by the State's administration of the antipsychotic drugs.

Riggins, 504 U.S. at 138. See also Sell, 539 U.S. at 189 (Scalia, J. dissenting) ("the Riggins Court held that forced medication of a criminal defendant that fails to comply with Harper creates an unacceptable risk of trial error and entitles the defendant to automatic vacatur of his conviction.").

Riggins, 504 U.S. at 138. The State argues that the Riggins presumption of prejudice only applies on direct review, not in post-conviction proceedings. The State provides no authority to support this argument. Furthermore, presumptions of prejudice have been used in other post-conviction contexts. See Burdine v. Johnson, 262 F.3d 336, 348-50 (5th Cir. 2001) (en banc), cert. denied, Cockrell v. Burdine, 535 U.S. 1120 (2002). Also, this part of the State's argument seems to challenge the state trial court's findings of fact regarding the effects of the medication on Willis. However, the State does not mention that the state trial court made findings of fact regarding this issue, nor argue that those findings are unreasonable in light of the evidence presented. See 28 U.S.C. § 2254(e)(1).

Harper, 494 U.S. at 229-30 (identifying the "serious, even fatal, side effects" of antipsychotic drugs). See also Riggins, 504 U.S. at 134; Sell, 539 U.S. at 185-86 (Whether a particular drug will tend to sedate a defendant, interfere with communication with counsel, prevent rapid reaction to trial developments, or diminish the ability to express emotions are matters important to determinating the permissibility of medication).

Riggins, 504 U.S. at 137.

Ex parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 10-11.

Ex Parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 10.

Id. at 11.

The State also argues that even if the medication were involuntary and harmful, it was medically necessary. As discussed above, the state trial court made detailed findings of fact that the medication of Willis was without medical need and those findings are properly before this Court. To the extent that the State challenges the finding that the administration of medication lacked necessity, the State fails to engage in the requisite analysis outlined in the AEDPA. The State has not rebutted the presumption of correctness afforded to state court factual findings by clear and convincing evidence, and a review of the record reveals that the factual findings of the state court are reasonable in light of the evidence presented.

For the reasons provided above, the medication of Willis during trial violated his right to due process, both because it was without medical need and also because it was involuntary. Willis is entitled to relief on the claim because the CCA's denial of the claim was contrary to, and an unreasonable application of, clearly established federal law.

Willis argues that the State's administration of the medication violated a number of other constitutional rights: right to confront witnesses, remain free from self-incrimination, effective assistance of counsel, and an individualized sentencing determination. These arguments were raised to the state trial court and to the CCA. The trial court found that the administration of medication violated all these rights. The CCA did not address any of these additional constitutional claims. Because this Court has granted relief on due process grounds, the Court declines to address the other bases for relief.

V. Prosecutorial Suppression of Evidence

Prosecutorial suppression of evidence that is favorable to an accused "violates due process where the evidence is material either to guilt or punishment, irrespective of the good faith or bad faith of the prosecution." A defendant need not request such evidence to trigger the prosecutor's duty to disclose. To establish a Brady claim, a petitioner must demonstrate that 1) the prosecution suppressed or withheld evidence 2) favorable to the defense and 3) material to guilt or punishment.

Brady v. Maryland, 373 U.S. 83, 87 (1963).

United States v. Bagley, 473 U.S. 667, 682 (1985).

East v. Johnson, 123 F.3d 235, 237 (5th Cir. 1997).

Evidence is material if "there is a reasonable probability that, had the evidence been disclosed to the defense, the result of the proceeding would have been different." Four aspects of materiality govern the inquiry. First, a petitioner need not prove by a preponderance that disclosure of the suppressed evidence would have resulted ultimately in a sentence less than death. "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. A reasonable probability of a different result is accordingly shown when the government's evidentiary suppression undermines confidence in the outcome of the trial."

Kyles v. Whitley, 514 U.S. 419, 433 (1995).

See id. at 434.

Id.

Id. (internal citations and quotations omitted).

Second, the materiality analysis is not a sufficiency of the evidence test.

A defendant need not demonstrate that after discounting the inculpatory evidence in light of the undisclosed evidence, there would not have been enough left to convict. . . . One does not show a Brady violation by demonstrating that some of the inculpatory evidence should have been excluded, but by showing that the favorable evidence could reasonably be taken to put the whole case in such a different light as to undermine confidence in the verdict.

Id.

Id. at 435.

Third, harmless error analysis does not apply. And, fourth, materiality is assessed in terms of all suppressed evidence considered collectively, not item by item.

Id.

Id. at 436.

Judge Jones found that the State violated Brady by affirmatively or negligently failing to turn over the Wright report to the defense. The CCA overturned Judge Jones, stating that the Wright report was not favorable or material. The CCA did not question the trial court's determination that the Wright report had been suppressed, nor did it reject the trial court's findings of fact. The CCA based its ruling on a determination that the facts, as found by the trial court, did not meet the standard of favorability or materiality. Because the CCA's overruling of the trial court was not inconsistent with the trial court's factual findings, this Court must defer to those trial court findings of fact.

See Craker, 756 F.2d at 1213-14; Westley, 83 F.3d at 721 n. 2.

The CCA determined the Wright report was not favorable for two reasons: first, because at the evidentiary hearing Wright testified that he was unable to gather sufficient information during the examination of Willis to make a future dangerousness determination, and second, because the conclusions in the report were "hypotheticals." Dr. Wright's report states that "the data I was able to collect concerning Willis was [sic] insufficient for determining whether he would pose a continuing threat to society." The CCA offered no authority for the proposition that a report with a conditional conclusion fails the Brady standard for determining whether evidence is favorable.

Ex Parte Willis, No. 27, 787-01, Order at 4.

Willis argues that the CCA unreasonably applied Brady in finding the Wright report was not favorable. The Wright report contained two hypothetical scenarios, differing on the issue of the nature of the evidence produced at trial. One of the scenarios was favorable and one was not. The favorable scenario was: if sworn evidence indicates that his behavior until the time of the current alleged offense was no worse than previous behaviors, we could probably say with safety that the current alleged behavior was an isolated event which he probably will not repeat. The other scenario was as follows:

Recent years may have seen more and more irresponsibility or increasingly violent behaviors toward others. If testimony reflects this to a significant degree, we would certainly seem correct in assessing that he has passed through a behavioral door and that he will continue to commit vicious, violent type behaviors. A deterioration over the years would certainly seem to suggest that he would represent a continued threat to society.

Wright Report at 6. See Pet. at 166.

Id.

The State presented no evidence during the penalty phase of the trial that would have triggered the second scenario. The only prior criminal history presented a trial involved non-violent offenses.

Ex Parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 6.

In support of his argument that the CCA erred in holding the Wright report to be not favorable, Willis relies upon a Fifth Circuit case, holding that evidence meets the Brady standard of materiality, if it is both inculpatory and exculpatory. Willis also argues that the CCA's determination on favorability was unreasonable because it ignores the ongoing nature of the State's obligations under Brady. The State's obligation to produce Brady material continues throughout trial. Willis argues that the Wright report was clearly favorable and should have been disclosed because the hypothetical scenarios in the report were conditioned on the evidence presented at trial and that evidence did not ultimately include other violent behaviors.

See Sellers v. Estelle, 651 F.2d 1074, 1077 (5th Cir. 1981), cert. denied, 455 U.S. 927 (1982).

Jackson v. Johnson, 194 F.3d 641, 649 n. 18 (5th Cir. 1999), cert. denied, 529 U.S. 1027 (2000), citing United States v. Miranne, 688 F.2d 980 (5th Cir. 1982), cert. denied, 459 U.S. 1109 (1983).

See Ex Parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 4; Ex Parte Willis, No. 27, 787-01 Order at 4 (stating that the testimony presented during the penalty phase was "relatively brief with two law enforcement officers providing reputation testimony.").

Willis also suggests the Wright report was favorable because, even if the report itself were inconclusive, disclosure of the report would have led the defense to Dr. Wright, whose testimony would have been favorable. In determining whether evidence is material under Brady, the effect of the suppression of the evidence on the preparation or presentation of the defense case is relevant. The suppression of inadmissible evidence is material if the disclosure of the inadmissible evidence might have led defense counsel to admissible evidence.

Bagley, 473 U.S. at 683.

Sellers, 651 F.2d at 1077 n. 6; Spence v. Johnson, 80 F.3d 989, 1005 n. 14 (5th Cir. 1996), cert. denied, 519 U.S. 1012 (1996).

With these guidelines in mind, the Court finds that Wright's testimony would have been favorable and the prosecution's failure to disclose the Wright report violated Willis's due process right. As offered by the State in its Motion for Summary Judgment, Dr. Wright was "committed to his opinion that Willis would not pose a future danger when he testified during the state hearing in 1998." Furthermore, Dr. Wright visited with the District Attorney about his examiniation of Willis and said: "I didn't think this was a good death penalty case." Dr. Wright reiterated his belief that Willis's case was not a good death penalty case at the state habeas hearing.

Resp.'s Mot. Summ. J. at 76.

Id. at 77.

Id. (noting that Dr. Wright answered "yes" at the state habeas hearing when asked whether this was not a good death penalty case).

