Introduction
Since the “discovery” of child sex abuse as a social problem in the 1970s and 1980s (Weisberg Reference Weisberg1984, 1–3), it has been clear that the legal system must find ways to address it—whether by criminal prosecutions, civil lawsuits, regulation of institutions dealing with small children, mandatory reporting, education, or other means. The problems posed have been considerable, especially if a sexual abuse victim is very young. Most sexual abuse is not reported, primarily because of nondisclosure by the child due to shame or threats by the perpetrator. When a case does enter the legal system, moreover, it poses substantial problems for police, prosecutors, and judges as they struggle to deal with child witnesses while also protecting the rights of accused persons.
The traditional common law approach to child witnesses was to regard their testimony with extreme skepticism, especially when the topic was sexual abuse (Goodman Reference Goodman1984: 9–15). Requirements such as corroboration or cautionary warnings about the credibility of children claiming sexual abuse were common. However, as public attention was drawn to the issue of child sex abuse, most of these requirements were abolished, and statutes of limitation were extended so that sexual abuse offenders might be held accountable (Bickel Reference Bickel1991).
Legal rules and public policy on this issue have tended to go through massive swings in response to media attention to especially egregious cases—always from one extreme to the other, never pausing to consider shades of gray. After the widespread coverage given to the string of daycare center abuse cases in the 1980s, the pendulum swung in the direction of believing that children were extremely prone to suggestion and should not be regarded as credible witnesses. Groups to support persons who had been accused of child abuse formed and developed a narrative that caught on in the media, claiming that the accusations of child sex abuse were a “witch hunt” analogous to that in Salem, Massachusetts in the late 1600s.
The script of the witch‐hunt narrative goes something like this: an allegation of sexual abuse is made by a small child in a daycare setting, which then spreads like wildfire, presumably because the children are coached in some way by the adults investigating the situation, be they parents, psychologists, doctors, police, or prosecutors. The press takes up the cry for vengeance, focusing on the most sensational incidents to sell their media products. The result is that innocent people are put on trial for child sex abuse and sent to prison.
This narrative, which caught on both in the popular press and academic literature, basically attributed the rash of cases alleging abuse of small children to a kind of irrational moral panic. Calls for legal change followed, swinging the pendulum far in the opposite direction, changing the outcome in some cases and leading to procedural changes and legislation with unanticipated consequences. A much more nuanced understanding of the preschool cases is essential as a basis for sound public policy.
Ross Cheit's The Witch‐Hunt Narrative: Politics, Psychology, and the Sexual Abuse of Children seeks to inform the debate over the appropriate legal response to allegations of child sex abuse. The book is the culmination of fifteen years of research, during which Cheit undertook a detailed review of the trial court record and evidence—much of it very difficult to access—in more than twenty‐five daycare sex abuse cases from the 1980s and 1990s (Cheit Reference Cheit2014, 413–15). Cheit found that there was substantial evidence of sex abuse in many of these cases, enough that claims of moral panic are not justified.
Cheit's book illuminates many issues of importance to social scientists and lawyers, such as (1) the appropriate use of social science in litigation; (2) the impact of the mass media in influencing public opinion on important social issues; (3) the relationship between social scientific research and public policy; and (4) the suitability of criminal versus civil remedies for child sex abuse. I discuss each of these topics in this essay, after providing background on the child sex abuse scandals.
Background
The McMartin Preschool Case
The origin of the witch‐hunt narrative appears to have been the McMartin daycare center case in California, which began in 1983 and was finally resolved in 1990. The case arose when the mother of a child attending the McMartin Preschool suspected that her son, who was bleeding from the anus, had been sodomized by the son of the school's owner. She took the child to three different pediatricians, each of whom concluded that there was significant evidence of sexual abuse and notified the police (Cheit Reference Cheit2014, 24–25). After receiving a letter from the police department detailing possible criminal sex abuse at the school, other parents were understandably alarmed, and at least eight other families reported positive signs or suspicions of abuse (40). The district attorney then engaged a nonprofit agency to investigate, which carried out numerous interviews with very little supervision (74). In part because of massive media attention and in part for political reasons, the district attorney rushed to file multiple charges without adequately reviewing all the evidence (75–77). The McMartin case ultimately involved as many as forty‐one complaints against seven defendants but resulted in no convictions (charges against five were dismissed, two were acquitted, and the jury deadlocked on the charges against the remaining defendant).
Cheit concludes that the McMartin case did in fact involve egregious errors by the investigators and prosecutors, which resulted in many injustices. On the other hand, he also found that many of the allegations against the primary defendant were credible, based on inspection of the trial court record. In short, the case was not a witch hunt based on the fantasies of small children and their parents, but the legal system failed to sort out credible from incredible claims of abuse, thus both harming innocent people and probably allowing provable abuse to go unpunished (except by the excruciating legal proceedings themselves).
