Law & Contemporary Problems: Dorothy F. Marsil, Jean Montoya, David Ross, and Louise Graham, Child Witness Policy: Law Interfacing with Social Science, 65 Law & Contemp. Probs. 209 (Winter 2002)

Cited: 65 Law & Contemp. Probs. 209 (Winter 2002)
[*pg 209]

CHILD WITNESS POLICY: LAW INTERFACING WITH SOCIAL SCIENCE

DOROTHY F. MARSIL,^* JEAN MONTOYA,^** DAVID ROSS,^*** AND LOUISE GRAHAM^****

I. INTRODUCTION

II. CHILDREN'S SHIELDED TESTIMONY
   A. Child Witness Protection
   B. Defendant Prejudice
   C. Perception of Fairness
   D. Child Witness Prejudice
   E. Reliability of Shielded Testimony
   F. Ability of Jurors to Detect Deception

III. CHILDREN'S HEARSAY
   A. Child Witness Protection
   B. Defendant Prejudice
   C. Reliability of Children's Hearsay Evidence

IV. CONCLUSION

FOOTNOTES

I. INTRODUCTION

The number of children testifying in court has posed serious practical and legal problems for the judicial system. One problem confronting the courts is how to protect children from experiencing the psychological trauma resulting from a face-to-face confrontation with a defendant who may have physically harmed the child or threatened future harm to the child. Another concern is that this trauma may impair children's memory performance and their willingness to disclose the truth. In response to these concerns, child witness innovations proliferated throughout the United States in the 1980s and 1990s.¹ Among the innovations were: placing a screen between child witnesses and the defendant during children's testimony;² transmitting children's testimony into the [*pg 210] courtroom by closed-circuit television;³ and admitting children's otherwise inadmissible hearsay,⁴ including children's videotaped interviews.⁵ These innovations spawned a fair amount of appellate litigation regarding their constitutionality. Much of the litigation focused on whether a given innovation violated the Confrontation Clause,⁶ but questions about due process arose as well.

[*pg 211]

Court decisions regarding the constitutionality of child witness innovations rest on a number of assumptions that are subject to empirical testing. This article examines many of these assumptions and evaluates whether they are supported by social science evidence.

Part II of the article examines the use of shielding procedures in child sexual abuse prosecutions. It begins by exploring the Supreme Court's analysis of state laws providing protection by shielding the child witness from the defendant. Next, it explores various questions: Do child witnesses need protection from confrontational stress? Does shielding prejudice the defendant? Does it impact juror perception of the proceedings' fairness? Does shielding impact juror perception of the child witness? How reliable is children's shielded testimony? Does shielding impair juror ability to detect deception in the child witness?

Part III examines the use of hearsay testimony in child sexual abuse prosecutions. As with Part II, it begins by painting a picture of the legal landscape. Specifically, it considers the evidentiary and constitutional implications of using hearsay when children are witnesses. Part III then addresses various questions: Does admitting hearsay testimony protect the child witness? Does admitting hearsay testimony prejudice the defendant? How reliable is hearsay testimony offered in trials involving child witnesses? How accurate are the hearsay witnesses? Are they able to reconstruct details of their out-of-court exchange with the child witness? Does the eyewitness report deteriorate as it is transmitted down the hearsay chain from the child to the hearsay witness?

Part IV concludes the article. It highlights the insights gained from the interface of law and social science and makes suggestions for legal practice and future social science research.

II. CHILDREN'S SHIELDED TESTIMONY

Using shielding procedures is distinct from admitting hearsay statements. With shielding procedures, the child's view of the defendant is obstructed during the child's testimony at trial. When a screen is used, the child testifies from behind a screen placed between the child and the defendant. When one-way closed-circuit television is used, the child testifies from a separate testimonial room, and the child's testimony is transmitted into the courtroom where the defendant, jury, and judge are able to view the testimony.⁷ In either case, the child witness testifies under oath and is subject to cross-examination. In contrast, when hearsay is admitted in lieu of the child's testimony, not only does the child avoid physical confrontation with the defendant, but the child witness is also not [*pg 212] subject to cross-examination, because the prosecution relies on the testimony of a hearsay witness who reports the child's out-of-court statements.

The United States Supreme Court first addressed the issue of children's shielded testimony at trial in Coy v. Iowa.⁸ In Coy, the Court considered an Iowa statute that allowed children to testify from behind a screen blocking their view of the defendant.⁹ The trial court permitted a large screen to be placed between the defendant and the witness stand during the testimony of the child victims.¹⁰ Coy held that the defendant's right to face-to-face confrontation was violated by use of the screen.¹¹ The opinion, delivered by Justice Scalia, described the "irreducible literal meaning" of the Confrontation Clause as the "'right to meet face to face all those who appear and give evidence at trial.'"¹² In seeming contradiction, the Court stated in dicta that any exceptions to the "irreducible literal meaning" of the Confrontation Clause "would surely be allowed only when necessary to further an important public policy."¹³ The State argued that it had established necessity because the Iowa statute implied a legislative finding that child witnesses suffer trauma from testifying in their assailant's presence.¹⁴ The Court rejected this argument and indicated that showing necessity requires "individualized findings that . . . particular witnesses needed special protection."¹⁵ Such specific findings had not been made in Coy.

In Maryland v. Craig,¹⁶ this individualized showing of trauma was found by the trial court.¹⁷ Craig involved a Maryland statute that allowed children to testify by one-way closed-circuit television if the judge first found that the child would suffer "serious emotional distress such that the child [could not] reasonably communicate" at trial.¹⁸ After receiving expert testimony,¹⁹ the trial court made the requisite findings and allowed the children to testify using the one-way closed-circuit television procedure.²⁰ The children thereafter testified in a separate room in the presence of counsel but outside the presence of the judge, jury, and defendant, all of whom remained in the courtroom where a television [*pg 213] displayed the children's testimony.²¹ As in Coy, the child witnesses could not see the defendant while they testified.²² Unlike Coy, however, the child witnesses did not testify in the presence of the judge and jury. The Craig Court upheld the constitutionality of the Maryland procedure inasmuch as it required a showing of necessity to protect the particular child witness²³ and, unlike the Iowa procedure in Coy, did not rely on a "legislatively imposed presumption of trauma."²⁴ Justice O'Connor, writing for a five-member majority, reasoned that the defendant's right to physical confrontation was not absolute.²⁵ Her opinion further asserted that the state's interest in protecting child witnesses from the trauma of testifying in a child abuse case was sufficiently important to outweigh the right to a face-to-face meeting.²⁶ Justice Scalia and others dissented on the ground that the "text of the Sixth Amendment is clear" and requires face-to-face confrontation.²⁷

A. Child Witness Protection

Shielding procedures are primarily premised on the idea that psychological trauma occurs when child witnesses testify in the defendant's presence. In Coy v. Iowa,²⁸ the dissent noted that a child's fear and trauma of testifying in a confrontational setting had two "serious" identifiable consequences: "They may cause psychological injury to the child, and they may so overwhelm the child as to prevent the possibility of effective testimony, thereby undermining the truth-finding function of the trial itself."²⁹ Later, in Maryland v. Craig, the Supreme Court recognized the "growing body of academic literature documenting the psychological trauma suffered by child abuse victims who must testify in court."³⁰ Indeed, the phenomenon of confrontational stress experienced by children is amply supported by social science evidence.³¹

[*pg 214]

Research shows that most child witnesses primarily fear a face-to-face confrontation with the defendant.³² For example, Gail Goodman and her colleagues interviewed actual child witnesses and reported that confrontation with the defendant was the most stressful part of having to testify.³³ This evidence was supported and extended by the findings of Louise Dezwirek-Sas, who found that, in addition to fearing a confrontation with the defendant, child witnesses feared being hurt by the defendant, testifying on the stand, crying during testimony, being sent to jail, and failing to understand questions that were asked of them.³⁴

Requiring children to testify in the defendant's presence has significant consequences. These consequences were identified by the Coy dissent, namely, that children's stress may produce inaccurate or incomplete testimony.³⁵ Social science strongly supports this concern. In fact, research has shown that the high level of stress and anxiety experienced by child witnesses can decrease children's ability and/or willingness to provide complete and accurate evidence.

For example, a study by Douglas Peters provides support for the negative effects of confrontational stress on child witnesses' ability to disclose the truth.³⁶ In his study, half of the children individually observed a simulation of a man stealing money.³⁷ The children were then asked to identify the thief either from a photo array or a live lineup including the perpetrator.³⁸ The children who viewed a photo array accurately identified the perpetrator seventy-five percent of the time, with only eight percent of the children incorrectly responding that the thief was not in the lineup.³⁹ When the child was in the presence of the thief viewing the live lineup, however, his or her willingness to identify the perpetrator decreased.⁴⁰ In fact, only thirty-three percent of the children viewing the live lineup accurately identified the perpetrator, with fifty-eight percent of them incorrectly claiming that the perpetrator was not in the lineup.⁴¹ These results strongly indicate that children are less willing or able to accuse a person when that person is present.

Kay Bussey and her colleagues also studied the impact of the defendant's presence on a child's ability to provide accurate and complete testimony.⁴² In [*pg 215] this study, children ages three, five, and nine years old individually witnessed a man, the perpetrator, break a valued glass.⁴³ The children were then interviewed about the event in the presence or absence of the perpetrator. Younger children -- three- and five-year-olds -- were significantly less likely than older children to disclose the event, and even less likely in the presence of the perpetrator.⁴⁴ Even though the nine-year-old participants were equally likely to disclose the event in the presence or absence of the perpetrator, they experienced anxiety as a result of having to tell on the perpetrator in his presence.⁴⁵ Thus, the presence of the perpetrator resulted in young children being less willing to give complete and accurate evidence.⁴⁶ These children appear to be at a greater risk of experiencing a high level of stress and anxiety as a result of confronting a perpetrator than are older children.

The evidence makes clear that children who are required to confront the accused experience more stress than children who are not required to do so. Because the state has a legitimate interest in protecting the welfare of the child witness, it is important to consider employing protective measures for children due to possible psychological trauma resulting from testifying in the defendant's presence. Moreover, high levels of trauma can indeed impair children's memory and willingness to disclose the truth. Shielding the child witness from the defendant, by allowing the child to testify in the courtroom from behind a screen or by allowing the child's testimony to be transmitted into the courtroom by closed-circuit television, addresses the problem of children's confrontational stress. The source of the child's stress, the defendant, is simply removed from the child's presence.

B. Defendant Prejudice

In Coy v. Iowa,⁴⁷ the child witnesses, two thirteen-year-old girls, testified in the courtroom from behind a screen that shielded them from seeing the defendant.⁴⁸ The screen did not prevent the defendant from seeing the child witnesses.⁴⁹ Nor did the screen prevent the child witnesses from seeing and being seen by the judge, counsel, and jury.⁵⁰ The defendant argued that allowing the children to testify from behind a screen violated his right to due process, because the procedure would make him appear guilty.⁵¹ The trial court rejected the defendant's claim, but instructed the jury to draw no inference of guilt from [*pg 216] the screen.⁵² The Supreme Court did not reach the defendant's due process claim because it reversed the conviction on Confrontation Clause grounds.⁵³ The dissent, however, assumed that the defendant would not be prejudiced by the shielding device: "A screen is not the sort of trapping that generally is associated with those who have been convicted. It is therefore unlikely that the use of the screen had a subconscious effect on the jury's attitude toward [the defendant]."⁵⁴

Social science evidence supports the dissent's position that shielding does not prejudice the defendant. Researchers have studied the impact of shielding procedures on conviction rates.⁵⁵ In these studies, testimony using various protective measures is compared to open-court testimony given by the alleged child victim.⁵⁶ This research is an important step in understanding how the presentation of children's testimony may influence a jury.