The State responds that the Wright report was not favorable because it contained negative information about Willis's drinking habits and convictions for obscene phone calls and drunk driving. The report contained information that Willis admitted to drinking after age seventeen, that Willis was accused of indecent exposure at age seventeen and several times later, that Willis was convicted twice for obscene phone calls, and that Willis was convicted four or five times for driving while intoxicated. Willis's convictions for driving while intoxicated and a felony conviction for "immoral conduct" were already before the jury. Thus, the only additional negative information contained in the report was the indecent exposure accusations. Considering that the report led to the highly favorable testimony of a state-sanctioned medical expert, who determined that Willis was not a future danger, the Court finds the overall character of the report is favorable, even though it also contained unfavorable information. The jury had to answer a specific question on future dangerousness to impose the death penalty, and the report would have favorably addressed this issue. The Wright report's overall character is favorable.

See Willis, 785 S.W.2d at 387.

The CCA also found that, even if the report were favorable, it was not material because no expert testimony was presented during the penalty phase on the issue of future dangerousness and because the penalty phase was relatively brief, with two law enforcement officers providing reputation testimony. Because a challenge to the sufficiency of the evidence on future dangerousness was raised and rejected on direct appeal, the CCA found that "in view of the evidence presented at trial, it is exceedingly difficult to conclude applicant has demonstrated that there is a reasonable probability the jury would have returned a negative answer on the future dangerousness finding if they had been aware of Wright's report." The CCA found Dr. Wright's report "inconclusive" and found that Willis had made no showing that the "verdict is unworthy of confidence."

Ex Parte Willis, No. 27, 787-01 Order at 4.

Id.

Id., citing Kyles, 514 U.S. at 433-35.

Willis argues that the CCA's conclusion on materiality should be rejected because it was "contrary to clearly established law." This Court agrees. The CCA's finding that the Wright report failed to meet the materiality standard was erroneous because it took into account the sufficiency of the evidence, in direct contrast to Kyles v. Whitley. There, the Supreme Court explicitly stated that the materiality analysis under Brady is not a sufficiency of the evidence test.

Id. See also Williams, 529 U.S. at 414 (O'Connor, J., concurring) (recognizing that the Virginia Supreme Court also applied the appropriate Strickland standard); East, 123 F.3d at 239 ("The Supreme Court has warned that the Brady materiality analysis is not a sufficiency of evidence test.").

Willis also argues that the CCA's use of the sufficiency of the evidence test to reject the materiality of the Wright report is contrary to clearly established federal law, even if the CCA did not exclusively rely on that test. In Williams v. Taylor, the Supreme Court held that, because it was impossible to tell how much the state court's use of the wrong standard affected its final determination, the state's determination was contrary to law.

Williams, 529 U.S. at 414. The State argued in Williams that even though the Virginia Supreme Court relied on the incorrect standard, the analysis was not contrary to law because the Virginia court had also cited Strickland Brief of Resp. in Williams v. Taylor, N 98-8384, 1999 WL 642451 at *37-38.

Similarly, the CCA's use of the incorrect legal standard is particularly problematic in this case because the two other factors the CCA used to judge materiality were also questionably applied. In holding that the report was not material, the other factors considered by the CCA were 1) no expert testimony was presented at trial on the issue of future dangerousness and 2) the punishment phase was "relatively brief with two law enforcement officers providing reputation testimony." The materiality standard depends "almost entirely on the value of the evidence relative to the other evidence mustered by the State." Thus, the fact that the evidence admitted at the penalty phase was limited — devoid of any expert testimony and consisting solely of two witnesses, two Pecos County law enforcement officers who provided conclusory and unsubstantiated descriptions of Willis's reputation in unspecified communities — supports, rather than undermines, a finding of materiality.

Ex Parte Willis, No. 27, 787-01 Order at 4.

Spence, 80 F.3d at 995. See also United States v. Agurs, 427 U.S. 97, 112 (1976).

The State argues that the report was not suppressed because defense counsel should have obtained it themselves and did not exercise due diligence in attempting to acquire the report. To establish a Brady violation, a petitioner must show that the information allegedly withheld was not available through due diligence. In support of its argument on this point, the State argues facts expressly rejected by the state trial court. Under section 2254(e)(1), state court findings of fact are presumed to be correct, and the party rebutting the presumption of correctness must do so by clear and convincing evidence. The State does not claim the state trial court's factual findings should not be presumed correct. Moreover, the state trial court's finding that the report had been suppressed under Brady was not rejected by the CCA.

United States v. Mmahat, 106 F.3d 89, 94 (5th Cir. 1997).

Pondexter, 346 F.3d at 146. See also Burden v. Zant, 498 U.S. 433, 436 (1991) (per curiam) (finding that presumption of correctness of state court fact findings applies when factual determination supports petitioner as well as when factual determination supports the State); Valdez v. Cockrell, 274 F.3d 941, 947 (5th Cir. 2001).

Nonetheless, for the following reasons, this Court finds that the state trial court's findings of fact are supported by the record. First, prior to trial, defense counsel successfully moved for disclosure of all evidence relevant to mitigation or exoneration of Willis. Second, although defense counsel was aware of psychological evaluation for the purpose of determining competency, counsel was not told and the prosecution did not reveal that an assessment of Willis's future dangerousness had also been conducted. Defense counsel must have actual notice that a psychological examination will encompass the issue of future dangerousness. Considering that the State was obliged to inform defense counsel of the scope of the evaluation, defense counsel did not fail to meet the standard of due diligence by relying on the State's representations regarding the scope of the examination. Furthermore, contrary to the State's assertion, the record supports the trial court's finding that Attorney DeHart did not receive the Wright report. Thus, the Wright report was not available through due diligence of defense counsel.

Pet.'s Reply at 58, citing Blank Aff., Ex. 8, 9.

Ex Parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 4.

Powell v. Texas, 492 U.S. 680, 685 (1989); Satterwhite v. Texas, 486 U.S. 249, 255-56 (1988).

Strickler v. Greene, 527 U.S. 263, 283-84, 288 (1993); Banks v. Dretke, 124 S.Ct. 1256, 1273 (2004) (petitioner cannot be faulted for relying on State's representations).

Defense Attorney Woolard testified that DeHart did not receive the report. Dr. Wright was not contacted by DeHart. Dr. Wright did not forward a copy of the report to DeHart. The testimony eliminates the possibility that the State gave DeHart a copy of the report because Prosecutor Johnson claims he did not know the Wright report existed.

The Court finds that the Wright report was suppressed and was both favorable and material under clearly established law. Moreover, the disclosure of the report would have led defense counsel to Dr. Wright's favorable testimony. That additional benefit to defense counsel further supports a finding that the report is both favorable and material. The Wright report presented an opinion by a qualified mental health expert, approved and hired by the State, who believed Willis was not a good candidate for the death penalty and who would have testified that Willis was not a future danger. Considering the lean evidence the State presented at the penalty phase, had the jury been aware of Dr. Wright's conclusions, there is a reasonable probability that at least one juror would have answered "no" to the question on future dangerousness, and Willis would not have been sentenced to death. Absent Dr. Wright's report and testimony, the Court does not have confidence in the outcome of the penalty phase.

Cf. East, 55 F.3d at 1003 (Prosecution had a duty to disclose a punishment phase witness' rapsheet because if the prosecution had revealed it, defense counsel would have investigated the witness' criminal history and eventually uncovered the witness' mental records in the files of the Bexar County Court.).

During a deposition before the state habeas hearing, the lead trial prosecutor, J.W. Johnson denied that he had ever met or heard of Dr. Wright at the time of Willis's trial. Evidence produced during the state habeas hearing showed that Johnson had worked with Dr. Wright on two other cases before Willis's trial. Johnson could not explain why, if Dr. Wright was not conducting the examination at the request of the State, Willis was given Miranda warnings before the examination.

See Kirkpatrick v. Whitley, 992 F.2d 491, 497 (5th Cir. 1993).

Because of the numerous errors the CCA made in addressing this claim: applying the sufficiency of the evidence test for materiality; erroneously stating that the brief nature of the evidence presented at the penalty phase undermined, rather than supported, a finding of materiality; and failing to consider that disclosure of the report would have led to the favorable testimony of Dr. Wright, the CCA's finding that the Wright report was not favorable was contrary to and an unreasonable application of clearly established federal law.

VI. Ineffective Assistance of Counsel

The constitutional standard for determining whether a criminal defendant has been denied the effective assistance of counsel was announced by the Supreme Court in Strickland v. Washington. "The benchmark for judging any claim of ineffectiveness must be whether counsel's conduct so undermined the proper functioning of the adversarial process that the trial cannot be relied on as having produced a just result." A two-prong test guides the inquiry:

466 U.S. 668 (1984).

Id. at 686. See also Nealy v. Cabana, 764, F.2d 1173, 1177 (5th Cir. 1985).

First, the defendant must show that counsel's performance was deficient. This requires showing that counsel made errors so serious that counsel was not functioning as the "counsel" guaranteed the defendant by the Sixth Amendment. Second, the defendant must show that the deficient performance prejudiced the defense. This requires showing that counsel's errors were so serious as to deprive the defendant of a fair trial, a trial whose result is reliable.

Strickland, 466 U.S. at 687.

