Oversimplified beliefs about child abuse allegations in custody cases with alienation rebuttals – review of empirical data

Abstract

This article is a companion piece to a prior article that analyzed the implications of oversimplified beliefs about alienation. Those beliefs support excessive skepticism about child abuse (CA) allegations in custody cases (Milchman, 2022a). Alienation refers to a child’s unjustified rejection of one parent and alliance with the other (Bernet, 2010). Responsible theoreticians recognize that there are many legitimate reasons a child might reject a parent (Drozd et al., 2013). However, alienation rebuttals of children’s abuse allegations in custody cases claim that the preferred parent has manipulated the child to make a false allegation. The plausibility of this argument rests on accepting untested assumptions about CA allegations in custody cases. This article empirically tests three assumptions: (1) CA and child sexual abuse (CSA) allegations are common in custody cases, (2) typically false, and (3) malicious. The empirical findings reviewed here address the frequency of these allegations in custody cases; child protective services (CPS) substantiation rates; judicial decisions; children’s disclosure patterns (prior disclosures, non-disclosures, delayed disclosures, and recanted disclosures); residual suspicions of unsubstantiated abuse; and deliberately false allegations. Results show that support for the assumptions about CA that underlie alienation rebuttals consists of small-to-moderate differences between custody and non-custody cases. On their own, these findings could indicate that custody cases are somewhat more likely to have false CA allegations. However, other empirical findings undermine support for that interpretation.

Keywords:

• Alienation
• child abuse
• custody
• empirical data

Introduction

When child abuse (CA) and child sexual abuse (CSA) allegations are raised in custody cases, the accused often rebuts them with alienation allegations. Alienation refers to a child’s unjustified rejection of one parent and alliance with the other (Bernet, 2010). Alienation claims rebut CA and CSA allegations by arguing that the preferred parent has led the child to make a false allegation. A prior article by the author, which is a companion piece to this article, analyzes oversimplified beliefs about alienation that are implicitly relied upon to support the argument that an allegedly abused child is really an alienated child (Milchman, 2022a). This article reviews empirical data related to assumptions about CA allegations in custody cases that buttress the credibility of the alienation argument by creating skepticism about the CA allegations. Focusing on assumptions about alienation that could lead to excessive or indiscriminate support for CA allegations are beyond the scope of this article but merit attention as well.

The first assumption is that CA and CSA allegations are common in custody cases. Clawar and Rivlin (2013) declare, “Given the epidemic proportion of incest allegations in contested custody or access disputes ….” The second assumption is that they are likely to be false. Clawar and Rivlin (2013) continue, “there is a heavy burden on experts to cautiously explore the veracity of the alleged victim” (p. 88) and “take great care not to brand those who are innocent” (p. 89). The third is that allegations are likely to be malicious (Gardner, 1986). Even though today sources of mistaken allegations other than intentional fabrications are well recognized (Kuehnle et al., 2016), concern about high rates of deliberately false allegations is ongoing (Campbell, 2013).

While these assumptions have gained credence in forensic arenas, the Diagnostic and Statistical Manual (DSM-5) refused to accept alienation as an identifiable mental health disorder or condition (Bernet, 2010). It only included some descriptions of alienating behaviors as examples of problematic parent-child relationships, comparable to other problems that are not mental disorders or conditions such as parental overprotection and parental distress (American Psychiatric Association, 2013). The need to bring custody and parenting time decisions more in line with the difficulty of identifying alienation cases has led to calls for more rigorous cross-examination (Myers & Mercer, 2022).

These assumptions about CA are not limited to cases with alienation allegations. They affect contested custody cases in general. After several children who made abuse allegations were murdered by parents following custody evaluations, the New York State Blue-Ribbon Commission on Forensic Custody Evaluations was convened. It was tasked in part with making recommendations as to whether forensic custody evaluations should be used in New York courts (Blue Ribbon Commission, 2021). While it concluded that custody evaluations were valuable, it was a close call. Accepting unproven assumptions risks the credibility of the entire profession, especially when they prove wrong. This article reviews empirical data related to the three assumptions that undermine the credibility of CA claims in custody cases. It begins by discussing reasons why judges might decide to accept alienation allegations and reject CA allegations without an evidentiary basis. It then conceptualizes the three assumptions as hypotheses and tests them with empirical data.

The data is drawn from scientific research studies in multiple countries and from the Canadian Incidence Studies (CIS), which are the official government CA reports. There are no comparable statistics from the U.S. because the official U.S. annual Child Maltreatment reports do not identify custody cases and differentiate them from non-custody cases. The article concludes that some empirical data could be interpreted to provide some support for alienation arguments about CA. However, that interpretation requires isolating that data from other empirical data that undermines the alienation-supportive interpretation. While the empirical review indicates that CA allegations in custody cases should not be met with skepticism, the differences between custody and non-custody cases, while not large, indicate that careful judicial scrutiny of case-specific evidence for both the alienation and the abuse allegations is needed (Milchman, 2022b).

In many studies, the term “substantiated” refers to both judicial and child protective services (CPS) decisions. Throughout this article, the terms “credited/uncredited” refer to judicial decisions, and the terms “substantiated/unsubstantiated” refer to CPS decisions to distinguish differences in decision-making procedures, criteria, and possible levels of expertise. The term “reports” refers to parents’ communications of their children’s abuse allegations to family courts and/or to CPS in addition to official agency reports. CA refers to multiple types of child abuse combined. CSA is specific to child sexual abuse. CPA is specific to child physical abuse. The words “child abuse” in the text are used when the meaning is generic rather than a reference to specific types of CA. The heading “custody disputes” is used when the section addresses either generic custody issues or custody cases and custodial parents. “T” refers to tables in cited sources. “Table” refers to tables in this article.

Judicial decisions crediting alienation over child abuse claims in custody disputes

Gender bias could affect judicial decisions to credit alienation over child abuse

When no abuse is alleged, judges do not appear to indiscriminately accept alienation allegations. In “no abuse” cases, Bala and Hunter (2015) found that Ontario judges credited and discredited alienation allegations roughly equally. Consistently, Meier (2019) found that judges who believed that parents had engaged in alienation, punished them in a gender-neutral manner with loss of custody and unfavorable parenting time. However, in cases where alienation claims rebutted CA claims, many studies have found that judges accept the alienation claims more readily than they accept the CA claims (Death et al., 2019; Ferguson et al., 2018; Meier, 2019, 2020; Milchman, 2017). Meier (2019) concludes that findings in “no abuse” alienation cases support assertions that alienation claims are not inherently “pro-father,” while findings in “abuse vs. alienation cases” support mothers’ assertions that countering a fathers’ alienation rebuttal is ineffective despite evidence to the contrary.

