What Have We Learned about Children as Victims and Witnesses in the Criminal Trial Process?

The Conundrum of Children, Confrontation, and Hearsay
by Richard D. Friedman*

Adjudicating claims of child abuse poses an excruciatingly difficult conundrum. The crime is a terrible one, but false convictions are abhorrent, and often the evidence does not point one way or the other with sufficient clarity to allow a decision free of gnawing doubt. In many cases, a large part of the problem is that the prosecution case depends critically on the statement or testimony of a young child. Even with respect to adult witnesses, the law of hearsay and confrontation is very perplexing, as anyone who has studied American evidentiary law and read the Supreme Court opinions on the subject knows. Juxtaposing problems of hearsay and confrontation with those of child sexual abuse yields one of the most intractable problems that a system of criminal justice can face.

A. Restructuring the General Law of Confrontation and Hearsay

I do not believe we can reach a satisfactory resolution of the problem until we restructure the general law of confrontation and hearsay.¹ Noting that these two bodies of law "are generally designed to protect similar values and stem from the same roots,"² the Supreme Court has tended to meld the two together, treating both as aimed at facilitating accuracy in truth-finding by weeding out unreliable evidence. As a result, the confrontation right has been given very little ambit independent of hearsay law. If a statement fits within a "firmly rooted" hearsay exemption – and inclusion in the Federal Rules of Evidence seems to qualify an exemption, and the bounds of that exemption, as firmly rooted³ – then the Confrontation Clause is unlikely to pose an obstacle to admissibility.⁴ As a result, whether a statement by a child alleging sexual abuse should be admitted at the trial of the alleged abuser often depends on whether the court deems the statement to fit within one of the hearsay exemptions listed in the Federal Rules, a list supposedly determined by considerations of reliability.

This approach is objectionable on numerous grounds.

1. The Difficulties of Determining Reliability

Reliability is very difficult to determine. To confirm this point, we need look no further than the current debate over the reliability of child witnesses.⁵ Any attempt to sort out reliable from unreliable evidence must run aground. If the determination is made according to broad categories of evidence, the law will be too blunt, not only sweeping in many statements that are not reliable but basing decision on generalizations that may have little bearing on the particular case. If the determination is made case-by-case, it will not necessarily be any better, but it will inevitably be very subjective, and so either effectively immune from appellate control or extremely demanding on appellate resources.

As an example of how reliability testing does not work, consider the decision of the United States Court of Appeals for the Tenth Circuit on remand in United States v. Tome.⁶ The panel majority held that a five-year-old girl's statements to pediatricians alleging that her father had sexually abused her more than a year before were within the hearsay exception for statements made for purposes of medical diagnosis or treatment, and therefore reliable for purposes for the Confrontation Clause, because the identity of the perpetrator was important information in determining whether there was a risk of sexually transmitted disease. What a remarkable way to determine admissibility this is! The broad premise of the hearsay exception – that statements made for the purposes of medical diagnosis or treatment are reliable, because a declarant would not want to endanger her health by giving false information in this context – is dubious enough, given the frequency with which patients lie to their doctors. But even accepting that premise, it is simply nonsense to conclude that the particular statements at issue were reliable for the reasons stated by the majority of the Tenth Circuit panel. Do we really believe that a five-year-old girl almost certainly recognizes that it is imperative that she identify correctly the person who abused her (assuming abuse did occur) more than a year earlier, because otherwise she would risk not being treated properly for sexually transmitted disease?

2. The Inappropriateness of the Reliability Criterion

More significantly, even if we had a sound way of determining reliability, it would be an unsuitable criterion for determining admissibility. Trials are not supposed to be based on winnowing evidence to that which is reliable;⁷ trials are attempts to determine the truth from an aggregation of evidence some of which may be very unreliable. Live testimony under oath and subject to cross-examination – the paragon of acceptable evidence – is not necessarily reliable; that is why we have conflicting testimony at trial.

