Mandate Now review of :
Published April 2016
Click on image to enhance quality
The Mandate Now position has always been that we follow the evidence. If sound research were to show that mandatory reporting in Regulated Activities (e.g. schools, hospitals etc.) were unhelpful to detecting child sex abuse, then we would stop our campaign for the introduction of such a law.
The government has consistently opposed mandatory reporting, and has cited various academic research papers which it claims support its position. These include:
- NSPCC, Wallace & Bunting (2007) included no original research, stated that there is “little empirical evidence” on the subject and yet came out against mandatory reporting, even to the extent of recommending that the existing mandatory reporting arrangements in Northern Ireland should be dismantled.
- Ainsworth (2001) looked at child abuse in Western Australia and talked a lot about the surge in “unsubstantiated” cases and that they are a significant burden to the system. Ainsworth does not recognize that close to half of all reports are made by non-mandated reporters; that a large proportion are multiple reports about the same children; that many reports are screened out and are not investigated, hence resulting in very little burden; and that in any event the substantiation rate of investigated cases following the introduction of MR is significantly higher (Mathews 2012).
- Lamond (1989) and Mathews (2007) actually supported mandatory reporting but were still cited by the DfE as evidence against its introduction.
- Harries and Clare (2002) was a literature review which commented on the lack of quantitative data but nonetheless claimed the existence of “overwhelming evidence” that mandatory reporting systems worldwide are “in chaos” without specifying what that evidence was. We undertook a critical examination of the value of Harries and Clare which you can read by following this link.
- Munro and Parton (2007) looked at the history of legislative arrangements in the UK, a 16-page document which contained a single paragraph of figures on substantiation rates, but otherwise no quantitative data at all. Nevertheless it opposed mandatory reporting on the grounds of insufficient resources to address an anticipated increase in the number of reports.
The possible arguments against mandatory reporting can be summarised as follows.
- The reporting rate is already so high that no compulsion is required.
- Although the reporting rate is not very high, mandatory reporting would not increase it significantly
- A low rate of reporting is acceptable or even desirable for various reasons (such as the unavailability of resources to investigate the referrals), and that no effort should be made to increase it.
It is notable that these reasons conflict with each other. If point 1 is true then points 2 and 3 cannot be, and if point 2 is true then point 3 is not. None of the research cited by the DfE offered particularly strong support for any of these reasons. This hasn’t prevented the NSPCC from using all three in its arguments against mandatory reporting, even though they clearly conflict with each other.
So this is an area where much rhetorical arm-waving has occurred in the absence of facts. But finally we have some. “Impact of a new mandatory reporting law on reporting and identification of child sexual abuse: A seven year time trend analysis” has been written by Ben Mathews, Xing Ju Lee and Rosana E. Norman, and is published in the leading international journal in the field: Child Abuse & Neglect. It is a longitudinal quantitative study of child sexual abuse reports in Western Australia over a seven year period. On 1st January 2009, mandatory reporting of suspicions of child abuse was introduced in Western Australia for certain categories of professionals, specifically doctors, nurses, midwives, police officers and teachers. The study has a consistent set of child abuse reporting data for the years 2006-2012, i.e. three years before the introduction of the law and four years after.
From a survey of existing literature, Mathews early on identifies that there is a low reporting rate of child sexual abuse. Giving Australia as an example, the figures for the most recent year available indicate that cases substantiated by official government child protection agencies after their investigations run at about 111.62/100,000 children, or 0.1%. But population survey data indicates that the actual incidence is much higher. An Australian meta-analysis found prevalence in children aged under 16 in the region of 14% for females and 5% for males, and other research figures in Australia and other countries are comparable. Clearly any argument that the reporting rate is already high enough is not based on facts.
The aim of the Mathews research was threefold. These are the aims as stated in the paper itself.
- To explore the change in numbers of reports of CSA after introduction of the mandatory reporting law by mandated reporters as a combined group, and by distinct groups of mandated reporters (and the rate of reports per 100,000 children).
