COMMONWEALTH vs. ELVIN BASTALDO.

GANTS, C.J. In the parking lot of a night club in Springfield, the defendant, Elvin Bastaldo, punched the victim, Juan Benito, several times in the face using brass knuckles, blinding him in one eye, while the victim was standing near a police officer who was arresting the defendant's brother, Juan Bastaldo (Juan). [Note 1] The defendant was convicted by a Superior Court jury of mayhem, in violation of G. L. c. 265, § 14, and resisting arrest, in violation of G. L. c. 268, § 32B. [Note 2], [Note 3]

On appeal, the defendant claims that he is entitled to a new trial because (1) the judge abused her discretion in denying the defendant's requested cross-racial and cross-ethnic eyewitness identification jury instruction where two of the three eyewitnesses were "Caucasian" and the defendant was a "dark-skinned Hispanic of Dominican descent"; (2) the admission of three in-court eyewitness identifications created a substantial risk of a miscarriage of justice where it was the first time any of them had formally identified the defendant; [Note 4] and (3) the judge committed prejudicial error by giving a consciousness of guilt instruction

Page 18

that suggested to the jury that the defendant was the assailant. [Note 5]

We conclude that because this case was tried before our opinion issued in Commonwealth v. Gomes, 470 Mass. 352, 376, 382 (Appendix) (2015), where we prospectively required that a jury instruction on cross-racial eyewitness identification be given in these circumstances, the judge did not abuse her discretion in declining to give the defendant's requested cross-racial and cross-ethnic instruction. We now revise the content of the provisional model jury instruction regarding cross-racial identification that we issued in Gomes, as well as our guidance as to when such an instruction should be given. In criminal trials that commence after the issuance of this opinion, a cross-racial instruction should always be included when giving the model eyewitness identification instruction, unless the parties agree that there was no cross-racial identification. We authorize judges in their discretion to include a cross-ethnic eyewitness identification instruction in appropriate circumstances.

We further conclude that where this case was tried prior to the issuance of Commonwealth v. Crayton, 470 Mass. 228 (2014), and Commonwealth v. Collins, 470 Mass. 255 (2014), the admission of the in-court eyewitness identifications did not create a substantial risk of a miscarriage of justice. Finally, although under the circumstances of this case the judge erred in instructing the jury regarding consciousness of guilt, we conclude that the error was not prejudicial. We therefore affirm the judgments of conviction. [Note 6]

Background. The jury could have found the following facts from the evidence admitted at trial. At approximately 12:30 A.M. on September 2, 2012, Juan and three companions (not including the defendant) attempted to enter a night club in Springfield. The victim, who, by his description, served as the "doorman, security, [and] host" of the club, denied their entry because the companions with Juan were under twenty-one years of age. A brief verbal

Page 19

and physical altercation ensued in which Juan punched the victim in the chest, and the victim countered by punching Juan below the eye. Springfield police officer Thomas Liebel, who was working a security detail at the club, ordered Juan to leave the area, which he did.

The club closed at 2 A.M. As Liebel walked to his vehicle to leave, Juan and two or three other men appeared from a nearby alleyway and headed toward the main entrance of the club. [Note 7] When they attempted to enter the club, Liebel approached them and ordered them to leave. The victim was standing inside the club near the entrance, along with Ronald Kenniston, a club employee who worked as a "bar back-up." As soon as the victim opened the entrance door, Juan punched the victim in the side of the face.

Liebel moved to arrest Juan, but when Juan "went for" Liebel, Liebel sprayed him with mace. Juan and the other men ran away, but Liebel gave chase and caught Juan. The victim followed to make sure Liebel was all right, and stood near Liebel as he struggled to handcuff Juan. The victim was then suddenly struck in the face. He did not see from where the blow came, but it rendered him dazed and blind in his left eye. When he turned around to defend himself, he saw the defendant, whom he had never seen before, standing a foot or two in front of him. The defendant punched the victim in the face two or three more times.