Could it be thought, though, that reliability acts as a sufficient proxy for these conditions? The theory would be that if a statement is so reliable that there is virtually no doubt about its truth then we can overlook the fact that it was not made under the conditions we usually expect of testimony. I do not believe this theory is persuasive. In order for us to accept an out-of-court statement as essentially conclusive it would have to be extraordinarily reliable indeed. There would have to be no more than a minuscule possibility that the statement would have been made if it were not true. Despite continued expressions of optimism on this score,⁸ we really cannot have such faith in any broad category of statement. Statements by children are no different. If a child makes a statement alleging that the accused abused her, then in most circumstances the probability that she would have made the statement even though it was false, though perhaps small, is potentially large enough to be of concern.

Suppose, to take a simple numerical example, that apart from the allegation of abuse made by a child, the fact-finder would assess the odds that the defendant is guilty of abuse to be 1:2. In other words, though other evidence makes guilt appear plausible, it appears about twice as likely that the defendant is not guilty. And suppose further that the fact-finder estimates that if the abuse occurred the probability that the child would say that it did is 70% and that if the abuse did not occur the probability that the child would nevertheless say that it did is only 5%. (These numbers, though presented for heuristic purposes only, are quite realistic.⁹) Then, given the allegation, the fact-finder should assess the odds that the defendant committed the abuse as 7:1,¹⁰ which is plainly insufficient for conviction.¹¹ If the odds of guilt as assessed without the child's statement are significantly lower, as they well may be in a case largely dependent on that statement, so too will the odds given that statement. If, as assessed apart from the child's statement the odds that the defendant are guilty appear quite small, and there is more than a minuscule probability that the child would make the statement even though the defendant was not guilty, the odds of guilt even given the statement will not be sufficient to support a conviction.

3. The Essence of the Confrontation Right

Most fundamentally, the focus on reliability misses the point of the confrontation right. The Confrontation Clause is not a constitutionalization of the law of hearsay, with all its oddities. It does not speak of reliability or of exceptions. The confrontation right reflects a belief central to our system of criminal justice, that a witness against a criminal defendant should give testimony under prescribed conditions – under oath, in the presence of the accused, subject to cross-examination, and, if reasonably possible, in open court. And this right should be recognized, as the language of the Clause suggests, as categorical, not subject to exceptions. When a witness testifies in court against an accused, the court does not say, "This testimony was so reliable that you really have no need to cross-examine." And the same should be true when a person acts as a witness by making a testimonial statement out of court that is later offered against the accused.

Under this approach, the confrontation right extends not to all hearsay statements but only to those made by "witnesses" against the accused. Put another way, to come within the Confrontation Clause, a statement must be testimonial in nature. What makes a statement testimonial? That, of course, is sometimes a very difficult question. The right is meant to prevent the creation of a system in which witnesses can provide evidence for prosecution of a crime without submitting themselves to cross-examination under oath in the presence of the accused.¹² Basically, if an out-of-court declarant knows that she is likely providing evidence for investigation or prosecution of a crime, then she is testifying – for if the law permitted admissibility of the statement then this would be a means for her to present evidence without submitting to what we regard as the usual conditions of testimony. If the declarant makes an accusation to the authorities, then presumably – in the case of adults – it is a testimonial statement. But testimony should not be limited to statements made directly to the authorities,¹³ for witnesses can speak through intermediaries who pass the testimony on.

In most cases, though not all, I think it is reasonably clear whether a statement should be considered testimonial. If it is, the confrontation right should attach, and categorically bar admissibility of the statement unless the accused has had an adequate opportunity to examine the witness.¹⁴

Like almost any right, however, the confrontation right should be considered subject to forfeiture by misconduct. If wrongdoing of the accused accounts for the unavailability of the witness, then the accused cannot complain that his confrontation right renders improper use of a statement made by the witness but not under the usual conditions for testimony.¹⁵

Once the nature of the confrontation right is recognized, and the right is protected by a doctrine independent of the law of hearsay, the importance of hearsay law diminishes vastly. No longer is there a need for a broad presumptive rule of exclusion, with a long list of exemptions. Instead, courts should recognize that if live testimony of the declarant would be more probative than prejudicial, usually the hearsay would be too.¹⁶ And therefore, when hearsay has significant probative value and does not raise confrontation problems, courts should generally be receptive to it, excluding it only for a good reason – such as that exclusion will induce the production of better evidence, in this case or over the long run.