The media's response to the case was a simplistic one: children are not credible witnesses. A more appropriate message would have been that criminal justice actors need to be trained in how to effectively investigate and prosecute child sex abuse.
A Panoply of Other Cases
Cheit follows his detailed discussion of the McMartin case with an examination of twenty cases that are used as evidence of the alleged witch hunt in the press or books decrying this phenomenon, such as Tom Charlier and Shirley Downing's six‐part series Justice Abused: A 1980's Witch‐Hunt (Reference Charlier and Downing1988) and Debbie Nathan and Michael Snedeker's book Satan's Silence (Reference Nathan and Snedecker1995). Cheit's reaches this nuanced conclusion:
The claims of a national witch‐hunt are much stronger than the evidence that has been offered in support. There are no national data; instead, there are scores of comparisons to McMartin and various lists of people who were allegedly falsely accused or convicted. But the comparisons do not stand up to close analysis. There were clearly some poorly investigated and poorly charged cases, but fewer and far between what the witch‐hunt narrative claims. (Cheit Reference Cheit2014, 150)
Cheit continues with a chapter discussing well‐substantiated daycare abuse cases that took place both during and prior to the McMartin case, and also provides evidence that there was no atmosphere of hysteria or moral panic in the communities where the allegations of abuse arose (151–95). If anything, he argues, there was a continuation of the traditional response of denial of child sex abuse claims (177–85) even in the face of overwhelming evidence of abuse at several daycare centers (153). He acknowledges the great difficulties confronting the prosecution of child sex abuse, but also points out that child molesters who are clearly guilty have been dealt with extremely leniently by the justice system (185–95).
Cheit sees the Michaels case, tried in New Jersey in 1988, as the turning point of the witch‐hunt narrative. The defendant, Kelly Michaels, a twenty‐three‐year‐old woman working at a daycare center in Maplewood, abruptly left her position in 1985. Several days later, a four‐year‐old boy spontaneously told a medical assistant at his pediatrician's office who was taking his temperature anally that Kelly did this to him at school. Questioned by their parents, two more boys from the center said that Michaels had fondled their genitals (Cheit Reference Cheit2014, 204–12). An indictment was based on these allegations.
The New Jersey Department of Youth and Family Services then became involved, holding several parent meetings and ultimately interviewing virtually every child at the school. One particularly overinvolved investigator interviewed some children repeatedly, including those who had denied experiencing abuse, using suggestive and overly aggressive techniques (Cheit Reference Cheit2014, 212–22). The case then mushroomed, and ultimately went to trial with 165 criminal counts involving twenty children (226).
Nineteen children testified via closed‐circuit television from the judge's chambers, describing various sexual games Michaels had played with them (228). There was also evidence based on the children's behavior, especially hyper‐sexualized behavior, and medical evidence of genital soreness that would be expected from the conduct described (228–31). Michaels testified and denied all the charges (239), supported by three expert witnesses—one of whom was later revealed to be pro‐pedophile (236–37), one who testified that women did not fit the alleged profile of child molesters (237), and one who viewed the videotaped testimony of only two of the children and found them suspicious because of the children's lack of affect (238). The judge dismissed thirty‐eight counts and, after thirteen days of deliberation, during which they repeatedly reviewed videotapes of the children's interviews, the jury found Michaels guilty of most of the remaining counts. They rejected claims that seemed incredible and also charges by the children whose testimony appeared most likely to have been tainted (239–42).
Many journalists characterized Michaels's conviction as injustice resulting from the witch hunt. In the Village Voice, Debbie Nathan (Reference Nathan1988) argued that the investigative interviews all used leading and suggestive questions, an assertion that Cheit refutes (245–66). Another article two years later in Harper's discredited all of the evidence in the case, attributing the guilty verdict to a national panic akin both to Salem and to McCarthyism (Rabinowitz Reference Rabinowitz1990, 63). A Kelly Michaels Defense Committee was formed (Cheit Reference Cheit2014, 243). Other journalists and television programs began to spread the witch‐hunt narrative, with particular attention to the Michaels and McMartin cases (244).
The New Jersey Appellate Court reversed the Michaels verdict in 1993, completely accepting the arguments about repetitive interviews and child suggestibility and holding that the case could not be retried unless a “taint hearing” were held to assess the reliability of each child's statement (Cheit Reference Cheit2014, 269). The New Jersey Supreme Court sustained this decision in 1994; Michaels was released from prison and all charges against her were dropped (279–80).
