Opinion

Introduction

At a critical hearing in this juvenile dependency case, the juvenile court judge was under a serious misimpression about what was at stake. The hearing concerned a subsequent petition which alleged that the father, Rogelio, had molested his three-year-old daughter Daisy. The judge, however, mistakenly believed that the matter had already been decided against Rogelio. At the outset he thought the hearing was a simple six-month review; he was not disabused of his misimpression until after the evidentiary portion of the hearing had been completed. The record also indicates that the judge had not read the subsequent petition alleging the molestation. Nor was any oral testimony concerning the alleged molestation presented at the hearing. The only evidence that might have supported a finding that Rogelio had molested his daughter would have to have been painstakingly extracted from a series of lengthy reports written by a social worker.

The court sustained the petition at the conclusion of the hearing and ordered a forensic psychologist to do an in-depth study of the entire family, including Rogelio’s two stepdaughters. A few months later the psychologist exonerated Rogelio of any child molestation, or any tendency toward child molestation. Even so, the fact the court had sustained a petition alleging child abuse remained; consequently, at an 18-month review conducted by another judge, the juvenile court determined that—without ever examining whether any molestation ever really occurred—it would be detrimental to return the children of Rogelio and Blanca to them. From the orders made in the wake of that review, Blanca has brought this writ proceeding, contending that the evidence is insufficient to support the detriment finding.

As we explain below in greater detail, the only support for the detriment finding is the fact that the juvenile court sustained the subsequent petition alleging child molestation. However, after examining the record of the hearing at which the petition alleging molestation was sustained, we have concluded that a new 18-month hearing must be held, one which deals squarely with the question of whether any molestation was ever committed in the first place.

Facts

In April 1994, the minors Nora, thirteen, Juan, seven, Daisy, three, and Isaías, one, the children of Rogelio and Blanca P., were detained at Orange-wood Children’s Home after Juan was struck in the face several times by his mother, petitioner Blanca, hard enough to leave finger marks. The children were made dependents of the juvenile court based on an initial petition charging excessive corporal punishment and were eventually placed in foster care.

During that summer Daisy’s foster parent noticed a vaginal opening the foster mother herself opined was too large for a three-year-old. A child abuse services team (CAST) interview was then conducted with the child.

She was asked if anyone touched her “pee-pee.” The girl first answered, “my mom.” Daisy was asked again. This time the answer was “boy.” She was asked a third time. Having exhausted “mom” and “boy,” she answered, “yes, my Pappy.” On October 17, subsequent petitions were filed alleging that Daisy’s father, Rogelio, had, prior to the detention, digitally penetrated her vagina.

The hearing on the subsequent petition was not held until almost eight months later, on April 12, 1995. The hearing was a short one. The evidentiary portion took fewer than 30 pages of trial transcript (which, for readers not familiar with the size of reporter’s transcripts, is quite short). The trial judge was not the same judge who would later hear the 18-month review. We describe what happened in detail.

The deputy county counsel introduced a variety of reports written by the social worker assigned to the case. Rogelio’s attorney then cross-examined the social worker and elicited the comment that the social worker still recommended reunification. However, the social worker also stated the children should not be returned immediately because there was an “allegation of sexual abuse” and the parents had not “accept[ed] that the molest ha[d] occurred.” The social worker further testified the parents had only monitored visitation because “in the past mother has told the minors not to say, you know, what happened in the home in the past because they won’t ever go back home and issues of that nature.”

The social worker was also asked whether Blanca believed “Daisy’s allegations against the father,” and the social worker’s answer was that the mother waffled. The social worker said Blanca said there “might be a possibility,” but, then again, she would “later” say “it didn’t happen.”

After the court ascertained the social worker spoke Spanish, the father was called to the stand. An interpreter was needed. Rogelio’s attorney asked him directly: “And since the day that Daisy was bom, have you ever taken your finger and inserted it into Daisy’s vagina?” Rogelio’s answer: “Never.”

On cross-examination, Rogelio had to admit that there were no other “male figures” that took care of Daisy in the first three years of her life, and that he was called “pappy” at home. Rogelio’s cross-examination ended with a colloquy initiated by the deputy county counsel’s question, “Do you and your wife teach Daisy to tell the truth?” Rogelio tried to explain that “we didn’t teach her because she would not understand.”

The evidentiary portion of the hearing concluded with the judge asking a number of questions of Rogelio. “It says in the report that you have forced yourself upon your wife, forcing her to have sexual intercourse with you even if she did not want to. [<30 Have you ever done that?” Answer: “Not that I remember. You can ask her.” Next, “Mr. P[.], it says that there have been times when you have withheld money from your wife so that she could not feed herself or her children. [TD Do you agree that it is your responsibility as a man, as a husband and as a father to provide money for the mother to feed herself and the children?” Answer: “I never took it away, I always give her as much as I could.” The court then struck that answer as nonresponsive, and asked the father, “Do you agree that you, as a man, as a father and as a husband have a responsibility to provide money to the mother for her to feed herself and her children.” This time: “Of course, yes.”

At this point, the trial judge said something very strange.

“Do you understand, Mr. P[.], that today is not a trial to determine if you have done those things, if you have, in fact, done this to Daisy[;] that you are asked whether you have hit your wife, hit your children or supplied them with money for food, that is not what we are deciding today, because that has already been decided. [

After Rogelio responded that, “well, [he] believe[d] so,” the trial judge continued: “Do you understand that we in this system, the court, the social services are acting as if these things did happen, they have—it has already been established that those things have happened, are we not of [sic] a six-month review?” (Italics added.)

The deputy county counsel spoke up. “Yes, but we are also here on a subsequent petition.”

The judge: “I’m sorry?”