4. The Basic Framework

The discussion above suggests the basic framework that I believe should govern the law of confrontation and hearsay. It does not leave room for anything like the hearsay exceptions – no attempts to stretch the exceptions for excited utterances or for statements made for medical diagnosis out of shape to accommodate statements by children or other declarants.¹⁷ Instead, we should ask first whether the statement was a testimonial one. If so, then the confrontation right attaches, categorically. And so the statement must be excluded, unless the accused has had an adequate opportunity to examine the witness – or unless the reason the accused has not had such an opportunity is attributable to the accused's own wrongdoing. If the statement was not a testimonial one, then the confrontation right does not apply, and presumably the statement ought not be excluded if it is significantly probative. But other considerations, particularly the possibility of inducing the presentation of better evidence, might yet require the exclusion of the evidence.

At the most general level, this framework should be applied to statements by children as well as to those by adults. But of course the application must be sensitive to context, and so the results may be profoundly affected by the fact that the declarant is a child. I will focus on two questions that will often be particularly important in the case of children: Was the statement testimonial? And did the accused forfeit the right to confront the child?

B. Is the Statement Testimonial?

Suppose a ten-year-old child of ordinary intelligence makes a statement to an investiga-tive social worker accusing an adult of sexual abuse. Under the approach I have advocated here, it is clear that she is acting as a witness against the accused, just as much as if she were an adult, and the confrontation right attaches to her statement: If she does not testify subject to oath and cross-examination, then her statement should not be admitted, unless the accused is held to have forfeited the confrontation right.

And the same result would presumably follow if instead the child made the statement to her mother. The child presumably realized that her mother would likely take responsive conduct that might well have significant punitive consequences for the accused. In this setting, I believe the child should be deemed to be testifying just as if she were adult, and just as if she made her statement in court rather than through an intermediary.

Now consider a polar opposite case: Suppose a bloodhound barks at the accused, effectively identifying him as the source of an item of clothing. Evidence of the bloodhound's conduct may very well be admitted – and we would not dream of excluding it because the bloodhound cannot be cross-examined.¹⁸

Given those two polar treatments, how should the law of confrontation treat astatement by a very young child alleging abuse? It is tempting to conclude without analysis that, given that the child is a human, she should be treated as a witness, like adult humans or older children who make statements alleging criminal conduct. I am not so sure this is so.

Suppose a child is so young that she has no sense that what she is reporting is wrongful conduct (if in fact her words are properly understood as a report of conduct that adults know to be wrongful), and no sense that the person she asserts committed the conduct is subject to punishment for it. Consider, for example, these two cases:

· Webb: An 18-month-old, girl, on being lowered into the bath, said, "Ow bum," and then after the bath, while her mother was examining her, "Ow bum daddy."¹⁹

· Rhea: A 2 ½ year-old child was interviewed by a child protection specialist of the Department of Human Resources. She denied in response to questions that her father had done anything wrong to her but nodded in agreement to the question, "Did daddy make you put his tee-tee in your mouth?"²⁰

It seems to me dubious to say that the children in these cases were acting as witnesses. Without some sense of social organization that would be beyond the ken of children so young, they could not have understood that they were providing information that could lead to the punishment of their fathers, who presumably they regarded as being virtually immune to punishment. Indeed, it appears that the child in Rhea did not regard the conduct as wrongful, and the same may be true of the child in Webb. I doubt that a child this young is capable of recognizing that any conduct by an adult is wrongful and may lead to punishment.

