Cossins, Annie --- "Time Out for Longman: Myths, Science and the Common Law" [2010] MelbULawRw 3; (2010) 34(1) Melbourne University Law Review 69

	Home \| Databases \| WorldLII \| Search \| Feedback Melbourne University Law Review

[Because a sexual assault complainant’s testimony is often uncorroborated, sexual assault trials are replete with common law warnings that are designed to draw attention to the unreliability of the complainant’s evidence, particularly where there has been a delay in complaint. This article examines the validity of such common law safeguards in child sexual assault trials. A detailed review of the psychological literature about the patterns of disclosure of sexually abused children reveals that delay in complaint is a typical feature of child sexual abuse. Accordingly, common law warnings about delay in complaint, such as the Longman warning, do not adequately consider the context in which child sexual abuse, and its disclosure, occur. This article argues that the Longman warning, and other similar warnings, should be abolished. It examines the limitations of the warning and the legislative attempts that have been made to mitigate its impact and makes recommendations for future reform.]

I INTRODUCTION

Medical, eyewitness and other corroborating evidence is usually unavailable in most child sexual abuse cases^[1] which means that the child complainant’s testimony will be central to the prosecution’s case. Because of longstanding beliefs about the reliability of the evidence of sexual assault complainants, sexual assault trials are replete with so-called safeguards in the form of specific common law warnings that are aimed at highlighting the unreliability of the complainant’s evidence particularly where there is no corroborating evidence and/or when there has been a delay in complaint. In the Australian context, these warnings include the Longman,^[2] the Crofts^[3] and the corroboration warnings.^[4]

Recent studies have shown that the consistency and credibility of child complainants is significantly associated with verdict.^[5] They also show that the pre-existing beliefs of mock jurors about sexual assault have more impact on verdicts than the evidence presented at trial.^[6] This means that negative suggestions about children’s evidence made either during cross-examination by the defence or by way of judicial warnings are likely to have an impact on verdicts by reinforcing a range of misconceptions that jurors hold about children and child sexual abuse.^[7] Such warnings are given in a context where it is not yet common practice in Australia for expert witness testimony to be admitted to correct these misconceptions.^[8]

This article examines the validity of these common law safeguards in light of what is known about delay in complaint. In particular, it argues that one of the common law warnings that is likely to have an impact on juror perceptions of complainant credibility, the Longman warning, should be abolished. The article begins with a review of the psychological literature about the patterns of disclosure of sexually abused children. After surveying 11 studies over a 23-year period, for the first time in the literature this article documents the key features of children’s patterns of disclosure as a result of being sexually abused. It then considers the implications of these findings for the continued use of the Longman warning in child sexual assault trials. In particular, it documents the criticisms that have been made of the Longman warning, its limitations and the various legislative attempts that have been made to ameliorate its impact. After a review of these attempts the article concludes that further reforms are required and makes recommendations for future reform.

II STUDIES THAT REVEAL THE PATTERNS OF DISCLOSURE OF SEXUALLY ABUSED CHILDREN

An examination of several studies conducted over a period of more than 20 years shows that there are particular patterns of disclosure by children who have been sexually abused, contrary to the assumptions underpinning the legal interpretation of evidence of delayed disclosure. This summary draws on information reported in studies on the prevalence of child sexual abuse within the general community, as well as studies that have specifically examined the patterns of disclosure of sexually abused children using representative samples from the general population (also known as retrospective studies) which allow ‘longer periods of delayed disclosure … [to] be identified.’^[9]

Not all prevalence studies examine the issue of disclosure and its timing so only those studies that have reported such information are discussed here. For example, Russell only reported on disclosure to the authorities, but did not otherwise report the timing of disclosure.^[10] Of the women in her American community sample who reported at least one experience of sexual abuse before the age of 18, only 2 per cent of intra-familial sexual abuse cases and 6 per cent of extra-familial sexual abuse cases were ever reported to the police.^[11] Similarly, Baker and Duncan found that, out of the men and women in their British study who reported being sexually abused before the age of 16, only 12 per cent of female respondents and 8 per cent of male respondents disclosed the abuse.^[12]

The first national prevalence study of child sexual abuse in the United States reported that a majority of respondents (56 per cent of men and 57 per cent of women) did not report the abuse within a year of its occurrence and a significant proportion of respondents never reported the abuse to anyone (42 per cent of men and 33 per cent of women).^[13]

A study of a community sample of New Zealand women reported that only 37 per cent of victims disclosed within one year of the abuse, 10 per cent disclosed between one to 10 years after the abuse, 24 per cent disclosed 10 or more years after the abuse and 28 per cent had not disclosed before the survey. Only 7.5 per cent of victims ‘had the abuse reported to either social work or police investigators.’^[14] The authors found that ‘[t]here were differences in reporting patterns for relationship with the abuser, with those abused by a close family member being significantly less likely to report the abuse within a year, compared with other victims’.^[15] When respondents to the survey were asked what prevented them from disclosing the abuse, 29 per cent said they expected to be blamed, 25 per cent said embarrassment, 24 per cent said not wanting to upset anyone, 23 per cent expected they would be disbelieved, 18 per cent said they were not bothered by the abuse, 14 per cent said they wished to protect the abuser, 11 per cent said fear of the abuser and 3 per cent said obedience to adults.^[16]

In a study of a community sample of 710 Australian women, in which 35 per cent reported some sexual abuse or experience that was unwanted or distressing during childhood, Fleming found that only 10 per cent of abuse victims reported the abuse to the police, a doctor or other agency, such as a sexual assault service.^[17] Just over half of the women who disclosed sexual abuse by an adult involving sexual contact (52 per cent) had revealed the abuse whilst another 5 had tried to disclose unsuccessfully, although the patterns of disclosure varied.^[18] Of the 80 women who disclosed or tried to disclose, 23 did so at the time of the abuse, 7 within the first year, 14 between 1 and 10 years after the abuse, whilst 36 ‘did not disclose until at least 10 years after the first abuse episode.’^[19] Overall, only 21 per cent had disclosed or attempted to disclose within the first year of the abuse, as seen in Table 1.

Table 1: Percentage of Women Who Disclosed or Tried to Disclose (n = 144)^[20]

Fleming also reported that ‘there were significant differences in the timing of disclosure … by age at time of abuse. Girls aged under 12 years at the time of the abuse were less likely to tell someone within a year of the abuse than were girls aged over 12 years.’^[21] In 49 per cent of cases, mothers were the person most frequently told of the abuse, followed by friends (32 per cent) and siblings (29 per cent).^[22] When asked what prevented disclosure

Rates of disclosure ‘showed a significant decrease with age’, with 83 per cent of women aged 17–24 years having disclosed the abuse, compared with 59 per cent of women aged 25–35 years, 51 per cent aged 35–44 years and 38 per cent aged 45 years or over.^[24]

In a study of a national population sample of American women about their abuse experiences and patterns of disclosure, Smith et al ‘gather[ed] information about the length of time that women who experienced a rape in childhood delayed before disclosing their experiences to others, and to whom such disclosures were typically made.’^[25]

Out of 3220 women, 288 (9 per cent) retrospectively reported having experienced one rape prior to their 18^th birthday, with child rape being defined as any type of penetration involving threat or force.^[26]

In relation to disclosures, 28 per cent revealed they had never told anyone about being sexually assaulted before being interviewed for the study.^[27]

The remaining 72 per cent (207 women) had told at least one person before the interview, with a majority of women (52.1 per cent) reporting delays in disclosure of 60 months (five years) or more,^[28] as set out in Table 2. Only 17.8 per cent had reported within 24 hours of being raped whilst 37.7 per cent had reported within the first year.^[29] Thus, four out of five victims did not tell anyone within the first 24 hours and only just over one quarter had reported the rape within one month.^[30]

Table 2 includes the 81 women who had never disclosed until the study but excludes 52 women who could not estimate with certainty when they disclosed (as a group, they were significantly younger at the time of the rape than the rest of the sample).^[31] The data in Table 2 shows that, similar to the data in Table 1, less than 20 per cent of victims make an immediate report and that a majority do not make a report for several years.

Table 2: Delays between Sexual Penetration and Initial Disclosure (n = 236)^[32]

For those who disclosed, the most common types of confidants were friends (22.5 per cent), mothers (20.7 per cent) or other relatives (20.9 per cent), with only 6.6 per cent disclosing to someone in authority (the police, a social worker or clergy) as the first confidant.^[33] Overall, only 12 per cent stated that the assaults were eventually reported to the authorities at some stage,^[34] a figure that is similar to other studies in the literature.^[35]

Smith et al also sought to identify the predictors associated with delayed disclosure, in terms of victim, crime and perpetrator characteristics.^[36]

They found that victims who fell into the long delay group (delay in disclosure of more than one month) were on average significantly younger (by 2.3 years) than the short delay group (disclosure within one month).^[37] The long delay group was also more likely to have experienced more than one rape over a period of months or years.^[38] Perpetrators of the short delay group of victims were over four times more likely to be strangers than those of the long delay group.^[39] Long delay perpetrators were twice as likely to be immediate or extended family members of the victim.^[40] Thus, ‘being related to the perpetrator was associated with longer delays before telling’ whereas the absence of a familial relationship between victim and perpetrator was related to more immediate disclosure.^[41]

All in all, Smith et al found that four variables — age, single versus multiple rapes, biological relationship and a perpetrator who was a stranger to the victim — were significantly different between the long delay and short delay groups.^[42] Rape by a stranger was found to be ‘the best individual predictor of whether a child would tell someone … relatively quickly’,^[43] although it must be remembered that stranger abuse is the least common type of sexual abuse.^[44]

When the effects of other variables were controlled, ‘older age … was significantly predictive of disclosure within 1 month’ of the rape^[45]

and ‘penile–vaginal penetration was more common among those women who told within 1 month than among those who did not’, with digital penetration being more common among non-disclosers.^[46] This is contrary to findings by Sauzier^[47] and Arata^[48] that less intrusive forms of abuse are more likely to be reported than penetration but may be explicable on the ground that Smith et al were only comparing incidents of rape, as opposed to all types of sexual abuse.^[49]

