CHILD WITNESSES AND THE CONFRONTATION CLAUSE

Abstract

After the Supreme Court’s ruling in Crawford v. Washington that a criminal defendant’s right to confront the witnesses against him is violated by the admission of testimonial hearsay that has not been cross-examined, lower courts have overturned convictions in which hearsay from children was admitted after child witnesses were either unwilling or unable to testify. A review of social scientific evidence regarding the dynamics of child sexual abuse suggests a means for facilitating the fair receipt of children’s evidence. Courts should hold that defendants have forfeited their confrontation rights if they exploited a child’s vulnerabilities such that they could reasonably anticipate that the child would be unavailable to testify. Exploitation includes choosing victims on the basis of their filial dependency, their vulnerability, or their immaturity, as well as taking actions that create or accentuate those vulnerabilities.

I. Introduction

In State v. Waddell,¹ a seven-year-old child named J.M.J. disclosed sexual abuse to a social services investigator in a videotaped interview.² In the interview, J.M.J. described “sexual intercourse, anal sex, the touching of her breast, and touching of [defendant’s] penis.”³ Waddell was J.M.J.’s next-door neighbor. J.M.J. would visit Waddell to watch T.V., to use his bathroom, and to play with his puppies.⁴ She would also visit his daughter and granddaughter when they were in his home.⁵ J.M.J. also revealed the abuse to her grandmother, a teacher, a nurse, a day-care provider, and a therapist.⁶ She explained that she had not immediately disclosed the abuse because she wanted to continue playing with Waddell’s puppies and because Waddell had threatened her with a knife that she should not tell.⁷ At trial, J.M.J. refused to testify. Her videotaped interview was admitted into evidence.

At trial it was also revealed that Waddell had previously abused his own daughter. She testified to eleven years of sexual abuse and recalled Waddell’s warning that “she would be sent to an orphanage if she told their secret.”⁸ She also testified that Waddell had taken naked pictures of her when she was a child.⁹ Waddell admitted abusing his daughter, but emphasized that it was “nonforcible.”¹⁰ He also admitted taking naked pictures, but asserted that she was seventeen and that she had “volunteered to be a photography model.”¹¹ A picture of a naked young girl was found in the defendant’s bedroom,¹² and J.M.J. reported that Waddell showed her a picture of a naked child.¹³ Waddell denied abusing J.M.J.

The jury convicted Waddell, but his conviction was overturned because of the Supreme Court’s decision in Crawford v. Washington.¹⁴ The appellate court held that because J.M.J. refused to testify and because her videotaped statement to a social services investigator constituted testimonial hearsay, admitting the videotape violated the defendant’s constitutional right to cross-examine J.M.J.¹⁵ The fact that the trial court had admitted the videotape only after assessing its reliability mattered naught,¹⁶ because Crawford emphasized that a defendant’s Confrontation Clause right to confront witnesses was a procedural right, rendering the reliability of the videotape irrelevant. Because the videotaped interview was central to the state’s case, the error could not be considered harmless and the conviction was overturned.¹⁷

Waddell is one of many cases around the country in which criminal convictions have been overturned because children’s out-of-court statements were admitted after they failed to testify.¹⁸ These cases include allegations of sexual abuse, physical abuse, and domestic violence, the types of cases in which child witnesses are most often called to testify.¹⁹ Typically, children’s statements were admitted after the trial courts found the children unavailable to testify and assessed the statements’ reliability under special hearsay exceptions for children’s complaints. Most states have such exceptions, which were promulgated to address the difficulties of proving child abuse while remaining mindful of the need for individualized assessments of the trustworthiness of children’s reports.²⁰

Crawford radically altered the treatment of hearsay under the Confrontation Clause of the Sixth Amendment.²¹ In Crawford, the Court held that a criminal defendant’s confrontation rights are violated by the admission of testimonial hearsay that has not been cross-examined. “Testimonial hearsay” includes statements “in which state actors are involved in a formal, out-of-court interrogation of a witness to obtain evidence for trial.”²² Unless an emergency exists, many if not most statements to government officials are likely to be considered testimonial. Statements made to social workers and the police when they interview children during investigations of alleged child abuse and domestic violence are usually deemed testimonial. Testimonial hearsay also includes most statements made to agents of the police.

In many cases involving child witnesses, children are interviewed at child advocacy centers by specially trained forensic interviewers who work closely with social services and the police.²³ In Waddell, a social worker interviewed the child at a child advocacy center while a police detective operated the video camera. The worker took several breaks, during which she consulted with the detective, and when the interview was complete, she forwarded the tape to law enforcement.²⁴ At the best child advocacy centers, interview protocols are followed, the interviews are videotaped, and both social services and the police observe the interviews in order to minimize the need for multiple interviews. Because the interviews are recorded, the exact words used by the interviewer and by the child can be closely scrutinized for evidence of suggestion, confabulation, or misinterpretation. Ironically, because of the formality and the input provided by state actors, the statements are almost sure to be deemed testimonial, triggering application of Crawford.²⁵

One of the few exceptions to the rule announced in Crawford is the doctrine of forfeiture by wrongdoing. In Giles v. California,²⁶ the state sought to introduce against the defendant statements that the murder victim previously made to the police.²⁷ The defendant successfully argued that his confrontation rights were presumptively denied because he could not cross-examine the victim at trial.²⁸ However, Giles held that a defendant forfeits his right to confront an unavailable hearsay declarant if the court finds that he engaged in wrongdoing that was designed to and did in fact cause the declarant’s unavailability.²⁹ A majority of the Court expressed the view that repeated acts of domestic violence against the declarant should suffice to prove that her murder was motivated by a desire to control the declarant and render her unavailable.³⁰ The opinion constitutes an unusual willingness by the Court to consider the dynamics of a crime in assessing the rights of criminal defendants.

The Giles opinion provides an opportunity to apply the forfeiture doctrine to the special challenges facing the prosecution in child-witness cases. Assessing the dynamics of child sexual abuse, we will argue that forfeiture should apply if the defendant exploited a child’s vulnerabilities such that he could reasonably anticipate that the child would be unavailable to testify. Exploitation includes choosing victims on the basis of their filial dependency, their vulnerability, or their immaturity. Exploitation also includes taking actions that create or accentuate those vulnerabilities.

Section II explores how Crawford altered the prosecution of child abuse. Section III describes the forfeiture-by-wrongdoing exception to the Confrontation Clause and discusses how a majority of the Court exhibited a willingness to consider the dynamics of abuse in applying the forfeiture doctrine. Section IV describes the dynamics of child sexual abuse and how perpetrators exploit the foreseeable unavailability of their victims. Section V demonstrates how the lower courts have missed the opportunity to apply forfeiture to child-witness cases. Section VI concludes.

II. How Crawford Altered the Prosecution of Child Abuse

Crawford v. Washington changed the status of hearsay evidence under the Confrontation Clause. Prior to Crawford, such evidence was assessed under Ohio v. Roberts,³¹ which admitted hearsay statements from unavailable declarants if they bore “indicia of reliability.”³² Reliability could be assumed either if the statements fell within a “firmly-rooted hearsay exception” or if they had “particularized guarantees of trustworthiness.”³³ Many hearsay statements by children had to satisfy the trustworthiness standard, because they were admitted under exceptions that were not firmly rooted, such as the residual exception or special statutory exceptions for children’s abuse complaints. Hence, in Idaho v. Wright,³⁴ the Court upheld the reversal of a sexual abuse conviction because statements made by a two-and-a-half-year-old to a physician were admitted under the “residual” exception to the hearsay rule and lacked guarantees of trustworthiness.³⁵ The Court held that the Confrontation Clause required exclusion of such statements “unless an affirmative reason, arising from the circumstances in which the statement was made, provides a basis for rebutting the presumption that a hearsay statement is not worthy of reliance at trial.”³⁶ The prosecution would have to prove that “the declarant’s truthfulness [was] so clear from the surrounding circumstances that the test of cross-examination would be of marginal utility.”³⁷ Although the Court rejected a procedural requirement that the statements be recorded,³⁸ it endorsed an assessment that is facilitated by recording: analysis of the spontaneity, consistency, and age-appropriateness of the child’s responses.³⁹

Crawford and its progeny overruled Roberts.⁴⁰ Crawford emphasized that the Confrontation Clause was about procedure, rather than substance: The defendant has the right to be confronted with the witnesses against him. Hence, the question is not whether the hearsay is true or false, but whether the hearsay declarant could be called a “witness” or not and whether the defendant had been given the right to confront the declarant.

To be sure, the Clause’s ultimate goal is to ensure reliability of evidence, but it is a procedural rather than a substantive guarantee. It commands, not that evidence be reliable, but that reliability be assessed in a particular manner: by testing in the crucible of cross-examination.⁴¹

The Court coined the term “testimonial hearsay,”⁴² which, as the Court has since clarified, is hearsay that has “a primary purpose of creating an out-of-court substitute for trial testimony.”⁴³ Crawford and subsequent cases have held that testimonial hearsay includes most statements made to the police,⁴⁴ other governmental officials,⁴⁵ or agents of the police,⁴⁶ unless there is an emergency⁴⁷ or the primary purpose of the interaction is something other than pursuing prosecution.⁴⁸ If the testimonial declarant fails to testify, the defendant is deprived of his right to confront a witness against him. In short, Crawford held that admission of testimonial hearsay from a declarant who cannot be cross-examined violates the Confrontation Clause. Crawford abandoned any application of the Clause to nontestimonial hearsay, and abandoned any attempt by the courts to analyze the reliability of hearsay under the Clause.⁴⁹

It immediately became clear that Crawford would have a major effect on the prosecution of crimes within the family and the home because of the frequency with which family members fail to testify. Crawford itself involved a case in which the defendant’s wife was the hearsay declarant and failed to testify because the defendant claimed the spousal privilege. Davis v. Washington,⁵⁰ in which the Court carved out an exigency exception to the rule, involved two domestic violence cases: one in which a woman gave statements to the police in her home while her husband was detained in another room, and the other in which a woman called 911 to report her boyfriend’s abuse as he was fleeing the home.⁵¹ In Giles v. California, in which the Court considered the forfeiture-by-wrongdoing exception, the defendant was charged with murdering his ex-girlfriend, and the challenged hearsay involved her statements to a police officer complaining of an attack after Giles accused her of having an affair.⁵²

A. FAILED PROSECUTIONS POST-CRAWFORD

In addition to the Waddell case described in the Introduction, the appellate reports are replete with cases in which application of Crawford led to reversals of convictions in which children’s hearsay had been admitted after the children failed to testify. These are typically cases in which the child’s inability to testify was unexpected and in which other evidence of guilt convinced the prosecutor to attempt to go forward without the child’s testimony.