First, a study by Janet Swim and colleagues (the "Swim study") examined the use of videotaped depositions versus live testimony by a child witness in court.⁵⁷ Mock jurors watched a videotape of a simulated child sexual abuse trial and completed pre- and post-deliberation questionnaires. They deliberated on charges of criminal sexual assault in the first degree, attempted criminal sexual assault in the first degree, and criminal sexual assault in the second degree.⁵⁸ The modality of the child's testimony impacted pre-deliberation verdicts with the results showing that the defendant was less likely to be found guilty on the charge of criminal sexual assault in the first degree when the videotaped depositions were used than when the child testified in open court.⁵⁹ The study found no difference between trial conditions on the other charges.⁶⁰ The finding suggests that when the charge is more serious, the jurors would rather hear from the child in open court before rendering a guilty verdict. Accordingly, any difference found in pre deliberation verdicts actually favored the defendant.⁶¹ No difference, however, in post-deliberation verdicts was found based on trial condition.⁶²

A study by David Ross and colleagues (the "Ross study") supported and extended these findings. In the Ross study, the trial stimulus was based on the [*pg 217] transcript of an actual child sexual abuse case involving a ten-year-old victim.⁶³ Mock jurors watched one of three videotaped trials that varied as to whether the child testified in open court confronting the defendant, in court with a protective shield, or via closed-circuit television.⁶⁴ In the first experiment, mock jurors watched the entire trial, including the testimony of the child and other witnesses, and then rendered a verdict.⁶⁵ The results showed that the verdict was not influenced by whether the child testified in open court, with a protective shield, or via closed-circuit television.⁶⁶ In the second experiment, mock jurors watched the same trial except that the child was the first and only witness to testify.⁶⁷ These mock jurors were more likely to find the defendant guilty when the child testified in open court than when the child testified from behind a protective shield or via closed-circuit television.⁶⁸ Thus, the presentation of the child's testimony made a difference when it was the only evidence presented: Shielding reduced the likelihood of conviction.

Finally, in a study by Gail Goodman and her colleagues (the "Goodman study"), children from two age groups -- five- to six and eight- to nine years old -- were videotaped while individually interacting with a male confederate.⁶⁹ In the guilty condition, the male asked the child to place stickers on exposed body parts, such as their toes, arms, and belly button.⁷⁰ In the not-guilty condition, the male asked the child to place stickers on items of clothing, such as their shoe, shirt sleeve, or belt.⁷¹ After a two-week delay, mock jurors participated in a simulated trial in which a child testified about the event in either open court confronting the defendant or by closed-circuit television.⁷² The male confederate was charged with videotaping a child displaying exposed body parts.⁷³ Results showed that the trial condition did not impact the conviction rate.⁷⁴

These studies suggest that mock jurors are not biased against the defendant when shielding procedures are employed. Moreover, protective devices do not appear to imply guilt. Not only do protective devices not increase convictions rates, they sometimes reduce them.

C. Perception of Fairness

In Coy v. Iowa, the Supreme Court acknowledged "[t]he perception that confrontation is essential to fairness."⁷⁵ Similarly, in Maryland v. Craig, the [*pg 218] Court recognized the "strong symbolic purpose served by requiring adverse witnesses at trial to testify in the accused's presence."⁷⁶ Craig concluded that the symbolic purpose of the Confrontation Clause was not impinged when shielding procedures were used.⁷⁷ Nevertheless, if jurors viewed shielding procedures as unfair to the defendant, the prosecution's case could be prejudiced.

Social science research has documented jurors' perception of trial fairness when shielding procedures are used. For example, Rod Lindsay, David Ross, and their colleagues showed mock jurors a videotaped re-creation of a child sexual abuse trial where the child's testimony was presented either in open court, from behind a protective shield, or via closed-circuit television.⁷⁸ In certain instances, the judge warned jurors that the use of the shield or closed-circuit television should not be used as evidence of the defendant's guilt; in others, the judge did not.⁷⁹ In addition, the participants were asked to place themselves in the role of either a juror, a sibling of the defendant, or a sibling of the victim's mother.⁸⁰ The results indicated that the use of the protective devices did not impact jurors' perceptions of the fairness of the trial.⁸¹ The judge's instructions did have an impact, however, in that mock jurors who received the implied guilt warning were significantly more likely to agree that the use of the protective procedures was fair compared with jurors who did not receive the warning.⁸² Finally, the perspective of the participant had a significant impact on perceived fairness.⁸³ Participants who played the role of the defendant's sibling perceived the protective devices to be more biased and unfair than participants who played the role of juror or sibling of the victim's mother.⁸⁴

These results are consistent with findings from other studies. The same results were found in the Ross⁸⁵ and Swim⁸⁶ studies that used a videotape of a simulated child sexual abuse trial as the stimulus, as well as in the Goodman study that used a highly realistic mock trial as its stimulus.⁸⁷ In particular, juror ratings of trial fairness were found not to differ whether the child testified using a shield, by closed-circuit television, or in open court. Thus, the evidence is consistent across a number of different studies. When protective measures are used, the trial is generally perceived to be as fair as when a child testifies in open court. This is especially true when the judge provides a warning to the ju- [*pg 219] rors not to allow the use of the shield or closed-circuit television to imply or suggest that the defendant is guilty.

D. Child Witness Prejudice

In addition to being concerned with the jury's perception of the trial's fairness in general, the prosecution is also concerned with the jury's perception of the child witness. In Maryland v. Craig, the Court noted "the many subtle effects face-to-face confrontation may have on an adversary criminal proceeding."⁸⁸ The Court was nevertheless satisfied that the presence of the other aspects of confrontation -- oath, cross-examination, and observation of witness demeanor, even if conducted by video monitor -- rendered shielded testimony the functional equivalent of live, in-person testimony.⁸⁹

Prosecutors, however, bear the burden of proving their case beyond a reasonable doubt, and "subtle effects" can make all the difference. The testimony of the child witness may be the centerpiece of the prosecutor's case. The prosecutor thus rightly questions whether shielded children's testimony is the functional equivalent of children's live, in-person testimony for purposes of jury perception of the child witness.

First, the Ross study also examined the use of protective shields and their impact on mock jurors' perceptions of child credibility.⁹⁰ Results showed that mock jurors rated the child's credibility the same regardless of the testimony condition.⁹¹

Similarly, the Swim study found that the medium of presentation did not significantly impact jurors' perceptions of the witnesses, including the child witness.⁹² Both of these studies, however, involved videotaped simulations of a child sexual abuse trial.⁹³ In other words, mock jurors in these studies were never presented with a live child witness, but only with a videotape of a live child witness.

Research using a live child witness has produced different results. The Goodman study examined mock jurors' perceptions of the child witness's credibility.⁹⁴ The results showed that while children gave more accurate testimony when protective measures were used, mock jurors viewed the child witness as less credible.⁹⁵ This effect was somewhat mediated in that children who testified more accurately were believed more credible by the mock jurors.⁹⁶ In a related [*pg 220] study by Ann Tobey and her colleagues,⁹⁷ jurors rated children who testified using closed-circuit television as less believable, less accurate for both the prosecution and the defense, less accurate in recalling the event, "more likely to have made up the story, less able to testify based on fact . . . than fantasy, less attractive, less intelligent, and less confident."⁹⁸ Interestingly, jurors also noted that children testifying in the closed-circuit condition were less stressed than children testifying in open court.⁹⁹

A study by Graham Davies and Elizabeth Noon evaluated the "Livelink" project, which involved the use of closed-circuit television for children's testimony in England and Wales.¹⁰⁰ Among other topics, they explored the perception of a child's credibility when he or she testifies via live link.¹⁰¹ Court professionals expressed concern that such testimony detracts from the impact of that witness, including the perception of diminished emotion on the part of the victim, reduced eye contact with the jury, and a loss of rapport with the jury.¹⁰² This reduced impact could lead a jury to form negative perceptions of a child's credibility.

These studies support the view that the credibility of the child witness is not enhanced, but suffers when protective procedures are used.¹⁰³ Therefore, jurors are likely to have a higher opinion of the child witness when the child appears in person.¹⁰⁴ Thus, the contrary findings of Ross et al. and Swim et al. may be attributable to the experimental stimulus.¹⁰⁵

E. Reliability of Shielded Testimony

In finding no violation of the Confrontation Clause, the Craig Court assumed that children's shielded testimony was "adequately" reliable.¹⁰⁶ Even if the purpose of the Confrontation Clause is to ensure that the evidence against a criminal defendant is reliable,¹⁰⁷ evidentiary reliability is not a monolithic concept.¹⁰⁸ The Coy Court recognized this:

[*pg 221]

The State can hardly gainsay the profound effect upon a witness of standing in the presence of the person the witness accuses, since that is the very phenomenon it relies upon to establish the potential "trauma" that allegedly justified the extraordinary procedure in the present case. That face-to-face presence may, unfortunately, upset the truthful rape victim or abused child; but by the same token it may confound and undo the false accuser, or reveal the child coached by a malevolent adult.¹⁰⁹

What does it mean to say that children's shielded testimony is "adequately" reliable? Is the appropriate standard the overall accuracy of children's shielded testimony? Does adequate reliability mean that children's shielded testimony is at least as accurate as their unshielded courtroom testimony? Or does perspective matter? Is the appropriate standard the incidence of false accusations? Does adequate reliability mean that children's false accusations occur no more frequently when children are shielded than when children are unshielded in the courtroom? Is the appropriate standard the incidence of false accusations as compared to false recantations?¹¹⁰ How many false recantations would offset a false accusation?¹¹¹

In Coy, the Court conceded that false allegations are less likely with face-to-face meetings:

The perception that confrontation is essential to fairness has persisted over the centuries because there is much truth to it. A witness "may feel quite differently when he has to repeat his story looking at the man whom he will harm greatly by distorting or mistaking the facts." . . . It is always more difficult to tell a lie about a person "to his face" than "behind his back."¹¹²

Even Craig recognized that "face-to-face confrontation enhances the accuracy of fact-finding by reducing the risk that a witness will wrongfully implicate an innocent person."¹¹³

Social science has only begun to answer questions about the reliability of shielded testimony. This nascent research has studied the performance of children who witnessed an event and were asked to testify about that event in either the presence or absence of the perpetrator, an experimental confederate.

[*pg 222]

For example, Paula Hill and Samuel Hill had seven- to nine-year-old children watch a video of an unpleasant exchange between a father and daughter.¹¹⁴ Children were then questioned about the video either in a courtroom setting with the father present or in a small room with the father absent.¹¹⁵ The children who testified in the small room were able to provide more detailed and accurate testimony than the children who testified in a courtroom facing the father.¹¹⁶ Additionally, the children in the protective condition were less likely to say "I don't know" or give no response than the children who testified in the courtroom setting in front of the father.¹¹⁷

In part, the Goodman study sought to determine whether testimony given via closed-circuit television versus in open court would improve children's accuracy.¹¹⁸ The researchers found that, in general, the modality of the child's testimony did not impact the accuracy of their free recall.¹¹⁹ One positive effect, however, was found: specifically, that younger children, the five- to six-year-olds, who testified in open court made more errors of omission than children of the same age who testified via closed-circuit television.¹²⁰

These studies provide some evidence that shielded children are more accurate, detailed, and make fewer errors than children who are required to testify in open court. It is important to note that in the studies described above, children were not testifying about a traumatic sexual abuse incident. They were testifying under conditions expected to be less stressful than if the child was testifying in a sexual abuse trial. The studies may therefore actually under-predict how shielding improves children's accuracy.

One concern regarding the use of shielded testimony is that the child may make more false statements. The evidence that follows shows that this concern may be unwarranted. Holly Orcutt and her colleagues conducted a study (the "Orcutt study") in part to address children's ability to lie when they testify in open court versus via closed-circuit television.¹²¹ In this study, seventy children ranging in age from seven to nine years old individually made a video with a male confederate.¹²² The children participated in one of three conditions: guilty, not guilty, or deception.¹²³ During the making of the video, children in the guilty condition placed stickers on exposed body parts -- for example, arms, toes, or belly button -- while children in the not-guilty condition placed stickers on [*pg 223] clothing and accessories -- for example, shirt sleeves, shoes, or belt.¹²⁴ In the deception condition, children made the same video as children in the not-guilty condition, but were coached to lie in their testimony and say they had placed stickers on exposed body parts, as occurred in the guilty condition.¹²⁵ This study is the first to employ this design while randomly assigning children to testify in either open court or via closed-circuit television. Community members were paid to act as mock jurors for the simulated trials.¹²⁶

Results showed that children provided equally incriminating testimony in the guilty and deception conditions and that children in both these conditions provided significantly more incriminating testimony than children in the not- guilty condition.¹²⁷ In general, children who were asked to lie in the deception condition implicated the defendant about as much as the children in the guilty condition.¹²⁸ Notably, children in the deception condition, regardless of whether they testified in open court or via closed-circuit television, did not differ in their implication of the defendant.¹²⁹ Therefore, children's lying was not facilitated by the use of protective measures.¹³⁰

In sum, the current research on whether shielding procedures facilitate lying indicates that children do not lie better when they are protected from a face-to-face confrontation with the alleged defendant. Given that the Orcutt study is the only study to address this issue, additional research is needed to replicate and extend its findings. In addition, conscience had no role to play for the children in the deception condition in the Orcutt study. Presumably, it is harder for a person to falsely accuse the defendant in a confrontational setting, because the defendant's presence triggers the accuser's conscience. Children in the deception condition of the Orcutt study were told that it was a "pretend trial."¹³¹ They were told to think of themselves as actors.¹³² They also knew that the "defendant" would not be in trouble because of anything they said.¹³³ These problems are understandable given the ethical constraints on this sort of study. Nevertheless, if shielding did not facilitate their false implication of the defendant, it could be because children in the deception condition were comfortable falsely implicating the defendant without being shielded.

