Reset A A Font size: Print

Court of Criminal Appeals of Texas.


NO. AP-77,036

Decided: November 02, 2016


This Court's majority opinion upholds the conviction against Juan Balderas, appellant, despite the facts that his photo was the sole one in the photo array matching the physical description of the shooter; that it took the eyewitness two days of discussions with a police officer who showed her the array for her to make a positive identification of appellant, even though she had previously known appellant as “Apache”; and that the in-court identification of appellant at his trial that took place over eight years after the offense appears to have been tainted by the procedures used to obtain the earlier identification from the photo array. I disagree with this Court's majority opinion's conclusion that the pretrial identification procedure in this case was not impermissibly suggestive. I also disagree that the in-court identification that was made over eight years later was reliable. I would sustain appellant's eighth issue, find the error harmful, reverse appellant's conviction and death sentence, and remand for a new trial.

I. The Highly Suggestive Photo Spread Violated Appellant's Due-Process Rights

After reviewing the applicable law for eyewitness-identification evidence, I explain why I conclude that the photo-spread lineup that was used in this case was extremely unfair in that it included only one photo that matched the description of the shooter, and I will show that there was a substantial likelihood of misidentification in the later in-court identification of appellant.

A. Applicable Law

The Due Process Clause bars the admission of identification evidence when the introduction of that evidence is “so extremely unfair that its admission violates fundamental conceptions of justice.” Perry v. New Hampshire, 132 S. Ct. 716, 723 (2012) (quoting Dowling v. United States, 493 U.S. 342, 352 (1990)). Admission of an in-court identification after pretrial identification procedures that are so impermissibly suggestive as to be conducive to misidentification constitutes a denial of due process. Simmons v. United States, 390 U.S. 377, 384 (1968). Accordingly, “[a]n in-court identification is inadmissible when it has been tainted by an impermissibly suggestive pretrial photographic identification.” Gamboa v. State, 296 S.W.3d 574, 581 (Tex. Crim. App. 2009); see also Luna v. State, 268 S.W.3d 594, 605 (Tex. Crim. App. 2008); Ibarra v. State, 11 S.W.3d 189, 195 (Tex. Crim. App. 1999). A pretrial identification procedure may be impermissibly suggestive if the suspect is the only individual in a photo array who closely resembles the pre-procedure description. Barley v. State, 906 S.W.2d 27, 33 (Tex. Crim. App. 1995). The test for determining whether an identification is admissible under these circumstances is “whether, considering the totality of the circumstances, the photographic identification procedure was so impermissibly suggestive as to give rise to a very substantial likelihood of irreparable misidentification.” Loserth v. State, 963 S.W.2d 770, 772 (Tex. Crim. App. 1998) (quoting Simmons, 390 U.S. at 384). Reliability is the “critical question”:

If the totality of the circumstances reveals no substantial likelihood of misidentification despite a suggestive pretrial procedure, subsequent identification testimony will be deemed reliable, reliability being the linchpin in determining the admissibility of identification testimony.

Id. (citations and quotations omitted). In assessing reliability under the totality of the circumstances, the following five non-exclusive factors should be “weighed against the corrupting effect of” the suggestive pretrial procedure: (1) the opportunity of the witness to view the suspect at the time of the crime, (2) the witness's degree of attention, (3) the accuracy of the witness's prior description of the criminal, (4) the level of certainty demonstrated by the witness at the confrontation, and (5) the length of time between the crime and the confrontation. Id. (citing Neil v. Biggers, 409 U.S. 188, 199 (1972)). The party challenging the identification bears the burden to prove, by clear and convincing evidence, that the in-court identification has been irreparably tainted before a court will reverse a conviction on that basis. Barley, 906 S.W.2d at 34.

B. The Pretrial Identification Procedure Was Impermissibly Suggestive

Wendy Bardales was present at the time of the December 2005 shooting of Eduardo Hernandez. Hernandez was killed when he was shot at least nine times in the back and head by a gunman who entered the apartment where Hernandez was socializing with several friends. Wendy was one of the several witnesses present at the time of the shooting, but she was the only one who claimed that she could identify the gunman. Wendy was later interviewed at the police station on the night of the shooting.

