ELLSWORTH v. WARDEN NEW HAMPSHIRE STATE PRISON (2003)

Raymond Ellsworth (“Ellsworth”) appeals the district court's denial of his 28 U.S.C. § 2254 petition for a writ of habeas corpus.   A New Hampshire jury convicted Ellsworth of seven counts of sexual assault, all based on the testimony of one emotionally disturbed eleven-year-old boy whom Ellsworth supervised at a residential treatment facility.   After the New Hampshire Supreme Court upheld his convictions, Ellsworth filed a § 2254 petition in federal district court, alleging three violations of his Sixth and Fourteenth Amendment rights at trial:  (1) He was wrongfully prohibited from cross-examining his juvenile accuser about prior sexual experiences that would have provided sufficient sexual knowledge to fabricate realistic accusations;  (2) he was denied access to exculpatory impeachment materials relating to the complainant, which the trial court reviewed in camera but refused to disclose to the defense;  and (3) he was not allowed to present testimony from a witness who observed the complainant make false accusations of sexual voyeurism against others, after the complainant, on cross-examination, denied ever making such accusations.

After two rounds of summary judgment motions, the district court (DiClerico, J.) granted summary judgment to the respondents on all three claims, finding that insofar as the state courts had erred in curtailing Ellsworth's cross-examination of his accuser, they had done so “harmlessly.”   As to the denial of access to impeachment evidence, a claim not subject to harmless error review, Kyles v. Whitley, 514 U.S. 419, 435-36, 115 S.Ct. 1555, 131 L.Ed.2d 490 (1995), the district court found no error because the materials in question, viewed in context, were “not of significant value.”   Finally, the district court held that the trial court's exclusion of the impeachment testimony was neither contrary to, nor an unreasonable application of, Supreme Court precedent.   Because we disagree with all three characterizations, we now reverse.

This case tremendously disturbs this Court.   The crimes, if they occurred, deserve substantial punishment-and, indeed, they may have happened.   But if we were to uphold the conviction, we would be doing so on the testimony, and only the testimony, of a very disturbed young boy.   There was no corroboration of the boy's testimony as to the criminal acts themselves.   The jury had no information that would help it to test the boy's credibility because much relevant, probative, and useful information was withheld from it.   The possibility that an innocent man was convicted is very real.

I. FACTS

From October 1988 through September 1992, Ellsworth worked at Spaulding Youth Center, a residential school and treatment facility for children with emotional, behavioral, and/or neurological impairments.   His position was that of “cottage teacher” for Colcord Cottage, a dormitory in the boys' program that housed emotionally disturbed boys between approximately six and twelve years of age.   Ellsworth supervised the boys in his group after school, during meals, and at other times during the day and night, organized sports activities, took students on off-campus bicycling, swimming, and field trips, and met with several students on a weekly basis for “kid meetings.”   Approximately once a week, Ellsworth slept in the staff room at the cottage and, when necessary, was awakened to handle any problems that arose during the night.

During a therapy session that took place in November 1992, Matthew, an eleven-year-old boy living in Colcord Cottage who participated in weekly “kid meetings” with Ellsworth, accused Ellsworth of sexually abusing him.   Matthew described three incidents of molestation that he said occurred during the spring and summer of 1992:  one while he was returning home with Ellsworth from a bicycle trip;  one during a swimming outing with Ellsworth and Stephen, another boy from Spaulding;  and a third late at night at Colcord Cottage after Matthew returned from a home visit.   Ellsworth denied all the charges, but was indicted on four counts of aggravated felonious sexual assault and eight counts of felonious sexual assault.

Before trial, the defense sought discovery of privileged materials relating to the charges against Ellsworth:  a dormitory incident report for the night of the alleged third incident of sexual assault, records relating to Matthew's prior sexual victimization and his first accusation against Ellsworth, and records relating to Matthew's disciplinary history.   The trial court reviewed these materials in camera, but refused to disclose them to defense counsel.

At trial, in January 1995, Matthew testified about the three alleged incidents of sexual abuse.   As to the first, Matthew testified that, during the summer of 1992, he and Ellsworth went on a bicycle ride near the Spaulding campus.   On the way back, he related, Ellsworth lured him into the woods by claiming he heard a noise, and then pulled down his pants and told Matthew to touch and put his mouth on his penis.   Matthew testified that Ellsworth's penis was soft when he first took his pants down, but was hard when it was in Matthew's mouth.   Matthew also testified that Ellsworth took off Matthew's shorts and touched his (Matthew's) private parts.   After it was over, Matthew stated, Ellsworth told him not to tell anyone about the incident, or he would get hurt.

Matthew was questioned, both on direct and on cross-examination, about the location in the woods where the alleged incident occurred.   On direct examination, Matthew testified that, in order to get into the woods with Ellsworth, he had to step over part of a barbed wire fence that was “pushed down,” and that there was a broken tree stump along with three other trees at the spot where Ellsworth molested him.   On cross-examination, Matthew explained that, at some point after he reported the abuse, Police Chief Leary (“Leary”), Trooper Nolan (“Nolan”), and Doug Beaton (“Beaton”), who worked at the Division of Children and Youth Services, took him into the woods and asked him to point out where Ellsworth had molested him.   Matthew testified that he went with Leary, Beaton, and Nolan in a car and told the driver to stop the car when it neared the point where the incident occurred.   Matthew and the three adults then got out of the car, where Matthew stated that he recognized the trees and the barbed wire, and Beaton marked the spot by placing a knife mark on a nearby telephone pole.

Ellsworth, who testified at trial, denied that he had ever taken Matthew into the woods alone, much less sexually molested him.   He explained that he had often taken groups of students (which might have included Matthew) on bicycle trips to the woods, but that he had never done so alone with Matthew.   In addition, Stephen, another resident, testified that Matthew had told him a different story:  that the abuse incident occurred when Matthew stopped on the bicycle trip to urinate and Ellsworth followed him into the woods.

As to the second incident, Matthew testified that Ellsworth had taken him and Stephen on a swimming trip to Sandoggerdy Pond, near Spaulding.   Matthew related that it was a warm summer day, and that there was a beach with picnic tables next to the pond, as well as little houses nearby, but that there was no one else at the pond that day other than Ellsworth and the two boys.   Matthew described the water itself as “mucky,” with leaves on the bottom-so mucky that, if he went underwater and opened his eyes, he couldn't see very much.   Matthew related that, after playing with both boys in the water, Ellsworth asked Stephen to swim away.   Then, with Matthew standing in the water up to his chest, Ellsworth pulled down both of their bathing suits, touched Matthew's penis and buttocks, and told Matthew to put his mouth on Ellsworth's penis (which he did by going underwater).  (On cross-examination, but not on direct, Matthew testified that Ellsworth anally raped him underwater as well.)   Matthew testified that Ellsworth again told him not to tell anyone about the incident and threatened him.

Ellsworth testified that he remembered going swimming at the pond with Matthew and Stephen, but denied that any sexual incident had occurred.   Stephen testified that he had been on two such trips with Matthew and Ellsworth, and related that he had seen nothing of a sexual nature transpire between Matthew and Ellsworth.   He also related that there were other people at the beach playing with children and swimming, that he was swimming and talking to people within twenty feet of Matthew and Ellsworth but not paying close attention to them, that Matthew was playing with Ellsworth in the same way he did with other staff members, and that he neither saw nor heard anything unusual.

Finally, Matthew testified that the third incident of abuse occurred when he returned early to Spaulding from a weekend home-visit because he had been misbehaving at home.   When he arrived back at Colcord Cottage, Ellsworth was the only staff person on duty in the cottage, and no one was asleep in the staff bedroom.   Matthew testified that, while he was putting on his pajamas, Ellsworth came into his room and began touching him, and, at Ellsworth's direction, Matthew touched and put his mouth on Ellsworth's penis, as he had done at the pond.   Ellsworth testified that he had never sexually abused Matthew at this or any other time.   The Executive Director of Spaulding testified that the policy at Spaulding was not to have one staff member alone in a cottage, and another Spaulding resident, who was Matthew's roommate beginning in October 1992, testified that Matthew had told him that Ellsworth molested him in the afternoon at Colcord Cottage during a “kid meeting,” and that nothing happened at night.

Of relevance here are two bodies of evidence that the defense sought to introduce during trial.   The first is evidence pertaining to Matthew's prior sexual abuse, which occurred when he was about three years old.   The district court related that the man accused in 1985 of abusing Matthew was charged with acts of fellatio and digital-anal intercourse, but the charges against him were dropped when it was determined that Matthew was not competent to testify against him.   The trial court excluded the evidence on relevance grounds because (1) the abuse was different from the abuse alleged against Ellsworth and (2) the defense had not proved its theory that the prior abuse gave Matthew an alternative source for the information used in his allegations against Ellsworth.

The second body of evidence at issue here is the testimony of Craig Klare, a counselor who supervised Matthew at the Pine Haven School, another residential treatment and educational facility for emotionally disturbed boys, where Matthew lived after he left Spaulding.   The defense proffered that Klare would testify that Matthew made two false accusations of sexual voyeurism-other boys attempting to peep at him in the shower or the toilet-and a false accusation of theft against other residents at Pine Haven.   The trial court excluded the testimony, but permitted cross-examination of Matthew about these incidents.   Matthew denied that any of them had ever occurred.

Following trial, Ellsworth was convicted of two counts of aggravated felonious sexual assault and five counts of felonious sexual assault;  he was acquitted of one count of aggravated felonious sexual assault, and the state nol prossed one other count of aggravated felonious sexual assault and three counts of felonious sexual assault.   Ellsworth was sentenced to eighteen and one half to thirty-seven years in prison, stand committed, and fourteen to twenty-eight years, deferred.   Both his conviction and his sentence rested entirely on the testimony of Matthew, an emotionally disturbed thirteen-year-old boy;  there was no other evidence-no physical evidence, no doctor's report, no witness testimony-that any sexual assault had ever occurred.