The defendant then approached Liebel and yelled in English, "I am going to fuck you up, Officer." The defendant came within three feet of Liebel before police sirens sounded and the defendant "bolted." Liebel watched the defendant run through a large parking lot, transmitted a description of the defendant's clothing and location over the police radio, and learned one minute later that the defendant had been arrested. Liebel soon saw the defendant again before he was placed inside a police transport vehicle with Juan, where they threatened and cursed Liebel in English.

Kenniston had been standing approximately fifteen feet away from the victim when a person "came up from behind [the victim] and sucker punched his eye a few times." He identified the defendant at trial as the person who threw the "sucker" punches. Kenniston testified that he got a good look at the defendant's face,

Page 20

and observed a silver object in the defendant's hand that covered three of his fingers. He also watched the defendant throw an object across the street, which sounded like metal when it landed, before the defendant ran away. [Note 8] Kenniston saw police officers catch up to the defendant, tackle him, arrest him, and bring him back to Liebel.

Kenniston then drove the victim to a local hospital. On the way, he passed the police transport vehicle and saw that the defendant was in custody. [Note 9] The victim was later transferred to Massachusetts General Hospital, where he underwent surgery on his eye. At the time of trial, the victim was still blind in his left eye.

The defendant testified at trial that he and Juan had arrived at the club at approximately 9:05 P.M. by themselves and remained inside until 2 A.M. He then left with Juan but Juan stayed near the entrance to talk with someone while the defendant continued walking. The defendant had not walked far when he turned around to see that there was fighting and that a police officer had handcuffed Juan. He saw that Juan had lost a shoe, so he retrieved it and walked over toward him and the officer who had handcuffed him, asking, "What happened?" When the police were about to take Juan away, the defendant walked through a parking lot in the direction of his house. Before he reached the street, he was grabbed by the police and thrown to the ground. The defendant stated that he did not see anyone strike the victim, and did not punch the victim himself. [Note 10]

Discussion. 1. Cross-racial and cross-ethnic eyewitness identification instruction. At the charge conference, the defendant's attorney requested the following cross-racial and cross-ethnic eyewitness identification instruction:

"In this case, the identifying witnesses are of a different race or ethnicity than the defendant. Scientific studies have shown that it is more difficult to identify members of a different race

Page 21

or ethnicity than members of one's own. In addition, studies reveal that even people with no prejudice against other races and substantial contact with persons of other races still experience difficulty in accurately identifying members of a different race or ethnicity. Quite often people do not recognize this difficulty in themselves. You should consider this in evaluating the reliability of the witnesses' identification of the defendant."

As to the race or ethnicity of the eyewitnesses, Kenniston and Liebel testified that they are Caucasian; no evidence was offered regarding their ethnicity. The victim testified that his father is Puerto Rican and his mother is Italian, and he considers himself Hispanic. No evidence was offered regarding the race of the victim; based on his photograph, his skin color appears to be brown.

As to the defendant's race or ethnicity, the defendant testified that he is from the Dominican Republic but did not discuss his racial identity. Liebel testified that the person who struck the victim was Hispanic. Kenniston was also asked if the person who struck the victim was Hispanic, and he responded:

"Yeah . . . . Well, I mean I don't know the classification because . . . I have friends that are . . . black, Puerto Rican, and they can speak two languages, so . . . just because they are a certain color, I'm not going to say they are Spanish or Black. . . . I really need to talk to them to know what they are."

The defendant's written request for a cross-racial and cross-ethnic instruction stated that he is Hispanic. On appeal, he characterizes himself as a "dark-skinned Hispanic of Dominican descent." Based on his photograph, his skin color appears to be black.

The Commonwealth objected to the request and questioned whether the identifications were truly cross-racial or cross-ethnic, as the evidence only showed that the witnesses may have different ethnic backgrounds. The Commonwealth also asked that if a cross-racial or cross-ethnic instruction were given, it not apply to the victim because he was also Hispanic.