The deputy county counsel then disabused the judge of his evident misimpression of the nature of the proceedings. “For Juan, Daisy and Isaías and that’s on the allegation the minors’ father digitally penetrated the minor sibling Daisy.”

The judge corrected himself. “I beg your pardon. [*][] Mr. P[.], I was wrong. We are trying to determine if, in fact, you did this to Daisy. However, with regard to the other issues, they have already been established. So I was wrong, we are here to determine this issue with regard to whether you inserted your finger into the vagina of Daisy. So I apologize for that statement. [CJ[] We do, though, need to understand each other so we can work together to bring, to reinforce your relationship with your children and with your wife. I’m sorry. [U I have no further questions.” (Italics added.)

The judge asked if any of the attorneys wanted to present any more evidence. Then he made another strange comment.

“Just give me a moment again, please. I hadn’t even looked at that supplemental petition, and then momentarily forgot.” (Italics added.)

After the deputy county counsel pointed out to the judge that the hearing was on a “subsequent,” as distinct from “supplemental” petition, the judge rendered a tentative decision, finding “based upon the evidence presented by a preponderance of the evidence that the allegations have been established with regard to Daisy and then by relations to the siblings to them.” He elaborated about the parents: he knew both parents loved their children, but said that he thought Rogelio’s two older stepdaughters (Nora and Fabiola) were “afraid he will molest them like they believe he has molested Daisy”; accordingly, the children should not be returned. He also noted Rogelio’s and Blanca’s involvement in therapy, but expressed his concern that they were not “really putting out and discussing their feelings.” They were attending therapy sessions “without participation, and that is not particularly helpful.”

The deputy county counsel then mentioned that she was requesting an expert evaluation under Evidence Code section 730 to “help to focus on what would need to be done.” The judge thought it was “a great idea for both” parents. After oral argument the judge adopted his prior statements, and found that “the allegations in the subsequent petitions in each of these matters have been sustained.”

That summer Rogelio and Blanca faithfully continued with their therapy and compliance with their reunification plan. In August Dr. Jose LaCalle filed his Evidence Code section 730 evaluation. Dr. LaCalle’s report exonerated Rogelio of any propensity to sexually abuse children:

The psychologist also noted that both stepdaughters of Rogelio, Nora and Fabiola, independently denied that Rogelio had ever molested them, and, indeed, spoke well of the man. Nora, according to Dr. LaCalle, stated “Rogelio has always treated her well. It is her mother who is the disciplinarian in the home.” Nora further volunteered, “I think [the social worker] does not want us to go back home.” Fabiola said Rogelio “is a nice man; he is working all the time doing yard work.” And “[s]he added that she does not believe that her dad ever molested Daisy and she knows that her dad never did anything wrong to her.”

Dr. LaCalle opined that much of the family’s troubles were the result of a lack of rapport with the social worker assigned to the case. He related Blanca’s statement concerning the social worker: “[The social worker] humiliates us. She called us illiterates. She laughs at me when I cry. She told me that my husband had raped my Daisy, and that is not true. I am the one that changed her, bathed her everyday, changed her panties, and I never noticed blood in them.” He also related Rogelio’s belief that his wife had been told by social workers “that if she leaves him, she might have a chance to get her children back.”

Dr. LaCalle concluded: “There is no rapport between [the social worker] and the P[.] parents. Without reflecting in any manner on the performance of [the social worker], a new social worker should be assigned to this case. The mistrust of Mr. and Mrs. P[.] toward [the social worker] is too deep and certainly will interfere with any attempts at positive family intervention. . . . The P[.] family perceives SSA, at the present time, as an agency of the government determined to ‘steal’ their children away."

The assigned social worker was not replaced; in fact, she prepared the report for the 18-month review. She noted that Blanca was willing to “ ‘accept the lie’ ” that her daughter was molested, but had not “totally accepted” the purported “fact.” Then again, Blanca was reported to have said (the report is not clear to whom, but presumably it was the social worker herself) there was a “possibility” the father could have molested Daisy. Then yet again, Blanca was not “fully convinced about the minor Daisy’s sexual abuse” and therefore, in the social worker’s opinion, would be “unable to protect her.” Rogelio was reported to have stated (apparently to either his therapist or the social worker or both) that he was “willing to accept that he molested Daisy as long as he is not prosecuted," because he was applying for permanent residence in the United States.

While the parents had attended numerous counseling and therapy sessions, the social worker did not believe that the parents had sufficiently “internalized” proper parenting skills. Their “pain of not having their children with them” had “unabled [sic] them to gain insight about their past behavior.” Both parents were “not willing to make any changes until the other party approves it. Furthermore, they seem to relate to the world in a stance of them against the rest of the world.”

Nor was the social worker impressed with Rogelio’s continued refusal to admit he molested his daughter. She wrote of how “the parents’ denial about Daisy’s sexual abuse in session with minors Daisy and Juan present was [not] in the minors best interest.” She elaborated, “The minors’ father denies he molested Daisy. It was expected that he would assume responsibility for the molest, demonstrate remorse for such inappropriate behavior, and learn skills to prevent similar incidents from re-occurring. Thus far, since such progress never occurred, he continues to pose a risk to any child placed under his custody.”