Social scientists played a significant role in this drama. Two in particular, Stephen Ceci and Maggie Bruck, published articles emphasizing the suggestibility and unreliability of children's memories, using the Michaels case as an example (Cheit Reference Cheit2014, 271–76). The articles were picked up by the mainstream press and had a major impact on public opinion on this issue during the pendency of the Michaels appeal. In addition, Bruck and Ceci submitted an amicus brief in that appeal—styled the “Concerned Social Scientists’ Brief”—which I discuss in the next section.
The use of Social Science in Litigation
Social scientific research is clearly relevant to sound public policy and to legal rules and procedures as well. I have been a strong proponent of using social science findings in formulating, for example, the legal approach that should be taken to cohabitation in the United States (Bowman Reference Bowman2007). I have also written amicus curiae briefs in a number of cases involving sexual abuse, though from a legal, not social scientific, perspective. Yet the use of social scientific studies in the Michaels case is troubling to me. To explore why, I discuss in this section what is known from social scientific studies of young children about their capacity as witnesses regarding sexual abuse and then evaluate the impact of the “Concerned Social Scientists” brief in the Michaels case.
What Do We Know About Child Witnesses?
As noted above, children were traditionally regarded as unreliable witnesses; their testimony was either not admitted at trial or regarded with extreme suspicion and required corroboration (Klemfuss and Ceci Reference Klemfuss, Ceci, Holliday and Marche2013, 181; Odegard and Toglia Reference Odegard, Toglia, Holliday and Marche2013, 96–97). By the mid‐1980s, however, there were studies suggesting that children could in fact be credible witnesses, as good as—or at least no worse than—adults in many cases (Goodman Reference Goodman1984). At the time of the daycare cases, however, other studies appeared, specifically invoking the Salem witch trials and concluding that preschool children were particularly susceptible to misleading postevent information and exceedingly suggestible, out of a desire to please adult authority figures posing questions to them (e.g., Ceci, Ross, and Toglia Reference Ceci, Ross and Toglia1987, 38, 46). Many of these studies were based on experiments that were far from the context of child sex abuse. We now have the benefit of twenty‐five more years of increasingly good empirical work on this subject. In earlier experiments, children were shown videos and tested for recall while the experimenter attempted to manipulate their memories; later studies have focused on events more similar to the context of sex abuse—for example, young children's memories of intrusive medical procedures that involved genital and anal examinations or urinary catheterization (Goodman et al. Reference Goodman, Quas, Batterman-Faunce, Riddlesberger and Kuhn1994; Pezdek and Taylor Reference Pezdek, Taylor, Eisen, Quas and Goodman2002).
All witnesses—adults and children—have some trouble accurately recalling events, but the particular problems of testimony by children aged three to five are obvious. Their brains are immature, their vocabularies just developing, and their knowledge of sexual matters is extremely limited. To be competent to testify in court, a witness must have the ability to retain memory, to encode it, and to recall it within an acceptable range of accuracy.Footnote 1 Moreover, their lack of understanding of what constitutes sexual abuse leaves young children without a framework to interpret experiences of this sort, with a detrimental effect on accurate memory (Goodman et al. Reference Goodman, Quas, Batterman-Faunce, Riddlesberger and Kuhn1994, 287).
Research now shows that preschool children are capable of recalling events, especially distinctive events, with a high degree of accuracy (Pezdek and Taylor Reference Pezdek, Taylor, Eisen, Quas and Goodman2002, 171). In part because of the difference between their receptive and productive vocabularies (i.e., they can understand many more words than they can produce), they are unlikely to communicate this information in the form valued in the courtroom (Saywitz and Lyon Reference Saywitz, Lyon, Eisen, Quas and Goodman2002, 86–88). If questioned with what might be considered leading questions to prompt their recall, however, they are capable of giving accurate information about sexual abuse (Pezdek and Taylor Reference Pezdek, Taylor, Eisen, Quas and Goodman2002, 189).
A major downside of leading questions is, of course, suggestibility. Children are reluctant to say “I don't know” and will thus attempt to answer questions they do not understand; and they defer to authority figures who question them—parents, policemen, and prosecutors (Klemfuss and Ceci Reference Klemfuss, Ceci, Holliday and Marche2013, 190). Yet many recent experiments demonstrate that children are also able to resist suggestions and thus are more impervious to the implantation of false memories than previously thought: “[A]lthough it is relatively easy to suggest to a child a change in a touch that he or she actually experienced, … it is less likely that a completely new touch can be planted in memory …” (Pezdek and Roe Reference Pezdek and Roe1997, 103). Moreover, adults are also subject to manipulation and inaccurate memories—or memories that combine accurate and inaccurate elements—yet we rely on the finder of fact and our adversarial legal system, with its engine of cross‐examination, to sort out these problems.