In State v. Pitt,⁵³ the four-year-old victim, while living with her mother and the defendant, began to resist being alone with the defendant and disclosed sexual abuse to her mother.⁵⁴ She made consistent statements to a physician, a psychologist, and a forensic interviewer in a videotaped interview.⁵⁵ The physician found physical evidence of abuse.⁵⁶ The child also disclosed having seen the defendant sexually abuse the child’s five-year-old cousin, who herself confirmed abuse of both girls in a videotaped interview.⁵⁷ The state presented both girls at trial, but they appeared too upset and frightened to answer questions and were declared unavailable.⁵⁸ The videotaped interviews of both children were admitted, and the conviction was reversed on appeal because the interviews were testimonial hearsay.⁵⁹

In State v. Noah,⁶⁰ an eleven-year-old broke down during the preliminary hearing.⁶¹ Her hearsay was allowed under an exception requiring reliability.⁶² She had told her brother and mother that the defendant, a “longtime family friend,”⁶³ had touched her “private spot” and recounted seven specific incidents of abuse occurring over several years to a social worker and police.⁶⁴ The Kansas Supreme Court upheld reversal of the conviction on the ground that the statements to the social worker and the police were testimonial hearsay.⁶⁵

In People v. Sharp,⁶⁶ a five-year-old was found to be unavailable because she was “too traumatized.”⁶⁷ The trial court admitted a videotaped statement in which the child disclosed to a forensic interviewer her father’s sexual abuse, which was consistent with what she previously had told her mother.⁶⁸ The appellate court reversed the conviction because the statement was testimonial hearsay.⁶⁹

In In re S.R.,⁷⁰ a four-year-old was deemed unavailable after becoming “hysterical” at trial, but her hearsay was admitted after the trial court assessed its reliability.⁷¹ The mother had heard the child say, “Do you want me to do it to you?” during play, and the victim disclosed that the defendant, her uncle, had digitally penetrated her anus.⁷² The victim later disclosed other details of abuse to a forensic interviewer.⁷³ The appellate court reversed the conviction on the ground that the forensic interview was testimonial hearsay.⁷⁴

In several of the aforementioned cases, the excluded statements were videotaped interviews with the child. Hence, under Roberts, it was particularly easy for a trial court to review the tape and assess the reliability of the child’s statements. These cases provide the starkest examples of how Crawford has changed the reception of a child’s statements when the child fails to testify. The exact words spoken by the child and interviewer were available, but because the child could not (or would not) testify in court, the recorded statement could not be presented to the jury.

These cases represent just the tip of the iceberg, because Crawford has changed the nature of the cases that prosecutors choose to pursue, rendering its full effects invisible to readers of the appellate reports. Reversals of cases in which child interviews were admitted into evidence are likely to have deterred prosecutors in two respects. In cases in which the interview provided the most compelling evidence of abuse, prosecutors will have dropped charges or watered down charges in exchange for guilty pleas. In cases in which there was other admissible hearsay, prosecutors will have proceeded without introducing the interviews. Neither type of case would have led to appellate assessment of whether these interviews were testimonial.

In sum, Crawford has made it difficult to prosecute cases in which the child witness initially reported the crime to a state actor but later is afraid or intimidated by the prospect of testifying. Ironically, in these cases the most reliable hearsay evidence is the least likely to be admitted, because structured interviews that are captured on videotape are most likely to be deemed testimonial.

B. ATTEMPTS TO LIMIT CRAWFORD

The Supreme Court has refused to review a series of Confrontation Clause cases dealing with child witnesses,⁷⁵ leading the lower courts (and commentators) to come up with a variety of approaches. One approach is to argue that children below a certain (as yet indeterminate) age are incapable of making testimonial statements because they do not understand the implications of their accusations.⁷⁶ This position has not fared well in the lower courts, particularly when the statements were made to law enforcement or their agents.⁷⁷ However, the Supreme Court may have breathed new life into the argument in Michigan v. Bryant, in which it held that the perspective of the declarant is relevant in determining whether statements are testimonial.⁷⁸ Another approach is to find that the purpose of the interview was not to create evidence for prosecution, but to protect the child or facilitate medical treatment.⁷⁹ Finally, the exigency exception created by the Court in Davis (and applied in Bryant) has been applied to cases in which children were conceivably in immediate danger.⁸⁰

It is beyond the scope of this Article to discuss these proposals in detail. Suffice it to say that many child-witness statements are likely to be characterized as testimonial, even if all of these approaches ultimately succeed. For example, when a child has made an initial disclosure, and actions are taken to protect her from immediate harm, she is often taken to a child advocacy center to be interviewed by a specially trained child interviewer. The interview will be videotaped, and both law enforcement and child-protection workers will observe and provide input (typically from behind a two-way mirror).⁸¹ The interview will follow an interview protocol so that the interviewer avoids leading questions and elicits the most complete report possible. This type of interview is almost always held to be testimonial.⁸² Because of law enforcement involvement, it is unlikely to be characterized as medical or protective. Because a perpetrator has been identified and initial steps have been taken to protect the child, it is unlikely to be viewed as an emergency. And because many children are aware of the potential consequences of their disclosures, it is unlikely that it will be characterized as nontestimonial due to the child’s naiveté. Ironically, this type of interview tends to be the best documentation of the child’s report, and the strongest evidence that abuse occurred.

III. The Forfeiture-by-Wrongdoing Exception to the Confrontation Clause

Crawford and subsequent cases hinted at another possible approach when child witnesses are too scared or too young to testify: forfeiture by wrongdoing.⁸³ In Crawford, the Court mentioned the concept in passing. Justifying its position that the protections of the Confrontation Clause are procedural (testimonial hearsay must be subjected to cross-examination), rather than substantive (hearsay must be reliable), it noted that forfeiture as an exception to the confrontation right makes “no claim to be a surrogate means of assessing reliability,” but rather is founded “on essentially equitable grounds.”⁸⁴ In Davis, in which the Court considered a pair of domestic violence cases, the Court sought to reassure critics concerned about the effects of the holding on domestic violence prosecutions by again referring to forfeiture by wrongdoing. It expanded on the concept by noting that defendants “have the duty to refrain from acting in ways that destroy the integrity of the criminal-trial system.”⁸⁵ The Court noted that under the forfeiture doctrine, “one who obtains the absence of a witness by wrongdoing forfeits the constitutional right to confrontation.”⁸⁶

Both Crawford and Davis cited an 1878 case, Reynolds v. United States, the only previous Supreme Court case to give any serious attention to the concept of forfeiture by wrongdoing.⁸⁷ In Reynolds, George Reynolds was indicted on bigamy charges for marrying Amelia Jane Schofield, who lived with him in his home.⁸⁸ The sheriff attempted to serve Schofield several times at Reynolds’s home.⁸⁹ On one occasion, he was told by Reynolds’s first wife that Schofield had not been there for several weeks.⁹⁰ On another occasion, Reynolds told the sheriff that he would not help him find Schofield, and that she would “not appear” in the case.⁹¹ The sheriff, after several attempts and failed inquiries in the neighborhood, never served Schofield.⁹²

The Court approved admission of Schofield’s former testimony under the forfeiture-by-wrongdoing doctrine, concluding that “[t]he Constitution does not guarantee an accused person against the legitimate consequences of his own wrongful acts; it grants him the privilege of being confronted with the witnesses against him. But if he voluntarily keeps the witnesses away, he cannot insist on his privilege.”⁹³ Reynolds’s intent to keep Schofield from testifying was established by the fact that he failed to offer any assistance to the sheriff and in fact boasted to the sheriff that she would not appear.⁹⁴ The only evidence that Reynolds caused Schofield’s unavailability was that Reynolds cohabited with Schofield, thus giving him privileged access to her whereabouts.⁹⁵ This was enough for the Court to conclude that the burden shifted to Reynolds to demonstrate that he had not kept her away.⁹⁶

The Court squarely faced forfeiture by wrongdoing in Giles. In Giles, the defendant, Dwayne Giles, was charged with murdering his ex-girlfriend, Brenda Avie.⁹⁷ At trial, the defendant claimed self-defense and testified to prior acts of violence by Avie.⁹⁸ The hearsay at issue involved statements that Avie had made to the police responding to a domestic violence call three weeks before her murder. Avie, who was crying when she spoke, told the officer that Giles had accused her of having an affair, and that after the two began to argue, Giles grabbed her by the shirt, lifted her off the floor, and began to choke her. According to Avie, when she broke free and fell to the floor, Giles punched her in the face and head, and after she broke free again, he opened a folding knife, held it about three feet away from her, and threatened to kill her if he found her cheating on him.⁹⁹