On the night of the shooting, police officers obtained a description of the shooter from Wendy before she was shown any photo spread. Wendy said that she saw the gunman enter the apartment, that her eyes followed him until he left, and that he wore a black jacket with a hood pulled over his head, but that, at one point, when his hood fell down, she got a good look at his face. She said that she had never seen the gunman before, and that he had a mark on his face but she did not recall where it was. She stated,

I got a good look at his face. I have never seen him before. He was Hispanic and about 16-17 years old. He was around 5 foot 5 inches to 5 foot 7 inches tall. I remember him having a dark birth mark on his face but I can't remember exactly where. He was very skinny and clean shaven. He had black hair, it was short. He had a fade type haircut. He was wearing a black sweat shirt hooded jacket and khaki pants.

Also on the night of the shooting, a police officer showed Wendy a photo-spread array that did not include appellant's photo. Wendy did not identify anyone as the shooter, but she said that she recognized one of the people shown, Israel Diaz, who was a friend of Hernandez's. At that time, Wendy changed her earlier description of the shooter by claiming that the gunman had a dark mark on his cheek, which was different from her prior claim that she did not know where the facial mark was located.

Six days after the shooting, a police officer showed Wendy a different photo array with six photos, one of which was appellant's photo. Rather than create a unique photo spread for this particular case, the officer used a photo spread from a prior investigation that had included appellant's photo. Appellant's photo was the only one depicting a person with a dark mark on his cheek, wearing a black hooded sweatshirt, and matching Wendy's physical description of the shooter. When she saw the photo spread, Wendy immediately pointed at appellant's photo and identified him as “Apache,” describing him as a friend of Hernandez's and Diaz's. Wendy did not positively identify appellant as the shooter at that time. Rather, she made more tentative statements that appellant “could be the shooter,” that he “looked like the shooter,” and that his “face looked exactly like the shooter's face.” Wendy's indefinite remarks about appellant's photo left the officer unable to characterize her identification as “positive” when he left from his meeting with her.

Unsatisfied with his not having obtained a positive identification from Wendy, the officer visited her again the next day to further discuss the same photo spread. Wendy again told the officer that appellant's photo had the same face as the shooter, but she did not positively identify him as the shooter. The officer then told her to use her hands to cover the hair of each subject because the gunman had worn a hood over his head. Wendy placed her hands over the hair on each of the photos. When she put her hands over appellant's hair, her eyes “grew wide” and “began to water.” She then said she was absolutely positive in her identification of appellant as the shooter.

I conclude that the pretrial identification procedure in this case was impermissibly suggestive, and I, therefore, disagree with the trial court's assessment that, because all subjects were light-skinned, short-haired Hispanic males of the same general age and build, appellant's photo did not stand out in the six-photo array. Appellant's photo was the sole one in the photo spread that matched Wendy's description of the shooter as having a dark mole or birthmark on his face or cheek, a fade-style haircut, and wearing a black sweatshirt or jacket with a hood. The fact that appellant's photo was the only one in the photo spread possessing two of the distinctive characteristics of the shooter—a dark birth mark or mole and a black hooded sweatshirt—coupled with the suggestive nature of the procedure itself that involved the officer showing Wendy the same photo array twice to obtain a positive identification, rendered the procedure impermissibly suggestive. See Barley, 906 S.W.2d at 33-34 (explaining that suggestiveness “may be created by the manner in which the pre-trial identification procedure is conducted, for example by police pointing out the suspect or suggesting that a suspect is included in the line-up or photo array,” by the “content of the line-up or photo array itself if the suspect is the only individual closely resembling the pre-procedure description,” or by the “cumulative effect” of the suggestive procedures).

C. The In-Court Identification Was Unreliable Under the Totality of the Circumstances

Having concluded that the pretrial identification procedure in this case was impermissibly suggestive, it is necessary to determine whether Wendy's in-court identification of appellant that occurred over eight years after the offense was nevertheless reliable under the totality of the circumstances. See Loserth, 963 S.W.2d at 772 (explaining that the relevant inquiry is whether the procedure was so impermissibly suggestive as to give rise “to a very substantial likelihood of irreparable misidentification”) (quoting Simmons, 390 U.S. at 384). As explained above, this inquiry requires a weighing of the five non-exclusive factors established by Biggers, 409 U.S. at 199. “The underlying Biggers factors are, taken individually, historical facts and, as such, should be viewed deferentially.” Loserth, 963 S.W.2d at 773. The reviewing court should therefore consider the historical facts underlying the five Biggers factors in a light favorable to the trial court's ruling. Id. The factors, viewed in this light, are then weighed de novo against “the corrupting effect” of the suggestive pretrial identification procedure. Id. at 773-74; see also Gamboa, 296 S.W.3d at 581 (explaining that this Court “review[s] de novo a trial court's ruling on how the suggestiveness of a pre-trial photo array may have influenced an in-court identification”).