The 18-month review was finally heard in January 1996. As related above, the hearing was before a different judge than the one who adjudicated the subsequent petition. The court took judicial notice of the minute orders made after the April 12, 1995, hearing, but otherwise the question of whether Rogelio had really committed any child molestation was not litigated. Rather, the evidence focused on how Blanca and Rogelio reacted to reunification services in light of the assumption that Rogelio had committed child molestation. In that regard the social services agency presented evidence that Blanca and Rogelio had not made sufficient “progress” in therapy and had not “internalized” proper parenting skills, based mostly the couple’s refusal to admit, in therapy, that Rogelio had molested Daisy. When asked what “issues” should have been, but were not, addressed in therapy, the social worker testified: “Well, in terms of the sexual abuse, the parents have not acknowledged—they are not convinced, or at least the mother, and she is not able to protect. And father has not assumed responsibility.” And when asked why she thought there was risk of physical and sexual abuse, the social worker reiterated the point that “the father has not assumed responsibility” and “the mother does not believe.” The new judge, without elaboration, found it would be detrimental to return the children to the parents and reunification services should be terminated. The matter was set over for a Welfare and Institutions Code section 366.26 hearing for May 1996. In the meantime, Blanca’s counsel filed this writ petition. There is no dispute that Rogelio and Blanca have complied with their service plans by all objective measures.

Discussion

The Evidence Apart From the Alleged Molestation Cannot Support the Detriment Finding

We first consider whether the evidence apart from the alleged molestation can support the detriment finding. If so, the obvious problem of the reliability of the molestation finding in light of the prior juvenile judge’s misimpression about the nature of the hearing and Dr. LaCalle’s subsequent exoneration need not be dealt with. Any error would be harmless.

At the 18-month review hearing the dependent child must be returned to his or her parents unless it would be detrimental to do so; further, the burden is on the state to show such detriment.

When a child is removed from the family home, a plan is usually developed to allow his or her return. The purpose of the plan is to overcome the problem that led to removal in the first place. (See, e.g., In re Samkirtana S. (1990) 222 Cal.App.3d 1475, 1488 [ 272 Cal.Rptr. 489 ].) In determining whether it would be detrimental to return the child at the 18-month review, the court must consider whether the parent participated regularly in any treatment program set forth by the plan, the “efforts or progress” of the parent, and the “extent” to which the parent “cooperated and availed himself or herself of services provided.” (§ 366.22, subd. (a).)

In deciding whether it would be detrimental to return a child, the easy cases are ones where there is a clear failure by the parent to comply with material aspects of the service plan. In Angela S. v. Superior Court (1995) 36 Cal.App.4th 758, 763-764 [ 42 Cal.Rptr.2d 755 ], for example, a mother continued to test positive for illegal drug use, continued to move from place to place, failed to “regularly” attend therapy, and failed to complete her parenting class. This was obviously enough to support a finding of detriment.

The harder cases are, like the one before us, where the parent has complied with the service plan, but for some reason has not convinced a psychologist or social worker that it would be safe to return the child to the parent. The problem is not, as it were, quantitative (that is, showing up for counseling or therapy or parenting classes, or what have you) but qualitative (that is, whether the counseling, therapy or parenting classes are doing any good). These are sensitive cases, fraught with emotional overtones, because they invariably deal with an evaluation of the personality, character and attitudes of the parent.

Three cases, In re Jasmon O. (1994) 8 Cal.4th 398 [ 33 Cal.Rptr.2d 85 , 878 P.2d 1297 ], In re Brian R. (1991) 2 Cal.App.4th 904 [ 3 Cal.Rptr.2d 768 ], and In re Heather P. (1988) 203 Cal.App.3d 1214 [ 250 Cal.Rptr. 468 ], are instructive in making some legal sense of this area, particularly showing the quantum of evidence concerning a parent’s personality, character and attitudes required to sustain a detriment finding.

We begin with Jasmon O., which, while it did not arise out of a section 366.22 detriment finding, discussed the question of psychological evaluations. There the court reviewed the evidence of whether a parent had failed to “maintain an adequate parental relationship” under former section 232, subdivision (a)(7). The majority and the dissent disagreed, among other things, on the use of certain psychological evidence to support a finding that the father had not done so. Writing for the majority, Justice Mosk noted that “[wjithout the testimony of psychologists, in many juvenile dependency and child custody cases superior courts and juvenile courts would have little or no evidence, and would be reduced to arbitrary decisions based upon the emotional response of the court.” (In re Jasmon O., supra, 8 Cal.4th at p. 430.)

And the psychological evidence in Jasmon O. was indeed formidable. The seven-and-one-half-year-old child had spent virtually her entire life in foster care, a point made forcefully in the opening paragraph of the opinion. (See 8 Cal.4th at p. 407.) The father suffered from drug dependency and mental illness, but eventually turned his life around. No less than four psychologists —the child’s treating psychologist, another child psychologist, a psychoanalyst (who was also an expert in child psychology) and even the father’s own psychologist—all agreed that the child would suffer severe psychological harm if placed (“returned” would not be the right word) in the father’s care. The psychological trauma of being so placed meant “serious lifelong emotional detriment” and “serious mental illness” for the child. (See id. at p. 416.) Moreover, the child’s “separation anxiety” was objectively demonstrable: there was disruption of toileting and sleeping patterns, violent fantasies involving kidnapping, and difficulty relating to other children. (Id. at p. 417.)

Brian R. dealt with the issue of detriment more directly. There, the father had “made substantial strides” in curing the problems that led to the dependency. He now tested “clean” for illegal drugs, had completed a parenting class, was receiving domestic abuse counseling, was consistently visiting his child, and had stable employment and housing. ( 2 Cal.App.4th at p. 913 .) He had, like the father in Jasmon O., turned his life around.

However, a psychologist concluded that “despite this reversal” the father “did not have the capacity to parent.” ( 2 Cal.App.4th at p. 913 .) Based on “testing and on clinical evaluations” the father had a “nonspecific, longstanding personality disorder with marked antisocial, dependent, and passive-aggressive features.” He “rationalized his abuse of his domestic partners; and he had a problem with taking responsibility.” Thus if the child were returned, there was still the risk of abandonment. Plus, the father was “likely” to change residences and caretakers and to develop “unstable, unhealthy relationships” which would entail “possible domestic abuse.” (Ibid.) This evidence, held the Court of Appeal, was not too speculative, because it was “expert opinion testimony, based on psychological testing and clinical evaluation.” (Id. at p. 915.)