In order to identify the proper scope of forfeiture, the Court looked to the common law at the time that the Bill of Rights was enacted and found that hearsay from unavailable witnesses was admissible when they were “detained” or “kept away” by the “means or procurement” of the defendant.¹⁰⁰ Giles agreed with Reynolds that the rationale of the rule is that “a defendant should not be permitted to benefit from his own wrong,” and added that “[t]he absence of a forfeiture rule covering this sort of conduct would create an intolerable incentive for defendants to bribe, intimidate, or even kill witnesses against them.”¹⁰¹

Because murder renders a declarant unavailable, the Court did not question whether the causation element of forfeiture was satisfied. Rather, the question was intent. The Court rejected an approach in which the defendant must simply have been aware of the effect of his actions on the declarant’s unavailability, which would mean the forfeiture rule would apply in all cases in which the defendant murdered the declarant. Rather, the plurality (and the concurrence) held that the exception “applied only when the defendant engaged in conduct designed to prevent the witness from testifying.”¹⁰²

Crucially, Giles acknowledged the role that the dynamics of domestic violence may play in assessing forfeiture.¹⁰³ The plurality opinion acknowledged that the domestic violence context in which the case occurred was relevant because domestic violence is often “intended to dissuade a victim from resorting to outside help.”¹⁰⁴ Hence, it may be possible to prove that, in committing the crime, the defendant “expressed the intent to isolate the victim and to stop her from reporting abuse to the authorities or cooperating with a criminal prosecution.”¹⁰⁵ The plurality opinion concluded that “[e]arlier abuse, or threats of abuse, intended to dissuade the victim from resorting to outside help would be highly relevant to this inquiry.”¹⁰⁶ The concurrence argued that such intent is equivalent to the “intent inferred on the part of the domestic abuser in the classic abusive relationship, which is meant to isolate the victim from outside help, including the aid of law enforcement and the judicial process.”¹⁰⁷ The dissent argued that mere knowledge, rather than purpose, should be sufficient and thus agreed with a possible reading of the concurrence: purpose could be “based on no more than evidence of a history of domestic violence.”¹⁰⁸ A majority of the Court thus believed that a history of domestic violence would support a forfeiture finding based on the theory that repeated violence is motivated by a desire to exert control over the victim.

A. LESSONS OF REYNOLDS AND GILES

There are several principles that one can extract from Reynolds and Giles that are significant for considering forfeiture in the context of child abuse cases. First, intent can be proven based on the dynamics of the crime with which the defendant is charged. In Giles, five Justices (the three dissenters and the two concurring Justices) endorsed the view that a court can presume intent based on a pattern of domestic violence. Two of these Justices (Souter and Stevens) have since been replaced (by Sotomayor and Kagan), and the new Justices’ views are unknown. Importantly, however, all seven of the Justices who have remained on the Court since Giles recognized the importance of understanding the dynamics of the charged crime in determining whether forfeiture should occur. We will explore the dynamics of child sexual abuse in order to see how perpetrators intend to silence their victims.

Second, the relationship between the defendant and the declarant is important. This fact was clearly recognized by the Court in Giles, when it discussed the unique dynamics of domestic violence because, by definition, domestic violence occurs between family members and intimates. In Reynolds, the Court recognized that a relationship and cohabitation equates with influence, so that if one proves intent and a means of carrying out that intent, causation can be presumed. Following Reynolds, the lower courts have also recognized the importance of the relationship between the defendant and the declarant in assessing forfeiture.¹⁰⁹

Moreover, particularly when there is a preexisting relationship between the defendant and the declarant, the lower courts recognize that the defendant’s actions need not be threatening in order to influence the declarant: cajolery can be as powerful a tool.¹¹⁰ Thus, wrongdoing may be established where the defendant promised gifts of value or otherwise bribed the witness¹¹¹ or where the defendant played upon the witness’s sympathy and pleaded with the witness not to testify.¹¹² Hence, courts applying the forfeiture-by-wrongdoing rule often find that outwardly benevolent conduct is evidence for forfeiture where the defendant and the witness have a preexisting relationship of trust, affection, or authority. We will see how child sexual abuse perpetrators use positive inducements to take advantage of their victims and deter them from disclosing abuse.

Third, the actions that give rise to forfeiture may occur before the charged crime. It is not necessary that the crime be completed or charges filed. All of the Justices in Giles believed that the defendant’s actions well in advance of the charged crime were relevant in assessing his intent. The lower courts have not gone as far, but they have recognized that the actions giving rise to forfeiture need not occur post-arrest or post-indictment.¹¹³ Child sexual abuse perpetrators plan their approaches; victim selection and grooming are designed to eliminate the need for post-crime threats.

Fourth, forfeiture can apply even if the declarant had other reasons to avoid testifying. Neither Reynolds nor Giles considered whether the declarants had motives to stay silent independent of the defendants’ actions.¹¹⁴ Commentators have argued that a declarant’s independent reasons might undermine forfeiture,¹¹⁵ but support for their position is lacking.¹¹⁶ Moreover, the lower courts have recognized that declarants’ purported reasons for their unavailability are often themselves the product of the defendants’ influence¹¹⁷ or simply not credible.¹¹⁸ Extending the argument, we will argue that perpetrators should be held responsible for their victims’ unavailability when they exploit or cultivate preexisting frailties.

Fifth, forfeiture can apply even if the defendant had other reasons to commit the wrongdoing.¹¹⁹ When Reynolds kept his wife’s whereabouts a secret, his only apparent motivation for doing so was to prevent her from being served. But when Giles murdered his ex-girlfriend, he was surely motivated as much by jealousy as by a desire to prevent her from testifying to the prior abuse; her statement to the police notes that he “threatened to kill her if he found her cheating on him.”¹²⁰ Perpetrators obviously abuse children because of their sexual interest and not because they wish to render them unavailable. Our argument for forfeiture will not be that perpetrators abuse children in order to silence them. Rather, we will show that the way in which they abuse children is designed to maintain their cooperation and silence.

Finally, forfeiture is not limited to cases in which the hearsay statements were themselves under oath and subject to cross-examination. Historical analysis¹²¹ and Reynolds suggested that this was so,¹²² but none of the Justices in Giles expressed this view. The prototypical hearsay statements in child sexual abuse cases are structured, videotaped interviews. The child may have been given a child-friendly version of an oath,¹²³ but cross-examination is obviously lacking.

Reynolds and Giles are complementary. One might read Reynolds to suggest that direct evidence of intent must be provided, but, given Giles, it is more consistent to assume that the Court in Reynolds believed that the strong facts of intent before it were sufficient rather than necessary. Giles illustrates how the dynamics of the charged acts can inform an analysis of what the defendant intended, even when direct evidence is lacking. Similarly, one might read Giles to suggest that clear evidence of causation must be provided (murdering a declarant clearly renders her unavailable); but, given Reynolds, it is more consistent to assume that Giles found the evidence of causation sufficient rather than necessary. Reynolds presumed causation merely based on cohabitation. Read together, the opinions suggest a flexible approach for considering whether a defendant forfeited his right to cross-examine an unavailable declarant.

Flexibility with respect to applying forfeiture is justifiable for two reasons. First, proof of forfeiture becomes more difficult as the defendant’s wrongdoing becomes more successful. If the defendant keeps the declarant off the stand altogether, then she cannot explain her absence. If the defendant successfully threatens the victim, she will be deterred from explaining to the court the reasons for her uncooperativeness.¹²⁴ If the defendant is particularly clever and colludes with the declarant, then she will provide alternative explanations for her uncooperativeness, which the defendant can point to as undermining causation.

Second, it is important to reiterate that forfeiture is an equitable principle: one should not be permitted to benefit from one’s wrong. From the perspective of fairness, intent is more important than causation. If the defendant desired that the declarant fail to testify and took actions to fulfill that desire, then the defendant should not be heard to complain if the declarant does not testify.

A focus on intent admittedly conflicts with the forfeiture-by-wrongdoing hearsay exception in the Federal Rules of Evidence, which requires both intent and causation.¹²⁵ However, the forfeiture-by-wrongdoing hearsay exception is not synonymous with constitutional doctrine, because the hearsay exception is concerned with both equity and reliability.¹²⁶ Causation is most important when forfeiture is used as an exception to the hearsay rule. A clear causal connection between the defendant’s actions and the declarant’s failure to testify increases the reliability of the statement. If the defendant’s actions are not the cause, an alternative explanation for the declarant’s failure to testify is that her statements were unreliable and she feared committing perjury.

Proof of intent should presumptively satisfy a constitutional forfeiture claim, but need not suffice as an exception to the hearsay rule. Forfeiture may make it fair to admit testimonial hearsay from an unavailable declarant, but the statutory rules regarding hearsay and the defendant’s due process rights may still apply to assess the reliability of the hearsay.¹²⁷ We will argue that it is not always necessary that a defendant cause a declarant’s unavailability for forfeiture to apply; rather, it is sufficient if the defendant exploited the declarant’s unavailability.

B. FORFEITURE BY EXPLOITATION

We propose that criminal defendants forfeit their rights to cross-examine a child witness if they exploit a child’s reasonably foreseeable unavailability. Exploitation includes taking advantage of vulnerabilities as well as creating or accentuating those vulnerabilities through one’s actions. Child sexual abusers exploit their victims’ vulnerabilities and immaturity. Perpetrators choose vulnerable victims, escalate the abuse over time, and cajole and threaten children into continued silence.

Technically, the defendant who exploits a child’s foreseeable unavailability need not cause that unavailability. One can either take advantage of preexisting vulnerabilities or create them. In either case, one’s intent is to ensure that the victim will be unavailable. Practically, defendants will usually take actions to guarantee that vulnerable victims remain so. When they take those steps, they have clearly both taken advantage of and caused unavailability.