- To explore the change in numbers of investigated reports of CSA after introduction of the mandatory reporting law by mandated reporters as a combined group, and by distinct groups of mandated reporters (and the proportion of reports that were investigated).
- To explore the change in numbers of substantiated reports of CSA after introduction of the mandatory reporting law by mandated reporters as a combined group, and by distinct groups of mandated reporters (and the proportion of investigated reports that were substantiated).
Mathews obtained and analysed a complete and consistent set of data both pre- and post-law change. Data earlier than 2006 were not used because it was coded using a different system. He goes into detail that needn’t be repeated here as to the design and methods of the study. Suffice to say that they are thorough, much more so than any of the other papers mentioned at the start of this review. This is real research on real primary data.
The results are very interesting.
Taking all the mandated reporters together, there were 3 times as many reports per year after the introduction of the law compared to the level before. The level was stable in the three years prior to the change. There was a jump by a factor of 2.7 in the year the law was introduced and then some small further rises thereafter. Even at its highest level, the annual proportion of children reported as suspected victims of CSA was 1 in 210.
The number of reports which resulted in an investigation (again stable in the 3 years before the law change) rose year on year thereafter, with a mean per year also 3 times the pre-law average. However, in the immediate aftermath of the change, the proportion of reports investigated dropped sharply and then recovered in 2011.
The number of reports that were substantiated was stable in the 3 pre-law years, and increased and remained constant for the years after. The annual number of substantiated reports doubled in the post-law years, compared with the pre-law years. Once the number of reports investigated rose in 2011, the total number of substantiations also rose from about 160 substantiated cases per year to about 380 per year in the years 2011-2012. However, the proportion of investigated reports that were substantiated showed a modest fall.
Among the mandated groups, the largest increase in reports came from teachers. Before the law change, teachers generated fewer reports than the police (about 220 vs 360 per year) but by 2011-12 this had changed, with teachers generating about 2600 per year against the police 1287 per year.
In previous papers on the subject, there has been much discussion about substantiation rates, and the idea that children’s services would be wading through a huge number of trivial and ultimately unsubstantiated reports. This is not what has happened. Yes, the substantiation rate has dropped modestly, but the overall number of substantiated cases has risen.
But Mathews makes the point that substantiation rates are not all that useful a measure of effectiveness. Other research has already discovered that “Many reports of abuse and neglect that are investigated but unsubstantiated do involve abuse and provide opportunities for early intervention” and as a result have concluded that substantiation is “a distinction without a difference”. For instance, a report of suspected CSA is frequently based on the reporter observing the child’s adverse health symptomatology, behaviour, and social context. In such circumstances, there is often a health and welfare need whether or not the CSA is substantiated. Mathews also points out that substantiation rates can also be markedly affected by such things as policy variations in setting evidentiary thresholds, the capacity of agencies to investigate, and whether there is an emphasis on evaluating existing harm or assessing the risk of future harm.
As Mathews says, “there is not firm ground for concluding that when exploring trends in reporting and report outcomes, the sole measure of the soundness of a report of suspected CSA is whether it is substantiated. Outcomes such as actual service provision to the child, and perceived need for service provision even if this is unable to be provided, are among those that are also relevant.”
Unfortunately, the Mathews study did not obtain data on either the perceived or recorded need for, or actual provision of, services to the child and her or his family. Mathews states that “The question of service provision need is crucial when assessing efficacy of reporting practice. Future research to explore this question is urgently needed, using appropriate and sensitive research methods.”
Mathews’ overall conclusion is “The results of this research suggest a mandatory reporting law for CSA is associated with a substantial and sustained increase in identification of cases of CSA. Societies which are considering the introduction of a mandatory reporting law for CSA should find support for this policy intervention from these findings, while recognizing the associated needs for reporter education, investment in agency capacity and service provision, and the need to implement responses to reports with sensitivity.”
Mandatory reporting is not a magic bullet. It will not cure the problem of child sex abuse all by itself. But well implemented, along with other measures to complement it, mandatory reporting is a vital component of a functioning child protection system. These are children’s lives we can avoid being wrecked.