Inter-agency relationships could affect judicial decisions to credit alienation over child abuse

Relationships between family courts and CPS agencies in specific jurisdictions may affect judges’ decisions to give less credibility to CA claims. In some jurisdictions in the U.S. and Canada, custody investigations might not proceed in family courts until CPS investigations are complete (Brown et al., 2000; Ferguson et al., 2018). However, even if that is true, in the U.S. and Canada judges would not be legally mandated to defer to CPS. In Australia, the family court system has recognized that CPS’s failure to substantiate abuse is not equivalent to “no abuse” and that family court judges may need more information from other sources to decide on safe living arrangements for the child (Higgins, 2007). They deemed the problem of insufficient information from CPS sufficiently serious to warrant the development of a special case management system to obtain and coordinate a greater body of evidence. In these countries, at the least, judicial decisions to defer to CPS are not likely to be determined solely by inter-agency relationships because, absence a legal mandate, judges would or should rely on the totality of evidence presented to them.

Unreliable expert opinions could affect judicial decisions to credit alienation over child abuse

Several researchers have discussed judicial over-reliance on expert opinions (Bala & Hunter, 2015; Bala et al., 2010; Death et al., 2019; Saunders et al., 2011). Many experts do not assess the causal complexity of children’s rejection of their parents, even though there is a theoretical consensus in the field that this should be done (Drozd et al., 2013; Friedlander & Walters, 2010; Johnston & Sullivan, 2020; Milchman, 2022b, 2022c; Rudd & Beidas, 2021). Experts may conduct superficial “checklist” assessments of alienation (Death et al., 2019; Neilson, 2018; Warshak, 2021). They may misinterpret the failure of CPS to substantiate abuse as proof that abuse did not occur, erroneously equating the absence of evidence with evidence of absence and ignoring the many institutional obstacles to conducting thorough CPS investigations in custody cases (Houston et al., 2017; Milchman, 2022a; Saini et al., 2019; Sudland, 2021). Such experts may be concerned that giving weight to unsubstantiated abuse risks punishing innocent parents, as Clawar and Rivlin (2013) warn, it would be reckless to opine that abuse could have occurred despite a lack of substantiation if that opinion were only based on allegations. However, when that opinion is based on evidence that is present but insufficient to meet legal burdens, then it is reckless to ignore that evidence. Moreover, doing so contradicts CPS policies worldwide for protecting children.

In the U.S., Canada, and Australia, CPS agencies recognize the presence of abuse evidence in cases where there is insufficient evidence for legal substantiation (Brown, 2003; Ferguson et al., 2018). In Canada, CPS terms these partially supported cases “suspected,” defined as those with “insufficient evidence to substantiate abuse … but maltreatment cannot be ruled out” (CIS-98, Trocme et al., 2001; CIS-2003, Trocme et al., 2005; CIS-2008, Public Health Agency, 2010, p. 24). CIS-2008 reported that suspected cases made up 8% (17,918) of all investigations—not a trivial number (Public Health Agency, 2010, p. 24). In the U.S., though different states use different terms, the national Child Maltreatment 2021 report uses the term “indicated,” also defined as “maltreatment [that] could not be substantiated under state law or policy, but there is a reason to suspect that at least one child may have been maltreated or is at risk of maltreatment” (U.S. Department of Health & Human Services [USDHHS], 2021, p. 17). This article uses the term “suspected” for these kinds of cases.

Not all states in the U.S. differentiate between substantiated and suspected cases. In states that do not, abuse might be underestimated because suspected cases could be counted as unsubstantiated ones. However, CPS definitions of suspected cases make it clear that some affirmative evidence of abuse must exist. In both the U.S. and Canada, suspected cases contribute to official national counts of abuse victims. In the U.S., cases deemed to present low or moderate risk of abuse are not classified as suspected cases (USDHHS, 2021). Caseworkers are instructed to check on whether an alleged victim had previously been suspected of being abused (USDHHS, 2021). Suspected cases institutionalize “an important clinical distinction” that separates legal burdens of proof from the amount of evidence needed to protect a child (CIS-2008, Public Health Agency, 2010; USDHHS, 2021, p. 15), which is also the concern of family courts. Thus, whereas CPS agencies take pains to avoid the risk of undetected repeated victimization that is created by ignoring, dismissing, or minimizing evidentiary support for abuse allegations that could not be substantiated, experts in custody cases create this risk out of a misguided concern about supporting baseless allegations. Allegations that are partially supported by evidence are not baseless allegations.

Judges may be overly accepting of unreliable expert opinions that alienation and not abuse is the cause of a CA allegation because they mistakenly believe that experts know alienation when they see it. The wealth of widely disseminated anecdotes purported to illustrate alienation creates the illusion that alienation is obvious, easy to detect, and “it looks like that” (e.g., Bernet, 2010; Clawar & Rivlin, 2013; Darnall, 2013; Warshak, 2013; Worenklein, 2013). Since alienation proponents have not addressed the problem that the anecdotes they provide could just as logically be interpreted as indicating children’s defensive avoidance of abuse or of its disclosure (Milchman, 2022a, 2022b), judges may not be aware of the ambiguity in alienation “evidence” and may not see the need for in-depth analysis of the foundation for an expert’s opinion.

Frequency of judicial decisions crediting alienation vs. child abuse claims

While there are no official government statistics in the U.S. on the frequency with which judges favor alienation over CA claims in custody cases, Meier’s (2020) study is a recent large-scale quantified study of CA and alienation claims in U.S. family courts. She analyzed published judicial opinions in appellate custody cases as well as published trial opinions from 2005–2014, yielding a multi-jurisdictional sample of 4,338 cases. There were 1,946 cases in which mothers accused fathers of CA or domestic violence and there were no cross-claims of alienation (p. 95). There were 669 cases with alienation claims, which included 312 cases in which the alienation claims rebutted abuse claims. Of these 312 cases, there were 222 in which mothers reported abuse and fathers cross-claimed alienation and 90 cases in which fathers reported abuse and mothers cross-claimed alienation. Whereas judges credited 41% of mothers’ abuse reports when fathers did not cross-claim alienation (p. 96) - consistent with the substantiation rate of 45% in CIS-2003 (Trocme et al., 2005) and 36% in CIS-2008 (Public Health Agency, 2010) - they credited only 23% (Meier, 2020, p. 97) when fathers did cross-claim alienation (p. 98). Meier (2020) does not report the rate of judicial crediting of mothers’ alienation rebuttals when fathers accuse them of abuse in the 90 cases in which this occurred.

The low rate of alienation claims in general and fathers’ alienation rebuttals of abuse allegations specifically in Meier’s (2020) study could be interpreted to support the idea that alienation rebuttals are offered cautiously and are therefore likely to be well-founded. Alternatively, it could be interpreted to mean that alienation claims that were successful in trial courts are not likely to be appealed, especially by mothers. Parents, especially mothers, might be reluctant to appeal because their resources have been depleted, and/or they fear losing custody if they insist on the merit of CA claims that a judge has already rejected. Meier’s (2020) analysis of custody losses indicates that this fear is realistic.

Other studies have also found judicial rejection of CA claims that are rebutted by alienation claims (Table 1; Ferguson et al., 2018; Milchman, 2017; Silberg & Dallam, 2019). Neilson (2018) found that some Canadian judges ordered expert evaluations of alienation claims but not of abuse claims. These claims might have been unsubstantiated and so mistakenly considered false. Alternatively, the judges might have mistakenly believed that finding alienation disproves abuse (Milchman, 2022b, 2022c).