In short, many problems arise in introducing the experience of small children into the courtroom. Children's need for cues to recall has led to the relaxation of rules against leading questions with child witnesses (Goodman Reference Goodman1984, 15), but there are indications that jurors find testimony that appears to follow suggestions less credible (Klemfuss and Ceci Reference Klemfuss, Ceci, Holliday and Marche2013, 200). Moreover, jurors expect to hear a coherent narrative, a story, accompanied by detail about not only facts but also evaluative and emotional responses; that is what they consider to be credible (Lyon et al. Reference Lyon, Scurich, Choi, Handmaker and Blank2012, 448). Children are unlikely to offer this type of detail without being explicitly asked for it, by the use of “How did you feel when …”– type questions, for example (Lyon et al. Reference Lyon, Scurich, Choi, Handmaker and Blank2012, 453–54).
What lessons should we draw from all these studies? Certainly not that preschoolers are incapable of providing reliable testimony about sex abuse. Yet the fact remains that many of the young children who are victims of sexual abuse are very difficult to interview and obtain coherent answers from. Conflicting interests collide here. Protection of children, the due process rights of criminal defendants, and the integrity of the justice system are all at stake. To some extent, one's response to the difficulties may depend on how one prioritizes these goals.
Is there some way to optimize all these important values? One obvious task—one to which many social scientists have responded—is to attempt to understand how to improve the questioning of children about sexual abuse, both in and out of court. Recent articles and books about how to investigate and present these cases include many useful suggestions, such as how to “cue” a child to recall without suggesting the answer desired, what kinds of questions to use and to avoid, and how to improve a child's recall while at the same time reducing the susceptibility to suggestion (Goodman and Bottoms Reference Goodman and Bottoms1993; Saywitz and Lyon Reference Saywitz, Lyon, Eisen, Quas and Goodman2002, 96–105)—for example, by telling the child to “Tell me as much as you can remember about what really happened, even the little things, without guessing or making anything up” or saying that it is all right to say “I don't know” (Saywitz and Lyon Reference Saywitz, Lyon, Eisen, Quas and Goodman2002, 98, 105–06).
Repeated interviews are inadvisable, for reasons related both to their tendency to suggest the questioner's desire for different answers than a child has previously given (Odegard and Toglia Reference Odegard, Toglia, Holliday and Marche2013,110–11) and to avoid traumatizing the child witness. Yet more than one interview and repeated questions may be unavoidable for persons charged with child protection, given the immediate need to protect children who may be in danger of abuse. As Saywitz and Lyon conclude, “an interviewer may be forced to ask more specific questions in order to make necessary decisions about safety and protection in a given case” (Saywitz and Lyon Reference Saywitz, Lyon, Eisen, Quas and Goodman2002, 90). Videotaping the initial interview is a good idea, to cut down on the number of interviews necessary, as well as to preserve the evidence soon after the abuse took place.
One important finding of the research is that the use of anatomically correct dolls to ask young children about abuse of their bodies is contraindicated. Dolls in general—and these dolls in particular, given children's naiveté about human bodies—fascinate young children, both those who have been abused and those who have not. Their curiosity and desire to play with the dolls seems to produce inaccurate information (Pezdek and Taylor Reference Pezdek, Taylor, Eisen, Quas and Goodman2002, 181; Odegard and Toglia Reference Odegard, Toglia, Holliday and Marche2013, 111–12). Although such dolls may have some use in therapy, they are not appropriate for use by law enforcement personnel for that reason.
The “Concerned Social Scientists Brief” in the Michaels Case
As I mentioned earlier, in 1993 Maggie Bruck and Stephen Ceci, two well‐respected social scientists, coauthored an amicus curiae, or friend of the court, brief in the Michaels appeal.Footnote 2 In it they summarized recent studies showing the suggestibility of young children and the effects of interviewer bias, of repeated questions, of the emotional tone of the interview, of repeating misinformation across interviews, of peer pressure on children's reports, of children's tendency to defer to interviewers of high status, of stereotype inducement and source attribution errors, and of using anatomically detailed dolls. The summary accurately describes critical studies available at that time, but as a legal brief, it does not give serious consideration to extant research showing that children can provide accurate information about sexual abuse and that they can resist leading and suggestive questions in this context (Lyon Reference Lyon1995, 435). Instead, the brief presents its views on suggestibility and inaccuracy of child testimony as though they represented a social scientific consensus.