Ellsworth's convictions were affirmed on appeal to the New Hampshire Supreme Court.  State v. Ellsworth, 142 N.H. 710, 709 A.2d 768 (1998).   Ellsworth filed the instant habeas petition on March 23, 1999, alleging three grounds for relief:  (1) The trial court violated his Sixth Amendment right to confront his accuser by prohibiting him from cross-examining Matthew about prior sexual experiences;  (2) the trial court violated his Fourteenth Amendment rights under Pennsylvania v. Ritchie, 480 U.S. 39, 107 S.Ct. 989, 94 L.Ed.2d 40 (1987), and Brady v. Maryland, 373 U.S. 83, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963), by denying him access to exculpatory materials that the court reviewed in camera;  and (3) the trial court violated his Fourteenth Amendment right to present impeachment evidence by excluding the testimony of Craig Klare.

After several attempts to ensure that the record before the district court was complete, the parties filed for summary judgment in the fall of 2000.   On February 23, 2001, the district court granted summary judgment to the respondents on Ellsworth's third claim, but denied both parties' summary judgment motions on the other two grounds, inviting them to file renewed motions “presenting a record expanded by the disclosed information and including the harmless error prong of habeas review.”   On January 23, 2002, the district court granted summary judgment to the respondents on Ellsworth's other two claims.   The district court issued a certificate of appealability (“COA”) as to all three claims on February 21, 2002.

II. LEGAL ANALYSIS

A. Standard of Review

 “We review the district court's denial of habeas relief de novo.”  Nadeau v. Matesanz, 289 F.3d 13, 15 (1st Cir.2002).   The general rule for habeas petitions governed by the Antiterrorism and Effective Death Penalty Act (“AEDPA”), 28 U.S.C. § 2254, is that, in order to prevail, the petitioner must show that the state court decision was either contrary to federal constitutional law or an unreasonable application of clearly established federal law as established by the Supreme Court.  28 U.S.C. § 2254(d)(1).¹  In Williams v. Taylor, 529 U.S. 362, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000), the Supreme Court provided further elaboration on the AEDPA standards, as follows:  A state court decision is contrary to clearly established precedent if the state court either (1) applies a rule that contradicts the governing law set forth in Supreme Court cases, or (2) confronts a set of facts materially indistinguishable from those in a Supreme Court decision and nevertheless arrives at a different result.  Id. at 405-06, 120 S.Ct. 1495;  see also, e.g., Hurtado v. Tucker, 245 F.3d 7, 15 (1st Cir.2001);  Phoenix v. Matesanz, 233 F.3d 77, 80 (1st Cir.2000).   The question of whether a state court's determination is an “unreasonable application” of the precedent in question is based on an objective standard-one that requires more than mere incorrectness-but beyond that the Supreme Court left it up to the federal judiciary to determine objective reasonableness on a case-by-case basis.  Williams, 529 U.S. at 410-11, 120 S.Ct. 1495.

 In this Circuit, however, the AEDPA deferential standard of review applies only where the state court actually addressed the merits of the petitioner's federal claim.  Fortini v. Murphy, 257 F.3d 39, 47 (1st Cir.2001).   Where the state court did not address the federal issue, the standard is de novo review.  Id.

Accordingly, two different standards of review govern this case.   Because, as the district court observed, the New Hampshire Supreme Court did not address the federal aspects of Ellsworth's claims based on the Confrontation Clause and the denial of access to exculpatory materials, these claims are subject to de novo review by this Court.   However, as to Ellsworth's claim regarding his right to present extrinsic impeachment evidence,² the New Hampshire Supreme Court expressly declined to undertake a separate federal analysis on the grounds that the state law claim failed and federal law provided no additional protection that would have dictated a different outcome.  142 N.H. at 718, 709 A.2d 768.   Thus, this claim is subject to heightened deferential review under AEDPA.

B. Confrontation Clause

1. Violation of the right

 The district court found that Ellsworth's Confrontation Clause rights were in fact violated by the trial court's refusal to allow cross-examination of Matthew on the issue of his prior sexual abuse.   Prior to trial, the defense learned that Matthew had been sexually abused by a babysitter when he was approximately three years old.   In December 1994, defense counsel moved to be allowed to cross-examine Matthew at trial about his prior sexual abuse, arguing that Matthew's prior experience was relevant both (1) as an explanation for why he could describe sexual acts in such detail and (2) to show the possibility that he was fabricating his claim against Ellsworth.   Matthew's guardian ad litem maintained that the abuse should not be brought up at trial due to its being too far in the past and the possible detrimental effect of such cross-examination on Matthew.

The trial court ultimately ruled that the prior abuse was “irrelevant” because it was too different in nature from the conduct with which Ellsworth was charged, and because Ellsworth did not prove that it was the prior abuse, rather than some other source of information, that might have provided Matthew with his sexual knowledge.   The trial judge allowed defense counsel to cross-examine Matthew only as to the general facts of his participation in group therapy, and the fact that the group discussed the difference between “good touch” and “bad touch.”   As a result, at trial, Matthew testified that he was taught about “good touch” and “bad touch” at Spaulding and that he participated in group discussions about good and bad touch, families, and feelings.   No other evidence of Matthew's prior sexual abuse was introduced at trial;  the jury knew only that Matthew was at Spaulding and had participated in group therapy, not that he had been sexually abused earlier in his life.

 As the district court explained, Davis v. Alaska, 415 U.S. 308, 94 S.Ct. 1105, 39 L.Ed.2d 347 (1974), and Delaware v. Van Arsdall, 475 U.S. 673, 106 S.Ct. 1431, 89 L.Ed.2d 674 (1986), establish the parameters of the criminal defendant's rights under the Confrontation Clause.   The Davis Court held:

“The main and essential purpose of confrontation is to secure for the opponent the opportunity of cross-examination.”  ․ Cross-examination is the principal means by which the believability of a witness and the truth of his testimony are tested.   Subject always to the broad discretion of a trial judge to preclude repetitive and unduly harassing interrogation, the cross-examiner is not only permitted to delve into the witness' story to test the witness' perceptions and memory, but the cross-examiner has traditionally been allowed to impeach, i.e., discredit, the witness.

415 U.S. at 315-16, 94 S.Ct. 1105;  accord Van Arsdall, 475 U.S. at 680, 106 S.Ct. 1431.   Thus, the Davis Court went on to explain, in order to make cross-examination effective, “defense counsel should [be] permitted to expose to the jury the facts from which jurors, as the sole triers of fact and credibility, could appropriately draw inferences relating to the reliability of the witness.”  415 U.S. at 318, 94 S.Ct. 1105.   Because defense counsel in Davis, although permitted to ask the witness whether he was biased, was not allowed to “make a record from which to argue why [the witness] might have been biased or otherwise lacked that degree of impartiality expected of a witness at a trial,” the Supreme Court found that Davis's rights under the Confrontation Clause had been violated.  Id. Significantly, as we have observed, “a criminal defendant's entitlement to cross-examine a witness increases in sensitivity in direct proportion to the witness's importance to the prosecution's case.”  Bui v. DiPaolo, 170 F.3d 232, 241-42 (1st Cir.1999).

 Of course, the defendant's right of confrontation is not unlimited.   As the Van Arsdall Court put it:

[T]rial judges retain wide latitude insofar as the Confrontation Clause is concerned to impose reasonable limits on such cross-examination based on concerns about, among other things, harassment, prejudice, confusion of the issues, the witness' safety, or interrogation that is repetitive or only marginally relevant.   And as we observed earlier this Term, “the Confrontation Clause guarantees an opportunity for effective cross-examination, not cross-examination that is effective in whatever way, and to whatever extent, the defense might wish.”  Delaware v. Fensterer, 474 U.S. 15, 20, 106 S.Ct. 292, 88 L.Ed.2d 15 (1985) (per curiam) (emphasis in original).

475 U.S. at 679, 106 S.Ct. 1431;  accord Bui, 170 F.3d at 242.   Nonetheless, the Court concluded, “[b]y [ ] cutting off all questioning about an event that the State conceded had taken place and that a jury might reasonably have found furnished the witness a motive for favoring the prosecution in his testimony, the court's ruling violated respondent's rights secured by the Confrontation Clause.”  475 U.S. at 679, 106 S.Ct. 1431.

In this case, the district court correctly rejected the trial court's rationale for denying Ellsworth the right to cross-examine Matthew concerning his earlier sex abuse.   The district court observed that the state courts had all ignored the common element of fellatio in both Matthew's earlier experiences and the charged conduct in this case, and concluded that the evidence at issue “might have provided facts from which jurors could have appropriately drawn inferences related to the reliability of Matthew as a witness.”   We agree with the district court that Ellsworth's rights under the Confrontation Clause were violated when the trial judge failed to allow him to cross-examine Matthew on his earlier sexual abuse.

2. Harmless error

 Constitutional violations resulting from trial error, such as a violation of the Confrontation Clause, are subject to harmless error review.   Van Arsdall, 475 U.S. at 682, 106 S.Ct. 1431;  see also Brecht v. Abrahamson, 507 U.S. 619, 629-30, 113 S.Ct. 1710, 123 L.Ed.2d 353 (1993).   In Brecht, the Supreme Court adopted for habeas cases the harmless error standard from Kotteakos v. United States, 328 U.S. 750, 776, 66 S.Ct. 1239, 90 L.Ed. 1557 (1946);  accordingly, the test is “whether the error had substantial and injurious effect or influence in determining the jury's verdict.   Under this standard, habeas petitioners may obtain plenary review of their constitutional claims, but they are not entitled to habeas relief based on trial error unless they can establish that it resulted in ‘actual prejudice.’ ”  507 U.S. at 637, 113 S.Ct. 1710 (internal citation omitted).   Accord O'Neal v. McAninch, 513 U.S. 432, 436, 115 S.Ct. 992, 130 L.Ed.2d 947 (1995) (“When a federal judge in a habeas proceeding is in grave doubt about whether a trial error of federal law had substantial and injurious effect or influence in determining the jury's verdict, that error is not harmless.”);  Sanna v. Dipaolo, 265 F.3d 1, 14 (1st Cir.2001).

 The Van Arsdall Court enumerated several factors that go into assessing whether error is harmless, including “the importance of the witness' testimony in the prosecution's case, whether the testimony was cumulative, the presence or absence of evidence corroborating or contradicting the testimony of the witness on material points, the extent of cross-examination otherwise permitted, and, of course, the overall strength of the prosecution's case.”  475 U.S. at 684, 106 S.Ct. 1431.