Table 1. Frequency of judicial decisions crediting alienation vs. CA and CSA claims.

Study, Characteristics	Findings	Evidence cited by Judges
Bala et al. (2010), Canada, 1989–2008, N = 175 (p. 166)	PA − 61% credited (p. 166) PA − 39% rejected (p. 167) CA − 7% credited of 39% PA rejections	Expert opinions Behavioral details
Bala and Hunter (2015), Ontario, 2013–2014, N = 33 (p. 12)	PA– 48 % (16) credited (p. 12) PA − 52 % (17) not credited	Not reported
Bala and Schuman (1999), Canada, 1990–1998, N = 196	CPA + CSA − 23% credited	Not reported
Ferguson et al. (2018), Australia, 2013–2015, N = 154	CSA credited − 10.3% (T. 2, p. 104) CSA disbelieved − 61.7% CSA disbelieved/perceived lack of evidence − 32.6% CSA disbelieved/perceived PA − 27.4% (T.3, p. 106)	(Disbelieved vs. credited) (T. 3, p. 105) Reports of accuser and child (55 vs. 11) Child’s behavior change (6 vs. 1) Witnesses (7 vs. 1 Injury (5 vs. 0) Sex knowledge (6 vs. 0) Multiple sources of evidence (14 vs. 3) PA (26 vs. 0)
Johnston et al. (2005), California, 1989–2002, N = 120	CA − 34%	Not reported
Meier (2019), U.S., N = 222 (p. 12–p. 13)	CPA − 27% credited PA/CPA − 18% credited CSA − 15% credited PA/CSA − 2% (1) credited	Not reported
Milchman (2017), U.S., N = 20	CSA rejected − 15/20 (75%) cases, no affirmative evidence against CSA PA credited/CSA discredited − 12/20 (60%) cases, no affirmative evidence of PA	No evidence cited in judicial record
Neilson (2018), Canada PA/DV/CA cases, N = 142 (p. 33)	PA evaluations − 43.7% DV/CA evaluations −2.8%	Expert evaluations Children’s views
Silberg and Dallam (2019), U.S., N = 27	27 cases with custody/visitation awarded to abuser; decisions overturned	Re-abuse
Thoennes and Tjaden (1990)	CSA − 50% credited (multiple data sources: judges, court administrators, custody evaluators, mediators)	Not reported

Note. PA = alienation, CA = child abuse (all types combined), CPA = child physical abuse, CSA = child sexual abuse. Percentages were computed from numbers provided by authors when authors did not provide percentages.

Some of these decisions might have been well-founded. Meier (2020) did not assess the evidence on which the judges based their decisions. A few other studies did and the results are mixed. Bala et al. (2010) found that Canadian judges discredited alienation claims when they had evidence of poor parenting (35%) or abuse and violence (7%). However, Death et al. (2019), Ferguson et al. (2018), and Milchman (2017) found that judges disregarded or minimized evidence of CSA when they credited alienation (Table 1).

The study by Ferguson et al. (2018, T. 2) is a small one (N = 103), but it is particularly important because they reported the evidence on which judges relied to believe or disbelieve¹ CSA allegations. Most often, they relied solely on the reports of accusers and children (T. 3). Sometimes inconsistency between these reports led them to disbelieve the allegations (4 cases; T. 4). This would be rational, though overvaluing consistency is an error because inconsistency could be related to numerous factors other than the truth of the allegations (Bala & Schuman, 1999; Faller, 2007; Fivush, 1993; Walker, 2013). Other decisions were not rational because they went against the weight of the evidence. While about a third of the judges said they disbelieved the CSA allegations because they lacked evidence (Ferguson et al., 2018, T. 4), disbelieved cases actually had more corroborating evidence of various kinds than believed ones (T. 3). Children’s behavior was reported to change more often in disbelieved than believed cases (6 vs. 1), though the specificity of these changes to CSA was not reported and they might have been caused by divorce stresses. Other corroborating evidence was less ambiguous. Physical witnesses were present in more disbelieved than believed cases (7 vs. 1). Injuries and sexual knowledge were only present in disbelieved cases. More disbelieved than believed cases had multiple kinds of corroborating evidence. Why didn’t this greater level of corroborating evidence influence these judges to believe the CSA allegations?

Belief that alienation was present appeared to drive some of their decisions to reject the CA allegations. In 26 of the disbelieved cases (27.4%), but no believed ones, the judges thought that alienation was present and reasoned accordingly (Ferguson et al., 2018, T. 4). In 16 of these cases (16.8%) they believed that the accuser was unreliable and in 10 cases (10.5%) they believed that the child was unreliable. Judges believed that the accuser suffered from mental health issues in 4 cases (4.2%) or was deliberately dishonest in 6 cases (6.3%). They believed that the accuser’s refusal to accept innocent explanations for the alleged sexual abuse was unwarranted in 16 cases (16.8%). They believed that the accuser had influenced or coached the child in 6 cases (6.3%). While they relied primarily on accusers’ reports and did so at the expense of other evidence, they nevertheless did not trust those reports as often as they trusted the reports in the cases they believed (57.0% vs. 68.8%).

Consistent with Ferguson et al. (2018), Death et al. (2019) found that the image of mothers as vengeful, manipulative, and mentally ill was present in alienation discourse and discredited mothers’ reports of CSA. Meier’s (2020, T. 7) results suggested that judges who found mothers to be alienating did not believe that fathers could have committed CPA or CSA, except when they had also committed domestic violence. Perhaps Ferguson et al.’s (2018) judges, believing that alienation was present, just could not believe that CSA was also present no matter what the evidence said. The abuse skepticism that is fomented by alienation thinking was tragically apparent in 27 cases where trial courts accepted alienation claims at the expense of abuse claims and the children were re-abused (Silberg & Dallam, 2019). Does the frequency of CA allegations in custody disputes justify skepticism?

Frequency of child abuse allegations in custody cases in family court

Studies that compare the percentage of family court custody cases with CA allegations to the percentage of family court custody cases overall find CA allegations to be rare. An older study by Thoennes and Tjaden (1990) that investigated 169 cases from 12 jurisdictions across the U.S. found only 2% of custody cases in the U.S. with CSA allegations. The low rate could plausibly be attributed to cultural reticence about discussing incest at that time. However, a more recent study of California cases by Johnston et al. (2005, T. 2) reported findings from statewide data in 1999 involving 18,000 family court cases in California and found virtually the same percentage of CSA allegations as did Thoennes and Tjaden (1990; 1% by mothers and 2% by fathers), along with 5% of CPA allegations. In the data they collected on families with domestic violence and alcohol abuse allegations, they found much higher rates: 15% of cases had CPA allegations against mothers and 21% against fathers, and 6% had CSA allegations against mothers and 23% against fathers. These more violent and troubled families engaged in extensive litigation and so may create an impression that custody cases contain high rates of CA allegations. Even so, these findings do not support the assumption that there is an “epidemic” (Clawar & Rivlin, 2013) of CSA allegations in custody cases.