States should amend their special hearsay exceptions for children’s complaints of abuse. The statutes currently require that, in order to admit this hearsay, courts find indicia of reliability and, when the child is unavailable, corroborative evidence of abuse. In order to comport with Crawford and Giles, the statutes could additionally require that if the child is unavailable and the statements are testimonial, the court must find that the defendant exploited the child’s foreseeable unavailability before admitting this hearsay.¹²⁸

Proof of exploitation will entail an examination of the relationship between the perpetrator and the child. Parents and adult household members enjoy authority and private access, as do professionals who care for and interact with children. In extrafamilial abuse, it is important to consider the extent and kinds of interaction between the child and the perpetrator. Child interviewers should ask the child about the pre-abuse relationship; the progression of abuse; the perpetrator’s statements to the child about the abuse and the consequences of disclosure; the child’s reasons for disclosing (and, if appropriate, for delaying disclosure); and the child’s feelings about the effects of disclosure. Recipients of the child’s disclosure can provide information about the context in which the disclosure occurred and the child’s explanations, if any, for delays.

IV. The Dynamics of Child Abuse

Understanding the dynamics of child sexual abuse is helpful in applying the principles of forfeiture by wrongdoing to child witnesses. The dynamics of abuse speak to both intent and causation. Intent is addressed from the perpetrator’s perspective, and looks to the process of victim selection, seduction, and silencing. Causation is addressed from the child’s perspective, and focuses on immaturity, filial dependency, self-blame, and secrecy. Although the focus here is on sexual abuse, many of the factors discussed here also keep child witnesses to other crimes off the stand.¹²⁹

There are several sources of information regarding the nature of child sexual abuse, including interviews with admitted perpetrators; population surveys; and clinical samples drawn from medical contexts, social service investigations, criminal investigations, dependency court, and criminal court.¹³⁰

A popular conception of the molester is a stranger who grabs a child off the street. This type of perpetrator’s strategy for avoiding detection is to conceal his identity. He might threaten the child not to tell, but such a threat may carry little weight because the child has no desire to protect the perpetrator. The perpetrator is unlikely to have continuing contact with the child, and by virtue of the violent abduction, the child does not feel complicit in the crime. The child may fail to disclose the abuse because of her immaturity or her inherent sense that she was somehow to blame, but from an intent perspective, the stranger perpetrator has acted more out of impulse than cunning.

In contrast, in the typical case of child abuse, the perpetrator is a parent, a parent figure, or a familiar and authoritative adult. The perpetrator selects his victim on the basis of immaturity, vulnerability, and private access. The perpetrator befriends the child before he abuses the child and introduces more serious sexual acts only gradually, so as to maintain the child’s trust and monitor the child’s continuing compliance and secretiveness. Once abuse has begun, the perpetrator maintains secrecy through admonishments and occasional threats. The success of the perpetrator’s approach is demonstrated by the child’s failure to disclose the abuse immediately.

A. VICTIM CHOICE: EXPLOITATION OF THE VULNERABLE CHILD

Virtually all sexual abuse is perpetrated by someone the child knows. In Smallbone and Wortley’s survey of 182 child sex perpetrators, only “6.5 percent of offenders had their first sexual contact with a stranger.”¹³¹ Population surveys confirm that with the exception of noncontact offenses (such as exhibitionism), strangers are rarely the perpetrators.¹³² Similarly, criminal samples are made up primarily of perpetrators familiar to the child, with the most common single type a parent or parent figure.¹³³ Evans and Lyon examined transcripts of over 400 children who testified in felony sexual abuse cases prosecuted in Los Angeles County over a five-year period and found that the defendant was a stranger to the child only 13% of the time.¹³⁴ This is also true with respect to the production of child pornography: only 4% of images confiscated by the police were photographed by strangers to the child, whereas 37% were photographed by parents, stepparents, or other relatives.¹³⁵

It is not an accident that perpetrators target children they know. Perpetrators choose victims on the basis of their accessibility; children living in perpetrators’ homes or with whom perpetrators work are most accessible. Sullivan and Beech interviewed forty-one perpetrators who molested children with whom they worked and found that 15% chose their profession exclusively to provide them access to victims; another 42% acknowledged that this partially motivated their job choice.¹³⁶

Within the group of potential victims, perpetrators often acknowledge that they look for the most vulnerable children. Conte and colleagues found that perpetrators “claimed a special ability to identify vulnerable children.”¹³⁷ Vulnerability was defined both in terms of a child’s status (e.g., living in a divorced home or being young) and in terms of her emotional or psychological state (e.g., a needy child, a depressed or unhappy child).¹³⁸ Forty-nine percent of the sex perpetrators interviewed by Elliott and colleagues stated that they targeted children who lacked self-confidence or self-esteem.¹³⁹ Beauregard and colleagues noted that child sex perpetrators often targeted “a child with family problems, without supervision, always on the street and in need of help.”¹⁴⁰

By choosing a victim from among family and friends, perpetrators ensure that their contact with the victim will be perceived by others as prosocial, and therefore will not arouse suspicion. This is also true of the child’s perception of the perpetrator: the perpetrator is in a position of trust and authority, and the child will interpret interest as paternalistic rather than predatory.

B. GROOMING OF THE VICTIM

Perpetrators emphasize the extent to which they seduce their victims over time rather than commit isolated assaults. Most child molestation typically includes attempts to obtain the assent and cooperation of victims. The first step for the sex perpetrator is to befriend the child, typically before any kind of physical contact is attempted. Leclerc and colleagues noted that child sex perpetrators adopt strategies “that are similar to prosocial behaviors which consist of demonstrating love, attention and appreciation.”¹⁴¹ Both intrafamilial and extrafamilial sex perpetrators describe spending time with the child¹⁴² and giving the child gifts,¹⁴³ sometimes introducing children to alcohol and pornography.¹⁴⁴ The first sexual contact often does not occur for a substantial period of time, particularly given the speed with which children, particularly younger children, can form attachments to adults. In Smallbone and Wortley’s study, 76% of the intrafamilial perpetrators, 28% of the extrafamilial perpetrators, and 39% of the mixed-type perpetrators knew the child for more than one year before initiating abuse.¹⁴⁵

The second step is to desensitize the child to sexual touch through progressively more invasive sexual touch and talk. Kaufman and colleagues¹⁴⁶ found this to be the most-often-endorsed means of obtaining the child’s compliance by both intrafamilial and extrafamilial child sex perpetrators.¹⁴⁷ This approach has several purposes. The perpetrator can test the child’s willingness to acquiesce¹⁴⁸ and the likelihood that the child will disclose.¹⁴⁹ If the child discloses at an early stage of the process, the perpetrator can claim that the touch was merely affectionate, accidental, or otherwise nonsexual.¹⁵⁰ As the abuse progresses, the perpetrator can assure the child of the harmlessness and morality of his actions.¹⁵¹

When the sexual abuser is the child’s parent, the extra attention paid to the child not only has the effect of making the child feel special, but also isolates the child and the offending parent from the other family members. Christiansen and Blake found that “[p]otential victims become alienated from the mothers because these daughters are placed by their fathers in their mothers’ traditional role of confidante, intimate friend, and sex partner. Alienation from siblings occurs because of the privileges and special favors potential victims receive.”¹⁵²

Third, the perpetrator initiates overtly sexual acts. In part because of careful victim selection and preparation, this need not involve violent force. In Fleming’s population survey, only 7% of victims recalled actual violence, whereas 64% recalled verbal threats (including threats of violence), and 72% stated that some form of coercion was used.¹⁵³ In Hershkowitz’s large study of sexual abuse cases investigated in Israel, children reported coercion in 30% of the cases, and threats in only 10%.¹⁵⁴ In Lang and Frenzel’s sample, two-thirds of the sex perpetrators “frightened the children in some way.”¹⁵⁵

Much of the persuasive power comes from the perpetrator’s status as an adult. Kaufman and colleagues pointed out that adults’ “greater physical sizes, statuses afforded by their age (i.e., ‘When adults tell you to do something, you listen’), and greater perceived credibility may reduce the need for explicit threats to gain victim compliance in abusive sexual activity.”¹⁵⁶ They found that when comparing adolescent to adult perpetrators, adults endorsed fewer strategies for obtaining compliance and in particular, adults were less likely to have threatened the child with a weapon.