 In this case, the district court concluded that the trial court's error was harmless.   Despite the fact that “Matthew's credibility was crucial to the state's case,” the district court found that there was little evidence to explain why Matthew would fabricate the charges against Ellsworth.   It also found the “graphic detail of Matthew's testimony [ ] compelling,” and stated that “[t]he detail with which the incidents are recounted militates against the possibility they were fabricated.”   Particularly in light of Matthew's young age at the time of his original sexual abuse, the district court found it unlikely that his earlier abuse supplied the factual basis for Matthew's testimony about Ellsworth.   The court also noted that Matthew testified that he had liked Ellsworth until the abuse began.   Thus, the district court determined that, had the jury been supplied with more information about Matthew's sexual history, that information would not have “undermined” Matthew's graphic descriptions of what Ellsworth allegedly had done to him, and moreover that it might have suggested that Matthew was ready prey for someone who, like Ellsworth, knew his sexual history.

We disagree with the district court's conclusion that this error was harmless, and find that the trial court's bar on cross-examination of Matthew does indeed give rise to “grave doubt about whether a trial error of federal law had substantial and injurious effect or influence in determining the jury's verdict.”  O'Neal, 513 U.S. at 436, 115 S.Ct. 992.   The trial court made a critical mistake when it denied Ellsworth the opportunity to present the jury with facts from which it could draw a more accurate picture of Matthew's emotional and psychological condition.

The jury at Ellsworth's trial heard no evidence whatsoever that Matthew had been sexually abused prior to the incidents involving Ellsworth;  the most that it knew was that Matthew was at Spaulding, that he had emotional and/or behavioral problems of some sort, and that he had participated in a group that discussed good and bad touching, among other things.   The facts surrounding Matthew's prior molestation are crucial for several reasons.   First, the district court concluded that the prior molestation could not have provided Matthew with knowledge or experience enabling him to make a false accusation of Ellsworth.   Matthew's prior abuse is not only salient because it might have provided him with conscious knowledge of details about sex that could be used to fabricate charges, but, more importantly, knowing of Matthew's early abuse would have given the jury some insight into the mental and emotional damage he sustained as a result of that abuse.   The long-term effects of that damage are enhanced by his emotional difficulties.   The jury would have known that Matthew was a child who had been forcibly “awakened” sexually at a young age, and that there were major reasons other than his exposure to Ellsworth why he might have a heightened awareness of, and pathological absorption with, sexual matters.   This information was critical to the jury's appraisal of Matthew's credibility as a witness.   But the trial judge excluded it.³

Equally problematic is the district court's suggestion, echoing the arguments of both the defendants and the trial judge in this case, that the amount of “graphic” detail in Matthew's testimony makes that testimony less likely to be fabricated and less likely to be outweighed by knowledge of Matthew's sexual history-in other words, that the error was harmless because of the overwhelming strength of Matthew's testimony.   We do not find this argument persuasive.   First, the corroboration found in the extensive details in Matthew's testimony pertained not to the sexual acts themselves, but to the details about the locale of the bike trip, the physical characteristics of Sandoggerdy Pond and who was along on the trip, and Matthew's early return to Colcord Cottage.   No one in this case disputes that Matthew had been on bike trips to the woods and swim trips to the pond on multiple occasions, and no one disputes that he did return early from a weekend visit home due to his misbehavior.   The issue is as to whether Ellsworth sexually abused Matthew in those settings.   As to that there is no corroboration.⁴

As to the sexual details, the fact that Matthew knew words like “penis,” “dick,” and “butt,” and was able to describe an erection, distinguishes him not at all from most eleven-year-old boys.   The bottom line is this:  We can certainly believe that it was compelling, and sad, to watch a vulnerable young boy (thirteen at the time he testified) recounting graphic details about horrific sexual abuse.   But we do not find the evidence as strong as the district court would have it.   The corroborating evidence goes only to matters other than the sexual assault.   Matthew's testimony was the sole evidence against Ellsworth, and the jury had absolutely no indication that Matthew's perspective about sex might have been skewed by anything other than his interaction with Ray Ellsworth.   Knowledge about Matthew's past would have provided context for his accusations and possible motive;  it would have enabled the jury to see precisely how early and extensive Matthew's abuse was, and would have presented them with more information about the etiology of his current mental and emotional state.

Simply put, this is not harmless error.   Looking at this case in light of the harmless-error factors set forth by the Supreme Court in Van Arsdall, the harmfulness of the error becomes even clearer:

(1) The importance of the accusing witness's testimony in the prosecution's case is impossible to overstate.   The district court itself recognized this:  Matthew was Ellsworth's only accuser and the only witness to the alleged incidents.   Matthew's testimony was the whole of the prosecution's case.

(2) The testimony was not cumulative.   The district court itself observed that there was no other testimony about Matthew's sexual history.   Moreover, in light of the district court's emphasis on the graphic details Matthew recounted, the source of those details was extremely significant and not reached by other testimony.

(3) There was evidence corroborating Matthew's testimony, but not on material points.   The district court thought the corroboration to be material, but it was only marginally so.   The other evidence was only “corroborating” to the extent that there were occasions when Ellsworth took the boys on trips, took them swimming, and stayed overnight at Colcord Cottage.   There was no corroborating evidence of the abuse itself.

(4) The extent of permitted cross-examination was very limited.   Matthew was permitted to testify in an extremely general way that he talked about “good touch” and “bad touch” in group therapy.   This did not provide the jury with any information about his background and prior abuse.

(5) The overall strength of the prosecution's case was not overwhelming.   It involved a sympathetic victim and a lurid accusation, but, at bottom, it turns on the word of an unstable witness-a child at that.

Accordingly, we reject the district court's conclusion that the trial court's error in curtailing Ellsworth's cross-examination of Matthew was “harmless.”   The district court is hereby REVERSED on this claim.

C. Access to Exculpatory Evidence 5

The evidence that the trial court reviewed in camera but refused to disclose to the defense consisted of incident reports pertaining to the Colcord Cottage incident, records of Matthew's early sexual abuse and subsequent treatment, records of Matthew's discipline and behavior in 1992, and notes made by Matthew's therapist.   It included, inter alia, records that Matthew had expressed concern before even arriving at Spaulding that he would be abused there;  that he had made allegations of abuse by staff at Hampstead Hospital, where he had been treated before coming to Spaulding;  that one of the directors at Spaulding, Jan Smith, noted that “special precautions would be necessary to minimize the risk of false allegations by Matthew”;  that Matthew had threatened his parents, saying he knew how to get them, and that his mother was concerned about false allegations of abuse;  that Matthew had accused other Spaulding residents of engaging in or asking him to engage in sexual conduct;  that Matthew was reliving his abuse and having sexual dreams about it;  and that Matthew was participating in an attempt to reopen the case against the perpetrator of his earlier abuse.

The district court agreed with Ellsworth that these materials “pertain[ed] to Matthew's credibility in a manner that is favorable to the defense,” but characterized much of it as information that Ellsworth already knew from other sources (e.g., as one of Matthew's three caseload managers in Colcord Cottage)-information that thus cannot form the basis for a Brady violation.   The district court did acknowledge that three pieces of information-Jan Smith's note that Matthew might make false accusations, Matthew's sexual dreams and recent reliving of his abuse, and the Spaulding staff's interest in Matthew's participation in reopening the case against his earlier perpetrator-were all new information favorable to Ellsworth.   However, the district court found that the information was not material because Ellsworth knew of Matthew's parents' concern that he might make false accusations, and because the added details of Matthew's dreams and the Spaulding staff's focus on reopening Matthew's case “[add] little that would help the defense.”

 In United States v. Bagley, 473 U.S. 667, 676-78, 105 S.Ct. 3375, 87 L.Ed.2d 481 (1985), the Supreme Court held that the test for reversible error for failure to disclose impeachment evidence is the same as that for failure to disclose exculpatory evidence under Brady:  “[A] constitutional error occurs, and the conviction must be reversed, only if the evidence is material in the sense that its suppression undermines confidence in the outcome of the trial.”   Thus the test for a constitutional violation is whether “there is a reasonable probability that, had the evidence been disclosed to the defense, the result of the proceeding would have been different.”  Kyles, 514 U.S. at 433, 115 S.Ct. 1555 (citing Bagley, 473 U.S. at 682, 105 S.Ct. 3375).  The Court elaborated:

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, 514 U.S. at 434, 115 S.Ct. 1555 (citing Bagley, 473 U.S. at 678, 105 S.Ct. 3375).  As the Kyles Court observed, the issue of materiality is to be assessed cumulatively as to all suppressed evidence.  514 U.S. at 437, 115 S.Ct. 1555.

 The district court found that there were essentially two new and undisclosed pieces of evidence-the Jan Smith note expressing concern that Matthew might make false accusations of sexual abuse and the reports bearing on the extent to which Matthew had been reliving and dreaming about his earlier abuse-but concluded that “[t]aken as a whole, in the context of the other information known to the defense, that information was not of significant value.”   We again disagree.   The information that the Spaulding director recognized, and was concerned, that Matthew had a tendency to accuse people of sexual abuse does seriously undermine one's confidence in the outcome of the trial.   The district court maintains that this information is insignificant because Ellsworth knew that Matthew's parents had such a concern, but it is an entirely different matter that a Spaulding director, a professional who works with emotionally disturbed boys on a daily basis, highlighted this as a potential problem.   Especially in light of the extent to which the trial judge curtailed cross-examination of Matthew himself, this information becomes even more important.   Had the jury been presented with this information, along with the information that Matthew was reliving his abuse, it seems to us far from clear that it would have reached the same conclusion as to Matthew's credibility.

Likewise, the fact that Matthew had recently been reliving and having nightmares about his earlier abuse is also significant here.   The jury reached its conclusion that Matthew's allegations were credible with absolutely no information about a crucial part of his life that bore significantly on his mental and emotional state.   We find it extremely troubling that the jury was totally unaware, not only of Matthew's history, but of the extent to which that history was having a particular impact on him at the time when he made his accusations against Ellsworth.

Looking at the evidence cumulatively, as Kyles requires, 514 U.S. at 437, 115 S.Ct. 1555, we find that the suppressed evidence is sufficient to undermine confidence in the jury's verdict.   Accordingly, we REVERSE the district court on the Bagley claim as well.