Courts are extremely deferential in scrutinizing the performance of counsel and make every effort to eliminate the distorting effects of hindsight. It is strongly presumed that counsel rendered adequate assistance and made all significant decisions in the exercise of reasonable professional judgment. An attorney's strategic choices informed by a thorough investigation of relevant facts and law are virtually unchallengeable. Thus, Willis must overcome a strong presumption that the conduct of his trial counsel falls within a wide range of reasonable professional assistance.

See Lockhart v. Fretwell, 506 U.S. 364, 372 (1993); Burger v. Kemp, 483 U.S. 776, 789 (1987); Strickland, 466 U.S. at 689; Green v. Johnson, 116 F.3d 1115, 1122 (5th Cir. 1997).

See Strickland, 466 U.S. at 690; Duff-Smith v. Collins, 973 F.2d 1175, 1182 (5th Cir. 1992).

See Boyle v. Johnson, 93 F.3d 180, 187-88 (5th Cir. 1996).

See Strickland, 466 U.S. at 687-91; Belyeu v. Scott, 67 F.3d 535, 538 (5th Cir. 1995).

To establish he has sustained prejudice, Willis "must show that there is a reasonable probability that, but for counsel's unprofessional errors, the result of the proceeding would have been different. A reasonable probability is a probability sufficient to undermine confidence in the outcome."

Strickland, 466 U.S. at 694; Cantu v. Collins, 967 F.2d 1006, 1016 (5th Cir. 1992).

A. The Texas CCA's Analysis

The state trial court held that Willis was entitled to relief under Strickland The CCA overruled the trial court's recommendation of relief on this basis. The CCA divided the analysis of ineffective assistance for Willis's two trial attorneys: Attorney DeHart and Attorney Woolard. However, the CCA cited no federal authority requiring a petitioner to show that each attorney's conduct separately meets the Strickland standard as opposed to the defense representation as a whole. Citing its own case, the CCA stated that "[i]n view of the multiple counsel representation of applicant, it was incumbent upon applicant to prove deficient performance by all counsel." The CCA also stated that the record did not reflect the two defense attorneys' respective duties, responsibilities and division of labor.

Ex Parte Willis, No. 27, 787-01, Order at 5 (citing McFarland v. State, 928 S.W.2d 482 (Tex.Crim.App. 1995)).

For Attorney DeHart, the CCA conducted an overview of DeHart's background. The CCA mentioned that Woolard had faith in DeHart's ability, that he had been licensed for twenty-one years at the time of Willis's trial, that he had previously been employed as an Assistant District Attorney for four years, that he was then the Presiding Judge of the 384th District Court in Alpine, and that he was considered a "seasoned veteran," due to his criminal law experience. Thus, the CCA held that on the record before it, Willis could not overcome the presumption that DeHart provided effective assistance of counsel.

For Attorney Woolard, the CCA noted that Woolard had been licensed to practice law for four years, and that Willis's case was his first capital trial. The CCA also stated that Woolard was surprised Willis was found guilty, and that Woolard had "loaded his guns" for the guilt-innocence phase and decided not to present mitigation evidence. The CCA mentioned that Woolard spoke with a number of Willis's friends and relatives and that Investigator Caspari also spoke with friends and relatives. Then the CCA found that Woolard made all significant decisions in the exercise of reasonable professional judgment. Thus the CCA held that Willis did not overcome the presumption that Woolard provided effective assistance of counsel.

Id. at 6.

Id. at 5.

The CCA's overruling of the trial court was consistent with the trial court's factual findings. The CCA relied on the record from the post-conviction trial court but attached a different legal significance to facts found by that court. For example, both the CCA and the trial court noted that defense counsel spoke with a number of friends and relatives of Willis in preparation for the penalty phase. The CCA also found that facts that were not relied upon by the trial court, such as defense counsel's experience, were legally significant. Furthermore, the CCA based its decision in part on its legal determination that Willis was required to show that each defense counsel individually met the standard for ineffectiveness. Because the CCA's resolution of the claim is not directly contrary to the trial court's factual findings, this Court must, as detailed above, defer to the state trial court's findings of fact.

See id. at 6; Ex Parte Willis, No. 27, 787-01, Find. of Fact and Conc. of Law at 20.

Before addressing Willis's specific allegations of ineffectiveness, the Court finds that the CCA violated clearly established federal law in holding that Willis had to show each attorney's performance, as opposed to the defense representation as a whole, met the Strickland standard. Strickland does not require that the applicable analysis be conducted separately for each attorney. Furthermore, later Supreme Court opinions applying Strickland, in which the petitioner was represented by more than one attorney at trial, conduct one Strickland analysis for the performance of defense counsel as a whole. There is no support for the CCA's holding that Willis must prevail on separate analyses of deficient performance and prejudice for each attorney. The CCA's ruling in this regard was therefore contrary to clearly established law.

Strickland, 466 U.S. at 687.

See Williams, 529 U.S. at 370 (alleging "trial attorneys had been ineffective during sentencing"); Wiggins v. Smith, 539 U.S. 510 (2003) (engaging in one Strickland analysis for petitioner's two defense attorneys, two public defenders in the same office).

Respondent's Motion for Summary Judgment states that Willis cannot prevail on his ineffective assistance of counsel claim because he did not present any testimony from Attorney DeHart at the state habeas hearing. There is no post- Strickland case requiring the testimony of both trial counsel as a prerequisite to an ineffectiveness claim. On the contrary, federal law requires that the analysis for an ineffectiveness claim is conducted as to defense counsel performance as a whole, not separately for each attorney. Thus, Respondent's argument in this regard fails.

B. Ineffective Assistance of Counsel At the Guilt-Innocence Phase

First, the Court considers Willis's allegations that defense counsel's performance was deficient on various grounds during the guilt-innocence phase of trial. The Court then separately considers the issue of prejudice as required by Strickland

1. Failure to Investigate Demeanor Failure to Discover Unnecessary Medication

The CCA overruled the trial court without addressing Willis's substantive allegation of ineffectiveness based upon defense counsel's failure to investigate the jail records or discover the unjustified use of antipsychotic medications. As stated above, the CCA based its overruling of the trial court on defense counsel's legal experience and its legal determination that Willis was required to show that each attorney met the standard for ineffectiveness.

The other factors mentioned by the CCA are relevant to defense counsel's performance during the penalty phase.

After the habeas hearing, the state trial court found that defense counsel recognized a problem with Willis's demeanor and suspected the problem could be related to medication. Despite counsel's awareness and suspicion, Judge Jones found defense counsel made no effort or inquiry to determine the cause of Willis's appearance or demeanor, even though defense counsel had the right to access Willis's medical records and it is "rudimentary" and "basic" for counsel to gather records. Willis now claims this failure to investigate constituted deficient performance and ineffective assistance of counsel.

Ex Parte Willis, No. 27, 787-01, Find. of Fact and Conc. of Law at 17.

In response, the State first argues counsel was not unreasonable to believe that Willis's flat affect and lack of emotion was caused by medications for his back pain. The State points to the Pecos County Jail medical log, which reflects Willis took a number of medications for back pain. The medical log does not support the State's argument as to defense counsel's belief because defense counsel did not obtain Willis's Pecos County Jail medical records. Defense counsel could not have known what medications Willis was taking, for back pain or otherwise. Nor could defense counsel have known the effect or potential effect of those medications. Therefore, counsel could neither have based an understanding of Willis's manner on that information, nor have made strategic trial decisions based thereon.

Id. at 16-17.

The critical failing of counsel with respect to Willis's demeanor was the failure to pursue or in any manner respond to counsel's admitted concern over Willis's demeanor, whether by gathering Willis's jail medical records or speaking with an expert. Strickland requires that the Court defer to counsel's decisions when those decisions are both fully informed and strategic, in the sense that it is expected, on the basis of sound legal reasoning, to yield some benefit or avoid some harm to the defense. Defense counsel cannot make informed or strategic decisions in the absence of a reasonable investigation and thus Strickland does not require deference to decisions that are not informed by an adequate investigation into the controlling facts and law. Interpreting Strickland, the Supreme Court stressed that a decision based on less than a complete investigation is reasonable only to the extent that the limits on the investigation were reasonable.

Cf., Roberts v. Dretke, 356 F.3d 632, 639 (5th Cir. 2004). "Where, as here, counsel is aware of the client's history of mental problems, the reasonableness of a decision made by counsel not to investigate that history is suspect." Id.

Moore v. Johnson, 194 F.3d 586, 615 (5th Cir. 1999).

Id. See also Andrews v. Collins, 21 F.3d 612, 623 (5th Cir. 1994) (counsel's strategic decision entitled to deference because supported by an adequate investigation which included contact with at least twenty-seven people); Drew v. Collins, 964 F.2d 411, 423 (5th Cir. 1992) (counsel's strategic decision entitled to deference because counsel made "reasonable inquiries" into defendant's mental state); Bouchillon v. Collins, 907 F.2d 589, 597 (5th Cir. 1990) ("Tactical decisions must be made in the context of a reasonable amount of investigation, not a vacuum."); Wiggins, 539 U.S. at 533 ("`strategic choices made after less than complete investigation are reasonable' only to the extent that `reasonable professional judgments support the limitations on investigation.'") (citation omitted).

Wiggins, 539 U.S. at 533.