When perpetrators do endorse strategies for inducing compliance, they mention a mixture of bribes and threats, and the strategies are similar for both intrafamilial and extrafamilial perpetrators.¹⁵⁷ In Lang and Frenzel’s sample, bribery was as common as physical force.¹⁵⁸ Kaufman and colleagues found that the most common form of bribery was giving gifts, and that the most common threat—particularly among intrafamilial perpetrators—was to prey on children’s helplessness by threatening to “tell on them about having sex with [the perpetrator] or by making them feel as if there was nothing they could do to stop it.”¹⁵⁹ Researchers have speculated that the efficacy of such a threat is founded on the desensitization process: “[V]ictims’ repeated acquiescence early in the grooming process (e.g., to nonsexual touch) may lead victims to believe that they have granted permission for more intrusive sexual contact.”¹⁶⁰

The fact that perpetrators fail to use force does not mean that they are unwilling to use it or incapable of doing so. Most of the perpetrators in Elliott and colleagues’ study stated that if the child resisted, they would stop and try to initiate contact later (61%), but a substantial minority (39%) admitted that they would then resort to threats or actual violence in order to complete the act.¹⁶¹ Perpetrators may also understate their use of force in an attempt to minimize the seriousness of their acts. In Christiansen and Blake’s sample of fathers who abused their daughters, less than one-fourth acknowledged using threats or physical punishment, but almost half of the victims (45%) claimed they had.¹⁶²

With respect to criminal cases, prosecutors may be more willing to charge when force is involved, as this is likely to be more convincing to jurors. Nevertheless, in criminal samples, abuse without the use of force also predominates. Smith and colleagues, for example, emphasized that “in the vast majority of cases, the sexual abuse was imposed on the child by the defendant simply by using his/her authority, or stature, as an adult.”¹⁶³ In Sas and Cunningham’s sample of children who testified in sexual abuse prosecutions, 30% did not even realize that the sexual act was wrong when it first occurred.¹⁶⁴

C. THE EFFECTS OF VICTIM CHOICE AND GROOMING: NONDISCLOSURE

Because of the means by which the perpetrator has selected and groomed his victim, disclosure is unlikely. Population surveys reveal that most respondents who report having been abused as children delayed disclosing the abuse for more than a year,¹⁶⁵ a large percentage had never told anyone before the survey,¹⁶⁶ and 90% of the abuse was never reported to authorities.¹⁶⁷

Obviously, because disclosure is the primary means by which abuse is discovered, abuse that is never disclosed to authorities will rarely if ever find its way into clinical samples or criminal case samples of abuse. Indeed, perpetrators typically admit having had a number of victims whose abuse was never brought to the attention of the authorities.¹⁶⁸

However, the extent to which perpetrators succeed in silencing their victims in clinical and criminal samples can be assessed by examining delays in disclosure. Clinical samples confirm that delays are common.¹⁶⁹ In criminal samples, delays are common as well. In Sas and Cunningham’s sample, two-thirds of the child witnesses reported having delayed reporting more than forty-eight hours after abuse, and one-third delayed more than a year after the first time abuse occurred.¹⁷⁰

Further evidence of nondisclosure can be found in the statistics on repeated abuse. If the child fails to report the abuse when it first occurs, it is likely to occur again.¹⁷¹ Over two-thirds of perpetrators report abusing the same victim over time.¹⁷² Charges of repeated abuse are also the norm in criminal samples,¹⁷³ and charges provide a conservative measure of whether abuse was repeated, because prosecutors often charge only what they are most confident they can prove and specifying individual acts is often difficult for child witnesses.¹⁷⁴

If the perpetrator has selected a child within his family or under his care, he can rely on the natural bonds between child and family to prevent disclosure. The most common factor that predicts delay in reporting abuse is the relationship between the perpetrator and the child: the closer the relationship, the longer the delay. This is true in population surveys,¹⁷⁵ clinical samples,¹⁷⁶ and criminal samples.¹⁷⁷ The relationship also affects the likelihood that the child will be inconsistent in her reports and ultimately recant the allegation of abuse.¹⁷⁸ Indeed, when asked, children endorse different rates of disclosure against parents and against strangers. By four years of age, children will predict less disclosure of parental transgressions than stranger transgressions (when asked what children “would” do), and by six years of age, children will endorse this difference as a norm (when asked what children “should” do).¹⁷⁹ By six years of age, children also make distinctions among the recipients of their disclosures, particularly disfavoring the reporting of parental transgressions to the police.¹⁸⁰ In laboratory research where children observe a parent or a stranger commit a minor transgression (such as stealing a book), children are more likely to keep secrets for the parent.¹⁸¹

If the perpetrator is close to the child’s mother, the mother is less likely to believe her child when the child discloses abuse.¹⁸² If the mother is unsupportive, the child is less likely to disclose in the first place,¹⁸³ more likely to delay reporting,¹⁸⁴ and more likely to recant her allegations.¹⁸⁵

Of course, not all parents are positive figures in children’s lives, and many perpetrators can expect secrecy from their victims because of fear rather than respect. In their sample of sexual abuse cases tried in criminal courts, Sas and Cunningham found that in about half the cases, the child had been exposed to domestic violence¹⁸⁶ and that “overt threats were not necessary [to deter immediate disclosure] if the man had a history of violence within the home.”¹⁸⁷

If the perpetrator chooses a child outside his family, but nevertheless under his influence, he can develop bonds that make him a father figure. Those bonds often extend beyond the child to the child’s family. Hence, the child looks up to the perpetrator and holds him in high esteem. The child is deterred from reporting the abuse for fear of harming the perpetrator or those in the child’s family who are close to the perpetrator. The child also has reason to doubt that others will believe his report because they will have difficulty believing the perpetrator is a child molester.

The gradual introduction of sexual acts will increase the likelihood that the child feels complicit in the abuse and will thus feel guilt, shame, and embarrassment, further deterring disclosure. The more manipulative forms of abuse are likely to increase children’s perceptions of self-blame. If the child fails to resist, she is more likely to believe that she consented. If she delays in reporting, she is more likely to believe that subsequent acts of abuse were consensual, or at least that her failure to disclose is responsible for their reoccurrence. The child may intuit these beliefs, and the perpetrator is likely to encourage them explicitly both to minimize the perpetrator’s own responsibility for the abuse and to help maintain secrecy.

In their sample of criminal cases, Sas and Cunningham found that delay was more likely if the child experienced pre-abuse grooming and “had been subjected to subtle and non-aggressive techniques to secure compliance with the sexual act.”¹⁸⁸ Clinical studies also find that manipulation is more likely than coercion to lead to a delay in disclosure.¹⁸⁹ Further, the younger the child, the more likely the self-blame,¹⁹⁰ and children abused by someone within the family exhibit more self-blame than children abused by someone outside the family.¹⁹¹ Self-blame has been found to delay disclosure.¹⁹²

When asked why they delayed disclosing or never disclosed, victims report many of the factors discussed above. In Anderson’s population survey:

When asked what had prevented disclosure, 65% of the victims gave these reasons: expected to be blamed (29% of cases), embarrassment (25%), not wanting to upset anyone (24%), expected disbelief (23%), not bothered by abuse (18%), wished to protect the abuser (14%), fear of abuser (11%), and wanting to obey adults (3%).¹⁹³

In Fleming’s population survey:

When the women were asked what prevented disclosure, by far the most common reason given was embarrassment or shame (47/80 [46%]), followed by the belief that the other person would not be able to help them (23/80 [23%]), or would somehow blame or punish them for the abuse (19/80 [18%]).¹⁹⁴

And in Sas’s criminal sample:

In order of frequency, their responses were: fear of harm to self or others; fear of bad consequences for self (e.g., rejection by parent); concern for family and to protect them from disruption; fear of disbelief; never thought about telling; embarrassment/stigma; concern for bad consequences for abuser; lack of someone trusted to tell; and altruism—enduring abuse to protect other children from abuser.¹⁹⁵

D. OVERT THREATS

Notably, the discussion thus far has not assumed that perpetrators overtly threaten their victims. A successful perpetrator does not need to make any overt threats because the child will be sufficiently motivated to keep the abuse a secret. Perhaps more than any other factor, this distinguishes nondisclosure in sexual abuse cases from the classic forfeiture situation, in which the perpetrator threatens the victim in order to enforce secrecy.

The percentage of perpetrators who report specifically warning the child not to tell varies widely across the studies. In Budin and Johnson’s sample, 25% acknowledged threats not to disclose.¹⁹⁶ In Elliott and colleagues’ study, 33% acknowledged telling the child not to tell.¹⁹⁷ In Lang and Frenzel’s study, 40% of extrafamilial perpetrators and 85% of incest perpetrators acknowledged telling the child not to tell.¹⁹⁸

Of course, these may be underestimates. Kaufman and colleagues compared what child sex perpetrators admitted in interviews to what their therapists recalled from perpetrators’ records (and prior admissions) and found that the sex perpetrators consistently underreported their use of threats to induce both compliance and secrecy.¹⁹⁹ However, victims in criminal samples also report perpetrator threats to keep the abuse secret in no more than half the cases (the percentages range from about 25%²⁰⁰ to 50%).²⁰¹

When they do acknowledge discussing disclosure with the child, perpetrators report a wide variety of inducements to secrecy. Perpetrators often refer to serious consequences from disclosure. Sixty-one percent of the perpetrators in Smallbone and Wortley told children that the perpetrators would go to jail or get in trouble.²⁰² Forty-three percent of the incest perpetrators in Lang and Frenzel threatened that the family would split up.²⁰³ Twenty-four percent of the perpetrators in Elliott and colleagues’ study used anger and the threat of physical force.²⁰⁴

Oftentimes, however, the threats are much milder and simply refer to the loss of the perpetrator’s love and attention. Kaufman and colleagues reported that perpetrators most often endorsed strategies that involved giving or withdrawing benefits for nondisclosure, such as giving children special rewards or privileges, and telling children that the perpetrator (or caretaker(s)) would no longer love them if they disclosed.²⁰⁵ Similarly, Smallbone and Wortley found that perpetrators endorsed giving children special rewards or privileges (21%) and relied on children’s fears that they would lose the perpetrators’ affection (36%).²⁰⁶ In Elliott and colleagues’ study, 20% of perpetrators endorsed threatening the loss of love or stating that the child was to blame.²⁰⁷ Lang and Frenzel found that these sorts of threats—expressing love for the child, giving the child special favors, and avoiding punishing the child—were more common among incest perpetrators than among extrafamilial perpetrators, perhaps because these threats relied on the use of parental authority and control.²⁰⁸

Similar types of threats are reported by victims in the criminal samples. Smith and colleagues found that:

[W]arnings ranged from pleas that the abuser would get into trouble if the child told (or that the abuser would be sent away and the child would never see them again—a powerful message to a young child whose abuser is also a “beloved” parent), to threats that the child would be blamed for the abuse (especially troubling were children who were told that the defendant’s intimate—the child’s mother—would blame the child for “having sex” with the defendant and would thus turn against him or her), to ominous warnings that the defendant would hurt or kill the child (or someone he or she loved) if they revealed the abuse.²⁰⁹

Although it might seem obvious that overt threats deter disclosure, the evidence is actually mixed. In her large clinical sample, Hershkowitz found that threats increased the likelihood that disclosure was delayed.²¹⁰ In their criminal sample, Sas and Cunningham found that “threats were far more common” among children who delayed reporting.²¹¹ On the other hand, two other criminal samples have not found clear evidence of the effect of threats. Gray did not find a relation between threats and nondisclosure, although she failed to look at delays, which is a more sensitive measure.²¹² In another criminal sample, Goodman-Brown and colleagues found that fear of harm to self or the perpetrator did not predict delays in disclosure, although fear of harm to others did.²¹³

There are a number of reasons why threats may not always correlate with nondisclosure. One problem is that researchers do not always distinguish between threats to comply with the perpetrator’s sexual demands and threats to keep the abuse a secret.²¹⁴ Threats to comply suggest abuse by force rather than seduction, which is less likely to effectively silence a child. Second, as noted above, threats are only one of many means by which perpetrators silence children, and it is likely that perpetrators use threats less often when other forces are effective in maintaining secrecy. Third, threats can only be documented based on children’s reports (or perpetrators’ admissions), which may lead to misestimation.