Significant problems arise when Bruck and Ceci apply their conclusions to the specifics of the Michaels case, using excerpts from transcripts in the case to demonstrate the ill effects of each of the practices listed above. However, because they did not have access to the transcripts themselves, which were sealed at the time, they relied on edited excerpts provided by Michaels's defense attorney, Robert Rosenthal. This led to significant errors in their characterization of the trial record.
When they subsequently published Jeopardy in the Courtroom, a book that relied heavily on the Michaels case, Ceci and Bruck included more of the excerpted transcript from the Michaels case (Ceci and Bruck Reference Ceci and Bruck1995, 116–18). Cheit, who later obtained access to the entire transcript in the case, carefully compares the excerpts included with the complete transcript, and demonstrates how sections of it have been omitted—sometimes pages’ worth—to make investigators’ questions appear to be leading or suggestive when inclusion of the omitted portions show that they are not (Cheit Reference Cheit2014, 255–61). Examples include omitting open questions, like “Where did this touching take place?” (after the child has said there was touching), and eliding a series of questions about the instrument with which the touching occurred, thereby concealing that the investigator in fact asked a series of questions, with the child rejecting all the misleading suggestions (257):
Q: Did she touch them with toys?
A: No.
Q: Did she touch them with telephones?
A: No.
Q: Did she touch them with spoons?
A: Yes.
Q: What kind of spoons?
A: Teaspoons.
The edited version, by contrast, read as follows (256):
Q: Did she touch them with telephones? Did she touch them with spoons? What kinda spoons?
A: Teaspoons. [Plastic spoons and other utensils were used at the preschool (263).]
In short, a series of questions, with the child alternating between refusing to be led to a suggested answer and giving a positive answer, followed by an open‐ended question (“What kind of spoons?”), to which the child responded with a more detailed answer, has been shortened to read as a single question that gives the impression the investigator pressed the child to give the desired answer.
Although it can be helpful to provide the court with a balanced overview of social science research (better yet if it is presented at trial and thus subject to cross‐examination), applying laboratory studies to an actual case in an appellate brief is problematic. The “Concerned Social Scientists” brief did not provide information, for example, about whether the children allegedly subject to unreliable modes of questioning were in fact the ones whose cases went to trial and, more important at the appellate level, whether they were the ones the jury believed or disbelieved. Moreover, the brief makes unsupported claims that the interviews on which it focuses are characteristic of the manner in which all the evidence was obtained in the case (Cheit Reference Cheit2014, 273), when the authors had no way of drawing this conclusion without reviewing the entire trial record. Cheit, who has reviewed all the evidence, shows that the single interview excerpted was very unusual in a number of respects, leading them to an incorrect conclusion.
While the amicus brief accuses the investigators in Michaels of assuming the existence of abuse, it appears, Cheit argues, that Ceci and Bruck had come to assume the opposite: that abuse never happens in a daycare setting, or at least that there can never be sufficient evidence to prove it. For example, another case, tried in 1984–1985 in Florida, involved careful investigative techniques, as well as medical evidence of abuse, including STDs, and a history of sexual assault on the part of the primary defendant; in addition, that defendant's wife pled guilty and testified against her husband (Cheit Reference Cheit2014, 283–325). The jury had no trouble finding the defendant guilty, a verdict that was upheld throughout the Florida appellate courts and by the federal courts on habeas. Yet Ceci and Bruck included this case in their 1995 book, twinning it with the Michaels case as an example of the witch hunt (Ceci and Bruck Reference Ceci and Bruck1995, 122–23, 136–37, 146–49, 259–61), and Bruck filed an affidavit in the defendant's federal appeal (Cheit Reference Cheit2014, 327). Cheit describes this as an example of something he calls “disconfirmation bias.”
What conclusions should be drawn about amicus briefs from this example? As noted above, I myself have drafted and filed amicus briefs in sex abuse cases; I have also signed on to many amicus briefs advocating same‐sex marriage as part of a group of family law professors. I believe that amicus briefs can be helpful to a court in a number of circumstances, especially where background information is necessary to understand some aspect of a case. The Michaels brief falls into this category, offering a summary of social scientific findings about child witnesses. If it stopped there, I would put it into a category with helpful amicus briefs such as the “Historians’ Brief” filed in abortion cases in the Supreme Court, which discussed the historical treatment of abortion in the United States (Amicus Brief 1992).
But a problem arises when social scientists try to apply their findings to the evidence in a particular case. Perhaps if they have reviewed that evidence in its entirety I would think differently, but I am not sure; this type of argument is the realm of the lawyers. In the Michaels case, the defense presented three expert witnesses at trial, whom the jury did not find convincing. The “Concerned Social Scientists” brief on appeal essentially gave the defendant a second bite at the apple by injecting a lengthy discussion by two even more reputable social science experts, who were not subject to cross‐examination. This is not an appropriate use of an amicus brief on appeal.