The judge declined the defendant's request for an instruction, stating that Kenniston and Liebel are of different ethnicity from the defendant, but "[w]e are not talking about a cross-racial identification here." The judge also stated that, even if she were to

Page 22

assume that the identification was similar to a cross-racial identification, it is "far from settled" that such an instruction should be given, and giving such an instruction "is hardly the standard in the courts at this point." She acknowledged that we were considering the Report and Recommendations of the Supreme Judicial Court Study Group on Eyewitness Evidence (July 25, 2013) (Study Group Report), [Note 11] and that the proposed jury instructions regarding cross-racial identification were "very controversial" in the Superior Court. The judge instead instructed the jury in accordance with the then-existing model eyewitness identification instruction, based on Commonwealth v. Rodriguez, 378 Mass. 296, 310-311 (Appendix) (1979), as modified in Commonwealth v. Cuffie, 414 Mass. 632, 640-641 (Appendix) (1993), and Commonwealth v. Santoli, 424 Mass. 837, 845 (1997), including the instruction regarding the possibility of a good faith mistake. See Commonwealth v. Pressley, 390 Mass. 617, 620 (1983). Because the defendant objected at the close of the instructions to the absence of the requested instruction, we review its denial for prejudicial error. See Commonwealth v. Meas, 467 Mass. 434, 454, cert. denied, 135 S. Ct. 150 (2014).

In Gomes, 470 Mass. at 366-367, we concluded that a principle of eyewitness identification may be appropriate for inclusion in a model jury instruction "where there is a near consensus in the relevant scientific community adopting that principle." Because it was not argued that the identifications in that case were cross-racial, we did not address that issue, but we included an instruction on cross-racial identification in the provisional model jury instruction that we required to be given, where appropriate, in trials that commence after the issuance of that opinion. Id. at 376, 382 (Appendix). That instruction provided that, in deciding whether a witness's identification is accurate, a jury should consider "whether the witness and the offender are of different races -- research has shown that people of all races may have greater difficulty in accurately identifying members of a different race than they do in identifying members of their own race." Id. at 382 (Appendix). Including that instruction reflected our conclusion that this principle had been adopted by a near consensus in the relevant scientific community. Id. at 382 n.10 (citations providing

Page 23

support for near consensus on cross-racial identification). But we declined to give the new provisional model jury instruction any retroactive application, id. at 376, so it has no bearing on this case, which was tried one year before the issuance of Gomes.

Under our case law at the time of trial, a judge was not precluded "in the exercise of discretion from instructing a jury that, in determining the weight to be given eyewitness identification testimony, they may consider the fact of any cross-racial identification and whether the identification by a person of different race from the defendant may be less reliable than identification by a person of the same race." Commonwealth v. Hyatt, 419 Mass. 815, 819 (1995). But a defendant "was not entitled to such an instruction." Commonwealth v. Bly, 448 Mass. 473, 496 (2007) ("While we acknowledge the significant body of scientific literature on the problems inherent in cross-racial identification, . . . we have never held that those problems require a jury instruction when cross-racial identification testimony is offered, and we decline to do so here" [citation omitted]). Therefore, the judge did not err in declining to read the requested instruction.

Although it was not error before Gomes for the judge to decline to give a cross-racial instruction, such an instruction must be given in trials that commence after Gomes where there is a cross-racial identification. See Gomes, 470 Mass. at 376, 382 (Appendix). The existence of the "cross-race effect" (CRE) -- that people are generally less accurate at identifying members of other races than they are at identifying members of their own race -- has reached a near consensus in the relevant scientific community and has been recognized by courts [Note 12] and scholars [Note 13] alike. We remain convinced that jurors who are asked to evaluate the accuracy of

Page 24

an identification should be informed of the CRE. [Note 14]

We take this opportunity, however, to consider when a cross-racial instruction should be given. In Bly, 448 Mass. at 496, we declared that it is within a judge's discretion to give a cross-racial instruction "when warranted by the evidence," and our provisional instruction in Gomes, 470 Mass. at 382 (Appendix), provides

Page 25

that the instruction should be given when the "witness and offender are of different races." See Study Group Report, supra at 134 (proposed supplemental cross-racial and cross-ethnic instruction should be given "[i]f the witness and the perpetrator are of a different race or ethnicity"). But we have yet to discuss when the evidence warrants such an instruction and who, if anyone, should determine whether the witness and the person identified are of different races, perhaps because the cross-racial character of an identification is often not contested. See, e.g., Commonwealth v. Engram, 43 Mass. App. Ct. 804, 805 n.1 (1997) (defendant was black and at argument "it was agreed that the identifying witnesses were white").