It is notable that the subtler methods of exploiting and inducing compliance are more consistently found to deter or delay disclosure. Perpetrators who do not have to threaten their victims are most likely to enjoy continued secrecy. The modus operandi of child molesters thus demonstrates how a perpetrator both exploits and nurtures children’s vulnerabilities. His purpose is to molest without the danger that the child will disclose the abuse. This both ensures continued access to the child and avoids discovery of the abuse. The efficacy of perpetrators’ methods is demonstrated by the facts that most children do not disclose their abuse and that the vast majority of abuse is never reported to authorities. The perpetrator reasonably anticipates that the means by which he accomplishes abuse ensures that the child will not speak out against him.

E. ADDRESSING COUNTERARGUMENTS

Several concerns might be raised with our depiction of the dynamics of child sexual abuse: (1) the data is tainted by false allegations; (2) children’s reluctance to disclose abuse is exaggerated; and (3) the focus should be on cases that are criminally prosecuted, not on sexual abuse generally. We will address each of these concerns in turn.

1. False allegations

Studies in which victims are asked about their experiences are subject to the concern that some of the allegations are false. In child sexual abuse cases, the child’s statement is often the strongest evidence that abuse occurred. However, the child may be lying, may have been coached, or may have formed false beliefs due to adult suggestion. Several factors minimize these concerns. First, interviews with admitted perpetrators provide convergent evidence for the dynamics of sexual abuse. Second, respondents who describe abuse in population surveys are not subject to the sorts of pressures that might undermine child witnesses’ claims. As noted above, population surveys consistently find that of those respondents who claim to have been sexually abused as children, only about 10% state that their abuse was ever reported to the authorities.²¹⁵ This reduces the likelihood of false positives, because overzealous parents and investigative authorities are the most commonly cited causes of false allegations. Similarly, only about 2% report that their memory of abuse was elicited with the support of a therapist,²¹⁶ another possible source of false reports. Third, even if these samples contain some false allegations, it is not obvious that this would undermine evidence that perpetrators are deliberately exploitative. For this to be the case, false allegations would have to contain more evidence of exploitation than true cases. This seems unlikely, if only because false allegations are usually attributed to the unfounded suspicions of lay persons, and lay conceptions of sexual abuse likely assume force.²¹⁷

2. Reluctance

Some researchers have argued that children’s reluctance to disclose sexual abuse is exaggerated. They primarily rely on the fact that in clinical samples of children questioned about suspected sexual abuse, the disclosure rate among children determined to be sexually abused is very high.²¹⁸ The problem with this argument is that clinical samples contain an inordinate number of victims who are uncommonly willing to disclose abuse. The most common evidence of abuse is a disclosure. Because a disclosure is what usually triggers suspicion of abuse and a report to the authorities, and disclosure to an investigator is usually necessary to substantiate abuse, disclosure rates are necessarily very high in substantiated abuse cases.²¹⁹

One solution is to identify children whose abuse could be identified and substantiated without questioning the child.²²⁰ If abused children are not reluctant to disclose, then those children should acknowledge being abused when directly asked. Lyon reviewed several decades of research examining disclosure rates among children with gonorrhea, most of whom had been diagnosed before ever being questioned about sexual abuse. Less than half of the children disclosed sexual abuse when first questioned.²²¹ Similarly low rates of disclosure have been found in the few studies examining cases in which evidence of sexual abuse surfaced before the child’s disclosure.²²²

Another solution to the problem with relying on clinical studies in order to assess children’s reluctance to disclose is to resort to population surveys. Unlike clinical samples, which enlist participants who already self-identified as former victims, population surveys can identify former victims who have never previously disclosed their abuse. As noted above, substantial percentages of survey respondents who acknowledge childhood abuse report having delayed disclosing or never having disclosed at all.²²³

Skeptics argue that survey respondents’ nondisclosure was simply due to the fact that they were never asked.²²⁴ This may be true, though whether they were ever asked is unknown. Regardless, the surveys provide other evidence of reluctance. First, substantiated abuse is often subsequently denied by survey respondents.²²⁵ Second, more persistent questioning elicits more reports of abuse.²²⁶ Third, respondents surveyed repeatedly are often inconsistent in acknowledging that abuse occurred.²²⁷

3. Criminal cases vs. sexual abuse generally

One might argue that since we are crafting a rule to apply in criminal cases, we ought to focus exclusively on criminal samples. However, if the goal is to identify what is reasonably foreseeable from the perpetrator’s perspective when he abuses a child, one needs to understand abuse in general. As one moves from clinical samples to criminal court samples, the children become less representative of abused children in general. In a number of jurisdictions, most substantiated cases of sexual abuse are never referred for prosecution.²²⁸ If one focuses on prosecution samples, one understates the difficulties that child victims have in coming forward. Prosecutors reject the cases that are most fit for forfeiture treatment. Prosecutors are less likely to file charges when the child is young, exhibits reluctance or inconsistency in reporting, and the family is unsupportive.²²⁹ Cases are more often dismissed when the child is young and the abuse occurs within the family.²³⁰ As a result, cases that go to trial typically involve children who successfully take the stand and maintain their allegations.²³¹

Nevertheless, testifying still constitutes a major hurdle for child witnesses. In their study of criminal trials, Goodman and colleagues found that children’s greatest fear was of seeing the defendant in court.²³² In a lab study, Goodman and colleagues conducted a mock trial in which four- to eight-year-olds were asked to testify. Despite their efforts to make the experiment as comfortable as possible for the children, 25% of the children “refused to testify, either by outright refusal or by appearing distressed so that the RA judged that the child should not continue.”²³³

In sum, even if almost all children who entered the court process managed to testify, this would not undercut the argument that a perpetrator can reasonably foresee a victim’s unavailability in cases in which he has selected, groomed, and admonished his victim to keep silent. The children for whom charges are filed are the tip of the iceberg. When a child ultimately refuses to take the stand, this only surprises those whose experience is limited to the cases that come to trial.

V. Forfeiture by Exploitation in the Courts

The Supreme Court has shown some appreciation of the dynamics of abuse. The Court has noted that “a child’s feelings of vulnerability and guilt and his or her unwillingness to come forward are particularly acute when the abuser is a parent.”²³⁴ Citing some of the research reviewed here, the Court has found that “[u]nderreporting is a common problem with respect to child sexual abuse. … [O]ne of the most commonly cited reasons for nondisclosure is fear of negative consequences for the perpetrator, a concern that has special force where the abuser is a family member. …”²³⁵

The lower courts have sometimes been amenable to the arguments underlying forfeiture by exploitation, but have trod carefully because of continuing uncertainties over the scope of the forfeiture doctrine. The typical case of forfeiture by wrongdoing involves ham-fisted silencing of witnesses through murder and overt threats of violence.²³⁶ These cases raise only a few of the issues that arise in child abuse prosecutions. Most of the cases cited above in which appellate courts reversed convictions due to the admission of testimonial hearsay from unavailable child witnesses did not even consider the issue of forfeiture.

A notable exception is State v. Waddell, the case discussed in the Introduction. The reader will recall that seven-year-old J.M.J. accused Waddell, her next-door neighbor, of intercourse, sodomy, and other sexual acts. The state anticipated the girl’s difficulty in testifying, and did what it could to either present her at trial or lay the groundwork for admission of her videotaped interview. Before trial, the state had filed “several motions designed to provide a careful structure for her courtroom appearance.”²³⁷ Shortly before the trial, a psychotherapist who spoke to J.M.J. observed that she “tried to avoid talking about the abuse,” but was willing to draw a picture depicting genital touching.²³⁸

The state then made a motion to find J.M.J. unavailable to testify, which was denied.²³⁹ “During the hearing on her availability … J.M.J. told defense counsel that she was scared, but said she was not afraid that anyone in the room would do anything to her. She stated she remembered what happened, but she did not want to talk about it.”²⁴⁰ As a result, “J.M.J. was required to appear at trial, but refused to talk about the events involving Waddell.”²⁴¹

The appellate court found that the facts of the case did not justify a forfeiture finding. It emphasized that “in forfeiture cases involving threats or coercion, the threats or coercion occurred after the events giving rise to the criminal charges,” and noted that “there was no evidence Waddell had any contact with J.M.J. after his detention.”²⁴² The arbitrary temporal limitation prevented the court from considering the ways in which Waddell had exploited the child.