Overall, these results support the conclusions from prevalence studies that delayed disclosure of child sexual abuse is common, including delays of one-year or more. In fact, Smith et al conclude that ‘the very long latencies prior to disclosure reported by women in this sample suggest that the phenomenon of delayed disclosure is more prevalent, and that the typical length of delay is longer, than previous research has revealed.’^[50]

In a second national population study, Kogan conducted telephone interviews of a random sample of adolescents aged 12–17 years as part of the National Survey of Adolescents (n = 4023) in the United States.^[51] 1958 were female and of these, a sub-sample of 263 (13 per cent) adolescent females reported at least one unwanted sexual experience (‘USE’) and provided information about the type of abuse and the perpetrator, as well as when and to whom they disclosed the abuse.^[52]

A majority of the respondents were adolescents at the time of onset of the USE (29 per cent were 11–13 years; 35 per cent were 14–17 years), with 55 per cent reporting that the perpetrator was a peer.^[53] Seventy-nine per cent of respondents knew the perpetrator, with 24 per cent reporting that the perpetrator was a family member.^[54] Most respondents reported single USEs although ‘a third (34 per cent) involved multiple events by the same perpetrator.’^[55] Penetration occurred in 37 per cent of incidents, whilst 30 per cent of respondents reported that they were ‘afraid they might be killed.’^[56]

Respondents were grouped into three categories according to their disclosure patterns:

• delayed disclosers (12 per cent disclosed within one year of the USE, whilst 19 per cent disclosed more than a year later, creating a total of 31 per cent disclosing more than a month after the USE);

Thus, a majority of respondents (57 per cent) did not disclose the USE within one month.

Authority figures such as police, teachers and clergy were least likely to be a victim’s confidant (6 per cent) with most confidants being friends (36 per cent), mothers (35 per cent), and other relatives (8 per cent).^[58] These findings accord with those reported by Smith et al^[59]

and Ullman and Filipas who found that out of 167 college students who reported sexual abuse in childhood, 81 per cent disclosed to informal sources with the remainder reporting to ‘both formal (eg, police, medical, religious or mental health professionals) and informal sources.’^[60]

The older the child, in particular those aged 14–17 years, the more likely they were to disclose to peers, and immediate disclosure occurred most often to peers.^[61]

However, for 14–17 year olds, the information they shared with peers was more likely to be an USE perpetrated by a peer, rather than other types of sexual abuse.^[62] This type of disclosure by adolescents, therefore, appears to be a result of the nature and importance of peer relationships in adolescence, as well as the topics adolescents talk about, such as sex and sexual relationships, something that is not likely to be the same for younger children.^[63]

Kogan undertook a series of analyses to determine the association, if any, between disclosure, age and other characteristics. He found a positive association between age of onset of the USE between 11–13 years and disclosure within one month of the USE and a negative association between immediate disclosure and age of onset of the USE under 7 years.^[64]

Serial incidents of sexual abuse were also associated with delayed disclosure.^[65]

Similar to the findings of Smith et al, Kogan found that if the perpetrator was a stranger this was positively associated with immediate disclosure.^[66] Furthermore, ‘[a] family member perpetrator was negatively associated with immediate disclosure and positively associated with non-disclosure.’^[67] In other words, the closer the relationship between child and perpetrator, the less likely the child will disclose immediately, which is consistent with the findings in a number of other studies.^[68]

Disclosure of sexual abuse by a stranger is probably less likely to cause family disruption or to invite blame, so for the child ‘there are fewer potential costs’ associated with disclosing stranger abuse.^[69]

Indeed, ‘[w]hen the perpetrator is a significant caregiver, then attachment issues, traumatic bonding, and the child’s need to protect the integrity of the family unit are … possible explanations for withholding or delaying disclosure’.^[70]

Using multi-variate analyses, Kogan found that age of onset, knowing the perpetrator, having a family member as perpetrator and having a drug user in the household were all factors associated with delayed disclosure.^[71]

• children aged 11–13 years at the onset of the USE were 67 per cent less likely to disclose compared to those aged 14–17 years at onset;

• children aged 0–6 years at time of onset were five times more likely to delay disclosure compared to those aged 14–17 years at onset;

• children who knew their perpetrator were three times more likely not to disclose and 3.7 times more likely to delay disclosure than to disclose within a month;

• children who were abused by a family member were 5.6 times more likely to delay disclosure than disclose within a month;

• children who had a drug user in the household were 78 per cent less likely to disclose than to disclose within a month.^[72]

Age was a key factor associated both with disclosure and to whom children disclosed, thus confirming the findings from a number of other studies that have shown that ‘older age has been associated with purposeful disclosures’.^[73]

Kogan speculates that the younger the child the less likely that he or she will be able ‘to surmount the barriers to disclosure including such factors as developmental stage and susceptibility to perpetrator tactics for maintaining secrecy.’^[74]

In a study involving 218 children whose cases of alleged sexual abuse had been referred to prosecutors’ offices, Goodman-Brown et al investigated the factors that contributed to children’s disclosures.^[75]

Children subject to intra-familial sexual abuse took longer to disclose compared to those who suffered extra-familial abuse.^[76] In summary, Goodman-Brown et al were able to confirm that ‘age, type of abuse (intra-familial or extra-familial), fear of negative consequences, and perceived responsibility all contributed either directly or indirectly to the length of time it took for children to disclose sexual abuse.’^[77]

Finally, London et al conducted an evaluation of retrospective studies in which adults had reported experiences of child sexual abuse (general population samples) and studies that specifically recruited adults with childhood histories of sexual abuse, in order to evaluate rates of disclosure of sexual abuse during childhood.^[78]

In 6 out of the 11 studies that were evaluated, they found that the modal rate of childhood disclosure of sexual abuse was just over 33 per cent while another 3 studies reported disclosure rates of 42 per cent to 54 per cent during childhood.^[79] However, the percentage of children reporting to the authorities was much lower, from 10 per cent to 18 per cent.^[80] London et al concluded that ‘[g]iven the differences in methodology, definitions of abuse, and sample characteristics [between 10 of the studies], the general consistency of these findings … is noteworthy.’^[81]

III SUMMARY OF THE LITERATURE

The above studies provide evidence of at least 10 consistent patterns of children’s reactions to sexual abuse (‘consistent’ is defined here as being reported in at least two studies), which are summarised in Table 3. Three other patterns are also listed, but they are not yet supported by sufficient data. Other studies are also cited where their findings support these patterns.

Table 3: Patterns of Disclosure of Sexually Abused Children

Patterns	Studies	Type of Sample
1. A majority of children do not disclose immediately or within one month of sexual abuse.	Baker and Duncan (1985)	Probability sample (UK)
	Finkelhor et al (1990)	Community sample (US)
	Anderson et al (1993)	Community sample (NZ)
	Fleming (1997)	Community sample (Australia)
	Smith et al (2000)	National population sample (US)
	Kogan (2004)	National population sample (US)
	Alaggia (2004)	Mixed clinical/non-clinical (US)
	Ullman and Filipas (2005)	College students (US)

Patterns	Studies	Type of Sample
2. A majority of children only disclose sexual abuse one or more years after it occurred, or not at all.	Finkelhor et al (1990)	Community sample (US)
	Anderson et al (1993)	Community sample (NZ)
	Roesler and Weissmann Wind (1994)	Clinical sample (US)
	Smith et al (2000)	National population sample (US)
	London et al (2005)	Review of 11 retrospective studies
3. Children abused by strangers are more likely to disclose within one month.	Arata (1998)	Clinical (US)
	Smith et al (2000)	National population sample (US)
	Kogan (2004)	National population sample (US)
3a. Stranger abuse is a key predictor of rapid disclosure.	Smith et al (2000)	National population sample (US)
4. Children abused by family members are more likely to delay disclosure longer than one month.	Sauzier (1989)	Clinical sample (US)
	Anderson et al (1993)	Community sample (NZ)
	Arata (1998)	College students (US)
	Smith et al (2000)	National population sample (US)
	Goodman-Brown et al (2003)	Forensic sample (US)
	Kogan (2004)	National population sample (US)
	Hershkowitz et al (2005)	Forensic sample (Israel)
5. Repeated abuse is more likely to occur if the abuser is a relative.	Fleming (1997)	Community sample (Australia)
Patterns	Studies	Type of Sample
6. Children who experience multiple abuse are less likely to disclose.	Arata (1998)	College students (US)
	Smith et al (2000)	National population sample (US)
	Kogan (2004)	National population sample (US)

7. Less intrusive forms of abuse are more likely to be reported.	Sauzier (1989)	Clinical sample (US)
	Arata (1998)	College students (US)
	cf Smith et al (2000)	National population sample (US)
8. The younger the child at onset of abuse, the less likely she or he will disclose.	Fleming (1997) (<12 yrs)	Community sample (Australia)
	Smith et al (2000)	National population sample (US)
	Kogan (2004) (<7 yrs)	National population sample (US)
	Hershkowitz et al (2005)	Forensic sample (Israel)
9. Threats and use of force may be associated with delay; data inconclusive.	Smith et al (2000)	National population sample (US)
	London et al (2005)	Review of 11 retrospective studies
10. Authority figures (eg, police) are the least common type of confidant.	Russell (1983)	Community sample (US)
	Anderson et al (1993)	Community sample (NZ)
	Fleming (1997)	Community sample (Australia)
	Arata (1998)	College students (US)
	Smith et al (2000)	National population sample (US)
	Kogan (2004)	National population sample (US)
	Ullman and Filipas (2005)	College students (US)
Patterns	Studies	Type of Sample
11. Friends are the most common type of confidant. 11a. Disclosure to peers increases with age.	Smith et al (2000)	National population sample (US)
	Kogan (2004)	National population sample (US)
	London et al (2005)	Review of 11 retrospective studies
	cf Fleming (1997) (mothers most common)	Community sample (Australia)
	cf Arata (1998) (mothers most common)	College students (US)
12. Embarrassment, shame, being blamed or feeling responsible for the abuse are key factors that prevent children from reporting.	Anderson et al (1993)	Community sample (NZ)
	Roesler and Weissmann Wind (1994)	Clinical sample (US)
	Fleming (1997)	Community sample (Australia)
	Goodman-Brown et al (2003)	Forensic sample (US)
13. Family dysfunction such as having a drug user in the household may contribute to delays.	Kogan (2004)	National population sample (US)

From the studies set out in Table 3, it is possible to conclude that delay is a typical, rather than an aberrant, feature of child sexual abuse. This broad range of studies is significant in that it spans a 23-year period and shows that the patterns of disclosure are remarkably similar in relation to both general population samples (representative samples) and clinical/forensic samples (unrepresentative samples). The findings from these studies indicate that if a child does not report immediately or within one month it is likely that he or she will either not disclose for some years or not disclose at all.