Here, although I acknowledge that we owe deference to the trial court's determination of historical facts underlying its ruling, the relevant facts are largely undisputed. See Loserth, 963 S.W.2d at 773. Thus, the primary question before this Court is one of law—that is, whether adequate indicia of reliability exist to outweigh the suggestiveness of the pre-trial photo array. Id. at 773-74. After addressing each factor below, I conclude, based on a de novo weighing of the factors, that appellant has satisfied his burden of showing by clear and convincing evidence that, under the totality of the circumstances, the impermissibly suggestive pre-trial identification procedure in this case gave rise to a substantial likelihood of misidentification. Barley, 906 S.W.2d at 33-34.

My conclusion that Wendy's eyewitness identification of appellant was wholly unreliable is based in large part on the same considerations as those addressed by Dr. Malpass, the eyewitness-identification expert who testified in this case. Dr. Malpass said that Wendy's identification of appellant is problematic because it evolved over time. Additionally, he explained that the viewing of successive photo spreads increases the possibility of memory contamination. Dr. Malpass determined that the officer's act of returning with the same photo array the day after Wendy had been unable to positively identify appellant would have conveyed a signal to Wendy that she needed to provide a more positive identification, which is exactly what she did in this case.

I note here that historically Texas has had a significant problem with the misidentification of suspects based on flawed pretrial identification procedures. Misidentification of people has been a large part of the reason for the high number of innocent people who have been wrongfully convicted in Texas. The problem with suggestive identifications was significant enough that it was recently addressed by the Texas Legislature though legislation that came into effect after this offense. For example, Article 38.20 of the Code of Criminal Procedure now requires law enforcement agencies to adopt policies for photograph and live lineup identification procedures that would require, if possible, that the photospread be shown to an eyewitness by someone unfamiliar with the identity of the suspect in the case so as to prevent opportunities to influence the witness. See TEX. CODE CRIM. PROC. Art. 38.20, § 3. It is highly unlikely that the manner in which the identification procedures were conducted in this case would comply with the requirements of this statute, even if the identification were otherwise admissible in a trial court. See id. § 5. Until this Court disallows tainted identifications based on suggestive photo spreads, as occurred in this case, Texas will continue to be a leader in the wrongful convictions of innocent people. Although I do not suggest that appellant is innocent of this offense, I conclude, as explained more fully below, that he is entitled to a new trial that should be conducted absent the tainted identification that occurred in this case.

Applying the relevant legal standard to the instant facts, here there were no factors that would make Wendy's identification otherwise reliable. The Supreme Court has determined that an identification based on a suggestive photo spread may be admitted if there is evidence that shows that the corrupting effect of the suggestive identification procedure was ameliorated by five other circumstances that are weighed for their persuasive value. See Biggers, 409 U.S. at 199. But these circumstances are unpersuasive in this case.

First, Wendy's opportunity to view the shooter at the time of the crime was impeded by the hoodie worn during the entire event, except for a short time during which the hood fell down. Although it may be true, as Wendy claimed, that she watched the shooter the whole time that he was in the apartment, the shooter wore a hoodie covering his head the entire time that he was there, except for the moment when his hood fell down. The length of Wendy's entire observation of the shooter, therefore, is immaterial because it has little value in discerning the reliability of her identification of appellant's face in a photo array. What is pertinent is the amount of time that Wendy had an unobstructed view of the shooter's face, which only occurred during the brief moment when his hood fell down. Because the shooter's face was obstructed by the hood he wore over his head during most of the offense, that fact weighs against the reliability of Wendy's identification of appellant's photo. Furthermore, Wendy knew appellant by the name of “Apache” prior to the night of the shooting, but she did not mention that to police officers when she gave a physical description of the shooter before she was shown a photo spread; instead she represented to police officers that she had never seen the shooter before. If her opportunity to view the shooter was adequate to give rise to a reliable identification, it would be reasonable to expect that Wendy, at a minimum, would have told police officers that she believed she had seen the shooter before, even if she could not recall that his name was “Apache.” Thus, this factor weighs against a determination of reliability.