Frequency of child abuse allegations in custody vs. non-custody cases in CPS investigations

Since the U.S. Child Maltreatment reports do not differentiate custody and non-custody cases, data discussed here is taken from the national Canadian Incidence Studies (CIS-98, Trocme et al., 2001; CIS-2003, Trocme et al., 2005; CIS-2008, Public Health Agency, 2010) and a regional study from Ontario (Table 2; OIS, 1993; Trocme et al., 1994).

Table 2. Frequency of CA and CSA allegations in custody vs. Non-custody cases Investigated by CPS.

Study, Characteristics CA	CPS Findings CA
	Custody	Non-Custody
Bala and Schuman (1999), OIS-1993, N = 42,000 (Section 4 (b)	9%	91%
Bala et al. (2007), CIS-2003, Oct.-Dec. 2003, N = 11,562 (p. 5)^1	10% (N = 1136/11,562, p. 6)	90%
Black et al. (2016), CIS-2008, Oct.-Dec. 2008, N = 15,980 (p. 148)	12.4% (estimated, p. 151, T. 4)	85.4% (estimated, p. 151, T. 4)^2
Saini et al. (2013), CIS-2003, Oct.-Dec., 2003, N = 11,562 (p. 115)	12% (p. 115)	88%
Trocmé and Bala (2005), CIS-98, N = 7509, Oct.-Dec. 1998, (p. 1340, T. 5)	12% (p. 1340, T. 5)	88% (p. 1340, T. 5)
Study, Characteristics CSA	CPS Findings CSA
Black et al. (2016), CIS-2008, Oct.-Dec. 2008, N = 10,173 (p. 151, T. 4)	5.3% (estimated CSA) (p. 151, T. 4)	4.1% (estimated CSA) (p. 151, T. 4)
Saini et al. (2013), CIS-2003, Oct.-Dec. 2003, N (estimated) = 12, 299 (p. 125, T. 1)	5% (estimated CSA) (p. 125, T. 1)	6% (estimated CSA) (p. 125, T. 1)

¹Bala et al. (2007) and Saini et al. (2013) are both analyses of CIS-2003. Differences in rates reflect sample weights (Bala et al., 2007) vs. national estimates (Saini et al., 2013).

²CIS-2008 estimates do not total 100% because 2% of cases had unknown custody status.

Note. CA = Aggregate of multiple types of child abuse; CPA = child physical abuse; CSA = child sexual abuse, CAN = neglect. When researchers did not provide percentage rates for non-custody cases, they were derived by subtracting the percent of custody cases from 100.

These studies show that CA and CSA allegations in custody cases are a small minority of CA investigations overall. They ranged from 9%–12% of all Canadian CA investigations between 1993–2008 (Bala & Schuman, 1999; Bala et al., 2007; Black et al., 2016; Saini, et al., 2013, T. 1; Trocmé & Bala, 2005). CSA cases were even less frequent and occurred at roughly comparable rates in custody (5–5.3%) and non-custody (4.1–6%) cases (Table 2). There was no empirical support for the claim that CA or CSA allegations are common in custody cases.

Frequency of child abuse allegations by custodial vs. non-custodial parents investigated by CPS and family courts

Empirical findings from the U.S., Canada, and Australia for custodial parents indicate that they communicate more abuse allegations than non-custodial parents to CPS and family courts (Table 3; Bala & Schuman, 1999; Ferguson et al., 2018; Johnston et al., 2005; Saini et al., 2013; Trocmé & Bala, 1999). Thoennes and Tjaden (1990) reported that mothers accused fathers more than eight times as often as fathers accused mothers (48% vs. 6%; T. 1). While they did not report the custodial status of the mothers, 35% of their cases were identifiable as mother-custody cases because they involved fathers’ motions to modify custody (T. 2). This large difference could be interpreted as supporting alienation claims. Bala and Hunter (2015) hypothesized that custodial or residential status allows more opportunity for alienation. Alternatively, custodial parents are usually mothers (Bala & Hunter, 2015; Neilson, 2018; Trocmé & Bala, 2005) and it has long been recognized that children are more likely to disclose abuse by their fathers when they no longer live with them (Bala & Schuman, 1999; Corwin et al., 1987).

Table 3. Frequency of CA and CSA reports by custodial vs. Non-custodial parents or mothers vs. Fathers when custodial status is not reported Investigated by CPS and family courts.

Study, Characteristics	Custodial Parents	Non-Custodial Parents
Bala and Schuman (1999), N = 196, CA reports to Canadian courts (Section 4c – custody reported)	70% (mothers: 64% + fathers: 6%)	17% (mothers: 6% + fathers: 11%)
Ferguson et al. (2018), N = 152, CSA reports to Australian courts (p. 101, T. 1 – custody not reported; mothers usually have residential custody, personal communication, 2/22/22)	79.2% (mothers)	11.6% (fathers)
Johnston et al. (2005), N = 120, CSA reports to San Francisco courts (p. 286, Table 2 – 54% of children lived with mothers)	44% (mothers, 21% CPA, 23% CSA	21% (15% CPA, 6% CSA)
Saini et al, 2013 (p. 124), N = 25,101 (estimated), CIS-2003, reports to Canadian CPS (p. 124 - custody reported)	24%	19%
Bala and Schuman (1999), OIS-93, reports to Ontario CPS, N = 3780 (9% of 42,000), custody not reported (Section 4 b), but Bala and Hunter (2015, p. 12) state that mothers are usually the custodial or residential parent	2/3	1/3

Custodial or residential status does not, however, seem to be a sufficient explanation for rates of abuse allegations in custody cases because non-custodial parents also communicate abuse allegations, though they do so less frequently (Table 3). In Johnston et al.’s (2005) study of violent and alcohol-abusing families where mothers communicated many allegations against fathers, 54% of the mothers had full custody and 45% did not. Bala and Schuman (1999, Section 4c) found that non-custodial fathers communicated abuse allegations more often than custodial fathers (11% vs. 6%). Is the frequency of CA allegations from different sources related to their perceived credibility?

Credibility of child abuse allegations in custody disputes

Family court decisions crediting child abuse allegations in custody cases

Brown (2003) found that family courts validated 22% of CA allegations and 52% when they were managed by a special program designed for CA cases. Between 1990 and 2021, family court judges’ crediting of CSA varied widely from 10.3% to 50% (Table 4).

Table 4. Judicial decisions about the validity of CSA allegations.

Study, Characteristics	CSA Credited	CSA Disbelieved^1
Bala and Schuman (1999), OIS-1993, N = 42,000 (Section 4 (b)	25% (approximately, Sec. 10)	Not reported
Ferguson et al. (2018), Australia, 2013–2015, N = 156 (p. 93)	10.3% (p. 102)	72.4% (p. 102)
Thoennes and Tjaden (1990), U.S., 1986, N = 9000 custody cases (p. 153), N with CSA = 2% (180 cases), N with validity determined = 129 (p. 151)	50% (65 cases)	33% (43 cases)
Webb et al. (2021), Australia, 2012–2019, N = 380 (p. 1)	14% (p. 9, T. 2)	Not reported

¹When the percentage of credited and discredited cases does not total 100%, it is because cases fall into other categories such as indeterminate cases. The percent of discredited cases reported here is that given by the authors of the study cited.