Failure to disclose is influenced by factors such as the type of perpetrator (for example, a stranger), closeness of the relationship between offender and victim, the child’s age at the onset of the abuse, repeated abuse and possibly threats, use of force and family dysfunction.^[82]

There is evidence that children use non-verbal behaviour to try to communicate that something is wrong, such as ‘clinging, temper tantrums … angry outbursts in adolescents, withdrawal, avoiding being at home and/or running away’, as well as mood swings and indirect verbal hints such as refusing to go home after school or asking a mother not to go to work.^[83]

However, the inability of children to directly express what has occurred to them and the failure of adults to interpret behavioural attempts to disclose compounds the problem of nondisclosure.^[84]

This means there are likely to be a substantial number of children in the community who have been abused, but are unable to directly disclose or unwilling to do so because of shame, fear or the need to protect their family. In particular, these findings ‘suggest that the abuse that may be of most harm (long-term abuse by a parent or other relative) is the type of abuse that is least likely to be disclosed.’^[85]

These patterns of disclosure have major implications in terms of the ongoing abuse of children and the selection of new victims over time by offenders, as well as for policing, timely prosecution and the conduct of the child sexual assault trial. In the context of the trial process, these characteristics raise a question mark about the legitimacy of warnings based on delay in complaint and the relevance of cross-examination about delay. Should the trial process be implicated in silencing the voices of victims through the use of warnings, which are based on the outdated premise that delay is indicative of fabrication?

IV JUDICIAL WARNINGS

In a sexual assault trial, a jury may receive a number of different directions from the trial judge which focus on the unique characteristics of sexual assault such as delay, the existence of only one witness to the offence and a lack of corroborating evidence.^[86] Historically, children and women who complained of sexual abuse were treated by the common law as unreliable witnesses, such that juries were warned of the danger of convicting on their uncorroborated evidence.^[87]

Corroboration warnings also applied to accomplices as a class of witness, creating an interesting analogy between the credibility of accomplices to a crime and women and children as victims of crime.^[88]

In fact, corroboration warnings ‘made categorical assumptions about the credibility of whole classes of witness irrespective of the circumstances of the case.’^[89]

The rationale for doubting the credibility of children and sexual assault complainants can be traced back to the late 17^th century when Hale wrote that rape ‘is an accusation easily to be made and hard to be proved, and harder to be defended by the party accused, tho’ never so innocent.’^[90] Indeed, since that time various formulations of the need for corroboration have existed both at common law and under legislation, with the oft-repeated judicial belief that allegations of sexual abuse ‘were very easy to fabricate, but extremely difficult to refute’ becoming, over time,^[91] a ritual incantation.^[92]

In the face of empirical evidence which shows that child sexual assault is significantly under-reported, that attrition rates are high and conviction rates are low,^[93] it is time for this myth to be consigned to the pages of legal history. Yet even in recent times judges have expressed long-held prejudices as facts. For example, in Bromley v The Queen, Brennan J stated that ‘courts have a sharpened awareness of the danger of acting on the uncorroborated evidence of … witnesses’ such as children and sexual assault complainants.^[94] In Longman v The Queen, Deane J stated that ‘[t]he possibility of child fantasy about sexual matters, particularly in relation to occurrences when the child is half-asleep or between periods of sleep, cannot be ignored.’^[95] Similar views were expressed by McHugh J, who believed that ‘[r]ecollection of events which occurred in childhood is particularly susceptible to error and is also subject to the possibility that it may not even be genuine’.^[96]

The common feature of each of these statements is that a particular subjective view is elevated to fact with no empirical evidence to support it. As Spigelman CJ has observed ‘[m]any judges share a conventional wisdom about human behaviour, which may represent the limitations of their background. This has been shown to be so in sexual assault cases.’^[97] After a review of the substantial psychological literature indicating that even very young children can give reliable evidence, his Honour noted that ‘[t]he complexity of these issues is not reflected in the observations of Deane J and McHugh J in Longman v The Queen, which should, accordingly, be treated with caution.’^[98]

Although the requirement to give a common law corroboration warning has been abolished in all Australian jurisdictions^[99] such warnings may still be given since the warning itself is not prohibited.^[100] Other attempts to ameliorate the effect of corroboration warnings were introduced into the Evidence Act 1995 (NSW) (in the form of s 165A and the former s 165B) but ‘they have been significantly undermined by the development of a new class of common law warnings which bear many of the hallmarks of the traditional corroboration warning.’^[101]

Indeed, this development, led by the decision in Longman v The Queen and expanded in Crampton v The Queen^[102] and Doggett v The Queen,^[103] has resulted in the reinstatement of ‘a near mandatory warning regime in relation to a number of categories of evidence, including … evidence of delayed complaint in sexual assault cases’.^[104]

This means that legislative attempts to circumvent the common law corroboration warning in relation to sexual assault complainants (and the prejudice inherent in it) have, in turn, been circumvented by the judicial obsession with the reliability of the evidence of sexual assault complainants and a new class of warnings. Since the uniform Evidence Acts (‘UEAs’)^[105] preserve, under s 165(5), the power of trial judges to give common law warnings, the Longman warnings and others are still given in both common law and UEA jurisdictions.^[106] Trial judges still retain the discretion to give a corroboration warning, although they may be restricted in what they can say.^[107] For example, s 294AA of the Criminal Procedure Act 1986 (NSW) prevents a judge from warning or suggesting to a jury that complainants are a class of unreliable witness and prohibits warning a jury of the danger of convicting on the uncorroborated evidence of a complainant.

The specific directions or warnings that a jury can receive in a sexual assault trial were summarised in R v BWT by Wood CJ at CL.^[108] Although this is a New South Wales case, it usefully highlights the complexity of a trial judge’s task when instructing the jury and the intellectual struggle that lay jurors face, with the possibility that they may have to absorb and understand up to eight common law warnings.^[109]

Along with warnings that may be requested under the UEA and ‘other standard directions customarily given in a criminal trial … a trial judge is faced with a somewhat formidable task in sufficiently directing a jury in this category of case.’^[110] Indeed, in R v LTP, Dunford J (with whom Simpson and Howie JJ agreed) stated that in sexual assault trials, judges ‘would be well advised to use the list of Wood CJ at CL in R v BWT … as a check list’ on the grounds that ‘it is preferable to give the directions, even if the judge considers one or more of them unnecessary in the particular case, rather than have convictions upset on appeal because of the failure to give them.’^[111] This view is supported by data that shows that in New South Wales between 2001 and June 2004, 54 per cent of successful appeals were allowed on the basis of a misdirection given by the trial judge.^[112]

Such advice, however, may result in juries being misled and confused where directions are not necessary or relevant to the circumstances of a particular case. For example, the New South Wales Law Reform Commission (‘NSWLRC’) has recently discussed the limits of juror understandings of ‘long and complex sets of instructions … and the failure to use English that lay people can understand easily’ as well as the increasing length of summings-up.^[113] After a survey of the available jury research, the Commission concluded that ‘jurors do not have the high level of comprehension they thought they had, or that they did, in reality, misunderstand or have problems with specific directions.’^[114] This means that ‘[i]ncomprehensible jury instructions may prevent justice from being seen to be done’ and raise the more practical question of ‘whether justice can actually be done.’^[115]

The typical features of child sexual assault are delay in complaint, the existence of only one witness to the crime and lack of corroboration.^[116]

Because these factors have been interpreted as being indicative of fabrication (rather than being associated with children’s fears of reporting and how child sex offenders perpetrate sexual abuse), the criminal justice system is limited in its ability to tackle the incidence of child sexual abuse in the Australian community. The common law’s apparent wish for perfect child sexual assault cases in which there are victims who report immediately, eye-witnesses and other corroborating evidence is unlikely to be granted. Rather than continuing to view children as unreliable witnesses because of these factors, it is important to examine how the criminal justice system can be reformed to accommodate the reality of child sexual abuse.