Second, regarding Wendy's degree of attention, although she was focused on the events and shooter during the crime, this does not necessarily correlate with a reliable identification in this case in which the facts show that the shooter's head was covered by a hood during most of the events and that Wendy was in a state of shock over the events. Wendy claimed to have fixated on the shooter, but, as explained above, she could only see his entire face during the instance when his hood fell off. It is reasonable to believe that Wendy's attention was not focused on the shooter because, on the night of the offense, Wendy did not tell police that she recognized the shooter as a person she knew as “Apache,” the name that she used to refer to appellant. Even a week later, when Wendy was shown the photospread that contained appellant's photo, she did not immediately identify him as the shooter; instead it took her two viewings of the photo spread, along with a suggestion from the police officer that she manipulate the photos by placing her hands over the individuals' hair, before she made a positive identification. Other circumstances that suggest that Wendy's degree of attention was minimal are that she misdescribed the murder weapon and misstated that she had been shot at. Dr. Malpass explained why an identification by an eyewitness can be impeded or degraded by the shock of the criminal events. In sum, despite Wendy's claim that she fixated on the shooter, the record fairly demonstrates that, when she spoke to police officers on the night of the offense, she did not identify appellant as the shooter even though she knew him as “Apache,” it took her two days of looking at the same photo spread before the officer showing her the photo spread would characterize her identification as a positive one, and she misdescribed other key aspects of the shooting including the color of the weapon and who had been shot at. Thus, this factor also weighs against a finding of reliability.

Third and fourth, Wendy lacked accuracy in her prior description of the shooter, and her level of certainty in her identification of appellant is unconvincing. As the chart below demonstrates, Wendy's claims about her ability to identify the shooter lacked consistency. At first, Wendy said that she had never seen the gunman before, but when she first saw appellant's photo in the photo spread, she identified appellant as a friend of Hernandez and Diaz rather than as the shooter. At first, Wendy said that the shooter had a mark on his face but she did not know where on his face. Later, Wendy specified that the mark was on the shooter's cheek. At first, Wendy could not positively identify appellant as the shooter when she was shown the second photo spread, but the next day when the same officer returned to her with the same photo spread asking her to be more definitive in her identification, she positively identified appellant as the shooter. These questionable aspects of Wendy's identification are summarized in the chart below:

Issue Wendy's First Statements Wendy's Later Statements Whether she had seen the She initially told police She identified appellant in gunman before. before she saw the photo the second photo spread as spread that she had never a friend of her friend. seen the gunman before. Whether she could identify She initially said the She later said that the mark the mark on the gunman's gunman had a mark on his was on the gunman's face. face but she could not say cheek. where on his face. Whether the identification When she first saw the When she saw the same of appellant in the second photo spread, she said photo spread the next day, photo spread was positive. appellant's photo “could be she was positive that the shooter” and that he appellant was the shooter. looked exactly like the shooter.

As the chart above illustrates, Wendy's identification of appellant as the shooter is unreliable because she said she had never seen the shooter before, and yet when she saw appellant's photo she recognized him as someone she knew who was called “Apache.” She picked the only person who matched her description of the shooter: the photo of the person with a facial mark on his cheek and wearing a hoodie. And her level of certainty was weak in that it took the officer two days to elicit from her a positive identification of appellant as the shooter.

Fifth, although the length of time between the crime and Wendy's identification of appellant in the photo spread was minimal, with Wendy picking appellant's photo out of the photo array at seven days after the offense, the in-court identification is totally unreliable as it occurred more than eight years after Wendy witnessed the crime. After such a long period, it is highly unlikely that Wendy's in-court identification was completely independent from the photo spread. That is, eight years after the murder, Wendy was not likely truly identifying appellant as the shooter based on her independent memory of that night, but rather was simply identifying appellant in court because he was the person sitting at counsel table who had been arrested and charged for this offense based on her prior photo spread identification of him.