Lower rates of CSA crediting in custody cases might reflect increased cultural acceptance of reporting incest, bringing more ambiguous cases into family courts. Webb et al. (2021, T. 4) found that judges believed that 49% of cases in which the accused continuously denied CSA allegations were genuine errors on the part of the accusers.

Thoennes and Tjaden (1990, T. 6) and Johnston et al. (2005) investigated differences in rates of judicial crediting of mothers’ and fathers’ abuse claims and found them equal. Thoennes and Tjaden’s (1990) used court records, mail surveys, and telephone interviews with court administrators, court counselors, mediators, CPS workers, court evaluators, and judges. Thus, their respondents included relatively untrained people. It is not apparent that court administrators, for example, would be trained in identifying CA cases. Several results suggest that the validity ratings they made reflect an uncritical acceptance of alienation assumptions. One court worker said she was “pretty skeptical of the cases where dad becomes abusive only after the divorce” (p. 159). Her skepticism appears to have been shared with other respondents because CSA allegations were more likely to be perceived as valid if they were made more than 2 years after the divorce (66% v. 41%) and as part of a visitation modification rather than a custody modification.

Johnston et al. (2005) failed to find gender differences for individual types of CA analyzed separately (T. 3). However, for all types of CA combined (T.4), fathers’ allegations against mothers were substantiated more frequently than mothers’ allegations against fathers (46% vs. 26%), which could be interpreted as supporting alienation claims accusing mothers of false CA allegations. The authors contradicted this interpretation because when all types of abuse were combined (i.e., adult, child, and substance) gender disparities disappeared, and they interpreted that to mean that women were not more likely to make false allegations in general.

However, alienation rebuttals of CA allegations do not need to argue that women are generally liars. It is enough to argue that they lie about CA in custody cases. It seems meaningful to ask whether there are family dynamics in these cases that cause mothers to be more distrustful of fathers’ ability to be a safe parent than is warranted. In Johnston et al. (2005) a likely answer could lie in their own victimization by fathers, which was substantiated much more frequently than fathers’ victimization by mothers (41% vs. 15%; T. 3). It is possible that their own victimization made them unreasonably distrustful of their children’s safety with their father. It is also possible that their observations of their children’s reactions to the domestic violence that they experienced gave them a deep appreciation of their children’s fear. Exposure to domestic violence terrorizes a child and is a type of psychological maltreatment (American Professional Society on the Abuse of Children, 2017). Johnston et al.’s (2005) study did not include children’s exposure to domestic violence in their assessment of types of CA. This research cautions against assuming that fathers who are violent with mothers are necessarily violent with their children, but it does not lead to the conclusion that they are less abusive toward their children when terrorizing them is recognized as a type of CA.

CPS substantiation decisions in custody cases

CPS substantiation decisions play a particularly important role in disputes about the likely validity of CA and CSA allegations in custody cases. Two types of statistics are relevant in the various jurisdictions studied. Substantiation rates in custody cases could be compared to general national or state-wide substantiation rates for all cases and/or they could be compared to national or state-wide substantiation rates for non-custody cases. Comparisons between custody and non-custody cases are more specific and provide better data but are not always available. Both statistics are relevant to determinations of whether rates of CA and CSA allegations in custody cases are unusually high and warrant skepticism.

In Canada, CIS-2008 (Public Health Agency, 2010, p. 24) reported that 36% of all investigations were substantiated. In the U.S., Child Maltreatment 2021 reported that 17.8% of investigations had victimized children, of which 16.7% were substantiated and another 1.1% were indicated (USDHHS, 2021, p. 20). Alcohol abuse and domestic violence were identified as risk factors, which is consistent with Johnston et al.’s (2005, T. 4) finding of relatively high judicial crediting of CA (34%) for families with these problems. The low U.S. rate is believed to reflect the impact of the COVID-19 pandemic on reducing services, including investigations, to families.

The only recent large-scale studies that directly compare CA substantiation rates in custody and non-custody cases come from the Canadian Incidence Studies. In 1998 and 2003, these studies substantiated custody cases at lower rates than non-custody cases: 40% vs. 47% (CIS-98, Trocmé & Bala, 2005, T. 5) and 45% vs. 49% respectively (CIS-2003, Saini et al., 2013, T. 2). Lower substantiation rates in custody cases could indicate more false reports, consistent with alienation claims. Alternatively, lower substantiation rates could be caused by children’s resistance to disclose abuse by family members, especially parents (Hershkowitz et al., 2005) and biological fathers (Pipe et al., 2007); less adequate statements when they do disclose (Everson & Boat, 1989; Haskett et al., 1995); and/or less corroborating evidence (London et al., 2007).²

CPS substantiation decisions for custodial parents

The 2021 U.S. Child Maltreatment report found that parents were the most frequent abusers (USDHHS, 2021, T. 3–10), which undermines any perception that CA and/or CSA by parents in custody cases is unusual. Despite lower CPS substantiation rates for custody cases in the Canadian Incidence Studies just discussed, CPS substantiation rates for custodial parents indicate that their reports are substantiated more often than those of non-custodial parents, indicating that they should not elicit automatic skepticism. The Ontario Incidence Study-1993 found that while custodial mothers reported abuse allegations at nearly double the rate that custodial fathers did (2/3 vs. 1/3), the rate at which those allegations were substantiated more than doubled (23% vs. 10%) as well (Bala & Schuman, 1999, Section 4b). CIS-98 found higher substantiation rates for custodial parents, usually mothers, than non-custodial parents (47% vs. 33%; Trocmé & Bala, 2005, T. 2). Substantiation rates for custodial parents in Australia dramatically favored mothers’ compared to fathers’ reports (75% vs. 12.5%; Ferguson et al., 2018). While the lower substantiation rate for alienation cases could be interpreted to support alienation claims, the higher substantiation rate for custodial parents compared to non-custodial parents does not. Rather, it suggests that the breakup of the marriage either provokes child abuse or encourages its disclosure, particularly, but not only, in cases with domestic violence.

Judges’ and CPS caseworkers’ ability to assess CA and CSA allegations accurately depends in part on what children are or are not willing to tell them. Consideration of the factors that affect children’s willingness to disclose is beyond the scope of this article. Regardless of those factors, what does the empirical evidence say about children’s expectable disclosure patterns? To whom do they disclose, when, and how consistently? A plausible interpretation of these findings would be that children confide abuse in one or the other parent depending on the quality of their relationship with that parent, not on their residential status. Do children’s disclosure patterns help or hinder legal efforts to protect them?

Children’s abuse disclosure patterns in custody disputes

Prior disclosures to confidantes

Pipe et al. (2007) found that children typically disclose to personal confidantes before official interviews. Almost all children subject to forensic interviews at every age (4–13 years) had made prior disclosures (78%–82%). Further, almost all the children who made prior disclosures also disclosed in official interviews, though younger children did so less frequently (75%) than did older ones (91%, 6–8 years; 98% 9–13 years; T. 5.2). If children fail to disclose in interviews when parents have claimed that the children disclosed to them, accused parents could plausibly argue that the children never made the prior disclosures. However, this argument is weakened by the finding that only 8% of 4–6-year-old children confided in people outside the family (e.g., teachers; T. 5.3). Young children might not have any understanding of why they should talk to strangers, including interviewers, about something hurtful that they had already revealed to their families.