In Heather P., on the other hand, a mother with a “history of mental problems,” a “very transient lifestyle” (she had been described as a “bag lady”), mental illness, and a suicide attempt, turned her life around relatively late in the dependency process. She met the elements of the service plan, obtained a stable residence, and began working in a church nursery. The social worker, however, still considered it detrimental to return the child to her mother because the mother’s therapist, although noting progress, did not give her a positive evaluation for neglect or endangerment. Further, another psychologist had written somewhat earlier in the case that the mother did “not exhibit adequate, nondetrimental or healthy parenting capabilities.” (See In re Heather P., supra, 203 Cal.App.3d at pp. 1219-1223, 1227, 1229.) The Court of Appeal declared that the psychologist’s opinion was insufficient to justify a detriment finding. The earlier psychologist had provided merely “general statements” about the parent’s psychological condition, but had not demonstrated “with specificity how the minor would be harmed.” The other psychologist’s statements were not only “outdated,” but were “not the type of evidence which could be deemed credible and of solid value and from which the juvenile court could conclude that [the] minor’s physical or emotional well-being would be threatened if she were returned” at the time of the review hearing. (Id. at pp. 1229-1230.)

If Jasmon O. (indirectly) and Brian R. (directly) instruct us that psychological evaluations can serve as credible evidence to sustain a detriment finding, Heather P. and, indeed, by implication Jasmon O., show they must at least be reasonably specific and objective. Significantly, the very justification for psychological evaluations afforded by the Jasmon O. court is the need for decisions based on evidence rather than an emotional response. To paraphrase Gertrude Stein’s famous dictum, there must be something there there. In Jasmon O. there were specific and objective factors which the court could point to within the psychological evaluations themselves as well as other evidence corroborating those evaluations. In Brian R. there was actual testing and a clinical evaluation. Heather P., on the other hand, involved mere subjective impressions, albeit professional subjective impressions.

The evidence in the present case is not only closer to Heather P. than either Brian R. or Jasmon O., it is weaker than that in Heather P. There is nothing there here. We have no clinical evaluation, no testing to indicate mental illness, just the opinion of the mother’s social worker and a therapist that she has not “internalized” what she has learned in parenting classes.

This failure to “internalize,” moreover, is seen largely by the social worker and therapist as the result of Blanca’s refusal to believe her husband is a child molester. When we look at just the subject of excessive corporal punishment, interestingly enough, the evidence is undisputed that Blanca has said she has learned that she should not use excessive force in child discipline. Indeed, her comments as related (but not believed) by the social worker are that she is willing to forswear corporal punishment altogether. Thus apart from the molestation, the case against Blanca boils down to the fuzzy notion that she does not truly believe what she has been taught about child discipline—a subject about which there is controversy even among experts and which is hard enough for any parent to master.

Let us be plain. The idea that, despite enduring countless hours of therapy and counseling (much of it predicated on the possibly erroneous assumption that her husband is a child molester), a parent who has faithfully attended required counseling and therapy sessions must still relinquish her child because she has not quite “internalized” what she has been exposed to has an offensive, Orwellian odor. The failure to “internalize” general parenting skills is simply too vague to constitute substantial, credible evidence of detriment. To hold otherwise would come perilously close to allowing legal decisions of monumental importance to the persons involved to be based on nebulous ideas more appropriate to an afternoon talk show than a court of law. (Cf. Simms v. State, Dept. of Health & Rehab. (Fla.Dist.Ct.App. 1994) 641 So.2d 957, 962-963 (dis. opn. of Jorgenson, J.) [describing as “psychobabble” vague testimony that mother lacked “sense of ‘self’ ” or ability or motivation to improve parenting skills]; see also Alexander, Big Mother: The State’s Use of Mental Health Experts in Dependency Cases (1993) 24 Pacific L.J. 1465, 1493-1494 [“When adult patients fail to agree with their therapists, psychological terms help the therapist explain that the patient is wrong. She may be seen as projecting her own problems or denying reality, to illustrate two professional buzzwords.”] (hereinafter Alexander).)

The Problem of the “Confession Dilemma” Demonstrates the Need for Clear, Reliable Findings Where Sexual Abuse Is Alleged

The lack of evidence of detriment apart from the molestation allegation underscores the critical importance of the jurisdictional hearing in this case. In the hindsight afforded appellate courts, the outcome of the April 12 hearing—a sustained subsequent petition based on allegations of child molestation—now looms as the event which would otherwise sever Rogelio and Blanca from their children permanently. Yet the very fact that Blanca and Rogelio have continually denied that Rogelio is a child molester is now asserted by the social services agency as evidence supporting the detriment finding.

The agency’s argument forces us to examine one of the most troublesome problems in juvenile dependency jurisprudence—the dilemma faced by a parent who is falsely accused of sexually molesting his or her child. If the parent denies what any decent person must regard as a horrible act, that denial itself—as the agency’s argument here illustrates—may end up preventing reunification.

In considering this problem of the “confession dilemma,” a few basic (and for the most part commonplace) truths must be kept in mind. Few crimes carry as much (or as much deserved) social opprobrium as child molestation. Most people would rather be accused of bank robbery. The crime is usually done in secret. Proof is often difficult. Perpetrators are not likely to admit their guilt. The victims of molestation may be too young, too frightened, too embarrassed or too dependent to provide credible evidence against the molester. And innocent children need protection.