Had the court felt comfortable assessing the relationship between the defendant and J.M.J. and the means by which he accomplished the abuse, it could have reached a different conclusion regarding her failure at the availability hearing to testify to her fears and her ultimate refusal to testify. As the reader will recall, Waddell was J.M.J.’s next-door neighbor, and J.M.J. “played frequently in Waddell’s yard and with his pets. She also occasionally went inside Waddell’s trailer.”²⁴³ According to Waddell, she would come into his trailer unannounced, they had watched television together, and she had used his bathroom.²⁴⁴ He testified that J.M.J had been mad at him for failing to get her a Christmas present and for not buying her a new pet turtle when her old turtle died.²⁴⁵ J.M.J. was also acquainted with his daughter and granddaughter, and she visited them at his trailer (while he was in jail awaiting trial).²⁴⁶ When J.M.J. first disclosed Waddell’s abuse to her grandmother, she said, “‘I still want to play with the puppies, but Kenny touched me in my privates.’”²⁴⁷ She also told two other adults that she had not told because she wanted to continue seeing the puppies.²⁴⁸

In subsequent interviews, J.M.J. described more serious worries. According to the daycare provider, J.M.J. said that “Waddell put a knife to her forehead or throat and threatened to kill her if she told anyone or did not come back.”²⁴⁹ In the videotaped interview:

J.M.J. stated that the first time Waddell touched her, it continued until she slapped him and pulled his hands off; then she ran to her grandmother’s house. J.M.J. described an incident where Waddell had a knife in his hand and said he would kill her unless she stayed … . J.M.J. said several times that she screamed loudly. After the touchings, Waddell let J.M.J. get dressed and run home. Later in the interview, J.M.J. said she had been outside playing with Waddell’s dogs when he grabbed her and forced her inside the house to commit these acts. She screamed. He also grabbed her when she took over Girl Scout cookies. When asked whether Waddell touched her with the knife in his hand, J.M.J. said, “No.” At the conclusion of the interview, however, J.M.J. spontaneously said Waddell had two knives in one hand and held them to her face. J.M.J. said Waddell would kill her if she did not stay with him.²⁵⁰

According to J.M.J.’s grandmother, after J.M.J.’s initial disclosure, she did not want to go outside because “‘[s]he was afraid he might be outside and know she told.’ J.M.J. did not want to even see Waddell’s trailer, so [her grandmother] covered up the windows on that side of her house.”²⁵¹

J.M.J.’s fears were no less relevant to a forfeiture determination because they were attributable to actions taken by the defendant during the course of repeated abuse. Waddell’s exploitation of the child was as deliberate and as wrongful as any defendant’s actions taken to silence a witness after a crime has been committed. Although many of Waddell’s actions were described by the child as violent and forceful, she originally described her desire to play with his puppies as the primary reason for her delayed disclosure. Hence, the evidence suggests Waddell utilized a combination of positive and coercive inducements in order to continue the abusive relationship.

On the other hand, not all sexual abuse should be subject to a forfeiture claim. The question is whether the perpetrator took advantage of and nurtured the child’s frailties in order to successfully commit (and recommit) the crime. We have emphasized the factors that perpetrators use to exploit children and effectively delay reporting: (1) a familial or otherwise close relationship between the perpetrator and the child and (2) grooming behavior by the perpetrator, which includes befriending the child, desensitizing the child to touch, using bribes and threats in order to induce compliance with sexual acts, and admonishing the child not to disclose. The efficacy of the defendant’s exploitation is evinced by nondisclosure, delays in disclosure, and inconsistencies in willingness to disclose over time.

Many of these elements were missing in In re Rolandis G., a case in which the Illinois Supreme Court rejected a forfeiture claim. In that case, the perpetrator was an eleven-year-old neighborhood boy who allegedly “forced [Von, a six-year-old boy] into a nearby wooded area and threatened him with a stick if he did not do what Rolandis wanted.”²⁵² Rolandis made Von “pinky swear” not to tell anyone about the abuse.²⁵³ Within minutes of coming home from the assault by Rolandis, and ignoring Rolandis’s pleas to come back outside, Von disclosed the abuse to his mother.²⁵⁴ True, Von told his mother that because of the pinky swear, he “did not want her to tell anyone else what Rolandis had done.”²⁵⁵ But within ten minutes a police officer arrived, and Von disclosed to him as well.²⁵⁶ Even armed with an understanding of the dynamics of abuse, a court could reasonably conclude that there was limited evidence that the way in which Rolandis chose his victim and accomplished the abuse effectively silenced the boy.

The Illinois Supreme Court acknowledged that “sexual abusers sometimes select children as their victims because children are generally more vulnerable to threats and coercion due to their age and immaturity.”²⁵⁷ However, the court then rejected forfeiture on the grounds that “there is no indication that when respondent sexually assaulted Von, his assault was motivated in any way by a desire to prevent Von from bearing witness against him at trial.”²⁵⁸ The reach of the court’s language is unclear. It is certainly reasonable to doubt that the eleven-year-old who wields a stick chooses and grooms his sexual abuse victims. But if the court assumed that sexual motivation for the act undermined forfeiture, then it misread Giles. The fact that Giles was a jealous ex-lover did not mean he could not intend to silence his victim; such mixed motives did not defeat the forfeiture claim. Moreover, in child sexual abuse cases, the question is not whether the perpetrator abused the child in order to silence him. The question is whether the perpetrator’s modus operandi was designed to maintain the child’s cooperation and silence.

With respect to the pinky swear, the Illinois Supreme Court held “there is nothing in the record to indicate that when respondent extracted the promise from Von, he did so in contemplation of some future trial.”²⁵⁹ Again, the reach of the court’s statement is unclear. Perhaps it was implying that an eleven-year-old fears only detection, but never imagines being prosecuted. If this is so, then it is reasonable to refuse to apply forfeiture. But surely defendants’ threats can be aimed at deterring any disclosure, above and beyond disclosures that lead to legal liability, and still provide evidence of forfeiture. Significantly, Giles referred to the defendant’s potential efforts to deter the victim from “resorting to outside help.”²⁶⁰ The defendant hoped that the victim would not report his abuse to anyone; thus, it would be odd to hold that his actions were not intended to deter reporting to the police.

Perhaps the most significant fact about In re Rolandis G. is that the child disclosed abuse to a police officer shortly after the first and only assault. The relation between threats and unavailability may be undermined by intervening testimonial statements. This consideration may explain the New York Court of Appeal’s uncertainty about forfeiture in People v. Johnson.²⁶¹ In Johnson, the defendant was a fifty-two-year-old pastor convicted of intercourse and sodomy with a twelve-year-old girl, one of his parishioners. The abuse was discovered when the mother found sexually explicit letters from the girl to the defendant. At her mother’s urging, the child placed a pretextual call to the defendant that was recorded by the police. The child also testified about the abuse before the grand jury, but “[w]hen called at trial … she had ‘nothing to say’ and otherwise refused to testify.”²⁶² In the recorded conversation, the defendant asked the child to “lie so that he would not ‘go to jail.’”²⁶³ The court acknowledged that his statements could “support an inference of domination or improper influence, providing the requisite link to the girl’s eventual refusal to testify at trial.”²⁶⁴ However, mentioning the fact that the child had both reported the abuse to the police and testified before a grand jury, the court ordered a hearing so that the “defendant would have … an opportunity to challenge the People’s evidence by raising questions as to defendant’s role in securing the victim’s unavailability at trial.”²⁶⁵ Hence, the court implied that the child’s testimonial statements, which included formal testimony, cast doubt on the causal connection between the defendant’s actions and the child’s refusal to testify at trial. Of course, the child’s willingness to testify before the grand jury (at which the defendant could not be present) does not preclude a finding that her unwillingness to testify at trial (at which the defendant had a right to be present) was attributed to the defendant’s actions. Nevertheless, whether and when the child disclosed to authorities are certainly relevant factors to the court’s assessment.

Commentators have gone so far as to claim that any disclosure of the abuse breaks the causal link between predisclosure threats and the victim’s failure to testify.²⁶⁶ However, if there is a close relationship between the defendant and the victim, the abuse continues over time, and the child delays reporting, then the victim’s ultimate failure to testify can fairly be attributed to the defendant. The efficacy of the defendant’s actions is demonstrated by the delays in reporting. The child’s disclosure likely reflects a fragile resolve; the same factors that increase delays in reporting increase the likelihood that the child’s report will be inconsistent over time. Recantations are more likely when the perpetrator is closer to the child and the child is younger.²⁶⁷ Furthermore, although legal doctrine since Crawford treats a child’s testimonial hearsay as equivalent to testimony, this formal equivalence says nothing about the psychological difference between disclosing abuse to a police officer in an interviewing room and testifying to abuse in an open courtroom in front of the defendant.

In assessing whether pretrial disclosures should affect the forfeiture inquiry, it is useful to consider the different perspectives of adults and children. When adults report crimes in the face of threats, they probably understand that reporting to the police can ultimately lead to testifying in court against the defendant. If they are unwilling to testify, they will simply fail to report or refuse to cooperate with the police. Hence, the fact that an adult reports a crime to police may weigh against finding a causal connection between the defendant’s prior threats and the declarant’s unavailability at trial.

Even if she does cooperate, an adult’s continuing fears of testifying may support a forfeiture claim. In Ware v. Harry,²⁶⁸ a post-Giles case, an eyewitness who described the crime and named the defendant spoke several times with the police. However, the police officers who interviewed the witness reported that she was “visibly frightened,” and she repeatedly told them that she was afraid to testify.²⁶⁹ The court found forfeiture on the basis of threats that were made at the time of the crime.²⁷⁰

Children have a less sophisticated understanding of the legal process, such that cooperation with the police is only weak evidence that they anticipate testifying. They are unlikely to recognize that by talking to the police, they are initiating a process that will not be complete unless they are capable of taking the stand and testifying against the defendant in the defendant’s presence.