D. Extrinsic Impeachment Evidence

At trial, Ellsworth had proffered that Craig Klare's testimony would show that Matthew had made false accusations against students at Pine Haven, the school he attended after Spaulding.   Klare, one of Matthew's counselors, related that Matthew falsely accused other boys at the school of peeking at him in the shower, of peeking under his toilet, and of stealing toys that he himself had hidden.   The trial court ruled that Ellsworth could cross-examine Matthew about these incidents, but excluded Klare's testimony.   On cross-examination, Matthew denied that the incidents ever occurred.

The district court found that it was neither contrary to nor an unreasonable application of Supreme Court precedent for the New Hampshire state courts to exclude the testimony of Craig Klare, which would have rebutted Matthew's denials, under New Hampshire Rules of Evidence 6 404(b) (barring introduction of evidence of propensity evidence) and 608(b), which provides:

Specific instances of the conduct of a witness, for the purpose of attacking or supporting the witness' credibility, other than conviction of crime as provided in rule 609, may not be proved by extrinsic evidence.   They may, however, in the discretion of the court, be inquired into on cross-examination of the witness (1) concerning the witness' character for truthfulness or untruthfulness, or (2) concerning the character for truthfulness or untruthfulness of another witness as to which character the witness being cross-examined has testified.

 As the Supreme Court has observed, “[f]ew rights are more fundamental than that of an accused to present witnesses in his own defense.”   Chambers v. Mississippi, 410 U.S. 284, 302, 93 S.Ct. 1038, 35 L.Ed.2d 297 (1973);  accord Crane v. Kentucky, 476 U.S. 683, 690, 106 S.Ct. 2142, 90 L.Ed.2d 636 (1986) (“Whether rooted directly in the Due Process Clause of the Fourteenth Amendment, or in the Compulsory Process or Confrontation clauses of the Sixth Amendment, the Constitution guarantees criminal defendants ‘a meaningful opportunity to present a complete defense.’ ”(internal citations omitted)).   The district court correctly noted that the Supreme Court has established that, at least under some circumstances, due process may require that a criminal defendant be permitted to introduce extrinsic impeachment evidence, even despite contrary evidentiary rules, in order to protect a weighty or critical defense interest.   E.g., Chambers, 410 U.S. at 302, 93 S.Ct. 1038 (holding that trial court's exclusion of particular evidence on hearsay grounds deprived defendant of a fair trial “in accord with traditional and fundamental standards of due process”);  Crane, 476 U.S. at 690-91, 106 S.Ct. 2142 (1986) (holding that defendant's constitutional rights would be violated “if the State were permitted to exclude competent, reliable evidence bearing on the credibility of a confession when such evidence is central to the defendant's claim of innocence”).   The Supreme Court delineates the doctrine as follows:

A defendant's right to present relevant evidence is not unlimited, but rather is subject to reasonable restrictions.   A defendant's interest in presenting such evidence may thus “bow to accommodate other legitimate interests in the criminal trial process.”   As a result, state and federal rulemakers have broad latitude under the Constitution to establish rules excluding evidence from criminal trials.   Such rules do not abridge an accused's right to present a defense so long as they are not “arbitrary” or “disproportionate to the purposes they are designed to serve.”   Moreover, we have found the exclusion of evidence to be unconstitutionally arbitrary or disproportionate only where it has infringed upon a weighty interest of the accused.

United States v. Scheffer, 523 U.S. 303, 308, 118 S.Ct. 1261, 140 L.Ed.2d 413 (1998) (internal citations omitted).   Accord Montana v. Egelhoff, 518 U.S. 37, 43, 116 S.Ct. 2013, 135 L.Ed.2d 361 (1996) (state evidentiary rule “not subject to proscription under Due Process Clause unless ‘it offends some principle of justice so rooted in the traditions and conscience of our people as to be ranked as fundamental’ ”).   Whether a principle of justice is “fundamental,” according to the Supreme Court, is to be assessed historically, with reference to the common law.  Egelhoff, 518 U.S. at 43-44, 116 S.Ct. 2013.

 The question here, then, is whether the trial court's decision to exclude the Klare testimony under Rules 404(b) and 608(b) was either contrary to, or an unreasonable application of, the rule established in Scheffer and Egelhoff:  namely, that states may formulate their own evidentiary rules so long as those rules do not infringe on a weighty interest of the accused or offend a fundamental common-law principle of justice.   The district court concluded that the trial court's exclusion of the Klare testimony did not rise to this standard.   First, Ellsworth did not establish that there was a sufficiently weighty defense interest in the testimony to outweigh the evidentiary rules precluding its introduction.   Second, in any case, the testimony appeared to be proffered only as a “general” attack on Matthew's credibility rather than one targeted at exposing potential bias or motivation to lie.   This distinction is significant because the Supreme Court and the lower courts have held that the latter is a right entitled to the full protection of the Confrontation Clause whereas the former is not.   E.g., Van Arsdall, 475 U.S. at 678-80, 106 S.Ct. 1431;  Davis, 415 U.S. at 316-17, 94 S.Ct. 1105;  see also Boggs v. Collins, 226 F.3d 728, 736-38 (6th Cir.2000);  Hughes v. Raines, 641 F.2d 790, 793 (9th Cir.1981).

We find that the district court's application of Scheffer and Egelhoff to bar Klare's testimony, viewed in the context of the trial, was objectively unreasonable.   To conclude, as the district court did, that Klare's proffered testimony did not bear on Matthew's potential bias or motive for lying is simply incorrect.   There can be little question both that the testimony at issue bore on witness bias and motivation, and that Ellsworth's interest in introducing that testimony is quite substantial indeed.

At trial, defense counsel was permitted only to ask Matthew himself whether he had accused other children at Pine Haven of peeking at him or stealing his toys-all of which Matthew flatly denied.   Thus, the jury was left only with Matthew's unrebutted denials, and no evidence at all that the incidents in fact occurred-in some ways, an outcome far more prejudicial to Ellsworth than if counsel had not been allowed the cross-examination at all.

Accordingly, it is hard to conceive how Ellsworth's interest in impeaching Matthew's testimony could have been more weighty.   The jury was left to draw the conclusion that there was no rebuttal to Matthew's denials.   As to the issue of impeachment to show adverse motive or bias versus “general impeachment” of a witness's credibility, Klare's proffered testimony would have done much more than simply demonstrate Matthew's alleged “propensity to make false allegations about voyeurism and theft.”   The nature of the cross-examination at issue here is far more complex and subtle than a mere exploration of “general credibility”-it bears not only on any conscious motive that Matthew might have had for lying, but also on his unconscious, emotional motivation.   There was far more to be taken away from Klare's testimony than simple propensity evidence or Matthew's general untruthfulness;  it would have demonstrated to the jury the kinds of circumstances that unsettled Matthew enough to evoke untrue accusations-that constituted his motivation and his own set of biases.   We therefore REVERSE the district court on this claim as well.

III. CONCLUSION

By this decision, we do not minimize or disregard the genuine suffering that Matthew has gone through in his short life.   Nor do we conclude that he was not Ellsworth's victim.   However, the mere fact that a defendant stands accused by a sympathetic victim of a repulsive crime does not justify depriving him of his constitutional right to a fair trial.   The evidence against the defendant was not so strong that any errors made to his detriment were perforce “harmless” or insignificant.   Rather, because of the trial judge's limitations on Ellsworth's right to present a defense, he stands convicted and sentenced to many years in prison based entirely on the largely unrebutted testimony of an unstable, albeit sympathetic, complainant.

Accordingly, we hereby reverse the district court on all three claims, and remand with instructions to issue the writ of habeas corpus unless the State within sixty (60) days indicates that it will retry petitioner.

Raymond Ellsworth argues that the New Hampshire trial court violated his Sixth and Fourteenth Amendment rights by 1) prohibiting him from introducing evidence that Matthew was sexually victimized approximately eight years prior to his alleged abuse by Ellsworth, 2) refusing to disclose to the defense certain impeachment materials that the court reviewed in camera, and 3) excluding testimony that after Matthew was transferred to another school he made groundless complaints that students were stealing his toys and “peeking at him” while he was in the shower and on the toilet.

Emphasizing that Matthew is a “very disturbed young boy,” the majority attaches great weight to the absence of corroboration for Matthew's testimony about the criminal acts themselves, and ultimately agrees with all of Ellsworth's contentions.   Although I respect the majority's careful attention to the difficult issues before us, I cannot agree with any of its conclusions.   Disturbed young victims and uncorroborated victim testimony about acts of sexual abuse are commonplace in sexual abuse trials.   For the most part, the majority has transformed unremarkable evidentiary rulings into constitutional violations of such severity that, in the majority's view, Ellsworth's state court convictions must be vacated.   Because I do not believe that these transformations are warranted, I respectfully dissent.

I. DISCUSSION

A. Evidence of Matthew's Prior Victimization1. The Claimed Violation

At trial, the defense sought to introduce evidence that Matthew was subjected to fellatio and digital-anal intercourse when he was approximately three years old.   Matthew's prior victimization was relevant, according to the defense, because it provided him with the requisite sexual knowledge to fabricate the accusations he made against Ellsworth.   The trial court disagreed, finding that the nature of this prior abuse differed so substantially from Matthew's accusations against Ellsworth that the relevance of the prior abuse was minimal.

In appealing his convictions to the New Hampshire Supreme Court, Ellsworth argued that the judge's refusal to permit defense counsel to cross-examine Matthew about his prior abuse violated Ellsworth's Sixth Amendment right to confront the prosecuting witness.   Although the New Hampshire Supreme Court upheld Ellsworth's convictions, it did not address the federal elements of his Sixth Amendment claim.   Accordingly, our review of this question is de novo.   Fortini v. Murphy, 257 F.3d 39, 47 (1st Cir.2001).   Applying this non-deferential standard, the majority concludes that Ellsworth had a Sixth Amendment right to explore Matthew's prior abuse in his cross-examination of Matthew, asserting that the trial court's ruling to the contrary “ignored the common element of fellatio in both Matthew's earlier experiences and the charged conduct in this case.”   While the New Hampshire courts did overlook a common thread between Matthew's prior abuse and his accusations against Ellsworth, this error does not rise to constitutional dimensions.