Prior disclosures were the most common trigger for official investigations in every age group. They were also more likely to be associated with allegations in the official interviews than any other trigger (Pipe et al., 2007, T. 5.4), especially as children grew older (83%, 4–5 years; 89%, 6–8 years; 99% 9–13 years; T. 5.5). Alienation thinking would cast doubt on whether a child’s statements to authorities after talking with confidantes were really their disclosures or whether they were the confidantes’ suggestions (Death et al., 2019). There is a small empirical research literature that demonstrates suggestibility in response to parents’ informal conversations with their children (Lawson et al., 2018; Warren & Peterson, 2014), but it is minimally, if at all, applicable to abuse suggestions because the events suggested were positive or only mildly stressful.

If parents or other family members were suggesting abuse, then younger and presumably more suggestible children would be expected to make more allegations in interviews than would older children when the interviews were triggered by family “concerns” rather than by the children’s prior statements. This did not happen in Pipe et al.’s (2007) study. The same percentage of children in every age group (68%) made allegations to interviewers in cases that were triggered by immediate family concerns (T. 5.5). However, for younger children (4–8 years), the concerns of immediate family triggered many more allegations than did any other trigger (40% vs. 2%–5%) except for the children’s own prior disclosures (47%; T. 5.4). This might appear to indicate family pressure, except that the children in this age group as well as the older ones also failed to make allegations to official interviewers more often in response to immediate family concerns than in response to any other trigger (32%, 4–6 years; 42%, 6–8 years; 32%, 9–13 years vs. 4%–26% across all ages). The interpretation that appears most consistent with both the allegation and non-allegation rates would be that immediate family members are often sensitive in recognizing abuse-related behavior even when children did not disclose to them, especially when children are young. They can also be overanxious, and when this occurs, children resist their anxiety and do not make allegations to interviewers.

Non-disclosure to authorities

Empirical findings indicate that children—especially, but not only when young—were very resistant to disclosing abuse by an immediate family member, a parent, and, particularly, a biological father accused of sexual abuse.

The research reported in Table 5 presents children’s rates of disclosure. Non-disclosure rates were not provided by the authors but are simply the inverse of disclosure rates. The disclosure rates reported by the authors are cited here to make it easier for the reader to refer to the original studies.

Table 5. Children’s CA and CSA disclosure rates (non-disclosure rates are the inverse of disclosure rates).

Study, Characteristics	Against Parents CPA	Against Others CPA
Hershkowitz et al. (2005), Israel, 1998–2002, N = 26,446, 3–14 years old, NICHD interviews (p. 1203)	60.5% (p. 1209, T. 3) 46.8% −3–6 years 63.4% − 7–10 years 66.9% − 11–14 years	63.1% (p. 1209, T. 3) 56.7% 3–6 years 66.3% − 7–10 years 63.3% − 11–14 years
	CSA	CSA
	20.9% (all ages) (p. 1209, T. 3) 15.4% −3–6 years 20.6% − 7–10 years 28.5% − 11–14 years	89.3% (all ages)) (p. 1209, T. 3) 75.2% − 3–6 years 90.0% − 7–10 years 93.2% − 11–14 years
Pipe et al. (2007), U.S., 1997–2000, N = 397, 4 years − 13 years 11 months, NICHD interviews (p. 81)	Against Immediate Family CSA	Against Other CSA
	72% (all ages) (p. 84, T. 5.1) 44% − 4–5 years 70% − 6–8 years 88% − 9–12 years	71%- 4–5 years (p.85, test) 82%- 6–8 years 85% − 9–13 years
	Against Biological Fathers CSA
	41% − 4–5 years (p. 85, text) 47% − 6–8 years 92% − 9–13 years

The findings that across all these studies more than half of young to school-age children in the U.S. (59%) did not disclose CSA by biological fathers, and nearly 80% or more of comparably aged children in Israel did not disclose CSA by parents (79.1%) are sobering facts for family courts. Fathers offering alienation rebuttals of CSA allegations argue that a child’s failure to tell “what happened” means that nothing did. Given children’s resistance to abuse disclosure, non-disclosure to authorities in response to family concerns cannot be assumed to indicate no abuse. As Bala et al. (2007) astutely observed, “while there are legitimate concerns about the possibility that accusing parents or children may be lying or mistaken, those who have abused children usually falsely deny or minimise their abuse.” When a child has not made abuse statements either to family or to authorities, their non-disclosure might indicate that family concerns are innocent mistakes, perhaps misinterpretations of normal sexual behavior or normal divorce-related distress, which would be consistent with Webb’s (2021) finding discussed above that judges tend to interpret continuously denied allegations as innocent errors on the part of the accusing parents. Young children might not disclose CSA because they are not developmentally competent to do so. However, since they, like older children, were more willing to disclose CSA by non-family suspects (Hershkowitz et al., 2005; Pipe et al., 2007), it appears that motivational factors are more likely to be responsible for non-disclosure. Here is where it is particularly important to have an in-depth abuse evaluation by a CA expert knowledgeable and respectful of the possibility that a child’s silence could indicate defensive concealment of abuse or its genuine absence.

Delayed disclosure to authorities

Children may overcome their reluctance to disclose, but only after a delay. If the delay extends beyond the legal decision-making period, it is, in effect, a non-disclosure. However, even if it is made within that period, an accused parent could argue that it was only made after the favored parent had exerted sufficient pressure on the child to compel a false allegation (London et al., 2007). While this interpretation is possible, it could be wrong because delayed disclosure, like non-disclosure, is also a normal response to sexual abuse. Nearly two-thirds of adult incest survivors say they did not disclose when they were children (London et al., 2007). Pipe et al. (2007, T. 5.6) found that 29%³ of young children (4–6 years old) did not disclose CSA within the month after the last abuse incident and 19% did not disclose up to six months after it. Delays of a year or more were infrequent for this age group (3%) but more common for older children (19%, 6–8 years old; 27%, 9–14 years old). For most children, delays were longer when the accused was an immediate or other family member, abuse was more severe, and it occurred over multiple incidents. Children’s normal delays in reporting sexual abuse by fathers would be precisely the circumstances that would make alienation rebuttals appear plausible when they are not.

Recanted disclosure to authorities

When children recant, they may be perceived as taking back statements that were not true in the first place (London et al., 2007). However, Malloy et al. (2007, p. 16) found that in U.S. custody cases, children with substantiated sexual abuse recanted less often (17.5%) than did children overall (23.1%) and at approximately the same rates as children in cases with corroborating evidence (i.e., medical evidence, perpetrator admissions, and multiple victims; 6.7–20%). Recantations were most likely when children were younger and abused by a parent, just as were non-disclosures and delayed disclosures. Support from nonoffending caregivers decreased the likelihood of recantation. However, in custody cases with alienation allegations, parental support could readily be misinterpreted as encouraging false allegations. When recantations, delays, and non-disclosures are used to call children’s credibility into question, their protection depends on the credibility of reporting parents, which is precisely what alienation thinking disparages, often successfully, as discussed above (Ferguson et al., 2018).