The social science research establishing the CRE often does not define race. See Chiroro, Tredoux, Radaelli, & Meissner, Recognizing Faces Across Continents: The Effect of Within-Race Variations on the Own-Race Bias in Face Recognition, 15 Psychonomic Bull. & Rev. 1089, 1091 (2008) ("Face recognition researchers have investigated the [CRE] for almost [forty] years, but few have attempted to provide a definition of race. This is not surprising, since the concept of race is notoriously unclear, with most biologists asserting that it has no defensible definition" [emphasis in original]). Roy S. Malpass, a leading scholar on the CRE, highlights the difficulty of defining race in this area of research, stating, "There seems to be no good and consistent way to refer to all the various 'races,'" and "the old racial names just don't seem to work, especially in complex multiethnic societies." They All Look Alike to Me, in The Undaunted Psychologist: Adventures in Research 77 (1993) ("This problem has not been solved in a satisfying way. We have to acknowledge it, and get on with the inquiry about facial recognition -- even if we have to communicate by using some not so terribly appropriate terminology"). For example, what is the race of a person whose grandparents on his father's side were an African-American and an Asian-American, and on his mother's side were a Caucasian and a Native American? See id. And what evidence would be admissible to ascertain the person's race? [Note 15]

Page 26

In facial recognition studies, the person making the identification is generally asked to self-identify his or her race, and that self-identification is accepted as the person's race for purposes of the study; [Note 16] the race of the person who is identified is generally determined based on the physical appearance of the person's face, including but not limited to skin color. [Note 17] Although social scientists "refer to the phenomenon as the [CRE] . . . the operative factor is perceived facial physiognomic characteristics, regardless of racial classification per se." Wells & Olson, The Other-Race Effect in Eyewitness Identification: What Do We Do About It?, 7 Psychol., Pub. Pol'y, & L. 230, 234 (2001). See McKone, Stokes, Liu, Cohan, Fiorentini, Pidcock, Yovel, Broughton, & Pelleg, A Robust Method of Measuring Other-Race and Other-Ethnicity Effects: The Cambridge Face Memory Test Format, 7 PLOS ONE, no. 10, Oct. 2012, at 1 (McKone) ("we use the term race of a face to refer to the relatively large physical differences in faces with ancestry from different major world regions, such as

Page 27

Europe, Asia, or Africa" [emphasis in original]). [Note 18] In short, when we speak of cross-racial identification in the context of eyewitness identification, we mean that based on facial appearance, the person who made the identification is likely to have perceived the person identified to be of a different race.

Because differences in race based on facial appearance lie in the eye of the beholder, we shall not ask judges to determine whether a reasonable juror would perceive the identification to be cross-racial. Rather, we shall direct that a cross-racial instruction be given unless all parties agree that there was no cross-racial identification. This obviates any need for the judge to decide whether the identification was actually cross-racial, or whether jurors might perceive it to be. If the jury receive such an instruction but do not think the identification was cross-racial, they may simply treat the instruction as irrelevant to their deliberations. Consequently, we amend our provisional instruction in Gomes to the extent that, in criminal trials that commence after the issuance of this opinion, the following instruction should be included when giving the model eyewitness identification instruction, unless all parties agree to its omission: "If the witness and the person identified appear to be of different races, you should consider that people may have greater difficulty in accurately identifying someone of a different race than someone of their own race." [Note 19]

Page 28

We also take this opportunity to consider whether a cross-ethnic instruction should be included with the cross-racial instruction. Ethnicity is generally distinct from race; for instance, a person who identifies as Hispanic may be of any race. [Note 20] Yet, in facial recognition studies, the terms "race" and "ethnicity" are often conflated and used interchangeably; when they are defined separately, ethnicity may refer to "the smaller physical differences that exist within a race, such as with ancestry from Norway versus Greece within Europe, or China versus Japan within Asia, or Nigeria versus Ethiopia within Africa." McKone, supra at 1. From our review of the social science, we are aware of studies that support the conclusion that people are better at recognizing the faces of persons of the same ethnicity than a different ethnicity. [Note 21] But there is not yet a near consensus in the relevant scientific