Neither the State, Willis, the state trial court, nor the CCA articulated any benefit to the defense case from Willis's being medicated with unnecessary antipsychotic drugs. To the contrary, the harm to Willis is well-documented, as discussed previously. Defense counsel could not have made a decision about the benefits or risks of Willis's medication because counsel did not go to the minimal effort required to investigate Willis's demeanor, that is, to gather Willis's jail medical records and discover he was being unnecessarily medicated. In this case, the limits on investigation — the failure to gather the jail medical records — are not merely unreasonable. Considering counsel's admitted concern for Willis's demeanor, the limits on investigation here are beyond explanation. Counsel's failure to address or rectify Willis's demeanor is thus not entitled to a presumption of reasonableness because it was neither informed by a reasonable investigation nor supported by any logical position that such failure would benefit Willis's defense, and thus cannot possibly be construed as strategic.

See Moore, 194 F.3d at 616.

The Court finds that defense counsel's failure to investigate Willis's demeanor was deficient performance under Strickland Counsel's failure to investigate Willis's demeanor was objectively unreasonable because: 1) counsel was concerned with Willis's demeanor; 2) counsel could have addressed that concern by obtaining Willis's jail medical records but did not do so, even in light the standard that gathering medical records is a "basic" part of defense counsel's duties in a capital case; and 3) no strategic decision supported the failure to gather the medical records.

The Court also finds that the CCA's rejection of this claim was an unreasonable application of Strickland In addition to errors made by the CCA already discussed, the CCA's determination that counsel made all significant decisions in the exercise of reasonable professional judgment is unreasonable. The CCA did not assess whether the failure to gather the jail medical records actually demonstrated reasonable professional judgment. Courts may not defer to decisions by counsel that are not strategic or are not informed by a reasonable investigation or reasonable limits on investigation. The CCA's assumption that the failure to investigate was adequate was thus an unreasonable application of clearly established federal law.

Strickland is clearly established federal law within the meaning of 28 U.S.C. § 2254. See Wiggins, 539 U.S. at 522 (referring to the "`clearly established' precedent of Strickland"); Dowthitt, 230 F.3d at 743 ("the merits of an ineffective assistance of counsel claim are governed by the well-established rule of Strickland v. Washington.").

See Wiggins, 539 U.S. at 527.

See Strickland, 466 U.S. at 690-91; Wiggins, 539 U.S. at 528; Moore, 194 F.3d at 615.

See Wiggins, 539 U.S. at 528.

2. Failure to Object to Prosecution's Use of Willis's Demeanor at Guilt-Innocence Phase

Willis also contends trial counsel violated his right to effective assistance of counsel by failing to object to the prosecutor's reference to his trial demeanor during closing arguments. Willis raises four statements by the prosecution as the basis for his claim: 1) reference to Willis's "dead pan, insensitive, expressionless face;" 2) description of Willis's "cold fish eyes on everybody and everything that has come in here, and he just merely stared and watched very impassively, very cold heartedly, much like he probably did that morning outside the fire when he watched and listened;" 3) commenting that "[t]his guy has been able to sit in here and observe everyone that took the stand, look at all of you throughout this proceeding;" and 4) stating that "[y]ou know, it's hard for us to even imagine the perverted thoughts and the fascination this Defendant must have had standing out there . . . observing and knowing what was going on inside . . . What kind of thoughts go through somebody's mind like that? You know, what he was thinking when he is watching this satanic deed that he did? People burning up in there . . . That's what he was doing, listening and watching . . . And he showed no mercy or remorse afterwards."

Vol. 28 at 83, ll. 1-3.

Vol. 28 at 83, ll. 8-12.

Vol. 28 at 65, ll. 14-16.

Vol. 28 at 82, ll. 4-24.

Before addressing the substance of Willis's arguments relative to these remarks, the Court finds two unworthy of review. Willis challenged the third remark on direct appeal. The CCA found the third remark was not a comment on Willis's demeanor but juxtaposed Willis's presence at trial with the absence of the deceased victim. The Court likewise finds that this prosecutorial remark was not a comment on Willis's trial demeanor, and therefore, should not be included in this analysis.

See Vol. 28 at 64, ll. 13-21 ("My clients aren't in the courtroom today. They are dead. Understand the distinction. . . .").

Next, the State argues Willis is barred from presenting the fourth remark because he did not cite the remark during the state habeas process. Although the Court will not consider the fourth remark for reasons explained below, the remark is not barred, under the Texas abuse-of-writ doctrine, as the State argues. The State relies upon two cases, both of which are properly distinguished from the instant case, to support its argument.

In Anderson v. Harless, the Supreme Court held that a claim was not exhausted when it was raised as a state law issue to the state courts, and thus the corresponding federal constitutional claim had not been presented to the state courts. Willis's case is distinguished from Anderson because Willis presents a federal claim relying on federal law. Therefore, the Court will not eschew consideration of the fourth remark based upon Anderson. In Nobles v. Johnson, the petitioner presented in the state courts a Sixth Amendment claim that he had been denied the effective assistance of a competent court-appointed psychiatrist. In federal court, the petitioner raised a claim of ineffective assistance of counsel based on failure to present mitigating evidence. Nobles argued the "gist" of the claims was the same and he should therefore be able to present the federal court with the "re-postured" claim. The Court rejected Nobles's argument, and held that when the two claims required "wholly different inquiries," the petitioner had not provided the state court with the requisite "fair opportunity to apply controlling legal principles to the facts bearing upon his constitutional claim." However, such is not the case for Willis's claim. Here, the state court was given the opportunity to consider precisely the same legal claim with the same facts. In Willis's case, the difference in the federal petition is the addition of supplemental factual examples of prosecutorial comments. Including new facts in a federal habeas petition does not render the federal claim based upon those facts unexhausted unless the facts materially alter the legal claim presented to the state courts. The facts must be material and must put the claim in a significantly different and stronger evidentiary posture than it was when presented to the state courts.

Anderson v. Harless, 459 U.S. 4, 6 (1982).

Nobles v. Johnson, 127 F.3d 409 (5th Cir. 1997).

Id. at 420.

Id.

Id. (internal citations omitted).

The additional facts are not new facts in the sense that the examples were part of the trial record that was presented to the trial court during the post-conviction hearing.

Vasquez v. Hillery, 474 U.S. 254, 260 (1986).

Dowthitt, 230 F.3d at 745-46 (finding a petitioner's claim exhausted despite the presentation of two additional expert psychological reports that were not presented to the state courts).

Willis exhausted his claim with regard to the fourth remark because its addition does not materially alter the legal claim presented to the state court, but the addition of the fourth remark does not place Willis's federal claim in a stronger evidentiary posture. It is a less dramatic example of prosecutorial comment on non-testimonial demeanor than either the first or second remarks. Consequently, the Court finds that the fourth remark is not material and does not make Willis's claim significantly stronger or different.

Because the fourth remark does not add to the claim, the Court will not consider the remark in determining the merits of Willis's claim. The merits of the claim will therefore be determined on the basis of the first and second remarks only.

To begin, Willis must demonstrate that counsel's performance fell below an objective standard of reasonableness. Willis argues that under Texas or federal law, the prosecutor's remarks constituted error, and thus, a reasonable defense attorney would have objected. Under state law, Willis argues that the CCA found error when the prosecution commented on the defendant's non-testimonial demeanor by describing the defendant as "cold, unnerved, uncaring . . . [and] unsympathetic." Willis argues that defense counsel's failure to object was objectively unreasonable because, under this precedent, the trial court would have committed reversible error by refusing to sustain an objection.

See Strickland, 466 U.S. at 687.

Good v. State, 723 S.W.2d 734, 736 (Tex.Crim.App. 1986).

See Vaughn v. State, 931 S.W.2d 564, 567 (Tex.Crim.App. 1996).

The State responds that prosecutorial comment on a defendant's non-testimonial demeanor is not error according to the Supreme Court. The State confuses the legal standard for reviewing a state court's determination of a claim under 28 U.S.C. § 2254(d), which requires a showing that the state court unreasonably applied clearly established federal law, with the standard for ineffective assistance of counsel. The proper inquiry is whether a reasonably effective attorney would have objected to the prosecutor's statements, not whether the prosecutorial statements themselves violated clearly established federal law.

See Bishop v. Wainwright, 511 F.2d 664, 667 (5th Cir. 1975) (prosecutor's comments about defendant's courtroom demeanor raise no habeas corpus issue).

See Strickland, 466 U.S. at 668.

On Willis's direct criminal appeal, the CCA held that the comments were improper under state law. On habeas review, the state trial court found that the prosecution commented on Willis's non-testimonial demeanor, that the prosecution urged jurors to infer lack of remorse from the non-testimonial demeanor and that defense counsel failed to object. Willis argues defense counsel's performance was deficient under the first prong of Strickland because a reasonable attorney would have objected to the comments as improper 1) under state law, given the CCA's determination on direct appeal that the prosecutor's comments violated state law, and 2) under federal law, as a violation of Willis's fundamental right against self-incrimination protected by the Fifth Amendment.

Willis, 785 S.W.2d at 386 n. 8.

Ex Parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 17.