V FORENSIC DISADVANTAGE AS A RESULT OF DELAY IN COMPLAINT: THE LONGMAN WARNING

The Longman warning has attracted a great deal of comment and criticism in recent times, from the New South Wales Court of Criminal Appeal in R v BWT,^[117] the New South Wales Standing Committee on Law and Justice,^[118] the Tasmania Law Reform Institute (‘TLRI’),^[119] as well as from the Victorian Law Reform Commission (‘VLRC’)^[120] and the recent joint discussion paper and report by the Australian Law Reform Commission (‘ALRC’), the New South Wales Law Reform Commission (‘NSWLRC’) and the VLRC.^[121]

The warning was also the subject of extensive discussion by the New South Wales Criminal Justice and Sexual Offences Taskforce, which made various recommendations to the New South Wales Attorney-General about the conduct of sexual assault trials in NSW.^[122]

There have been legislative attempts over the years to counter the myths and stereotypes which characterise sexual assault trials in the form of warnings that require a judge to inform the jury that delay in complaint does not necessarily indicate that an allegation is false.^[123] Despite these reforms, High Court judges have continued to perpetuate these cultural beliefs through the development of the Longman warning and later the Crofts warning, which together have reinstated ‘the traditional beliefs and prejudices about sexual assault complainants … and … created significant difficulties in practice for trial judges and appellate courts.’^[124]

The Longman warning means that where there has been a long delay in complaint the jury will be directed that, because of the passage of so many years, it would be dangerous to convict on the uncorroborated evidence of the complainant alone ‘unless the jury, scrutinizing the evidence with great care … were satisfied of its truth and accuracy.’^[125] The warning is based on the rationale that a significant delay puts the accused at a forensic disadvantage because he has lost the ‘means of testing the complainant’s allegations which would have been open to him had there been no delay’.^[126] Subsequent High Court cases have reaffirmed the necessity of the Longman warning where there has been a long delay,^[127] even where there is corroborating evidence of the complaint.^[128] Lewis observes:

Worryingly, the Longman warning has more or less developed into a mandatory warning, since it is ‘required to be delivered in almost every case involving delay’^[130]

such that a trial judge’s failure to give the warning in the terms prescribed by Longman, Crampton and Doggett is likely to result in a successful appeal if the accused is convicted.^[131] As the TLRI has observed:

This ‘retreat to safety’ by trial judges gives ‘the impression … that judges are again, by a back door, treating complainants … as ordinarily unreliable witnesses’.^[133] Indeed, the unequivocal and mandatory nature of the warning that it is ‘dangerous to convict’ on the complainant’s evidence ‘must cast all complainants who delay in making complaint, as unreliable’ and not to be believed.^[134]

In particular, the Longman warning must be couched in the form of a warning, not merely a guide or comment.^[135] Despite some views that no particular form of words is required,^[136] the warning is considered to be very specific in terms of its content:^[137]

One of the key problems with the wording is that it focuses on the truth and accuracy of the evidence of the complainant rather than the forensic disadvantage to the accused, so that the warning is ‘remarkably close to the full corroboration warning previously required at common law.’^[139]

This can be seen in the formulation by Sully J in R v BWT, which has been adopted in subsequent cases,^[140] that the Longman warning must include:

The VLRC considers that the Longman warning ‘offers almost no guidance to trial judges on the circumstances in which a warning might be required’^[142]

and the length of the delay that would trigger it, with a number of cases showing that there is broad variation between trial judges on whether the warning is to be given and the strength and content of it:

Similarly, Sully J has observed that the decisions of the High Court do not give trial judges any guidance as to the length of the delay where a Longman warning will not be required so that ‘the only prudent approach of a trial judge is one that regards any delay between offence and complaint as sufficient to raise for consideration the need for a Longman direction.’^[144] This means, somewhat irrationally, that delays in the order of days, weeks and months could invite the warning^[145] as judges strive to appeal-proof their decisions, thus amounting to ‘the effective removal of the discretion of the trial judge, who is arguably in a far better position than an appellate court to determine whether a warning is necessary in the particular circumstances of the case’.^[146]

More recent interpretations of the Longman warning mean that it has ‘moved a considerable way from the rationale underpinning the majority judgment in Longman’ which considered that after a long delay of 20 years the accused was likely to have faced significant obstacles in mounting his defence.^[147] This movement away from the original rationale means there is now a requirement for the warning wherever there are factors that are perceived to affect the reliability of the complainant’s evidence, such as age and inconsistencies in evidence and, quite oddly, circumstances that are peculiarly associated with sex offender behaviour, such as sexual abuse commencing whilst a child is asleep.^[148]

The import of the warning is such that it may convey ‘the unmistakable message … to a lay jury as to the Judge’s own assessment of the case’^[149]

and ‘as a not too subtle encouragement … to acquit’.^[150] The VLRC has also observed that a strongly worded Longman warning, particularly if followed by a direction about lack of corroborating evidence, may be seen as a direction to acquit.^[151]

In addition, the Longman warning can positively mislead the jury into believing that the accused has suffered a forensic disadvantage despite the possibility that he may not have.^[153] The Longman warning requires a direction to be given in every case involving substantial delay, irrespective of whether the accused has in fact been disadvantaged by the delay:

The difficulty with the reasoning behind the Longman warning is that it ‘elevates the presumption of innocence … to an assumption that the accused was in fact innocent’ and would have been able to rebut the complainant’s evidence had there been no delay in complaint.^[155] This reasoning is flawed if, in fact, the accused did commit the offence, since there would be no rebuttal evidence such as alibi evidence.^[156] Lewis considers that this presumption of forensic disadvantage ‘has the potential to mislead the jury and usurp its fact-finding function’^[157]

because the Longman warning essentially amounts to a finding of fact that the accused ‘was unable by reason of the delay to test or to meet the prosecution case.’^[158]

No appeal case has ever referred to any evidence in the form of documented miscarriages of justice to support the view that it is dangerous to convict on the basis of a complainant’s evidence after there has been a delay in complaint. Indeed, the traditional view that delay in complaint is linked to the credibility of the complainant is a ritual incantation that has been repeated so often over the decades that its mere repetition is now used as evidence for the proposition for which it stands. The contrary view, supported by empirical research, is that there is no demonstrable ‘nexus between delay in complaint and the credibility of the complainant’.^[159]

This means that the issue of forensic disadvantage is entirely separate to the quality and accuracy of the complainant’s evidence.

The difficulty in giving an adequate Longman warning is demonstrated in a study by the Judicial Commission of New South Wales in relation to successful appeals against conviction in sexual assault trials for the period 2001 to June 2004.^[160] In 60 per cent of these cases (22 out of 37), ‘there was a deficiency in the Longman warning resulting in an error of law.’^[161]

An inadequate Longman warning was the only error of law identified in 18 of those 22 cases.^[162] In the other 4 cases, inadequate Longman warnings, plus other misdirections, were found to have occurred. Of these 22 cases, a retrial was ordered in 14 whilst the New South Wales Court of Criminal of Appeal ordered an acquittal in the other 8 cases.^[163] Boniface has also reported that of all appeals to the New South Wales Court of Criminal of Appeal in 2004, 28.3 per cent were appeals against convictions for sex offences and of those appeals, more than half (55.5 per cent) involved appeals on the grounds of an ‘inadequate or incorrect judicial direction and or warning’, although no data was provided on the type of warning or direction.^[164]

Australia-wide, the TLRI has listed 66 appeal cases heard between 1994–2004 in which there has been a successful appeal on the ground of failure to give an adequate Longman warning.^[165]

1 The misleading nature of the warning, in that it amounts to ‘an irrebuttable presumption’ that the accused has in fact been prejudiced;^[166]

3 Its focus on the evidence of the complainant rather than the forensic disadvantage suffered by the accused;

4 The wording of the warning suggests that there is something unreliable about the complainant’s evidence which has the danger of perpetuating myths that children and women are a class of unreliable witnesses;

5 The lack of guidance as to the actual length of delay where a warning is required;

6 A requirement for the warning even where there is corroboration of the complainant’s evidence;

7 The mandatory nature of the warning as judges strive to make their directions appeal-proof;

8 The strength of the warning means that it is easily interpreted as a not too subtle hint to acquit.

The reasoning behind the Longman warning contradicts the public interest in encouraging children and adults to report sexual abuse and the public interest in prosecuting offenders. At trial, inappropriate warnings are not only unfair to complainants, they are

Because of the assumptions inherent in the Longman warning, there have been calls to either abolish it or limit its applicability to cases where it can be shown that the accused has suffered an actual disadvantage because of the delay in complaint.^[168]

Although some believe that the mere fact of delay gives rise to a disadvantage and that an accused may not be able to identify the actual disadvantage suffered (such as the loss of documents or the unavailability of particular witnesses),^[169]

the unequivocal nature of the Longman warning means that all cases of delay are presently being treated as if the accused were at a forensic disadvantage. If, however, ‘the accused did commit the offence charged, any such warning … would be misleading if not positively untrue.’^[170] It would also be untrue if the accused has otherwise not been prejudiced by the delay, such as a case in which rebuttal evidence does not exist because the accused concedes he was with the complainant on the occasion in question. Arguably, it makes sense to confine the warning to cases where there has been demonstrable disadvantage suffered by the accused, rather than rely on the vague assumption that disadvantage is inevitable. Interestingly, the Longman warning contrasts with the approach taken in the English case law, since ‘in order to gain an entitlement to a warning [based on forensic disadvantage as a result of delay], an English defendant must point to the ways in which his defence has been prejudiced by the delay.’^[171]

VI REFORMS TO THE LONGMAN WARNING

A number of commentators and reports have considered the ways in which the Longman warning could be amended. Most consider that the warning should only be given in terms that focus on the forensic disadvantage to the accused,^[172]

that the words ‘dangerous or unsafe to convict’ should be removed from the warning, and that it should only be given if the accused can show that he has suffered a disadvantage.

In 2002, the New South Wales Standing Committee on Law and Justice recommended that an amendment be made to the Criminal Procedure Act 1986 (NSW) ‘to prohibit the issuing of the Longman judicial warning where there is no evidence or good reason to suppose that the accused was prejudiced by the delay in complaint.’^[173]

Similarly, the TLRI considered that where there has been no disadvantage to the accused as a result of the delay in complaint, ‘application of the Longman warning is irrational’ and ‘should be limited to situations where an accused can show a specific disadvantage caused by the delay’ or where there are exceptional circumstances which cannot be established by delay alone.^[174]

The TLRI recommended repeal of s 165(5) of the Evidence Act 2001 (Tas) as the first step in abolishing the Longman warning, since s 165(5) preserves the power of a trial judge to give common law warnings and would ‘encourage trial judges to give warnings in accordance with s 165(1), (2), (3) and (4) rather than in accordance with the common law’.^[175] In jurisdictions which operate under the UEA, this reform will be insufficient because s 9 preserves the operation of the common law unless the Act expressly or by necessary intendment provides otherwise.^[176]

The VLRC recommended that the Longman warning should be restricted to situations

Specifically, the VLRC recommended that the following legislative provision be inserted into s 61(1) of the Crimes Act 1958 (Vic) as an alternative to the Longman warning:

(c) The judge must not warn, or suggest in any way to the jury that it is dangerous or unsafe to convict the accused, unless satisfied that:

(i) there is evidence that the accused has in fact suffered some specific forensic disadvantage due to a substantial delay in reporting; or

(ii) there is evidence that the accused has in fact been prejudiced as a result of other circumstances in the particular case.