But by the same token, it cannot be denied that it is an outrageous injustice to use the fact parents deny they have committed a horrible act as proof that they did it. That really is Kafkaesque. And by the same token it is also unjust to use the fact that a parent denies molesting his or her child as the reason to terminate reunification services—at least when (assuming we can be certain of such matters) the parent has been falsely accused. Further, it is undeniable that false accusations of child molestation do happen. In such a case, “denial”—in both its legal and psychological senses—should not become, perversely, the very fact which demonstrates the futility of reunification services.

In the context of California’s statutory scheme, the confession dilemma places an extraordinary premium on the correct adjudication of a petition alleging sexual abuse. If an injustice occurs there, the hard fact of life is that the very innocence of the parent will in all likelihood render the family asunder. And it is also irrefutable that no honorable person will want—or should have—to admit to a despicable sexual act of which he or she is innocent. The hearing on a contested petition alleging child sexual abuse is thus, to repeat, extraordinarily important. It is not the sort of thing to be rushed, or taken routinely. Allegations of child molestation are serious; they merit more than a rubber stamp. With the exception of death penalty cases, it is hard to imagine an area of the law where there is a greater need for reliable findings by the trier of fact. The consequences of being wrong—on either side—are too great.

In the present case we are confronted with a case where there is not only substantial new evidence that casts doubt on an initial jurisdictional finding of molestation, but also the very circumstances of that initial finding cast doubt on it. We must now consider whether there is any safety mechanism to remedy what could be a terrible and gross injustice.

Because of the Juvenile Judge’s Mistaken Impression About the Nature of the Hearing on the Molestation Allegation and the Subsequent Exoneration of Rogelio by a Psychologist, a New 18-month Review Hearing Must Be Held, in Which the Molestation Issue Can Be Confronted Directly

We now come to the procedural problem which lies at the very heart of this case—the collateral estoppel effect, if any, of the order sustaining the subsequent petition. Or, to put the question in more concrete and human terms: does the juvenile court’s order April 12 conclusively establish that Rogelio is a child molester?

A jurisdictional order sustaining a petition such as the one the juvenile court made April 12 is, while not appealable by itself, appealable by way of a challenge to a dispositional order made subsequent to it. (In re Megan B. (1991) 235 Cal.App.3d 942, 950 [ 1 Cal.Rptr.2d 177 ] [“A jurisdictional finding, while not appealable, may be reviewed in an appeal from the dispositional order.”].) The April 12 order sustaining the subsequent petition in this case is thus long-since final, was not timely appealed after it was made, and has not been the subject of a set-aside petition under section 388. At the same time, we must repeat, that order is the only basis on which the detriment finding could be upheld, even as to Blanca. Her refusal to leave her husband combined with her refusal to believe that her husband committed child molestation means, as a practical matter, that if the children are returned to her, they will also be returned to him and be at risk from any danger of molestation that he poses.

If the April 12 order is taken, for purposes of the 18-month review hearing, as having etched in concrete the idea that Rogelio molested Daisy, then of course there was substantial evidence at the latter hearing for a detriment finding, regardless of any misimpression of the juvenile court judge at the April 12 hearing and regardless of anything that Dr. LaCalle might have found which exonerated Rogelio. But the April 12 order should not be so taken. The reason is found in two Supreme Court cases, In re Carmaleta B. (1978) 21 Cal.3d 482 [ 146 Cal.Rptr. 623 , 579 P.2d 514 ], and Cynthia D. v. Superior Court (1993) 5 Cal.4th 242 [ 19 Cal.Rptr.2d 698 , 851 P.2d 1307 ], plus the opinion of the appellate court in In re Nathaniel P. (1989) 211 Cal.App.3d 660 [ 259 Cal.Rptr. 555 ].

In Carmaleta B., a mother appealed a judgment terminating her custody and control of her five children under former section 232. The children had been initially declared dependents based on allegations of cruelty, neglect and the father’s sexual abuse, and the mother challenged the section 232 findings on the ground that it was “predicated solely on the acts which initially caused the wardship proceedings.” (In re Carmaleta B., supra, 21 Cal.3d at p. 493 .) Confronting this challenge, the court first noted that orders under section 232 necessarily rested on both “present circumstances" as well as “past acts,” and then opined, that “[u]nder the principles of collateral estoppel” the mother might have been “precluded from denying the truth of the findings of the dependency proceeding that she failed to protect the children,” but the issues of cruelty and neglect could still be explored in the former section 232 termination proceeding. The court then quoted In re Zimmerman (1962) 206 Cal.App.2d 835, 843 [ 24 Cal.Rptr. 329 ], to the effect that section 232 proceedings were not so “rigid and inflexible” in character that the court in such a proceeding was merely “finalizing an antecedent interlocutory judgment.” (In re Carmaleta B., supra, 21 Cal.3d at pp. 493-494, quoting In re Zimmerman, supra, 206 Cal.App.2d at p. 843 .) Accordingly, the mother was “ ‘entitled to have the circumstances leading to’ ” the initial dependency orders “ ‘reviewed in the light of subsequent events.’ ” (In re Carmaleta B., supra, 21 Cal.3d at p. 494 , quoting In re Morrow (1970) 9 Cal.App.3d 39, 55 [ 88 Cal.Rptr. 142 ].)

Nathaniel P., like Carmaleta B., arose out of a judgment terminating parental rights under former section 232. The two children, ages five and three, were initially made dependents because they were found alone on a traffic island in San Francisco, but an amended petition alleged physical and sexual abuse. Five of the six counts alleged in that petition were found to be true. At the section 232 trial the father sought to testify that he had not physically or sexually abused his children. Counsel for the department of social services, however, objected, arguing that the father “could not go behind the findings already made after hearing on the section 300, subdivision (d) petition.” The court sustained the objections. The appellate court held that was error and reversed, relying on Carmaleta B.