To the extent that the State argues that the failure to object was not deficient performance because the objection would have been futile or without merit, the Court disagrees. The objection would have been neither futile nor meritless. To the contrary, the CCA determined on direct appeal that the prosecutor's comments violated state law, and therefore defense counsel's objection would have been objectively reasonable. An objectively reasonable attorney would have objected to the prosecutorial comments as improper under state law. Moreover, because of the CCA's determination on direct appeal, a determination that defense counsel's failure to object was sufficient performance would have been unreasonable under Strickland, had the CCA applied federal law to this particular allegation of ineffectiveness. Because a reasonable attorney would have objected to the comments as improper under state law, it is not necessary for the Court to decide whether a reasonable attorney would have objected under federal law. The Court holds that defense counsel performed deficiently under the first prong of Strickland

See Morlett v. Lynaugh, 851 F.2d 1521, 1525 (5th Cir. 1988), cert. denied, 489 U.S. 1086 (1989).

Willis, 785 S.W.2d at 386 n. 8.

3. Prejudice at the Guilt-Innocence Phase

The Court now considers whether Willis was prejudiced by his trial counsel's deficient performance during the guilt-innocence phase. The Court views together all instance of deficient performance by defense counsel during the guilt-innocence phase to determine whether Willis was prejudiced. To establish prejudice, Willis must show a reasonable probability exists that, but for counsel's unprofessional errors, the result of the proceeding would have been different.

See Williams, 529 U.S. at 399, 416 (holding that the state trial court was correct in determining prejudice based on "the entire post-conviction record, viewed as a whole and cumulative of mitigation evidence presented originally, and faulting the Virginia Supreme Court for its piecemeal approach to the ineffectiveness claim."); Moore, 194 F.3d at 619 (considering the cumulative errors of counsel and finding prejudice).

See Kimmelman v. Morrison, 477 U.S. 365, 375 (1986); Darden v. Wainwright, 477 U.S. 168, 184 (1986); United States v. Conley, 349 F.3d 837, 841-42 (5th Cir. 2003); Williams v. Collins, 16 F.3d 626, 631 (5th Cir. 1994); and United States v. Bounds, 943 F.2d 541, 544 (5th Cir. 1991).

Had defense counsel conducted a reasonable investigation into Willis's demeanor, or at the least gathered his jail medical records, counsel would have learned that Willis was being medicated, absent medical need, with inappropriately high doses of antipsychotic drugs. And, as stated in the section addressing Willis's involuntary medication claim, Willis was severely prejudiced by the administration of the unnecessary antipsychotic medications. The Supreme Court has recognized the harm that can arise from a defendant being medicated with antipsychotic drugs during trial.

See Riggins, 504 U.S. at 142 (Kennedy, J., concurring). "It is a fundamental assumption of the adversary system that the trier of fact observes the accused throughout the trial, while the accused is either on the stand or sitting at the defense table. . . . At all stages of the proceedings, the defendant's behavior, manner, facial expressions, and emotional responses, or their absence, combine to make an overall impression on the trier of fact, an impression that can have a powerful influence on the outcome of the trial. . . . The side effects of antipsychotic drugs may alter demeanor in a way that will prejudice all facets of the defense. . . . As any trial attorney will attest, serious prejudice could result if medication inhibits the defendant's capacity to react and respond to the proceedings and to demonstrate remorse or compassion."); Coy, 487 U.S. at 1016-20 (emphasizing the importance of the face-to-face encounter between the accused and the accuser).

In addition, here the State used Willis's demeanor and flat affect as an argument in support of his guilt. The state trial court found that the State referred to Willis's demeanor during trial as evidence of guilt and future dangerousness and that the State urged jurors to infer a lack of remorse based on Willis's demeanor. These factual findings, to which this Court must defer, further support that Willis was prejudiced by the deficient performance of counsel in failing to investigate Willis's demeanor or determine the medication that cause the demeanor.

The State also argues that Willis cannot prevail on his ineffective assistance claim grounded on counsel's failure to investigate Willis's demeanor and failure to detect the antipsychotic medications because he has not shown that had counsel investigated Willis's demeanor, counsel would have found an expert available to testify at that time regarding the alleged impropriety of antipsychotic medications. Testimony presented at Willis's post-conviction hearing demonstrated that, based on 1987 standards, the medication given to Willis was medically inappropriate, and Judge Jones found as much in fact. The Court finds that a reasonably qualified expert in 1987 would have testified to such and reasonably effective defense counsel would have obtained one.

Ex Parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 16-17; Lipman Dep. at 33, ll. 11-17; 52, ll. 3-53; 37, ll. 21-38, ll. 14-54; Tr. at 252, ll. 20-24; Tr. at 268, ll. 1-5.

Therefore, the Court finds that Willis was prejudiced by defense counsel's failure to investigate his demeanor. The CCA's determination that Willis was not prejudiced is objectively unreasonable considering the clarity of the Supreme Court's jurisprudence on the potential harm of medicating criminal defendants with antipsychotic drugs, as well as the evidence in the record regarding the harm to Willis. The deficiencies in counsel's performance during the guilt-innocence phase rendered the proceeding fundamentally unfair and the result of the proceeding unreliable. Willis received ineffective assistance of counsel during the guilt-innocence phase because Willis's trial counsel were deficient — by failing to investigate his demeanor and by failing to object to the prosecution's reference to his demeanor to establish guilt and future dangerousness — and because Willis was prejudiced by these deficiencies.

See Riggins, 504 U.S. at 127; Harper, 494 U.S. at 210; Sell, 539 U.S. at 176-77.

The CCA found that Willis did not demonstrate deficient performance of counsel, and thus, the CCA did not substantively analyze the prejudice requirement of Strickland beyond simply stating that Willis had failed to show prejudice. See Ex Parte Willis, No. 27, 787-01, Order at 5.

See Soffar v. Dretke, 368 F.3d 441, 471 (5th Cir. 2004), citing Lockhart, 506 U.S. at 372.

C. Ineffective Assistance at the Sentencing Phase

The Court turns now to Willis's claims of ineffective assistance of counsel at the sentencing phase.

1. Failure to Investigate and Discover the Wright Report

Willis argues counsel was ineffective for failing to investigate and discover the report of Dr. Wright, the psychologist who examined Willis before trial at the request of the prosecution. As detailed above, the Wright report indicated that Willis was not a future danger. For the reasons outlined in the following section addressing claims of prosecutorial suppression, the Court holds that the prosecution suppressed the Wright report. And therefore, defense counsel's performance was not deficient, nor counsel ineffective, for failing to investigate that which the State bore a duty to disclose and that which was hidden from the defense.

2. Failure to Object to the State's Descriptions of Willis as an Animal

Willis argues that counsel's failure to object to prosecutorial comments characterizing Willis as an animal constituted ineffective assistance of counsel. The state trial court found that the prosecution characterized Willis as a "pit bull," and "animal," and a "rat," during voir dire, closing arguments and at the penalty phase. During voir dire the following exchange took place between Prosecutor Johnson and one juror. Defense counsel did not object.

Vol. 5 at 15, ll. 4, 13-16.

Q: Okay. Well, let me give you a hypothetical here now. You are aware of that case out in San Diego where that old boy went to a McDonald's and killed 16 people in about 30 minutes.

A: Right.

Q: Did they ever develop a motive for that man going berserk?

A: No. I don't believe?

Q: Okay. There can be a lot of speculation.

A: Right.

Q: But unless that person tells you, you don't know.

A: That's right.

Q: And that's what I need to know from you. Are you going to require yourself to know why they did something?
A: No. I don't believe so. As long as they did it, I believe I would go ahead and vote for it.
Q: We get back to the premise that actions speak louder than words.

A: Right.

Q: Okay. Because these — you have been reading about these pit bull attacks?

A: Right.

Q: You know, we don't need — you can't talk to the dog and find out why it wanted to eat the little four year old baby, can you?

A: Right.

Q: You know it's a mean, vicious dog, and it's capable of hurting, crippling, and killing people?

A: Right, sir.

Q: And once it shows it has that propensity to do that to a human being, you want to find out why the dog went off its rocker and started doing that or you take action?

A: I think I would take action on that.

Q: Okay. I think most of us will, but I want to make sure that you understand that the motive of this Defendant in doing this act and premeditation are not elements that the State is required to prove in this courtroom to gain a guilty conviction and to gain a death sentence.

A: Right, sir.

Q: Okay?

A: Okay.

Prosecutor Johnson had the following exchange with another juror. Defense counsel did not object.

Vol. 4 at 76, ll. 13-77, 15.

Q: Okay. But when it comes to proof, now, his motive isn't one of them. That's not going to bother you?
A: I don't think so, if I have enough, like I said, enough proof to know that he did it.
Q: Okay. Because there are lots of times people do things and they don't tell you why they did it. Even though you want to know, they ain't going to tell you why they did it.
A: Yeah, I understand that. I'm that way to sometimes. I do things.
Q: But when that happens — and we don't know why it happened, and they won't tell us, or it is an animal and it hurt somebody, and it can't tell us either.

A: Right.

Q: But when that happens and we don't know what the motive was, we just say the actions of that person or animal speak loud and clear, don't they?

A: Right.

Q: That's when we go by the actions rather than you are going to explain it or say about it or whatever the words may be.

A: Uh-huh.