(d) If the judge is satisfied in accordance with subsection (c) that a jury warning is required, the judge may warn the jury in terms she or he thinks appropriate having regard to the circumstances of the particular case.

(e) In giving a jury warning pursuant to subsection (d), it is not necessary for the judge to use the words ‘dangerous or unsafe to convict’.^[178]

However, the TLRI questioned whether the VLRC’s recommendation will ‘actually achieve its legislative intent and displace the requirement to give a Longman warning.’^[179]

First, it noted that the above recommendation contains no standard by which the trial judge is to be satisfied that the accused has suffered a forensic disadvantage.^[180] Secondly, the requirement that a trial judge may use whatever terms he or she thinks appropriate in giving the alternative warning will probably see trial judges using the wording of the Longman warning in order to avoid appeals.^[181] Thirdly, sub-s (e) does not prohibit the use of the words ‘dangerous or unsafe to convict’ nor does it require trial judges to emphasise the forensic disadvantage suffered by the accused; it can thus be expected that, under this provision, the warning will continue to focus on the truth and accuracy of the complainant’s evidence.^[182] In fact, the TLRI considered that because of the shortcomings in the VLRC’s recommendation, prescriptive reform may be needed:

In fact, in the subsequent amendments to s 61 of the Crimes Act 1958 (Vic), the common law power to give the Longman warning appears to have been abrogated under s 61(1E):

(1A) If the judge, on the application of the accused in a proceeding to which subsection (1) applies, is satisfied that the accused has suffered a significant forensic disadvantage because of the consequences of the delay in making a complaint about the alleged offence by the person against whom the offence is alleged to have been committed, the judge must, in any terms that the judge considers appropriate having regard to the circumstances of the case—

(a) inform the jury of the nature of the forensic disadvantage suffered by the accused; and

(1B) Despite subsection (1A), a judge must not warn, or suggest in any way to, the jury that it would be dangerous or unsafe to find the accused guilty because of the delay.

(1E) A judge must not give a warning referred to in subsection (1A) or a warning to the effect of a warning referred to in subsection (1A) except in accordance with this section and any rule of law to the contrary is hereby abrogated.

(1F) Nothing in subsections (1A) to (1E) affects the power of a judge to give any other warning to, or to otherwise inform, the jury.

Very clear words of abrogation need to be included in a provision that is designed to remove the power to give a common law warning.^[184] For the purposes of a forensic disadvantage warning given under s 61(1A), it is clear that the content of the warning has to comply with that section, such that it cannot include the words ‘dangerous or unsafe to convict’. Section 61(1E) reinforces this fact. However, the words under sub-s (1E) that ‘any rule of law to the contrary is hereby abrogated’ could either mean that the Longman warning is abrogated or that the form of words known as the Longman warning cannot be given under s 61(1A). This would mean that the Longman warning could still be given in addition to a warning pursuant to s 61(1A). All other common law warnings are preserved under s 61(1F).

In R v Taylor [No 2], Ashley JA recognised that there were ‘statutory constraints’ that had been imposed by s 61(1A)–(1E) on the obligation of a judge to give a Longman warning but, since these provisions did not apply to the facts of the case, no interpretation of them was made.^[185]

A number of submissions to the ALRC, the NSWLRC and the VLRC considered that the Longman warning should be included in the UEA,^[186]

although there was also the view that ‘the large number of appeals concerning the Longman warning would not be alleviated’ by doing so.^[187] If the warning was to be included in the UEA, the ALRC, the NSWLRC and the VLRC considered that ‘further categories [could] be added to s 165(1) to deal with’ the type of situation envisaged by Longman and suggested the following addition: ‘evidence that may be unreliable but not demonstrably so because of the inability to test it adequately for any reason including the passage of time’.^[188]

However, this recommendation would do little to stem the mandatory nature of the Longman warning: given a choice between the common law warning and a s 165 warning, trial judges would still have their eye on the need to comply with the exact terms of the Longman warning.^[189]

The above suggested category also suffers from the flawed reasoning that unreliability and the presumed inability to test evidence because of delay go hand in hand.

In their final report on the UEA, the ALRC, NSWLRC and VLRC stated that there was considerable support for abolishing the Longman warning in its current form, legislating to clarify its operation and to limit ‘its application to cases where defence counsel demonstrates that a particular forensic disadvantage has been incurred.’^[190]

In light of the many concerns that have been expressed about the Longman warning, the ALRC and the VLRC (but not the NSWLRC) recommended the enactment of a provision in the UEA of general application in all criminal trials which would ‘assist to reinforce the fact that forensic disadvantage … is an issue which should be considered independently of the credibility of the complainant.’^[191] This recommendation means that a warning should only be given ‘where there is an identifiable risk to the accused’ so that ‘prejudice should not be assumed to exist merely because of the passage of time.’^[192] The NSWLRC dissented from this recommendation on the grounds that the Longman warning is essential to a fair trial.^[193]

Under the ALRC and VLRC’s recommendation, the trial judge would be required to outline ‘the particular circumstances which have created the forensic disadvantage and explain their significance for the accused’s case.’^[194] Although no particular form of words is required under the recommended warning, the words ‘dangerous to convict’ are prohibited because these words can be interpreted by juries as a direction to acquit.^[195] The ALRC and VLRC considered that if no specific forensic disadvantage could be identified by the defence then the mere fact of delay could be raised by the defence in its closing address to the jury.^[196] Lastly, the ALRC and the VLRC considered that the warning should only be given if requested by the defence, a factor that would then need to be taken into account in any subsequent appeal.

The ALRC and VLRC’s recommendation was subsequently enacted as s 165B of the Evidence Act 1995 (NSW) with some minor changes;^[197] in particular the warning can only be given if requested by a party, not just by the accused:

(2) If the court, on application by a party, is satisfied that the defendant has suffered a significant forensic disadvantage because of the consequences of delay, the court must inform the jury of the nature of that disadvantage and the need to take that disadvantage into account when considering the evidence.

(3) The judge need not comply with subsection (2) if there are good reasons for not doing so.

(4) It is not necessary that a particular form of words be used in informing the jury of the nature of the significant forensic disadvantage suffered and the need to take that disadvantage into account, but the judge must not in any way suggest to the jury that it would be dangerous or unsafe to convict the defendant solely because of the delay or the forensic disadvantage suffered because of the consequences of the delay.

(5) The judge must not warn or inform the jury about any forensic disadvantage the defendant may have suffered because of delay except in accordance with this section, but this section does not affect any other power of the judge to give any warning to, or to inform, the jury.

(a) delay includes delay between the alleged offence and its being reported; and

(b) significant forensic disadvantage is not to be regarded as being established by the mere existence of a delay.

(7) For the purposes of this section, the factors that may be regarded as establishing a significant forensic disadvantage include, but are not limited to, the following:

(a) the fact that any potential witnesses have died or are not able to be located;

(b) the fact that any potential evidence has been lost or is otherwise unavailable.^[198]

Although the VLRC recently concluded that s 165B deals appropriately with the issue of delay and forensic disadvantage and preferred it to s 61 of the Crimes Act 1958 (Vic), the Commission failed to highlight the limitations of the provision.^[199] The apparently robust nature of s 165B is undermined by the retention of the power of trial judges to give any other warnings under sub-s (5), which includes common law warnings. The reform may have little effect in practice since the power to give a Longman warning has not been removed and a s 165B warning is dependent on an application by a party, unlike the Longman warning. While the NSW version allows the prosecution to seek a warning under s 165B (which could be used as a tactic to prevent the Longman warning being given), a situation could arise where the prosecution seeks a s 165B warning while the defence reminds the judge of the mandatory nature of the Longman warning. As such, a trial judge could end up giving both warnings although he or she is likely to invoke s 165B(3) and refuse to give the s 165B warning on the grounds that a Longman warning will be given instead.^[200] If there is a conviction, it would be extremely unlikely for the defence to appeal against a failure by a trial judge to give the less advantageous s 165B warning. In Victoria and the Australian Capital Territory, where a warning under s 165B is dependent upon a request by the accused, it is doubtful that the defence would make such a request as an alternative to the more advantageous Longman warning.

The Criminal Justice and Sexual Offences Taskforce made a similar recommendation that the Longman warning should only be given where ‘the court is satisfied that there is some evidence that the accused has suffered a specific forensic disadvantage due to the delay.’^[201]

Subsequently, the NSW government implemented the Taskforce’s recommendation as an amendment to s 294 of the Criminal Procedure Act 1986 (NSW), which included the proviso that ‘the mere passage of time is not in itself to be regarded as establishing a significant forensic disadvantage.’^[202] However, the amendment did not abolish the words ‘dangerous to convict’ as recommended by the Taskforce. This amendment to s 294 has since been repealed.^[203]

The most radical reforms to the Longman warning have been enacted in South Australia. In 2006, the South Australian Attorney-General’s Office released a discussion paper in which it reviewed the various approaches that had been made to reforming the Longman warning.^[204] Although no final report was released, submissions in response to this discussion paper resulted in a number of reforms to sexual assault laws in South Australia. The reform to the Longman warning is based on criticisms made by the ALRC, the NSWLRC and the VLRC^[205]

and the subsequent amendments to s 165B of the UEA.^[206] However, the South Australian reform, in the form of s 34CB of the Evidence Act 1929 (SA), goes much further than the amendments to s 165B in that it appears to have abolished the Longman warning:

(1) A rule of law or practice obliging a judge in a trial of a charge of an offence to give a warning of a kind known as a Longman warning is abolished.^[207]

Instead of the Longman warning, a trial judge must give a direction about any forensic disadvantage suffered by the accused if the judge ‘is of the opinion that the period of time that has elapsed between the alleged offending and the trial has resulted in a significant forensic disadvantage to the defendant’.^[208] This direction must not take the form of a warning; it must be specific to the circumstances of the particular case and it must not include the wording of the Longman warning, ‘“dangerous or unsafe to convict” or similar words or phrases.’^[209] In other words, the trial judge is required to explain the nature of the forensic disadvantage to the jury and direct the jury that they must take that ‘disadvantage into account when scrutinising the evidence.’^[210] Although it is clear that the South Australian Parliament was intending to abolish the Longman warning with the enactment of s 34CB, arguably,^[211] the wording of s 34CB merely abolishes the obligation to give the Longman warning. In other words, the abolition of the rule of law or practice which obliges a trial judge to give the warning, arguably, means that the trial judge still has the power to give the warning. The approach under s 34CB could be similar to the abolition of the requirement to give a corroboration warning which has been interpreted as abolishing the requirement, but not the power, to do so.^[212]

VII FUTURE REFORMS

This paper endorses the views of the ALRC, the VLRC and the TLRI that there is an overwhelming need to reform the Longman warning, although it considers that prescriptive, legislative reform is required to abolish the power of trial judges to give that warning. The Longman warning should be replaced with an alternative warning which specifies a particular form of words to describe the disadvantages suffered by the accused because of delay in complaint. This would ensure that the accused receives a fair trial where delay has prejudiced their case but would also ensure the fairness of the trial from the point of view of complainants and the community, given that fairness is to be measured more widely than merely the interests of the accused.^[213] However, the alternative warning should only be given where the accused can show that he has suffered an actual forensic disadvantage or prejudice as a result of a delay in complaint.