Specifically, the Nathaniel P. court noted that different standards and burdens of proof meant that the doctrine of collateral estoppel did not apply. The earlier (i.e., § 300, subd. (d)) findings were based on a preponderance of the evidence standard. The former section 232 judgment required the higher clear and convincing evidence standard. The father was thus entitled to have the section 232 proceeding “redetermined” under the appropriate standard. (See In re Nathaniel P., supra, 211 Cal.App.3d at pp. 670-672.)

This brings us to Cynthia D., which grew out of the controversy over the datum that the standard of proof for detriment at 12- and 18-month review hearings is the lower preponderance of the evidence rather than, as with the old section 232 proceedings, clear and convincing. (See generally Cynthia D. v. Superior Court, supra, 5 Cal.4th at p. 250 .) The argument being made by parents’ counsel was that changes in the procedure for permanently terminating parental rights (outlined at id. at pp. 246-250) violated federal due process rights as articulated by Santosky v. Kramer (1982) 455 U.S. 745 [ 71 L.Ed.2d 599 , 102 S.Ct. 1388 ] because now a critical link in the chain—a finding it would be detrimental to return the child at a 12- or 18-month review—could be based on preponderance of the evidence rather than what Santosky seemed to require, clear and convincing evidence.

While, in rejecting the argument, the court did not cite Carmaleta B. (see Cynthia D. v. Superior Court, supra, 5 Cal.4th at pp. 253-256), the same thought animating Carmaleta B. also animated Cynthia D.\ the proceeding which actually terminates the parent’s relationship to the child is more than just the finalization of an “antecedent interlocutory order.” The key passage in Cynthia D., supra, 5 Cal.4th 242 , is found on page 253 of the opinion, in which the court pointed out that California’s termination statute, section 366.26, “cannot properly be understood except in the context of the entire dependency process of which it is a part.” The court then went on to point out that by the time dependency proceedings reach the section 366.26 termination stage, “there have been multiple specific findings of parental unfitness.” Moreover, initial removal required a clear and convincing standard, and, for the case to have gotten to the section 366.26 stage in the first place, the state had to have “continually” established over a period of time that return of the children to the parents would be detrimental.

Significantly, in discussing one of the factors that motivated the United States Supreme Court in Santosky to require an “elevated standard of proof’—the “risk of erroneous fact-finding”—Cynthia D. noted the requirement for repeated findings of detriment under the preponderance of the evidence standard were protection against an erroneous result. (See Cynthia D. v. Superior Court, supra, 5 Cal.4th at pp. 254-255.)

Considered together, Carmaleta B., Nathaniel P. and Cynthia D. establish that collateral estoppel effect should not be given, at a 12- or 18-month review, to a prior finding of child molestation made at a jurisdictional hearing when the accused parents continue to deny that any molestation ever occurred and there is new evidence supporting their denial. Applying collateral estoppel would be inconsistent with the rationale in Cynthia D., which emphasized that California’s dependency scheme considered as a whole affords parents “repeated” opportunities to challenge detriment findings to, among other things, diminish the risk of erroneous fact-finding. (See Cynthia D. v. Superior Court, supra, 5 Cal.4th at p. 254 .) In cases where child molestation is alleged and denied, and there is new evidence supporting the denial, to say that a parent is collaterally estopped from contesting the molestation itself at a 12- or 18-month review hearing is to make the “antecedent” jurisdictional finding virtually dispositive in terminating parental rights—and dispositive based on a prior finding made under a preponderance standard.

To limit the evidence at the hearing to just the issue of a parent’s propensity to commit molestation in the future is, under circumstances when there is new evidence that no molestation ever happened in the first place, not only grossly unfair, but also contrary to the logic of Cynthia D. The remedy for a factually erroneous molestation finding made at a jurisdictional hearing is, in the words of Carmaleta B., to allow it to be reviewed in light of subsequent events.

Our conclusion is also buttressed by the realities of our overcrowded juvenile dependency courts. We have emphasized the need for accurate and reliable findings of fact where child molestation is at issue. But we also know that jurisdictional findings are made under extreme time pressures, and with a certain degree of urgency necessary to protect children. The hard truth is that all too often (the facts in the case before us demonstrate the point as much as any other) juvenile courts and counsel do not have enough time to fully explore molestation issues in jurisdictional hearings, and psychological evidence about a parent’s propensity to commit molestation is likely to be unavailable, inadmissible or nonexistent.

In the case before us, however, the 18-month review was conducted on the assumption that it had already been conclusively established that Rogelio had molested Daisy. Dr. LaCalle did not testify, and counsel for the minors, for example, successfully objected to a question put to the social worker that if Rogelio had not molested Daisy, “there wouldn’t be anything for the mother to protect her from,” as irrelevant. But it was irrelevant only if the idea that Rogelio had actually molested his daughter had been already established and was immune, at the 18-month hearing, from any attack.

It was not immune, though the juvenile court mistakenly thought so. We will therefore grant the petition to vacate the orders made at the 18-month hearing, but not order the immediate return of the children. Rather, a new 18-month hearing which will focus on the molestation allegations against Rogelio should be held as soon as possible. In light of Dr. LaCalle’s report and this opinion, we trust that the issue will be fully explored and resolved.

Disposition

Let a peremptory writ of mandate issue commanding the juvenile court to vacate its January 10, 1996, order and to hold another 18-month review hearing on the molestation allegations set forth in the subsequent petition filed October 17, 1994. The hearing should be held as soon as possible, consistent with the right of all sides to prepare their case.