Q: Okay. That's all we are coming in here and doing is showing you this Defendant's actions on June 11, 1986, that resulted in the death of this woman. That's going to be all right?

A: Okay?

Q: Okay. Because we can't get into his mind.

A: Right.

Q: And, of course, he doesn't have to take the stand either and tell you why he did it because he has a right to remain silent. Can you go along with that?

A: Yeah.

Prosecutor Johnson also questioned another juror as follows. Defense counsel objected to this statement.

Vol. 11 at 64, ll. 13-24.

Q: . . . You have two children, eight and twelve. If they were playing out in the front yard and some person you had never seen before was walking a pit bull dog and that pit bull dog breaks his leash and attacks your eight your [sic] old and gets him down, hurts him real bad, you come running out of the house here and hearing all the commotion, you are not going to stop and find out the reasons why that dog is attacking your child, are you?

A: Well, no.

Q: You are just going to react.

A: Right.

Q: You are going to take care of that dog.

Finally, the State made the following statement during closing argument of the guilt-innocence phase:

[L]adies and gentlemen, this is an animal sitting right down here at the end of the table, just like one of them pit bull dogs in the back of the Robinson's [sic] yard. They attack and destroy stuff and you don't know why. You can't get in their mind. . . . You don't need to know the motive. Actions speak loud enough. This is an animal.

Vol. 28 at 70, ll. 3-10.

The statement during closing argument was objected to and thus was not an instance of deficient performance on the part of trial counsel.

This prosecutorial comment was not raised as a point of error on direct appeal. Willis does not argue that direct appeal counsel was ineffective for failing to raise it.

In a footnote, Willis raises additional comments by the prosecutor, to which defense counsel did not object, that are also part of Willis's claim of ineffective assistance of counsel. The comments fall into three categories: comments about Willis's exercise of his due process rights, inflammatory arguments, and arguments justifying the death penalty based on its deterrent effect. None of these specific comments were raised in the state courts. The State argues the statements are therefore not exhausted. Because this Court finds that Willis's claim is rejected on the merits, it is unnecessary for the Court to decide whether the additional remarks are exhausted.

See Pet. at 112, n. 43.

For example, the prosecutor stated:
"If it was what was fair and what was right, I submit to you back in the old days, our grandparents might have taken him out there and put him in the house, boarded it up, and set it on fire. That would have been justice. That would have been an eye for an eye, but today in our civilized society, even out here in West Texas where we are hard people, we have to live by the laws of our Constitution and our country, which many of us go to war for and defend for something like this to come in here and have his due process." Vol. 29 at 43, ll. 11-20.
"[H]e wanted his due process. He wanted his trial by 12 people. That's the type they are. They will be the judge and the jury and the executioner but when it comes to their turn, no, no, no. They want to run behind the Constitution, and then they want to run behind their rights, which they don't give to no one [sic] else." Vol. 29 at 46, ll. 18-23.
"Out here in West Texas, I have always taken great pride in the fact that we are pretty hard people. . . . And just two generations ago, ladies and gentlemen, our grandparents lived out here under the laws of Judge Roy Bean, who was a very famous jurist, and the law was swift and certain back in those days." Vol. 29 at 39, ll. 8-14.
"I'm sorry this proceeding has taken this long, ladies and gentlemen, but, once again, it's due process." Vol. 29 at 48, ll. 12-13.

The prosecution referred to Willis as: a "satanic demon," (Vol. 29 at 41, ll. 13-19); a "monster from a horror film," (Vol. 29 at 44, ll. 11-14); a "thing," (Vol. 29 at 47, ll. 12-13); and, the "most cowardly, most despicable thing that exists in our society," (Vol. 29 at 45, ll. 19-22).
The prosecution also made the following comments:
"I'm here to tell you . . . when they snap, they snap, and they are not human beings anymore. They have no utility to us. None. What he did was a cold, calculating, heartless act with methodical premeditated deliberation when you are doing something on the floor." Vol. 29 at 44, ll. 15-18.
"[I]t's hard for you to recognize those qualities that exist in a person that turns them into something other than a human being, but they have no compassion, no forgiveness in their hearts." Vol. 29 at 40, ll. 1-5.
"And forevermore, once a person reaches that snapping point in their brain where they don't have the ability to discipline themselves from doing violent acts like this, they forever, then, have the capability of hurting and killing us forever, because once you pass that line, you have committed your soul to the Devil." Vol. 29 at 41, ll. 13-19.

". . . I want you to consider the deterrent effect when you come back with your answer to these special issues, because there are people out here who have no compassion for their fellow man, who are coldhearted, bloody killers. . . . Let him and all other people that are like him that exist out here in our communities or around us or want them to be transients that come into our communities know that we believe in social vengeance. . . . We want them answered "Yes". . . . And anyone else like him that wants to come out here. . . . I want them to know that our juries out here will give it to them. Vol. 29 at 39, ll. 16, 20-47.

As to the merits of the remark, the State argues that the remarks were not improper and thus defense counsel was not deficient for failing to object to them. The State claims that the prosecution simply used animal imagery to ascertain whether any of the prospective jurors would hold the State to proving motive. Willis argues that the animal imagery was used to dehumanize him. Willis points to comments throughout trial describing Willis as a "rat," and "animal," a "satanic demon," a "monster from a horror film," a "thing," and someone who had "committed his soul to the devil." Willis argued defense counsel should have objected under and Lockett v. Ohio, and Eddings v. Oklahoma. Both cases discuss the fundamental respect for humanity underlying the Eighth Amendment, and both cases concern the right of the defendant to present mitigating evidence.

The State argues that these comments cannot form the basis of a claim of ineffectiveness at the penalty phase because they were made during voir dire or at closing arguments for the guilt-innocence phase. However, under Texas law, capital jury sentencing deliberations include evidence and arguments presented during both the guilt-innocence and penalty phases. See Banda v. State, 890 S.W.2d 42, 51 (Tex.Crim.App. 1994).

Vol. 11 at 68, ll. 18-21.

Vol. 28 at 70, ll. 3-10.

Vol. 29 at 41, ll. 13-19.

Vol. 29 at 44, ll. 11-14.

Vol. 29 at 47, ll. 12-13.

Vol. 29 at 41, ll. 13-19.

See id. at 113-14; Lockett, 438 U.S. at 604.

Comments, such as those made by the prosecutor here, do not violate Eddings or Lockett. While the Court finds the comments beyond poor taste and shameful, the Court must only decide whether the CCA's determination that the failure to object was not deficient performance is an unreasonable application of Strickland Willis has not cited, nor has the Court found on independent review, persuasive authority that the comments would have been error had defense counsel objected. It does not follow that, because the comments are distasteful and shameful, the CCA's determination that counsel was not deficient is unreasonable application of federal law. Our present rules are thus. Hence, as to this particular claim of ineffectiveness, the Court cannot say that defense counsel's performance was deficient. The Court need not reach the issue, then, of whether Willis was prejudiced by his counsel's failure to object to the State's descriptions of him as an animal.

3. Failure to Cross Examine and Present Mitigating Evidence

Willis argues that defense counsel was ineffective for failing to cross-examine the State's witnesses who provided testimony on aggravating factors and that defense counsel was ineffective for failing to present mitigating evidence. As stated above, the CCA addressed the claim of ineffectiveness as a whole and did not address the specific claim of ineffectiveness at the penalty phase of the trial, but a portion of the CCA's analysis refers to the penalty phase. The CCA stated that defense counsel was surprised Willis was found guilty and that defense attorney Woolard had "loaded his guns" for the guilt-innocence phase. The CCA mentioned that Woolard spoke with a number of Willis's friends and relatives and that Investigator Caspari also spoke with friends and relatives of Willis. The CCA stated that defense counsel decided not to present mitigation evidence. The CCA held that Willis did not overcome the presumption that defense counsel provided effective assistance of counsel. Also, the CCA divided the analysis of ineffective assistance for Willis's two trial attorneys which, as explained above, is contrary to clearly established law.

Ex Parte Willis, No. 27, 787-01, Order at 5.

The CCA's overruling of the trial court was not inconsistent with the trial court's factual findings. The CCA based its decision on its determination that defense counsel was reasonable to focus on the guilt-innocence phase and that defense counsel's mitigation investigation was reasonable, as was the decision to not present mitigating evidence. The CCA held that the record before it did not meet the standard for deficient performance. Because the CCA's decision was not inconsistent with the trial court's findings, this Court must defer to the state trial court's findings of fact.

The factual finding that could be perceived as inconsistent with the CCA's opinion is the trial court's determination that defense counsel did not prepare for the penalty phase. This could be construed as inconsistent with the CCA's statement that defense counsel Woolard and Investigator Caspari interviewed friends and relatives. However, the trial court also made a finding that defense counsel spoke with four or five people who knew Willis. Thus, the trial court determined that, despite interviewing some people, defense counsel was nonetheless unprepared for the penalty phase, and the CCA determined that the interviews conducted by defense counsel were sufficient to prevent a finding of deficient performance. Thus, the CCA's opinion is not inconsistent with the trial court's findings, but in fact relies upon them.

See Craker, 756 F.2d at 1213-14; Westley, 83 F.3d at 721 n. 2.