The majority proposes two theories of relevance that underscore the prejudice resulting from the exclusion of Matthew's prior abuse:  1) evidence of Matthew's prior victimization reveals an alternative source of the sexual knowledge he exhibited on the witness stand, and 2) the jury was entitled to assess whether Matthew's prior abuse had inflicted “mental and emotional pain” that bore on his general credibility as a witness.

a. Alternative source of sexual knowledge

This theory of relevance rests on the assumption that jurors are inclined to perceive children such as Matthew as sexually innocent.   Hence, the theory goes, evidence of the child's prior sexual experiences is relevant to rebut that perception.   If the defense is precluded from introducing this evidence, jurors might infer that the child victim could describe the sexual abuse alleged in the particular case only because the defendant had actually abused the child.

The proposition that jurors ascribe this sexual naïveté to child witnesses such as Matthew as a matter of course is nothing more than a hunch.   Arguably, in an age of omnipresent movies, television videos, magazines, and Internet sites with pervasive sexual content, there is no basis for assuming that jurors actually hold such views about the sexual innocence of children.   Instead, the relevance of the child's prior sexual experiences should depend on the extent to which the prosecution's trial tactics or the circumstances of the particular case put the child's innocence at issue.

Here, the record does not reveal a single instance in which the prosecution argued or even hinted that Matthew's sexual knowledge could only arise from the alleged abuse.   For this reason alone state appellate courts have affirmed rulings to exclude evidence of prior victimization when the prosecution declines to open the door by arguing the child's naïveté.   As the Georgia Court of Appeals remarked in McGarity v. State, 224 Ga.App. 302, 480 S.E.2d 319 (1997):

Although an exception to the exclusion of evidence regarding a child's sexual past exists ․ the State [n]ever argue[d] at trial that the child's testimony must be credible because she could not have learned about sexual matters other than through molestation by the defendant.   It was not error for the trial court to exclude the proffered testimony.

Id. at 321-22, 480 S.E.2d 319.

Indeed, the prosecution could not have credibly argued that Matthew was sexually naive before the alleged acts of abuse committed by Ellsworth.   On cross-examination, Matthew testified that as a resident of Spaulding Youth Center, he participated in group discussions featuring such topics as “good and bad touching.”   Brian Blake, Matthew's counselor at Spaulding, reaffirmed on direct examination that the program Matthew participated in “talks about good kinds of touch and bad kinds of touch ․ basically the kids are taught that anything that a bathing suit covers you shouldn't be touching on someone else.”   Finally, Ellsworth himself testified that “Matthew was drawing his past abuse through his art and he was rewarded for that and he felt good” (emphasis added).

Nor was the jury under the misimpression that Matthew was of such a “tender age” as to lack knowledge of oral sex apart from the alleged acts of abuse.   During the defendant's direct examination of Matthew's roommate, a boy of approximately Matthew's age, the jury heard the following exchange:

DEFENSE:  Matt said to you that Ray Ellsworth made him do some obscene things?

WITNESS:  Yes.

DEFENSE:  That was your testimony a minute ago.   What exactly did Matt say to you?

WITNESS:  He just said, Ray made me give him a blow job.

DEFENSE:  And you knew what that meant?

WITNESS:  Yes.

 *   *   *   *   *   *

DEFENSE:  Did you ask questions?

WITNESS:  No.

DEFENSE:  He just talked to you?

WITNESS:  Yeah.

DEFENSE:  So you didn't clarify what he meant when he said, Ray made me give him a blow job?

WITNESS:  Right.

DEFENSE:  You didn't ask him what he meant by that?

WITNESS:  No, but I'm pretty sure I knew.

DEFENSE:  You assumed that what you understood that obscene expression to mean is the same thing that Matthew assumed it to mean?

WITNESS:  Right.

DEFENSE:  You didn't have to ask him;  you knew what it meant?

WITNESS:  Yes.

Even if the Ellsworth jury was prepared to attribute Matthew's sexual knowledge to Ellsworth's acts of abuse on the assumption of Matthew's sexual innocence, the defense would have gleaned little benefit from introducing the details of his prior victimization.   Although the majority accurately observes that fellatio was a common element of Matthew's prior abuse and his current allegations, Matthew also testified to acts of penile-anal intercourse, a type of sexual behavior that was not a component of Matthew's prior abuse and one that would arguably lie further outside an eleven-year old child's range of knowledge than fellatio.   Evidence of prior victimization that incompletely accounts for the acts described by the prosecuting witness will not serve the purpose urged by the defense for its admission-to suggest a source of knowledge other than the defendant's sexual abuse.   The significance of this discrepancy between the two incidents of anal intercourse was not lost on the trial court.   In its written order denying defendant's motion in limine to admit evidence of Matthew's prior victimization, the court found that “the prior sexual abuse is irrelevant to this case ․ the two cases differ as to the type of anal intercourse.”

It is well settled that trial courts have broad discretion to exclude evidence of marginal relevance.  Delaware v. Van Arsdall, 475 U.S. 673, 679, 106 S.Ct. 1431, 89 L.Ed.2d 674 (1986) (“[T]rial judges retain wide latitude insofar as the Confrontation Clause is concerned to impose reasonable limits on such cross-examination based on concerns about, among other things ․ interrogation that is ․ only marginally relevant.”).   Here, the state trial court properly exercised that discretion.   The majority's ruling to the contrary presumes that in the minds of jurors sexual innocence automatically attaches to every child witness alleging sexual abuse.   That untested assumption has the effect of routinely denying the protection of state rape shield laws to child victims of sexual abuse.   The majority cites no unique fact or circumstance in the case at bar that distinguishes this case from other sexual abuse cases with child victims.   Accordingly, I do not believe that the Constitution compelled the introduction of evidence of Matthew's prior victimization.

b. “Mental and emotional damage” and the appraisal of credibility

The majority suggests in its harmless error analysis another rationale for constitutionally requiring the admission of evidence of Matthew's prior victimization:  “knowing of Matthew's early abuse would have given the jury some insight into the mental and emotional damage he sustained as a result of that abuse.”   Therefore, “this information was critical to the jury's appraisal of Matthew's credibility as a witness.”

While some courts have determined that the “alternative source of sexual knowledge” rationale triggers Sixth Amendment protections under certain circumstances, see, e.g., Lajoie v. Thompson, 217 F.3d 663, 668-73 (9th Cir.2000);  Shaw v. United States, 24 F.3d 1040, 1043 (8th Cir.1994), no court, to my knowledge, has invoked the Confrontation Clause to secure the admission of evidence that generally demonstrates a child witness's “mental and emotional damage.”   Indeed, even Ellsworth does not advance this novel theory of relevance on appeal.   The unfortunate reality is that the specter of mental and emotional trauma attends every rape, child abuse, and sexual assault prosecution where the complaining witness has suffered prior abuse.   Nonetheless, the Supreme Court has endorsed state rape shield laws designed to exclude evidence of prior victimization in the vast majority of cases, acknowledging “the valid legislative determination that rape victims deserve heightened protection against surprise, harassment, and unnecessary invasions of privacy.”  Michigan v. Lucas, 500 U.S. 145, 150, 111 S.Ct. 1743, 114 L.Ed.2d 205 (1991).   Moreover, “[t]he majority of states view prior rape or sexual abuse of a child as ‘sexual conduct’ within the ambit of state rape shield laws.”  Grant v. Demskie, 75 F.Supp.2d 201, 211 (S.D.N.Y.1999) (citing cases).

Applying the majority's logic, it is difficult to conceive of a case where the defense could not argue that the Confrontation Clause mandates the admissibility of prior victimization evidence to facilitate “the jury's appraisal of [an individual's] credibility as a witness.”   Hence, the rule espoused by the majority eviscerates an important function of state rape shield statutes by routinely interposing the Sixth Amendment into areas previously regarded as the sole province of state evidentiary rules.   Yet the Supreme Court's Confrontation Clause jurisprudence discourages these broad-based challenges to state evidentiary rules under the guise of constitutional error by limiting the Sixth Amendment's reach:  “[T]he Confrontation Clause guarantees an opportunity for effective cross-examination, not cross-examination that is effective in whatever way, and to whatever extent, the defense might wish.”  Delaware v. Fensterer, 474 U.S. 15, 20, 106 S.Ct. 292, 88 L.Ed.2d 15 (1985) (per curiam) (emphasis in original).   Expounding on this principle, the Eleventh Circuit has observed that

The [Confrontation] [C]lause emphatically does not confer upon criminal defendants a right to present any and all relevant substantive evidence․   The Confrontation Clause plays a discrete (albeit essential) role in criminal trials, but it simply is not coextensive with relevance.

Jones v. Goodwin, 982 F.2d 464, 469 (11th Cir.1993).

The generic applicability of the “mental and emotional damage” theory of relevance suggests its non-constitutional dimensions.   Moreover, the trial court in this case permitted defense counsel to develop numerous avenues of impeachment during Matthew's interrogation which cumulatively afforded Ellsworth an effective cross-examination:

- Chief Leary, Trooper Nolan, and Matthew's guardian ad litem inappropriately encouraged Matthew to accuse Ellsworth of sexual assault.

- Matthew described the Colcord Cottage incident and the bicycling incident to his peers at Spaulding in a manner that was inconsistent with the story he told to the police.

- Matthew's recollection of the Sandoggerdy Pond incident during his interview with Chief Leary conflicted with his testimony on direct examination.

- Matthew resented Ellsworth, who on two occasions placed him in a solitary confinement room, and who generally had the authority to deny Matthew important privileges by assigning him “negative points.”

- Matthew participated in discussions about good touching and bad touching at Spaulding.

- After leaving Spaulding, Matthew made false accusations of sexual voyeurism and theft against his peers at Pine Haven.

Accordingly, I do not find that the majority's second theory of relevance reveals constitutional error in the trial court's exclusion of evidence concerning Matthew's prior victimization.

2. Harmless Error Analysis

Although I conclude that no Sixth Amendment violation occurred, the district court's harmless error analysis provides an alternative basis for denying the petitioner's habeas claim arising from the exclusion of the prior victimization evidence.   The Supreme Court noted in Brecht v. Abrahamson, 507 U.S. 619, 113 S.Ct. 1710, 123 L.Ed.2d 353 (1993), that “the historic meaning of habeas corpus [is] to afford relief to those whom society has grievously wronged.”   Id. at 637, 113 S.Ct. 1710 (internal quotation marks omitted).   Accordingly, error is considered harmless in a habeas case unless it had “substantial and injurious effect or influence in determining the jury's verdict.”  Id. (quoting Kotteakos v. United States, 328 U.S. 750, 776, 66 S.Ct. 1239, 90 L.Ed. 1557 (1946)).