F. Ability of Jurors to Detect Deception

In Coy v. Iowa, the Court suggested that jurors are better able to detect deception when witnesses testify unshielded in the courtroom: "It is always more [*pg 224] difficult to tell a lie about a person 'to his face' than 'behind his back.' In the former context, even if the lie is told, it will often be told less convincingly."¹³⁴ A criminal defendant has a legitimate interest in exposing false allegations. If jurors are better at detecting deception when the witness testifies unshielded in the courtroom, defendants will prefer children's courtroom testimony to children's shielded testimony. The fact of the matter is that jurors are not particularly adept at detecting deception in either case.¹³⁵

In the Goodman study, mock jurors assessed the veracity of children's testimony in a shielded condition -- via closed-circuit television -- or in an open court condition.¹³⁶ Using a correlational analysis, the researchers examined the relationship between children's overall accuracy and mock jurors' perceptions of the child's accuracy.¹³⁷ They found that mock jurors were best at determining accuracy when the six-year-olds testified in open court and when eight-year-olds testified via closed-circuit television.¹³⁸ Overall, mock jurors were not impaired in discerning accuracy due to the use of protective measures.¹³⁹ The mock jurors, however, were not very good at determining whether a statement was actually correct under either condition.¹⁴⁰

In the Orcutt study, mock jurors' ability to evaluate the honesty of children's testimony in the closed-circuit television condition and the open court condition was examined.¹⁴¹ The results showed that mock jurors' ability to assess deception was not impaired by the use of closed-circuit television.¹⁴² That is, the mock jurors' verdicts did not differ with regard to whether the child testified in open court or via closed-circuit television.¹⁴³ Thus, the concern that protective measures may hinder a juror's ability to detect deception may be unfounded.¹⁴⁴

III. CHILDREN'S HEARSAY

The article thus far has focused on issues relating to protecting a child when he or she testifies in court. It will now examine issues relating to the use of children's hearsay in court.

[*pg 225]

Most lay people understand hearsay to be one person's report about what another person has said. As a legal matter, however, hearsay is a statement originally made out of court and later offered in court to prove the truth of the matter asserted.¹⁴⁵ The person who makes the out-of-court statement is the hearsay declarant.¹⁴⁶ The person who recounts in court the hearsay declarant's out-of-court statement is the hearsay witness. Usually, but not always, the hearsay witness is not the same person as the hearsay declarant.

Hearsay is generally not admissible.¹⁴⁷ A witness ordinarily testifies in court only about past events perceived personally by that witness and does not repeat in court what someone else has said about an event. Hearsay that falls within an exception to the hearsay rule is admissible as an evidentiary matter.¹⁴⁸ Children's hearsay may be admissible pursuant to well-established exceptions to the hearsay rule, like the excited utterance exception,¹⁴⁹ and the exception for statements made for purposes of medical diagnosis or treatment.¹⁵⁰ Children's hearsay may also be admissible pursuant to special child hearsay statutes.¹⁵¹

Some hearsay that is admissible as an evidentiary matter may be inadmissible as a constitutional matter.¹⁵² When hearsay, including children's hearsay,¹⁵³ [*pg 226] falls within a firmly rooted exception to the hearsay rule,¹⁵⁴ the Confrontation Clause poses no obstacle to admissibility; the prosecution need not first produce the declarant or demonstrate the declarant's unavailability.¹⁵⁵ The rationale for this rule is that the context of these out-of-court statements imbues the statements with substantial guarantees of trustworthiness that cannot be recaptured by later in-court testimony.¹⁵⁶ In White v. Illinois, the United States Supreme Court explained the significance of a statement's context:

A statement that has been offered in a moment of excitement -- without the opportunity to reflect on the consequences of one's exclamation -- may justifiably carry more weight with a trier of fact than a similar statement offered in the relative calm of the courtroom. Similarly, a statement made in the course of procuring medical services, where the declarant knows that a false statement may cause misdiagnosis or mistreatment, carries special guarantees of credibility that a trier of fact may not think replicated by courtroom testimony.¹⁵⁷

When out-of-court statements do not fall within a firmly rooted hearsay exception, the statements are presumptively unreliable and inadmissible for Confrontation Clause purposes.¹⁵⁸ This presumption, however, is rebuttable if the prosecution can demonstrate the hearsay's trustworthiness.¹⁵⁹ The hearsay statement must be "at least as reliable as evidence admitted under a firmly rooted hearsay exception," and "similarly [must] be so trustworthy that adversarial testing would add little to its reliability."¹⁶⁰ The court making the trustworthiness determination, however, considers only the circumstances surrounding the making of the statement and that render the declarant particularly worthy of belief.¹⁶¹ It is not enough that evidence of unreliability is absent; affirmative evidence of reliability must be present.¹⁶² The fact that other evidence corroborates the hearsay statement is irrelevant to the inquiry.¹⁶³ Whether the declarant's production or unavailability is required to admit hearsay not falling within a firmly rooted hearsay exception may be an open question.¹⁶⁴

[*pg 227]

A child hearsay statute providing for the admissibility of otherwise inadmissible children's hearsay is a relatively recent legislative experiment and not a firmly rooted exception to the hearsay rule.¹⁶⁵ Thus, these statutes generally require the child to testify or be unavailable, in addition to a showing that the hearsay statement is reliable.¹⁶⁶

As with the use of shielding procedures when children are witnesses, the admissibility of children's hearsay is based on a host of assumptions. Some of these assumptions are explored below in light of the available social science evidence.

A. Child Witness Protection

Receiving children's out-of-court hearsay statements in lieu of live testimony protects children from confrontational stress. Indeed, if the child is not called as a witness, the child is not even subjected to cross-examination, as child witnesses are when shielding procedures are used. In theory, children's hearsay can be offered in lieu of children's testimony: With firmly rooted hearsay exceptions, the prosecutor is not obligated to produce the child witness. Moreover, with most child hearsay statutes, the prosecutor also need not produce the child witness if he or she can demonstrate the child's unavailability.

Despite this theoretical possibility of receiving children's hearsay in lieu of children's live testimony, the reality in United States criminal courts is different. The admissibility of children's hearsay does not protect child witnesses from confrontational stress because children's hearsay is typically admitted in addition to, rather than in lieu of, their courtroom testimony.¹⁶⁷ The recitation of facts in appellate cases reviewing the admission of children's hearsay often indicates that the child victim testified.¹⁶⁸ Empirical research regarding child sexual [*pg 228] abuse prosecutions also indicates that children's hearsay is admitted in addition to the child's courtroom testimony.¹⁶⁹

A study by John E.B. Myers and his colleagues (the "Myers study") involved the collection of data from forty-two jury trials in criminal courts in Sacramento, California, and Phoenix, Arizona.¹⁷⁰ In each of the trials, at least one child testified live in court and at least one adult reported child hearsay.¹⁷¹ The production of these child witnesses in court means that the child witnesses physically confronted the defendant, which is a documented source of emotional trauma for child witnesses.

The Myers study does not indicate whether the children's hearsay was admitted pursuant to firmly rooted hearsay exceptions or special child hearsay statutes. Nor does the study indicate whether prosecutors attempted but failed to convince the trial court of the child's unavailability to proceed under a special child hearsay statute. If prosecutors produce the child witness in addition to offering the child's hearsay pursuant to a firmly rooted hearsay exception, or offer hearsay evidence pursuant to special child hearsay statutes and call the child witness without attempting to demonstrate the child's unavailability, it could indicate the prosecutor's belief that the child does not require protection from confrontation.¹⁷² It could also indicate a belief that they need the child's [*pg 229] live testimony to prevail and are willing to sacrifice child protection to that end.¹⁷³

B. Defendant Prejudice

A survey of state trial court judges asked about their perceptions of the fairness of child witness innovations. The judges expressed that certain innovations, such as allowing hearsay evidence, were unfair.¹⁷⁴ Nevertheless, they admitted such hearsay evidence in their own courts.¹⁷⁵

Children's hearsay may be unfairly overused by prosecutors. A review of the appellate cases discussing the admission of hearsay in child abuse prosecutions indicates that triers of fact are hearing from multiple hearsay witnesses.¹⁷⁶ The Myers study provides additional evidence that prosecutors call multiple hearsay witnesses in child abuse cases. Data from forty-two trials indicated the following: One adult hearsay witness testified in one trial; two adult hearsay witnesses testified in twenty-one trials; three adult hearsay witnesses testified in thirteen trials; four adult hearsay witnesses testified in six trials; and five adult hearsay witnesses testified in one trial.¹⁷⁷

At some point, enough is enough from a due process perspective. In Felix v. State, the Nevada Supreme Court declared that "the unlimited admission of repetitive hearsay testimony can jeopardize the fundamental fairness of the entire trial."¹⁷⁸ Indeed, the Felix court waxed philosophical when it asserted:

[*pg 230]

The repetition of multiple child hearsay accusations through several adult witnesses presents the very real possibility of placing the person accused of [child sexual abuse] at an unfair disadvantage. Although some may opine that this is what such an offender deserves, we must remember that a trial is a procedure to determine whether the accused is the offender and not simply a process to confirm allegations made by the State. No person, neither the alleged victim nor the accused, should be placed at a substantial disadvantage in a criminal trial by the rules of procedure and evidence. We believe that the appellants were disadvantaged in this case by the unbridled introduction into evidence of repetitive hearsay allegations . . . when much of this evidence was merely cumulative and unnecessary.¹⁷⁹

In Felix, the child testified in court. Her out-of-court statements were then recounted by five adult hearsay witnesses: her mother, two psychologists, a pediatrician, and a detective.¹⁸⁰ The court concluded that hearsay allegations should be restricted "once the child's [sexual abuse] accusations have been fairly presented . . . and any challenges to the victim's credibility are fairly met."¹⁸¹ The court held that the limitation on the admission of multiple hearsay was reasonable because it amounted to the admission of unnecessary prior consistent statements,¹⁸² it amounted to vouching for the child victim's credibility, and it was more prejudicial than probative.¹⁸³

A few child hearsay statutes limit the number of hearsay statements that are admissible. The Michigan statute limits the proponent of children's hearsay to the child declarant's first statement.¹⁸⁴ The Texas statute similarly admits only the child's statements "to the first person, 18 years of age or older."¹⁸⁵ The Massachusetts statute may indirectly prevent the admission of multiple hearsay statements through its requirement that the statement be "more probative on the point for which it is offered than any other evidence which the proponent can procure through reasonable efforts."¹⁸⁶ Presumably, a repetitive hearsay statement would not be more probative.

Social scientists have begun to explore the effects of using hearsay testimony, combining the child victim's live testimony with hearsay testimony, and admitting the testimony of multiple hearsay witnesses. These studies represent a first step toward understanding how hearsay testimony can impact a trial.