Pipe et al. (2007) did not investigate custody cases, the proportion of prior disclosures to immediate family that were made to parents specifically, nor whether the prior disclosures were credible. As a result, their findings do not apply directly to custody cases. However, to the extent that their findings generalize to custody cases, then custodial parents should make more abuse reports than non-custodial parents and their reports would be credible, as the data substantiating custody parents’ abuse reports, discussed above, has found.

Substantiation rates might reveal only a fraction of true abuse cases, reduced by children’s disclosure resistance. Unsubstantiated and substantiated cases are re-reported and substantiated at equal rates within three years of the incident being categorized as unsubstantiated (Kohl et al., 2009). How can the need to compensate for children’s disclosure resistance, on the one hand, and the contrary need to protect innocent accused parents, on the other, be balanced?

CPS and judges’ suspicions about unsubstantiated child abuse in custody disputes

One answer that has already been institutionalized in CPS systems, whether or not there is a custody dispute, and that is acknowledged by judges in custody disputes, is to give increased weight to suspicions of abuse. Canadian judges found residual CA suspicions in 31%⁴ of custody cases (Bala & Schuman, 1999, Section 4c). Canadian CPS caseworkers also found residual abuse suspicions, but results for custody vs. non-custody cases varied. CIS-98 found that custody cases had fewer residual abuse suspicions (14%) than non-custody cases (19%; Trocmé & Bala, 2005, T. 5). However, CIS-2003 found that custody cases had more residual abuse suspicions (19%) than non-custody cases (12%; Saini et al., 2013, T. 2).

Mothers’ CA reports appear to contain residual abuse suspicions more often than fathers’ reports do. OIS-93 also found that 27% of custodial mothers’ reports left residual abuse suspicions whereas 18% of non-custodial fathers did (Bala & Schuman, 1999, Section 4b). Australian judges found residual CSA suspicions in 70% of custody cases where mothers made allegations against fathers compared to 20% of cases where fathers made them against mothers (Ferguson et al., 2018, T. 1). Since mothers are usually the custodial parents, higher rates of residual CA suspicions, like higher rates of CA substantiations, suggest that children are confiding in them or that they are more likely to detect CA cues and become concerned.

Combining CPS substantiations or judges’ credited cases and abuse suspicions

Arguments that question the importance of suspected cases because they are a minority (London et al., 2007) do not respond to the difficulty of obtaining abuse evidence from children against their parents. If residual evidence of abuse reflects the evidentiary obstacles created by children’s resistance to disclose abuse by parents, particularly biological fathers, then a more realistic estimate of CA in custody cases would be achieved by a combined metric that includes substantiated or credited cases and suspicious cases (Table 6).

Table 6. Rates of substantiated/credited and suspected CA in custody cases combined.

Study, Characteristics	Measure	Substantiated/Credited Cases	Suspected Cases	Total
Bala and Schuman (1999, Section 4c), Judges’ decisions − 1990–1998	custody cases	23%	35%	58%
Bala and Schuman (1999), OIS-93 (Section 4b)	custodial mothers + non-custodial fathers	33% (23% + 10%)	45% (27% + 18%)	78%
Trocmé and Bala (2005), CIS-98 (p. 1340, T. 5)	custody cases	40% (custodial parents − 47%; non-custodial parents − 33%)	14% (custodial parents − 24%; non-custodial parents − 15%	54%
Trocmé and Bala (2005), CIS-2003 (p. 1339, T. 2)	custody cases	45%	19%	64%

Consistent with results showing that custodial status does not fully explain the frequency of abuse reports, discussed above (Table 3), Table 6 suggests that custodial status also does not fully explain the level of evidence that is associated with those reports because non-custodial parents make substantiated or credited and suspected reports as well, though at a lesser rate than custodial parents. The distribution of substantiated, credited, and suspected cases across custodial and non-custodial parents suggests that the confiding quality of children’s relationships with their parents rather than their residence is a central factor related to disclosure in many cases.

As CPS systems in the U.S. and Canada recognize, the combined metric clearly increases the number of cases that would warrant child protective measures from family courts (Table 6). Across studies, the range of credible custody cases based only on substantiated/credited cases is 23%–45% and the range of suspected custody cases is 14%–45%. Within each study, the rate of substantiated/credited and suspected custody cases combined is between 54%–78%. This range does not consistently favor or disfavor custody cases. CIS-98 (Trocmé & Bala, 2005, T. 5) would find 66% of non-custody cases credible based on the combined metric. CIS-2003 (Saini et al., 2013, T. 5) would find 61% of non-custody cases credible based on the combined metric. The range of credible custody cases based on the combined metric here (54%–78%) would thus overlap with the range of credible non-custody cases (61%–66%), which suggests that the combined metric would not bias credibility decisions in favor of either custody or non-custody cases. However, the higher end of the combined range did favor custody cases (78% vs. 66%), which could indicate the evidentiary difficulties created by children’s disclosure resistance or more false allegations. Research is needed to investigate the evidence that leaves suspicions of abuse and the reasons why that evidence is considered insufficient to substantiate or credit it. Does the belief that custody disputes are likely to contain false abuse allegations support the need for caution about abuse suspicions?

Frequency of deliberately false CA or CSA allegations in custody disputes

Subjective judgment weighs heavily in determining whether an unsubstantiated or even disproven CA or CSA allegation is deliberately false because there are no objective criteria to assess intent (Black et al., 2016; Brown et al., 2000, 2001; Trocmé & Bala, 2005).

Rates of intentional false allegations by children

Case examples of purportedly false CSA allegations made by children, usually older ones, exist (Bala et al., 2007). However, empirical data indicates that they are very rare. CIS-98 found that only 4% of children’s abuse reports, in general, were considered intentionally false (Trocmé & Bala, 2005, T. 2). CIS-98 also found that in 43 investigations over three months, no child ever made a false CSA report (Trocmé & Bala, 2005, T. 4). CIS-98 further found that caseworkers believed that children’s abuse reports were infrequently deliberately false and virtually as often when there was a custody dispute (2%) and when there was not (3%; Trocmé & Bala, 2005, T. 6). This is the only research to investigate children’s actual false CSA reports. If parents are falsely suggesting CSA to their children, children do not appear to be repeating those suggestions to caseworkers.

Rates of intentional false allegations by adults

In CIS-98, caseworkers believed that 4% of all child abuse investigations nationally contained deliberately false allegations (Trocmé & Bala, 2005, T. 1), usually made by anonymous reporters (25%; Trocmé & Bala, 2005, T. 2). They believed that 6% of CSA allegations nationally had deliberately false allegations, somewhat higher than other types of CA (2%–4%; Trocmé & Bala, 2005, T.3).