All in all, a warning about forensic disadvantage ought to reflect its objective — to highlight the disadvantage that has been suffered by the accused as the result of delay. The proposal to restrict the form of words used by the trial judge is based on the fact that an inadequate or impermissible warning given under s 165B (that is, where a warning is given that favours the accused’s case by using the words ‘dangerous to convict’) will not be the subject of appellate review if the accused is acquitted. If convicted, the accused’s counsel is unlikely to appeal against impermissible wording if it favours the accused’s case.

Combining the above ideas with wording from ss 165 and 165B, this paper recommends, first, that s 165B of the UEA should be repealed and the following provision enacted in its stead and, secondly, that non-UEA jurisdictions enact a similar provision:

(1) This section applies to evidence given by a child in prescribed sexual offences proceedings before a jury.

(2) If, on application by the defendant, the court is satisfied on the balance of probabilities that the defendant has demonstrated that he or she has suffered an actual forensic disadvantage as a result of a delay in complaint, the court must:

(a) inform the jury that because of the passage of time, the defendant has shown he or she has suffered a disadvantage in meeting the prosecution’s case as a result of the delay in complaint;

(b) explain the nature of the disadvantage suffered by the defendant according to the circumstances of the case; and

(c) inform the jury that they may take that disadvantage into account in deciding whether the prosecution has proved its case beyond reasonable doubt.

(3) The mere fact of delay is not to be regarded as sufficient to give rise to a warning under subsection (2).

(4) No other form of words can be used by the court other than those set out in subsection (2) and a court must not warn, or suggest to the jury in any way that it is dangerous or unsafe to convict the accused.

(5) For the purposes of this section, delay includes delay between the alleged offence and the complainant’s first report to the police.

(6) The court need not comply with subsection (2) if there are good reasons for not doing so but must state those reasons.

(7) Despite any other provisions to the contrary, the judge must not warn or inform the jury about any forensic disadvantage the defendant may have suffered because of delay except in accordance with this section. This section abolishes the power of the court to give the common law warning known as the Longman warning (Longman v The Queen [1989] HCA 60; (1989) 168 CLR 79).

(8) For the purposes of this section, the factors that may be regarded as establishing an actual forensic disadvantage include, but are not limited to, the following:

(a) the fact that any potential witnesses have died or are not able to be located;

(b) the fact that any potential evidence has been lost or is otherwise unavailable.

2 the wording, ‘actual forensic disadvantage’ is used instead of ‘significant forensic disadvantage’. The word ‘significant’ is difficult to define^[214] and is inconsistent with the idea that there has either been a forensic disadvantage or there has not;

3 sub-s (2) prescribes the exact wording of the warning and sub-s (4) prohibits any other form of words being used;

4 sub-s (6) requires the trial judge to state her or his reasons for not giving a warning; and

5 clear words of intention to abrogate the Longman warning are used under sub-s (7).

The proposed s 165B also differs from the South Australian reform, s 34CB of the Evidence Act 1929 (SA), in various ways. First, the accused bears an evidential onus of demonstrating that he has suffered an actual forensic disadvantage, whereas the South Australian provision merely states that the court must be ‘of the opinion’ that a forensic disadvantage has occurred. Arguably, this judicial opinion could be formed merely because there has been a delay in complaint. Secondly, a specific form of words is included under the proposed s 165B to inform the jury that they may take the forensic disadvantage into account when deciding whether the prosecution has proved its case beyond reasonable doubt, which accords with the function of the jury to assess the weight and the reliability of all the evidence. The South Australian provision directs the jury that they must take the forensic disadvantage into account when scrutinising ‘the evidence’, although it does not specify whose evidence.^[215] This could be a reference to the complainant’s evidence, a focus that the Longman warning has always suffered from, as discussed previously.

Thirdly, the proposed s 165B states that the mere fact of delay is not to be regarded as sufficient to give rise to a warning, unlike the South Australian provision, and it also provides a different definition of delay. The South Australian definition (time between the alleged offending and the trial) is arguably too broad, since there can be situations where delays up to the time of trial are caused by the accused rather than the complainant’s failure to report the incident. Fourthly, the trial judge has a discretion to give a s 165B direction in that he or she can refuse to give the warning if there are good reasons for doing so, unlike the South Australian provision. Fifthly, the abolition of the rule of law or practice which obliges a trial judge to give the Longman warning under the South Australian provision arguably still leaves open the possibility that a trial judge can still give the warning. The proposed s 165B abolishes the power to give the Longman warning. Finally, the proposed s 165B lists the type of factors that would establish an actual forensic disadvantage, unlike the South Australian provision.

In addition, there is a particular problem that reform bodies have not addressed when proposing reforms to the Longman warning which means that the proposed s 165B, if adopted, would need to operate in conjunction with other reforms. There has been a failure to recognise that there is no avenue for appellate review of impermissible directions given by a trial judge; that is, where a trial judge continues to give the Longman warning or uses the ‘dangerous to convict’ wording under s 165B. Such impermissible directions will clearly benefit the defence so that if a accused is convicted, it is unlikely that he will lodge an appeal on those grounds. Although there are practical difficulties with allowing the prosecution to appeal against impermissible directions, in particular, the interruption of the trial, there is no other avenue for review of such directions. Presumably an expedited interlocutory hearing before the Court of Criminal Appeal would be able to dispose of what would be, in essence, a very simple appeal — whether the trial judge has, or has not, given directions which contravene s 165B.

The following recommendation is drafted as an amendment to the Criminal Appeal Act 1912 (NSW).^[216] It is recommended that a similarly worded provision be enacted in all other Australian jurisdictions to give the prosecution the power to appeal against impermissible warnings:

(2) The Attorney-General or the Director of Public Prosecutions may appeal to the Court of Criminal Appeal against any warning, direction or suggestion which contravenes s 165B of the Evidence Act 1995 (NSW) pronounced by the court of trial to which the Crown was a party.

(3) The Court of Criminal Appeal must determine whether the warning, direction or suggestion pronounced by the court of trial contravenes s 165B of the Evidence Act 1995 (NSW).

(4) If the Court of Criminal Appeal determines that there is a contravention of s 165B of the Evidence Act 1995 (NSW), the Court of Criminal Appeal must order the court of trial to either discharge the jury or withdraw the warning, direction or suggestion made to the jury.

VIII CONCLUSION

This paper has undertaken an extensive review of the studies that have investigated the patterns of disclosure of children who have been sexually abused. For the first time in the literature, the key patterns of disclosure by children have been identified and summarised in order to inform the legal debate about the significance of delay in complaint.^[217] In particular, this summary reveals that a majority of children do not report within one year of being sexually abused and that the closer the relationship between the child and the offender, the longer it will take for a child to disclose. This empirical research stands in stark contradiction to the long-held beliefs by various judges that delay in complaint ought to be the subject of a specific warning to inform jurors of the dangers of convicting in such circumstances. Rather, the review of the literature in this paper shows that delay in complaint is likely to be indicative of a child who has been abused by a family member; who has experienced multiple abuses; who has experienced more intrusive forms of abuse; who was first abused under the age of 12 years; or whose abuse was accompanied by threats and force.

This extensive review indicates that the judicial view that has, historically, associated delay in complaint with fabrication is a myth that should no longer be perpetuated by the criminal justice system. Because this myth has been couched in the form of the powerful Longman warning, it is quite possible that jurors receiving the instruction that it would be ‘dangerous to convict’ based on the uncorroborated evidence of the complainant alone have taken it as a perhaps not too subtle hint to acquit.

This paper has canvassed the extensive criticisms of the Longman warning, which have been made by sources varying from Court of Appeal judges to law reform commissions. Although a number of reforms to the warning have been recommended and some have been implemented, most notably s 165B of the UEA and s 34CB of the Evidence Act 1929 (SA), these reforms do not go far enough because they have not abrogated the power of trial judges to give the Longman warning. Because of the mandatory nature of the Longman warning, the various reforms that have been made are likely to have little impact on the frequency with which the Longman warning is given in child sexual assault trials.

Since prosecuting offenders is the most effective way our society has yet developed to protect children from sexual abuse, it is important to incorporate within the criminal justice system understandings about why children cannot make complaints in the way adults expect them to. This means dealing with the reality of child sexual abuse which typically involves: (a) an offender who is known or related to the child; (b) grooming by the offender over time to gain a child’s trust, as well as to maintain secrecy and future sexual access; (c) a relationship of dependence, control or power between the child and the offender; (d) no eyewitnesses to the actual sexual contact; (e) no forensic evidence such as medical evidence of penetration;^[218] and (f) delayed complaint. Inevitably, factors (a)–(d) will result in one of the most typical features of child sexual abuse — delayed complaint. In order to accommodate the reality of child sexual abuse, this paper has made two recommendations for reform in order to abolish the use of the Longman warning in child sexual assault trials. Many will see these as radical measures. Arguably, the more radical option is to allow the criminal justice system to continue to perpetuate injustice against children by giving juries a warning that conflicts with everything that is now known about why children do not disclose sexual abuse.