Weighing the relevant factors and considering the totality of the circumstances, I conclude that appellant has demonstrated by clear and convincing evidence that the corrupting effect of the suggestive pre-trial identification procedure in this case created a substantial risk that Wendy misidentified appellant as the shooter due to the lack of any factors to show that her identification of him in court was reliable. Consequently, Wendy's in-court identification of appellant as the gunman should not have been admissible. See Ibarra, 11 S.W.3d at 195. In light of the highly suggestive identification procedures that occurred in this case and irreparably tainted in-court identification by Wendy, I would hold that the trial court erred by permitting this identification and that this violated appellant's federal due-process rights.

II. Harm Analysis

Because appellant's complaint is premised on a violation of his due-process rights, the constitutional-error harm standard applies here. See Stovall v. Denno, 388 U.S. 293, 301-02 (1972); Hernandez v. State, 60 S.W.3d 106, 108 (Tex. Crim. App. 2001). Under that standard, “[i]f the appellate record in a criminal case reveals constitutional error that is subject to harmless error review, the court of appeals must reverse a judgment of conviction or punishment unless the court determines beyond a reasonable doubt that the error did not contribute to the conviction or the punishment.” TEX. R. APP. P. 44.2(a). Constitutional error may be harmless if there is “overwhelming” untainted evidence to support the conviction. See Harrington v. California, 395 U.S. 250, 254 (1969); see also Wesbrook v. State, 29 S.W.3d 103, 119 (Tex. Crim. App. 2000). Conversely, the error was not harmless if there is a reasonable likelihood that it “materially affected the jury's deliberations.” Neal v. State, 256 S.W.3d 264, 284 (Tex. Crim. App. 2008). Thus, the court must evaluate the reasonable possibility that the “constitutional error was actually a contributing factor in the jury's deliberations in arriving at [its] verdict—whether, in other words, the error adversely affected ‘the integrity of the process leading to the conviction.’ ” Scott v. State, 227 S.W.3d 670, 690 (Tex. Crim. App. 2007).

A constitutional-error harm analysis does not focus on the propriety of the outcome of the trial, that is, whether the jury verdict was supported by the evidence. Id. Rather the focus is on the probable impact of the constitutional error on the conviction in light of the existence of other evidence. See Wesbrook, 29 S.W.3d at 119. The entire record must be evaluated in a neutral, impartial, and even-handed manner—not in the light most favorable to the prosecution. Harris v. State, 790 S.W.2d 568, 586 (Tex. Crim. App. 1989). In analyzing harm, the court must assess whether there was a reasonable possibility that the error, either alone or in context, “moved the jury from a state of non-persuasion to one of persuasion.” Scott, 227 S.W.3d at 690. This examination may consider (1) the nature of the error, (2) the extent to which the error was emphasized by the State, (3) the probable implications of the error, and (4) the weight the jury would likely have assigned to it in the course of deliberations. Snowden v. State, 353 S.W.3d 815, 817 (Tex. Crim. App. 2011); see also Scott, 227 S.W.3d at 690 (noting that “how weighty the jury may have found the erroneously admitted evidence [should] be compared to the balance of the evidence with respect to the element or defensive issue to which it is relevant”). This list is not exclusive, and the harm analysis for constitutional error should account for “any and every circumstance apparent in the record that logically informs an appellate determination whether ‘beyond a reasonable doubt [that particular error] did not contribute to the conviction or punishment.’ ” Snowden, 353 S.W.3d at 822 (quoting TEX. R. APP. P. 44.2(a)). Accordingly, given my conclusion that appellant's due-process rights were violated by Wendy's tainted identification, I would reverse his conviction unless the record establishes, beyond a reasonable doubt, that admitting Wendy's identification of him as the shooter did not contribute to his conviction.

The first factor requires consideration of the nature of the error, but to understand that matter, a review of the record as a whole is required. See Snowden, 353 S.W.3d at 822. It is necessary to review the untainted evidence admitted against appellant, as well as the evidence produced by the defense in response to the State's case. Excluding Wendy's identification, the State's case consisted primarily of circumstantial evidence that appellant had motive to kill Hernandez, that appellant was at the crime scene moments after the shooting occurred, and that appellant was in possession of the murder weapon when he was arrested days after the shooting occurred. But without Wendy's identification, the State was left with evidence supplied by Israel Diaz, a gang member with significant motives to testify falsely, and appellant's possession of the murder weapon over a week after the shooting, making it much less persuasive as a meaningful link to the crime.