For example, a study by David Ross, Rod Lindsay, and Dorothy Marsil examined the impact of hearsay testimony on conviction rates in child sexual [*pg 231] abuse cases.¹⁸⁷ In Experiment 1, mock jurors watched a realistic videotape of a trial in a courtroom setting.¹⁸⁸ Half of the mock jurors saw the child victim testify in open court while the other half saw the child's mother, the hearsay witness, testify on behalf of the child victim.¹⁸⁹ In this experiment, the defendant was convicted significantly more often when the child testified instead of the hearsay witness.¹⁹⁰ In Experiment 2, mock jurors read a trial summary of a child sexual abuse case.¹⁹¹ The type of testimony varied and was either given by the child or a hearsay witness.¹⁹² The type of hearsay witness was also varied, with the witness taking the role of the child's mother, doctor, teacher, or neighbor.¹⁹³ Results showed that each hearsay condition, except the neighbor, produced more guilty verdicts than when the child testified.¹⁹⁴

Although discrepant results in conviction rates were found between Experiment 1 and Experiment 2, this might be explained by the different stimulus or the relationship between the victim and the hearsay witness.¹⁹⁵ First, Experiment 1 used a highly realistic videotape of a trial versus a trial summary.¹⁹⁶ Second, in Experiment 1, the hearsay witness was the child's mother, the defendant was the child's father, and the parents were divorced.¹⁹⁷ By contrast, in Experiment 2, the defendant was the child's neighbor.¹⁹⁸ Presumably, the mother in Experiment 1 may have a motive to lie against the father, whereas in Experiment 2, there is a less clear motive for the hearsay witness to lie.

Research by Anne Tubb and colleagues analyzed the impact of hearsay testimony on conviction rates using a written child sexual abuse trial summary.¹⁹⁹ Mock jurors were randomly assigned to one of three testimony conditions in which key testimony was provided.²⁰⁰ In the first condition, the child testified; in the second, a police officer testified as a hearsay witness in lieu of the child; the third condition varied from the second condition only in that the police officer also gave his positive opinion of the child's credibility.²⁰¹ The results showed that there was no difference in the verdict for any of these conditions.²⁰² Thus, [*pg 232] mock jurors were neither more nor less likely to convict when the child testified directly than when the hearsay witness testified, regardless of the credibility opinion.²⁰³

A study by Golding and colleagues (the "Golding study") examined the use of hearsay in a child sexual assault trial where mock jurors read a trial summary and then completed questionnaires.²⁰⁴ In Experiment 1, the victim alleged that her uncle had fondled her.²⁰⁵ The primary variable in this experiment was the condition under which the testimony was presented.²⁰⁶ In the first condition, the critical testimony was given by the child victim who was either six or fourteen years old.²⁰⁷ In the second condition, the critical testimony was given by the hearsay witness, either the child's mother or teacher.²⁰⁸ In the third condition, both the victim and the hearsay witness testified.²⁰⁹ In the fourth condition, no critical testimony was given; the only witness to testify for the prosecution was a cousin who said that the child had spent the night at the uncle's house with her.²¹⁰ The Golding study was the first to use a trial summary stimulus to examine the use of hearsay in a child sexual assault trial and the first to combine the testimony given by a child victim with testimony given by a hearsay witness. No significant differences in guilty verdicts were found among the following situations: when the child testified, when the child and hearsay witness testified, or when the hearsay witness testified. In fact, guilty verdicts were equal in the child victim and the hearsay witness situations.²¹¹ The only significant differences occurred when the cousin testified -- there were significantly fewer guilty verdicts than in the other conditions tested.²¹² The age of the victim did not influence the verdict.²¹³

Experiment 2 used a similar method but included a four-year-old victim and varied the sex of the hearsay witness, in this case a teacher.²¹⁴ In this experiment, the testimony conditions were the same as in Experiment 1, except that the condition in which both the victim and the hearsay witness testified was excluded.²¹⁵ Results showed that, in general, there was no difference among testimony conditions with regard to guilty verdicts except, again, there were significantly less guilty verdicts for the condition in which only the cousin testified.²¹⁶

[*pg 233]

In this experiment, however, the age of the victim had a small impact on guilty verdicts when the victim was four years old and a hearsay witness testified.²¹⁷ When the victim was four years old, ninety-five percent of the verdicts in the hearsay condition were guilty verdicts, compared to eighty-five percent when the victim was six years old, and seventy-six percent when the victim was fourteen years old.²¹⁸ Guilty verdicts did not differ by the victim's age when the child testified -- about eighty-five percent of the verdicts were returned guilty.²¹⁹ Interestingly, a hearsay witness may actually help get a conviction when the child victim is very young and unwilling or unable to testify.²²⁰

Finally, a second study by Golding and colleagues examined the effect of multiple hearsay witnesses on verdicts in a child sexual assault case.²²¹ There were three conditions varying the way in which the critical testimony for the prosecution was presented: (1) by the child victim and expert witness; (2) by a hearsay witness -- the child's sister -- and expert witness; or (3) only by the expert witness.²²² The results show that when multiple hearsay witnesses, in this case the victim's sister and an expert clinical psychologist, testified in lieu of the victim, there were significantly more convictions than when the critical testimony was given by the victim or by the expert alone.²²³

Taken together, these studies report dramatically conflicting results. One study reports higher conviction rates when the child versus a hearsay witness testifies;²²⁴ others report no difference.²²⁵ Another study reports higher conviction rates when hearsay witnesses rather than the child witness testify.²²⁶ Ross and his colleagues posed a credibility inflation/deflation model to explain these conflicting results.²²⁷ The underlying idea of the model is that the hearsay witness's perceived credibility is critical to the impact that his or her testimony will have on the trial.²²⁸ For example, in the Ross study described above, either the child testified or the child's mother testified on her behalf.²²⁹ The mother was also involved in a heated divorce with the child's father who was the defendant in the child sexual abuse case.²³⁰ Thus, in this situation the mother had a strong vested interest in the case, and jurors were reluctant to believe her testimony.²³¹ [*pg 234] Jurors were more likely to believe the child testifying on her own behalf than the mother testifying for her.²³² Therefore, the critical issue in interpreting these studies involves the effect of the context, and whether it increases or decreases the perceived credibility of the hearsay witness.

On a separate note, given the prevalence of admitting the testimony of multiple hearsay witnesses in child sexual abuse trials, more information is needed on the impact of such evidence. The initial research by Golding is striking and suggests that the admission of multiple hearsay witnesses may be unfair to defendants.²³³ Therefore, this important issue should be further addressed by social scientists through empirical research.

C. Reliability of Children's Hearsay Evidence

The fundamental premise of cases upholding the admission of hearsay evidence in the face of the Confrontation Clause is that the hearsay evidence is reliable, and statement context is considered critical to hearsay's reliability.²³⁴ In Idaho v. Wright, where children's hearsay did not fall within a firmly rooted hearsay exception, the United States Supreme Court provided a nonexclusive list of factors relevant to the requisite reliability determination: the child's spontaneity and consistent repetition of the allegations, the child's mental state when the allegations are made, the child's use of terminology unexpected of a child of similar age, and the child's lack of motive to fabricate.²³⁵ The Court also emphasized the suggestive manner in which the child may have been interviewed, and clarified that spontaneity can be an inaccurate indicator of trustworthiness when evidence exists of prior interrogation, prompting, or manipulation by adults.²³⁶ Although the Court emphasized the importance of the manner in which a child witness is interviewed,²³⁷ the Court declined to require the taping of child witness interviews as a prerequisite to admissibility under the Con- [*pg 235] frontation Clause.²³⁸ Nevertheless, the failure to record these interviews leaves the trial court at the mercy of the interviewer to reconstruct the manner in which the interview was conducted.

Social science has much to tell us about how child witnesses should be interviewed and whether interviewers accurately remember what children have told them and what questions they asked the children during the interview. The bulk of this research supports the conclusion that there is a deterioration in the quality of some, but not all, aspects of memory for the event experienced by the child as it passes down the hearsay chain from the child to the hearsay witness, and subsequently from the hearsay witness to the jury. Another issue involves the ability of the hearsay witness, often a professional who has interviewed the child, to recall accurately the nature of the interview and the specific questions asked. This issue has been examined in several large-scale studies.

For example, Bruck and her colleagues had young children play with a confederate, and several days later the mothers interviewed their children about the event.²³⁹ Half of the mothers were warned that the study involved examining memory, and the other half were not.²⁴⁰ Three days after interviewing their children, the mothers were questioned regarding their recall for what the child experienced at the preschool and aspects of the interview that they conducted with their children.²⁴¹

The results indicated that, while the mothers could recall approximately sixty-six percent of the primary events the child experienced, they had considerable difficulty with other aspects.²⁴² For example, mothers only recalled approximately thirty-five percent of the details that their children had given them, and they had serious difficulty reporting many of the specifics of the interview. Their memory was particularly lacking with respect to what statements were spontaneous or prompted and whether specific utterances were spoken by the child or the mother.²⁴³ Finally, the warning provided to the mothers had no impact on their memory.²⁴⁴

The finding that hearsay witness memory may be problematic with mothers can be generalized to professional interviewers as well. Warren and her colleagues had preschool children interviewed by professional interviewers concerning staged events that occurred in their preschool.²⁴⁵ After the child interview, the professional interviewers were questioned by an experimenter about [*pg 236] their interview with the child.²⁴⁶ Among other things, the professionals were asked to recall the major events the child experienced, the specific questions asked during the interview, and the child's exact responses.²⁴⁷ The professionals could recall the majority, eighty percent, of the major event activities, but they were very poor at recalling the characteristics of the interview, such as the number and content of the specific questions asked.²⁴⁸

Michael Lamb and his colleagues also studied the accuracy of professional interviewers of children.²⁴⁹ In their study, the researchers examined Israeli youth investigators' interviews with alleged child sexual abuse victims by comparing their "verbatim" written notes to an audio recording of the same interview.²⁵⁰ First, the results showed that interviewers recorded significantly fewer details than the child actually provided in the audiotaped interview.²⁵¹ Specifically, interviewers failed to report about twenty-five percent of the relevant details and about eighteen percent of the central details given by the child.²⁵² The interviewers, however, almost never reported incorrect or false information (also known as "errors of commission").²⁵³

Second, the interviewers were not very good at recalling what they had said during the interview.²⁵⁴ That is, they failed to record over half of the utterances they made to the child.²⁵⁵ In addition, the interviewers failed to record accurately when their utterances were meant to elicit specific information from the child rather than open-ended questions.²⁵⁶ In fact, they only recorded correctly about forty-four percent of their more focused questions.²⁵⁷ Thus, the interviewers reported information that frequently did not accurately represent the interview with the child and did not report some suggestive questioning included in the interview.²⁵⁸ These reports include the type of information that could be presented in court and could impair a jury's evaluation of the child's account as reported by the hearsay witness.

Finally, in a study by Maithilee Pathak and William Thompson, the researchers examined the accuracy of an adult hearsay witness reporting an observed interview with a child and then how mock jurors evaluated that testi- [*pg 237] mony.²⁵⁹ During the experiment, children individually watched a janitor clean or play with toys and after a one-hour delay were interviewed in one of two conditions: neutral or suggestive.²⁶⁰ Each child was videotaped during the interview.²⁶¹ Notably, the children in the neutral condition accurately reported the janitor's actions, but the children in the suggestive condition were significantly influenced by the questioning and inaccurately reported the janitor's actions.²⁶²

Later, adult participants were shown a videotape of the child's interview.²⁶³ Before watching the videotape, they were told that the child had either watched the janitor clean the toys or that the janitor had attempted to engage the child by playing with the toys.²⁶⁴ The participant, while being videotaped, reported what was said by the child during a structured interview.²⁶⁵

The results showed that the adult participants could recognize when suggestive questions were being used, and that they judged those questions to be more influential on the child's report than when children were questioned in the neutral condition, regardless of the janitor's actual actions or what the participant was told the janitor did.²⁶⁶ Those participants were also asked to decide what the janitor did based on the child's report.²⁶⁷ Participants in the neutral condition were better able to determine the janitor's actual actions than those in the suggestive condition.²⁶⁸

In the second experiment, participants from universities and the community watched a videotape of the structured interview with the first experiment's adult participants reporting what the child had said.²⁶⁹ Based on the hearsay witness, the participants in the second experiment were asked to determine whether the child's interview had been suggestive and what the janitor's actions had been.²⁷⁰ The results showed that these participants were not able to determine the influence of suggestive questioning on the child's report.²⁷¹ The participants were better able to discern the actions of the janitor, however, when the child had been questioned in the neutral condition rather than the suggestive condition.²⁷² These results demonstrate that the type of interview, suggestive or neutral, impacts the perception of the hearsay witness and the subsequent evaluation of the hearsay witness's testimony by mock jurors.