^[*] BSc (Hons), LLB, PhD (UNSW); Senior Lecturer, Faculty of Law, The University of New South Wales.

[1] Suzanne Blackwell, ‘Child Sexual Abuse on Trial in New Zealand’ (Paper presented at the Criminal Law Symposium, Auckland, New Zealand, November 2008).

^[6] Natalie Taylor and Jacqueline Joudo, The Impact of Pre-Recorded Video and Closed Circuit Television Testimony by Adult Sexual Assault Complainants on Jury Decision-Making: An Experimental Study (Australian Institute of Criminology, 2005) 57–60.

^[7] See Anne Cossins, Jane Goodman-Delahunty and Kate O’Brien, ‘Uncertainty and Misconceptions about Child Sexual Abuse: Implications for the Criminal Justice System’ (2009) 16 Psychiatry, Psychology and Law 435, 449.

^[9] Daniel W Smith et al, ‘Delay in Disclosure of Childhood Rape: Results from a National Survey’ (2000) 24 Child Abuse & Neglect 273, 275.

^[10] Diana E H Russell, ‘The Incidence and Prevalence of Intrafamilial and Extrafamilial Sexual Abuse of Female Children’ (1983) 7 Child Abuse & Neglect 133.

^[12] Anthony W Baker and Sylvia P Duncan, ‘Child Sexual Abuse: A Study of Prevalence in Great Britain’ (1985) 9 Child Abuse & Neglect 457, 459.

^[13] David Finkelhor et al, ‘Sexual Abuse in a National Survey of Adult Men and Women: Prevalence, Characteristics, and Risk Factors’ (1990) 14 Child Abuse & Neglect 19, 22.

^[14] Jessie Anderson et al, ‘Prevalence of Childhood Sexual Abuse Experiences in a Community Sample of Women’ (1993) 32 Journal of the American Academy of Child and Adolescent Psychiatry 911, 915.

^[26] Ibid 278. This prevalence rate of 9 per cent is considerably lower than that reported in other prevalence studies and is due to the narrow definition of child sexual abuse — that is, penetration involving threat or force. See, eg, John Briere and Diana M Elliott, ‘Prevalence and Psychological Sequelae of Self-Reported Childhood Physical and Sexual Abuse in a General Population Sample of Men and Women’ (2003) 27 Child Abuse & Neglect 1205. Out of a random sample of 1442 subjects in the United States, 14.2 per cent of men and 32.3 per cent of women reported childhood experiences of sexual abuse: at 1210.

^[35] Other studies have shown that sexual assaults were reported to authorities in only 10–18 per cent of cases: Kamala London et al, ‘Disclosure of Child Sexual Abuse: What Does the Research Tell Us about the Ways that Children Tell?’ (2005) 11 Psychology, Public Policy and Law 194, 199.

^[44] Rochelle F Hanson et al, ‘Factors Related to the Reporting of Childhood Rape’ (1999) 23 Child Abuse & Neglect 559, 565 also found that disclosure was more likely when the perpetrator was a stranger.

^[47] Maria Sauzier, ‘Disclosure of Child Sexual Abuse: For Better or Worse’ (1989) 12 Psychiatric Clinics of North America 455.

^[48] Catalina M Arata, ‘To Tell or Not to Tell: Current Functioning of Child Sexual Abuse Survivors Who Disclosed Their Victimization’ (1998) 3 Child Maltreatment 63, 67–8.

^[49] London et al, above n 35, 202, reported that there is ‘no consistent association between severity or method of coercion and disclosure’ of sexual abuse.

^[51] Steven M Kogan, ‘Disclosing Unwanted Sexual Experiences: Results from a National Sample of Adolescent Women’ (2004) 28 Child Abuse & Neglect 147, 150.

^[60] Sarah E Ullman and Henrietta H Filipas, ‘Gender Differences in Social Reactions to Abuse Disclosures, Post-Abuse Coping, and PTSD of Child Sexual Abuse Survivors’ (2005) 29 Child Abuse & Neglect 767. In their study, only 27.4 per cent of students reported immediately with 63.6 per cent disclosing a year or more after the abuse occurred. Even when disclosing, three quarters (74.5 per cent) of the students reported that they only gave vague or brief details of the abuse: at 774.

^[68] Marcellina Mian et al, ‘Review of 125 Children 6 Years of Age and under Who Were Sexually Abused’ (1986) 10 Child Abuse & Neglect 223, 226; Sauzier, above n 47; Gail Elizabeth Wyatt and Michael Newcomb, ‘Internal and External Mediators of Women’s Sexual Abuse in Childhood’ (1990) 58 Journal of Consulting and Clinical Psychology 758, 763–5; Arata, above n 48, 67–8; Smith et al, above n 9, 285; Tina B Goodman-Brown et al, ‘Why Children Tell: A Model of Children’s Disclosure of Sexual Abuse’ (2003) 27 Child Abuse & Neglect 525, 527–8; Irit Hershkowitz, Dvora Horowitz and Michael E Lamb, ‘Trends in Children’s Disclosure of Abuse in Israel: A National Study’ (2005) 29 Child Abuse & Neglect 1203, 1208, 1212. Cf Fleming, above n 17, 68, who found that the relationship to the abuser led to ‘[n]o significant differences in disclosure rates’.

^[69] Kogan, above n 51, 160. This view is supported by a study that has shown that parents’ reactions to girls who disclosed incest were generally negative: Thomas A Roesler and Tiffany Weissmann Wind, ‘Telling the Secret: Adult Women Describe their Disclosures of Incest’ (1994) 9 Journal of Interpersonal Violence 327, 328, 330.

^[70] Ramona Alaggia, ‘Many Ways of Telling: Expanding Conceptualizations of Child Sexual Abuse Disclosure’ (2004) 28 Child Abuse & Neglect 1213, 1216.

^[73] Ibid 148. See also Deborah E Nagel, Frank W Putnam and Jennie G Noll, ‘Disclosure Patterns of Sexual Abuse and Psychological Functioning at a 1-Year Follow-up’ (1997) 21 Child Abuse & Neglect 137, 144; Smith et al, above n 9, 283; Mary L Paine and David J Hansen, ‘Factors Influencing Children to Self-Disclose Sexual Abuse’ (2002) 22 Clinical Psychology Review 271, 281; Hershkowitz, Horowitz and Lamb, above n 68, 1212.

^[90] Sir Matthew Hale, Historia Placitorum Coronae (Profession Books, first published 1736, 1971 ed), quoted in Elisabeth McDonald, ‘Gender Bias and the Law of Evidence: The Link between Sexuality and Credibility’ (1994) 24 Victoria University of Wellington Law Review 175, 176.

^[92] See, eg, Judith A Allen, Sex & Secrets: Crimes Involving Australian Women since 1880 (Oxford University Press, 1990) 57–8; Boniface, ‘Ruining a Good Boy for the Sake of a Bad Girl’, above n 87; Jill Bavin-Mizzi, ‘Understandings of Justice: Australian Rape and Carnal Knowledge Cases, 1876–1924’ in Diane Kirkby (ed), Sex Power and Justice: Historical Perspectives of Law in Australia (Oxford University Press, 1995) 19.

^[93] Jacqueline Fitzgerald, ‘The Attrition of Sexual Offences from the New South Wales Criminal Justice System’ (2006) 92 Crime and Justice Bulletin: Contemporary Issues in Crime and Justice 1, 3. See also Joy Wundersitz, Child Sexual Assault: Tracking from Police Incident Report to Finalisation in Court (Office of Crime Statistics and Research, 2003).

^[101] Australian Law Reform Commission, New South Wales Law Reform Commission and Victorian Law Reform Commission, Uniform Evidence Law, ALRC Report No 102, NSWLRC Report No 112, VLRC Final Report (2005) 594–5 [18.25] (‘Uniform Evidence Law Report’).

^[103] (2001) 208 CLR 343, 355–7 (Gaudron and Callinan JJ). But see criticism of the expansion in the dissenting judgment of Gleeson CJ: at 348.

^[109] These warnings are: the Murray direction (R v Murray (1987) 11 NSWLR 12, 19 (Lee J)); the Longman direction (Longman v The Queen [1989] HCA 60; (1989) 168 CLR 79, 85, 94 (Brennan, Dawson and Toohey JJ)); the Crofts direction (Crofts v The Queen [1996] HCA 22; (1996) 186 CLR 427, 451 (Toohey, Gaudron, Gummow and Kirby JJ)); the KRM direction (KRM v The Queen [2001] HCA 11; (2001) 206 CLR 221, 234 [36] (McHugh J), 248 [79], 257 [106]–[107] (Kirby J), 263–4 [132] (Hayne J)); a warning limiting the use of evidence for credibility purposes (see, eg, Evidence Act 1995 (Cth) s 136); the Gipp warning (Gipp v The Queen (1998) 194 CLR 106, 133 (McHugh and Hayne JJ)); warnings concerning the use of coincidence (similar fact) evidence (see, eg, R v BWT [2002] NSWCCA 60; (2002) 54 NSWLR 241, 251 [32] (Wood CJ at CL); Evidence Act 1995 (Cth) s 98); and the BRS direction (BRS v The Queen [1997] HCA 47; (1997) 191 CLR 275, 298–301 (Gaudron J)). It is to be noted that in UEA jurisdictions, since R v BWT [2002] NSWCCA 60; (2002) 54 NSWLR 241, 251 (Wood CJ at CL) other warnings and directions, such as a warning pursuant to Evidence Act 1995 (Cth) ss 165A–165B, may be required at the request of a party. In addition, the Gipp warning is no longer required as a result of the decision in HML v The Queen [2008] HCA 16; (2008) 235 CLR 334, 353 [9] (Gleeson CJ), 498 [501] (Kiefel J).

^[112] Criminal Justice and Sexual Offences Taskforce, Attorney-General’s Department (NSW), Responding to Sexual Assault: The Way Forward (Oxford Publishing, 2006) 90.