Rates of allegations believed to be deliberately false in custody cases specifically were consistently higher than the national rates for children and all CA cases. Australian judges believed that 23% of CSA allegations in heavily litigated custody cases were deliberately false (Webb et al., 2021, T. 4). Canadian judges believed that 30% of CA allegations in custody cases were intentionally false (Bala & Schuman, 1999, Section 4c). These rates are much higher than Canadian caseworkers believed, though they too thought custody cases had more deliberately false allegations than non-custody cases. CIS-98 found that caseworkers believed 12% of deliberately false CA allegations were in custody cases compared to 3% in non-custody cases (Trocmé & Bala, 2005, T. 5). Similarly, CIS-2003 found that caseworkers believed 13% of deliberately false CA allegations were in custody cases compared to 4% in non-custody cases (Saini et al., 2013, T. 2). However, they believed that 3% of deliberately false CSA allegations were in custody cases compared to 4% in non-custody cases (Saini et al., 2013).

Again, as with substantiation rates, results for custodial parents were different than for custody cases. Caseworkers attributed malice more often to non-custodial parents or fathers than to custodial parents or mothers. In OIS-93, they believed 21% of non-custodial fathers’ CA allegations were deliberately false compared to 1.3% of custodial mothers’ allegations (Bala & Schuman, 1999, Section 4b). In CIS-98, they believed that 15% of non-custodial parents’ CA allegations were deliberately false compared to 2% of custodial parents’ allegations (Trocmé & Bala, 2005, T. 2). They also believed that 43% of the deliberately false allegations in custody disputes were made by non-custodial parents, compared to 14% by custodial parents (Trocmé & Bala, 2005, T. 6). In CIS-2003, they believed that 15% of deliberately false CA allegations were made by non-custodial parents compared to 9% by custodial parents, but they considered 14% of custodial parents to be malicious compared to 11% of non-custodial parents where CSA allegations were concerned (Bala et al., 2007). Non-custodial parents are usually fathers (Bala & Hunter, 2015; Neilson, 2018).

These findings contradict the alienation claim that custodial parents, usually mothers, weaponize CA and, especially, CSA allegations. Rather, they suggest the opposite, that non-custodial parents, usually fathers, are the ones who weaponize CA allegations. The perceived low rate of deliberately false CA allegations by custodial parents, usually mothers, lends further support to the hypothesis that they are usually reporting their children’s confidences. Results for judicial and caseworkers’ perceptions of intentionally false allegations indicate that allegations by custodial parents should not be suspected of malice so much as should the allegations of non-custodial parents, a result that is consistent with that found for substantiation rates.

Conclusion

The purpose of this article is to bring empirical findings about child abuse allegations in custody cases together to increase their visibility, impact, and accurate interpretation in light of assumptions about them in custody cases that unfairly and inaccurately bolster the credibility of alienation rebuttals. This review found some results that could appear to provide empirical support for alienation claims but interpreting them in that way would largely require exaggerating small-to-modest differences between custody and non-custody cases, failing to disentangle results for custody cases from results for custodial parents, and isolating some empirical findings from others that affect their meaning.

Results for Canada and the U.S. showed CA allegations in custody cases to be a small proportion of CPS and family court investigations overall. CSA allegations were equally frequent in custody and non-custody cases. The only results that showed unusually high rates of CA allegations in custody cases were for U.S. families with high rates of domestic violence and alcohol abuse. Alienation advocates initially claimed that mothers, subsequently custodial parents of either gender, were responsible for a high frequency of false CA allegations. Results for Ontario, Canada, Australia, and California could be interpreted as providing support for this claim because they showed that CA allegations were substantiated less often in custody cases than in non-custody cases and that custodial mothers reported abuse much more frequently than custodial fathers. However, this interpretation would erroneously equate a lack of substantiation with falsity. Moreover, equating reports in custody cases with reports by custodial parents is misleading because custodial mothers’ and custodial parents’ abuse reports were substantiated more often than non-custodial parents of either gender, which means that they should be reported more often.

The evidence indicating children’s resistance to disclosing abuse by immediate family and parents, especially biological fathers, is a compelling explanation of the difficulty in obtaining legally adequate statements from children, especially young ones. Children’s resistance to disclosing abuse by parents creates serious evidentiary deficiencies that introduce uncertainty into the judicial decision-making process. The imperative to make enormously consequential decisions in the face of these evidentiary deficiencies could make skepticism about CA claims welcome. Believing that CA allegations are common, usually false, and often malicious could provide a rationalization that allows decision-makers to protect accused parents’ constitutional rights while reducing their fears of putting children in danger. It would be better to reduce judicial uncertainty by recognizing that children’s resistance to disclosure requires giving greater weight to abuse suspicions. Using a combined metric would bolster the perceived credibility of CA allegations in custody cases in general and could empower judges to resist unwarranted skepticism in good conscience.

Alienation advocates could also interpret findings that malicious false allegations are more frequent in custody cases than in non-custody cases to support their claim that children in custody cases are often led to make false abuse allegations. However, that interpretation does not bear close scrutiny either because non-custodial parents make more malicious allegations than do custodial parents. Further, malicious false CSA allegations by adults have been found to be infrequent. Finally, alienation advocates’ reliance on suggestibility as the mechanism that could explain how children could be led to make false allegations flounders on the evidence that not one single child was found to make a false sexual abuse allegation in the Canadian research that addressed this. Finally, the higher rates of malicious allegations by adults, including parents, are difficult to interpret because there is a “culture of skepticism” (Death et al., 2019) about abuse allegations in custody cases, no objective criteria to counter it, and studies that document the perception of falsity do not provide the basis for or the accuracy of that perception.

The results of this empirical research review suggest that child custody evaluation practice in cases where alienation claims rebut CA claims is at a crossroads. If experts are going to be able to realistically assess danger to children and put their safety first, they must realize that CPS substantiation or its absence is a weak proxy for truth. They must recognize favored parents’ role as children’s confidantes and abandon ready skepticism of them. They cannot conclude that alienation exists because an allegedly abused child did not disclose convincingly to authorities. They must seek evidence that assesses the complexity of the case (Drozd et al., 2013; Milchman, 2022b, 2022c).

Alienation advocacy defeats these practices by supporting superficial evaluations based on oversimplified beliefs that lead to overconfidence in identifying alienation and dismissing or minimizing abuse allegations made by children and communicated by parents to legal authorities. Hopefully, the empirical findings reviewed here will contribute to improving evaluation practice by transforming alienation assumptions, sustained by the persuasive power of advocacy, into empirical realities tested by the mundane power of evidence.

Judges and experts must accept the need to rely on imperfect evidence if they are going to protect children. While science can inform the court about children’s ability to give reliable testimony and the limits to that ability, and good practice can help improve children’s testimonial reliability, ultimately judges must make a moral choice, not a scientific one. Science is an imperfect ally in identifying truth. Judges must choose whether to prioritize children’s safety in the context of compromised evidence or to prioritize parents’ constitutional rights. On this issue, advocacy shouts but science is silent.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 “Disbelieved cases” affirmatively rejected the CSA allegations. They are not unsubstantiated cases.

2 Parents were counted separately for each type of relationship they had with other abusers.

3 Pipe et al. (2007) report the statistic as 71% disclosed within the month. The non-disclosure rate is the inverse of the disclosure rate.

4 Bala and Schuman (1999) and Bala et al. (2007) cite this as 35%, based on 61/196 cases, which is 31%, Bala agreed to the correction in a personal communication, 2/20/22.