Page 29

community that people are generally less accurate at recognizing the face of someone of a different ethnicity than the face of someone of their own ethnicity. See American Bar Association Policy 104D: Cross-Racial Identification, 37 Sw. U. L. Rev. 917, 927 (2008) ("The research on cross-ethnicity identification is less clear-cut"); State v. Romero, 191 N.J. 59, 71 (N.J. 2007) ("The most that can be said is that research in the area has begun"). See also McKone, Hall, Pidcock, Palermo, Wilkinson, Rivolta, Yovel, Davis, & O'Connor, Face Ethnicity and Measurement Reliability Affect Face Recognition Performance in Developmental Prosopagonosia: Evidence from the Cambridge Face Memory Test -- Australian, 28 Cognitive Neuropsychol. 109, 135 (2011) ("The existence of other-race effects on face memory is well established. However, the question of whether . . . ethnicity of faces within a race influences face recognition has received less attention"). In Romero, supra at 66, a non-Hispanic Caucasian male identified a Hispanic male that the trial court determined was also Caucasian. The New Jersey Supreme Court investigated the social science research and concluded that "[s]ocial science research does not tie identification unreliability directly to ethnic differences in the same way that racial differences can affect identification reliability." Id. at 63.

For now, we leave the decision to add ethnicity to the cross-racial instruction in the judge's sound discretion. [Note 22] Where the persons involved in the identification self-identify as being of the same race but different ethnicity, and look as categorically different as people of different races, a cross-ethnic instruction will generally be appropriate, because the research suggests that cross-ethnic facial recognition in these circumstances has reliability issues similar to the CRE. See note 21, supra. Ethnicity should

Page 30

also generally be included in the instruction where, for example, a non-Hispanic eyewitness identifies a defendant who is Hispanic and looks multiracial, because the jury may not know whether to attribute the difference in appearance to race or ethnicity or both; an instruction that only references race might inadvertently lead the jury to ignore facial characteristics that are relevant to the CRE. [Note 23] But the addition of ethnicity may not be appropriate where, for example, the witness and the person identified appear to be of the same race, but one is Australian and the other North American. [Note 24] Until the social science reaches a near consensus, we will not require the inclusion of ethnicity in the model instruction but will leave its inclusion to the discretion of the judge based on the circumstances of the identification.

2. Admission of in-court eyewitness identifications. The first time that the victim, Kenniston, and Liebel made a formal identification of the defendant was in court during their testimony. The defendant now challenges the admission of these in-court identifications. Because there was no motion to suppress or objection at trial, "the error, if any, is reviewed for a substantial risk of a miscarriage of justice." Commonwealth v. Brown, 451 Mass. 200, 207 (2008).

We recently announced the following prospective rule in Commonwealth v. Crayton, 470 Mass. 228, 241-242 (2014), to be applied in trials that commence after the issuance of the opinion on December 17, 2014: "Where an eyewitness has not participated

Page 31

before trial in an identification procedure, we shall treat the in-court identification as an in-court showup, and shall admit it in evidence only where there is 'good reason' for its admission." See Commonwealth v. Collins, 470 Mass. 255, 265 (2014) (we shall prospectively apply new rule in Crayton where, before trial, witness made "something less than an unequivocal positive identification" during nonsuggestive procedure). Because the defendant's trial took place before the issuance of Crayton and Collins, those prospective rules do not apply in this case. Instead, we evaluate the alleged errors under the existing law at the time of trial. See Crayton, supra at 245 (no abuse of discretion in admission of in-court identifications where doing so was "in accord with the case law existing at the time of [the judge's] decision"). See also Collins, supra at 261 (defense counsel not ineffective for failing to object to admission of in-court identification when its admission "conformed to our case law").