Children’s limited understanding of the legal process has led some to argue that children’s statements to the police should not be classified as testimonial.²⁷¹ However, we believe this goes too far. It is likely that even young children understand the role of the police in arresting and incarcerating criminals; they know that the police put people in jail. Researchers have demonstrated that children exhibit an early appreciation of the power of the police. Lyon and colleagues found that even the youngest children they tested (four-year-olds) distinguished between reporting transgressions to the police and to teachers, and by six, children distinguished between reporting transgressions to the police and to parents (preferring parents as recipients).²⁷² Moore and colleagues found that “after the President, kindergartners view the policeman as the next most influential governmental figure.”²⁷³

What children probably fail to comprehend is the complicated process by which an arrest leads to conviction. It is because they do not understand the judicial process that young children are likely to believe that the police can summarily throw criminals in jail. Therefore, children’s statements to the police could be considered testimonial, at the same time that their willingness to talk to the police says little about their willingness to testify.²⁷⁴

In sum, when adults make testimonial statements, their willingness to do so is evidence that they are willing to cooperate with the prosecution, and that prior threats have not deterred them from coming forward. Of course, it is not conclusive evidence, but it helps to explain why courts would look for evidence that the defendant engaged in wrongdoing after the declarant’s initial disclosure. The fact that a young child has reported abuse to the police is much weaker evidence that threats have been ineffective. Rather, the efficacy of the defendant’s actions in silencing the child can be demonstrated through delays and inconsistencies in the child’s report.

VI. Conclusion

The dynamics of child abuse reveal how abusers exploit their victims’ vulnerabilities, making it unlikely that the victims will ever take the stand. Abusers silence children, and then complain when children are unable to speak out at trial. Our proposal takes into account the dynamics of child abuse and the vulnerability of child abuse victims, and it offers an approach that is both fair and consistent with Supreme Court doctrine.

A. EFFECTS OF THE PROPOSAL ON PROSECUTION

Adopting these views of forfeiture will not make it easy to prosecute child sexual abuse. Jurors intuit the frailties of child witnesses that researchers have emphasized: younger children have weaker memories and are more likely to be suggestible.²⁷⁵ Jurors expect to see medical evidence of sexual abuse,²⁷⁶ but even penile-vaginal penetration and sodomy usually do not leave physical signs.²⁷⁷ Even when there is medical evidence, it rarely points to a specific perpetrator, making the child’s identification essential.²⁷⁸ If the child discloses abuse with little emotion, which is typical, the jurors will be skeptical, expecting to see tears.²⁷⁹ And if the child reports that she didn’t resist, which is also typical, the jurors are less likely to convict.²⁸⁰ If the child is twelve or older, male jurors will see consent as a good defense, and if they are told not to consider consent, they are even more likely to do so.²⁸¹

Our proposal will not lead to a flood of child abuse prosecutions without the victim. Prosecutors believe that jurors want to see the child in court,²⁸² an intuition that is supported by research suggesting that jurors are more inclined to believe child witnesses when they see them in person.²⁸³ There is no evidence that prior to Crawford, prosecutors routinely went to trial without putting the child witness on the stand. Evans and Lyon examined five years of felony sexual abuse cases tried in Los Angeles County before Crawford and found that prosecutors relied on child hearsay only 3% of the time.²⁸⁴ Rather, our proposal makes it possible to proceed in the relatively rare cases in which there is a substantial amount of corroborative evidence, but the child’s statements are necessary to convince a jury beyond any reasonable doubt.

B. EXPANDING THE ANALYSIS TO OTHER TYPES OF CASES

When children testify in court, it is most often in sexual abuse cases. However, child witnesses also appear in physical abuse and domestic violence cases. In a number of cases post-Crawford, appellate courts reversed murder convictions in cases where children were the only witnesses to murders. In Bell v. State,²⁸⁵ a woman’s daughters (ages four and five) saw her killed. The younger daughter told police officers that Bell (her father) “asked [her mother] for money” and that her mother “emptied her purse out on the floor.”²⁸⁶ She then told the officers that “Bell pushed [her mother] down over a table, broke the table … broke a mirror in [the] bathroom … [and] used a small knife to put ‘blood on [her mother’s] back.’”²⁸⁷ The child’s statements were corroborated by the physical evidence—police found an overturned coffee table, a purse with its contents emptied, a broken mirror in the bathroom, and multiple knife wounds in the mother’s body.²⁸⁸ Both girls were found unavailable after they were unable to endure a mock pretrial practice session.²⁸⁹ The conviction was reversed because the statements were testimonial.²⁹⁰

In State v. Siler, a three-year-old reported witnessing his father beat and then hang his mother in their garage.²⁹¹ In response to questioning by a police detective, the child stated that his mother was “sleeping standing” in the garage.²⁹² The child told how “Daddy, mommy fighting” in the garage had scared him.²⁹³ When asked “if anyone was hurting mommy,” he responded, “Daddy did.”²⁹⁴ Although officers prevented the child from seeing his mother’s body, the child informed the detective that “the yellow thing” had held his mother upright in the garage.²⁹⁵ A yellow cord had been tied around his mother’s neck.²⁹⁶ When “asked who put the yellow thing on her, the child responded, ‘Daddy.’”²⁹⁷ The prosecution also presented evidence of threats the father made against the mother and of past incidents of domestic violence.²⁹⁸ The conviction was reversed because the child’s out-of-court statements were testimonial.²⁹⁹

In Flores v. State, the defendant’s five-year-old daughter told her foster mother and police that after her stepsister “peed on her [own] pants,” her mother hit her and “she never woke up.”³⁰⁰ The child refused to testify against her mother and the trial court admitted her hearsay after assessing its reliability.³⁰¹ The Nevada Supreme Court reversed the murder conviction on the ground that the child’s statements to police were testimonial hearsay.³⁰²

These cases suggest that Crawford has affected the prosecution of child physical abuse and domestic violence when children were the key witnesses. Prosecution of child physical abuse is often facilitated by documented physical injury, and thus may be easier than prosecuting sexual abuse, in which medical evidence is uncommon. However, the perpetrator must be identified, and the injuries must be proven to be intentional. Because of the privacy afforded people in their homes, the only potential witnesses are often family members. If the father is accused, then the defense is likely to be that the mother was to blame, and vice versa. Prosecution of domestic violence is notoriously difficult because the adult victims—typically mothers—are themselves often (if not usually) uncooperative.³⁰³ If the mother refuses to cooperate, it is likely that the child will be similarly unavailable.

The dynamics of child physical abuse and domestic violence are sure to parallel what we have documented here regarding child sexual abuse. Virtually all child physical abuse allegations involve parents.³⁰⁴ Domestic violence by definition involves assaults by family members. Spouse batterers frequently seek to control their victims, and violence is just one means by which they foster dependency.³⁰⁵ When domestic violence occurs, there is a high likelihood that child physical abuse is also occurring.³⁰⁶ Underreporting of domestic violence by adults is legion. Children’s underreporting of physical abuse is comparable to that of sexual abuse.³⁰⁷ Future work can explore the extent to which the dynamics of physical abuse and domestic violence offer similar explanations for child witnesses’ unavailability in these cases.³⁰⁸

C. EFFECTS OF THE PROPOSAL ON CHILD INTERVIEWING

The proposed elements of a forfeiture finding encourage a more complete interview with the child. The interviewer should explore the pre-abuse relationship between the perpetrator and the child; the progression of abuse over time; the child’s feelings about the abuse and its aftermath; any threats, bribes, or other inducements used by the perpetrator; and the child’s disclosure history. All of these facts are relevant to the determination of whether the defendant exploited the child’s foreseeable unavailability.

Our proposal encourages careful documentation of children’s reports soon after the child’s first statements about abuse. As we have discussed, many of the cases in which child abuse convictions were overturned because of Crawford involved videotaped interviews that had been assessed for their reliability pursuant to special hearsay exceptions for children’s complaints of abuse. By expanding defendants’ confrontation rights such that testimonial hearsay is always disallowed, the Court has discouraged the videotaping of formal interviews that benefit from the input and cooperation of all potential players, including social services and law enforcement. Hearsay is more likely to be classified as testimonial if it is carefully documented, if it is in response to structured questioning, and if it is conducted by agents of the government. However, all of these are elements of a state-of-the-art forensic interview and enhance both reliability and the ability to assess reliability.³⁰⁹ Videotaping enables viewers to assess the exact words used by the interviewer and by the child, whereas notes or verbal reports are sure to be less complete and less accurate.³¹⁰ When an interview is conducted well, the videotape will reveal appropriate instructions and rapport building, the use of open-ended questions, noncontingent encouragement, and the child’s free narrative of the alleged event. When an interview is conducted poorly, the videotape will reveal the interviewer’s bias, coercion, and suggestiveness on the one hand, and the child’s confusion, acquiescence, and misinterpretation on the other.

Videotaping also facilitates a reduction in repeated interviewing,³¹¹ because the interview can be shared by the different agencies (and the different players within the agencies) involved in the child’s case, including social services, the police, and attorneys in dependency court, family court, and criminal court. Repeated interviewing increases the risk that the child’s story will be distorted by suggestions and confabulation that become incorporated in the child’s narrative, making the child’s trial testimony a pale comparison to the original report.

In sum, our proposal accepts the proposition that defendants have a right to confront declarants whose words are gathered in anticipation of prosecution. At the same time, it recognizes the reality of child abuse: defendants subvert justice not only through overt threats and violent acts, but also through exploitation and manipulation of our most vulnerable citizens. When exploitation ensures that a child victim will not testify, a finding that the defendant has forfeited his confrontation rights is a fair means to let the child be heard.