The Supreme Court emphasized in Van Arsdall that courts engaging in harmless error analysis must consider a variety of factors:

Whether such an error is harmless in a particular case depends upon a host of factors, all readily accessible to reviewing courts.   These factors include the importance of the witness' testimony in the prosecution's case, whether the testimony was cumulative, the presence or absence of evidence corroborating or contradicting the testimony of the witness on material points, the extent of cross-examination otherwise permitted, and, of course, the overall strength of the prosecution's case.

Van Arsdall, 475 U.S. at 684, 106 S.Ct. 1431.   Of these five factors, the majority relies heavily on factors one and three in its harmless error analysis:  namely, the absence of corroboration for the testimony of the crucial witness for the prosecution.   However, this lack of corroboration, while not insignificant, is not unusual in sexual abuse cases.   Indeed, the Seventh Circuit has observed that “[d]etecting sexual abuse, and convicting its perpetrators, is problematic because of the lack of witnesses [and] the difficulty of obtaining corroborative physical evidence ․” Doe v. United States, 976 F.2d 1071, 1074 (7th Cir.1992) (citing Judy Yun, Note, A Comprehensive Approach to Child Hearsay Statements in Sex Abuse Cases, 83 Colum.  L.Rev. 1745, 1745 (1983)).   Even Ellsworth's attorney, in moving to dismiss the counts of the indictment that were not nol-prossed, conceded that

[i]t is true that the uncorroborated testimony of a victim of sexual assault is traditionally sufficient to go the jury․   The reason for that rule is the obvious one.   These offenses happen in secret and often there is no such thing as corroboration and but for that rule, the guilty would escape punishment.

Accordingly, the Van Arsdall test is structured to permit findings of harmless error even when the prosecuting witness's testimony is uncorroborated.   See Quinn v. Haynes, 234 F.3d 837, 850 n. 11 (4th Cir.2000) (“To treat the absence of corroboration as dispositive would undermine the Supreme Court's holding in Lucas.”).

Nevertheless, I believe that the majority understates the degree of corroboration for Matthew's testimony.   The State's brief describes at length the testimony offered by other witnesses to corroborate Matthew's detailed accounts of the circumstances surrounding the three incidents-the bike trip, the outing to Sandoggerdy Pond, and the incident at Colcord Cottage.   Although the majority dismisses this testimony because it does not corroborate the criminal acts themselves, the record also indicates that the only individual in the vicinity of Matthew and Ellsworth during an alleged assault provided strong circumstantial evidence that the jury could have viewed as corroborative of Matthew's testimony about the actual abuse.   Steven, a student at Spaulding who accompanied Matthew and the defendant to Sandoggerdy Pond, testified that he observed Matthew and the defendant “roughhousing and wrestling” in the pond, and that he saw Matthew go underwater to “sneak up on Ray from behind.”

Finally, the fifth Van Arsdall factor implicates the overall strength of the prosecution's case.⁷  The district court observed that “Matthew's descriptions of the three incidents include remarkably consistent graphic detail,” and that “[n]one of the record evidence contradicts the essential elements of Matthew's testimony.”   While these facts alone reflect the strength of the prosecution's case, the jury heard additional testimony that was probative of Ellsworth's culpability.   Edward DeForrest, the Executive Director of Spaulding, and Gary Lavallee, the Administrative Director, testified that Ellsworth acted contrary to Spaulding rules by taking students off campus for one-on-one bike trips and outings to Sandoggerdy Pond without having another staff member present.   The jury heard the following testimony from Lavallee:

STATE:  Ray Ellsworth, like every other cottage teacher, could have, say, for instance, undertaken a bicycle trip with one of his students in and around campus or around the loop?

WITNESS:  Around the loop area itself, a student?

STATE:  Well, several students, let's say.

WITNESS:  Yes.

STATE:  Could he have done it with one student?

WITNESS:  Could he have?

STATE:  Yes.

WITNESS:  He could have.

STATE:  Was he supposed to have?

WITNESS:  No.

STATE:  There were rules regarding the one-on-one relationship?

WITNESS:  Yes.

 *   *   *   *   *   *

STATE:  With respect to the policy of a residential teacher not being alone with a student, is that a written policy?

WITNESS:  Yes, it is.

Confronted with this testimony, Ellsworth acknowledged that he had taken one-on-one trips with students and admitted taking Matthew and Steven to Sandoggerdy Pond, but denied ever taking Matthew out for a one-on-one bike trip.   Nevertheless, this admission by Ellsworth that he had violated Spaulding's one-on-one policy added to the strength of the state's case and the force of the district court's harmless error analysis:

Based on the record as a whole, if the jury had been informed of Matthew's prior abuse, their more complete picture of Matthew would have been that he had some precocious sexual knowledge and experience.   Given the dissimilarities in the experiences, information about the prior abuse would not likely have undermined Matthew's graphic descriptions of the three separate incidents of abuse by Ellsworth.

Ellsworth v. Warden, Civil No. 99-132 at 37 (D.N.H. Jan. 23, 2002).   I agree.

B. Access to Exculpatory Materials

This is the most troubling issue presented in the defendant's habeas petition.   As the majority observes, our review of this question is de novo because the Supreme Court of New Hampshire did not address the federal elements of Ellsworth's due process claim.  Fortini, 257 F.3d at 47.   The defendant argues that the trial court violated his due process rights by withholding three documents containing admissible, exculpatory information.   The first document was an “intake note” generated by Jan Smith, a clinical director at Spaulding, after conducting an admission interview with Matthew prior to his enrollment.   The second document was an incident report indicating that Matthew was experiencing vivid dreams of his prior abuse.   The third document, according to Ellsworth, described the efforts of Spaulding staff members to reopen the legal case against the perpetrator of Matthew's prior abuse.   Of these three documents, the majority limits its discussion to the incident report concerning Matthew's dreams and Jan Smith's intake note.   The district court determined that both documents contained “new favorable arguments,” but concluded that they did not meet the materiality requirement of Brady.

1. The Critical Incident Report Describing Matthew's Dreams of Sexual Abuse

The majority expresses concern that the trial court did not turn over the incident report discussing Matthew's vivid dreams of his prior abuse:  “We find it extremely troubling that the jury was totally unaware, not only of Matthew's history, but of the extent to which that history was having a particular impact on him at the time when he made his accusations against Ellsworth.”   The district court, however, found that the incident report lacked materiality because it was duplicative of information that was known to Ellsworth.

The Second Circuit observed in United States v. Diaz, 922 F.2d 998 (2d Cir.1990), that “[e]vidence is not ‘suppressed’ [within the meaning of Brady ] if the defendant either knew, or should have known of the essential facts permitting him to take advantage of any exculpatory evidence.”  Id. at 1007 (internal quotation marks omitted).   Moreover, we have recognized that “[s]ince cumulative evidence is not material to either guilt or punishment, the unavailability of cumulative evidence does not deprive the defendant of due process.”  United States v. Sanchez, 917 F.2d 607, 618 (1st Cir.1990) (internal quotation marks omitted);  see also United States v. Rodriguez Alvarado, 985 F.2d 15, 19 (1st Cir.1993);  Diaz, 922 F.2d at 1007;  United States v. Weintraub, 871 F.2d 1257, 1264 (5th Cir.1989);  Smith v. Kelso, 863 F.2d 1564, 1573 (11th Cir.1989).

The record indicates that, as the residential teacher assigned to Matthew and only two other students, Ellsworth had untrammeled access to every document in Matthew's file during his work with Matthew at Spaulding.⁸  Edward DeForrest testified that “[s]taff is assigned to a particular child who would be intimately familiar with the entire case history of the child and is expected to be.   Further, they need that background in order to interpret information coming from the home or in fact how to report to the home.”   Gary Lavallee also asserted in an affidavit that Ellsworth not only had access to Matthew's file in theory, but that “there is evidence that he accessed Matthew's file.   Namely, Mr. Ellsworth generated reports relating to Matthew's residential care․   It is customary for cottage teachers to access resident files in order to collect this data.”   Lavallee added that “Mr. Ellsworth had access to all critical incident reports generated relating to Matthew․   As Matthew's cottage teacher, Mr. Ellsworth was encouraged to read all critical incident reports relating to Matthew.”   Thus, even if Ellsworth was unaware of particular details of Matthew's past abuse while Matthew was at Spaulding, he was sufficiently aware of its impact on Matthew's life to attempt to explore this impact as part of his defense.

The best evidence of this awareness is Ellsworth's own testimony at trial.   Contrary to the majority's assertion that the jury was “totally unaware ․ of Matthew's history,” the jury heard Ellsworth testify on direct examination that “Matthew was drawing his past abuse through his art and he would receive attention for that and he was rewarded for that and he felt good.”   Moreover, in light of Ellsworth's testimony that Matthew was reliving his prior abuse through his art, the withheld incident report would have been cumulative of other evidence adduced at trial about Matthew's attention to his prior abuse.   As we have previously observed, cumulative evidence lacks the requisite materiality to trigger a Brady violation.  Sanchez, 917 F.2d at 618.

2. Jan Smith's Intake Note

The trial court's decision to withhold Jan Smith's intake note poses a more complicated issue.   Specifically, the intake note indicates that 1) Matthew alleged that he was abused by staff members at Hampstead Hospital, where he resided prior to his arrival at Spaulding, 2) Matthew expressed concern that he would be sexually assaulted at Spaulding, and 3) Spaulding would need to take special precautions to protect staff members from false accusations if Matthew enrolled.⁹

As a threshold matter, evidence that is not admissible at trial is ipso facto immaterial under Brady:  “Inadmissible evidence is by definition not material, because it never would have reached the jury and therefore could not have affected the trial outcome.”  United States v. Ranney, 719 F.2d 1183, 1190 (1st Cir.1983);  see also United States v. Kennedy, 890 F.2d 1056, 1059 (9th Cir.1989) (“To be material under Brady, undisclosed information or evidence acquired through that information must be admissible.”).   Neither Ellsworth nor the majority addresses the admissibility of the intake note.   Instead, they assume its admissibility and focus their discussion exclusively on the note's materiality.   The defense argues that “[t]he withheld evidence demonstrates that Matthew was preoccupied with being a victim of child molestation.   He made false allegations against staff at another placement.   He feared abuse at Spaulding.”   Echoing these views, the majority remarks that “[t]he information that the Spaulding director recognized, and was concerned, that Matthew had a tendency to accuse people of sexual abuse does seriously undermine one's confidence in the outcome of the trial.” (emphasis in original).