To establish motive and opportunity, prosecutors called Israel Diaz to testify that Hernandez had betrayed their gang, La Tercera Crips (“LTC”), by affiliating with other gangs and cooperating with police. Diaz testified that certain LTC members agreed that Hernandez should be killed for his disloyalty. To explain why appellant would undertake to kill Hernandez, Diaz stated that appellant bore an unspoken responsibility for Hernandez as his sponsor into the LTC gang. Diaz further testified that he and appellant met minutes after the shooting just across the street from where it had occurred. Diaz said that, during that meeting, appellant remarked that he “finally got him,” and that appellant reloaded a silver semi-automatic handgun that looked similar to the handgun that was later proven to be the murder weapon.

The defense's cross-examination, however, showed that Diaz's testimony arguably lacked credibility for two reasons. First, Diaz also had a motive to kill Hernandez. Prior to Hernandez's death, Diaz had stolen a car at gunpoint. He later loaned that car to Hernandez. Police stopped Hernandez while he was driving the stolen car. When police questioned Hernandez about the car, Hernandez implicated Diaz in its theft, and Diaz was ultimately charged with aggravated robbery. Thus, jurors were presented with evidence that Diaz had a motive to kill Hernandez, either to prevent Hernandez from testifying against him for the robbery or to retaliate for the betrayal of gang loyalty. Second, Diaz's testimony at trial was procured by the State on the eve of appellant's trial in exchange for the State reducing Diaz's pending capital-murder charge in another case to aggravated robbery. Diaz, therefore, had an incentive to testify against appellant to secure a reduced charge. Moreover, the defense offered the testimony of Walter Benitez, another LTC member, who contradicted much of Diaz's testimony. Benitez testified that, in fact, it was an LTC member named Victor Arevalo who had killed Hernandez and that appellant actually advocated against killing Hernandez when LTC members discussed his disloyalty. In light of Diaz's motive to have committed this offense, the fact that his testimony was given in exchange for reduced charges on another offense, and the testimony that a different gang member killed Hernandez, the State's evidence from Diaz weakly connected appellant to Hernandez's murder.

It is true that evidence connects appellant to the murder weapon. But this connection to the murder weapon was not exclusive of other LTC members, who would have had similar motive to kill Hernandez and who had access to the same cache of weapons during the ten-day interval of time between the shooting and seizure of the weapon. Police officers testified that ten days after the shooting, appellant was arrested pursuant to a warrant. At that time, appellant and another individual were holding boxes when police arrived to make the arrest. When he saw police approaching him, appellant set the box down before both he and the other individual ran from the police. After the officers arrested appellant, they inspected the contents of the box that appellant had been holding moments earlier. Inside the box were various firearms, one of which was later identified as the murder weapon through ballistics testing. However, the defense introduced testimony that LTC was not a well-armed gang, and it was common for gang members to pool and share weapons. This is evidenced by the number of weapons in the box. Thus, while finding the murder weapon in appellant's possession at the time of his arrest ten days after the shooting is some evidence of appellant's guilt, its weight is less significant because it establishes only a loose connection between appellant and the murder weapon under these particular circumstances.

Additional circumstantial evidence admitted at trial suggested that Hernandez was killed due to his LTC gang association. The day he was murdered, Hernandez and an LTC gang member had a private discussion after which Hernandez was worried because he knew something was wrong or something bad was going to happen. Also, the day that Hernandez was killed, LTC-themed graffiti had been spray painted near the apartment where he was killed. Karen Bardales, Wendy's sister who was also present at the time of the murder, testified that Hernandez “knew something was going to happen” upon seeing the graffiti.

While the State's case showed that, in all likelihood, Hernandez was killed by an LTC member, the only evidence admitted to persuasively show appellant to be the specific LTC member who shot Hernandez was Wendy's identification. Without Wendy's eyewitness identification, it cannot be said beyond a reasonable doubt that the jury's deliberation upon the rest of the State's evidence would have remained unchanged and would still have produced the same guilty verdict.

The second factor requires consideration of the extent to which the error was emphasized by the State. Here, the State relied heavily on Wendy's identification, which was the sole piece of evidence directly linking a specific LTC gang member, appellant, to this offense. In the absence of this evidence, the State would have been forced to concede that other LTC gang members had a similar motive to kill Hernandez and also had access to the murder weapon.