[*pg 238]

The results of these studies should be of great concern to the courts. These studies demonstrate that adults, specifically mothers²⁷³ and professional interviewers,²⁷⁴ are not very good at recalling the details of their conversations with children, although the professional interviewers²⁷⁵ were more accurate than the mothers.²⁷⁶ Of greatest concern is the finding that both groups did not recall very well how they learned information -- whether the child spontaneously told them the information or whether it was the result of a specific question.²⁷⁷ In addition, one of the studies showed that even when mock jurors realized that an interview with a child was suggestive, they did not place sufficient weight on how that type of questioning had impacted the child's responses.²⁷⁸ Therefore, given the current research findings, the testimony of hearsay witnesses should, at the very least, be viewed with some skepticism.²⁷⁹

IV. CONCLUSION

Current normative legal standards for determining whether to shield testifying children attempt to resolve the tension between a defendant's constitutional rights under the Confrontation Clause and the state's interest in protecting children from significant harm. The Supreme Court, in Maryland v. Craig,²⁸⁰ treated those two interests as if they were poised in inevitable opposition. One critical issue will be whether social science research causes courts to view the interests at stake in new ways, ultimately permitting the development of rules that adequately protect defendants' interests, while also better protecting children.²⁸¹

Research conducted on shielded testimony makes several important points. First, children undergo significant stress when they are required to testify in court and in the presence of the defendant.²⁸² More important, children's stress under these circumstances leads to incomplete and potentially inaccurate testimony.²⁸³ Finally, some studies examining the reliability of shielded testimony [*pg 239] indicate that shielded testimony may not be inherently less reliable, if reliability is defined in terms of accuracy and honesty.²⁸⁴

Both prosecutors and defense attorneys will want to pay attention to findings that continue to support the notion that a child's fear of a defendant can have a significant impact on the child's ability to testify in court. For prosecutors, this information corroborates and strengthens the basis for Craig, namely the child's fear of a particular defendant. Given the need to show psychological trauma to the individual child, prosecutors would do well not to rely on generalized findings, but rather to provide expert testimony on the child's mental state and the impact of trial testimony on the particular child.²⁸⁵ Defense attorneys will be concerned that social science findings connecting children's trauma to the defendant's presence might form the basis for leading questions, judicial assumptions, or conclusory procedures to be used when a child is interviewed as a predicate to shielding.²⁸⁶ They will also be concerned that the determinative process itself not be used to create implications of heinous behavior by a particular defendant or to bestow any aura of special sympathy upon the alleged victim.

One disturbing conclusion following from these studies may give pause to those who consider shielding as the answer to the problem of children's testimony in court. Social science research indicates that children's shielded testimony is not equivalent to children's in-court testimony,²⁸⁷ but the difference may not point in the expected direction. Several studies show a possible negative side to the shielding procedure.²⁸⁸ While shielding appears not to affect the ver- [*pg 240] dict in most mock trial studies,²⁸⁹ shielding may have an important impact on the jurors' perception of the child witness. Shielding a child witness from the rigors of confrontation may permit jurors to discount the child more easily or to take the child's story less seriously.

Current research shows only that juror perceptions of shielded children were not as favorable as the same juror perceptions of children who testified in court.²⁹⁰ If this research is extended, it may prove important to child advocates and prosecutors, as well as to defense attorneys.²⁹¹ The legal system may need to recognize that, in the long run, using protective measures such as shielding or hearsay may not be as effective as more time-consuming and expensive methods like court preparation techniques and court schools, which have the potential to empower some child witnesses.²⁹² Social science research can help by making further and more detailed comparisons between different kinds of child protective methods.

Hearsay evidence regarding children's out-of-court statements presents even more complex problems than shielded testimony. Studies of current uses of child-based exceptions to the hearsay rule indicate that these exceptions do not always provide protection for children.²⁹³ Children's hearsay often appears to be admitted as additional testimony rather than in lieu of the particular child's testimony. At least one study indicates that a conviction for child sexual assault may be more likely when a hearsay witness testifies than when the child victim testifies.²⁹⁴ Other studies show different results, with conviction rates increasing with child testimony.²⁹⁵

Studies on the reliability of hearsay evidence also raise concerns. One of the primary defense concerns about allegations of sexual misconduct, particularly when there is no physical corroborating evidence, is whether the child is responding to an adult's suggestion of sexual wrongdoing by the alleged perpetrator or is truly relaying his own experience. Thus, the investigative technique used to elicit the child's response is important. Defense attorneys will certainly be concerned to know that mothers in recall studies had difficulty remembering [*pg 241] if they asked a question that elicited a response from the child or if the child gave the information spontaneously.²⁹⁶ These concerns may be heightened by findings that jurors are sometimes unable to recognize the impact of suggestive interviewing on the testimony of children.²⁹⁷ These concerns make the reliability area important for future research.

Finally, all of this evidence demonstrates one central feature that should concern both social scientists and lawyers: the tendency of adults to place the most vulnerable child witnesses in a double bind. The youngest, and therefore the most vulnerable, witnesses in terms of anxiety may be the most likely to find shelter under Craig. Having been shielded, however, those children are viewed less favorably by adults than older shielded children. These children were also viewed less favorably than children who testified in court.²⁹⁸ In the instances of child hearsay, a very young child victim's claim of abuse may be more credible to a jury when it is relayed by a hearsay witness.²⁹⁹ Both social scientists and the courts have a strong interest in protecting children from undue harm, but all adults should recognize that this protection appears to carry a significant negative impact, not only for the children involved, but for the society that values them. At the present time, social science has helped to uncover this duality. Once uncovered, it can be examined by both disciplines to produce policies that fairly protect and empower children while also respecting the values inherent in the constitutional rights of defendants.