Prior to Crayton, an in-court identification was excluded if, in the totality of the circumstances, it was "tainted by an out-of-court confrontation . . . that [was] 'so impermissibly suggestive as to give rise to a very substantial likelihood of irreparable misidentification.'" Crayton, supra at 238, quoting Commonwealth v. Carr, 464 Mass. 855, 877 (2013). An in-court identification was admissible in the absence of any prior out-of-court confrontation. Crayton, supra ("where there had been no out-of-court identification to taint the in-court identification, the judge's admission of the in-court identification conformed to our case law"). Because the defendant does not assert that any out-of-court confrontation took place involving Liebel, there was no error in the admission of his in-court identification.

Page 32

We also find no error in the admission of the in-court identification made by Kenniston. Kenniston saw the defendant in police custody near the crime scene when he drove the victim to the hospital, and a police officer directing traffic said to Kenniston, "Yeah, we got [the assailant]. He's right over there." But a few moments before this exchange, from close range and in a well-lit area, Kenniston not only observed the defendant punch the victim but also saw him flee and get chased, tackled, and arrested. Under these circumstances, the officer's confirmation that the police arrested the assailant told Kenniston nothing more than what he had seen with his own eyes, and was not so impermissibly suggestive as to create a substantial likelihood of irreparable misidentification.

We also find no error in the admission of the in-court identification made by the victim. The victim testified that between the original event and the trial, he saw a photograph of the defendant in a newspaper given to him by police. We recognize that it might be unnecessarily suggestive for police to provide a newspaper article about the relevant crime to a witness and ask whether the person shown in the newspaper photograph is the assailant. Compare Commonwealth v. Jules, 464 Mass. 478, 489-490 (2013), quoting Commonwealth v. Horton, 434 Mass. 823, 835 (2001) (in absence of police manipulation or prompting, "simple exposure to the media is not sufficient ground to suppress an identification [on constitutional grounds]"). But the level of suggestiveness ultimately depends on the context of the confrontation. See Commonwealth v. Cavitt, 460 Mass. 617, 632 (2011), quoting Commonwealth v. Miles, 420 Mass. 67, 77 (1995) (unnecessarily suggestive procedure must be proved "in light of the totality of the circumstances"). Here, there is no evidence in the record detailing the circumstances of the victim's viewing of the newspaper photograph, such as what the police said to him, or how long after the initial event he saw it. It was the defendant's burden to prove by a preponderance of the evidence that any out-of-court confrontation with the victim was so impermissibly suggestive as to create a substantial likelihood of irreparable misidentification. The defendant has failed to satisfy that burden, especially where the victim, after the first "sucker punch" blinded him in one eye, saw the assailant stand within "a foot or two" of him and punch him once or twice more in the face. [Note 25]

3. Consciousness of guilt instruction. The judge instructed the jury regarding consciousness of guilt based on the evidence that the assailant ran away after punching the victim and discarded the

Page 33

brass knuckles. [Note 26] Because the consciousness of guilt instruction was given over the defendant's objection, we review for prejudicial error. See Commonwealth v. Stuckich, 450 Mass. 449, 452-453 (2008).

An instruction on consciousness of guilt is appropriate where the jury may draw an inference of guilt "'from evidence of flight, concealment, or similar acts,' such as false statements to the police, destruction or concealment of evidence, or bribing or threatening a witness." Commonwealth v. Morris, 465 Mass. 733, 737-738 (2013), quoting Stuckich, supra at 453. A defendant's flight is often considered "classic evidence" of consciousness of guilt. Commonwealth v. Vick, 454 Mass. 418, 426 (2009). The inference of guilt may be drawn in part from the premise that a person flees "because he feels guilt concerning that act" and the person feels guilt concerning the act because he "committed that act." Commonwealth v. Toney, 385 Mass. 575, 584 (1982).