For the guidance of the juvenile court, we elaborate on the steps that should be taken in the wake of the hearing. If, at the new 18-month hearing, the juvenile court finds that Rogelio did not commit child molestation, then (assuming there are no new developments that would warrant otherwise) the children should be returned to their parents; that is the logic of our decision that there is no support for a detriment finding apart from the molestation allegations. If the court does find that molestation occurred, the court can make whatever orders are appropriate in light of that finding and then deal with any other issues raised by the parties.

Crosby, J., and Wallin, J., concurred.

A fifth child, Fabiola, nine, had been declared a dependent in 1992 based on excessive corporal punishment and was already living in a group home. The statute governing the 18-month review hearing is section 366.22 of the Welfare and Institutions Code. All statutory references are to the Welfare and Institutions Code unless otherwise designated, except for section 232, which is to the Civil Code. See section 361.5, subdivision (b) for a list of the relatively less common situations where reunification plans need not be provided. See also In re Heather B. (1992) 9 Cal.App.4th 535, 563 [ 11 Cal.Rptr.2d 891 ], which involved a challenge to an order denying visitation pending a permanency planning hearing. There, the father failed to maintain contact with his daughter, did not attend any parenting classes, did not attend any substance abuse or substance testing programs, and did not undergo a required psychological examination. Heather P. was overruled on an unrelated point in In re Richard S. (1991) 54 Cal.3d 857, 866 , footnote 5 [ 2 Cal.Rptr.2d 2 , 819 P.2d 843 ]. Justice Baxter’s comments in his dissent in Jasmon O. on the use of the particular psychological evidence in that case may not have carried the day, but applied to the case before us, where the evidence is far flimsier than in Jasmon O., they are particularly trenchant. (See In re Jasmon O., supra, 8 Cal.4th at p. 438 (dis. opn. of Baxter, J.) [characterizing as “Orwellian” the rule that a parent should lose a child “period and forever” because a “therapist thinks” the parent “should be more loving and demonstratively affectionate”].) Our comments should not be taken beyond their context, which here is that of a non-mentally ill parent whose child was removed because of excessive corporal punishment. We note there is an obvious difference between excessive corporal punishment and child sexual molestation. We are spared the much more difficult problem of the parent who unquestionably has molested a child, who admits the molestation, faithfully complies with the reunification plan, says all the right things, and yet there is still doubt as to whether he or she really has reformed. That situation is not before us. We also recognize that psychological evaluations can be extremely crucial in cases where, as in In re Andrea G. (1990) 221 Cal.App.3d 547 [ 270 Cal.Rptr. 534 ] (parent put frozen vomit in freezer and tried to withdraw “good blood” from 12-year-old child), a parent has a demonstrable mental illness. Judging from Dr. LaCalle’s report, Blanca is not likely to be acquainted with the tradition of dystopian literature exemplified by authors such as Orwell, Huxley and LeGuin. But consider the eloquence of this testimony, ironically elicited by the deputy county counsel at the 18-month review hearing in response to a question as to whether Blanca would believe her daughter “if she said she was molested”: “They keep telling me, just say—admit, say, say say it’s true that. [<]Q And to me, they—they wash my head here, say, say, say. But I don’t believe it. The truth is, I don’t believe it, because I know my husband. I have been with him for 12 years.” One of the most egregious cases of a false accusation of molestation is that of Alicia W., who was the subject of a 1992 report by the San Diego Grand Jury. The ordeal of the father in that case has been recounted in Professor Alexander’s article, 24 Pac. LJ. at page 1493, footnote 123: “A daughter of a military father was raped, and he was accused of having been the rapist. He denied the charge and was joined in the denial by the steadfast assertion of his eight-year-old daughter, Alicia, that a stranger raped her. Alicia was removed from the home while the case progressed. After what the Grand Jury described as 13 months of prodding by a psychologist, Alicia was willing to admit that her father was the rapist. After two and one half years, the Department finally tested the semen sample they had kept. It was incompatible with the father’s semen.” (Citing A Report of the 1991-92 San Diego County Grand Jury, The Case of Alicia W., Report No. 6 (June 23, 1992); see also Patton, Child Abuse: The Irreconcilable Differences Between Criminal Prosecution and Informal Dependency Court Mediation (1992) 31 U. Louisville J. Fam. L. 37, 47, fn. 37 [“For the most unconscionable case of an innocent father pleading to criminal charges and the parade of horribles that took place, see The Case of Alicia W.: Report No. 6, a Report by the 1991-92 San Diego County Grand Jury, June 23, 1992, at 1-49.”].) This point was also made by the San Diego Grand Jury and recounted by Professor Alexander: “If the [dependency] court believes a molest occurred and the family member could have been responsible a ‘true finding’ is made and wardship declared. If a father denies molest and a true finding is made, he suffers the ultimate Catch 22—he can either admit and take a chance that the department will allow him to begin reunification with his family or he can deny and no reunification will occur. [<¡0 But the irony does not end there. If the spouse supports her husband’s denial, she cannot be trusted to protect the child and she too will not be allowed to reunify with the child., a [sz'c] current assertion is that the mother must have known all along and failed to protect. That then becomes a protective issue and reason to remove the child from the mother. [^Q Still worse, if the child denies the molest, this can be seen as part of a ‘child abuse accommodation syndrome’ and an additional reasons [sic] why the child should have no contact with the parents. . . . Thus, all members of the family can deny a false molest allegation and, in each instance, the system uses the denial as evidence of guilt.” (Alexander, supra, 24 Pac. L.J. at p. 1482, fn. 81, quoting San Diego County Grand Jury, Child Sexual Abuse, Assault, and Molest Issues, Rep. Nos. 8, 2, 3 (June 29, 1992).) Compare, e.g., In re Jessica B. (1989) 207 Cal.App.3d 504, 516 [ 254 Cal.Rptr. 883 ], in which the court indicated that as long as a father who had been accused of physically