FOOTNOTES

*	Ph.D. Candidate, University of Kentucky. M.S. 1999, University of Tennessee at Chattanooga.
**	Professor of Law, University of San Diego School of Law. J.D. 1985, U.C.L.A.
***	Associate Professor of Psychology, University of Tennessee at Chattanooga. Ph.D. 1989, Cornell University.
****	Wendell H. Ford Professor of Law, University of Kentucky. J.D. 1977, University of Texas.
1.	Child witness innovations proliferated in other countries as well. See, e.g., Nicholas Bala, Child Witnesses in the Canadian Criminal Courts: Recognizing Their Capacities and Needs, 5 PSYCHOL. PUB. POL'Y & L. 323 (1999) (discussing the law in Canada); Helen L. Westcott et al., Children, Hearsay, and the Courts: A Perspective from the United Kingdom, 5 PSYCHOL. PUB. POL'Y & L. 282 (1999) (discussing the law in England and Wales).
2.	Using a screen to shield a child witness from the defendant is not as popular as the more high-tech procedure of using closed-circuit television. Statutes expressly providing for use of a screen to shield a child witness from the defendant are uncommon. See ALASKA STAT. § 12.45.046 (Michie 1998) (providing for the use of one-way mirrors or closed-circuit television); MICH. COMP. LAWS ANN. § 600.2163a(12) (West 2000) (providing for rearranging the courtroom and positioning the defendant "so that the defendant is as far from the witness stand as is reasonable and not directly in front of the witness stand"). But see CONN. GEN. STAT. § 54-86g(a) (1994) (allowing the defendant to be "screened from the sight and hearing of the child"). Some cases, however, record instances of trial courts permitting the use of a screen. E.g., State v. Welch, 760 So. 2d 317, 319 n.1 (La. 2000) (defendant ordered to sit behind clear glass wall nine to ten feet behind defense counsel's table, where paper was taped to the glass to block the child's view of the defendant); State v. Murphy, 542 So. 2d 1373, 1374 (La. 1989) (defendant present in the same room as the child witness during child's testimony by closed-circuit television, but defendant ordered to sit behind opaque screen that shielded him from child's view); State v. Davis, 830 P.2d 1309, 1312 (Mont. 1992) (defendant present in courtroom during children's testimony, but a "free-standing hinged space partition, apparently intended for use as a temporary room divider" was placed between the witness stand and the counsel table where the defendant sat); State v. Thomas, 442 N.W.2d 10, 17 (Wis. 1989) (defendant present in the same room as child witness during the recording of child's videotaped deposition, but screen was placed between defendant and child to block child's view of defendant).
3.	E.g., 18 U.S.C. 3509 (1994 & Supp. 1999) (allowing two-way closed-circuit television); ALASKA STAT. § 12.45.046 (Michie 1998); ARIZ. REV. STAT. § 13-4253 (1989); ARK. CODE ANN. § 16-43-1001 (Michie 1999); CAL. PENAL CODE § 1347 (West Supp. 2000); COLO. REV. STAT. § 18-3-413.5 (1999); CONN. GEN. STAT. § 54-86g (1994); DEL. CODE ANN. tit. 11, § 3514 (1995); FLA. STAT. ch. 92.54 (1999); GA. CODE ANN. § 17-8-55 (Harrison 1998) (allowing two-way closed-circuit television); HAW. REV. STAT. § 626-616 (1993) (same); IDAHO CODE § 19-3024A (Michie 1997) (same); 725 ILL. COMP. STAT. ANN. 5/106B-5 (West Supp. 2000); IND. CODE ANN. § 35-37-4-8 (West 1998) (same); KAN. STAT. ANN. § 22-3434 (1995); KY. REV. STAT. ANN. § 421.350 (Banks-Baldwin 1999); LA. REV. STAT. ANN. § 15:283 (West 1992); MD. ANN. CODE art. 27, § 774 (1996 & Supp. 2000) (same); MASS. GEN. LAWS ANN. ch. 278, § 16D (West 1998) (subsequently held unconstitutional for violating the defendant's state constitutional right to a "face-to-face" meeting with witnesses under Commonwealth v. Bergstrom, 524 N.E.2d 366 (Mass. 1988)); MINN. STAT. ANN. § 595.02(4) (West 2000); MISS. CODE ANN. § 13-1-405 (Supp. 2000); N.J. STAT. ANN. § 2A:84A-32.4 (West 1994); N.Y. CRIM. PROC. LAW § 65.30 (McKinney 1992) (allowing two-way closed-circuit television); OHIO REV. CODE ANN. § 2945.481 (West Supp. 2000) (same); OKLA. STAT. ANN. tit. 22, § 753 (West 1992); R.I. GEN. LAWS § 11-37-13.2 (1994); TENN. CODE ANN. § 24-7-120 (Supp. 1999) (allowing two-way closed-circuit television); TEX. CRIM. PROC. CODE ANN. § 38.071 (sec. 3) (West Supp. 2000); VA. CODE ANN. § 18.2-67.9 (Michie 1996 & Supp. 2000) (allowing two-way closed-circuit television); WASH. REV. CODE ANN. § 9A.44.150 (West 2000); UTAH R. CRIM. P. 15.5; VT. R. EVID. 807 (allowing two-way closed-circuit television, but permitting the court to proscribe transmitting the defendant's image to the child witness).
4.	E.g., ALA. CODE §§ 15-25-31, 15-25-32, 15-25-37 (1995); ALASKA STAT. § 12.40.110 (Michie 2000) (making child hearsay admissible before the grand jury); CAL. EVID. CODE § 1360 (West Supp. 2000); COLO. REV. STAT. ANN. § 13-25-129 (West 1997); DEL. CODE ANN. tit. 11, § 3513 (1995 & Supp. 1999); FLA. STAT. ANN. § 90.803(23) (West 1999); GA. CODE ANN. § 24-3-16 (Harrison 1998) (providing, in 1995 amendment, for the admissibility of child witness statements; held unconstitutional in Woodard v. State, 496 S.E.2d 896 (Ga. 1998)); HAW. REV. STAT. § 804(b)(6) (1993); 725 ILL. COMP. STAT. ANN. 5/115-10 (West 1992 & Supp. 1998); IND. CODE ANN. § 35-37-4-6 (Michie 1998); KAN. STAT. ANN. § 60-460(dd) (1994 & Supp. 1999); LA. CODE EVID. ANN. art. 804(B)(5) (West 1995); MD. ANN. CODE art. 27, § 775 (1996 & Supp. 2000); MASS. GEN. LAWS ANN. ch. 233, § 81 (West 2000); MINN. STAT. ANN. § 595.02(3) (West 2000); MO. ANN. STAT. § 491-075 (West 1996); NEV. REV. STAT. ANN. § 51.385 (Michie 1996); OKLA. STAT. ANN. tit. 12, § 2803.1 (West Supp. 2001); OR. REV. STAT. § 40.460(18b) (1988 & Supp. 1998); 42 PA. CONS. STAT. § 5985.1 (2000); S.D. CODIFIED LAWS § 19-16-38 (Michie 1995); TEX. CRIM. PROC. CODE ANN. § 38.072 (West Supp. 2000); UTAH CODE ANN. § 76-5-411 (1999); WASH. REV. CODE ANN. § 9A.44.120 (West 2000); ARK. R. EVID. 803(25), 804(b)(7); MICH. R. EVID. 803A; N.J. R. EVID. 803©(27); N.D. R. EVID. 803(24); OHIO R. EVID. 807; VT. R. EVID. 804a.
5.	E.g., KAN. STAT. ANN. § 22-3433 (1995); LA. REV. STAT. ANN. § 15:440.5 (West 1992); MINN. STAT. ANN. § 595.02(3) (West 2000) (including video, audio, or other recorded statements as permissible testimony); MO. ANN. STAT. § 492.304 (West 1996); OKLA. STAT. ANN. tit. 10, § 7003-4.2 (West 1998); TEX. CRIM. PROC. CODE ANN. arts. 38.071(2) & (5) (West Supp. 1999); VA. CODE ANN. § 63.1-248.13:3 (Michie 1988); WIS. STAT. ANN. § 908.08 (West 2000); UTAH R. CRIM. P. 15.5. But see ARIZ. REV. STAT. ANN. § 13-4252 (West 1989) (allowing videotaped testimony, but later held unconstitutional for conflicting with the Arizona Supreme Court's rulemaking authority in State v. Taylor, 2 P.3d 674 (Ariz. Ct. App. 1999)). For a critique of these statutes, see Jean Montoya, Something Not So Funny Happened on the Way to Conviction: The Pretrial Interrogation of Child Witnesses, 35 ARIZ. L. REV. 927, 940-50 (1993) (discussing the ineffectiveness of these statutes in protecting either the fact-finding process or the child witness from interviewing abuses).
6.	The Confrontation Clause provides that in "criminal prosecutions, the accused shall enjoy the right . . . to be confronted with the witnesses against him." U.S. CONST. amend. VI. It was apparently included in the Bill of Rights without debate. See California v. Green, 399 U.S. 149, 175-76 (1970) (Harlan, J., concurring). The Confrontation Clause applies to the states through the Fourteenth Amendment. Pointer v. Texas, 380 U.S. 400, 403 (1965).
7.	When two-way, as opposed to one-way, closed-circuit television is used, the defendant's image is also transmitted to the location where the child is testifying.
8.	487 U.S. 1012 (1988). Although physical confrontation was distinctly at issue in the earlier case of Kentucky v. Stincer, 482 U.S. 730 (1987), the Court sidestepped the issue there because physical confrontation was denied only for purposes of the children's competency hearing and was otherwise enjoyed during the trial. Id. at 740-44.
9.	Coy, 487 U.S. at 1014 (citing IOWA CODE ANN. § 910A.14 (West 1987)).
10.	Id.
11.	Id. at 1022.
12.	Id. at 1021 (quoting California v. Green, 399 U.S. 149, 175 (1970) (Harlan, J., concurring)) (emphasis in original).
13.	Id.
14.	Id.
15.	Id.
16.	497 U.S. 836 (1990).
17.	Id. at 842.
18.	Id. at 841 (citing MD. CODE ANN., CTS. & JUD. PROC. § 9-102(a)(1)(ii) (1989)).
19.	For a discussion of expert witness testimony and the necessity to shield a particular child witness, see Jean Montoya, Lessons from Akiki and Michaels on Shielding Child Witnesses, 1 PSYCHOL. PUB. POL'Y & L. 340, 342-43, 356-66 (1995) (arguing that trial court reliance on the testimony of the child's therapist is misplaced and that trial courts should personally examine the child witness).
20.	Craig, 497 U.S. at 842-43.
21.	Id. at 841.
22.	Id.
23.	Id. at 855.
24.	Id. at 845 (quoting Coy v. Iowa, 487 U.S. 1012, 1021 (1988)). The Court's holding resembles that of Globe Newspaper Co. v. Superior Court, 457 U.S. 596 (1982). In Globe, the Court acknowledged the state's compelling interest in protecting minor victims of sex crimes from the further trauma and embarrassment of testifying in open court, but held that a Massachusetts statute barring press and public access to criminal sex-offense trials during the testimony of minor victims violated the First Amendment because the statute mandated uniform closure rather than selective closure based on a case-by-case determination of the particular child's needs. Id. at 607-08, 610-11 & n.27.
25.	Craig, 497 U.S. at 844-50.
26.	Id. at 853.
27.	Id. at 861 (Scalia, J., dissenting).
28.	487 U.S. 1012 (1988) (holding that placement of a screen between the defendant and child witnesses violated the defendant's Confrontation Clause rights).
29.	Id. at 1032 (Blackmun, J., dissenting).
30.	497 U.S. 836, 855 (1990) (holding that transmitting children's testimony by closed-circuit television did not violate the defendant's Confrontation Clause rights when the prosecution demonstrated a need for shielding a particular child witness).
31.	See generally Louise Dezwirek-Sas, Empowering Child Witnesses for Sexual Abuse Prosecution, in CHILDREN AS WITNESSES 194 (Helen Dent & Rhona Flin eds., 1992); Rhona H. Flin, Hearing and Testing Children's Evidence, in CHILD VICTIMS, CHILD WITNESSES 289 (Gail S. Goodman & Bette L. Bottoms eds., 1993); Gail S. Goodman et al., Testifying in Criminal Court: Emotional Effects on Child Sexual Assault Victims, in 5 MONOGRAPHS OF THE SOCIETY FOR RESEARCH IN CHILD DEVELOPMENT 100 (1992).
32.	Id.
33.	Goodman et al., supra note 31, at 101.
34.	Dezwirek-Sas, supra note 31, at 187.
35.	487 U.S. 1012, 1032 (1988) (Blackmun, J., dissenting).
36.	Douglas P. Peters, The Influence of Stress and Arousal on the Child Witness, in THE SUGGESTIBILITY OF CHILDREN'S RECOLLECTIONS 60, 75 (John Doris ed., 1991).
37.	Id. at 68.
38.	Id.
39.	Id. at 69.
40.	Id.
41.	Id.
42.	Kay Bussey et al., Lies and Secrets: Implications for Children's Reporting of Sexual Abuse, in CHILD VICTIMS, CHILD WITNESSES 162 (Gail S. Goodman & Bette L. Bottoms eds., 1993).
43.	Id.
44.	Id.
45.	Id.
46.	Id.
47.	487 U.S. 1012 (1988).
48.	Id. at 1014.
49.	Id. at 1027 (Blackmun, J., dissenting). After certain lighting adjustments in the courtroom, the defendant was able to "dimly" perceive the witnesses. Id. at 1015.
50.	Id.
51.	Id. at 1015.
52.	Id.
53.	Id. at 1022.
54.	Id. at 1035 (Blackmun, J., dissenting).
55.	See Gail S. Goodman et al., Face-to-Face Confrontation: Effects of Closed-Circuit Technology on Children's Eyewitness Testimony and Jurors' Decisions, 22 LAW & HUM. BEHAV. 165, 190-91 (1998); David F. Ross et al., The Impact of Protective Shields and Videotape Testimony on Conviction Rates in a Simulated Trial of Child Sexual Abuse, 18 LAW & HUM. BEHAV. 553, 554 (1994); Janet K. Swim et al., Videotaped Versus In-Court Witness Testimony: Does Protecting the Child Witness Jeopardize Due Process?, 23 J. APPLIED SOC. PSYCHOL. 603, 606 (1993).
56.	Id.
57.	Swim et al., supra note 55, at 606.
58.	Id. at 607-08.
59.	Id. at 620.
60.	Id.
61.	Id.
62.	Id.
63.	Ross et al., supra note 55, at 553.
64.	Id. at 556-57.
65.	Id. at 558.
66.	Id.
67.	Id. at 561.
68.	Id. at 563.
69.	Goodman et al., supra note 55, at 176-77.
70.	Id. at 177.
71.	Id.
72.	Id. at 178.
73.	Id. at 177.
74.	Id. at 198.
75.	487 U.S. 1012, 1019 (1988).
76.	497 U.S. 836, 847 (1990).
77.	Id. at 852.
78.	Rod C.L. Lindsay et al., What's Fair When a Child Testifies?, 25 J. APPLIED SOC. PSYCHOL. 870, 873 (1995).
79.	Id. at 874-75.
80.	Id. at 875.
81.	Id. at 878.
82.	Id. at 883.
83.	Id. at 878.
84.	Id. at 879.
85.	Ross et al., supra note 55, at 559 tbl. 2.
86.	Swim et al., supra note 55, at 616.
87.	Goodman et al., supra note 55, at 191.
88.	497 U.S. 836, 851 (1990).
89.	Id.
90.	Ross et al., supra note 55, at 560.
91.	Id.
92.	Swim et al., supra note 55, at 617.
93.	Ross et al., supra note 55, at 556; Swim et al., supra note 55, at 606.
94.	Goodman et al., supra note 55, at 195.
95.	Id. at 196.
96.	Id.
97.	Ann E. Tobey et al., Balancing the Rights of Children and Defendants: Effects of Closed-Circuit Television on Children's Accuracy and Jurors' Perceptions, in MEMORY AND TESTIMONY IN THE CHILD WITNESS 214 (Maria S. Zaragoza et al. eds., 1995).