The Mass Media and Public Opinion about Child Sex Abuse
The mass media had a field day with the daycare cases and filtered reports of the social science accompanying them, disseminating not only the facts but also the message that young children were incurably suggestible and unreliable as witnesses about sex abuse. Cheit analyzes the swing in mass media coverage and opinions about the credibility of children's accounts of child sex abuse in the 1980s, starting from a balanced 1984 report about the problems of child witnesses in the New York Times, which described studies concluding that children's memory was not that much worse than that of adults (Goleman Reference Goleman1984). Next came the onslaught of articles published at the time of the McMartin and Michaels cases adopting the witch‐hunt narrative and impugning the testimony of children. These were followed by television coverage on shows like Frontline and movies such as Indictment and Capturing the Friedmans (Cheit Reference Cheit2014, 8–11, 130–33), presenting innocent persons as having been railroaded on imagined charges or ones implanted in memory by police or investigators—a type of show that still seems to fascinate the viewing public.Footnote 3
I am accustomed to inaccurate mass media reports. I have seen this tendency in cases I cared about, witnessed, or been close to.Footnote 4 By now, I have also seen how journalists do their research, often just contacting someone like me and asking for comments on a legal situation, in lieu of doing independent research at all. After all, they are working against a deadline and do not have time to become experts before they write. It surprised me, therefore, that even my own memories of the daycare cases Cheit discusses, all derived from press accounts, were that they involved fantastic elements and were incredible. Now that I have read his book, I realize that many allegations that appear at first glance to be fantastic may in fact be either true as descriptions of a child's experience, in a young child's way of perceiving and communicating that experience, or simply mischaracterized (Cheit Reference Cheit2014, 338–40). For example, Ceci and Bruck discredit one case by saying that it included fantastic allegations such as riding on sharks, when the interview transcripts show that the boy being interviewed was discussing a Jacques Cousteau television program he had seen (339).
The problem of media distortion and 180‐degree swings of opinion is particularly acute in the area of child sex abuse. One can see this, for example, in comparing early versus more recent media coverage of the clergy sex abuse scandals in the Catholic Church. As one recent article in the Daily Beast described: “Newsrooms turned on a dime” (Berry Reference Berry2014). The clergy abuse cases primarily involved testimony by adult males about remembered abuse, although some of it was subject to delayed recall, which is itself controversial,Footnote 5 but the young men's testimony did not invoke any of the traditional suspicions about the testimony of women and children about sexual matters. Elizabeth Mertz and I have documented a similar swing in media coverage of cases relying on recovered memory by women, concluding that it appears difficult for media reports to convey the more balanced, and more accurate, social science conclusion that there are both accurate and inaccurate claims of remembered sexual abuse (Bowman and Mertz Reference Bowman and Mertz1996, 618–22).
Much of the black‐and‐white character of mass media coverage of social issues can be explained by its purposes. The media publish what will sell to mass audiences; public attention is short, and dramatic stories attract attention. The journalists who presented the witch‐hunt narrative in a dramatic fashion won some of the most coveted prizes in their fields for the articles they published (Cheit Reference Cheit2014, 282). By contrast, a detailed social scientific report is unlikely to capture the public attention or reach a very large audience at all unless it relates to events currently in the news or presents a conclusion that startles by its contrast to the conventional wisdom on a subject.
Legal cases on lurid subjects have long been the subject of headlines. Reporters stop by the courthouse to assess whether any new stories might present themselves. This, of course, presents ample opportunity for capture of the media by litigators, who have grown accustomed to issuing press releases, holding press conferences, and regarding the jury of public opinion as one of their responsibilities. Yet the immense power of mass media over public opinion gives reason for concern because it is media outcry, rather than social scientific research, to which public policy is likely to respond. The results, as described below, are frequently unconsidered swings in policy, with unanticipated and sometimes counterproductive consequences.
The Relationship Among Media Coverage, Social Science Research, and Public Policy
When an egregious case or social problem is publicized in the press, policy makers come under pressure to do something about it. Added to this tendency, in the daycare and other sex abuse cases, the academics, journalists, and lawyers who supported the defendants also organized politically. Groups and Web sites devoted to “false allegations” sprang up in response to media coverage of cases, offering assistance to persons charged with sexual abuse of children (Cheit Reference Cheit2014, 374, 387–90). This sometimes created odd bedfellows, as individuals and groups more generally opposed to government intervention in “private life” joined with the false allegation groups to oppose the Child Abuse Prevention and Treatment Act (401). A “Day of Contrition” was held in Salem, Massachusetts on January 14, 1997, 300 years after the witch trials there, featuring many of the journalists and academics who had become associated with the “false memory” movement, along with defendants from the daycare cases, including Kelly Michaels and the McMartin defendants (389).Footnote 6
The impact of the witch‐hunt narrative on legal policy has been detrimental. While the law had supposedly moved beyond denial of the existence of child sex abuse and the traditional disbelief of children, the witch‐hunt narrative reaches similar results by other means, such as taint hearings and disbelief of children's accounts on the basis of unrepresentative social scientific evidence.