The third factor requires consideration of the probable implications of the error. I conclude there is a reasonable likelihood that the jury believed the testimony of Diaz and disregarded the testimony of Benitez because the jury had heard from Wendy—the only eyewitness to the murder who claimed to have seen the killer's face—that appellant was the shooter. Wendy's identification corroborated Diaz's testimony that appellant was present near the crime scene shortly after the killing and that he was carrying a handgun that appeared similar to the murder weapon. Additionally, Wendy's identification of appellant as the shooter provided the context for Diaz's testimony that appellant's remark that he “finally got him” was a reference to appellant having killed Hernandez.

Furthermore, Wendy's identification supports the inference that appellant had been in possession of the murder weapon since the night of the shooting, and, therefore, appellant had killed Hernandez. Absent Wendy's identification of appellant as the shooter, the jury would have likely attributed less weight to appellant's possession of a box with numerous weapons, one of which was the murder weapon, and would have given more weight to the testimony that LTC members pooled and shared their weapons. Appellant's possession of the box containing weapons while he was with another individual would have had little persuasive value absent Wendy's identification testimony. Because ten days had passed between the shooting and appellant's arrest while in possession of the murder weapon, the shooter would have had ample opportunity to either dispose of the murder weapon or to re-deposit the murder weapon in the LTC cache. Given this fact, appellant's possession of the murder weapon at the time of his arrest is only weak evidence of his guilt under these circumstances.

The fourth factor requires consideration of the weight the jury would likely have assigned to Wendy's identification in the course of deliberations. Although other evidence supports the conviction, the quality of that evidence was weak because it included the bartered-for testimony of Diaz, a fellow gang member with strong motives to kill Hernandez who was near the apartment where Hernandez was shot at the time of the shooting, and appellant's possession of the murder weapon over a week after the offense. Wendy's eyewitness identification is the only evidence directly connecting appellant to the murder. Without her testimony, the strength of the State's case was significantly undermined.

While a jury could have rationally reached a guilty verdict without Wendy's identification of appellant as the killer, that is not dispositive in finding that the constitutional error at trial was harmless. After applying the correct standard under Rule 44.2(a) and weighing the factors in view of the record in its entirety, I cannot conclude beyond a reasonable doubt that the jury's deliberation would have been unaltered and that the tainted identification did not contribute to the conviction. The State relied heavily on Wendy's identification of appellant as the shooter because there were no other eyewitnesses who could identify the shooter and no forensic evidence linking appellant to the murder scene. Diaz's testimony that appellant was near the location where Hernandez was killed shortly after the murder, that appellant had a handgun at that time, and that appellant alluded to having shot Hernandez would likely have been credited by the jury based on Wendy's identification that may have served as a basis for the jury to not only minimize Diaz's bias but also to disregard Benitez's identification of another person as Hernandez's killer. Moreover, the jury may have seized upon Wendy's identification to support the inference that, because appellant was arrested while carrying a number of weapons, one of which was the murder weapon, appellant had used the weapon to shoot Hernandez.

Because I cannot conclude beyond a reasonable doubt that the admission of Wendy's identification testimony did not lend significant support to the State's other circumstantial evidence of appellant's guilt, and because Wendy's testimony might have provided a reason for the jury to discount appellant's defensive evidence, I cannot say beyond a reasonable doubt that its admission did not contribute to the guilty verdict. I, therefore, would hold that the admission of this evidence at trial was not harmless error.

III. Conclusion

Because most of the facts surrounding the procedures that led to the identification of appellant are undisputed, this is not a case that requires deference to the trial court's decision to admit identification evidence. Rather, this is a case that requires this Court to apply the law to the largely undisputed facts. By appropriately applying the law to the facts, I conclude that the pretrial identification procedure in this case was so impermissibly suggestive as to give rise to a substantial likelihood of misidentification, and I further conclude that appellant has shown that there are no circumstances to show that Wendy's identification of him in court eight years after the offense was reliable so as to diminish the corrupting effect of the procedure. Because I cannot conclude that the error was harmless, I respectfully dissent from this Court's judgment that affirms appellant's conviction for capital murder and sentence of death. Accordingly, I would reverse the judgment of the trial court and remand for a new trial.

ALCALA, J., filed a dissenting opinion.

Copied to clipboard