98.	Id. at 232.
99.	Id.
100.	Graham Davies & Elizabeth Noon, Video Links: Their Impact on Child Witness Trials, in CHILDREN, EVIDENCE, AND PROCEDURE 24 (Noel K. Clark & Geoffrey M. Stephenson eds., 1993).
101.	Id.
102.	Id. at 24-25.
103.	Goodman et al., supra note 55, at 196; Davies & Noon, supra note 100, at 24-25.
104.	Id.
105.	Ross et al., supra note 55, at 556; Swim et al., supra note 55, at 606.
106.	Maryland v. Craig, 497 U.S. 836, 851 (1990).
107.	Id. at 845.
108.	The Court's emphasis on the "truth-seeking goal" of the Confrontation Clause is problematic and has engendered sharp criticism. See Philip Halpern, The Confrontation Clause and the Search for Truth in Criminal Trials, 37 BUFF. L. REV. 165, 200 (1988-89) ("[T]he emphasis on reliability marks a dangerous trend in criminal adjudication. Unqualified truth concerning facts disputed in litigation cannot consistently, if ever, be attained."); Toni M. Massaro, The Dignity Value of Face-To-Face Confrontations, 40 U. FLA. L. REV. 863, 888 (1988) ("The confrontation guarantee is a defendant-centered right and cannot reasonably be read as a general assurance of trials based on reliable evidence.").
109.	487 U.S. 1012, 1020 (1998).
110.	Jean Montoya, On Truth and Shielding in Child Abuse Trials, 43 HASTINGS L.J. 1259, 1290-91 (1992) (discussing the dilemma of shielding to prevent false negatives-children's incomplete and ineffective testimony and children's recantations-at the risk of facilitating false positives-children's inaccurate and untruthful allegations).
111.	The reasonable doubt standard in criminal trials is premised on the belief that it is better for some guilty to go free than to convict the innocent. In re Winship, 397 U.S. 358, 372 (1970) (Harlan, J., concurring). Compare Thomas D. Lyon, The New Wave in Children's Suggestibility Research: A Critique, 84 CORNELL L. REV. 1004 (1999) (arguing that false acquittals are a more serious problem than false convictions given the realities of child sexual abuse cases), with Stephen J. Ceci & Richard D. Friedman, The Suggestibility of Children: Scientific Research and Legal Implications, 86 CORNELL L. REV. 33 (2000) (using Bayesian probability theory to demonstrate that a false positive is not equivalent to a false negative and that even small probabilities that a child would make a false allegation of sexual abuse may be highly significant).
112.	487 U.S. at 1019 (quoting Jay v. Boyd, 351 U.S. 345, 375-76 (1956) (Douglas, J., dissenting)).
113.	497 U.S. 836, 846 (1990).
114.	Paula E. Hill & Samuel M. Hill, Videotaping Children's Testimony: An Empirical View, 85 MICH. L. REV. 809, 814 & n.19 (1987).
115.	Id. at 814.
116.	Id.
117.	Id. at 815.
118.	Goodman et al., supra note 55, at 183.
119.	Id. at 184.
120.	Id. at 185.
121.	See Holly K. Orcutt et al., Detecting Deception in Children's Testimony: Factfinders' Abilities to Reach the Truth in Open Court and Closed-Circuit Trials, 25 LAW & HUM. BEH. 339, 339-72 (2001).
122.	Id. at 343, 345.
123.	Id. at 343.
124.	Id. at 347.
125.	Id. at 351.
126.	Id. at 345.
127.	Id. at 356.
128.	Id.
129.	Id.
130.	Id.
131.	Id. at 351.
132.	Id.
133.	Id.
134.	487 U.S. 1012, 1019 (1988).
135.	See Paul Ekman, Why Lies Fail and What Behaviors Betray a Lie, in CREDIBILITY ASSESSMENT 71 (John C. Yuille ed., 1989) (finding that there are behavioral cues that indicate deception, but people rarely use these cues).
136.	Goodman et al., supra note 55, at 192.
137.	Id.
138.	Id. at 193.
139.	Id.
140.	Id.
141.	See Orcutt et al., supra note 121, at 357.
142.	Id.
143.	Id. at 366.
144.	Id.
145.	FED. R. EVID. 801©.
146.	FED. R. EVID. 801(b).
147.	FED. R. EVID. 802.
148.	See FED. R. EVID. 803 (listing exceptions for which the availability of the declarant is immaterial for admissibility); FED. R. EVID. 804 (requiring the unavailability of the declarant for admissibility); FED. R. EVID. 807 (including a residual exception).
149.	See FED. R. EVID. 803(2) (creating an exception for a "statement relating to a startling event or condition made while the declarant was under the stress of excitement caused by the event or condition"). In White v. Illinois, 502 U.S. 346, 349, 350 & n.1 (1992), for instance, statements made by a child sexual assault victim to her mother thirty minutes after her assault and while the child appeared "scared" and a "little hyper" were admitted under the Illinois spontaneous declaration exception, the state equivalent of the federal excited utterance exception. Id.
150.	See FED. R. EVID. 803(4). In White, for instance, statements made by a four-year-old child sexual assault victim to an emergency room nurse and doctor were admitted under the Illinois "medical examination" exception. 502 U.S. at 350-51. For a discussion of this hearsay exception in the context of child sexual abuse cases, see generally Robert P. Mosteller, Child Sexual Abuse and Statements for the Purpose of Medical Diagnosis or Treatment, 67 N.C. L. REV. 257 (1989) (untangling the rationales supporting the medical examination exception), and Robert P. Mosteller, The Maturation and Disintegration of the Hearsay Exception for Statements for Medical Examination in Child Sexual Abuse Cases, 65 LAW & CONTEMP. PROBS. 47 (2002) (updating his earlier analysis).
151.	See statutes cited supra note 4. Statutes vary, but the most common formulation requires the child's testimony or a showing of the child's unavailability and corroboration. See Jean Montoya, Child Hearsay Statutes: At Once Over-Inclusive and Under-Inclusive, 5 PSYCHOL. PUB. POL'Y & L. 304, 305-06 & n.10 (1999) (discussing the child hearsay statutes).
152.	See Idaho v. Wright, 497 U.S. 805, 814 (1990) (observing that hearsay and Confrontation Clause analyses are distinct).
153.	Citing Coy v. Iowa and Maryland v. Craig, the defendant-petitioner in White argued that children's hearsay should be admitted only upon a showing of necessity. White, 502 U.S. at 357-58. The Court in White, however, distinguished Coy and Craig as cases addressing Confrontation Clause requirements when shielding procedures are used and not when hearsay is offered. Id. at 358. It may, however, be appropriate to treat children's hearsay, even children's hearsay technically falling within a firmly rooted hearsay exception, differently than adult hearsay. See Montoya, supra note 5, at 977-86 (arguing that the child witness context raises distinct issues when assessing the reliability of hearsay).
154.	The excited utterance exception and the exception for statements made for purposes of medical diagnosis or treatment are among the firmly rooted hearsay exceptions. White, 502 U.S. at 355 n.8 (1992).
155.	Id. at 348-49.
156.	Id. at 355-56.
157.	Id. at 356.
158.	Idaho v. Wright, 497 U.S. 805, 817 (1990) (regarding Idaho's residual hearsay exception).
159.	Id.
160.	Id. at 821.
161.	Id. at 819.
162.	Id. at 821.
163.	Id. at 822-24. This approach to the trustworthiness determination was sharply criticized by the Wright dissenters. Id. at 827-35 (Kennedy, J., dissenting). See John E.B. Myers et al., Jurors' Perceptions of Hearsay in Child Sexual Abuse Cases, 5 PSYCHOL. PUB. POL'Y & L. 388, 390-91, 408, 415 (1999) (concluding from an empirical study that jurors consider the existence of corroborating evidence in assessing the trustworthiness of children's hearsay).
164.	In Ohio v. Roberts, the Court set forth "a general approach" for determining when hearsay statements admissible under an exception to the hearsay rule also meet the requirements of the Confrontation Clause. 448 U.S. 56, 65 (1980). Under this approach, "the prosecution must either produce, or demonstrate the unavailability of, the declarant whose statement it wishes to use against the defendant." Id. In White v. Illinois, the Court limited Roberts to its facts, asserting that "unavailability analysis is a necessary part of the Confrontation Clause inquiry only when the challenged out-of-court statements were made in the course of a prior judicial proceeding." White v. Illinois, 502 U.S. 346, 354 (1992). White, however, dealt with hearsay falling within firmly rooted hearsay exceptions. See supra note 154 and accompanying text. White, like Roberts, may ultimately be limited to its facts. Predating White, the Court addressed hearsay not falling within a firmly rooted hearsay exception in Idaho v. Wright, 497 U.S. 805, 817 (1990). Applying the Roberts approach, the Court in Wright assumed that the child declarant was unavailable within the meaning of the Confrontation Clause. Id. at 816.
165.	See Wright, 497 U.S. at 817 (observing that "[a]dmission under a firmly rooted hearsay exception satisfies the constitutional requirement of reliability because of the weight accorded longstanding judicial and legislative experience in assessing the trustworthiness of certain types of out-of-court statements"). For a brief history of the child hearsay statutes, see Robert G. Marks, Note, Should We Believe the People Who Believe the Children?: The Need for a New Sexual Abuse Tender Years Hearsay Exception Statute, 32 HARV. J. ON LEGIS. 207, 237-38 (1995).
166.	See statutes cited supra note 4.
167.	See Montoya, supra note 151, at 309-16 (discussing the overinclusiveness of child hearsay statutes).
168.	See Tome v. United States, 513 U.S. 150, 154 (1995) (indicating that the child testified, and the Government produced six witnesses who testified, to a total of seven statements made by the child describing the alleged sexual assaults); State v. Robinson, 735 P.2d 801, 805-06 (Ariz. 1987) (noting that one of two child victims testified, and out-of-court statements were introduced through a physician, a psychologist, and two babysitters); State v. Tucker, 798 P.2d 1349, 1351 (Ariz. Ct. App. 1990) (noting that a child testified, and the child's out-of-court statements were introduced through a police officer, detective, and social worker); Duvall v. State, 852 S.W.2d 144, 145 (Ark. Ct. App. 1993) (noting that the victim testified, and that the state introduced the child's hearsay statements through various witnesses such as a SCAN worker, a police officer, a doctor and his nurse, a babysitter, and the babysitter's daughter); People v. Brodit, 72 Cal. Rptr. 2d 154, 163 (Cal. Ct. App. 1998) (noting that the child and six hearsay witnesses testified including the child's grandmother, her aunt, a social worker, a police detective, a nurse practitioner, and the child's therapist); People v. Salas, 902 P.2d 398, 400 (Colo. Ct. App. 1994) (noting that the jury heard the child's description of the assault a total of four times when the child testified, two adult prosecution witnesses recounted the child's hearsay statements, and a police investigator played a recorded interview with the child); State v. Marshall, 694 A.2d 816, 818 (Conn. App. Ct. 1997) (noting that the trial court admitted into evidence the testimony of six constancy-of-accusation witnesses, in addition to the victim's videotaped testimony and two videotaped interviews), cert. granted in part, 697 A.2d 361 (Conn. 1997); Jenkins v. State, 508 S.E.2d 710, 711 (Ga. Ct. App. 1998) (noting that the child and seven hearsay witnesses testified including the child's mother, great-grandmother, a criminal investigator, a social worker, a psychologist, a nurse, and a physician); People v. Byron, 645 N.E.2d 1000, 1002-03 (Ill. App. Ct. 1995) (noting that the victim testified, and that the victim's mother, sister, and two police officers testified, pursuant to the child hearsay statute); State v. Silvey, 894 S.W.2d 662, 664-65, 672 (Mo. 1995) (noting that the child testified, and that the child's out-of-court statements were recounted by her mother, her aunt, and a social worker); State v. Gollaher, 905 S.W.2d 542, 545 (Mo. Ct. App. 1995) (noting that the child victim testified, and that the child's out-of-court statements were recounted by four witnesses, namely, the doctor who examined the child, the child's mother, the child's aunt, and an investigating officer); State v. White, 873 S.W.2d 874, 877 (Mo. Ct. App. 1994) (noting that the child testified, and that the child's out-of-court statements were recounted by four witnesses, namely, a pediatrician, two case workers, and his foster mother), overruled in part by State v. Redman, 916 S.W.2d 787 (Mo. 1996); State v. Tringl, 848 S.W.2d 29, 31 (Mo. Ct. App. 1993) (noting that the victim testified, and her out-of-court statements were recounted by four witnesses, namely, the child's stepmother, two detectives, and a doctor); Felix v. State, 849 P.2d 220, 229, 254 (Nev. 1993) (noting that the child testified, and that her out-of-court statements were recounted by five adult witnesses including her mother, two psychologists, a pediatrician, and a detective); State v. Cook, 881 P.2d 913, 916 (Utah Ct. App. 1994) (noting that the child victim testified, and that the child's father and the defendant's step-granddaughter recounted the child's hearsay statements).
169.	See Myers et al., supra note 163, at 416.
170.	Id. at 394, 396.
171.	Id. at 397.