As to Willis's claim that defense counsel was ineffective for failing to cross-examine the State's witnesses, the Court agrees with the State. To the extent Willis argues defense counsel should have challenged the State's witnesses, Willis does not specify what evidence a cross-examination would have uncovered. Thus, Willis has not shown defense counsel was deficient in this regard.

See United States v. Green, 882 F.2d 999, 1003 (5th Cir. 1989).

As to the argument that defense counsel was ineffective for failing to present mitigating evidence, the Court finds counsel's performance was deficient. "Mitigating evidence concerning a particular defendant's character or background plays a constitutionally important role in producing an individualized sentencing determination that the death penalty is appropriate in a given case." Defense counsel did not present any mitigating evidence during the punishment phase of the trial.

Moore, 194 F.3d at 612. See also Woodson v. North Carolina, 428 U.S. 280 (1976); Eddings, 455 U.S. 104.

In Moore v. Johnson, defense counsel failed to present any mitigating evidence because defense counsel felt that mitigating evidence was contrary to an alibi defense and that the case was a "guilt-innocence" case, rather than a "punishment" case. The Fifth Circuit held that while "counsel's failure to develop or present mitigating background evidence is not per se deficient performance . . . Strickland does not require deference to those decisions of counsel that, viewed in light of the facts known at the time of the purported decision, do not serve any conceivable strategic purpose." The Fifth Circuit declined to defer to counsel's decision not to present mitigating evidence because the decision "was neither informed by a reasonable investigation nor supported by any logical position that such failure would benefit [the] defense." "Given that counsel's failure to investigate was not supported by reasonably professional limits upon investigation, the Court finds that there is no decision entitled to a presumption of reasonableness under Strickland"

Moore, 194 F.3d at 614.

Id. at 615. See Strickland, 466 U.S. at 681 ("Counsel may not exclude certain lines of defense for other than strategic reasons."); Boyle, 93 F.3d 180 (explaining basis for counsel's strategic decision not to offer mitigating evidence identified by the defendant); Loyd v. Whitley, 977 F.2d 149, 158 (5th Cir. 1992) ("Whether counsel's omission served a strategic purpose is a pivotal point in Strickland and its progeny. The crucial distinction between strategic judgment calls and just plain omissions has echoed in the judgments of this court.") (footnote omitted); Profitt v. Waldron, 831 F.2d 1245, 1249 (5th Cir. 1987) (no required deference to decisions that do not yield any conceivable benefit to the defense); Bell v. Lynaugh, 828 F.2d 1085, 1090 (5th Cir. 1987) (stating that when counsel makes an informed and considered decision not to present mitigating evidence, the issue becomes whether the decision was reasonable); Wilson v. Butler, 813 F.2d 664, 672 (5th Cir. 1987) (remanding for evidentiary hearing because record did not reflect whether counsel made a sound strategic decision not to present mitigating evidence of troubled background and mental impairment); Lyons v. McCotter, 770 F.2d 529, 534-35 (5th Cir. 1985) (finding deficient performance because there was no sound strategic basis for counsel's failure to object to evidence of prior offenses); Mattheson v. King, 751 F.2d 1432, 1439-40 (5th Cir. 1985) (explaining strategic purpose motivating counsel's decision to exclude evidence of mental impairment from sentencing phase); Moore v. Maggio, 740 F.2d 308, 315-19 (5th Cir. 1984) (explaining basis of counsel's considered decision to limit investigation by excluding implausible lines of mitigating evidence).

Moore, 194 F.3d at 616.

Id. at 617. See also Wiggins, 539 U.S. at 522 ("[O]ur principal concern in deciding whether [defense counsel] exercised `reasonable professional judgment,' is not whether counsel should have presented a mitigation case. Rather, we focus on whether the investigation supporting counsel's decision not to introduce mitigating evidence of [defendant's] background was itself reasonable." (internal citations omitted)).

As in Moore, defense counsel's decision in this case not to present any mitigating evidence was not motivated or justified by any strategic or tactical rationale. Counsel's decision was instead borne out of poor planning and false hopes for the guilt-innocence phase of the trial. There was simply no "thorough investigation of the law and facts relevant to all plausible lines of defense," necessary to make a "strategic or tactical decision not to present mitigating evidence." Here, as in Moore, counsel was unprepared and did not expect to proceed to the punishment phase of Willis's trial immediately after the guilty verdict was returned. Also, counsel agreed to proceed rather than request a continuance, as was the case in Moore.

See Moore, 194 F.3d at 615; Whitley, 977 F.2d at 158-59, nn. 21-22; Profitt, 831 F.2d at 1249; Lyons, 770 F.2d at 534-35 ( Strickland does not require deference when there is no conceivable strategic purpose that would explain counsel's conduct).

Moore, 194 F.3d at 615.

Id. See also McCoy v. Lynaugh, 874 F.2d 954, 964 (5th Cir. 1989) (counsel's decision not to present mitigating evidence is entitled to deference when based upon an informed and reasoned practical judgment); Wilkerson v. Collins, 950 F.2d 1054, 1064-65 (5th Cir. 1992) (affording strategic decision deference where record established counsel retained an investigator to explore whether mitigating evidence relating to defendant's background or mental ability was available); McCoy, 874 F.2d at 964 (finding scope of investigation reasonable where counsel investigated possibility of mitigating evidence by interviewing everyone on a list provided by the capital defendant and determined none of them had anything good to say about the defendant); Jones v. Thigpen, 788 F.2d 1101, 1103 (5th Cir. 1986) ("counsel either neglected or ignored critical matters of mitigation").

See Moore, 194 F.3d at 615. See also Ex Parte Willis, No. 27, 787-01 at 6.

Moore, 194 F.3d at 615, n. 9.

In many situations, ineffective assistance claims are rejected "because the record established counsel conducted an adequate investigation, but made an informed trial decision not to use the potentially mitigating evidence because it could have a prejudicial backlash effect on the defense." This is not such a case. The mitigating evidence here — testimony of Willis's heroic acts and good behavior — could only have helped and could not have harmed the case. Thus, the decision to forego mitigation could not be expected to "yield some benefit or avoid some harm to the defense."

See id. at 617. See also Darden, 477 U.S. 168 (counsel's failure to present mitigating evidence relating to defendant's character, psychiatric evaluation and history as a family man did not constitute deficient performance where such evidence would have opened the door to otherwise excluded evidence that defendant had prior criminal convictions, was diagnosed as a sociopathic personality, and had in fact abandoned his family); Mattheson, 751 F.2d 1439, 1440 (counsel made reasonable strategic decision to omit presentation of mitigating evidence of mental impairment where such evidence would have opened door to known evidence that defendant was a violent sociopath).

Moore, 194 F.3d at 615.

Finally, it is well established that the type of mitigating evidence that could have been presented in Willis's case is relevant to the sentencing determination. In Skipper v. South Carolina, the Supreme Court held that "evidence that the defendant would not pose a danger is spared (but incarcerated) must be considered potentially mitigating," and that a "jury could have drawn favorable inferences from . . . testimony regarding [defendant's] character and his probable future conduct if sentenced to life in prison." The Court also stated that "a defendant's disposition to make a well-behaved and peaceful adjustment to life in prison is itself an aspect of his character that is by its nature relevant to the sentencing determination." Furthermore, information showing a defendant as a good family member is mitigating evidence.

Skipper v. South Carolina, 476 U.S. 1, 4 (1986).

Id. at 7.

Hitchcock v. Dugger, 481 U.S. 393, 397 (1987) (vacating death sentence for failure of trial judge to consider, in part, that petitioner had been a fond and affectionate uncle to the children of one of his brothers).

Defense counsel's decision to not present mitigating evidence was deficient performance, based on counsel's failure to investigate, failure to prepare, failure to follow-up and the fact that there could be no benefit, and thus no strategic reason, to not present mitigation.

Willis also argues that defense counsel was ineffective for failing to make an individualized closing argument. Because this claim addresses the failure of defense counsel to acquire knowledge of Willis and present that knowledge at trail, the claim is incorporated into the claim of failure to investigate and present mitigating evidence.

The CCA's determination that counsel's failure to present mitigating evidence was not deficient performance is an unreasonable application of Strickland The CCA based its decision, without discussion of federal authority, on the fact that counsel focused on the guilt-innocence phase of the trial instead of the punishment phase, that counsel spoke with some people who knew Willis, and on the fact that Willis failed to show each attorney separately met the Strickland standard.

See Ex Parte Willis, No. 27, 787-01, Order at 6.

Clearly established federal law requires defense counsel to prepare for and investigate mitigating evidence. While the CCA correctly noted that defense counsel spoke with friends and relatives, the CCA did not determine whether the decision to limit the investigation at that point actually demonstrated reasonable professional judgment. The CCA did not address the trial court's factual finding that the limits on the investigation were due to a failure to follow-up and a lack of preparation.

See Williams, 529 U.S. at 393 ("[It] is undisputed that Williams had a right — indeed, a constitutionally protected right — to provide the jury with the mitigating evidence that his trial counsel either failure to discover or failure to offer.") Moore, 194 F.3d at 615; Stafford v. Saffle, 34 F.3d 1557 (10th Cir. 1994) (finding deficient performance and rejecting argument that an alibi defense during the guilt phase is per se inconsistent with mitigating evidence relating to the defendant's personal background); Brewer v. Aiken, 935 F.2d 850 (7th Cir. 1991) (granting relief on claim that counsel failed to offer mitigating evidence during the sentencing phase in case involving an alibi defense at the guilt phase).