In contrast, a consciousness of guilt instruction regarding flight is generally inappropriate where there is no dispute that the crime was committed by the person fleeing from the crime scene, and the only contested issue is the identification of the defendant as

Page 34

the fleeing offender. See Commonwealth v. Pina, 430 Mass. 266, 272 (1999), citing Commonwealth v. Groce, 25 Mass. App. Ct. 327, 331-332 (1988). Under these circumstances, if the jury finds that it was the defendant who fled, the jury will find him guilty, not because flight evidences consciousness of guilt but because flight reveals the defendant to be the assailant. See Vick, 454 Mass. at 439 (Botsford, J., concurring in part and dissenting in part) ("for the jury to consider the evidence that the assailant 'fled' from the immediate scene of the shooting as consciousness of guilt on the defendant's part, they would need first to conclude, based on separate evidence, that the defendant was in fact the shooter; otherwise, they would have no basis on which to ascribe the act of fleeing to the defendant at all"); Groce, supra at 331-332 (consciousness of guilt instruction was "inapposite" where "[t]here [was] no dispute that the same individual committed the offense and fled from the scene"). In these circumstances, a consciousness of guilt instruction would provide no relevant guidance to the jury but would pose the risk that the jury might think the judge was suggesting that the defendant was the person who fled and therefore the person who committed the crime. Groce, supra at 332 (judge who gave consciousness of guilt instruction "may well have conveyed the notion to the jury that he believed that it was the defendant who fled and, thus, that the victim's identification testimony was accurate"). Where, as here, the only live issue at trial was identification and it was plain that the person who fled was the assailant, the risk that a consciousness of guilt instruction might imply that the defendant was the person who fled outweighed the negligible benefit of instructing the jury that flight may be evidence of consciousness of guilt. Therefore, the judge erred in giving the instruction.

The error, however, was not prejudicial for two reasons. First, the evidence of the defendant's guilt in this case was overwhelming. Although the defendant contends the case rests solely on three unreliable eyewitness identifications, the most compelling evidence of guilt comes from Kenniston's testimony that the person he saw "sucker punch" the victim was the same person whom he saw flee and be tackled by the police. This testimony did not rest on facial recognition; it would have mattered little if Kenniston had never seen the assailant's face. Where the defendant was the only person tackled by the police, and Kenniston saw that it was the assailant who was tackled, Kenniston's testimony

Page 35

provided compelling evidence of the defendant's guilt. [Note 27] The eyewitness identifications of Liebel and the victim corroborated Kenniston's testimony. Liebel's testimony was stronger because, unlike the victim, he had not been blinded by a punch and saw the defendant menacingly approach within a few feet of him after having beaten the victim. Although Liebel lost sight of the defendant after watching him run, Liebel reported the direction where the defendant fled and the defendant was soon tackled, arrested, and brought back to where Liebel was holding Juan. There was no formal show-up identification, but Liebel saw that the defendant was the person the police had arrested, and he did not tell his fellow officers that they arrested the wrong man. The defendant's testimony is also so inconsistent with all the other evidence in the case and so improbable that it "adds to our confidence in the jury's verdict." Commonwealth v. Rosado, 428 Mass. 76, 81 (1998) ("sheer implausibility of defendant's own alibi" supported conclusion of no prejudicial error). In view of the overwhelming weight of the evidence against the defendant, we are confident that the jury's verdict "was not substantially swayed by the error" in giving the consciousness of guilt instruction. See Commonwealth v. Flebotte, 417 Mass. 348, 353 (1994).

Second, the risk that the jury understood the judge to be suggesting that the defendant was the person who fled was small. The judge began her instruction by saying that "[t]here has been evidence in this case alleging that [the defendant] may have fled when he was about to be arrested . . . and/or that he may have intentionally tried to conceal . . . what is alleged to be a dangerous weapon." The inclusion of the word "alleging" made clear that the jury needed to evaluate the quality of that evidence, and that the judge did not intend to suggest that the evidence should be credited. Moreover, the judge gave the consciousness of guilt instruction immediately following the eyewitness identification instruction, in which the judge emphasized several times that the Commonwealth bears the burden of proving the defendant's identity beyond a reasonable doubt. See Pina, 430 Mass. at 272 (no prejudicial error in giving consciousness of guilt instruction where judge emphasized that prosecutor bore burden of proving identity of defendant as perpetrator). Therefore, giving the consciousness of guilt instruction did not constitute prejudicial error.

Conclusion. The judgments of conviction against the defendant are affirmed.

So ordered.