Certainly these children were providing information, but people do that all the time – completing business records, advancing conspiracies, and so forth – without the communication being testimonial. It seems doubtful to me that the children's communications in these cases should be considered testimonial.

That conclusion is weaker in Rhea than in Webb, not only because the child was older but also because her communication was elicited by a state officer. At the very least, that is a factor to be taken into account in determining what the child's understanding was; it is possible that a child would not, and perhaps could not, have made a testimonial statement on her own initiative, but still would have enough understanding of the nature and purpose of the questioning that her responses should be characterized as testimonial.

Arguably, though, a broader proposition should prevail – that elicitation by the state is in itself enough to treat a person's statement as testimonial.²¹ I am reluctant to draw that conclusion – which, if consistently applied in the adult realm, would presumably render subject to the Confrontation Clause many statements made by conspirators not knowing that their listeners were informants or undercover police officers. Even if the government is trying to draw information from a source, that does not mean that the source is properly thought to be testifying. Indeed, prosecutorial authorities often try to elicit information from non-human sources, and the information they gain cannot be characterized as testimonial. I do not think there is a persuasive reason to treat differently the human source who, while recognizing that she is providing information in the course of a conversation, does not regard the conversation as testimonial in nature.

If the communications in cases like Rhea and Webb are considered non-testimonial, that means that the confrontation right as I have conceived it should not apply.. But it does not mean the statements should necessarily be admitted. A court might determine that the methods of questioning by authorities were so unduly suggestive that they should not be countenanced – a decision that would be motivated largely by the desire to deter future investigators from acting in such an irresponsible way. Or the court might conclude that the authorities failed to follow procedures – such as videotaping the interview – to which it is reasonable to hold repeat players like themselves.²² Or the court might even hold that the evidence could have been rendered far more satisfactory if the investigators had allowed a representative of the defendant to take part, perhaps not by formal questioning of the child but by planning the interview and suggesting follow-up questions. Indeed, such participation might be far more productive than cross-examination in court, which is typically not only uninformative but also a gruelling, frustrating, and stressful affair for all concerned.

I acknowledge that there may appear to be a paradoxical aspect to the theory I am advancing here. I am suggesting that the younger and less mature and understanding a child is, the less likely her statement should be considered testimonial, subject to the Confrontation Clause, and therefore (all other things being equal) the more likely the statement should be admitted. But I do not think this is really paradoxical at all. Even statements by very young children may be highly probative. But very young children are not yet at a stage where we can expect them to take the responsibility of being a witness – the responsibility of speaking under oath, subject to questioning by the accused, under the implicit injunction, "Look me in the eye and say that."²³ With respect to very young children – I cannot say here just how young – we should admit their statements for what they are worth, without pretending that the children have the capacity to act like adults.

C. Did the Accused Forfeit His Right to Object?

The forfeiture principle prescribes that if the accused's own wrongful conduct is responsible for the inability of the witness to testify under the conditions ordinarily required the accused cannot object to use of the witness's prior statement. This principle is well established as a matter of confrontation and hearsay law, and it potentially has an important, though difficult, role to play in child abuse cases.

It seems plain enough that the forfeiture principle is not rendered inapplicable when the witness who was prevented from testifying at trial is the victim of the crime being charged. I have contended that the same logic should be taken one step further: The forfeiture principle remains applicable even when the conduct that allegedly rendered the witness unavailable to testify is the same criminal conduct for which the accused is now on trial.²⁴

At first glance, this application of the forfeiture principle might seem to be a bizarre instance of bootstrapping. But it isn't. Suppose, as would usually be so, that the case is being tried by a jury. The judge would decide for purposes of the forfeiture principle whether the accused had acted in such a way as to render the witness unavailable to testify. The jury would decide guilt or innocence of the crime. The judge would not tell the jury how she had decided. Not only would the reasons why the factual issue was presented to her and to the jury be different, but the two would decide according to different bodies of evidence²⁵ and different standards of proof.²⁶ The situation is closely analogous to that of a judge presiding over a conspiracy trial who decides that a statement may be admitted because it was made during the course of and in furtherance of a conspiracy – the very one being tried – of which the accused was a part.²⁷