abusing his infant daughter did not admit the abuse and display appropriate remorse, traditional treatment would be futile. A redundant dispositional order (the children were already in foster care) from which an appeal could have been taken was made in April 1995 at the time the subsequent petition was sustained. We note in this regard that because the children had already been removed from their parents’ home, we cannot infer that the evidence of child molestation presented at the April 12 hearing reached the level of clear and convincing as is required by section 361. Either that, or, if Blanca does leave Rogelio, the social services agency will have succeeded in breaking up an intact family at some cost to a financially strapped county government and immense human cost to the individuals involved. One of the most disturbing, although collateral aspects of this case is, as mentioned in Dr. LaCaile’s report, the attempt of social workers to get Blanca to separate from her husband Rogelio. “Mr. P[.] resented the attempts of the Social Services to break up his marriage. ‘They’ have told his wife that if she leaves him, she might have a chance to get her children back.” We are mindful that the Legislature has declared a prime purpose of California’s juvenile dependency statutes is the reunification of families, not their breakup. (See § 300 [protection for children “shall focus on the preservation of the family whenever possible”].) The exact quote was: “Under the principles of collateral estoppel, Mrs. B. may be precluded from denying the truth of the findings of the dependency proceeding that she failed to protect the children [citation]; however, ‘inquiry into the issue of cruelty or neglect [in a section 232 action] is not foreclosed by the order making the minor a dependent child of the juvenile court. [Citations].’ ” (In re Carmaleta B., supra, 21 Cal.3d at p. 493 .) Thus if a parent were falsely accused of molestation which was found true under a preponderance standard at a jurisdictional hearing, and the parent continued to deny it throughout the reunification period, the fact that the parent denied having committed child molestation would demonstrate that the reunification services had not worked, and the fact that the jurisdictional finding had been made would show that the parent had committed child molestation. As a member of this panel has written in another context, “Perhaps the strongest objections to the use of collateral estoppel in any setting are injustice, surprise, potential perpetuation of an earlier error, and untoward effects on the original tribunal. ...['][]... Collateral estoppel is strong legal medicine because, like full faith and credit, res judicata, and even law of the case, it can preclude relitigation of an issue even when the earlier decision was clearly erroneous.” (Crosby, Administrative Collateral Estoppel in California: A Critical Evaluation of People v. Sims (1989) 40 Hastings L.J. 907 , 940-941, fns. omitted.) It is precisely the insight of the Cynthia D. court that California’s juvenile dependency system affords parents opportunities to correct erroneous jurisdictional findings. In In re Brandon C. (1993) 19 Cal.App.4th 1168 [ 23 Cal.Rptr.2d 571 ] the court suggested that an erroneous finding of molestation made at a jurisdictional hearing could be challenged in a section 388 hearing. There, a petition alleging sexual abuse was sustained at a jurisdictional hearing which included an eyewitness account of the molestation by the father’s sister. The sister, however, recanted her testimony four months after the dispositional hearing. Seven months after the sister’s recantation the father brought a petition for writ of habeas corpus. The court noted that if the recantation was believed, a “crucial piece of evidence would drop out of the department’s case in the jurisdictional hearing,” and therefore one could not say that the jurisdictional petition would have been sustained without it. The appellate court opined that a section 388 petition would have been “a superior procedural vehicle for resolving the significant factual dispute.” ( 19 Cal.App.4th at p. 1174 .) Accordingly, in denying the petition for writ of habeas corpus, the appellate court specifically made its denial “without prejudice” to the father’s right to present new evidence in a section 388 motion. ( 19 Cal.App.4th at p. 1174 .) We express no opinion on the use of section 388 as a possible alternative remedy. No order denying a section 388 motion is before us. See Alexander, supra, 24 Pac. L.J. at page 1484 (“Currently, the attorneys and social workers involved in the dependency process have a staggering number of simultaneous cases”); Patton, Forever Torn Asunder: Charting Evidentiary Parameters, the Right to Competent Counsel and the Privilege Against Self-Incrimination in California Child Dependency and Parental Severance Cases (1987) 27 Santa Clara L.Rev. 299, 299 fn. 1 (noting that dependency petitions in Los Angeles County alone skyrocketed from 8,192 in 1980 to 17,913 in 1985). For example, under rule 1447(b) of the California Rules of Court, if a child is detained at the time the petition is filed, the jurisdictional hearing must be begun within 15 court days from the detention order. See Alexander, supra, 24 Pac. L.J. at page 1482, footnote 81, quoted above at footnote 10: “If the [dependency] court believes a molest occurred and the family member could have been responsible a ‘true finding’ is made and wardship declared.” The tendency on the part of courts to err on the side of the protection of children is both commendable and understandable. (Cf. In re Paul E. (1995) 39 Cal.App.4th 996 , 1003 [ 46 Cal.Rptr.2d 289 ] [noting need to err on side of child’s safety where there is doubt].) In re Mark C. (1992) 7 Cal.App.4th 433 [ 8 Cal.Rptr.2d 856 ] is illustrative. There, in passing, the appellate court noted that juvenile courts do not “routinely” consider psychological evaluations regarding risk to children at the jurisdictional stage. (See id. at p. 445.) There are two other issues besides detriment raised in Blanca’s writ petition, namely whether reasonable reunification services were in fact provided and whether reunification services should be extended. Those issues can be dealt with by the juvenile court in light of any new finding of molestation. Of course, if the court does not find molestation, those issues will be moot. We also note that Rogelio and Blanca did not have a full 12 months of reunification services in the wake of the sustained subsequent petition. (The subsequent petition was sustained in April 1995, the 18-month review terminating services was in January 1996.) That will be a problem only in the event the juvenile court finds that the molestation actually happened.