The implicit assumption in this observation that the substance of the intake note could have been introduced in its entirety through Jan Smith's testimony is not well grounded in the New Hampshire Rules of Evidence.   At most, pursuant to the state of mind exception to the hearsay rule set forth in Rule 803(3) of the New Hampshire Rules of Evidence, the trial court would have permitted Jan Smith to testify that Matthew had expressed concern for his safety and a fear of being sexually abused at a residential treatment center.¹⁰  The portion of the note referring to Matthew's allegations of sexual abuse by the staff at Hampstead Hospital could only have been used for cross-examination of Matthew under Rule 608(b) of the New Hampshire Rules of Evidence.¹¹  This rule permits inquiry on cross-examination into specific instances of the witness's conduct that are “probative of truthfulness or untruthfulness ․” N.H.R. Evid. 608(b).

Although Jan Smith does not refer to “false” allegations of sexual abuse towards staff at Hampstead Hospital in her intake note, the defense, the majority, and the district court all read her note to imply such falsity.   I think that is a fair interpretation of the note.   However, the claim that Matthew made false allegations of sexual abuse at Hampstead Hospital could only be used in cross-examining Matthew if the prior allegations were demonstrably untrue.   In State v. White, 145 N.H. 544, 765 A.2d 156 (2000), the Supreme Court of New Hampshire observed that

a defendant may introduce a victim's prior allegations of sexual assault by showing that the prior allegations were demonstrably false, which we interpret to mean “clearly and convincingly untrue.”   This approach requires greater proof of falsity than the “reasonable probability” standard proposed by the defendant, but less certitude than the “false-in-fact” test offered by the State.

Id. at 159 (emphasis added);  see also State v. Gordon, 146 N.H. 258, 770 A.2d 702, 704-05 (2001) (“[A] defendant in a sexual assault case may cross-examine the victim about a prior false allegation of sexual assault under Rule 608(b) only if the defendant ․ demonstrates clearly and convincingly that the prior allegations were false.”).   The district court concluded that “[t]aken in the context of the intake session and considering the participants, it appears that Matthew's prior allegations of abuse by Hampstead Hospital staff were false.   At the very least, it appears that Smith thought the allegations were false.”   The district court thus concluded that the defense was entitled to cross-examine Matthew about his false allegations of abuse against staff members at Hampstead Hospital.

I cannot agree with that conclusion for two reasons.   First, Jan Smith's intake note reveals nothing about the basis for her conclusion that Matthew actually accused Hampstead Hospital staff of abusing him.   Second, assuming that Matthew made such accusations, the hunch of falsity by a Spaulding administrator with no connection to Hampstead Hospital or its staff members and no exposure to the specific nature or circumstances of the allegations at issue falls substantially short of the “clear and convincing” proof of falsity test articulated in White and Gordon.   Consequently, Jan Smith's intake note does not provide a basis for cross-examining Matthew about the Hampstead Hospital allegations, and it fails the materiality standard under Brady.¹²

Anticipating Ellsworth's response that he could not demonstrate the fact of the Hampstead Hospital allegations and their falsity if he was not aware of these allegations in the first place, I find any such claim untenable for the same reason that the defendant could not claim ignorance of Matthew's efforts to deal with his past abuse.   Gary Lavallee asserted in his affidavit that

[a]s Cottage teacher responsible for Matthew's care, Mr. Ellsworth had access to Matthew's main file at SYC [Spaulding Youth Center].   In fact, cottage teachers are highly encouraged to familiarize themselves with patient files in order to provide thorough patient care ․ Matthew's file contains information about his past placements, including his treatment at Hampstead Hospital, references to Matthew's prior sexual abuse, and intake notes assessing Matthew's history and determining whether SYC was in a position to meet Matthew's needs.

(emphasis added).   Moreover, as the district court noted, Ellsworth was aware of the concerns of Matthews parents about previous false allegations of abuse.

Finally, any claim by the defendant that he was unaware of Jan Smith's intake note at the time of his trial is undermined by an affidavit he filed in the district court during the habeas proceeding.   On November 28, 2001, the district court issued an order permitting defense counsel to disclose to Ellsworth the substance of certain documents reviewed in camera, including the intake note.   One week later, on December 7, Ellsworth and Gary Lavallee filed opposing affidavits contesting the breadth of Ellsworth's knowledge with respect to the in camera documents.   Conspicuously absent from Ellsworth's affidavit is any claim that he was unaware of the intake note.   Ellsworth does, however, refer to a litany of other facts that he was unaware of, including 1) efforts by the Spaulding staff to assist Matthew in reopening the case against the perpetrator of his prior abuse, 2) Matthew's dreams of his prior sexual abuse, 3) efforts to prepare Matthew to testify against the perpetrator of the prior abuse, and 4) the fact that this preparation included reviewing the prior abuse.

In sum, after reviewing the record, I find that Ellsworth was sufficiently informed to try to establish before the trial court a basis for asking Matthew on cross-examination if he had falsely accused others of sexual abuse.   For this reason as well, his claim that the non-disclosure of the intake note violated Brady fails to meet the materiality requirement.

C. Extrinsic Impeachment Evidence

Evidence is extrinsic if it is proffered in support of a collateral matter, or a “matter [that] itself is not relevant in the litigation to establish a fact of consequence, i.e., not relevant for a purpose other than mere contradiction of the in court testimony of the witness.”  United States v. Marino, 277 F.3d 11, 24 (1st Cir.2002) (internal citations omitted).   Here, Craig Klare, a counselor at the Pine Haven School that Matthew attended after leaving Spaulding, was prepared to testify to the collateral fact that Matthew falsely accused other boys at the school of stealing his toys and peeking at him while he was in the shower and on the toilet.   The majority acknowledges that the trial court permitted Ellsworth to cross-examine Matthew about these incidents, but nonetheless concludes that the judge's exclusion of Clare's testimony despite Matthew's denials violated Ellsworth's rights under the Confrontation Clause.

As the majority observes, the New Hampshire Supreme Court did address the federal elements of Ellsworth's Sixth Amendment claim that the exclusion of this extrinsic evidence violated his rights under the Confrontation Clause.   Accordingly, our review of this issue is governed by the familiar AEDPA standard, which requires the petitioner to demonstrate that the state court decision was either contrary to or an unreasonable application of clearly established federal law as established by the Supreme Court.  28 U.S.C. § 2254(d)(1).   Elaborating on these standards, the Supreme Court stated in Williams v. Taylor that

under the “contrary to” clause, a federal habeas court may grant the writ 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 has 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 legal principle from the [Supreme] Court's decisions but unreasonably applies that principle to the facts of the prisoner's case.

Williams v. Taylor, 529 U.S. 362, 412-13, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000).

Commenting on the interplay between the Constitution and state evidentiary rules, the Supreme Court observed in United States v. Scheffer, 523 U.S. 303, 118 S.Ct. 1261, 140 L.Ed.2d 413 (1998), that

[S]tate and federal rulemakers have broad latitude under the Constitution to establish rules excluding evidence from criminal trials.   Such rules do not abridge an accused's right to present a defense so long as they are not “arbitrary” or “disproportionate to the purposes they are designed to serve.”   Moreover, we have found the exclusion of evidence to be unconstitutionally arbitrary or disproportionate only where it has infringed upon a weighty interest of the accused.

Id. at 308, 118 S.Ct. 1261 (internal citations omitted).   While determining whether the application of a state evidentiary rule has infringed a “weighty interest of the accused” is difficult in the abstract, federal courts have developed rules that govern a defendant's right to introduce extrinsic evidence under the Confrontation Clause.   Simply put, the Confrontation Clause does not secure a defendant's right to introduce extrinsic evidence for the purpose of attacking the general credibility of the witness, but does protect a defendant's right to use extrinsic evidence to reveal bias or motive:

[A] criminal defendant states a violation of the Confrontation Clause by showing that he was prohibited from engaging in otherwise appropriate cross-examination designed to show a prototypical form of bias on the part of the witness, and thereby to expose to the jury the facts from which jurors ․ could appropriately draw inferences relating to the reliability of the witness.

Van Arsdall, 475 U.S. at 680, 106 S.Ct. 1431 (internal citations omitted) (emphasis added);  see also Quinn v. Haynes, 234 F.3d 837, 845 (4th Cir.2000) (quoting Van Arsdall ).¹³  The phrase “prototypical form of bias” draws meaning from the Court's decision in Davis v. Alaska, 415 U.S. 308, 94 S.Ct. 1105, 39 L.Ed.2d 347 (1986), which distinguished general credibility attacks from those attacks warranting Sixth Amendment protection:  “A more particular attack on the witness' credibility is effected by means of cross-examination directed toward revealing possible biases, prejudices, or ulterior motives of the witness as they may relate directly to issues or personalities in the case at hand.”  Id. at 316, 94 S.Ct. 1105 (emphasis added).

Significantly, the threshold for admitting extrinsic evidence under Van Arsdall and Davis does not vary with the witness's importance in the case.   As the Sixth Circuit observed in Boggs v. Collins, 226 F.3d 728 (6th Cir.2000):

No matter how central an accuser's credibility is to a case-indeed, her credibility will almost always be the cornerstone of a rape or sexual assault case, even if there is physical evidence-the Constitution does not require that a defendant be given the opportunity to wage a general attack on credibility by pointing to individual instances of past conduct.   In other words, [a defendant's] argument that credibility is crucial to this case, and that therefore any evidence bearing on that credibility must be allowed in, simply does not reflect Sixth Amendment caselaw.

Id. at 740.

Applying AEDPA's deferential standard of review, the trial court's exclusion of Klare's testimony was not contrary to or an unreasonable application of this federal law for two reasons:  1) the New Hampshire Supreme Court did not unreasonably determine that Klare's testimony was proffered as a general attack upon Matthew's credibility, and 2) the probative value of Klare's testimony is minimal because no cogent theory of relevance establishes a nexus between Matthew's accusations of mischief against his peers at Pine Haven and his allegations of sexual assault against Ellsworth.