172.	Preparing the child witness to testify and giving child witnesses tours of the courtroom are strategies frequently used by prosecutors. See generally Gail S. Goodman et al., Innovations for Child Witnesses: A National Survey, 5 PSYCHOL. PUB. POL'Y & L. 255 (1999) (reporting the results of a nationwide survey of prosecutors). Adequate preparation of the child witness may ameliorate the child witness's fear of the confrontational setting. Louise Dezwirek-Sas, Empowering Child Witnesses for Sexual Abuse Prosecution, in CHILDREN AS WITNESSES 181, 196-97 (Helen Dent & Rhona Flin eds., 1992); Stephen Moston, Social Support and Children's Eyewitness Testimony, in CHILDREN AS WITNESSES 33, 43 (Helen Dent & Rhona Flin eds., 1992); Kathryn Sisterman Keeney et al., The Court Prep Group: A Vital Part of the Court Process, in CHILDREN AS WITNESSES 201, 207-08 (Helen Dent & Rhona Flin eds., 1992).
173.	See Myers et al., supra note 163, at 411 (surmising that "prosecutors are reluctant to take child sexual abuse cases to trial unless the victim is available to testify").
174.	See Thomas L. Hafemeister, Protecting Child Witnesses: Judicial Efforts to Minimize Trauma and Reduce Evidentiary Barriers, 11 VIOLENCE & VICTIMS 71, 77 (1996) (reporting that the use of hearsay exceptions was the innovation "most likely to be considered unfair" by surveyed judges).
175.	Id. at 75 (reporting that seventy-two percent of surveyed judges admitted children's out-of-court statements pursuant to hearsay exceptions).
176.	See cases cited supra note 168.
177.	See Myers et al., supra note 163, at 397.
178.	849 P.2d 220, 253 (Nev. 1993). A few other courts have addressed the problem of repetitive hearsay statements in child abuse cases. In Pardo v. State, for instance, the Florida Supreme Court recognized that the admission of repetitive hearsay statements pursuant to the child hearsay statute is problematic and subject to the balancing test in Florida's equivalent of Federal Rule of Evidence 403. 596 So. 2d 665, 667-68 (Fla. 1992). In that case, the state had filed notices of intent to rely on hearsay statements made by the child victim to nine separate individuals. Id. at 666. The trial court had found only three of the hearsay statements to be sufficiently reliable. Id. Similarly, in State v. D.G., the New Jersey Supreme Court reminded the state trial courts that repetitive child hearsay statements are subject to exclusion pursuant to New Jersey's equivalent of Federal Rule of Evidence 403. 723 A.2d 588, 595-96 (N.J. 1999). In that case, the child's credibility was at issue: The child had recanted, had made identical allegations against someone other than the defendant, and the prosecutor was forced to impeach her on the stand. Id. at 595. The court observed that the corroborative statements "may well have tipped the scale." Id.
179.	Felix, 849 P.2d at 254-55.
180.	Id. at 229, 254.
181.	Id. at 253.
182.	As a matter of federal evidence law, prior consistent statements are inadmissible unless offered to rebut a charge of recent fabrication or improper influence or motive. FED. R. EVID. 801(d)(1)(B). The statements are not admissible to bolster the declarant's trial testimony. Tome v. United States, 513 U.S. 150, 157-58 (1995).
183.	Felix, 849 P.2d at 253.
184.	MICH. R. EVID. 803A.
185.	TEX. CRIM. CODE ANN. § 38.072((2)(a)(2)) (West Supp. 2001).
186.	MASS. GEN. LAWS ANN. ch. 233, § 81(a) (West 2000).
187.	David F. Ross et al., The Impact of Hearsay Testimony on Conviction Rates in Trials of Child Sexual Abuse: Toward Balancing the Rights of Defendants and Child Witnesses, 5 PSYCHOL. PUB. POL'Y & L. 439, 441 (1999).
188.	Id. at 445.
189.	Id.
190.	Id. at 447.
191.	Id. at 449.
192.	Id. at 450.
193.	Id.
194.	Id. at 450-51.
195.	Id. at 452.
196.	Id.
197.	Id. at 445.
198.	Id. at 450.
199.	V. Anne Tubb et al., Effects of Suggestive Interviewing and Indirect Evidence on Child Credibility in a Sexual Abuse Case, 29 J. APPLIED SOC. PSYCHOL. 1111, 1114 (1999).
200.	Id. at 1115.
201.	Id.
202.	Id. at 1117.
203.	Id.
204.	Jonathan M. Golding et al., The Believability of Hearsay Testimony in a Child Sexual Assault Trial, 21 LAW & HUM. BEHAV. 299, 304 (1997).
205.	Id.
206.	Id.
207.	Id.
208.	Id.
209.	Id.
210.	Id.
211.	Id. at 308.
212.	Id.
213.	Id.
214.	Id. at 309.
215.	Id. at 310.
216.	Id. at 316.
217.	Id.
218.	Id.
219.	Id.
220.	Id. at 316-17.
221.	Jonathan M. Golding et al., The Effect of Hearsay Witness Age in a Child Sexual Assault Trial, 5 PSYCHOL. PUB. POL'Y & L. 420, 424 (1999).
222.	Id.
223.	Id. at 427.
224.	See Ross et al., supra note 187, at 447.
225.	See Golding et al., supra note 204, at 308; Tubb et al., supra note 199, at 1117.
226.	See Ross et al., supra note 187, at 450.
227.	Id. at 453.
228.	Id.
229.	Id. at 445.
230.	Id.
231.	Id. at 447.
232.	Id.
233.	See Golding et al., supra note 221, at 427.
234.	See supra notes 152-59 and accompanying text.
235.	Idaho v. Wright, 497 U.S. 805, 821-22 (1990). Various child hearsay statutes include their own nonexclusive list of factors relevant to the reliability inquiry. E.g., ALA. CODE § 15-25-37 (1995) (listing the following factors: (1) the child's personal knowledge of the event; (2) the age and maturity of the child; (3) certainty that the statement was made, including the credibility of the person testifying about the statement; (4) any apparent motive the child may have to falsify or distort the event, including bias, corruption, or coercion; (5) the timing of the child's statement; (6) whether more than one person heard the statement; (7) whether the child was suffering from pain or distress when making the statement; (8) the nature and duration of any alleged abuse; (9) whether the child's young age makes it unlikely that the child fabricated a statement that represents a graphic, detailed account beyond the child's knowledge and experience; (10) whether the statement has a "ring of verity," has an internal consistency or coherence, and uses terminology appropriate to the child's age; (11) whether the statement is spontaneous or directly responsive to questions; (12) whether the statement is suggestive due to improperly leading questions; (13) whether extrinsic evidence exists to show the defendant's opportunity to commit the act complained of in the child's statement).
236.	Wright, 497 U.S. at 826-27.
237.	The doctor who interviewed the child in Wright asked leading questions and questioned the child with a preconceived idea of what occurred. Id. at 818.
238.	Id. at 818-19. For a discussion about mandatory videotaping of child witness interviews, see Lucy S. McGough, Good Enough for Government Work: The Constitutional Duty to Preserve Forensic Interviews of Child Victims, 65 LAW & CONTEMP. PROBS. 179 (2002) (advocating videotaping).
239.	Maggie Bruck et al., The Accuracy of Mothers' Memories of Conversations with Their Preschool Children, 5 J. EXPERIMENTAL PSYCHOL.: APPLIED 89, 93 (1999).
240.	Id. at 94.
241.	Id. at 96.
242.	Id. at 99.
243.	Id. at 99-100.
244.	Id. at 96 tbl. 1.
245.	Amye R. Warren & Cara E. Woodall, The Reliability of Hearsay Testimony: How Well Do Interviewers Recall Their Interviews with Children?, 5 PSYCHOL. PUB. POL'Y & L. 355, 359 (1999).
246.	Id. at 360.
247.	Id.
248.	Id. at 363.
249.	See generally Michael E. Lamb et al., Accuracy of Investigators' Verbatim Notes of Their Forensic Interviews with Alleged Child Abuse Victims, 24 LAW & HUM. BEHAV. 699 (2000).
250.	Id. at 699-700.
251.	Id. at 703.
252.	Id.
253.	Id.
254.	Id.
255.	Id.
256.	Id. at 704.
257.	Id.
258.	Id. at 705.
259.	Maithilee K. Pathak & William C. Thompson, From Child to Witness to Jury: Effects of Suggestion on the Transmission and Evaluation of Hearsay, 5 PSYCHOL. PUB. POL'Y & L. 372, 374 (1999).
260.	Id. at 376.
261.	Id. at 375.
262.	Id. at 377-78.
263.	Id. at 376-77.
264.	Id. at 377.
265.	Id.
266.	Id.
267.	Id.
268.	Id. at 378.
269.	Id. at 379.
270.	Id.
271.	Id. at 381.
272.	Id.
273.	Bruck et al., supra note 239, at 99-100.
274.	Lamb et al., supra note 249, at 703; Warren & Woodall, supra note 245, at 363.
275.	Lamb et al., supra note 249, at 703; Warren & Woodall, supra note 245, at 363.
276.	Bruck et al., supra note 239, at 99-100.
277.	Id.; Lamb et al., supra note 249, at 704; Warren & Woodall, supra note 245, at 363.
278.	Pathak & Thompson, supra note 259, at 381.
279.	For additional discussion on this topic, see Ceci & Friedman, supra note 111, at 93-96.
280.	497 U.S. 836 (1990).
281.	The development of rules that strike the proper balance between an individual defendant's constitutional right to confront witnesses against him depends not only on social science research conclusions, but on the legal system's determination that those conclusions should be used to alter the framework currently used to analyze the problem. Some members of the Court have cited social science research in their analysis of the problem, see id. at 855, 857 (citing research by Gail Goodman), but others have labeled the conclusion in Craig "antitextual," and a "subordination of explicit constitutional text to currently favored public policy." Id. at 861, 863 (Scalia, J., dissenting) (indicating that resolution of a defendant's confrontation rights should not depend on policy developed from social science research).
282.	For more on this topic, see supra notes 28-46 and accompanying text.
283.	Bussey et al., supra note 42, at 162; Peters, supra note 36, at 69.
284.	Goodman et al., supra note 55, at 184; Hill & Hill, supra note 114, at 814-15.
285.	State court decisions vary on the persons permitted to serve as expert witnesses on trauma and on the kind of testimony required. Compare People v. Pesquera, 625 N.W.2d 407 (Mich. Ct. App. 2001) (using the testimony of experts as well as of the mothers of the children to determine that the children should testify under shielding procedure), and George v. Commonwealth, 885 S.W.2d 938 (Ky. 1994) (holding that the victim should not have been permitted to give shielded testimony when the only evidence of trauma was a social worker's testimony that the victim would be more traumatized than most children), with People v. Van Brocklin, 687 N.E.2d 1119 (Ill. App. Ct. 1997) (holding a social worker's testimony to be sufficient). See also State v. Welch, 744 So. 2d 64 (La. Ct. App. 1999) (requiring expert testimony by state statute), rev'd, 760 So. 2d 317 (La. 2000); Marx v. State, 987 S.W.2d 577 (Tex. App. 1999) (noting that the teacher, parent, and grandparent claimed the child's fear would prevent testimony, and that the treating mental health therapist testified that she could not be sure the child could testify). Some states have held that no expert testimony is necessary in a Craig hearing, e.g., State v. Bronson, 740 A.2d 458 (Conn. App. Ct. 1999), but prosecutors may still find it helpful.
286.	See generally Montoya, supra note 19 (suggesting that the trial court should not rely on descriptions offered by therapists or parents and should view the child). It is not clear that courts have a well-developed standard for "fearfulness." Some state statutes do give guidance to the trial court. See MICH. COMP. LAWS. ANN. § 600.2163a (West 2000) (listing factors for court consideration); WIS. STAT. ANN. § 967.04(7)(b) (West 1998) (detailing factors that the court must consider before shielding a child witness). But see 725 ILL. COMP. STAT. ANN. 5/106B-5(a)(2) (West Supp. 2000) (requiring only that the judge find that the child suffers from emotional distress sufficient to meet the Craig test); IOWA CODE ANN. § 915.38 (West Supp. 2000) (requiring only that the court find trauma that would meet Craig); HAW. REV. STAT. ANN. § 626 (Michie 1995) (accord); HAW. R. EVID. 616 (accord).
287.	Ross et al., supra note 55, at 563; see Swim et al., supra note 55, at 620.
288.	Ross et al., supra note 55, at 563; see Swim et al., supra note 55, at 620.
289.	Goodman et al., supra note 55, at 198; Ross et al., supra note 55, at 558; Swim et al., supra note 55, at 620.
290.	Davies & Noon, supra note 100, at 24-25; see Tobey et al., supra note 97, at 232.
291.	Defense attorneys may already know intuitively what this evidence tends to show: that jurors seem to have less connection with or empathy for a child who is shielded. Even with this knowledge, others factors will necessarily affect a defendant's assessment of the impact of shielding. It may be one thing to find the absence of bias and no increase in conviction rates in studies and quite another for the individual attorney to forego the defendant's rights under Craig. Not only does the criminal defendant have Confrontation Clause rights, but his right to effective assistance of counsel also has a constitutional dimension that plays an important role in the selection of trial tactics.
292.	See generally Louise Dezwirek-Sas, Sexually Abused Children as Witnesses: Progress and Pitfalls, in CHILD ABUSE: NEW DIRECTIONS IN PREVENTION AND TREATMENT ACROSS THE LIFESPAN 248-67 (David A. Wolfe & Robert Joseph McMahon eds., 1997); see also generally Dezwireck-Sas, supra note 32.
293.	Montoya, supra note 151, at 309-16.
294.	Ross et al., supra note 187, at 450.
295.	Id. at 447.
296.	Bruck et al., supra note 239, at 99-100.
297.	Pathak & Thompson, supra note 259, at 381.
298.	Goodman et al., supra note 55, at 196; Tobey et al., supra note 97, at 232.
299.	Golding et al., supra note 204, at 316-17.