Ex Parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 19-22. Willis's defense counsel failed to contact potential witnesses for the sentencing phase who ultimately spoke at the habeas evidentiary hearing. Some of the witnesses were present in the courtroom for Willis's trial. Some of the witnesses made it clear to defense counsel that they were able to testify on Willis's behalf. Defense counsel never followed up. See Pet. at 125.

Limits on investigation are reasonable only to the extent that reasonable professional judgments support the limitations. Because this principle constitutes clearly established federal law, the CCA's determination that defense counsel's investigation was adequate in this instance is an unreasonable application of clearly established federal law. While the CCA stated that defense counsel decided to forego mitigation and to "load guns" for the guilt-innocence phase, the CCA failed to address whether such a decision was reasonable considering the nature of the mitigating evidence available in this case. The available mitigation evidence included good acts by Willis and his good behavior while incarcerated. This is not a case in which mitigation would be inconsistent with the theory at the first phase of the trial or even a situation wherein mitigation would be damaging. Here, no reason exists to refrain from presenting evidence about the good deeds and nature of a defendant, particularly when the evidence includes testimony by law enforcement officers. Defense counsel's decisions to forego mitigation and focus on guilt was not strategic because it could not be expected to yield some benefit or avoid some harm to the defense. Therefore, the CCA's deference to defense counsel's decision to not present mitigation is an unreasonable application of Strickland

Strickland, 466 U.S. at 690-91. See also Wiggins, 539 U.S. at 524-26; Moore, 194 F.3d at 615.

See Moore, 194 F.3d at 615.

4. Prejudice at the Sentencing Phase

The testimony that could have been presented, but was not, at the penalty phase of Willis's trial pertained to Willis's propensity for future dangerousness. Law enforcement officers, including Pecos County Sheriff Bruce Wilson, would have testified on Willis's behalf. Sheriff Wilson, the Chief Deputy Sheriff, and two Pecos County jailers would have testified to Willis's good behavior in jail and that Willis was not a danger or threat in jail. In addition, defense counsel could have presented evidence that Willis surrendered himself to authorities when he learned of the charges against him; testimony describing Willis as a non-violent person; evidence of heroic acts by Willis; and testimony describing Willis as a loving family man.

During the sentencing phase of a Texas capital trial, the jury must answer two questions. The first concerns whether the crime was committed deliberately: Whether the conduct of the defendant that caused the death of the deceased was committed deliberately, and with the reasonable expectation that the death of the deceased or another would result. The second asks about the defendant's propensity for future dangerousness: Whether there is a probability that the defendant would commit criminal acts of violence that would constitute a continuing threat to society. TEX. CODE CRIM. P. art. 37.071 (Vernon 2004). See also Flores v. Johnson, 210 F.3d 456, 458 (5th Cir. 2000) (Garza, J., specially concurring) (thoroughly discussing the future dangerousness question and the lack of scientifically reliable evidence to support such a determination under federal law). "Overall, the theory that scientific reliability underlies predictions of future dangerousness has been uniformly rejected by the scientific community absent those individuals who routinely testify to, and profit from, predictions of dangerousness. . . . what separates the executioner from the murderer is the legal process by which the state ascertains and condemns those guilty of heinous crimes. If that process is flawed because it allows evidence without any scientific validity to push the jury toward condemning the accused, the legitimacy of our legal process is threatened." Id. at 465, 469-70. Nearly twenty-five years earlier, the Supreme Court indicated its disagreement in Jurek v. Texas, 428 U.S. 262, 274-76 (1976), but the issue will continue to demand the consideration of the federal courts.

Tr. at 85, ll. 21-23 (Wilson); Tr. at 47, ll. 11-12 (Harris); Tr. at 111, ll. 3-12 (Pringle); Tr. at 113, ll. 14 (Pringle). See also Tr. at 97, ll. 23-98 (Wilson); Tr. at 106, ll. 1-18 (Archer); Tr. at 114, ll. 3-5 (Pringle); Tr. at 49, ll. 2-6 (Harris).

This information could have been elicited from Deputy Jackson, one of the two prosecution witnesses during the penalty phase. Deputy Jackson met Willis in Odessa after Willis voluntarily came forward upon learning of the charges against him. Jackson did not have to restrain Willis on the drive to Fort Stockton. In fact, Willis sat in the front seat next to Deputy Jackson during the drive. Tr. at 118, ll. 22-25; Tr. at 119, ll. 16-24.

See e.g., Tr. at 54, ll. 23-55 (Officer Butts).

Several witnesses were available who could have testified to how Willis saved the life of a boy who was drowning by diving in and pulling the child out of a car which had accidentally backed into the lake. Tr. at 21, ll. 22-25; Tr. at 62, ll. 19-64; Tr. at 12, ll. 2-21. See also Tr. at 36, ll. 20-38; Tr. at 11, ll. 11-12, 21.

See e.g., Tr. at 51, ll. 5-53 (Officer Butts).

Thus, the mitigation evidence that could have been presented goes directly to the issue of Willis's propensity for future dangerousness, one of the two questions jurors must answer during the sentencing phase. Because of the extent of mitigating evidence concerning Willis's non-violent demeanor, the fact that law enforcement officers, including jailers, and the County Sheriff, were willing to testify to Willis's good behavior in jail, there is reasonable probability that, absent the failure of defense counsel, the jury would have concluded death was not the appropriate punishment for Willis. In addition, the State's aggravating evidence was less than substantial; prosecutors presented only two witnesses at the penalty phase, each testifying to Willis's "bad reputation" in the unspecified communities in which Willis lived. In a case in which innocence is a close question and in which the State's evidence of future dangerousness is weak, it is more likely that defense counsel's errors contributed to the jury's affirmative findings on issues of punishment. Here, Willis has shown that but for counsel's deficient performance, the result of the proceeding would have been different.

See Franklin v. Lynaugh, 487 U.S. 164, 177 (1988) (good conduct in prison is relevant to the special issue concerning future dangerousness under Texas capital sentencing scheme).

See Skipper, 476 U.S. at 8 (testimony of jailers would have likely been given great weight by the jury, since the jailers "would have had no particular reason to be favorably predisposed toward one of their charges").

See Strickland, 466 U.S. at 668.

See Ex Parte Willis, No. 27, 787-01 Find. of Fact and Conc. of Law at 20.

See Ex Parte Guzmon, 730 S.W.2d 724, 735 (finding defense prejudiced by ineffective assistance of counsel at capital penalty phase where the "State's evidence to prove future dangerousness was extremely weak"). See also Strickland, 466 U.S. at 696; Martinez-Macias, 979 F.2d at 1067 ("We are left with the firm conviction that [petitioner] was denied his right to adequate counsel in a capital case in which actual innocence was a close question").

Furthermore, the CCA did not reach the prejudice prong of the Strickland analysis and thus this Court is not constrained by section 2254(d) in determining whether Willis was prejudiced. However, because of the clarity of Supreme Court precedents holding that the type of mitigation evidence available in this case is relevant, and for the reasons stated above, a determination that Willis was not prejudiced is an unreasonable application of federal law. The Court holds that Willis received ineffective assistance of counsel at the sentencing phase because counsel's performance was deficient and Willis was prejudiced by counsel's deficiency.

Wiggins, 539 U.S. at 534 (finding that the Court's "review is not circumscribed by a state court conclusion with respect to prejudice, as neither of the state courts below reached this prong of the Strickland analysis").

See e.g., Skipper, 476 U.S. at 4, 7; Hitchcock, 481 U.S. at 397.

Finding reversible error at both the guilt-innocence phase and the sentencing phase, the Court need not address Willis's cumulative error claim.

CONCLUSION

Convinced, as stated above, that Willis's conviction and sentence both were obtained in violation of the United States Constitution, the Court grants Willis's request for relief as follows:

It is hereby ORDERED that the State's Motion for Summary Judgment is GRANTED IN PART and DENIED IN PART.

It is further ORDERED that Petitioner's Cross-Motion for Summary Judgment is GRANTED IN PART and DENIED IN PART.

It is further ORDERED that Willis's Petition for Writ of Habeas Corpus is GRANTED on the following grounds: 1) Petitioner's Due Process rights were violated by the State's administration of medically inappropriate antipsychotic drugs without Willis's consent; 2) the State suppressed evidence favorable and material to the sentencing determination; 3) Petitioner received ineffective assistance of counsel at the guilt-innocence phase; and 4) Petitioner received ineffective assistance of counsel at the sentencing phase.


Summaries of

Willis v. Cockrell

United States District Court, W.D. Texas, Pecos Division
Aug 9, 2004
No. P-01-CA-20, CAPITAL HABEAS (W.D. Tex. Aug. 9, 2004)
Case details for

Willis v. Cockrell

Case Details

Full title:ERNEST RAY WILLIS, Petitioner, v. JANIE COCKRELL, Director, Texas…

Court:United States District Court, W.D. Texas, Pecos Division

Date published: Aug 9, 2004

Citations

No. P-01-CA-20, CAPITAL HABEAS (W.D. Tex. Aug. 9, 2004)