1. General Versus Motive-Based Credibility Attacks

As the Supreme Court observed in Davis, extrinsic evidence reveals a “bias, prejudice or ulterior motive” only if it “relate[s] directly to issues or personalities in the case at hand.”  Davis, 415 U.S. at 316, 94 S.Ct. 1105 (emphasis added).   For example, a child might be motivated to fabricate sexual abuse allegations against her stepfather out of loyalty to her natural parent and/or resentment towards her natural mother's new spouse.   See, e.g., State v. Day, 538 A.2d 1166 (Me.1988).   If the defense had evidence of prior instances in which the child had exhibited this bias against the stepfather, it could inquire about these instances on cross-examination.   If the child disavowed the incidents, the defense could argue with some force that the constitution compels the admissibility of extrinsic evidence of the prior conduct notwithstanding a contrary evidentiary rule.   Here, if the defense had evidence of prior instances in which Matthew's behavior revealed a bias against Ellsworth, it could have pursued a similar course.   Such evidence would relate directly to “issues or personalities in the case at hand.”  Davis, 415 U.S. at 316, 94 S.Ct. 1105.   However, Craig Klare's proffered testimony, which does not involve Ellsworth at all, falls outside this category.   Nevertheless, the majority argues that Klare's testimony is the foundation of a motive-based credibility attack because “it bears not only on any unconscious motive that Matthew might have had for lying, but also on his unconscious, emotional motivation ․ it would have demonstrated to the jury the kinds of circumstances that unsettled Matthew enough to evoke untrue accusations.”

This formulation of Matthew's “motive” is simply another way of describing a general attack on credibility.   If “unconscious, emotional motivation” could form the basis of a specific credibility attack protected by the Sixth Amendment, this exception would swallow the rule proscribing the use of extrinsic evidence to support general credibility attacks.   Any general credibility attack can be recast as an inquiry into a witness's unconscious emotional motives or biases.   Just as the majority's “mental and emotional pain” rationale in the prior victimization context facilitates the Sixth Amendment preemption of state rape shield laws, its recognition of an “unconscious, emotional motivation” justification for the admission of Klare's testimony undermines state evidentiary rules limiting the admission of extrinsic evidence of other conduct.

As the offer of proof of Klare's testimony reveals, the practical import of this erosion is enormous.   The state cross-examined Klare at length about his limited exposure to Matthew, the fact that his testimony violated a confidential counselor-patient relationship, his breach of the agreement he signed at Pine Haven not to divulge confidential information about students, his sub-par job performance, and the nature of his relationship with the defendant.   This cross-examination was followed by the defense's efforts to rehabilitate Klare on re-direct and the state's rejoinder on re-cross.   If this extensive examination had been conducted in the presence of the jury, there would have been the distracting “trial within a trial” that states have attempted to eliminate through evidentiary rules that substantially limit the admissibility of extrinsic evidence.

2. The Probative Value Of Craig Klare's Testimony

As previously noted, Matthew's allegations at Pine Haven and his allegations of sexual abuse against Ellsworth are strikingly different.   First and most obviously, the nature and severity of the alleged acts are incomparable.   At Spaulding, Matthew accused the defendant of rape through such acts as fellatio and penile-anal intercourse.   At Pine Haven, Matthew complained that students were stealing his toys and peeking at him while he was in the bathroom and the shower.   Second, Matthew exhibited completely dissimilar responses to the Spaulding and Pine Haven incidents.   Brian Blake, Matthew's counselor at Spaulding, testified that Matthew was so reluctant to discuss the abuse that Blake had to use hand puppets to elicit the details of the alleged assaults from Matthew.   Yet Klare proffered testimony that Matthew immediately and publicly expressed his displeasure with the “peeking” as it was allegedly occurring.   Third, Matthew specifically implicated Ellsworth in the Spaulding incidents, while his allegedly false accusations of “voyeurism” at Pine Haven targeted no one in particular.   In fact, Matthew's only specific accusation at Pine Haven was leveled against a roommate and a boy living across the hall for the distinctly non-sexual offense of stealing his toys.   Finally, Klare's proffered testimony did not suggest that Matthew made any effort to follow up on his complaints with counselors or other staff members at Pine Haven.   At Spaulding, however, Matthew cooperated with one local chief of police, two state troopers, a Spaulding counselor and a guardian ad litem in recounting the details of the assault in audiotaped and videotaped conversations and retracing his steps through the scenes of the alleged crimes.

Not surprisingly, the New Hampshire Supreme Court observed that “[t]he subsequent alleged fabrications of sexual voyeurism and theft at Pine Haven and the charged crimes of fellatio and inappropriate touching completely lack similarity and evidentiary nexus.”  State v. Ellsworth, 142 N.H. 710, 709 A.2d 768, 774 (1998).   The district court agreed, observing that “[t]he allegedly false allegations at Pine Haven are quite different from his allegations against Ellsworth, even if the accusations of voyeurism were construed as sexual.”  Ellsworth v. Warden, Civil No. 99-132 at 15 (D.N.H. Feb. 23, 2001).   There is no basis for disturbing these findings under AEDPA's deferential standard of review.

II. CONCLUSION

Sexual abuse cases exact a tremendous emotional toll on victims, defendants, judges, counsel and witnesses.   When the victim is a child, there is the added burden of difficult evidentiary issues implicating the balance between the defendant's constitutional right to effectively confront the state's most damaging witness and the vitality of state evidentiary rules which set necessary limits on the cross-examination of the child.   The majority concludes that the New Hampshire trial court struck this balance so improperly that the verdicts at issue must be vacated.

I disagree.   In my view, the trial court, the Supreme Court of New Hampshire, and the federal district court demonstrated considerable skill in handling the difficult issues presented by this case.   After thoroughly examining the record, and applying AEDPA's deferential standard of review where appropriate, I find no basis for disturbing the verdicts.   Accordingly, I respectfully dissent from the majority's conclusion that the habeas petition should be granted.

FOOTNOTES

1.   A habeas petitioner may also, of course, prevail by demonstrating that the state court decision was “based on an unreasonable determination of the facts in light of the evidence presented in the State court proceeding,” § 2254(d)(2).   However, this provision is not at issue in this case.

2.   The district court adjudicated this claim under the AEDPA standard before the First Circuit issued Fortini.

3.   Our holding that the trial judge erred in excluding this evidence is in no way incompatible with either the rape shield laws or with Michigan v. Lucas, 500 U.S. 145, 111 S.Ct. 1743, 114 L.Ed.2d 205 (1991), as the dissent suggests.

4.   The dissent suggests that, because “disturbed young victims and uncorroborated victim testimony about acts of sexual abuse are commonplace in sexual abuse trials,” we should somehow be unconcerned about the complete lack of corroboration of Matthew's allegations in this case.   Yet this is precisely why we are concerned.   Of course, we do not mean to suggest that rape or sexual abuse victims should be required to adduce direct, eyewitness testimony recounting their abuse-we recognize fully that such testimony is usually nonexistent.   However, corroboration can be had in many other ways, such as physical evidence taken from the victim or his surroundings or the victim's behavior following the alleged incident.   Here, there is no such evidence;  the sole proffered corroboration is Matthew's own description of entirely collateral details.   The only evidence is the testimony of a seriously emotionally disturbed child.

5.   Review of this claim is de novo.  Fortini, 257 F.3d at 47.   Moreover, the Supreme Court has held that this type of claim (Bagley error) is not amenable to harmless error review under Brecht.  Kyles v. Whitley, 514 U.S. 419, 435-36, 115 S.Ct. 1555, 131 L.Ed.2d 490 (1995).

FN6. The text of these rules is identical to that of the corresponding Federal Rules of Evidence..  FN6. The text of these rules is identical to that of the corresponding Federal Rules of Evidence.

7.   The fourth factor in the Van Arsdall analysis considers the “extent of the cross-examination otherwise permitted.”   As noted above, the judge accorded defense counsel considerable latitude in pursuing at least six separate lines of impeachment.

8.   Gary Lavallee's affidavit states that Ellsworth was employed at Spaulding from October 1988 to September 1992.   It further reflects that Matthew was admitted to Spaulding on November 21, 1991, and discharged on November 11, 1993.   Ellsworth was assigned to Matthew's treatment team for the duration of the period that they overlapped-from November 21, 1991 to September 1992.

9.   The relevant portion of Jan Smith's intake note reads as follows:Note:  Matthew has expressed verbally his concerns regarding safety issues and the fact that he may be sexually abused at a residential treatment center.   It will be important for staff to be prepared for his concerns as well as the fact that he has made allegations towards staff at Hampstead Hospital of abusing him.   Special precautions will need to be taken regarding being alone with him and the possibility of him accusing staff of maltreating him.

10.   New Hampshire Rule of Evidence 803(3) creates an exception to the hearsay rule for “a statement of the declarant's then existing state of mind, emotion, sensation, or physical condition ․” N.H.R. Evid. 803(3).

11.   I recognize that the intake note also mentions the need for special precautions “regarding being alone with [Matthew] and the possibility of him accusing staff of maltreating him.”   The need for special precautions is premised on Jan Smith's belief that Matthew falsely accused staff at Hampstead Hospital of sexual abuse.   If evidence of Matthew's allegations of abuse at Hampstead Hospital was inadmissible under New Hampshire law, see infra, Ellsworth could not have established the requisite foundation for questioning Jan Smith about the special precautions.

12.   As I noted earlier, Jan Smith could probably have been called by the defense as a witness and asked if Matthew had expressed concerns to her about his safety and the possibility that he would be abused at a residential treatment center.   The defense could then have argued that these concerns were somehow predictive of the false allegations against Ellsworth.   Nevertheless, for the reasons set forth in my previous harmless error analysis, the absence of this evidence from the trial does not undermine my confidence in the guilty verdict or the fairness of Ellsworth's trial.   See Kyles v. Whitley, 514 U.S. 419, 434, 115 S.Ct. 1555, 131 L.Ed.2d 490 (1995) (“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.”).

13.   Technically, the defense's introduction of extrinsic evidence to rebut the prosecuting witness's denials during cross-examination is not a part of the cross-examination itself.   However, a defendant's qualified right to introduce such extrinsic evidence is treated in the caselaw as an element of the defendant's Sixth Amendment right to the opportunity to conduct an effective cross-examination of the primary witness.   See Davis, 415 U.S. at 316, 94 S.Ct. 1105;  Quinn, 234 F.3d at 848.

BETTY B. FLETCHER, Senior Circuit Judge.