Suppose that a child makes a statement alleging abuse and that the prosecutor, with the cooperation of the child's caretakers, makes a genuine effort by appropriate means to get the child to testify, but that she fails to do so. Suppose that it appears the child may have been intimidated, either by the abusive conduct itself or by a threatening statement – "Don't tell anyone!" – that accompanied or followed the conduct. In such a case, it may be appropriate to apply the forfeiture principle.²⁸

I do not minimize the difficulties of applying the principle in this context. First, as in any case in which the principle is invoked, the court must resolve a basic predicate question that is often very difficult: Did the accused actually commit, or acquiesce in, the wrongful conduct that assertedly rendered the witness unavailable? But in the case of a child witness, the complexities only begin there.

Another issue the court must resolve is whether, assuming the accused committed the conduct, that conduct really accounts for the child's silence. One alternative explanation may be that the cause is merely the child's age and immaturity; if that is so then, assuming the confrontation right applies (whether in accordance with or despite the views I have expressed above in Part B), the accused should not be deemed to have forfeited the right. The judge's task may be conceived as being to determine whether the child would be willing and able to testify about a matter that she did not find especially troubling;²⁹ if she would not be, then presumably it was not the accused's wrongful conduct that accounts for the child's silence and there should be no forfeiture. I do not pretend that resolving this issue would be a simple matter.

In some cases also, the court may have to resolve whether the child's failure to testify is attributable not to her own fears but to the desire of her caretaker, who – perhaps in a desire to protect her from trauma, perhaps from other motives – has refused to make the child available to testify or has instructed her not to testify. In such a case, the child should probably not be deemed unavailable by virtue of the defendant's wrongdoing. Thus, the judge, or an expert appointed by the judge, may need to speak with the child and try to determine whether the child's own fear, or some other influence, is preventing her from testifying under suitable conditions.³⁰

Finally, there is a question of remedy. Suppose the court concludes that the child should be deemed unavailable to testify in court under ordinary procedures, and that this unavailability should be attributed to the accused's wrongful conduct. There remains the issue of what steps, if any, the prosecution must take, as a prerequisite to introducing secondary evidence of the prior statement, to preserve as much as possible of the confrontation right. For example, perhaps even though the child is unable to testify in the courtroom at trial she could, or could have, testified in a less intimidating setting, perhaps without the accused present but electronically connected to the room where the child is testifying, and with a video transmission or recording of her testimony presented in the courtroom at trial.³¹

Conclusion

I began this essay by saying that adjudicating claims of child abuse poses an excruciatingly difficult conundrum. And I will end it the same way. I do not pretend that there are any easy answers, or at least any satisfactory easy answers, to the problem of determining when a child's statement alleging abuse should be admitted notwithstanding hearsay law and the accused's confrontation right. But it seems clear enough that the method by which we have approached the problem, relying on the prevailing structure of hearsay law and pretending to determine reliability, is utterly inadequate. So we must think anew.

Under the approach I have presented here, we would first ask, was the statement testimonial, so that the child was effectively acting as a witness in making the statement? If so, then presumably the statement should not be admitted unless the accused has had an adequate opportunity to examine her, or unless the accused is deemed to have forfeited his objection by wrongful conduct that prevented the child from testifying. If the statement was not testimonial, then the confrontation right should not block its admission. Exclusion may yet be warranted, though, by failure of the authorities to follow procedures, such as videotaping an interview, that would present the evidence in a satisfactory light

I readily acknowledge that in many cases the questions I pose here will be very hard to answer. But they are the right questions to ask. Trying to answer them will help us make the best of a dreadful situation

Back to Top