In New Jersey since the Michaels case, a defendant can request a pretrial taint hearing requiring the prosecution to show that a child's testimony has not been “tainted” by coaching or repetitive interviews and thus should be admitted at trial. Taint hearings go far beyond the competency hearings that may be required of witnesses more generally, which are held to determine whether a witness is incapable, perhaps because of youth or mental disability, of giving intelligible testimony (Klemfuss and Ceci Reference Klemfuss, Ceci, Holliday and Marche2013, 179–87). Taint hearings have been detrimental to the prosecution of child sex abuse cases in particular, by complicating and delaying the case and allowing cross‐examination prior to trial.
In one case, in which there was an eyewitness to the sexual abuse of a seven‐year‐old boy, the court was persuaded to hold such a hearing. It took more than two years, and Maggie Bruck testified as an expert witness criticizing the investigation on behalf of the defense. Although the court eventually ruled that the child's evidence was not tainted and was thus admissible, the experience was very negative for the abused child and his family, who refused to cooperate with any further prosecution, causing the case to be dropped. The defendant moved out of state and was later charged with nine counts of sexual abuse of other children; he pled guilty to possession of child pornography and child molestation (Cheit Reference Cheit2014, 390–93).
There is a stark contrast between the reluctance to believe child sexual abuse victims, on the one hand, and our society's embrace of extremely punitive responses to convicted sex offenders, on the other. What the two responses share, however, is their origin in the tendency of public policy to react quickly to cases in the media, with ill‐defined and untested “solutions.” Here I have in mind Megan's Law and the multitude of state laws requiring sex offenders to register and providing for community notification of their presence in an area. Quick passage of such laws made it look as though politicians were coming down hard on this type of crime, but in reality these laws have been not only ineffective but also unjust and sometimes self‐defeating. Recent research has found that sex offender registration and notification laws have the perverse effect of discouraging sexual assault reporting, driving down plea bargains, and increasing acquittals, among other problems (Corrigan Reference Corrigan2013, 218–35).
If dealing with child sex abuse in the criminal justice system presents so many problems, should we rely on civil remedies for child sex abuse instead? I consider this question in the next section.
Criminal Versus Civil Remedies for Child Sex Abuse
Although each system has advantages and disadvantages, there is a place for both the civil and criminal justice systems in addressing the problem of child sex abuse, especially in the institutional settings under examination in Cheit's book. Moreover, they are not alternatives. Indeed, a child (through his or her guardian) cannot simply bring a civil lawsuit against a sex abuse offender without triggering the involvement of other agencies. If the school is notified, or if a pediatrician sees signs of abuse, mandatory reporting laws come into effect; both teachers and health personnel have an obligation to report suspected abuse to the state. Of course, this might mean only that an investigation by the state department of children's services is launched, at least until the agency decides to bring formal criminal charges. The experience of the daycare cases Cheit describes indicates that the personnel in these state child welfare agencies may be less well trained than law enforcement agencies about how to appropriately question children without suggesting answers to them.
Civil remedies have a number of advantages over criminal prosecution. First, they are controlled by the party rather than the state. The upsides of this fact are autonomy and control over the litigation; the downside is the necessity to pay legal fees, unless one can obtain representation on a contingent‐fee basis. Second, there is a different standard of proof in civil cases—preponderance of the evidence instead of “beyond a reasonable doubt”—and fewer due process protections for the defendant because liberty is not at stake. Thus, civil cases should be easier to prove; the downside for a plaintiff and for society is that they do not take a sex abuser out of circulation. Damages, of course, can be useful, allowing not only for deterrence but also for the victim to have the ability to purchase therapeutic services, change schools, or take other steps to alleviate the effects of the abuse. Depending on the financial status of the offender, however, damages may not be effectively available at all, although even an uncollectable award can give the satisfaction of having been listened to, believed, and vindicated.