^[118] New South Wales Standing Committee on Law and Justice, Report on Child Sexual Assault Prosecutions, Parliamentary Paper No 208 (2002) (‘Report on Child Sexual Assault Prosecutions’).

^[119] TLRI, Warnings in Sexual Offences Cases Relating to Delay in Complaint, Final Report No 8 (2005) (‘Warnings in Sexual Offences Report’).

^[120] VLRC, Sexual Offences, Final Report (2004) (‘Sexual Offences Report’).

^[121] Uniform Evidence Law Report, above n 101, 614 [18.116]; ALRC, NSWLRC and VLRC, Review of the Uniform Evidence Acts, ALRC Discussion Paper No 69, NSWLRC Discussion Paper No 47, VLRC Discussion Paper (2005) 478–81 [16.111]–[16.121] (‘Review of the Uniform Evidence Acts Paper’).

^[129] Penney Lewis, ‘A Comparative Examination of Forensic Disadvantage Directions in Delayed Prosecutions of Childhood Sexual Abuse’ (2005) 29 Criminal Law Journal 281, 285 (citations omitted).

^[130] Justice James Wood, ‘Sexual Assault and the Admission of Evidence’ (Paper presented at Practice and Prevention: Contemporary Issues in Adult Sexual Assault in New South Wales, University of Technology, Sydney, 12 February 2003) [21]. See also R v Mazzolini [1999] VSCA 150; (1999) 3 VR 113, 130 [52] (Ormiston JA); Warnings in Sexual Offences Report, above n 119, 6–7 [1.2.5].

^[131] As discussed below, it is unclear whether the mandatory status of the Longman warning will continue as a result of recent reforms to the warning under the UEA (see, eg, Evidence Act 1995 (Cth) s 165B). At the time of writing, there has been no appellate review of that provision in jurisdictions using the Uniform Evidence Legislation.

^[132] Warnings in Sexual Offences Report, above n 119, 14–15 [2.2.11].

^[135] Crampton v The Queen (2000) 206 CLR 161, 181 [44]–[45] (Gaudron, Gummow and Callinan JJ), 208 [126] (Kirby J), 112–13 [142] (Hayne J); R v BWT [2002] NSWCCA 60; (2002) 54 NSWLR 241, 273 [95] (Sully J). This is in contrast to the approach of the Court of Appeal of Western Australia (see, eg, Angliss v Western Australia [2005] WASCA 162 (29 July 2005) [18]–[20] (Wheeler JA)) and the approach of the New South Wales Court of Criminal Appeal (see, eg, TJ v The Queen [2009] NSWCCA 257 (21 October 2009) [78] (Hidden J), [99] (McCallum J)).

^[136] DRE v The Queen (2006) A Crim R 400, 405 [29]–[31] (Spigelman CJ), 409 [59]–[60] (Simpson J).

^[138] R v BWT [2002] NSWCCA 60; (2002) 54 NSWLR 241, 273 [95] (Sully J). See also Review of the Uniform Evidence Acts Paper, above n 121, 475 [16.97]. Cf TJ v The Queen [2009] NSWCCA 257 (21 October 2009) [62] (McClellan CJ at CL), where it was considered that ‘a trial judge is not constrained to a particular formulation of the substance of the [Longman] warning’.

^[145] In New South Wales, a case involving a delay of only six months led one judge to find that some sort of warning about the effect of delay should have been given, but not a full Longman warning: R v Heuston (2003) A Crim R 422, 432 [50]–[52] (Hodgson JA). See also DRE v The Queen [2006] NSWCCA 280; (2006) 164 A Crim R 400 where a warning was given in relation to a delay of a few months for one count, although there was a delay of about three years in relation to four other counts. In other cases involving a two-year delay (Tully v The Queen (2006) 230 CLR 234, 253–4 [60] (Kirby J), 270–1 [123], 274 [132] (Callinan J), 287–9 [177]–[186] (Crennan J)) and a four-year delay (Perez v The Queen [2008] NSWCCA 46 (6 March 2008) [74], [78] (Kirby J)) a warning was not considered necessary.

^[148] See the facts in Robinson v The Queen [1999] HCA 42; (1999) 197 CLR 162, 170–1 [25] (Gleeson CJ, McHugh, Kirby, Hayne and Callinan JJ).

^[153] See, eg, R v BWT [2002] NSWCCA 60; (2002) 54 NSWLR 241, 248 [18]–[19] (Wood CJ at CL); Justice James Wood, ‘Complaint and Medical Examination Evidence in Sexual Assault Trials’ (2003) 15 Judicial Officers’ Bulletin 63.

^[158] R v BWT [2002] NSWCCA 60; (2002) 54 NSWLR 241, 252 (Wood CJ at CL) (emphasis altered). See also Uniform Evidence Law Report, above n 101, 615–16 [18.122].

^[160] Graham Hazlitt, Patrizia Poletti and Hugh Donnelly, Judicial Commission of New South Wales, Sentencing Offenders Convicted of Child Sexual Assault (Judicial Commission of New South Wales, 2004) 46–7.

^[167] Uniform Evidence Law Report, above n 101, 611 [18.97] (citations omitted).

^[176] Ibid 24 [3.3.1]; Review of the Uniform Evidence Acts Paper, above n 121, 480 [16.118].

^[185] [2008] VSCA 57; (2008) 18 VR 613, 615 [7]. The sufficiency of a direction given under s 61(1A) of the Crimes Act 1958 (Vic) was recently considered in R v Morrow [2009] VSCA 291 (17 December 2009) [26] (Redlich JA), in which it was stated that the trial judge had erred in giving a direction under s 61(1A)

by describing his direction on forensic disadvantage as ‘comments’. The imperative words used in s 61(1A) require that the jury must be directed that they are to take the relevant forensic disadvantage into consideration. The judge’s direction was therefore inadequate as it left ‘a discretion with the jury as to whether those matters ought be considered.

At the time of writing, there has been no appeal case on the question as to whether s 61(1B) has abrogated the Longman warning in Victoria.

^[186] Review of the Uniform Evidence Acts Paper, above n 121, 477–8 [16.104]–[16.110].

^[197] Section 165B is also included in the Evidence Act 2008 (Vic) and the Evidence Act 1995 (Cth). In these two jurisdictions, s 165B differs from the New South Wales version in that a warning can only be requested by a defendant, not a party. Also, the Victorian and Commonwealth versions of s 165B do not include sub-s (7). At the time of writing, this amendment had not been made to the Evidence Act 2001 (Tas).

^[200] The prediction that a trial judge could give both warnings has been realised in a recent judge-only trial in the Australian Capital Territory: see R v DF [2010] ACTSC 31 (15 April 2010) [254], [267] (Penfold J). See also R v Forsti [2010] ACTSC 85 (19 August 2010) [42] (Gray J).

^[204] Attorney-General’s Office (SA), Review of South Australian Rape and Sexual Assault Law, Discussion Paper (2006).

^[206] South Australia, Parliamentary Debates, House of Assembly, 25 October 2007, 1458 (Michael Atkinson, Attorney-General).

^[211] South Australia, Parliamentary Debates, House of Assembly, 25 October 2007, 1458 (Michael Atkinson, Attorney-General). Since its enactment, s 34CB has not been the subject of appellate review, although Vanstone J has noted ‘in passing’ that in the future ‘a trial judge will not be entitled to give a “warning” in respect of the features which would previously have led to warning in terms of Longman’: R v B [2009] SASC 110 (30 April 2009) [38]; R v Inston [2009] SASC 89; (2009) 103 SASR 265, 292 [107]. Nonetheless, it can be expected that in future cases, an accused will submit that the directions under s 34CB fall ‘well short of the requirements of Longman.’ See the arguments of the parties in R v Dawson-Ryan [2009] SASC 259; (2009) 104 SASR 571, 589 [74] (Gray, Layton and David JJ).

^[212] It is still permissible to give a corroboration warning where a trial judge considers that it is necessary to do so in the interests of justice: see, eg, Bromley v The Queen [1986] HCA 49; (1986) 161 CLR 315, 319 (Gibbs CJ); Longman v The Queen [1989] HCA 60; (1989) 168 CLR 79; Robinson v The Queen [1999] HCA 42; (1999) 197 CLR 162, 168–9 [20]–[21] (Gleeson CJ, McHugh, Kirby, Hayne and Callinan JJ); Conway v The Queen (2002) 209 CLR 203; Tully v The Queen (2006) 230 CLR 234.

^[213] Dietrich v The Queen [1992] HCA 57; (1992) 177 CLR 292, 335 (Deane J). See also Barton v The Queen [1980] HCA 48; (1980) 147 CLR 75, 101 (Gibbs ACJ and Mason J). ‘Indeed, the concept of fairness is not fixed and immutable and “may vary with changing social standards and circumstances”’: Annie Cossins, ‘Cross-Examination in Child Sexual Assault Trials: Evidentiary Safeguard or an Opportunity to Confuse?’ [2009] MelbULawRw 3; (2009) 33 Melbourne University Law Review 68, 72, quoting Dietrich v The Queen [1992] HCA 57; (1992) 177 CLR 292, 328 (Deane J). These circumstances include the increasing awareness of the interests of victims of sexual assault.

^[216] Appellate review in New South Wales is possible under s 5F(3A) of the Criminal Appeal Act 1912 (NSW) in relation to any decision or ruling on the admissibility of evidence, ‘if the decision or ruling eliminates or substantially weakens the prosecution’s case.’ It may be necessary to consider amendments to this provision as well as the above recommendation.

^[218] See, eg, Charles Felzen Johnson, ‘Child Sexual Abuse’ (2004) 364 Lancet 462. Johnson reported that ‘as many as 96 per cent of children assessed for suspected sexual abuse will have normal genital and anal examinations’: at 462.

Length of Time between Abuse and Disclosure	Percentage	Cumulative Percentage
Within 24 hours	17.8	17.8
1 month	8.9	26.7
6 months	7.2	33.9
12 months	3.8	37.7
2 years	5.1	42.8
3 years	2.5	45.3
4 years	2.5	47.8
5 years	4.3	52.1
After 5 years	47.9	100.0

Melbourne University Law Review