Civil suits have played a large role in recent sex abuse litigation, such as that brought by victims of past clergy abuse in the Roman Catholic Church. Adult survivors of childhood sex abuse can bring suit for damages either when they attain the age of majority or, with the extension of statutes of limitations in cases of delayed recall, upon recovery of the memory (Bickel Reference Bickel1991). The victim gets to tell his or her story; and, if the case is publicized, the offender is exposed and others put on notice that they can be called to account for similar conduct (Bowman Reference Bowman1998, 1481). Public attention is also focused on the problem of child sex abuse; and with enough cases, institutions begin to reform, as is apparently happening in the Catholic Church. Finally, the child is spared the trauma attendant upon a criminal prosecution, with its particularly aggressive cross‐examination; our Constitution limits the extent to which a court can protect a child from the pain involved in participating in such a case. Moreover, a child may not want to send his or her abuser to jail, especially if the offender is a family member or other loved person.
But what is lost in civil litigation that criminal prosecution may provide? If an abused child's parents were to bring a civil suit soon after the abuse takes place, the offender would be exposed and presumably removed from any position working with children; however, he or she would not be incapacitated from sexually abusing other children by being placed in a secure setting and/or ordered into sex offender therapy. Moreover, when civil lawsuits are brought by adult survivors of childhood abuse long after the abuse took place, although the experience may be empowering for the plaintiff, a great deal can be lost in the interval between the abuse and the lawsuit, especially if the child did not report the incident and receive appropriate treatment.
All this poses very difficult problems for society and the legal system, ones that can be addressed only in light of the values and goals we are seeking to effect, and then only imperfectly. The traditional goals of punishment, compensation, and deterrence point in different directions. Compensation clearly favors a choice of civil remedies, but they may be inadequately deterrent, given the tendency of sexual abuse offenders to repeat their patterns. Only the criminal justice system offers the possibility of taking offenders out of circulation, but at the cost of treating them overly punitively upon release, such that they are often unable even to find a place to live. And it is only the criminal law that articulates society's response to child sex abuse, through its function of prescribing and adjudicating exactly what the community sees as unacceptable behavior. However, it has also become abundantly clear, in this and other settings, that innocent people can be convicted of crimes.
This is a real dilemma: we have people in prison who do not belong there; we have child molesters being treated overly punitively; and we have children (and the adults they become) who have been molested who are not believed or protected. There is no easy answer.
One way to address this issue might be to ask what the appropriate courses of action are, at every level, so as both to deter and to punish the sexual abuse of small children, which undeniably does happen in some institutional settings. For example, technological options are rapidly becoming more sophisticated and affordable. Might state regulations require that devices such as closed circuit TVs or wrist voice‐recorders on children be employed in daycare centers to deter abuse and/or to provide evidence if it does occur? Then, once there is an allegation of abuse, the offender should be removed from the setting, but his or her rights demand a prompt and speedy legal proceeding—within the limits of due process—lest reputations and businesses be ruined during the pendency of the legal process. In the end, we may have to trust our adversarial system to discern between accurate and inaccurate claims, as we do in other types of cases, while holding the courts and other agencies to the highest of standards.
At least one thing appears clear from Cheit's book. Social scientists and lawyers must keep working on how to effectively investigate and prosecute cases of child sex abuse, and personnel at every level of the system should be trained in the best methods to elicit information from children without unacceptable damage to the rights of an accused person. In addition, continued research—and perhaps borrowing models from other countriesFootnote 7 —to determine the types of legal and institutional structures that can best handle these cases, as well as to reduce their incidence, is imperative.
Conclusion
The public message of the witch‐hunt narrative—that children are not credible witnesses—obviously played into the hands of, if it was not orchestrated by, groups that support persons who allege they have been falsely accused of child sex abuse. By examining in detail the evidentiary record of the daycare abuse cases, Cheit provides substantial evidence that there are situations where children's accounts should be believed. He also shows how the witch‐hunt narrative affected both public opinion and judges, making them exceedingly skeptical of the testimony of children about sex abuse. This is a throwback to an era when a child's testimony required corroboration, a requirement that had been abolished based on social science research and the movement to address the problem of child sex abuse.
Cheit's book is important for presenting a basis upon which to move forward. It shows the way that cases can burgeon out of control, starting with a real perpetrator and victims and then growing so as to include both children and accused persons who were not part of the problem, thus creating a situation where the initial victims are disbelieved. It elucidates what was done wrong in some cases and must be avoided. It also provides an important caution against making public policy based on the swings of media accounts and public opinion. Reading it reminds me how important it is for social scientists and legal scholars to do the kind of careful and time‐consuming research Cheit has done—to examine claims, to interrogate the effectiveness of remedies for this problem, to publicize those findings, and to seek out the best possible ways for the legal system to incorporate this knowledge and address the problem of child sex abuse.
